WO2025175620A1 - Smart air cabinet control method and system for improving air quality - Google Patents

Abstract

The present invention relates to the technical field of data processing. Provided are a smart air cabinet control method and system for improving air quality. The air entering an air cabinet is detected to obtain an air particle distribution state; analysis is then performed to determine a sensitive distribution particle size set and a sensitive distribution particle size and concentration set; optimization is performed on the basis of the sensitive distribution particle size set to generate multi-stage filter screen activation parameters, so as to execute control on a multi-stage filter screen of the air cabinet; and the air is introduced into the multi-stage filter screen of the air cabinet for filtering. The present invention solves the technical problems of the existing conventional air cabinet often lacking a smart regulation function in filtering control, being unable to adjust filtering parameters on the basis of real-time air quality data, and being unable to adapt to air pollution conditions in different environments. The present invention achieves the technical effect of improving the quality of air filtering of an air cabinet by means of selecting filter screens of the air cabinet and scientifically setting filtering parameters on the basis of the air particulate matter distribution condition and the air particulate matter concentration condition.

Description

Air cabinet intelligent control method and system for improving air quality Technical Field

The present invention relates to the technical field of data processing, and in particular to an intelligent control method and system for an air cabinet for improving air quality.

Background Art

The filtration control function of traditional air cabinets usually lacks intelligent regulation, cannot adjust parameters based on real-time air quality data, and cannot adapt to air pollution conditions in different environments. This limitation makes traditional air cabinets appear to be powerless when dealing with air quality issues and cannot achieve optimal filtration effects.

In summary, existing traditional air cabinet filtration controls often lack intelligent adjustment functions, are unable to adjust filtration parameters according to real-time air quality data, and are unable to adapt to technical problems such as air pollution conditions in different environments.

Summary of the Invention

The present application provides an intelligent control method and system for an air cabinet for improving air quality, which is used to solve the technical problems that existing traditional air cabinet filtration control often lacks intelligent adjustment functions, cannot adjust filtration parameters according to real-time air quality data, and cannot adapt to air pollution conditions in different environments.

In view of the above problems, the present application provides an intelligent control method and system for an air cabinet for improving air quality.

In a first aspect of the present application, an intelligent control method for an air cabinet is provided for improving air quality, the method comprising: activating a deployment In the first cavity laser caliper, an air particle distribution state is obtained, wherein the air particle distribution state includes an air particle distribution particle size list and an air particle distribution concentration list; the air particle distribution particle size list is traversed, and the reference distribution concentration list is matched based on the reference concentration calibration table; according to the air particle distribution concentration list and the reference distribution concentration list, a sensitive distribution particle size set and a sensitive distribution particle size concentration set whose air particle distribution concentration is greater than the reference distribution concentration are extracted; according to the sensitive distribution particle size set, activation optimization is performed on the air cabinet multi-stage filter to generate multi-stage filter activation parameters, wherein the multi-stage filter activation parameters include an activation filter number parameter, a cycle filtration frequency parameter and a cycle flow control parameter; according to the activation filter number parameter, the cycle flow control parameter and the cycle filtration frequency parameter, the air cabinet multi-stage filter is controlled, and air is passed into the air cabinet multi-stage filter for filtration, and after completion, the air is passed into the room.

The second aspect of the present application provides an air cabinet multi-stage filter device that realizes the above-mentioned air cabinet intelligent control method for improving air quality, the device comprising: a circulation passage, the circulation passage comprising: a main circulation passage; a first circulation branch passage, a second circulation branch passage until the Qth circulation branch passage, the first circulation branch passage, the second circulation branch passage until the Qth circulation branch passage are connected to the main circulation passage through a three-way valve respectively; an air supply passage, the air supply passage comprising: a main air supply passage; a first air supply branch passage, a second air supply branch passage until the Qth air supply branch passage, the first air supply branch passage, the second air supply branch passage until the Qth air supply branch passage are connected to the main air supply passage through a three-way valve respectively; a first filter cavity, the first air supply branch passage and the first circulation branch passage are connected through the first filter cavity, and the filter of the first filter cavity is deployed between the first air supply branch passage and the Qth A filter cavity is connected to a position, and the first filter cavity and the first circulation branch passage are connected through a one-way valve; a second filter cavity, the second air supply branch passage and the second circulation branch passage are connected through the second filter cavity, and the filter screen of the second filter cavity is arranged at the connection position between the second air supply branch passage and the second filter cavity, and the second filter cavity and the second circulation branch passage are connected through a one-way valve; until the Q filter cavity, the Q air supply branch passage and the Q circulation branch passage are connected through the Q filter cavity, and the filter screen of the Q filter cavity is arranged at the connection position between the Q air supply branch passage and the Q filter cavity, and the Q filter cavity and the Q circulation branch passage are connected through a one-way valve; wherein, both ends of the main circulation passage are sealed, one end of the gas flow direction of the main air supply passage is not sealed, and the other end is sealed.

The third aspect of the present application provides an intelligent control system for an air cabinet for improving air quality, the system comprising: a distribution state obtaining unit for activating a laser diameter meter deployed in a first cavity of the air cabinet when air enters the first cavity to obtain an air particle distribution state, wherein the air particle distribution state comprises an air particle distribution particle size list and an air particle distribution concentration list; a concentration list matching unit for traversing the air particle distribution particle size list and matching the reference distribution concentration list based on a reference concentration calibration table; a data extraction execution unit for extracting a sensitive distribution particle size set and a sensitive distribution particle size concentration set whose air particle distribution concentration is greater than the reference distribution concentration based on the air particle distribution concentration list and the reference distribution concentration list; an activation parameter generating unit for performing activation optimization on the multi-stage filter of the air cabinet based on the sensitive distribution particle size set to generate multi-stage filter activation parameters, wherein the multi-stage filter activation parameters comprise activation filter number parameters, cycle parameters, and activation filter number parameters. Ring filtration frequency parameters and circulation flow control parameters; an air filtration execution unit, which is used to control the multi-stage filter of the air cabinet according to the activation filter number parameters, the circulation flow control parameters and the circulation filtration frequency parameters, and pass the air into the multi-stage filter of the air cabinet for filtration, and then pass it into the room after completion.

