Model training method and device and readable storage medium

By using reinforcement learning models in the refrigeration equipment in the data center, training the target reinforcement model based on multiple sets of verification parameters, optimizing the energy consumption of the refrigeration equipment, solving the problem of poor energy consumption optimization effect in the data center, realizing energy consumption reduction and regional Temperature control.

CN118940624AActive Publication Date: 2024-11-12CHINA UNITED NETWORK COMM GRP CO LTD +1

Patent Information

Application Number
CN202410986167.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-12
Estimated Expiration
2044-07-22

Smart Images

  • Figure CN118940624A_ABST
    Figure CN118940624A_ABST
Patent Text Reader

Abstract

The invention discloses a model training method and device and a readable storage medium, relates to the technical field of computers, and is used for improving the optimization effect of energy consumption of refrigeration equipment. The method comprises the following steps: inputting multiple groups of verification parameters of the refrigeration equipment into a scene simulation model to obtain scene simulation data corresponding to each group of verification parameters; the refrigeration equipment is used for cooling the plurality of network equipment; the multiple groups of verification parameters comprise multiple groups of refrigeration parameters under different boundary conditions; the scene simulation data comprises energy consumption of refrigeration equipment and temperature data of each network equipment area; training the reinforcement model based on the multiple groups of verification parameters to obtain a target reinforcement model; the target strengthening model is used for determining a target refrigeration parameter based on the current boundary condition of the refrigeration equipment; the target refrigeration parameter is the refrigeration parameter, with the minimum energy consumption, of the refrigeration equipment in the multiple sets of refrigeration parameters to be selected; when the refrigeration equipment executes the refrigeration parameters to be selected, the temperature data of each network equipment area is smaller than a temperature threshold value.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computer technology, and in particular, to a model training method, device, and readable storage medium. Background Art

[0002] At present, the energy consumption of data centers is increasing. Among the total energy consumption of data centers, the consumption of refrigeration equipment accounts for a high proportion. In order to reduce the energy consumption of refrigeration equipment, it is necessary to optimize the energy consumption of quality equipment according to the temperature of network equipment areas at different locations in the data center.

[0003] However, the number of measurement points in the data center is limited, and it is impossible to obtain the scenario parameters of each branch. The scenario parameters may include at least one of the following: water flow, temperature, air volume, etc. In this way, when the traditional optimization model uses limited branch parameters to optimize the energy consumption of the refrigeration equipment, it is easy to cause overheating in local areas of the data center, or the energy consumption of the refrigeration equipment is reduced too little, resulting in poor optimization effect of the energy consumption of the refrigeration equipment. Summary of the invention

[0004] The present application provides a model training method, device and readable storage medium for improving the optimization effect of energy consumption of refrigeration equipment.

[0005] In order to achieve the above objectives, this application adopts the following technical solutions:

[0006] In a first aspect, a model training method is provided, the method comprising: inputting multiple groups of verification parameters of a refrigeration device into a scene simulation model to obtain scene simulation data corresponding to each group of verification parameters; the refrigeration device is used to cool multiple network devices; the multiple groups of verification parameters include multiple groups of refrigeration parameters under different boundary conditions; the scene simulation data include energy consumption of the refrigeration device and temperature data of each network device area; training an enhanced model based on the multiple groups of verification parameters to obtain a target enhanced model; the target enhanced model is used to determine a target refrigeration parameter based on the current boundary conditions of the refrigeration device; the target refrigeration parameter is a refrigeration parameter with the lowest energy consumption for the refrigeration device among multiple groups of refrigeration parameters to be selected; the temperature data of each network device area is less than a temperature threshold when the refrigeration device executes the selected refrigeration parameter.

[0007] Optionally, training the reinforcement model based on multiple groups of verification parameters to obtain a target reinforcement model includes: determining a reward value for executing each group of verification parameters based on the scene simulation data and the reward function corresponding to each group of verification parameters;

[0008] The reinforcement model is trained based on multiple groups of verification parameters and the reward value of each group of verification parameters to obtain a target reinforcement model.

[0009] Optionally, multiple groups of verification parameters of the refrigeration equipment are input into the scene simulation model to obtain scene simulation data corresponding to each group of verification parameters, including: inputting multiple groups of verification parameters into the mechanism model to obtain air supply parameters corresponding to each group of verification parameters; the air supply parameters include air volume and air temperature; the air supply parameters corresponding to each group of verification parameters, outdoor temperature, outdoor humidity and power of multiple network devices are input into the artificial intelligence computational fluid dynamics AI-CFD model to obtain scene simulation data corresponding to each group of verification parameters.

[0010] Optionally, the method also includes: generating multiple sets of training data based on a target computational fluid dynamics (CFD) model; the training data includes air supply parameters, power of network equipment, and temperature data of each corresponding network equipment area; based on the training data, training a preset neural network model to obtain an AI-CFD model.

[0011] Optionally, the method also includes: obtaining historical temperature data, historical air supply parameters and three-dimensional structural data of the target area; the three-dimensional structural data includes the position of each device in the target area and the configuration of the flow field; the equipment includes network equipment and refrigeration equipment; and constructing a target CFD model based on the historical temperature data, historical air supply parameters and three-dimensional structural data.

[0012] Optionally, a target CFD model is constructed based on historical temperature data, historical air supply parameters and three-dimensional structural data, including: constructing an initial CFD model based on historical temperature data, historical air supply parameters and three-dimensional structural data; determining a loss value of the initial CFD model, the loss value being used to characterize the degree of loss between temperature data output by the initial CFD model and actual temperature data; when the loss value of the initial CFD model is less than a loss value threshold, determining the CFD model as a target CFD model.

