[go: up one dir, main page]

WO2024154320A1 - Dispositif d'apprentissage, dispositif de commande basse température et système basse température - Google Patents

Dispositif d'apprentissage, dispositif de commande basse température et système basse température Download PDF

Info

Publication number
WO2024154320A1
WO2024154320A1 PCT/JP2023/001642 JP2023001642W WO2024154320A1 WO 2024154320 A1 WO2024154320 A1 WO 2024154320A1 JP 2023001642 W JP2023001642 W JP 2023001642W WO 2024154320 A1 WO2024154320 A1 WO 2024154320A1
Authority
WO
WIPO (PCT)
Prior art keywords
inventory
low
unit
temperature
history information
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/JP2023/001642
Other languages
English (en)
Japanese (ja)
Inventor
悠二 永田
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Mitsubishi Electric Corp
Original Assignee
Mitsubishi Electric Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Mitsubishi Electric Corp filed Critical Mitsubishi Electric Corp
Priority to PCT/JP2023/001642 priority Critical patent/WO2024154320A1/fr
Publication of WO2024154320A1 publication Critical patent/WO2024154320A1/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F25—REFRIGERATION OR COOLING; COMBINED HEATING AND REFRIGERATION SYSTEMS; HEAT PUMP SYSTEMS; MANUFACTURE OR STORAGE OF ICE; LIQUEFACTION SOLIDIFICATION OF GASES
    • F25D—REFRIGERATORS; COLD ROOMS; ICE-BOXES; COOLING OR FREEZING APPARATUS NOT OTHERWISE PROVIDED FOR
    • F25D23/00—General constructional features