One or more technical solutions provided in this application have at least the following technical effects or advantages:

The method provided in an embodiment of the present application is as follows: when air enters the first cavity of the air cabinet, activating a laser diameter meter deployed in the first cavity to obtain an air particle distribution state, wherein the air particle distribution state includes an air particle distribution particle size list and an air particle distribution concentration list; traversing the air particle distribution particle size list, and matching the reference distribution concentration list based on a reference concentration calibration table; extracting a sensitive distribution particle size set and a sensitive distribution particle size concentration set whose air particle distribution concentration is greater than the reference distribution concentration according to the air particle distribution concentration list and the reference distribution concentration list; performing activation optimization on the multi-stage filter of the air cabinet according to the sensitive distribution particle size set to generate multi-stage filter activation parameters, wherein the multi-stage filter activation parameters include an activation filter number parameter, a cycle filtration frequency parameter and a cycle flow control parameter; controlling the multi-stage filter of the air cabinet according to the activation filter number parameter, the cycle flow control parameter and the cycle filtration frequency parameter, passing air into the multi-stage filter of the air cabinet for filtration, and then passing air into the room after completion. The technical effect of calling the filter screen of the air cabinet and scientifically setting the filtering parameters according to the distribution and concentration of air particles is achieved, thereby improving the air filtration quality of the air cabinet is achieved.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a flow chart of an intelligent control method for an air cabinet for improving air quality provided by the present application;

FIG2 is a schematic diagram of a process for generating multi-stage filter activation parameters in the air cabinet intelligent control method for improving air quality provided by the present application;

FIG3 is a schematic structural diagram of the multi-stage filter device for an air cabinet provided in this application;

FIG4 is a schematic structural diagram of an intelligent control system for an air cabinet for improving air quality provided in this application.

Description of reference numerals:
Main circulation passage 1, first circulation branch passage 2, second circulation branch passage 3, three-way valve 4, main air supply passage 5, first air supply branch passage 6, second air supply branch passage 7, first filter cavity 8, one-way valve 9, second filter cavity 10, air intake passage 11, distribution state acquisition unit T11, concentration list matching unit T12, data extraction execution unit T13, activation parameter generation unit T14, air filtration execution unit T15.

DETAILED DESCRIPTION

This application provides an intelligent control method and system for air cabinets to improve air quality. This method addresses the technical issues that existing traditional air cabinet filter controls often lack intelligent adjustment capabilities, are unable to adjust filter parameters based on real-time air quality data, and are unable to adapt to air pollution conditions in different environments. This method achieves the technical effect of improving the air cabinet's air filtration quality by enabling the air cabinet's filter to be activated and filter parameters to be scientifically set based on the distribution and concentration of air particles.

The acquisition, storage, use, and processing of data in the technical solution of the present invention comply with relevant regulations.

Below, the technical solutions of the present invention will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments described herein. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. It should also be noted that, for the convenience of description, only the parts related to the present invention, rather than all, are shown in the accompanying drawings.

Example 1

As shown in FIG1 , the present application provides an intelligent control method for an air cabinet for improving air quality, which is applied to an intelligent control system for an air cabinet for improving air quality. The system is communicatively connected to the air cabinet, and the air cabinet includes a first cavity and an air cabinet multi-stage filter. The air cabinet multi-stage filter can be selectively activated in combination, including:

A100: When air enters the first cavity of the air cabinet, activate the laser diameter measuring instrument deployed in the first cavity to obtain the air particle distribution state, wherein the air particle distribution state includes an air particle distribution particle size list and an air particle distribution concentration list;

In one embodiment, when air enters the first cavity of the air cabinet, a laser diameter meter deployed in the first cavity is activated to obtain an air particle distribution state, wherein the air particle distribution state includes an air particle distribution particle size list and an air particle distribution concentration list. Step A100 of the method provided in this application also includes:

A110: The first cavity includes a first valve and a second valve, wherein the first valve and the second valve are initially closed, and the first cavity is initially in a vacuum state;

A120: Based on the Internet of Things, receives outdoor air density;

A130: When a ventilation command is received, the first valve is opened to let in air. When the density deviation between the air density in the first cavity and the outdoor air density is less than or equal to the density deviation threshold, the first valve is closed and the first cavity laser diameter meter is activated to perform multi-position detection to obtain the air particle distribution state, wherein the air particle distribution state includes an air particle distribution particle size list and an air particle distribution concentration list.

Specifically, in this embodiment, an intelligent control method for an air cabinet for improving air quality is applied to an intelligent control system for an air cabinet for improving air quality. The system is communicatively connected to the air cabinet, and the operating parameters of the air cabinet are adjusted and controlled based on the system to ensure that the air cabinet is operating to effectively filter suspended impurities in the air.

The air cabinet is composed of a first cavity, a first valve, a second valve and a multi-stage filter screen of the air cabinet. The multi-stage filter screen arranged inside the first cavity and participating in air filtration can be selectively combined and activated to achieve indoor air filtration with different filtration fineness levels.

Air enters the first cavity through the first valve, and becomes clean air after being filtered for suspended impurities by the multi-stage filter of the air cabinet. The clean air leaves the air cabinet through the second valve, completing the air circulation filtration.

When the air cabinet is in an initial state of not being used, the first valve and the second valve are initially closed, so that the first cavity is initially in a vacuum state.

At the same time, in addition to the air cabinet multi-stage filter screen that actually participates in air filtration, the laser diameter gauge for detecting the distribution state of suspended particles in the air entering the first cavity of the air cabinet is also deployed in the space formed by the first valve and the air cabinet multi-stage filter screen. The laser diameter gauge can detect the particle distribution of the air. Concentration and particle size distribution.

Based on the Internet of Things, the outdoor air density is received. When the system receives a ventilation command, it will open the first valve and keep the second valve closed to introduce outdoor air into the first cavity. At this time, the system will also monitor the air density of the first cavity based on the Internet of Things, and then compare it with the outdoor air density received based on the Internet of Things to calculate the density difference between the two.

A density deviation threshold is preset, and it is considered that when the density deviation between the two is less than or equal to the density deviation threshold, it can be considered that the outdoor air distribution is the same as the air distribution in the first cavity. At this time, the system will control the closure of the first valve of the air cabinet and enter the air particle distribution detection stage.

Multiple predetermined detection locations and multiple air particle size thresholds are preset within the first air cabinet. During the air particle distribution detection phase, the system activates a first cavity laser caliper to detect particles in the air at local locations within the multiple predetermined detection locations within the first air cabinet. Multiple sets of air particle size distributions and air particle concentrations are obtained for the multiple predetermined detection locations. Each set of air particle size distributions and air particle concentrations represents the concentration of air particles at a certain particle size threshold at a detection location, for example, 100 particles/ ^cm³ for 0.3-0.4 microns and 80 particles/ ^{cm³ for} 0.4-0.5 microns.

