CN116684330A - Traffic prediction method, device, equipment and storage medium based on artificial intelligence - Google Patents

Abstract

The embodiment of the present application belongs to the field of artificial intelligence and relates to a traffic prediction method based on artificial intelligence, including: obtaining network traffic; constructing a training set and a test set based on the network traffic; preprocessing the training set and the test set to obtain a target training set and the target test set; build an initial prediction model based on BP neural network structure; optimize the weight threshold of the initial prediction model based on the target cuckoo search algorithm to obtain the specified prediction model; use the training set and test set to train and test the specified prediction model , to obtain a well-trained traffic forecasting model; based on the traffic forecasting model, traffic forecasting processing is performed on the network traffic data. The present application also provides an artificial intelligence-based flow prediction device, computer equipment, and storage media. In addition, this application also relates to blockchain technology, and the traffic forecasting model can be stored in the blockchain. The present application can effectively improve the accuracy of the traffic prediction of the traffic prediction model and the iterative efficiency of the model training.

Description

Flow prediction method, device, equipment and storage medium based on artificial intelligence

technical field

This application relates to the field of artificial intelligence development technology and the field of financial technology, and in particular to an artificial intelligence-based flow prediction method, device, computer equipment and storage medium.

Background technique

With the rapid development of network communication and computer technology, various business systems, such as insurance systems, banking systems, etc., have various forms of online application services. These applications require stable network support, and put forward high requirements for network service quality, flow control and network management. Therefore, it is very necessary to analyze and predict network traffic. At present, there are already many analysis and prediction models for network traffic, but there are still many problems: on the one hand, network traffic is complex in time and space; on the other hand, network characteristics in various application scenarios are quite different. Complexity of predictive model building and training.

BP network is currently the most widely used forecasting model, and it stands out in dealing with nonlinear forecasting problems. At present, business systems need to use BP network to predict network traffic. However, the traditional BP network relies on the backpropagation of errors, which requires continuous iterations, and the prediction accuracy of the prediction model is not high.

Contents of the invention

The purpose of the embodiments of the present application is to propose an artificial intelligence-based flow prediction method, device, computer equipment, and storage medium to solve the problem that the existing way of using BP network to predict network flow depends on the backpropagation of errors. Continuous iterations are carried out, and the prediction accuracy of the prediction model is not high.

In order to solve the above technical problems, the embodiment of the present application provides an artificial intelligence-based traffic forecasting method, which adopts the following technical solutions:

Obtain network traffic within a preset historical time period;

Constructing samples of the network traffic based on a preset time period division unit to obtain corresponding training sets and test sets;

performing data preprocessing on the training set and the test set to obtain corresponding target training sets and target test sets;

Determine the BP neural network structure, and build an initial prediction model based on the BP neural network structure;

Based on the target cuckoo search algorithm, the weight threshold of the initial prediction model is optimized to obtain the optimized specified prediction model; wherein, the target cuckoo search algorithm is an improved step size update method of the original cuckoo search algorithm The obtained optimized cuckoo search algorithm;

using the training set to train the designated prediction model, and testing the trained designated prediction model through the test set to obtain a trained traffic prediction model;

Perform traffic prediction processing on the network traffic data to be processed based on the traffic prediction model.

Further, the step of constructing samples of the network traffic based on the preset time period division unit to obtain corresponding training sets and test sets specifically includes:

Obtain the preset time period division unit;

dividing each day in the historical time period into corresponding multiple unit time periods based on the time period division unit;

Constructing a training set corresponding to the network traffic by using the network traffic in the same unit time period as the same type of samples;

Get the preset time value;

randomly selecting specified data corresponding to the time value from the training set based on the time value;

Use the specified data as the test set.

Further, the step of performing data preprocessing on the training set and the test set to obtain the corresponding target training set and target test set specifically includes:

Performing data cleaning on the training set and the test set to obtain corresponding first training set and first test set;

performing normalization processing on the first training set and the first test set to obtain corresponding second training sets and second test sets;

The second training set is used as the target training set, and the second test set is used as the target test set.

Further, before the step of optimizing the weight threshold of the initial prediction model based on the target cuckoo search algorithm to obtain the optimized specified prediction model, it also includes:

Get the original cuckoo search algorithm;

Obtain the step size update method in the initial cuckoo search algorithm;

The step size update method of the cuckoo search algorithm is improved based on a preset formula to obtain the optimized target cuckoo search algorithm.

Further, the step of optimizing the weight threshold of the initial prediction model based on the target cuckoo search algorithm to obtain the optimized specified prediction model specifically includes:

Carry out parameter initialization to described target cuckoo search algorithm;

Randomly generate a plurality of bird's nest positions, and encode the initial weight threshold of the initial prediction model as the initial bird's nest position of the target cuckoo search algorithm;

Determine the fitness function of the target cuckoo search algorithm, and calculate the fitness of each nest position based on the fitness function;

Using the target cuckoo search algorithm, perform global iterative optimization processing according to the fitness, and find the corresponding global optimal position from all the bird's nest positions;

Determine whether the current number of iterations meets the preset maximum number of iterations;

If yes, the global optimal position is used as the optimal weight threshold of the initial prediction model to obtain the optimized specified prediction model.

Further, the step of performing traffic prediction processing on the network traffic data to be processed based on the traffic prediction model specifically includes:

Obtain pending network traffic data;

inputting the network traffic data into the traffic prediction model;

Prediction processing is performed on the network traffic data by the traffic prediction model, and a prediction result corresponding to the network traffic data is output.

Further, after the step of using the training set to train the designated prediction model, and using the test set to test the trained designated prediction model to obtain a trained traffic prediction model, further include:

Obtain the preset model ID;

determining a target storage sub-area matching the model identifier from multiple storage sub-areas included in the blockchain;

The traffic forecasting model is stored in the target storage sub-area.