[0013] Optionally, the boundary conditions include at least one of the following: outdoor air dry-bulb temperature, outdoor air wet-bulb temperature, chilled water supply temperature, chilled water flow rate, cooling load power; fan power, temperature threshold; the refrigeration parameters include at least one of the following: return air valve opening, fresh air valve opening, return air fan frequency, supply air fan frequency, chilled water valve opening, supply air temperature setting value.

[0014] Based on the technical solution provided by the present application, multiple groups of verification parameters of the refrigeration equipment are input into the scene simulation model to obtain scene simulation data corresponding to each group of verification parameters; the reinforcement model is trained based on the multiple groups of verification parameters to obtain the target reinforcement model. In this way, by constructing different simulation data in the scene simulation model, the number of parameters of the optimization model can be effectively increased, and the reinforcement model can be trained based on more simulation data to ensure the training effect of the reinforcement model. While effectively reducing the energy consumption of the refrigeration equipment, it can avoid overheating of the local network equipment area, thereby ensuring the effect of optimizing the energy consumption of the refrigeration equipment.

[0015] In a second aspect, a model training device is provided, which includes: an input unit and a processing unit; the input unit is used to input multiple groups of verification parameters of a refrigeration device into a scene simulation model to obtain scene simulation data corresponding to each group of verification parameters; the refrigeration device is used to cool multiple network devices; the multiple groups of verification parameters include multiple groups of refrigeration parameters under different boundary conditions; the scene simulation data includes energy consumption of the refrigeration device and temperature data of each network device area; the processing unit is used to train the enhancement model based on the multiple groups of verification parameters to obtain a target enhancement model; the target enhancement model is used to determine the target refrigeration parameter based on the current boundary conditions of the refrigeration device; the target refrigeration parameter is the refrigeration parameter with the lowest energy consumption for the refrigeration device among multiple groups of refrigeration parameters to be selected; the temperature data of each network device area is less than a temperature threshold when the refrigeration device executes the selected refrigeration parameter.

[0016] Optionally, the processing unit is specifically configured to: determine a reward value for executing each set of verification parameters based on the scene simulation data corresponding to each set of verification parameters and the reward function;

[0017] The reinforcement model is trained based on multiple groups of verification parameters and the reward value of each group of verification parameters to obtain a target reinforcement model.

[0018] Optionally, the processing unit is specifically used to: input multiple groups of verification parameters into the mechanism model to obtain air supply parameters corresponding to each group of verification parameters; the air supply parameters include air volume and air temperature; the air supply parameters corresponding to each group of verification parameters, outdoor temperature, outdoor humidity and power of multiple network devices are input into the artificial intelligence computational fluid dynamics AI-CFD model to obtain scene simulation data corresponding to each group of verification parameters.

[0019] Optionally, the processing unit is also used to generate multiple sets of training data based on the target computational fluid dynamics (CFD) model; the training data includes air supply parameters, the power of the network equipment, and the corresponding temperature data of each network equipment area; the processing unit is also used to train a preset neural network model based on the training data to obtain an AI-CFD model.

[0020] Optionally, the device includes an acquisition unit, which is used to acquire historical temperature data, historical air supply parameters and three-dimensional structural data of the target area; the three-dimensional structural data includes the position of each device in the target area and the configuration of the flow field; the equipment includes network equipment and refrigeration equipment; the processing unit is also used to construct a target CFD model based on the historical temperature data, historical air supply parameters and three-dimensional structural data.

[0021] Optionally, the processing unit is specifically used to: construct an initial CFD model based on historical temperature data, historical air supply parameters and three-dimensional structure data; determine a loss value of the initial CFD model, the loss value is used to characterize the degree of loss between the temperature data output by the initial CFD model and the actual temperature data; when the loss value of the initial CFD model is less than a loss value threshold, determine the CFD model as a target CFD model.

[0022] Optionally, the boundary conditions include at least one of the following: outdoor air dry-bulb temperature, outdoor air wet-bulb temperature, chilled water supply temperature, chilled water flow rate, cooling load power; fan power, temperature threshold; the refrigeration parameters include at least one of the following: return air valve opening, fresh air valve opening, return air fan frequency, supply air fan frequency, chilled water valve opening, supply air temperature setting value.

[0023] In a third aspect, a model training device is provided, which can implement the functions performed by the model training device in the above-mentioned aspects or possible designs. The functions can be implemented through hardware, such as: in one possible design, the model training device may include: a processor and a communication interface, and the processor can be used to support the model training device to implement the functions involved in the above-mentioned first aspect or any possible design of the first aspect.

[0024] In another possible design, the model training device may further include a memory for storing computer-executable instructions and data necessary for the model training device. When the model training device is running, the processor executes the computer-executable instructions stored in the memory so that the model training device performs the first aspect or any possible model training method of the first aspect.

[0025] In a fourth aspect, a computer-readable storage medium is provided, which may be a readable non-volatile storage medium, and which stores computer instructions or programs. When the computer-readable storage medium is run on a computer, the computer can execute the first aspect or any possible model training method of the above aspects.

[0026] In a fifth aspect, a computer program product comprising instructions is provided, which, when executed on a computer, enables the computer to execute the model training method of the first aspect or any possible design of the above aspects.