Based on a plurality of preset air particle size thresholds, data aggregation of a plurality of groups of air particle distribution particle sizes - air particle distribution concentrations is performed to obtain a plurality of air particle distribution concentrations for each air particle size distribution, and then the average of the plurality of air particle distribution concentrations is calculated to obtain the air particle distribution particle size - air particle distribution concentration of the concentration condition of particulate matter of a particle size threshold as a whole in the first air cabinet.

Similarly, based on the mean value, multiple groups of air particle distribution particle sizes - air particle distribution concentrations of the concentration conditions of various particle size thresholds in the first air cabinet as a whole are obtained, and then air particles are extracted to generate an air particle distribution particle size list, and particle distribution concentration is extracted to generate an air particle distribution concentration list.

Based on the air particle distribution size list and air particle distribution concentration list, you can understand the concentration levels of particles of different particle sizes in the current air cabinet as a whole.

This embodiment achieves the technical effect of digitizing the concentration status of suspended particulate matter in unfiltered air before the air cabinet performs air filtration by constructing an air cabinet including a first valve, a second valve, and a first cavity laser caliper, thereby providing benchmark reference information for the subsequent setting of filtration control parameters of the air cabinet.

A200: traverse the air particle distribution particle size list, and match the reference distribution concentration list based on the reference concentration calibration table;

A300: extracting a sensitive distribution particle size set and a sensitive distribution particle size concentration set whose air particle distribution concentration is greater than the reference distribution concentration according to the air particle distribution concentration list and the reference distribution concentration list;

Specifically, in this embodiment, the air particle distribution particle size list records the particle size composition of the suspended particulate matter currently in the first cavity, including multiple particle size thresholds.

The multiple particle size thresholds of the air particle distribution particle size list are used to interact with the Internet or industry standards to match and obtain multiple benchmark distribution concentrations of the concentration limits of particulate matter in the air that does not cause air pollution at multiple particle size thresholds. The multiple benchmark distribution concentrations constitute the benchmark distribution concentration list.

The multiple reference distribution concentrations in the reference distribution concentration list are used as comparison standards, according to Multiple particle size threshold mappings are compared with the multiple air particle distribution concentrations in the air particle distribution concentration list to obtain the sensitive distribution particle size set composed of multiple air particle distribution particle sizes whose air particle distribution concentration is greater than the benchmark distribution concentration, and the corresponding sensitive distribution particle size concentration set of the multiple air particle distribution concentrations actually in the first cavity.

A400: Based on the sensitive distribution particle size set, performing activation optimization on the multi-stage filter of the air cabinet to generate multi-stage filter activation parameters, wherein the multi-stage filter activation parameters include an activation filter number parameter, a cycle filtration frequency parameter, and a cycle flow control parameter;

In one embodiment, as shown in FIG2 , based on the sensitive distribution particle size set, activation optimization is performed on the multi-stage filter of the air cabinet to generate multi-stage filter activation parameters, wherein the multi-stage filter activation parameters include an activation filter number parameter, a cycle filtration frequency parameter, and a cycle flow control parameter. Step A400 of the method provided in this application also includes:

A410: Matching a filter number set and a filter particle size range set according to the multi-stage filter of the air cabinet;

A420: Compare the sensitive distribution particle size set with the filter particle size interval set to obtain the activation filter number parameter;

A430: According to the activation filter number parameter, activate the filter control optimization component of the cloud application server cluster embedded in the air cabinet intelligent control system for improving air quality to perform filter control optimization and generate the cycle filter frequency parameter and the cycle flow control parameter.

In one embodiment, according to the activation filter number parameter, the filter of the cloud application server cluster embedded in the air cabinet intelligent control system for improving air quality is activated. The control optimization component performs filtering control optimization to generate the cyclic filtering frequency parameter and the cyclic flow control parameter. Previously, step A430 of the method provided by this application also includes:

A430-1: Load the filter number set and perform combination enumeration to generate a plurality of filter number combinations, wherein the number of filter number combinations is greater than or equal to 2 and less than or equal to the total number of numbers;

A430-2: Traverse the plurality of filter screen number combinations and collect a plurality of filtration control test data sets, wherein any filtration control test data set includes a cycle filtration frequency record data set, a cycle flow record data set, and sensitive particle filtration degree record data, where the sensitive particle filtration degree is equal to the sum of the ratio of the sensitive particle concentration to the concentration reduction;

A430-3: configuring a first filtration degree prediction network based on a first cyclic filtration frequency record data set of a first filtration control test data set of the plurality of filtration control test data sets, the first cyclic flow record data set, and the first sensitive particle filtration degree record data;

A430-4: configuring an Mth filtration degree prediction network based on the Mth cyclic filtration frequency record data set, the Mth cyclic flow record data set, and the Mth sensitive particle filtration degree record data of the Mth filtration control test data set;

A430-5: Based on the first filtering degree prediction network up to the M-th filtering degree prediction network, combined with the cross-leading optimization algorithm, configure the filtering control optimization component and embed it in the cloud application server cluster.

In one embodiment, according to the activation filter number parameter, the filter control optimization component of the cloud application server cluster embedded in the air cabinet intelligent control system for improving air quality is activated to perform filter control optimization, generate the cycle filter frequency parameter and the Circulation flow control parameters, the method step A430 provided in this application also includes:

A431: activating matching filter degree prediction networks from the first filter degree prediction network to the Mth filter degree prediction network according to the activated filter mesh number parameter;

A432: Configure a cycle flow constraint interval and a cycle number constraint interval, assign values to filter control parameters, and generate multiple filter control parameter assignment results. Any filter control parameter assignment result includes a cycle flow assignment result and a cycle filtering frequency assignment result.

A433: traversing the plurality of filter control parameter assignment results, and generating a plurality of filter degrees based on the filter degree prediction network;

A434: When any one of the plurality of filtration degrees is greater than or equal to a filtration degree threshold, the cyclic filtration frequency parameter and the cyclic flow control parameter are selected from the plurality of filtration control parameter assignment results.

Specifically, in this embodiment, each of the multiple filters constituting the multi-stage filter of the air cabinet has a corresponding number, and each filter can filter particles within a certain range. The filter number set composed of the plurality of filter numbers is matched to obtain the filter particle size range set composed of the particle size ranges that can intercept and filter particulate matter that can be intercepted by the plurality of filters.

The filter number set is loaded for combination enumeration to generate several filter number combinations. The number of filter numbers in each filter number combination is greater than or equal to 2 and less than or equal to the total number of numbers. For example, when there are a total of 5 filter numbers in the filter number set, a total of 15 filter number combinations can be obtained based on the combination enumeration.