In order to solve the above technical problems, the embodiment of the present application also provides an artificial intelligence-based flow prediction device, which adopts the following technical solutions:

A first acquisition module, configured to acquire network traffic within a preset historical time period;

The first building module is used to construct samples of the network traffic based on the preset time period division unit, and obtain corresponding training sets and test sets;

A processing module, configured to perform data preprocessing on the training set and the test set to obtain corresponding target training sets and target test sets;

The second building block is used to determine the BP neural network structure, and construct an initial prediction model based on the BP neural network structure;

The optimization module is used to optimize the weight threshold of the initial prediction model based on the target cuckoo search algorithm to obtain an optimized specified prediction model; wherein, the target cuckoo search algorithm is the step size of the original cuckoo search algorithm The optimized cuckoo search algorithm obtained after the update method is improved;

A training module, configured to use the training set to train the designated prediction model, and test the trained designated prediction model through the test set to obtain a trained traffic prediction model;

A forecasting module, configured to perform traffic forecasting processing on the network traffic data to be processed based on the traffic forecasting model.

In order to solve the above technical problems, the embodiment of the present application also provides a computer device, which adopts the following technical solutions:

Obtain network traffic within a preset historical time period;

Constructing samples of the network traffic based on a preset time period division unit to obtain corresponding training sets and test sets;

performing data preprocessing on the training set and the test set to obtain corresponding target training sets and target test sets;

Determine the BP neural network structure, and build an initial prediction model based on the BP neural network structure;

Based on the target cuckoo search algorithm, the weight threshold of the initial prediction model is optimized to obtain the optimized specified prediction model; wherein, the target cuckoo search algorithm is an improved step size update method of the original cuckoo search algorithm The obtained optimized cuckoo search algorithm;

using the training set to train the designated prediction model, and testing the trained designated prediction model through the test set to obtain a trained traffic prediction model;

Perform traffic prediction processing on the network traffic data to be processed based on the traffic prediction model.

In order to solve the above technical problems, the embodiment of the present application also provides a computer-readable storage medium, which adopts the following technical solution:

Obtain network traffic within a preset historical time period;

Constructing samples of the network traffic based on a preset time period division unit to obtain corresponding training sets and test sets;

performing data preprocessing on the training set and the test set to obtain corresponding target training sets and target test sets;

Determine the BP neural network structure, and build an initial prediction model based on the BP neural network structure;

Based on the target cuckoo search algorithm, the weight threshold of the initial prediction model is optimized to obtain the optimized specified prediction model; wherein, the target cuckoo search algorithm is an improved step size update method of the original cuckoo search algorithm The obtained optimized cuckoo search algorithm;

using the training set to train the designated prediction model, and testing the trained designated prediction model through the test set to obtain a trained traffic prediction model;

Perform traffic prediction processing on the network traffic data to be processed based on the traffic prediction model.

Compared with the prior art, the embodiments of the present application mainly have the following beneficial effects:

In the embodiment of the present application, the network traffic in the preset historical time period is first obtained; then, based on the preset time period division unit, the network traffic is sample-constructed to obtain the corresponding training set and test set; and the training set Carry out data preprocessing with the test set to obtain the corresponding target training set and target test set; then determine the BP neural network structure, and build an initial prediction model based on the BP neural network structure; follow-up based on the target cuckoo search algorithm for Perform weight threshold optimization on the initial prediction model to obtain an optimized designated prediction model; further use the training set to train the designated prediction model, and use the test set to test the trained designated prediction model , to obtain a well-trained traffic prediction model; finally, perform traffic prediction processing on the network traffic data to be processed based on the traffic prediction model. In the embodiment of the present application, after constructing the corresponding training set and test set based on the obtained network traffic in the historical time period, the target cuckoo search algorithm obtained by improving the step size update method of the original cuckoo search algorithm is used to The weight threshold of the initial prediction model is optimized to construct a specified prediction model, and then the training set and test set are used to train and test the specified prediction model to obtain a trained traffic prediction model, which can effectively improve the network traffic prediction performance of the traffic prediction model Accuracy and iterative efficiency of model training.

Description of drawings

In order to illustrate the solution in this application more clearly, a brief introduction will be given below to the accompanying drawings that need to be used in the description of the embodiments of the application. Obviously, the accompanying drawings in the following description are some embodiments of the application. Ordinary technicians can also obtain other drawings based on these drawings on the premise of not paying creative work.

FIG. 1 is an exemplary system architecture diagram to which the present application can be applied;

Fig. 2 is a flow chart of an embodiment of the artificial intelligence-based traffic forecasting method according to the present application;

FIG. 3 is a schematic structural diagram of an embodiment of an artificial intelligence-based flow prediction device according to the present application;

Fig. 4 is a schematic structural diagram of an embodiment of a computer device according to the present application.

Detailed ways

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field of the application; the terms used herein in the description of the application are only to describe specific embodiments The purpose is not to limit the present application; the terms "comprising" and "having" and any variations thereof in the specification and claims of the present application and the description of the above drawings are intended to cover non-exclusive inclusion. The terms "first", "second" and the like in the description and claims of the present application or the above drawings are used to distinguish different objects, rather than to describe a specific order.

Reference herein to an "embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the present application. The occurrences of this phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is understood explicitly and implicitly by those skilled in the art that the embodiments described herein can be combined with other embodiments.

In order to enable those skilled in the art to better understand the solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

As shown in FIG. 1 , a system architecture 100 may include terminal devices 101 , 102 , 103 , a network 104 and a server 105 . The network 104 is used as a medium for providing communication links between the terminal devices 101 , 102 , 103 and the server 105 . Network 104 may include various connection types, such as wires, wireless communication links, or fiber optic cables, among others.

Users can use terminal devices 101 , 102 , 103 to interact with server 105 via network 104 to receive or send messages and the like. Various communication client applications can be installed on the terminal devices 101, 102, 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, and the like.

The terminal devices 101, 102, 103 can be various electronic devices that have display screens and support web browsing, including but not limited to smart phones, tablet computers, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III, moving picture Expert Compression Standard Audio Layer 3), MP4 (Moving PictureExperts Group Audio Layer IV, Moving Picture Experts Compression Standard Audio Layer 4) Players, Laptop Portable Computers and Desktop Computers, etc.