Definitions

  • the present disclosure relates to a learning device, a low-temperature management device, and a low-temperature system that generate a machine learning model that outputs operating settings for low-temperature equipment.
  • Patent Document 1 discloses a device for managing portable cold and heat source equipment that performs machine learning on a power consumption model that estimates power consumption per unit time, estimates the amount of power consumption required to operate the low-temperature equipment, and selects a power storage device that is as light as possible.
  • the power consumption of low-temperature equipment is also affected by the items stored in the storage facility.
  • the lightest possible power storage device is selected, so the selected power storage device may not be appropriate for the stored items.
  • the storage facility may not be maintained at a low temperature unless it is managed by an administrator.
  • the present disclosure has been made to solve the problems described above, and aims to provide a learning device that generates a machine learning model that outputs operation settings that maintain a storage facility at a low temperature without the need for management by an administrator.
  • the learning device includes an inventory history acquisition unit that acquires inventory history information indicating the past inventory state of a storage facility that stores stored items as learning input data, an operation setting history acquisition unit that acquires operation setting history information indicating the past operation settings of low-temperature equipment that keeps the storage facility at a low temperature, and a learning unit that generates a first machine learning model for estimating operation settings based on the learning input data and the operation setting history information, using inventory schedule information indicating the planned inventory state of the storage facility as input data.
  • the learning device disclosed herein learns based on inventory history information and operational setting history information, so it can generate a machine learning model that outputs operational settings that maintain a storage facility at a low temperature, even without management by an administrator.
  • FIG. 1 is a schematic configuration diagram showing a low-temperature system according to a first embodiment
  • FIG. 11 is a diagram showing inventory history information according to the first embodiment
  • 2 is a hardware configuration diagram showing a low-temperature management server according to the first embodiment.
  • FIG. 4 is a diagram showing a learning method according to the first embodiment
  • FIG. 4 is a diagram showing a method for controlling a low-temperature device according to the first embodiment.
  • FIG. 11 is a schematic configuration diagram showing a low-temperature system according to a second embodiment.
  • FIG. 11 is a diagram showing a learning method according to the second embodiment;
  • FIG. 11 is a diagram showing a method for controlling a low-temperature device according to the second embodiment.
  • FIG. 11 is a schematic configuration diagram showing a low-temperature system according to a third embodiment.
  • FIG. 13 is a diagram showing a learning method according to the third embodiment;
  • FIG. 11 is a diagram showing a method for controlling a low-temperature device according to the third embodiment.
  • Embodiment 1. 1 is a schematic diagram showing a low-temperature system 1 according to the first embodiment.
  • the low-temperature system 1 manages low-temperature equipment 5 and includes a low-temperature management server 2 and an inventory management system 3.
  • the low-temperature management server 2 creates a control schedule for controlling the low-temperature equipment 5 from information acquired from the inventory management system 3 and the low-temperature equipment 5, and controls the low-temperature equipment 5 based on this control schedule.
  • the low-temperature management server 2 is connected to the inventory management system 3 and the low-temperature equipment 5 via a network 4 such as the Internet.
  • the inventory management system 3 is a system for managing the inventory of the storage facility.
  • the inventory management system 3 records information related to past inventory of the storage facility and planned inventory of the storage facility.
  • the inventory management system 3 provides the information recorded in the low-temperature management server 2 based on a request from the low-temperature management server 2.
  • the storage facility is, for example, a refrigerator, a freezer, or a showcase in which stored items such as food or medicine are stored. Note that, if inventory can be managed by the inventory management system 3, a food processing plant or the like may also be considered as a storage facility.
  • the low-temperature equipment 5 is equipment for keeping the storage facility at a low temperature, and is, for example, a unit cooler or a condensing unit. Although multiple low-temperature equipment 5 are shown in FIG. 1, the number of low-temperature equipment 5 is not particularly limited.
  • the low-temperature management server 2 has an inventory history acquisition unit 11, an operation setting history acquisition unit 12, a learning unit 13, an inventory schedule acquisition unit 14, an estimation unit 15, a schedule generation unit 16, and a control unit 17.
  • the low-temperature management server 2 also has an inventory memory unit 21, an operation memory unit 22, and a schedule memory unit 23.
  • the inventory history acquisition unit 11 acquires inventory history information from the inventory management system 3 as learning input data for machine learning in the learning unit 13.
  • the inventory history information is a compilation period of past inventory status of the storage facility. The compilation period is, for example, the most recent one month.
  • FIG. 2 is a diagram showing inventory history information according to the first embodiment. As shown in FIG. 2, the inventory status includes the items, arrangement, quantity for each item, or total weight, best-before date, or expiration date for each item for each day, and is defined by one or more combinations of these parameters for each day. Note that the unit time for compiling and managing the inventory status may be shorter, such as 30 minutes, rather than by day as shown in FIG. 2. Also, the items are, for example, classifications such as meat, fish, vegetables, confectionery, or medicine.