[0027] In a sixth aspect, an electronic device is provided, the electronic device comprising one or more processors and one or more memories. The one or more memories are coupled to the one or more processors, the one or more memories are used to store computer program codes, the computer program codes include computer instructions, and when the one or more processors execute the computer instructions, the electronic device executes the model training method as described in the first aspect or any possible design of the first aspect.

[0028] In the seventh aspect, a chip system is provided, the chip system includes a processor and a communication interface, and the chip system can be used to implement the functions performed by the model training device in the above-mentioned first aspect or any possible design of the first aspect. In one possible design, the chip system also includes a memory, the memory is used to store program instructions and / or data. The chip system can be composed of a chip, or it can include a chip and other discrete devices, without limitation. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 A schematic diagram of the structure of a model training system provided in an embodiment of the present application;

[0030] Figure 2 A schematic diagram of the structure of another model training system provided in an embodiment of the present application;

[0031] Figure 3 A schematic diagram of the structure of a model training device provided in an embodiment of the present application;

[0032] Figure 4 A flowchart of a model training method provided in an embodiment of the present application;

[0033] Figure 5 A flowchart of another model training method provided in an embodiment of the present application;

[0034] Figure 6 A flowchart of another model training method provided in an embodiment of the present application;

[0035] Figure 7 A schematic diagram of the structure of another model training device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0036] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0037] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the embodiments of the present application as detailed in the attached claims.

[0038] It should also be understood that the term “comprising” indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements and / or components.

[0039] The terms involved in this application are explained below.

[0040] Reinforcement learning: It includes four elements: agent, environment, state, action and reward. The input of the agent is state and the output is action.

[0041] The training process of reinforcement learning is: through multiple interactions between the agent and the environment, the action, state, and reward of each interaction are obtained; these multiple sets of (action, state, reward) are used as training data to train the agent once. The above process is used to train the agent for the next round until the convergence condition is met.

[0042] Agent: refers to an entity that can think and interact with the environment. For example, an agent can be a computer system or part of a computer system in a specific environment. An agent can achieve the set goals autonomously in its environment based on its own perception of the environment, following existing instructions or through self-learning, and communicating and cooperating with other agents. An agent can be an entity that is software or a combination of software and hardware.

[0043] With the rapid development of cloud computing, new communication services and Internet data services, computing power is becoming a new productivity, and the layout scale of data centers as the physical carrier of computing power is expanding rapidly. The air conditioning system consumes as much as 40% to 60% of the total energy consumption of data centers. The central air conditioning system consists of two main components: the terminal and the cold source system. The energy consumption of the terminal cabinet can account for about 40% of the total energy consumption of the entire air conditioning system. Therefore, energy-saving optimization of the terminal system of the data center can effectively reduce the energy consumption of the data center.

[0044] In the related art, the operation mode of the terminal system of the data center is still mainly based on the traditional proportional integral derivative control (PID) control, and a more conservative control strategy is set by increasing the air supply volume and reducing the air supply temperature, which causes energy waste.

[0045] When adapting to the needs of modeling, simulation and optimization control of the terminal system in the data center, the preset methods that can be used include: AI-CFD method based on AI learning and simulation of temperature field and velocity field, terminal mechanism model modeling method based on Modelica, terminal system optimization control method based on reinforcement learning, deep learning modeling algorithm, model-free reinforcement learning optimization algorithm, etc.

[0046] Since there are few measurement points in the terminal system of the data center at present, there is a lack of information on the water flow and temperature, air volume and temperature of each branch and air duct. Therefore, it is difficult to use model-based optimization methods for optimization, such as heuristic search algorithms based on genetic algorithms, MPC algorithms based on the dynamic characteristics model of the controlled system, etc.

[0047] Model-free reinforcement learning algorithms require a large amount of historical data for training. However, many data center end systems do not have sufficient training data due to missing measurement points.

[0048] The traditional data center terminal optimization control method only monitors the return air temperature of the terminal system and uses the return air temperature for strategic optimization control, but the return air temperature of the terminal system cannot accurately represent the temperature of the cabinet.

[0049] The optimization of the terminal system of the data center cannot only consider the temperature of a single cabinet, but also needs to consider the temperature field cloud map of the entire data center room, and the temperature field cloud map must be linked with the optimization control algorithm. However, the traditional Transformer model is not good at handling long-distance dependency problems, and the multi-layer Transformer model has a large memory footprint.

[0050] In view of this, an embodiment of the present application provides a model training method, which includes: inputting multiple groups of verification parameters of a refrigeration device into a scene simulation model to obtain scene simulation data corresponding to each group of verification parameters; the refrigeration device is used to cool multiple network devices; the multiple groups of verification parameters include multiple groups of refrigeration parameters under different boundary conditions; the scene simulation data includes the energy consumption of the refrigeration device and the temperature data of each network device area; training the enhancement model based on the multiple groups of verification parameters to obtain a target enhancement model; the target enhancement model is used to determine the target refrigeration parameter based on the current boundary conditions of the refrigeration device; the target refrigeration parameter is the refrigeration parameter with the lowest energy consumption for the refrigeration device among multiple groups of refrigeration parameters to be selected; the temperature data of each network device area is less than the temperature threshold when the refrigeration device executes the selected refrigeration parameter.

[0051] The method provided in the embodiment of the present application is described in detail below with reference to the accompanying drawings.