The server 105 may be a server that provides various services, such as a background server that provides support for pages displayed on the terminal devices 101 , 102 , 103 .

It should be noted that the artificial intelligence-based traffic prediction method provided in the embodiment of the present application is generally executed by a server/terminal device, and accordingly, the artificial intelligence-based traffic prediction device is generally set in the server/terminal device.

The embodiments of the present application may acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. .

Artificial intelligence basic technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation/interaction systems, and mechatronics. Artificial intelligence software technology mainly includes computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning/deep learning.

It should be understood that the numbers of terminal devices, networks and servers in Fig. 1 are only illustrative. According to the implementation needs, there can be any number of terminal devices, networks and servers.

Continuing to refer to FIG. 2 , it shows a flow chart of an embodiment of an artificial intelligence-based traffic forecasting method according to the present application. The artificial intelligence-based traffic forecasting method comprises the following steps:

Step S201, acquiring network traffic within a preset historical time period.

In this embodiment, the electronic device (such as the server/terminal device shown in FIG. 1 ) on which the traffic prediction method based on artificial intelligence runs can obtain the network traffic in the historical time period through a wired connection or a wireless connection. . It should be pointed out that the above wireless connection methods may include but not limited to 3G/4G/5G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other wireless connections known or developed in the future. connection method. The traffic prediction method based on artificial intelligence provided in this application can be applied in the business scenario of network traffic prediction in business scenarios. To illustrate the business scenario of the insurance system, the network data of the current time period can be input into the traffic forecasting model built by this application, so that the network traffic forecast of the next time period after the current time period can be output through the traffic forecasting model data. Wherein, the above-mentioned network traffic specifically refers to network traffic in the business system. Specifically, the business system may be any system in an insurance system, a banking system, a transaction system, an order system, and the like. In addition, there is no specific limitation on the selection of the above-mentioned historical time period, which can be set according to actual business usage requirements, for example, it can be within one month before the current time.

Step S202, constructing samples of the network traffic based on the preset time period division unit to obtain corresponding training sets and test sets.

In this embodiment, the above-mentioned division unit based on the preset time period performs sample construction on the network traffic, and obtains the specific implementation process of the corresponding training set and test set. Further detailed descriptions will not be elaborated here.

Step S203, performing data preprocessing on the training set and the test set to obtain a corresponding target training set and target test set.

In this embodiment, the above data preprocessing may specifically include data cleaning processing and normalization processing. The training set and the testing set are collectively referred to as a sample set. By cleaning the data of the sample set, the noise data in the sample set can be eliminated. By normalizing the sample set, the sample set is converted to [0,1], so as to improve the training speed of the subsequent traffic prediction model.

Step S204, determining the BP neural network structure, and constructing an initial prediction model based on the BP neural network structure.

In this embodiment, the BP neural network can be divided into two parts, BP and neural network. BP is the abbreviation of BackPropagation, which means backpropagation. The BP neural network can learn and store a large number of input-output pattern mapping relationships without revealing the mathematical equations describing the mapping relationship in advance. Its learning rule is to use the steepest descent method to continuously adjust the weights and thresholds of the network through backpropagation to minimize the sum of squared errors of the network. Its main characteristics are: the signal is propagated forward, and the error is propagated backward. Among them, the BP neural network structure includes: the input layer is each time period and the corresponding network traffic, the output is the predicted traffic value, and the parameters such as the weight threshold and the maximum number of iterations of the BP neural network are initialized.

Step S205, based on the target cuckoo search algorithm, optimize the weight threshold of the initial prediction model to obtain an optimized specified prediction model. Wherein, the target cuckoo search algorithm is an optimized cuckoo search algorithm obtained by improving the step size update method of the original cuckoo search algorithm.

In this embodiment, the target cuckoo search algorithm is an optimized cuckoo search algorithm obtained by improving the step size update method of the original cuckoo search algorithm. The cuckoo search algorithm, also known as the CS algorithm, is a swarm intelligence optimization algorithm that solves optimization problems by simulating the cuckoo parasitic brood. Because it uses the Levy flight search strategy, it is more effective than other optimization algorithms. The path and position update strategy of the traditional CS algorithm is as follows: x _i (t) = x _i (t-1) + a⊙Levy (λ); in the above formula: x _i (t) and x _i (t-1) respectively is the i-th solution at the t-th and t-1 iterations; a is the step factor, used to determine the search scale; ⊙ is the dot product; Levy(λ) is the probability distribution that obeys Levy. Among them, the above-mentioned target cuckoo search algorithm is used to optimize the weight threshold of the initial prediction model to obtain the specific implementation process of the optimized specified prediction model. This application will further describe this in detail in subsequent specific embodiments. I won't elaborate too much here. By using the target cuckoo search algorithm to optimize the weight threshold of the initial prediction model, the accuracy and iteration efficiency of the parameter optimization of the initial prediction model can be guaranteed. In this embodiment, by integrating the swarm intelligence optimization algorithm and the deep neural network, a CS-BP traffic prediction model that can be used in complex network environments is constructed, which can provide guarantees for online system network traffic monitoring, network performance analysis, and network management.

Step S206, using the training set to train the designated prediction model, and using the test set to test the trained designated prediction model to obtain a trained traffic prediction model.

In this embodiment, after using the training set to train the specified prediction model, the test set is used to test the trained specified prediction model, and the actual value (ie predicted value) and the real value are calculated. Between the accuracy index, and evaluate the accuracy index (for example, evaluate whether the accuracy rate is greater than the set threshold), if the evaluation is passed, then use the specified prediction model that passed the evaluation as the trained traffic prediction model.

Step S207, performing traffic prediction processing on the network traffic data to be processed based on the traffic prediction model.

In this embodiment, the specific implementation process of performing traffic prediction processing on the network traffic data to be processed based on the traffic prediction model above will be further described in detail in subsequent specific embodiments in this application, and will not be described here Explain more.