  • the inventory history acquisition unit 11 stores the acquired inventory history information in the inventory storage unit 21.
  • the operation setting history acquisition unit 12 acquires operation setting history information from the low-temperature equipment 5 as teacher data for machine learning in the learning unit 13.
  • the operation setting history information is past operation settings of the low-temperature equipment 5 accumulated over a collection period.
  • the operation settings include the ON/OFF, set temperature, number of operating units, or power consumption of the low-temperature equipment 5 for each unit time, and are defined by one or more combinations of these parameters for each unit time.
  • the operation settings may include the operating frequency of a compressor provided in the low-temperature equipment 5.
  • the operation setting history acquisition unit 12 stores the acquired operation setting history information in the operation memory unit 22.
  • the manager of the low-temperature equipment 5 changes the operation settings according to the inventory status of the storage facility so that the storage facility can maintain a low temperature to maintain the quality of the stored items. Therefore, it can be said that the past operation settings are the result of reflecting the operation settings for maintaining the storage facility at a low temperature in response to the past inventory status of the storage facility.
  • the learning unit 13 learns the relationship between the inventory history information stored in the inventory memory unit 21 as learning input data and the operation setting history information stored in the operation memory unit 22 as teacher data, and makes an association. Specifically, the learning unit 13 generates or modifies a first machine learning model based on the inventory history information stored in the inventory memory unit 21 and the operation setting history information stored in the operation memory unit 22.
  • the first machine learning model is a model for estimating operation settings from inventory schedule information indicating the planned inventory of the storage facility, which will be described later.
  • the learning algorithm used by the learning unit 13 is not particularly limited.
  • a machine learning model is trained using a learning dataset that includes multiple combinations of past inventory states divided into unit times and corresponding past operation settings.
  • a collection of learning input data and corresponding teacher data becomes a learning dataset.
  • the weight parameters of the neural network are adjusted so that the operation settings obtained by the model using the past inventory states as input become the past operation settings that correspond to the past inventory states.
  • the learning unit 13 reflects the weight parameters in the machine learning model used by the estimation unit 15.
  • the inventory plan acquisition unit 14 acquires inventory plan information as input data from the inventory management system 3.
  • the inventory plan information indicates the planned inventory status of the storage facility, and is acquired for each unit time.
  • the inventory plan acquisition unit 14 may also acquire multiple pieces of inventory plan information within the prediction period within the range in which the planned inventory status has been input. For example, if the unit time is one day, the prediction period is one week, and if the unit time is 30 minutes, the prediction period is one day.
  • the estimation unit 15 uses the first machine learning model to input the inventory plan information acquired by the inventory plan acquisition unit 14 and outputs operation settings.
  • the past operation settings are the result of reflecting operation settings for maintaining the storage at a low temperature in response to the past inventory state of the storage. Therefore, it can be said that the operation settings output using the first machine learning model learned based on the inventory history information and the operation setting history information satisfy the conditions required for the storage to maintain the quality of the stored items.
  • the estimation unit 15 may perform processing based on multiple pieces of inventory plan information acquired by the inventory plan acquisition unit 14 and output operation settings for each unit time.
  • the schedule generation unit 16 generates a control schedule for controlling the low-temperature equipment 5 based on the multiple operation settings output by the estimation unit 15.
  • the control schedule is a compilation of operation settings for a specified period. In other words, if the unit time is, for example, 30 minutes and the prediction period is one day, the control schedule indicates the start and stop times of the low-temperature equipment 5 for one day, as well as the set temperature at start-up, in 30-minute units.
  • the schedule generation unit 16 stores the generated control schedule in the schedule storage unit 23.
  • the control unit 17 sends control commands to the low-temperature equipment 5 based on the control schedule generated by the schedule generation unit 16.
  • the low-temperature equipment 5 operates based on the received control commands.
  • the inventory storage unit 21 stores the inventory history information acquired by the inventory history acquisition unit 11.
  • the operation storage unit 22 stores the operation setting history information acquired by the operation setting history acquisition unit 12.
  • the schedule storage unit 23 stores the control schedule generated by the schedule generation unit 16.
  • FIG. 3 is a hardware configuration diagram showing the low-temperature management server 2 according to the first embodiment.
  • the low-temperature management server 2 is composed of a processor 111 such as a CPU and a GPU, and a memory 112.
  • the processor 111 and the memory 112 are connected to each other so as to be able to communicate with each other via a bus 113.
  • the functions of the low-temperature management server 2 are realized by software, firmware, or a combination of software and firmware.
  • the software and firmware are written as programs and stored in the memory 112.