[0052] It should be noted that the network system described in the embodiment of the present application is to more clearly illustrate the technical solution of the embodiment of the present application, and does not constitute a limitation on the technical solution provided in the embodiment of the present application. Ordinary technicians in this field can know that with the evolution of network systems and the emergence of other network systems, the technical solution provided in the embodiment of the present application is also applicable to similar technical problems.

[0053] Figure 1 FIG. 1 is a schematic diagram of a model training system 10 provided in an embodiment of the present application. Figure 1 As shown, the model training system 10 may include a terminal device 11 and a model training apparatus 12.

[0054] The terminal device 11 and the model training device 12 are connected to each other. For example, they can be connected wirelessly or by wire.

[0055] Among them, the terminal device 11 involved in the embodiment of the present application is used to send training data to the model training device 12. For example, the terminal device 11 can be any computer device or server, wherein the computer device includes but is not limited to a mobile phone, a tablet computer, a desktop computer, a laptop computer, a vehicle-mounted terminal, a handheld terminal, an augmented reality (AR) device, a virtual reality (VR) device, etc., and the present embodiment of the disclosure does not impose any special restrictions on the specific form of the terminal device 11. It can interact with the user through one or more methods such as a keyboard, a touchpad, a touch screen, a remote control, voice interaction or a handwriting device. In the embodiment of the present application, the specific technology and specific device form adopted by the terminal device 11 are not limited.

[0056] The model training device 12 involved in the embodiments of the present application can be used to receive training data of a terminal device and perform model training based on the training data. It can be an electronic device with processing functions such as a computer or a server. For example, the model training device 12 can be a computer, a server, etc. Among them, the server can be a single server, or it can be a server cluster composed of multiple servers. In some implementations, the server cluster can also be a distributed cluster. The embodiments of the present application do not limit the specific technology, specific number, and specific device form of the terminal device 11. In the embodiments of the present application, the specific technology and specific device form adopted by the model training device 12 are not limited.

[0057] Figure 1 This is just an exemplary framework diagram. Figure 1 The names of the individual devices included in the Figure 1 In addition to the functional nodes shown, other nodes may also be included, which is not limited in the embodiments of the present application.

[0058] Figure 2 FIG. 1 is a schematic diagram of the structure of another model training system 10 provided in an embodiment of the present application. Figure 2 As shown, the model training system 10 may include a CFD model, an AI-CFD model, a mechanism model, and a reinforcement learning model.

[0059] Among them, the AI-CFD model is obtained through training based on the CFD model. The model training system is used for joint simulation based on the AI-CFD model, mechanism model, and reinforcement learning model to optimize the energy consumption of refrigeration equipment.

[0060] When implementing it specifically, Figure 1 Each device in the Figure 3 The structure shown, or including Figure 3 Parts shown. Figure 3 The structure diagram of a model training device 200 provided in an embodiment of the present application is shown in FIG. 2 . The model training device 200 may be a network device, or the model training device 200 may be a chip or a system on a chip in a network device. Figure 3 As shown, the model training device 200 includes a processor 201 , a communication interface 202 and a communication line 203 .

[0061] Furthermore, the model training device 200 may also include a memory 204. The processor 201, the memory 204 and the communication interface 202 may be connected via a communication line 203.

[0062] The processor 201 is a CPU, a general processor, a network processor (NP), a digital signal processor (DSP), a microprocessor, a microcontroller, a programmable logic device (PLD), or any combination thereof. The processor 201 may also be other devices with processing functions, such as circuits, devices, or software modules, without limitation.

[0063] The communication interface 202 is used to communicate with other devices or other communication networks. The communication interface 202 can be a module, a circuit, a communication interface or any device capable of implementing communication.

[0064] The communication line 203 is used to transmit information between the components included in the model training device 200.

[0065] The memory 204 is used to store instructions, where the instructions may be computer programs.

[0066] The memory 204 may be a read-only memory (ROM) or other types of static storage devices that can store static information and / or instructions, or a random access memory (RAM) or other types of dynamic storage devices that can store information and / or instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), magnetic disk storage media or other magnetic storage devices, etc., without limitation.

[0067] It should be noted that the memory 204 can exist independently of the processor 201, or can be integrated with the processor 201. The memory 204 can be used to store instructions or program codes or some data, etc. The memory 204 can be located in the model training device 200, or can be located outside the model training device 200, without limitation. The processor 201 is used to execute the instructions stored in the memory 204 to implement the model training method provided in the following embodiments of the present application.

[0068] In one example, the processor 201 may include one or more CPUs, for example, Figure 3 CPU0 and CPU1 in.

[0069] As an optional implementation, the model training device 200 includes multiple processors, for example, Figure 3 In addition to the processor 201, a processor 205 may also be included.

[0070] It should be pointed out that Figure 3 The composition shown in the Figure 1 The limitations of each device in Figure 3 In addition to the parts shown, Figure 1 The individual devices in the Figure 3 More or fewer components, or combinations of certain components, or different arrangements of components.

[0071] In the embodiment of the present application, the chip system may be composed of a chip, or may include a chip and other discrete devices.

[0072] In addition, the actions, terms, etc. involved in the various embodiments of the present application can refer to each other without limitation. The message name or parameter name in the message exchanged between the various devices in the embodiments of the present application is only an example, and other names can also be used in the specific implementation without limitation.

[0073] In order to clearly describe the technical solutions of the embodiments of the present application, in the embodiments of the present application, words such as "first" and "second" are used to distinguish the same or similar items with substantially the same functions and effects. Those skilled in the art can understand that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit the difference.