The application first obtains the network traffic in the preset historical time period; then constructs samples of the network traffic based on the preset time period division unit to obtain the corresponding training set and test set; and compares the training set and the Perform data preprocessing on the test set to obtain the corresponding target training set and target test set; then determine the BP neural network structure, and build an initial prediction model based on the BP neural network structure; follow-up based on the target cuckoo search algorithm for the described Optimizing the weight threshold of the initial prediction model to obtain the optimized specified prediction model; further using the training set to train the specified prediction model, and testing the trained specified prediction model through the test set to obtain A trained traffic prediction model; finally, traffic prediction processing is performed on the network traffic data to be processed based on the traffic prediction model. In this application, after constructing the corresponding training set and test set based on the obtained network traffic in the historical time period, the initial prediction is made by using the target cuckoo search algorithm obtained by improving the step size update method of the original cuckoo search algorithm. The weight threshold of the model is optimized to build a specified prediction model, and then use the training set and test set to train and test the specified prediction model to obtain a trained traffic prediction model, which can effectively improve the network traffic prediction accuracy of the traffic prediction model And the iterative efficiency of model training.

In some optional implementation manners, step S202 includes the following steps:

Get the preset time period division unit.

In this embodiment, there is no limitation on the value of the division unit of the above time period, which can be set according to actual usage requirements, for example, it can be set to 1 hour.

Divide each day in the historical time period into corresponding multiple unit time periods based on the time period division unit.

In this embodiment, if the time segment division unit is 1 hour, then 24 hours per day may be divided into 24 unit time segments with a corresponding time length of 1 hour.

A training set corresponding to the network traffic is constructed by using the network traffic in the same unit time period as samples of the same class.

In this embodiment, if the historical time period is that the business system is in the previous week from the current time, then the network traffic of the business system in the week can be selected as the training set, 24 hours a day, and the same unit time period every day is regarded as a type, that is, it can be divided into 24 training sets within a week.

Get the preset time value.

In this embodiment, there is no limitation on the value of the division unit of the above time period, which can be set according to actual usage requirements, for example, it can be set to 6.

Based on the time value, randomly select specified data corresponding to the time value from the training set.

In this embodiment, referring to the above example, after dividing the network traffic of the business system in the week into 24 training sets, the network traffic corresponding to a unit time period of 6 hours is randomly selected from the 24 training sets as a test set.

Use the specified data as the test set.

This application obtains the preset time period division unit, and divides each day in the historical time period into corresponding multiple unit time periods based on the time period division unit; and then uses the same unit time period The network traffic is used as the same type of samples to construct a training set corresponding to the network traffic; after that, the preset time value is obtained; and based on the time value, the time value corresponding to the time value is randomly selected from the training set The specified data of , and use the specified data as the test set. This application divides and processes the network traffic in the historical time period based on the use of the time period division unit, so as to realize the rapid construction and obtain the required training set, and then processes the constructed training set based on the use of the time value to realize Quickly construct the required test set, improve the intelligence and standardization of training set and test set construction, and ensure the randomness of the obtained test set.

In some optional implementation manners of this embodiment, step S203 includes the following steps:

Data cleaning is performed on the training set and the test set to obtain a corresponding first training set and a first test set.

In this embodiment, the above data cleaning may include removing noise data in the training set and the testing set.

Perform normalization processing on the first training set and the first testing set to obtain corresponding second training sets and second testing sets.

In this embodiment, the training set and the testing set can be normalized by using a normalization formula, so as to transform the training set and the testing set into a range between [0,1].

The second training set is used as the target training set, and the second test set is used as the target test set.

The present application obtains the corresponding first training set and first test set by performing data cleaning on the training set and the test set; then normalizes the first training set and the first test set , to obtain a corresponding second training set and a second test set; subsequently, the second training set is used as the target training set, and the second test set is used as the target test set. In this application, by performing data cleaning and normalization processing on the training set and the test set, the feature data that meets the requirements of the model can be quickly and accurately obtained, so that when the training set and the test set are used to construct the traffic prediction model, It can effectively improve the training speed of the traffic prediction model.

In some optional implementation manners, before step S205, the above-mentioned electronic device may also perform the following steps:

Get the original cuckoo search algorithm.

Obtain the step size update method in the initial cuckoo search algorithm.

In this embodiment, the path and position update strategy of the traditional cuckoo search algorithm are as follows: x _i (t) = x _i (t-1) + a⊙Levy (λ); in the above formula: x _i (t) and x _i (t-1) are the i-th solution at the t and t-1 iterations respectively; a is the step size factor, which is used to determine the search scale; ⊙ is the point product; probability distribution.

The step size update method of the cuckoo search algorithm is improved based on a preset formula to obtain the optimized target cuckoo search algorithm.

In this embodiment, because the cuckoo search algorithm adopts the Levy search strategy, its step size has strong randomness. Although it has a strong global search ability, its local optimization ability is weak. In order to balance the global search and local optimization capabilities in the iterative process, it is considered to improve the step size update method, so that the step size can be adaptively adjusted dynamically during iteration. The improvement method is as follows: where x _best is the current optimal value. According to the improvement of the step size factor mentioned above, the CS algorithm has a larger search step size in the early stage and a strong global search ability; as the approach to the global optimal solution, the step size gradually becomes smaller, which improves the local optimization ability. To sum up, through the above-mentioned adaptive adjustment of the step size of the CS algorithm, the global search and local optimization capabilities of the CS optimization algorithm are balanced, and the convergence speed is improved. In this embodiment, by introducing a dynamic adaptation factor to improve the position update of the CS algorithm, the problem of unbalanced global optimization in the early stage and local optimization in the later stage can be solved. The subsequent use of the improved CS algorithm to optimize the initial weights and thresholds of the BP network can further improve the accuracy of network traffic prediction and the iterative efficiency of model training.