  • the processor 111 realizes the functions of the low-temperature management server 2 by reading and executing the programs stored in the memory 112.
  • FIG. 4 is a diagram showing a learning method according to the first embodiment.
  • the learning of the operation settings of the low-temperature equipment 5 may be performed less frequently than the control of the low-temperature equipment 5 described in FIG. 5.
  • the inventory history information acquisition unit acquires inventory history information from the inventory management system 3 (step S1).
  • the operation setting history acquisition unit 12 acquires operation setting history information from the low-temperature equipment 5 (step S2).
  • the learning unit 13 learns the association between the inventory history information and the operation setting history information (step S3).
  • the learning unit 13 generates or modifies a first machine learning model (step S4).
  • FIG. 5 is a diagram showing a method for controlling the low-temperature equipment 5 according to the first embodiment.
  • the inventory schedule acquisition unit 14 acquires inventory schedule information from the inventory management system 3 (step S11).
  • the estimation unit 15 estimates operation settings by inputting the inventory schedule information into the first machine learning model acquired from the learning unit 13 (step S12).
  • the schedule generation unit 16 creates a control schedule based on the operation settings estimated by the learning unit 13 (step S13).
  • the control unit 17 controls the low-temperature equipment 5 based on the control schedule generated by the schedule generation unit 16 (step S14).
  • the low-temperature management server 2 of embodiment 1 learns based on inventory history information and operation setting history information, and can generate a machine learning model that outputs operation settings that maintain the storage facility at a low temperature even without management by an administrator. Therefore, the low-temperature management server 2 can control the low-temperature equipment 5 by simplifying the series of tasks, such as considering and setting the necessary operation settings in advance.
  • the operating settings generated by the machine learning model can contribute to energy savings.
  • the inventory history acquisition unit 11, the operation setting history acquisition unit 12, the learning unit 13, the inventory storage unit 21, and the operation storage unit 22 constitute a learning device 10 that generates a machine learning model.
  • the learning device 10 may be provided outside the low-temperature management server 2.
  • FIG. 6 is a schematic diagram showing a low-temperature system 1A according to embodiment 2.
  • the low-temperature system 1A according to embodiment 2 differs from embodiment 1 in that the low-temperature management server 2A has an internal environment history acquisition unit 31 and further generates a second machine learning model.
  • the description of the points in common with embodiment 1 will be omitted, and the description will focus on the points that are different from embodiment 1.
  • the storage facility is provided with a temperature sensor or humidity sensor connected to the low-temperature equipment 5 by wire or wirelessly.
  • the internal environment history acquisition unit 31 acquires internal environment history information indicating the past environmental state inside the storage facility from the low-temperature equipment 5 as training data for machine learning in the learning unit 13.
  • the environmental state includes the air temperature or humidity inside the storage facility acquired using a temperature sensor or humidity sensor.
  • the internal environment history acquisition unit 31 acquires internal environment history information for each unit time of the aggregation period, for example.
  • the internal environment history acquisition unit 31 stores the acquired internal environment history information in the internal environment memory unit 24.
  • the learning unit 13 generates or modifies the first machine learning model as described in the first embodiment.
  • the learning unit 13 also learns the relationship between the inventory history information stored in the inventory storage unit 21 and the internal environment history information stored in the internal environment storage unit 24, and associates them.
  • the learning unit 13 generates or modifies the second machine learning model based on the inventory history information stored in the inventory storage unit 21 and the internal environment history information stored in the internal environment storage unit 24.
  • the second machine learning model is a model for estimating the future internal environmental state of the storage facility from the inventory schedule information.
  • the method for generating or modifying the second machine learning model is the same as the method for generating or modifying the first machine learning model described in the first embodiment.
  • the internal environment history information indicates the track record of the internal environmental state of the storage facility, that is, the internal environmental state of the storage facility as a result of control based on the operation settings determined by the manager of the storage facility to maintain the storage facility at a low temperature. Therefore, the estimated future internal environmental state of the storage facility conforms to the actual state of the internal environmental state of the storage facility so far.
  • the estimation unit 15 outputs the operation setting from the input data using the first machine learning model.
  • the estimation unit 15 also outputs the future internal environmental state of the storage from the input data using the second machine learning model, and outputs the operation setting corresponding to the future internal environmental state of the storage. For example, if the output operation setting is a set temperature or a set humidity, the estimation unit 15 outputs a set temperature or a set humidity of a value equivalent to the internal air temperature or humidity of the storage indicated by the internal environmental state of the future storage output by the second machine learning model, as the operation setting corresponding to the internal environmental state of the future storage.