[0074] It should be noted that, in this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in this application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0075] In the present application, "at least one" means one or more, and "plurality" means two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0076] Combine the following Figure 1 The model training system shown describes the model training method provided in the embodiment of the present application.

[0077] Figure 4 A model training method is provided for an embodiment of the present application, which is applied to a model training device and can also be applied to devices in the model training device, such as chips.

[0078] The present application embodiment is described by taking the application in a model training device as an example. Figure 4 As shown, the method includes the following S301-S302:

[0079] S301. Input multiple groups of verification parameters of the refrigeration equipment into a scene simulation model to obtain scene simulation data corresponding to each group of verification parameters.

[0080] The refrigeration equipment is used to cool multiple network devices, for example, it can be an air conditioner, etc. The multiple sets of verification parameters include multiple sets of refrigeration parameters under different boundary conditions; the scenario simulation data includes the energy consumption of the refrigeration equipment and the temperature data of each network device area.

[0081] Among them, the boundary conditions include at least one of the following: outdoor air dry-bulb temperature, outdoor air wet-bulb temperature, chilled water supply temperature, chilled water flow, cooling load power; fan power, temperature threshold; refrigeration parameters include at least one of the following: return air valve opening, fresh air valve opening, return air fan frequency, supply air fan frequency, chilled water valve opening, supply air temperature setting value.

[0082] As a possible implementation method, the model training device can input multiple groups of verification parameters into the mechanism model to obtain the air supply parameters corresponding to each group of verification parameters; the air supply parameters, outdoor temperature, outdoor humidity and power of multiple network devices corresponding to each group of verification parameters are input into the artificial intelligence computational fluid dynamics AI-CFD model to obtain the scene simulation data corresponding to each group of verification parameters.

[0083] Among them, the air supply parameters include air supply volume and air supply temperature.

[0084] It should be noted that the mechanism model is based on the situation of the refrigeration equipment terminal in the data center computer room, and the data center air conditioning terminal system is modeled using Modelica. The modeling content includes the air handling unit (AHU), the fresh air terminal (MAU), the fan coil unit (FCU), valve equipment, fan equipment, water and air pipes, etc.

[0085] It should be noted that the model training device can determine the maximum and minimum values ​​of each boundary condition and different cooling parameters and the data discrete step length, and combine different boundary conditions and different cooling parameters to obtain multiple groups of verification parameters.

[0086] In one example, when the boundary conditions include the first boundary condition and the second boundary condition, and the refrigeration parameters include the first refrigeration parameter and the second refrigeration parameter. If the parameter values ​​of the first boundary condition are a1 and a2, the parameter values ​​of the second boundary condition are b1 and b2, the parameter values ​​of the first refrigeration parameter are c1 and c2, and the parameter values ​​of the second refrigeration parameter are d1 and d2, then the multiple groups of verification parameters can be (a1, b1, c1, d1), (a1, b1, c1, d2), (a1, b1, c2, d1), (a1, b1, c2, d2), (a1, b2, c1, d1), ( a1, b2, c1, d2), (a1, b2, c2, d1), (a1, b2, c2, d2), (a2, b1, c1, d1), (a2, b1, c1, d2), (a2, b1, c2, d1), (a2, b2, c1, d1), (a2, b2, c1, d1), (a2, b2, c1, d2), (a2, b2, c2, d1), (a2, b2, c2, d2).

[0087] S302: Train the reinforcement model based on multiple groups of verification parameters to obtain a target reinforcement model.

[0088] The target enhancement model is used to determine the target cooling parameter based on the current boundary conditions of the cooling device. The target cooling parameter is the cooling parameter with the lowest energy consumption of the cooling device among the multiple groups of candidate cooling parameters. When the cooling device executes the candidate cooling parameter, the temperature data of each network device area is less than the temperature threshold.

[0089] The temperature threshold can be set as required, for example, 40 degrees, 50 degrees, etc.

[0090] As a possible implementation method, the model training device can divide the verification parameters into a training set and a test set; use a deep Q network (DQN) algorithm to build a reinforcement model, input the verification parameters of the training set into the DQN algorithm for training, adjust the model parameters and functions to obtain the trained reinforcement model, input the verification parameters of the test set into the trained reinforcement model, and verify the accuracy of the trained reinforcement model; and when the accuracy is greater than the accuracy threshold, determine the trained reinforcement model as the target reinforcement model.

[0091] As another possible implementation method, the model training device can determine the reward value for executing each set of verification parameters based on the scene simulation data and reward function corresponding to each set of verification parameters; train the reinforcement model based on multiple sets of verification parameters and the reward value of each set of verification parameters to obtain the target reinforcement model.

[0092] It should be noted that the termination condition of the enhanced model training may be that the temperature data of the network device area is greater than or equal to the temperature threshold.

[0093] It should be noted that the reward function is used to reward the reinforcement learning model (such as increasing the reward value of the verification parameter) when a set of cooling parameters makes the temperature data of the network device area less than the temperature threshold and the energy consumption of the cooling equipment decreases.

[0094] In one embodiment, the reward function is also used to punish the reinforcement learning model (such as reducing the reward value of the verification parameter) when a set of cooling parameters causes the temperature data of the network device area to be greater than or equal to the temperature threshold, or when the energy consumption of the cooling equipment does not decrease.