The first filter number combination is obtained by randomly calling the plurality of filter number combinations. The first filter number combination includes a plurality of filter numbers, and the plurality of filter numbers are used to traverse the historical operating parameters of the first air cabinet to obtain a first filtration control test data set for filtering air particulate matter using the first filter number combination.

The first filtration control test data set includes multiple sets of cycle filtration frequency, cycle flow rate records, and sensitive particle filtration degree. Each set of data represents the sum of the reduction ratios of the post-filtration air sensitive particle concentration (sensitive particle filtration degree) after a certain number of cycles (cycle filtration frequency) at a certain unit time of filtered air flow rate (cycle flow rate). The same method used to obtain the first filtration control test data set is used to traverse the multiple filter screen number combinations and collect multiple filtration control test data sets.

A standard filtration degree prediction network is pre-built based on a back propagation neural network. The input data of the standard filtration degree prediction network are the cycle filtration frequency and the cycle flow rate, and the output data are the prediction results of the sensitive particle filtration degree.

Based on the several filter control test data sets, the first filter control test data set is called, and then the first cycle filter frequency record data set, the first cycle flow record data set and the first sensitive particle filtration degree record data of the first filter control test data set are split into multiple groups of cycle filter frequency record data-cycle flow record data and sensitive particle filtration degree record data.

The above multiple data sets are divided into training set, test set and validation set. The standard filtration degree prediction network is trained based on the training set, the standard filtration degree prediction network is tested based on the test set, and the output accuracy of the standard filtration degree prediction network is verified based on the validation set until the filtration degree prediction accuracy of the standard filtration degree prediction network is stably higher than 97%. It is believed that the training results can be reflected based on the filtration cycle frequency and circulating air flow under the condition of fixed filter composition. The first filtration degree prediction network of the sensitive particle filtration degree is promoted.

Using the same method, the Mth filtration degree prediction network is configured according to the Mth cycle filtration frequency record data set, the Mth cycle flow record data set and the Mth sensitive particle filtration degree record data of the Mth filtration control test data set of the plurality of filtration control test data sets.

Based on the M filtration degree prediction networks above, it is possible to accurately predict the filtration degree of sensitive particles under any filter combination when setting the filtration cycle frequency and air flow rate during the filtration process, thereby providing a reference for the selection of multi-stage filters for the first air cabinet.

In this embodiment, an association mapping is constructed between several filter mesh number combinations and the first filtration degree prediction network up to the Mth filtration degree prediction network, and then combined with the cross-leading optimization algorithm, the filter control optimization component is constructed, and the filter control optimization component includes M groups of filter mesh number combinations-filtration degree prediction networks.

After constructing and obtaining the filtering control optimization component, this embodiment embeds it into the cloud application server cluster. This embodiment elaborates in detail in the subsequent description how to determine the multi-stage filter control parameters for effective outdoor air filtration based on the cloud application server cluster optimization.

In this embodiment, the multiple filters required for the current air filtration are first determined. The specific method is to compare the sensitive distribution particle size set with the filter particle size interval set to obtain the activation filter number parameters composed of several filter numbers corresponding to several filter particle size intervals that can effectively filter several particle size particles of the sensitive distribution particle size set.

According to the activation filter number parameter, the filtering control optimization component of the cloud application server cluster embedded in the air cabinet intelligent control system for improving air quality is activated.

In the filtering control optimization component, M filter number combinations are traversed based on the activation filter number parameter to obtain a filter degree prediction network corresponding to a filter number combination that is consistent with the filter number combination of the activation filter number parameter, and activated as the matching filter degree prediction network.

It should be understood that the more air filtration cycles there are or the lower the flow rate during the cycle, the higher the air filtration quality will be. However, correspondingly, the time cost of filtering particulate matter in the air will be higher. Therefore, this embodiment presets a cycle flow constraint interval and a cycle number constraint interval, and controls the air filtration time cost based on the constraint interval. The numerical setting of the constraint interval in this embodiment is not limited in this embodiment.

Within the constraints of the cycle flow rate constraint interval and the cycle number constraint interval, the filtration control parameters are assigned values to generate a plurality of filtration control parameter assignment results, each of which includes a cycle flow rate assignment result and a cycle filtration frequency assignment result. The plurality of filtration control parameter assignment results are then synchronized with the filtration degree prediction network to perform air particle filtration degree prediction, thereby generating a plurality of filtration degrees.

Furthermore, in this embodiment, the sensitive reference particle size concentration set is obtained by mapping the sensitive distribution particle size set to the reference distribution concentration list, and the sum of the excess concentration ratios is calculated based on the sensitive distribution particle size concentration set and the sensitive reference particle size concentration set, and the calculation result is used as the filtration degree threshold.

When any one of the plurality of filtration degrees is greater than or equal to a filtration degree threshold, the cyclic filtration frequency parameter and the cyclic flow control parameter are selected from the plurality of filtration control parameter assignment results.

This embodiment determines the particle size of the air passing through the filter by analyzing the filter The filtration control parameters and the required filter combination are determined by the degree prediction network, achieving the technical effect of obtaining the activation filter number parameters, circulation flow control parameters and circulation filtration frequency parameters that can scientifically and effectively filter the external air.

A500: According to the activation filter number parameter, the circulation flow control parameter and the circulation filtration frequency parameter, the multi-stage filter of the air cabinet is controlled, and the air is passed into the multi-stage filter of the air cabinet for filtration. After completion, the air is passed into the room.

Specifically, in this embodiment, multiple filters of the air cabinet multi-stage filter are activated according to the activation filter number parameter mapping, and then the circulation flow control parameter and the circulation filtration frequency parameter are used to perform filtration control of the first air cabinet, so that air is passed into the air cabinet multi-stage filter through the first valve for filtration. After completion, the air is passed into the room through the second valve, so as to reduce the concentration of all particulate matter in the air to below the limit value of the benchmark concentration calibration table.

This embodiment achieves the technical effect of calling the filter screen of the air cabinet and scientifically setting the filtering parameters according to the distribution and concentration of air particles, thereby improving the air filtration quality of the air cabinet.

In one embodiment, the method steps provided by the present application further include:

A435: When the plurality of filtering degrees are all less than the filtering degree threshold, clustering the plurality of filtering control parameter assignment results according to the distance evaluation function and the clustering distance threshold to generate a plurality of clusters of filtering control parameter assignment results, wherein the plurality of clusters of filtering control parameter assignment results have a one-to-one correspondence with a plurality of head filtering control parameter assignment results of the maximum filtering degree;

Wherein, the distance evaluation function is:
d＝ln[(v ₁ -v ₂ ) ² +(f ₁ -f ₂ ) ² ],

Wherein, d represents the distance between any two filter control parameter assignment results, _v1 represents the cyclic flow assignment result of the first filter control parameter assignment result, _v2 represents the cyclic flow assignment result of the second filter control parameter assignment result, _f1 represents the cyclic filtering frequency assignment result of the first filter control parameter assignment result, and _f2 represents the cyclic filtering frequency assignment result of the second filter control parameter assignment result;

A436: Obtain a first cluster head filter control parameter assignment result of the plurality of head filter control parameter assignment results, obtain a second cluster non-head filter control parameter assignment results of the plurality of cluster filter control parameter assignment results, up to an Lth cluster non-head filter control parameter assignment results, wherein the second cluster up to the Lth cluster are different from the first cluster;

A437: Using the first cluster head filter control parameter assignment result as the moving target, and using the second cluster non-head filter control parameter assignment results up to the Lth cluster non-head filter control parameter assignment results as the moving starting points;

A438: configuring a moving distance step constraint interval and a moving constraint number, mutating the second cluster non-head filter control parameter assignment results up to the Lth cluster non-head filter control parameter assignment results based on the moving starting point and the moving target, and generating an expanded solution for the filter control parameters;

A439: Based on the filtration degree prediction network, according to the filtration control parameter expansion solution and the several filtration control parameter assignment results, sort the cyclic filtration frequency parameters and the cyclic flow control parameters that are greater than or equal to the filtration degree threshold.

In one embodiment, the method steps provided by the present application further include:

A439-1: When the filter control parameter expansion solution and the several filter control parameters As a result of the assignment, the number of solutions greater than or equal to the filtering degree threshold is equal to 0, and based on the cycle flow constraint interval and the cycle number constraint interval, the assignment results of the plurality of filtering control parameters are updated.

Specifically, it should be understood that if the several filtration degrees are all less than the filtration degree threshold, it proves that within the constraints of the current circulation flow constraint interval and the circulation number constraint interval, the filtration control parameters are assigned, and the several filtration control parameter assignment results generated are unable to effectively filter the current air.

Based on this, in this embodiment, when the several filtering degrees are all less than the filtering degree threshold, a distance evaluation function is preset to evaluate the similarity of the two sets of filtering control parameter assignment results, and the clustering distance threshold is used to judge whether the two sets of filtering control parameter assignment results are the same group of data based on the distance obtained by calculating the distance evaluation function.

The distance evaluation function is as follows:
d＝ln[(v ₁ -v ₂ ) ² +(f ₁ -f ₂ ) ² ],

d represents the distance between any two filter control parameter assignment results, _v1 represents the circulation flow assignment result of the first filter control parameter assignment result, _v2 represents the circulation flow assignment result of the second filter control parameter assignment result, _f1 represents the circulation filtering frequency assignment result of the first filter control parameter assignment result, and _f2 represents the circulation filtering frequency assignment result of the second filter control parameter assignment result.

This embodiment does not limit the value of the cluster distance threshold, and uses the cluster distance threshold as a unit to obtain multiple equidistant value intervals.

From several filter control parameter assignment results, any filter control parameter assignment result is extracted as a benchmark, and combined with the remaining filter control parameter assignment results to form several groups of filter control parameters. The numerical assignment results are substituted into the distance evaluation function to obtain several distances.

Then, a number of distances are traversed through the aforementioned division to obtain a number of equidistant numerical intervals, so as to perform clustering on the said number of filter control parameter assignment results and generate multiple clusters of filter control parameter assignment results, wherein the distance between any two filter control parameter assignment results in the multiple filter control parameter assignment results in each cluster of filter control parameter assignment results satisfies the said cluster distance threshold.

Then, based on the mapping relationship between the plurality of filtering degrees and the plurality of groups of filtering control parameter assignment results, a plurality of groups of filtering degrees are obtained for the plurality of clusters of filtering control parameter assignment results. The filtering control parameter assignment results within the cluster are serialized based on the filtering degrees, and the filtering control parameter assignment result with the largest filtering degree is used as the head filtering control parameter assignment result of the cluster of filtering control parameter assignment results. This is repeated in this manner, so that the plurality of clusters of filtering control parameter assignment results have a one-to-one correspondence with the plurality of head filtering control parameter assignment results with the largest filtering degrees.

Obtain a first cluster head filter control parameter assignment result of the multiple head filter control parameter assignment results, obtain a second cluster non-head filter control parameter assignment result of the multiple cluster filter control parameter assignment results until the Lth cluster non-head filter control parameter assignment result, wherein the second cluster until the Lth cluster are different clusters from the first cluster.

The first cluster head filter control parameter assignment result is used as a moving target, and the second cluster non-head filter control parameter assignment results up to the Lth cluster non-head filter control parameter assignment results are used as moving starting points.

The single data variable amount and the total variable times of the circulation flow assignment result and the circulation filtering frequency assignment result in the preset filtering control parameter assignment result are used as the moving distance step constraint interval and the moving constraint times.

The second cluster non-head filter is controlled based on the moving starting point and the moving target. The parameter assignment results are mutated until the L-th cluster non-head filter control parameter assignment results, to generate the filter control parameter extended solution including the second cluster filter control parameter extended solution set to the L-th cluster filter control parameter extended solution set.

Based on the filtration degree prediction network, the cyclic filtration frequency parameter and the cyclic flow control parameter that are greater than or equal to the filtration degree threshold are sorted according to the filtration control parameter expansion solution and the plurality of filtration control parameter assignment results.

When the number of solutions of the expanded filtering control parameter solution and the plurality of filtering control parameter assignment results that is greater than or equal to the filtering degree threshold is equal to 0, the plurality of filtering control parameter assignment results are assigned and updated based on the circulation flow constraint interval and the circulation number constraint interval.

This embodiment achieves the technical effect of quickly optimizing and obtaining circulation flow control parameters and circulation filtration frequency parameters that can scientifically and effectively filter external air based on assignment update.

Example 2

As shown in FIG3 , in order to more clearly explain the air cabinet intelligent control method for improving air quality, the embodiment of the present application provides an air cabinet multi-stage filter device, which is as follows:

The circulation passage includes: a main circulation passage 1; a first circulation branch passage 2, a second circulation branch passage 3 until the Qth circulation branch passage, the first circulation branch passage 2, the second circulation branch passage 3 until the Qth circulation branch passage are connected to the main circulation passage 1 through a three-way valve 4 respectively.

Air supply passage, the air supply passage includes: a main air supply passage 5; a first air supply branch passage 6, a second air supply branch passage 7 until the Qth air supply branch passage, the first air supply branch passage 6, the second air supply branch passage 7 until the Qth air supply branch passage respectively It is connected to the main air supply passage 5 through the three-way valve 4.

The first filter cavity 8, the first air supply branch passage 6 and the first circulation branch passage 2 are connected through the first filter cavity 8, and the filter screen of the first filter cavity 8 is deployed at the connection position between the first air supply branch passage 6 and the first filter cavity 8, and the first filter cavity 8 and the first circulation branch passage 2 are connected through a one-way valve 9.

The second filter cavity 10, the second air supply branch passage 7 and the second circulation branch passage 3 are connected through the second filter cavity 10, and the filter screen of the second filter cavity 10 is deployed at the connection position between the second air supply branch passage 7 and the second filter cavity 10, and the second filter cavity 10 and the second circulation branch passage 3 are connected through a one-way valve 9.

Until the Qth filter cavity, the Qth air supply branch passage and the Qth circulation branch passage are connected through the Qth filter cavity, and the filter of the Qth filter cavity is deployed at the connection position between the Qth air supply branch passage and the Qth filter cavity, and the Qth filter cavity and the Qth circulation branch passage are connected through a one-way valve 9.

The main circulation passage is sealed at both ends, one end of the main air supply passage in the gas flow direction is not sealed, and the other end is sealed. The air intake passage 11 is connected to the first filter cavity 8 through a one-way valve.

In this embodiment, outdoor air enters the multi-stage filter device of the air cabinet through the air intake passage 11 in the first cavity of the air cabinet, and is circulated and filtered in the corresponding activated filter cavity through the air supply branch passage and the circulation branch passage to achieve effective filtration of particulate matter in the air.

Example 3

Based on the same inventive concept as the air cabinet intelligent control method for improving air quality in the aforementioned embodiment, as shown in FIG4 , the present application provides an air cabinet intelligent control system for improving air quality, wherein the system includes:

A distribution state obtaining unit T11 is configured to activate a laser diameter measuring instrument deployed in the first cavity of the air cabinet when air enters the first cavity to obtain an air particle distribution state, wherein the air particle distribution state includes an air particle distribution particle size list and an air particle distribution concentration list;

a concentration list matching unit T12, configured to traverse the air particle distribution size list and match the reference distribution concentration list based on the reference concentration calibration table;

A data extraction execution unit T13 is configured to extract, based on the air particle distribution concentration list and the reference distribution concentration list, a sensitive distribution particle size set and a sensitive distribution particle size concentration set whose air particle distribution concentration is greater than the reference distribution concentration;

an activation parameter generating unit T14, configured to perform activation optimization on the multi-stage filter of the air cabinet according to the sensitive distribution particle size set, and generate multi-stage filter activation parameters, wherein the multi-stage filter activation parameters include an activation filter number parameter, a cycle filtration frequency parameter, and a cycle flow control parameter;

The air filtration execution unit T15 is used to control the air cabinet multi-stage filter according to the activation filter number parameter, the circulation flow control parameter and the circulation filtration frequency parameter, pass the air into the air cabinet multi-stage filter for filtration, and then pass the air into the room after completion.

In one embodiment, the distribution status obtaining unit T11 includes:

The first chamber includes a first valve and a second valve, wherein the first valve and The second valve is initially closed, and the first cavity is initially in a vacuum state;

Based on the Internet of Things, receive outdoor air density;

When a ventilation command is received, the first valve is opened to let air in. When the density deviation between the air density in the first cavity and the outdoor air density is less than or equal to the density deviation threshold, the first valve is closed and the first cavity laser diameter meter is activated to perform multi-position detection to obtain the air particle distribution state, wherein the air particle distribution state includes an air particle distribution particle size list and an air particle distribution concentration list.

In one embodiment, the activation parameter generation unit T14 further includes:

According to the multi-stage filter screen of the air cabinet, a filter screen number set and a filter particle size range set are matched;

Comparing the sensitive distribution particle size set with the filter particle size interval set to obtain the activation filter number parameter;

According to the activation filter number parameter, the filter control optimization component of the cloud application server cluster embedded in the air cabinet intelligent control system for improving air quality is activated to perform filter control optimization and generate the cycle filter frequency parameter and the cycle flow control parameter.

In one embodiment, the activation parameter generation unit T14 further includes:

Loading the filter number set and performing combination enumeration to generate a plurality of filter number combinations, wherein the number of filter number combinations is greater than or equal to 2 and less than or equal to the total number of numbers;

Traversing the plurality of filter screen number combinations, collecting a plurality of filtration control test data sets, wherein any filtration control test data set includes a cycle filtration frequency record data set, a cycle flow record data set, and sensitive particle filtration degree record data, where the sensitive particle filtration degree is equal to the sum of the ratio of the sensitive particle concentration to the concentration reduction;

configuring a first filtration degree prediction network according to a first cycle filtration frequency record data set of a first filtration control test data set of the plurality of filtration control test data sets, the first cycle flow record data set, and the first sensitive particle filtration degree record data;

until an Mth filtration degree prediction network is configured according to the Mth cycle filtration frequency record data set, the Mth cycle flow record data set and the Mth sensitive particle filtration degree record data of the Mth filtration control test data set of the plurality of filtration control test data sets;

According to the first filter degree prediction network up to the Mth filter degree prediction network, combined with the cross-leading optimization algorithm, the filter control optimization component is configured and embedded in the cloud application server cluster.

In one embodiment, the activation parameter generation unit T14 further includes:

activating matching filter degree prediction networks from the first filter degree prediction network to the Mth filter degree prediction network according to the activation filter mesh number parameter;

Configure a cycle flow constraint interval and a cycle number constraint interval, assign values to the filter control parameters, and generate a plurality of filter control parameter assignment results. Any filter control parameter assignment result includes a cycle flow assignment result and a cycle filtering frequency assignment result.

Traversing the plurality of filter control parameter assignment results, and generating a plurality of filter degrees based on the filter degree prediction network;

When any one of the plurality of filtration degrees is greater than or equal to a filtration degree threshold, the cyclic filtration frequency parameter and the cyclic flow control parameter are selected from the plurality of filtration control parameter assignment results.

In one embodiment, the activation parameter generation unit T14 further includes:

When the several filtering degrees are all less than the filtering degree threshold, according to the distance evaluation function and a clustering distance threshold, clustering the plurality of filter control parameter assignment results to generate a plurality of cluster filter control parameter assignment results, wherein the plurality of cluster filter control parameter assignment results have a one-to-one corresponding plurality of head filter control parameter assignment results of the maximum filtering degree;

Wherein, the distance evaluation function is:
d＝ln[(v ₁ -v ₂ ) ² +(f ₁ -f ₂ ) ² ],

Wherein, d represents the distance between any two filter control parameter assignment results, _v1 represents the cyclic flow assignment result of the first filter control parameter assignment result, _v2 represents the cyclic flow assignment result of the second filter control parameter assignment result, _f1 represents the cyclic filtering frequency assignment result of the first filter control parameter assignment result, and _f2 represents the cyclic filtering frequency assignment result of the second filter control parameter assignment result;

Obtaining a first cluster head filter control parameter assignment result of the plurality of head filter control parameter assignment results, obtaining a second cluster non-head filter control parameter assignment results of the plurality of cluster filter control parameter assignment results up to an Lth cluster non-head filter control parameter assignment results, wherein the second cluster up to the Lth cluster are different from the first cluster;

Taking the first cluster head filter control parameter assignment result as the moving target, and taking the second cluster non-head filter control parameter assignment result up to the Lth cluster non-head filter control parameter assignment result as the moving starting point;

Configuring a moving distance step constraint interval and a moving constraint number, mutating the second cluster non-head filter control parameter assignment results up to the Lth cluster non-head filter control parameter assignment results based on the moving starting point and the moving target, and generating an expanded solution for the filter control parameters;

Based on the filtering degree prediction network, according to the filtering control parameter expansion solution and the The result of assigning several filtering control parameters is selected to select the cyclic filtering frequency parameter and the cyclic flow control parameter that are greater than or equal to the filtering degree threshold.

In one embodiment, the activation parameter generation unit T14 further includes:

When the number of solutions of the expanded filtering control parameter solution and the plurality of filtering control parameter assignment results that is greater than or equal to the filtering degree threshold is equal to 0, the plurality of filtering control parameter assignment results are assigned and updated based on the circulation flow constraint interval and the circulation number constraint interval.

Any of the methods or steps described above may be stored as computer instructions or programs in various types of computer memories, and the computer instructions or programs may be recognized by various types of computer processors to implement any of the methods or steps described above.

Based on the above specific embodiments of the present invention, any improvements and modifications made to the present invention by those skilled in the art without departing from the principles of the present invention shall fall within the scope of patent protection of the present invention.

Claims (10)