  • the estimation unit 15 calculates and outputs the operation setting that can realize the internal air temperature or humidity of the storage indicated by the internal environmental state of the future storage output by the second machine learning model.
  • the calculation method is not particularly limited, but for example, a calculation is performed taking into account the volume of the storage or the capacity of the low-temperature equipment 5, etc.
  • the estimation unit 15 may use the second machine learning model when the operation setting candidate output using the first machine learning model deviates from the operation settings output so far. This is because it is considered that the operation settings used as the teacher data are not appropriate when the operation setting candidate output using the first machine learning model deviates from the operation settings output so far.
  • the estimation unit 15 may compare the operation amount of each device when control is performed at the set temperature or set humidity indicated by the operation setting candidate output using the first machine learning model with the operation amount of each device when control is performed at the set temperature or set humidity output using the second machine learning model to select a machine learning model that contributes more to improving energy saving performance. Furthermore, the estimation unit 15 may use the operation setting candidates output using the first machine learning model and the operation setting candidates output using the second machine learning model depending on the time of year, or may average the set temperature or set humidity output by each of them. However, the user may select which of the machine learning model-based operation setting candidates is to be used as the operation setting actually used for control.
  • the internal environment storage unit 24 stores the internal environment history information acquired by the internal environment history acquisition unit 31.
  • FIG. 7 is a diagram showing a learning method according to the second embodiment.
  • the inventory history information acquisition unit acquires inventory history information from the inventory management system 3 (step S1).
  • the operation setting history acquisition unit 12 acquires operation setting history information from the low-temperature equipment 5 (step S2).
  • the internal environment history acquisition unit 31 acquires internal environment history information from the low-temperature equipment 5 (step S21).
  • the learning unit 13 learns the association between the inventory history information and the operation setting history information, and learns the association between the inventory history information and the internal environment history information (step S22).
  • the learning unit 13 generates or modifies a first machine learning model and a second machine learning model (step S23).
  • FIG. 8 is a diagram showing the control method of the low-temperature equipment 5 according to the second embodiment.
  • the inventory schedule acquisition unit 14 acquires inventory schedule information from the inventory management system 3 (step S11).
  • the estimation unit 15 inputs inventory schedule information to each of the first machine learning model and the second machine learning model acquired from the learning unit 13, thereby estimating operation setting candidates for each machine learning model (step S31).
  • the estimation unit 15 compares the operation setting candidates estimated using each machine learning model from the viewpoint of energy saving, determines whether the operation setting candidate estimated using the first machine learning model or the second machine learning model is to be the operation setting, and outputs the result (step S32).
  • the operation setting candidates are compared to determine the operation setting, but as described above, there is no particular limitation on how to determine the operation setting to be used for controlling the low-temperature equipment 5.
  • the schedule generation unit 16 creates a control schedule based on the operation setting output by the learning unit 13 (step S13).
  • the control unit 17 controls the low-temperature equipment 5 based on the control schedule generated by the schedule generation unit 16 (step S14).
  • the low-temperature management server 2A of the second embodiment estimates the operation settings using the second machine learning model in addition to the first machine learning model. This makes it possible to maintain the storage facility at a low temperature with even higher accuracy and improve energy-saving performance.
  • the inventory history acquisition unit 11, the operation setting history acquisition unit 12, the internal environment history acquisition unit 31, and the learning unit 13, as well as the inventory storage unit 21, the operation storage unit 22, and the internal environment storage unit 24, constitute a learning device 10A that generates a machine learning model.
  • the learning device 10A may be provided outside the low-temperature management server 2.
  • Fig. 9 is a schematic configuration diagram showing a low-temperature system 1B according to embodiment 3. As shown in Fig. 9, the low-temperature system 1B according to embodiment 3 differs from embodiment 1 in that a low-temperature management server 2B has a weather acquisition unit 32 and an external environment history acquisition unit 33 and is capable of communicating with a weather forecasting system 6. In embodiment 3, a description of the points in common with embodiment 1 will be omitted, and the description will focus on the points that are different from embodiment 1.
  • the low temperature management server 2B is connected to a weather forecasting system 6 via a network 4 such as the Internet.
  • the weather forecasting system 6 is a system that provides a service of distributing predicted weather information for at least the area in which the storage facility is located.
  • the weather information includes weather, temperature, humidity, etc.
  • the weather forecasting system 6 provides the low temperature management server 2B with weather information based on a request from the low temperature management server 2B.
  • the weather acquisition unit 32 acquires weather information for the area in which the storage facility is located from the weather forecasting system 6.