[0095] Based on the technical solution provided by the present application, multiple groups of verification parameters of the refrigeration equipment are input into the scene simulation model to obtain scene simulation data corresponding to each group of verification parameters; the reinforcement model is trained based on the multiple groups of verification parameters to obtain the target reinforcement model. In this way, by constructing different simulation data in the scene simulation model, the number of parameters of the optimization model can be effectively increased, and the reinforcement model can be trained based on more simulation data to ensure the training effect of the reinforcement model. While effectively reducing the energy consumption of the refrigeration equipment, it can avoid overheating of the local network equipment area, thereby ensuring the effect of optimizing the energy consumption of the refrigeration equipment.

[0096] A possible embodiment is as follows Figure 5 As shown, in order to determine the AI-CFD model, the model training method of the present application may also include the following S401-S402.

[0097] S401, generating multiple sets of training data based on a target computational fluid dynamics (CFD) model.

[0098] The training data includes air supply parameters, power of network equipment and temperature data of each corresponding network equipment area.

[0099] As a possible implementation method, the model training device can input the air supply parameters and the power of the network equipment into the CFD model, so that the target CFD model outputs the temperature data of each network equipment area corresponding to the air supply parameters and the power of the network equipment. Further, the air supply parameters, the power of the network equipment, and the corresponding temperature data of each network equipment area are used as training data.

[0100] As a possible implementation method, the model training device can input air supply parameters and the power of network equipment into the CFD model, so that the target CFD model outputs the temperature data of each network equipment area corresponding to the air supply parameters and the power of network equipment. Further, the temperature data of each network equipment area is feature processed to obtain a temperature characteristic diagram of each network equipment area. The air supply parameters, the power of network equipment, and the corresponding temperature characteristic diagram of each network equipment area are used as training data.

[0101] It should be noted that the temperature characteristic graph is data that can be recognized by the neural network model.

[0102] In some embodiments, the training data may further include defining a signed distance characteristic map, a velocity characteristic map, and a pressure characteristic map.

[0103] S402: Based on the training data, the preset neural network model is trained to obtain an AI-CFD model.

[0104] Among them, the preset neural network model can be a multi-feature Reformer neural network.

[0105] As a possible implementation method, the model training device can add constraints of thermal physics equations to the preset neural network intermediate layer and output layer based on the training data to train the AI-CFD model.

[0106] A possible embodiment is as follows Figure 6 As shown, in order to construct the target CFD model, the model training method of the present application may also include the following S501-S502.

[0107] S501. Obtain historical temperature data, historical air supply parameters and three-dimensional structure data of a target area.

[0108] The target area may be an internal area of ​​a data center. The three-dimensional structure data includes the location of each device in the target area and the configuration of the flow field; the devices include network devices and refrigeration devices. The historical temperature data and the historical air supply parameters may refer to the temperature data of a preset historical period and the air supply parameter history of a preset historical period, respectively. For example, the preset historical period may be one month before the current moment, or one week before the current moment, etc.

[0109] As a possible implementation method, the refrigeration equipment is connected to a cloud server, and the cloud server can store the working condition data of the refrigeration equipment. The model training device can send a request message to the cloud server to obtain the historical temperature data and historical air supply parameters of the target area. The model training device obtains the three-dimensional structure data based on the design drawings of the target area, or performs a three-dimensional scan on the target area to obtain the three-dimensional structure data.

[0110] S502: Construct a target CFD model based on historical temperature data, historical air supply parameters and three-dimensional structure data.

[0111] Among them, the process of building the target CFD model can include geometric modeling, meshing, boundary condition setting, etc.

[0112] As a possible implementation method, the model training device can simulate the temperature field distribution of the target area according to the three-dimensional structure data, the historical temperature data and the CFD simulation software for the target area to determine the initial CFD model. Further, the loss value of the initial CFD model is determined, and when the loss value of the initial CFD model is less than the loss value threshold, the CFD model is determined as the target CFD model.

[0113] It should be noted that the loss value is used to characterize the degree of loss between the temperature data output by the initial CFD model and the actual temperature data.

[0114] In one example, the model training device can use the rich component library (air conditioning components, network equipment components, chillers, cooling towers, etc.) of the CFD simulation software based on the computer room layout CAD drawings to carefully model and simulate the spatial structure of the refrigeration equipment (including the spatial layout of server IT equipment, the spatial layout of hot and cold channels, the spatial layout of refrigeration equipment, the spatial layout of temperature sensors, the system structure of refrigeration equipment, etc.) and the refrigeration system (refrigeration equipment and network equipment models, various parameter indicators of chillers and cooling towers), thereby establishing a computer room temperature field distribution model and a cooling system model, which can more accurately characterize the changes in the temperature of each measuring point in the target area over time and spatial distribution, as well as the relationship between various variables of the cooling system.

[0115] In one embodiment, when it is determined that the loss value of the initial CFD model is greater than or equal to the loss value threshold, the model training device may update the model parameters until the loss value of the CFD model is less than the loss value threshold.

[0116] It can be understood that when the loss value of the initial CFD model is greater than or equal to the loss value threshold, it means that the accuracy of the initial CFD model is low. At this time, the model training device can update the model parameters until the loss value is less than the loss value threshold, and a target CFD model with higher accuracy can be obtained.

[0117] The various solutions in the above embodiments of the present application can be combined without contradiction.

[0118] The embodiment of the present application can divide the model training device into functional modules or functional units according to the above method example. For example, each functional module or functional unit can be divided corresponding to each function, or two or more functions can be integrated into one processing module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of software functional modules or functional units. Among them, the division of modules or units in the embodiment of the present application is schematic, which is only a logical function division, and there may be other division methods in actual implementation.

[0119] In the case of dividing each functional module into corresponding functional modules, Figure 7 A structural schematic diagram of a model training device 700 is shown. The model training device can be a model training device or a chip used in a model training device. The model training device 700 can be used to execute the functions of the model training device involved in the above embodiments. Figure 7 The model training device 700 shown may include: an input unit 701 and a processing unit 702; the input unit 701 is used to input multiple groups of verification parameters of the refrigeration equipment into the scene simulation model to obtain scene simulation data corresponding to each group of verification parameters; the refrigeration equipment is used to cool multiple network devices; the multiple groups of verification parameters include multiple groups of refrigeration parameters under different boundary conditions; the scene simulation data include the energy consumption of the refrigeration equipment and the temperature data of each network device area; the processing unit 702 is used to train the enhancement model based on the multiple groups of verification parameters to obtain the target enhancement model; the target enhancement model is used to determine the target refrigeration parameter based on the current boundary conditions of the refrigeration equipment; the target refrigeration parameter is the refrigeration parameter with the lowest energy consumption among the multiple groups of refrigeration parameters to be selected; the temperature data of each network device area is less than the temperature threshold when the refrigeration equipment executes the selected refrigeration parameter.

[0120] Optionally, the processing unit 702 is specifically configured to: determine a reward value for executing each set of verification parameters based on the scene simulation data corresponding to each set of verification parameters and a reward function;

[0121] The reinforcement model is trained based on multiple groups of verification parameters and the reward value of each group of verification parameters to obtain a target reinforcement model.

[0122] Optionally, the processing unit 702 is specifically used to: input multiple groups of verification parameters into the mechanism model to obtain air supply parameters corresponding to each group of verification parameters; the air supply parameters include air volume and air temperature; the air supply parameters corresponding to each group of verification parameters, outdoor temperature, outdoor humidity and power of multiple network devices are input into the artificial intelligence computational fluid dynamics AI-CFD model to obtain scene simulation data corresponding to each group of verification parameters.

[0123] Optionally, the processing unit 702 is also used to generate multiple sets of training data based on the target computational fluid dynamics (CFD) model; the training data includes air supply parameters, the power of the network equipment, and the corresponding temperature data of each network equipment area; the processing unit 702 is also used to train the preset neural network model based on the training data to obtain an AI-CFD model.

[0124] Optionally, the device includes an acquisition unit 703, which is used to acquire historical temperature data, historical air supply parameters and three-dimensional structure data of the target area; the three-dimensional structure data includes the position of each device in the target area and the configuration of the flow field; the equipment includes network equipment and refrigeration equipment; the processing unit 702 is also used to construct a target CFD model based on the historical temperature data, historical air supply parameters and three-dimensional structure data.

[0125] Optionally, the processing unit 702 is specifically used to: construct an initial CFD model based on historical temperature data, historical air supply parameters and three-dimensional structure data; determine a loss value of the initial CFD model, the loss value is used to characterize the degree of loss between the temperature data output by the initial CFD model and the actual temperature data; when the loss value of the initial CFD model is less than a loss value threshold, determine the CFD model as a target CFD model.

[0126] Optionally, the boundary conditions include at least one of the following: outdoor air dry-bulb temperature, outdoor air wet-bulb temperature, chilled water supply temperature, chilled water flow rate, cooling load power; fan power, temperature threshold; the refrigeration parameters include at least one of the following: return air valve opening, fresh air valve opening, return air fan frequency, supply air fan frequency, chilled water valve opening, supply air temperature setting value.

[0127] The embodiment of the present application also provides a computer-readable storage medium. All or part of the processes in the above method embodiments can be completed by a computer program to instruct the relevant hardware, and the program can be stored in the above computer-readable storage medium. When the program is executed, it may include the processes of the above method embodiments. The computer-readable storage medium can be an internal storage unit of the model training device (including a data sending end and / or a data receiving end) of any of the above embodiments, such as a hard disk or memory of the model training device. The above computer-readable storage medium can also be an external storage device of the above terminal device, such as a plug-in hard disk equipped on the above terminal device, a smart memory card (smart media card, SMC), a secure digital (secure digital, SD) card, a flash card (flash card), etc. Further, the above computer-readable storage medium can also include both the internal storage unit of the above model training device and an external storage device. The above computer-readable storage medium is used to store the above computer program and other programs and data required by the above model training device. The above computer-readable storage medium can also be used to temporarily store data that has been output or is to be output.

[0128] It should be noted that the terms "first" and "second" in the specification, claims and drawings of the present application are used to distinguish different objects rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or devices.

[0129] It should be understood that in the present application, "at least one (item)" means one or more, "more than one" means two or more, "at least two (items)" means two or three and more than three, and "and / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0130] Through the description of the above implementation methods, technical personnel in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0131] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0132] The units described as separate components may or may not be physically separated, and the components shown as units may be one physical unit or multiple physical units, that is, they may be located in one place or distributed in multiple different places. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0133] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0134] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solution of the embodiment of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions to enable a device (which can be a single-chip microcomputer, chip, etc.) or a processor (processor) to perform all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard drives, ROM, RAM, magnetic disks or optical disks.

[0135] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto, and any changes or substitutions within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A model training method, characterized in that: The method comprises: Input multiple sets of verification parameters of the refrigeration equipment into the scenario simulation model to obtain scenario simulation data corresponding to each set of verification parameters; the refrigeration equipment is used to cool multiple network devices; the multiple sets of verification parameters include multiple sets of refrigeration parameters under different boundary conditions; the scenario simulation data includes energy consumption of the refrigeration equipment and temperature data of each network device area; The reinforcement model is trained based on the multiple groups of verification parameters to obtain a target reinforcement model; the target reinforcement model is used to determine the target refrigeration parameter based on the current boundary conditions of the refrigeration equipment; the target refrigeration parameter is the refrigeration parameter with the lowest energy consumption for the refrigeration equipment among the multiple groups of refrigeration parameters to be selected; when the refrigeration equipment executes the refrigeration parameter to be selected, the temperature data of each network device area is less than the temperature threshold.

2. The method according to claim 1, characterized in that The step of training the reinforcement model based on the multiple groups of verification parameters to obtain a target reinforcement model includes: Determine a reward value for executing each set of verification parameters based on the scenario simulation data and the reward function corresponding to each set of verification parameters; The reinforcement model is trained based on the multiple groups of verification parameters and the reward value of each group of verification parameters to obtain the target reinforcement model.

3. The method according to claim 1 or 2, characterized in that: The step of inputting multiple groups of verification parameters of the refrigeration equipment into the scene simulation model to obtain scene simulation data corresponding to each group of verification parameters includes: Input the multiple groups of verification parameters into the mechanism model to obtain air supply parameters corresponding to each group of verification parameters; the air supply parameters include air supply volume and air supply temperature; The air supply parameters, outdoor temperature, outdoor humidity and power of the multiple network devices corresponding to each set of verification parameters are input into the artificial intelligence computational fluid dynamics AI-CFD model to obtain scene simulation data corresponding to each set of verification parameters.

4. The method according to claim 3, characterized in that: The method further comprises: Based on the target computational fluid dynamics (CFD) model, multiple sets of training data are generated; the training data include air supply parameters, the power of the network equipment, and the temperature data of each corresponding network equipment area; Based on the training data, a preset neural network model is trained to obtain the AI-CFD model.

5. The method according to claim 4, characterized in that The method further comprises: Acquire historical temperature data, historical air supply parameters and three-dimensional structure data of the target area; the three-dimensional structure data includes the location of each device in the target area and the configuration of the flow field; the devices include network devices and refrigeration equipment; The target CFD model is constructed based on the historical temperature data, the historical air supply parameters and the three-dimensional structure data.

6. The method according to claim 5, characterized in that The constructing the target CFD model based on the historical temperature data, the historical air supply parameters and the three-dimensional structure data includes: Building an initial CFD model based on the historical temperature data, historical air supply parameters and the three-dimensional structure data; Determining a loss value of the initial CFD model, wherein the loss value is used to characterize a degree of loss between temperature data output by the initial CFD model and actual temperature data; When the loss value of the initial CFD model is less than a loss value threshold, the CFD model is determined as the target CFD model.

7. The method according to claim 1, characterized in that The boundary conditions include at least one of the following: outdoor air dry-bulb temperature, outdoor air wet-bulb temperature, chilled water supply temperature, chilled water flow, cooling load power; fan power, temperature threshold; the refrigeration parameters include at least one of the following: return air valve opening, fresh air valve opening, return air fan frequency, supply air fan frequency, chilled water valve opening, supply air temperature setting value.

8. A model training device, characterized in that: The device comprises: an input unit and a processing unit; The input unit is used to input multiple groups of verification parameters of the refrigeration equipment into the scene simulation model to obtain scene simulation data corresponding to each group of verification parameters; the refrigeration equipment is used to cool multiple network devices; the multiple groups of verification parameters include multiple groups of refrigeration parameters under different boundary conditions; the scene simulation data includes energy consumption of the refrigeration equipment and temperature data of each network device area; The processing unit is used to train the reinforcement model based on the multiple groups of verification parameters to obtain a target reinforcement model; the target reinforcement model is used to determine the target refrigeration parameter based on the current boundary conditions of the refrigeration equipment; the target refrigeration parameter is the refrigeration parameter with the lowest energy consumption for the refrigeration equipment among the multiple groups of refrigeration parameters to be selected; when the refrigeration equipment executes the refrigeration parameter to be selected, the temperature data of each network device area is less than a temperature threshold.

9. A computer-readable storage medium, characterized in that: The readable storage medium stores instructions, and when the instructions are executed, the method according to any one of claims 1 to 7 is implemented.

10. A model training device, characterized in that: include: A processor, a memory and a communication interface; wherein the communication interface is used for the model training device to communicate with other devices or networks; The memory is used to store one or more programs, which include computer-executable instructions. When the model training device is running, the processor executes the computer-executable instructions stored in the memory so that the model-based training device performs the method described in any one of claims 1-7.

Citation Information

Patent Citations

  • Data center machine room energy-saving control method and device, equipment and storage medium

    CN112888268A

  • Data center energy consumption joint optimization method, system, medium and equipment

    CN112966431A

  • Temperature control method, equipment, modular data center and storage medium

    CN113821903A

  • Data center refrigeration equipment control method based on multi-agent reinforcement learning

    CN115408957A

  • Data center energy consumption optimization method and device and storage medium

    CN116795198A

Cited By

  • Boundary backbone route determination method and device, electronic equipment and storage medium

    CN119341970A

  • Model training method and apparatus, and readable storage medium

    WO2026020731A1