An intelligent control method for an air cabinet for improving air quality is characterized in that the method is applied to an intelligent control system for an air cabinet for improving air quality, the system is communicatively connected to the air cabinet, the air cabinet includes a first cavity and an air cabinet multi-stage filter, and the air cabinet multi-stage filter can be selectively combined and activated, including: When air enters the first cavity of the air cabinet, a laser diameter measuring instrument deployed in the first cavity is activated to obtain an air particle distribution state, wherein the air particle distribution state includes an air particle distribution particle size list and an air particle distribution concentration list; Traversing the air particle distribution particle size list, and matching the reference distribution concentration list based on the reference concentration calibration table; Extracting a sensitive distribution particle size set and a sensitive distribution particle size concentration set whose air particle distribution concentration is greater than the reference distribution concentration according to the air particle distribution concentration list and the reference distribution concentration list; According to the sensitive distribution particle size set, performing activation optimization on the multi-stage filter of the air cabinet to generate multi-stage filter activation parameters, wherein the multi-stage filter activation parameters include an activation filter number parameter, a cycle filtration frequency parameter, and a cycle flow control parameter; According to the activation filter number parameter, the circulation flow control parameter and the circulation filtering frequency parameter, the multi-stage filter of the air cabinet is controlled, and the air is passed into the multi-stage filter of the air cabinet for filtration, and after completion, the air is passed into the room. The method according to claim 1, characterized in that when air enters the first cavity of the air cabinet, a laser diameter measuring instrument deployed in the first cavity is activated to obtain the air particle distribution state, wherein the air particle distribution state includes an air particle distribution particle size list and List of air particle distribution concentrations, including: The first cavity comprises a first valve and a second valve, wherein the first valve and the second valve are initially closed, and the first cavity is initially in a vacuum state; Based on the Internet of Things, receive outdoor air density; When a ventilation command is received, the first valve is opened to let air in. When the density deviation between the air density in the first cavity and the outdoor air density is less than or equal to the density deviation threshold, the first valve is closed and the first cavity laser diameter meter is activated to perform multi-position detection to obtain the air particle distribution state, wherein the air particle distribution state includes an air particle distribution particle size list and an air particle distribution concentration list. The method according to claim 1 is characterized in that, based on the sensitive distribution particle size set, activation optimization is performed on the multi-stage filter of the air cabinet to generate multi-stage filter activation parameters, wherein the multi-stage filter activation parameters include an activation filter number parameter, a cycle filtration frequency parameter, and a cycle flow control parameter, including: According to the multi-stage filter screen of the air cabinet, a filter screen number set and a filter particle size range set are matched; Comparing the sensitive distribution particle size set with the filter particle size interval set to obtain the activation filter number parameter; According to the activation filter number parameter, the filter control optimization component of the cloud application server cluster embedded in the air cabinet intelligent control system for improving air quality is activated to perform filter control optimization and generate the cycle filter frequency parameter and the cycle flow control parameter. The method according to claim 3 is characterized in that, according to the activation filter number parameter, the cloud embedded in the air cabinet intelligent control system for improving air quality is activated The filtering control optimization component of the application server cluster performs filtering control optimization to generate the cyclic filtering frequency parameter and the cyclic flow control parameter, which includes: Loading the filter number set and performing combination enumeration to generate a plurality of filter number combinations, wherein the number of filter number combinations is greater than or equal to 2 and less than or equal to the total number of numbers; Traversing the plurality of filter screen number combinations, collecting a plurality of filtration control test data sets, wherein any filtration control test data set includes a cycle filtration frequency record data set, a cycle flow record data set, and sensitive particle filtration degree record data, where the sensitive particle filtration degree is equal to the sum of the ratio of the sensitive particle concentration to the concentration reduction; configuring a first filtration degree prediction network according to a first cycle filtration frequency record data set of a first filtration control test data set of the plurality of filtration control test data sets, the first cycle flow record data set, and the first sensitive particle filtration degree record data; until an Mth filtration degree prediction network is configured according to the Mth cycle filtration frequency record data set, the Mth cycle flow record data set and the Mth sensitive particle filtration degree record data of the Mth filtration control test data set of the plurality of filtration control test data sets; According to the first filter degree prediction network up to the Mth filter degree prediction network, combined with the cross-leading optimization algorithm, the filter control optimization component is configured and embedded in the cloud application server cluster. The method according to claim 4 is characterized in that, based on the activation filter number parameter, activating the filter control optimization component of the cloud application server cluster embedded in the air cabinet intelligent control system for improving air quality to perform filter control optimization and generate the cycle filtration frequency parameter and the cycle flow control parameter, including: According to the activation filter number parameter, from the first filter prediction network to the The Mth filter degree prediction network activates the matched filter degree prediction network; Configure a cycle flow constraint interval and a cycle number constraint interval, assign values to the filter control parameters, and generate a plurality of filter control parameter assignment results. Any filter control parameter assignment result includes a cycle flow assignment result and a cycle filtering frequency assignment result. Traversing the plurality of filter control parameter assignment results, and generating a plurality of filter degrees based on the filter degree prediction network; When any one of the plurality of filtration degrees is greater than or equal to a filtration degree threshold, the cyclic filtration frequency parameter and the cyclic flow control parameter are selected from the plurality of filtration control parameter assignment results. The method according to claim 5, further comprising: When the plurality of filtering degrees are all less than the filtering degree threshold, clustering the plurality of filtering control parameter assignment results according to a distance evaluation function and a clustering distance threshold is performed to generate a plurality of clusters of filtering control parameter assignment results, wherein the plurality of clusters of filtering control parameter assignment results have a one-to-one correspondence with a plurality of head filtering control parameter assignment results of the maximum filtering degree; Wherein, the distance evaluation function is:
d＝ln[(v ₁ -v ₂ ) ² +(f ₁ -f ₂ ) ² ], Wherein, d represents the distance between any two filter control parameter assignment results, _v1 represents the cyclic flow assignment result of the first filter control parameter assignment result, _v2 represents the cyclic flow assignment result of the second filter control parameter assignment result, _f1 represents the cyclic filtering frequency assignment result of the first filter control parameter assignment result, and _f2 represents the cyclic filtering frequency assignment result of the second filter control parameter assignment result; Obtain the first cluster head filter control parameter of the plurality of head filter control parameter assignment results obtaining the second cluster non-head filter control parameter assignment results of the plurality of clusters of filter control parameter assignment results, up to the Lth cluster non-head filter control parameter assignment results, wherein the second cluster up to the Lth cluster are different from the first cluster; Taking the first cluster head filter control parameter assignment result as the moving target, and taking the second cluster non-head filter control parameter assignment result up to the Lth cluster non-head filter control parameter assignment result as the moving starting point; Configuring a moving distance step constraint interval and a moving constraint number, mutating the second cluster non-head filter control parameter assignment results up to the Lth cluster non-head filter control parameter assignment results based on the moving starting point and the moving target, and generating an expanded solution for the filter control parameters; Based on the filtration degree prediction network, the cyclic filtration frequency parameter and the cyclic flow control parameter that are greater than or equal to the filtration degree threshold are sorted according to the filtration control parameter expansion solution and the plurality of filtration control parameter assignment results. The method according to claim 6 is characterized in that it further includes: when the number of solutions of the expanded solution of the filtering control parameter and the multiple filtering control parameter assignment results that is greater than or equal to the filtering degree threshold is equal to 0, based on the cycle flow constraint interval and the cycle number constraint interval, the multiple filtering control parameter assignment results are assigned and updated. An air cabinet multi-stage filter device, characterized by comprising: A circulation path, the circulation path comprising: Main circulation pathway; The first circulation branch path, the second circulation branch path until the Qth circulation branch path, the first circulation branch path, the second circulation branch path until the Qth circulation branch path The branch passages are connected to the main circulation passage through three-way valves respectively; An air supply passage, the air supply passage comprising: Main air supply passage; The first air supply branch passage, the second air supply branch passage, and the Qth air supply branch passage are connected to the main air supply passage through a three-way valve. a first filter cavity, wherein the first air supply branch passage and the first circulation branch passage are connected through the first filter cavity, and the filter screen of the first filter cavity is disposed at a position where the first air supply branch passage and the first filter cavity are connected, and the first filter cavity and the first circulation branch passage are connected through a one-way valve; a second filter cavity, wherein the second air supply branch passage and the second circulation branch passage are connected through the second filter cavity, and the filter screen of the second filter cavity is disposed at a position where the second air supply branch passage and the second filter cavity are connected, and the second filter cavity and the second circulation branch passage are connected through a one-way valve; Until the Qth filter cavity, the Qth air supply branch passage and the Qth circulation branch passage are connected through the Qth filter cavity, and the filter of the Qth filter cavity is disposed at the communication position between the Qth air supply branch passage and the Qth filter cavity, and the Qth filter cavity and the Qth circulation branch passage are connected through a one-way valve; Wherein, both ends of the main circulation passage are sealed, and one end of the main air supply passage in the gas flow direction is not sealed, while the other end is sealed. The multi-stage filter device for an air cabinet according to claim 8, further comprising: An air intake passage is connected to the first filter cavity through a one-way valve. An intelligent control system for an air cabinet for improving air quality, characterized in that the system comprises: a distribution state obtaining unit, configured to activate a laser diameter measuring instrument disposed in the first cavity when air enters the first cavity of the air cabinet, and obtain an air particle distribution state, wherein the air particle distribution state includes an air particle distribution particle size list and an air particle distribution concentration list; a concentration list matching unit, configured to traverse the air particle distribution size list and match the reference distribution concentration list based on the reference concentration calibration table; a data extraction execution unit, configured to extract, based on the air particle distribution concentration list and the reference distribution concentration list, a sensitive distribution particle size set and a sensitive distribution particle size concentration set whose air particle distribution concentration is greater than the reference distribution concentration; an activation parameter generating unit, configured to perform activation optimization on the multi-stage filter of the air cabinet according to the sensitive distribution particle size set, and generate multi-stage filter activation parameters, wherein the multi-stage filter activation parameters include an activation filter number parameter, a cycle filtration frequency parameter, and a cycle flow control parameter; The air filtration execution unit is used to control the multi-stage filter of the air cabinet according to the activation filter number parameter, the circulation flow control parameter and the circulation filtration frequency parameter, pass the air into the multi-stage filter of the air cabinet for filtration, and then pass the air into the room after completion.