  • the weather acquisition unit 32 acquires weather information for each unit time for the forecast period, for example.
  • a temperature sensor or humidity sensor is provided outside the storage facility and is connected to the low-temperature equipment 5 by wire or wirelessly.
  • the external environment history acquisition unit 33 acquires external environment history information indicating past environmental conditions outside the storage facility from the low-temperature equipment 5 as learning input data for machine learning in the learning unit 13.
  • the environmental conditions include the temperature or humidity of the air outside the storage facility acquired using a temperature sensor or humidity sensor.
  • the external environment history acquisition unit 33 acquires external environment history information for each unit time of the aggregation period, for example.
  • the external environment history acquisition unit 33 stores the acquired external environment history information in the external environment memory unit 25.
  • the learning unit 13 learns and associates the inventory history information stored in the inventory storage unit 21 and the environmental history information stored in the external environment storage unit 25 with the operation setting history information stored in the operation storage unit 22. Specifically, the learning unit 13 generates or modifies a third machine learning model based on the inventory history information stored in the inventory storage unit 21 and the environmental history information stored in the external environment storage unit 25, and the operation setting history information stored in the operation storage unit 22.
  • the third machine learning model is a model for estimating operation settings from inventory schedule information and weather information.
  • the method of generating or modifying the third machine learning model is the same as the method of generating or modifying the first machine learning model described in the first embodiment.
  • the estimation unit 15 uses the third machine learning model to input inventory forecast information and weather information and output operation settings.
  • the external environment memory unit 25 stores the external environment history information acquired by the external environment history acquisition unit 33.
  • FIG. 10 is a diagram showing a learning method according to the third embodiment.
  • the inventory history information acquisition unit acquires inventory history information from the inventory management system 3 (step S1).
  • the operation setting history acquisition unit 12 acquires operation setting history information from the low-temperature equipment 5 (step S2).
  • the external environment history acquisition unit 33 acquires external environment history information from the low-temperature equipment 5 (step S41).
  • the learning unit 13 learns the association between the inventory history information and the external environment history information and the operation setting history information (step S42). Then, the learning unit 13 generates or modifies a third machine learning model (step S43).
  • FIG. 11 is a diagram showing the control method of the low-temperature equipment 5 according to the third embodiment.
  • the inventory schedule acquisition unit 14 acquires inventory schedule information from the inventory management system 3 (step S11).
  • the weather acquisition unit 32 acquires weather information from the weather forecast system 6 (step S51).
  • the estimation unit 15 estimates the operation settings by inputting the inventory schedule information and the weather information into the third machine learning model acquired from the learning unit 13 (step S52).
  • the schedule generation unit 16 creates a control schedule based on the operation settings estimated by the learning unit 13 (step S13).
  • the control unit 17 controls the low-temperature equipment 5 based on the control schedule generated by the schedule generation unit 16 (step S14).
  • the low-temperature management server 2B of embodiment 3 performs learning by adding external environmental history information to the learning parameters, and inputs weather information in conjunction with the weather forecast system 6 to estimate the operation settings.
  • the manager of the low-temperature equipment 5 changes the operation settings according to the environmental conditions so that the storage facility can maintain a low temperature to maintain the quality of the stored items. Therefore, it can be said that the past operation settings are the result of reflecting the operation settings for maintaining the storage facility at a low temperature in response to the past environmental conditions in the vicinity where the storage facility is located. Therefore, when the low-temperature equipment 5 is controlled based on the operation settings output by the third machine learning model, the storage facility can be maintained at a low temperature with even higher accuracy.
  • the inventory history acquisition unit 11, the operation setting history acquisition unit 12, the external environment history acquisition unit 33, and the learning unit 13, as well as the inventory storage unit 21, the operation storage unit 22, and the external environment storage unit 25, constitute a learning device 10B that generates a machine learning model.
  • the learning device 10B may be provided in the external environment of the low-temperature management server 2.
  • the low temperature management server 2 may be a terminal device such as a PC, rather than a server device.
  • the low temperature management server 2 or a device having functions equivalent to the low temperature management server 2 corresponds to the "low temperature management device" of the present disclosure.
  • multiple computer devices may be combined to have functions equivalent to those of the low temperature management server 2.
  • the algorithm used for learning may be of any type, such as a random forest model or SVM (Support Vector Machine), in addition to a neural network.
  • SVM Small Vector Machine
  • the first machine learning model may be replaced with the third machine learning model described in the third embodiment.
  • the second machine learning model may be configured to add external environment history information as learning input data and to add weather information as input data.