The present application obtains the original cuckoo search algorithm; then obtains the step size update method in the initial cuckoo search algorithm; subsequently improves the step size update method of the cuckoo search algorithm based on the preset formula, and obtains the optimized The target cuckoo search algorithm. This application uses the preset formula to improve the step size update method of the traditional cuckoo search algorithm, and can quickly obtain the optimized target cuckoo search algorithm, and realizes the adaptive adjustment of the step size of the traditional cuckoo search algorithm , which balances the global search and local optimization capabilities of the traditional cuckoo search algorithm, and effectively improves the convergence speed. The subsequent use of the target cuckoo search algorithm to optimize the weight threshold of the initial prediction model to construct the traffic prediction model can further improve the network traffic prediction accuracy of the traffic prediction model and the iterative efficiency of model training.

In some optional implementation manners, step S205 includes the following steps:

Perform parameter initialization on the target cuckoo search algorithm.

In this embodiment, the above parameters for parameter initialization of the target cuckoo search algorithm may include at least the number of nests of the target cuckoo search algorithm, initial step update factor, maximum number of iterations, error standard and other parameters. Wherein, there is no limitation on the value of the parameter, which can be set according to actual business requirements.

A plurality of nest positions are randomly generated, and the initial weight threshold of the initial prediction model is encoded as the initial nest position of the target cuckoo search algorithm.

Determine the fitness function of the target cuckoo search algorithm, and calculate the fitness of each nest location based on the fitness function.

In this embodiment, the output error of the initial prediction model can be obtained, and the output error can be used as the fitness function of the above-mentioned target cuckoo search algorithm.

Using the target cuckoo search algorithm, a global iterative optimization process is performed according to the fitness, and a corresponding global optimal position is found from all the bird's nest positions.

In this embodiment, by using the target cuckoo search algorithm, the current optimal position is obtained according to the above fitness, and then the position is updated according to the improved position update method to obtain a new bird's nest position. Then compare the fitness values of these bird's nest locations, discard the poorer locations, and obtain the current optimal location. Repeat the above iterations to find the global optimal position.

Determine whether the current number of iterations meets the preset maximum number of iterations.

In this embodiment, there is no limitation on the value of the above-mentioned maximum number of iterations, which can be set according to actual business requirements.

If yes, the global optimal position is used as the optimal weight threshold of the initial prediction model to obtain the optimized specified prediction model.

In this embodiment, if the current number of iterations does not meet the preset maximum number of iterations, or does not meet the error standard, the above-mentioned global iterative optimization process is cyclically executed, and the global optimum that meets the maximum number of iterations or the error standard is always found. Location.

The present application initializes the parameters of the target cuckoo search algorithm; then randomly generates a plurality of nest positions, and encodes the initial weight threshold of the initial prediction model as the initial nest position of the target cuckoo search algorithm; then determines The fitness function of the target cuckoo search algorithm, and calculate the fitness of each nest position based on the fitness function; subsequently use the target cuckoo search algorithm to perform global iterative optimization processing according to the fitness , find the corresponding global optimal position from all the bird’s nest positions; finally judge whether the current iteration number meets the preset maximum number of iterations; if so, use the global optimal position as the optimal position of the initial prediction model The weight threshold is used to obtain the optimized specified prediction model. This application optimizes the weight threshold of the initial prediction model based on the target cuckoo search algorithm to obtain the optimized specified prediction model. Since the target cuckoo search algorithm realizes the adaptive adjustment of the step size of the traditional cuckoo search algorithm , which balances the global search and local optimization capabilities of the traditional cuckoo search algorithm, and effectively improves the convergence speed, thereby effectively improving the iterative efficiency of traffic forecasting model training. Further, the accuracy of the network traffic prediction processing of the traffic prediction model is improved.

In some optional implementation manners of this embodiment, step S207 includes the following steps:

Get pending network traffic data.

In this embodiment, the above-mentioned network traffic data to be processed may be network traffic data of the service system within a specified time period.

The network traffic data is input into the traffic prediction model.

Prediction processing is performed on the network traffic data by the traffic prediction model, and a prediction result corresponding to the network traffic data is output.

In this embodiment, after the traffic prediction model predicts the network traffic data, it will output the network traffic forecast value of the service system in the target time period; wherein, the target time period is after the specified time period period.

The present application obtains the network traffic data to be processed; then inputs the network traffic data into the traffic prediction model; subsequently uses the traffic prediction model to perform prediction processing on the network traffic data, and outputs the same as the network traffic The prediction results corresponding to the data. In this application, by using the traffic prediction model trained by the target cuckoo search algorithm to perform traffic prediction processing on the network traffic data to be processed, the accuracy of the network traffic prediction of the traffic prediction model can be effectively guaranteed.

In some optional implementation manners of this embodiment, after step S206, the above-mentioned electronic device may also perform the following steps:

Get the preset model ID.

In this embodiment, the block chain is divided into multiple storage sub-areas corresponding to each other in advance according to multiple identifiers. IDs can include model IDs, data table IDs, file IDs, and so on. Each storage sub-area is used to store data corresponding to the identification.

A target storage sub-area matching the model identifier is determined from multiple storage sub-areas included in the block chain.

The traffic forecasting model is stored in the target storage sub-area.

This application obtains the preset model identification; then determines the target storage sub-area matching the model identification from multiple storage sub-areas contained in the block chain; subsequently stores the traffic prediction model in the target storage sub-area Inside. This application uses the training set to train the specified prediction model, and uses the test set to test the trained specified prediction model, and after obtaining the trained traffic prediction model, it will also intelligently according to the The model identification corresponding to the traffic prediction model transfers the traffic prediction model to the target storage sub-area that matches the model identification in the blockchain, which effectively improves the standardization and intelligence of the model storage, and is conducive to subsequent access from the target storage sub-area. The required model can be quickly called out from the region, thereby improving the efficiency of model calling.

It should be understood that the sequence numbers of the steps in the above embodiments do not mean the order of execution, and the execution order of each process should be determined by its functions and internal logic, and should not constitute any limitation to the implementation process of the embodiment of the present invention.

It should be emphasized that, in order to further ensure the privacy and security of the above-mentioned flow prediction model, the above-mentioned flow prediction model can also be stored in a block chain node.

The blockchain referred to in this application is a new application mode of computer technologies such as distributed data storage, point-to-point transmission, consensus mechanism, and encryption algorithm. Blockchain (Blockchain), essentially a decentralized database, is a series of data blocks associated with each other using cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify its Validity of information (anti-counterfeiting) and generation of the next block. The blockchain can include the underlying platform of the blockchain, the platform product service layer, and the application service layer.

The embodiments of the present application may acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. .

Artificial intelligence basic technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation/interaction systems, and mechatronics. Artificial intelligence software technology mainly includes computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning/deep learning.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. , when the program is executed, it may include the procedures of the embodiments of the above-mentioned methods. Wherein, the aforementioned storage medium may be a nonvolatile storage medium such as a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM), or a random access memory (Random Access Memory, RAM).

It should be understood that although the various steps in the flow chart of the accompanying drawings are displayed sequentially according to the arrows, these steps are not necessarily executed sequentially in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the flowcharts of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages may not necessarily be executed at the same time, but may be executed at different times, and the order of execution is also It is not necessarily performed sequentially, but may be performed alternately or alternately with at least a part of other steps or sub-steps or stages of other steps.

Further referring to FIG. 3 , as an implementation of the method shown in FIG. 2 above, the present application provides an embodiment of an artificial intelligence-based flow prediction device, which corresponds to the method embodiment shown in FIG. 2 , The device can be specifically applied to various electronic devices.

As shown in Figure 3, the artificial intelligence-based flow prediction device 300 described in this embodiment includes: a first acquisition module 301, a first construction module 302, a processing module 303, a second construction module 304, an optimization module 305, a training module 306 and prediction module 307. in:

The first acquisition module 301 is configured to acquire network traffic within a preset historical time period;

The first construction module 302 is configured to construct a sample of the network traffic based on a preset time period division unit, and obtain a corresponding training set and a test set;

A processing module 303, configured to perform data preprocessing on the training set and the test set to obtain corresponding target training sets and target test sets;

The second building block 304 is used to determine the BP neural network structure, and construct an initial prediction model based on the BP neural network structure;

The optimization module 305 is used to optimize the weight threshold of the initial prediction model based on the target cuckoo search algorithm to obtain an optimized specified prediction model; wherein, the target cuckoo search algorithm is a step of the original cuckoo search algorithm The optimized cuckoo search algorithm obtained by improving the long update method;

A training module 306, configured to use the training set to train the designated prediction model, and use the test set to test the trained designated prediction model to obtain a trained traffic prediction model;

The prediction module 307 is configured to perform traffic prediction processing on the network traffic data to be processed based on the traffic prediction model.

In this embodiment, the operations performed by the above-mentioned modules or units correspond one-to-one to the steps of the artificial intelligence-based traffic forecasting method in the foregoing embodiments, and details are not repeated here.

In some optional implementations of this embodiment, the first building module 302 includes:

The first acquisition sub-module is used to acquire the preset time segment division unit;

A division submodule, configured to divide each day in the historical time period into corresponding multiple unit time periods based on the time period division unit;

Constructing a sub-module for constructing a training set corresponding to the network traffic by using the network traffic in the same unit time period as the same type of samples;

The second acquisition sub-module is used to acquire a preset time value;

A screening submodule, configured to randomly filter out specified data corresponding to the time value from the training set based on the time value;

The first determining submodule is configured to use the specified data as the test set.

In this embodiment, the operations performed by the above-mentioned modules or units correspond one-to-one to the steps of the artificial intelligence-based traffic forecasting method in the foregoing embodiments, and details are not repeated here.

In some optional implementations of this embodiment, the processing module 303 includes:

The cleaning submodule is used to perform data cleaning on the training set and the test set to obtain the corresponding first training set and first test set;

A first processing submodule, configured to perform normalization processing on the first training set and the first test set to obtain corresponding second training sets and second test sets;

A second determining submodule, configured to use the second training set as the target training set, and use the second test set as the target test set.

In this embodiment, the operations performed by the above-mentioned modules or units correspond one-to-one to the steps of the artificial intelligence-based traffic forecasting method in the foregoing embodiments, and details are not repeated here.

In some optional implementation manners of this embodiment, the traffic prediction device based on artificial intelligence further includes:

The third acquisition sub-module is used to acquire the original cuckoo search algorithm;

The fourth acquisition sub-module is used to acquire the step size update method in the initial cuckoo search algorithm;

The improvement sub-module is used to improve the step size update method of the cuckoo search algorithm based on a preset formula to obtain the optimized target cuckoo search algorithm.

In this embodiment, the operations performed by the above-mentioned modules or units correspond one-to-one to the steps of the artificial intelligence-based traffic forecasting method in the foregoing embodiments, and details are not repeated here.

In some optional implementations of this embodiment, the optimization module 305 includes:

The second processing submodule is used to initialize parameters of the target cuckoo search algorithm;

The third processing submodule is used to randomly generate a plurality of nest positions, and encode the initial weight threshold of the initial prediction model as the initial nest position of the target cuckoo search algorithm;

Calculation sub-module, used to determine the fitness function of the target cuckoo search algorithm, and calculate the fitness of each nest position based on the fitness function;

The fourth processing submodule is used to use the target cuckoo search algorithm to perform global iterative optimization processing according to the fitness, and find the corresponding global optimal position from all the bird's nest positions;

A judging sub-module is used to judge whether the current number of iterations satisfies the preset maximum number of iterations;

The third determination sub-module is configured to use the global optimal position as the optimal weight threshold of the initial prediction model to obtain the optimized specified prediction model if yes.

In this embodiment, the operations performed by the above-mentioned modules or units correspond one-to-one to the steps of the artificial intelligence-based traffic forecasting method in the foregoing embodiments, and details are not repeated here.

In some optional implementations of this embodiment, the prediction module 307 includes:

The fifth acquisition sub-module is used to acquire network traffic data to be processed;

an input submodule, configured to input the network traffic data into the traffic prediction model;

The prediction sub-module is configured to perform prediction processing on the network traffic data through the traffic prediction model, and output a prediction result corresponding to the network traffic data.

In this embodiment, the operations performed by the above-mentioned modules or units correspond one-to-one to the steps of the artificial intelligence-based traffic forecasting method in the foregoing embodiments, and details are not repeated here.

In some optional implementation manners of this embodiment, the traffic prediction device based on artificial intelligence further includes:

The second obtaining module is used to obtain a preset model identification;

A determining module, configured to determine a target storage sub-area matching the model identifier from multiple storage sub-areas included in the blockchain;

A storage module, configured to store the traffic forecasting model in the target storage sub-area.

In this embodiment, the operations performed by the above-mentioned modules or units correspond one-to-one to the steps of the artificial intelligence-based traffic forecasting method in the foregoing embodiments, and details are not repeated here.

In order to solve the above technical problems, the embodiment of the present application further provides computer equipment. Please refer to FIG. 4 for details. FIG. 4 is a block diagram of the basic structure of the computer device in this embodiment.

The computer device 4 includes a memory 41 , a processor 42 and a network interface 43 connected to each other through a system bus. It should be noted that only the computer device 4 with components 41-43 is shown in the figure, but it should be understood that it is not required to implement all the components shown, and more or fewer components may be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculation and/or information processing according to preset or stored instructions, and its hardware includes but is not limited to microprocessors, dedicated Integrated circuit (Application Specific Integrated Circuit, ASIC), programmable gate array (Field-Programmable GateArray, FPGA), digital processor (Digital Signal Processor, DSP), embedded devices, etc.

The computer equipment may be computing equipment such as a desktop computer, a notebook, a palmtop computer, and a cloud server. The computer device can perform human-computer interaction with the user through keyboard, mouse, remote controller, touch panel or voice control device.

The memory 41 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a hard disk, a multimedia card, a card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static Random Access Memory (SRAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), Programmable Read Only Memory (PROM), Magnetic Memory, Magnetic Disk, Optical Disk, etc. In some embodiments, the memory 41 may be an internal storage unit of the computer device 4 , such as a hard disk or memory of the computer device 4 . In other embodiments, the memory 41 can also be an external storage device of the computer device 4, such as a plug-in hard disk equipped on the computer device 4, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, flash memory card (FlashCard), etc. Certainly, the memory 41 may also include both an internal storage unit of the computer device 4 and an external storage device thereof. In this embodiment, the memory 41 is generally used to store the operating system and various application software installed in the computer device 4 , such as computer-readable instructions of a flow prediction method based on artificial intelligence. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or will be output.

The processor 42 may be a central processing unit (Central Processing Unit, CPU), controller, microcontroller, microprocessor, or other data processing chips in some embodiments. This processor 42 is generally used to control the general operation of said computer device 4 . In this embodiment, the processor 42 is configured to execute computer-readable instructions stored in the memory 41 or process data, for example, execute computer-readable instructions of the artificial intelligence-based flow prediction method.

The network interface 43 may include a wireless network interface or a wired network interface, and the network interface 43 is generally used to establish a communication connection between the computer device 4 and other electronic devices.

Compared with the prior art, the embodiments of the present application mainly have the following beneficial effects:

In the embodiment of the present application, the network traffic in the preset historical time period is first obtained; then, based on the preset time period division unit, the network traffic is sampled to obtain the corresponding training set and test set; and the Perform data preprocessing on the training set and the test set to obtain the corresponding target training set and target test set; then determine the BP neural network structure, and build an initial prediction model based on the BP neural network structure; follow-up based on the target cuckoo search The algorithm optimizes the weight threshold of the initial prediction model to obtain an optimized designated prediction model; further uses the training set to train the designated prediction model, and uses the test set to train the designated prediction model A test is performed to obtain a trained traffic prediction model; finally, traffic prediction processing is performed on the network traffic data to be processed based on the traffic prediction model. In the embodiment of the present application, after constructing the corresponding training set and test set based on the obtained network traffic in the historical time period, the target cuckoo search algorithm obtained by improving the step size update method of the original cuckoo search algorithm is used to The weight threshold of the initial prediction model is optimized to construct a specified prediction model, and then the training set and test set are used to train and test the specified prediction model to obtain a trained traffic prediction model, which can effectively improve the network traffic prediction performance of the traffic prediction model Accuracy and iterative efficiency of model training.

The present application also provides another implementation manner, which is to provide a computer-readable storage medium, the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to The at least one processor is made to execute the steps of the aforementioned artificial intelligence-based traffic forecasting method.

Compared with the prior art, the embodiments of the present application mainly have the following beneficial effects:

In the embodiment of the present application, the network traffic in the preset historical time period is first obtained; then, based on the preset time period division unit, the network traffic is sampled to obtain the corresponding training set and test set; and the Perform data preprocessing on the training set and the test set to obtain the corresponding target training set and target test set; then determine the BP neural network structure, and build an initial prediction model based on the BP neural network structure; follow-up based on the target cuckoo search The algorithm optimizes the weight threshold of the initial prediction model to obtain an optimized designated prediction model; further uses the training set to train the designated prediction model, and uses the test set to train the designated prediction model A test is performed to obtain a trained traffic prediction model; finally, traffic prediction processing is performed on the network traffic data to be processed based on the traffic prediction model. In the embodiment of the present application, after constructing the corresponding training set and test set based on the obtained network traffic in the historical time period, the target cuckoo search algorithm obtained by improving the step size update method of the original cuckoo search algorithm is used to The weight threshold of the initial prediction model is optimized to construct a specified prediction model, and then the training set and test set are used to train and test the specified prediction model to obtain a trained traffic prediction model, which can effectively improve the network traffic prediction performance of the traffic prediction model Accuracy and iterative efficiency of model training.

Through the description of the above embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus a necessary general-purpose hardware platform, and of course also by hardware, but in many cases the former is better implementation. Based on such an understanding, the technical solution of the present application can be embodied in the form of a software product in essence or the part that contributes to the prior art, and the computer software product is stored in a storage medium (such as ROM/RAM, disk, CD) contains several instructions to make a terminal device (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) execute the methods described in the various embodiments of the present application.

Apparently, the embodiments described above are only some of the embodiments of the present application, but not all of them. The drawings show preferred embodiments of the present application, but do not limit the patent scope of the present application. The present application can be implemented in many different forms, on the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure content of the present application more thorough and comprehensive. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or perform equivalent replacements for some of the technical features . All equivalent structures made using the contents of the description and drawings of this application, directly or indirectly used in other related technical fields, are also within the scope of protection of this application.

Claims (10)

1. The artificial intelligence-based flow prediction method is characterized by comprising the following steps of:

acquiring network traffic in a preset historical time period;

sample construction is carried out on the network traffic based on a preset time period dividing unit, and a corresponding training set and a corresponding testing set are obtained;

performing data preprocessing on the training set and the testing set to obtain a corresponding target training set and a corresponding target testing set;

determining a BP neural network structure, and constructing an initial prediction model based on the BP neural network structure;

performing weight threshold optimization on the initial prediction model based on a target cuckoo search algorithm to obtain an optimized specified prediction model; the target cuckoo searching algorithm is an optimized cuckoo searching algorithm obtained by improving the step length updating mode of the original cuckoo searching algorithm;

training the appointed prediction model by using the training set, and testing the trained appointed prediction model by using the testing set to obtain a trained flow prediction model;

and carrying out flow prediction processing on the network flow data to be processed based on the flow prediction model.

2. The artificial intelligence based traffic prediction method according to claim 1, wherein the step of constructing samples of the network traffic based on a preset time period division unit to obtain a corresponding training set and test set specifically includes:

Acquiring a preset time period dividing unit;

dividing each day in the history time period into a corresponding plurality of unit time periods based on the time period dividing unit;

constructing a training set corresponding to the network traffic by taking the network traffic in the same unit time period as the same sample;

acquiring a preset time value;

randomly screening appointed data corresponding to the time value from the training set based on the time value;

and taking the specified data as the test set.

3. The artificial intelligence based flow prediction method according to claim 1, wherein the step of performing data preprocessing on the training set and the test set to obtain a corresponding target training set and a target test set specifically includes:

data cleaning is carried out on the training set and the testing set, and a first training set and a first testing set which correspond to each other are obtained;

normalizing the first training set and the first testing set to obtain a corresponding second training set and second testing set;

the second training set is taken as the target training set, and the second testing set is taken as the target testing set.

4. The artificial intelligence based flow prediction method according to claim 1, further comprising, before the step of optimizing the weight threshold for the initial prediction model based on the target cuckoo search algorithm to obtain an optimized specified prediction model:

acquiring an original cuckoo searching algorithm;

acquiring a step length updating mode in the initial cuckoo searching algorithm;

and improving a step length updating mode of the cuckoo searching algorithm based on a preset formula to obtain the optimized target cuckoo searching algorithm.

5. The artificial intelligence based flow prediction method according to claim 1, wherein the step of optimizing the weight threshold of the initial prediction model based on the target cuckoo search algorithm to obtain an optimized specified prediction model specifically comprises:

initializing parameters of the target cuckoo search algorithm;

randomly generating a plurality of bird nest positions, and encoding an initial weight threshold of the initial prediction model into initial bird nest positions of the target cuckoo search algorithm;

determining a fitness function of the target cuckoo search algorithm, and calculating fitness of each nest position based on the fitness function;

Using the target cuckoo searching algorithm, performing global iterative optimization according to the fitness, and searching out a corresponding global optimal position from all the bird nest positions;

judging whether the current iteration number meets the preset maximum iteration number or not;

if yes, the global optimal position is used as an optimal weight threshold of the initial prediction model, and the optimized appointed prediction model is obtained.

6. The artificial intelligence based traffic prediction method according to claim 1, wherein the step of performing traffic prediction processing on the network traffic data to be processed based on the traffic prediction model specifically comprises:

acquiring network traffic data to be processed;

inputting the network traffic data into the traffic prediction model;

and carrying out prediction processing on the network traffic data through the traffic prediction model, and outputting a prediction result corresponding to the network traffic data.

7. The artificial intelligence based traffic prediction method according to claim 1, further comprising, after the step of training the specified prediction model using the training set and testing the trained specified prediction model by the test set to obtain a trained traffic prediction model:

Acquiring a preset model identifier;

determining a target memory subarea matched with the model identification from a plurality of memory subareas contained in the blockchain;

and storing the flow prediction model into a target storage subarea.

8. An artificial intelligence based flow prediction device, comprising:

the first acquisition module is used for acquiring network traffic in a preset historical time period;

the first construction module is used for carrying out sample construction on the network flow based on a preset time period dividing unit to obtain a corresponding training set and a corresponding testing set;

the processing module is used for carrying out data preprocessing on the training set and the testing set to obtain a corresponding target training set and a corresponding target testing set;

the second construction module is used for determining a BP neural network structure and constructing an initial prediction model based on the BP neural network structure;

the optimization module is used for optimizing the weight threshold value of the initial prediction model based on a target cuckoo search algorithm to obtain an optimized specified prediction model; the target cuckoo searching algorithm is an optimized cuckoo searching algorithm obtained by improving the step length updating mode of the original cuckoo searching algorithm;

The training module is used for training the appointed prediction model by using the training set, and testing the trained appointed prediction model by using the testing set to obtain a trained flow prediction model;

and the prediction module is used for carrying out flow prediction processing on the network flow data to be processed based on the flow prediction model.

9. A computer device comprising a memory having stored therein computer readable instructions which when executed implement the steps of the artificial intelligence based flow prediction method of any of claims 1 to 7.

10. A computer readable storage medium having stored thereon computer readable instructions which when executed by a processor implement the steps of the artificial intelligence based flow prediction method of any of claims 1 to 7.