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • Physics & Mathematics (AREA)
  • Mechanical Engineering (AREA)
  • Thermal Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Air Conditioning Control Device (AREA)

Abstract

Un dispositif d'apprentissage comprend : une unité d'acquisition d'historique d'inventaire qui acquiert des informations d'historique d'inventaire indiquant un état d'inventaire passé d'une unité de stockage qui stocke des articles à stocker en tant que données d'entrée d'apprentissage ; une unité d'acquisition d'historique de réglage de fonctionnement qui acquiert des informations d'historique de réglage de fonctionnement indiquant des réglages de fonctionnement passés d'un équipement basse température qui maintient l'unité de stockage à basse température ; et une unité d'apprentissage qui génère un premier modèle d'apprentissage automatique pour estimer un réglage de fonctionnement sur la base des données d'entrée d'apprentissage et des informations d'historique de réglage de fonctionnement, à l'aide d'informations de calendrier d'inventaire indiquant un calendrier de l'état d'inventaire de l'unité de stockage en tant que données d'entrée.
PCT/JP2023/001642 2023-01-20 2023-01-20 Dispositif d'apprentissage, dispositif de commande basse température et système basse température Ceased WO2024154320A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/JP2023/001642 WO2024154320A1 (fr) 2023-01-20 2023-01-20 Dispositif d'apprentissage, dispositif de commande basse température et système basse température

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2023/001642 WO2024154320A1 (fr) 2023-01-20 2023-01-20 Dispositif d'apprentissage, dispositif de commande basse température et système basse température

Publications (1)

Publication Number Publication Date
WO2024154320A1 true WO2024154320A1 (fr) 2024-07-25

Family

ID=91955569

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2023/001642 Ceased WO2024154320A1 (fr) 2023-01-20 2023-01-20 Dispositif d'apprentissage, dispositif de commande basse température et système basse température

Country Status (1)

Country Link
WO (1) WO2024154320A1 (fr)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20210093615A (ko) * 2020-01-20 2021-07-28 엘지전자 주식회사 인공 지능 냉장고 및 그의 동작 방법
JP2021169910A (ja) * 2020-04-17 2021-10-28 パナソニックＩｐマネジメント株式会社 冷蔵庫、プログラム、端末装置、及び、冷蔵庫制御システム
US20220099362A1 (en) * 2019-03-19 2022-03-31 Daikin Industries, Ltd. Apparatus for calculating temperature setting, system for low temperature treatment, method of calculating temperature setting, and program for calculating temperature setting
JP2022064576A (ja) * 2020-10-14 2022-04-26 東芝ライフスタイル株式会社 情報処理システム、及び品質推定方法
JP2022092488A (ja) * 2020-12-10 2022-06-22 東芝ライフスタイル株式会社 発注システム
JP2022094595A (ja) * 2020-12-15 2022-06-27 東芝ライフスタイル株式会社 家電システム
JP2023003692A (ja) * 2021-06-24 2023-01-17 三菱重工サーマルシステムズ株式会社 運転プラン生成システム及びその方法並びにプログラム

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20220099362A1 (en) * 2019-03-19 2022-03-31 Daikin Industries, Ltd. Apparatus for calculating temperature setting, system for low temperature treatment, method of calculating temperature setting, and program for calculating temperature setting
KR20210093615A (ko) * 2020-01-20 2021-07-28 엘지전자 주식회사 인공 지능 냉장고 및 그의 동작 방법
JP2021169910A (ja) * 2020-04-17 2021-10-28 パナソニックＩｐマネジメント株式会社 冷蔵庫、プログラム、端末装置、及び、冷蔵庫制御システム
JP2022064576A (ja) * 2020-10-14 2022-04-26 東芝ライフスタイル株式会社 情報処理システム、及び品質推定方法
JP2022092488A (ja) * 2020-12-10 2022-06-22 東芝ライフスタイル株式会社 発注システム
JP2022094595A (ja) * 2020-12-15 2022-06-27 東芝ライフスタイル株式会社 家電システム
JP2023003692A (ja) * 2021-06-24 2023-01-17 三菱重工サーマルシステムズ株式会社 運転プラン生成システム及びその方法並びにプログラム

Similar Documents

Publication Publication Date Title
US11836599B2 (en) Optimizing data center controls using neural networks
CN113748386B (zh) 散热控制与模型训练方法、设备、系统及存储介质
US12013673B2 (en) Building control system using reinforcement learning
US9429921B2 (en) Method and system for energy control management
US20210294399A1 (en) Device Temperature Control
JP2017110904A (ja) 環境維持システムの制御
JP2007014179A (ja) デマンド制御装置および消費電力予測方法ならびにプログラム
JP7342833B2 (ja) 制御装置、コントローラ、制御システム、制御方法、および制御プログラム
CN108488969A (zh) 用于冷水机组的控制方法和控制装置
JP2019212131A (ja) 予測装置、電気機器、管理システム、予測方法、及び制御プログラム
CN110135655B (zh) 一种用于确定能源站的运行控制策略的方法与设备
US11371741B2 (en) Air conditioning apparatus and method for controlling using learned sleep modes
CN119103684A (zh) 温度智能控制方法及装置
EP4083723B1 (fr) Appareil d'évaluation, procédé d'évaluation, programme d'évaluation et appareil de commande
WO2024155507A2 (fr) Exploitation de la technologie domestique intelligente pour la gestion de l'efficacité énergétique et de la demande
Boubouh et al. Efficacy of temporal and spatial abstraction for training accurate machine learning models: A case study in smart thermostats
Khan et al. Advanced data analytics modeling for evidence-based data center energy management
WO2024154320A1 (fr) Dispositif d'apprentissage, dispositif de commande basse température et système basse température
CN114326987B (zh) 制冷系统控制及模型训练方法、装置、设备及存储介质
CN110737322A (zh) 信息处理方法及电子设备
US20230070842A1 (en) Systems and methods of semantic tagging
CN119844971A (zh) 一种冰箱制冷量调节方法
CN119806302A (zh) 边缘智能终端交互管理方法、装置、设备及存储介质
Ribault et al. Economic optimisation of cold production: a matheuristic with artificial neural network approach
CN116954329A (zh) 制冷系统的状态调节方法、装置、设备、介质及程序产品

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23917533

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE