Temperature control method and device of formation and capacity grading system, medium and equipment

Abstract

The invention relates to the technical field of battery temperature control, in particular to a temperature control method and device for a capacity grading system, a medium and equipment, and the method comprises the steps: S1, obtaining real-time data in the charging and discharging process of a battery module; s2, constructing a dynamic thermal model based on a neural network, preprocessing the real-time data, taking the preprocessed real-time data as input, and predicting a temperature prediction value after the next three seconds through the dynamic thermal model; s3, the current working mode of the formation and capacity grading system is judged based on the temperature predicted value, a PID control signal is generated, and meanwhile an optimal air volume scheme is calculated through an alternating direction multiplier method; and S4, the PID control signal is output to a thermoelectric refrigerating unit for segmented adjustment of the refrigerating power, meanwhile, the rotating speed of a plurality of independent fans is adjusted according to the optimal air volume scheme, and the temperature fluctuation range is reduced while the dynamic response speed and the energy efficiency ratio are increased.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of battery temperature control, and particularly relates to a temperature control method, device, medium and equipment of a component container system. BACKGROUND

[0002] In the manufacturing process of secondary batteries such as lithium ion batteries, battery charge-discharge activation (formation) and capacity test (component test) are crucial core processes. These processes directly determine the initial performance, consistency, long-term cycle life and safety reliability of the battery. During this process, complex electrochemical reactions occur inside the battery, and the progress and effect of the reactions are highly sensitive to the temperature environment. Therefore, accurate, rapid and uniform temperature control of the production environment (especially the temperature control cavity) is a key factor in ensuring the final quality of the battery. Small deviations or instability in temperature parameters can directly lead to an increase in the dispersion of battery performance parameters (such as capacity, internal resistance, self-discharge rate) (a decrease in consistency), an acceleration of life decay, and even hidden safety hazards.

[0003] Currently, the industry widely uses temperature control systems based on traditional proportional-integral-derivative (PID) algorithms to maintain the required control temperature of the temperature control cavity (such as the formation cabinet and component test cabinet). However, this technology has significant technical defects in actual application, especially in complex scenarios where the battery test charge-discharge current changes rapidly and the working conditions switch frequently, which severely restricts the improvement of battery production efficiency and quality:

[0004] Poor dynamic response performance and long recovery time: The traditional PID control algorithm performs poorly when dealing with sudden changes in working conditions (such as instantaneous step changes in charge-discharge current (the rate can reach 50A / s or more) or rapid switching during the test phase). Its inherent response lag characteristic causes the temperature control system to be unable to quickly track changes in the set value, making it prone to significant temperature overshoot (for example, within a target range of ±2℃, the actual overshoot can reach ±3.5℃ or even higher, i.e., the overshoot amplitude exceeds 15% of the set value range). Moreover, the time required for the system to recover from the overshoot state to the stable set value range is also too long (≥30 seconds). This large temperature fluctuation and long deviation from the steady state have irreversible negative effects on the batteries being formed or component tested, severely damaging the consistency of the batch of batteries.

[0005] Temperature field uniformity is insufficient, and spatial gradient is large: The existing temperature control system is difficult to achieve highly uniform temperature distribution in a complex cavity structure. There is a significant temperature difference between different positions in the cavity (such as the central region and the edge region), and the highest temperature difference actually measured can reach 7.2°C, and a temperature gradient of more than 5°C generally exists. This non-uniform temperature environment causes the same batch of batteries to be in different microclimates, resulting in differences in electrochemical reaction rate and degree, which is one of the main process reasons for the discreteness (i.e., poor consistency) of the performance parameters (capacity, voltage, etc.) of the battery monomers.

[0006] Multiple-source interference lacks effective dynamic compensation: The battery charging and discharging process itself generates Joule heat, which is strongly related to the charging and discharging current (size, rate of change); at the same time, environmental temperature and humidity fluctuations are also important interference sources. The existing traditional PID control system lacks the ability to model and compensate for real-time and accurate dynamic coupling of such multi-parameter (current, environmental temperature and humidity) interference. The system cannot actively and prospectively adjust the control output according to the real-time changes in the charging and discharging current and the changes in the environmental conditions, resulting in further deterioration of temperature control accuracy under interference.

[0007] Low energy utilization efficiency and high operating cost: To maintain the set constant temperature environment, especially in cases where rapid cooling is required, the traditional solution often relies on a continuously high-power operating refrigeration system (such as a compressor). This "extensive" constant temperature control strategy results in an abnormally large energy consumption of the cooling system, which can account for more than 40% of the total energy consumption of the entire battery production line. This not only significantly increases production costs, but also goes against the current trend of green and low-carbon development in the manufacturing industry.

[0008] Temperature control hysteresis is significant: The above problems (response lag, insufficient interference compensation) are a manifestation of the hysteresis of the overall temperature control. Under dynamic conditions such as rapid step changes in current, the temperature control system cannot respond in time, and the temperature control action significantly lags behind the occurrence of thermal disturbance, which is one of the root causes of the above problems such as large overshoot and long recovery time.

[0009] In summary, the existing battery production temperature control technology based on traditional PID control has a series of technical bottlenecks that are difficult to overcome when dealing with high dynamics and multiple interference in battery formation and capacity distribution, such as response lag, large overshoot, long recovery time, poor temperature uniformity, weak anti-interference ability, and high energy consumption. These defects directly lead to difficulties in improving the consistency and yield of battery products, high production costs (especially energy consumption costs), and may affect the long-term reliability and safety of the battery. SUMMARY

[0010] In order to overcome the deficiencies of the prior art, the present application provides a temperature control method for a formation and capacity distribution system, which aims to improve the dynamic response speed and energy efficiency ratio while reducing the temperature fluctuation range.

[0011] The technical means adopted by the present application to solve its technical problems is: a temperature control method of a component storage system, which is improved in that it comprises:

[0012] Step S1, acquiring real-time data in the battery module charging and discharging process;

[0013] Step S2, constructing a dynamic thermal model based on a neural network, pre-processing the real-time data as input, and predicting the temperature prediction value after 3 seconds through the dynamic thermal model;

[0014] Step S3, judging the current working mode of the formation and storage system based on the temperature prediction value, generating a PID control signal, and calculating an optimal air volume scheme using the alternating direction multiplier method;

[0015] Step S4, outputting the PID control signal to the thermoelectric refrigerator for segmented adjustment of refrigeration power, and adjusting the speed of several independent air fans according to the optimal air volume scheme.

[0016] The step S2 in the above technical solution comprises:

[0017] Step S201, using historical charging and discharging data as sample data, dividing the sample data into a training set, a validation set and a test set according to 7:2:1; the historical charging and discharging data includes charging and discharging current I(t), ambient temperature Tenv(t) and battery SOC value;

[0018] Step S202, constructing a 5-dimensional vector X(t) = [I(t), Tavg(t), dI / dt, SOC, Tenv(t)] according to the historical charging and discharging data in the training set, and using a 5-dimensional vector time series with a length of 10 seconds and a step of 100 ms as input after normalizing the 5-dimensional vector X(t); Tavg(t) is the average temperature, and dI / dt is the current change rate;

[0019] Step S203, constructing an LSTM neural network model to obtain the temperature prediction value of the 5-dimensional vector time series after 3 seconds in the training set, and adjusting the parameters of the LSTM neural network model according to the data in the validation set until the training termination condition is reached.

[0020] Step S204, using the trained LSTM neural network model to predict the temperature prediction value after 3 seconds.

[0021] The training termination condition of step S203 in the above technical solution comprises:

[0022] Calculate the error |ΔT| between the predicted temperature value Tpred(t+3s) 3 seconds later and the actual measured value Treal(t+3s) in the test set;

[0023] If the error |ΔT| > 0.5℃, online incremental learning is used to update the weights of the LSTM neural network model until the proportion of errors |ΔT| ≤ 0.5℃ is > 92%, thus completing the training of the LSTM neural network model.

[0024] The LSTM neural network model described in the above technical solution comprises two stacked LSTM layers, each with 128 neurons; the two stacked LSTM layers are followed by a fully connected output layer; the LSTM layers use ReLU as the activation function, and the output layer uses a linear activation function; mean squared error is used as the loss function, and the Adam optimizer is used for training.

[0025] Step S3 in the above technical solution includes:

[0026] Step S301: When the error |ΔT|>=1℃, it is an emergency mode. All thermoelectric coolers (TECs) are started to the maximum cooling power. A PID control signal is generated according to the on / off state of the thermoelectric coolers (TECs). At the same time, all fans run to the highest speed.

[0027] Step S302: When the error |ΔT| is between 0.3℃ and 1℃, it is in economic mode. Part or all of the thermoelectric coolers (TEC) are started. A PID control signal is generated according to the on / off state of the thermoelectric coolers (TEC). The fan runs at the base speed calculated by the alternating direction multiplier method.

[0028] Step S303: When the error |ΔT| <= 0.3℃, it is in equilibrium mode. The thermoelectric cooler TEC is turned off. A PID control signal is generated according to the on / off state of the thermoelectric cooler TEC. The fan operates completely according to the optimal air volume distribution scheme (Qi) calculated by the alternating direction multiplier method.

[0029] Step S3 in the above technical solution further includes: adaptively adjusting the PID dynamic optimization algorithm parameters according to the current change rate dI / dt, including:

[0030] Kp = Kp0 * (1 + α|dI / dt|); Ki = Ki0 / (1 + β|dI / dt|); Kd = Kd0 * e - γ|dI / dt|; where α = 0.02, β = 0.05, γ = 0.03; when |dI / dt| > 5 A / s, switch to enhanced proportional mode (Kp increases by 20%); when |dI / dt| > 10 A / s, the differential term weights decay exponentially to suppress oscillations.

[0031] The optimal air volume scheme calculated by the alternating direction multiplier method comprises:

[0032] Step S401, acquiring the temperature T of each cabinet i And the air volume u of each air outlet of the air conditioner i Wherein i=1,…,N;

[0033] Step S402, minimizing the sum of the square deviations of all area temperatures from the set value according to the following objective function, while constraining the temperature difference between any two areas to be no more than 0.5℃:

[0034]

[0035] Wherein the control quantity constraint condition is: the total air volume is constant (15m3 / h), and the single area air volume ranges from 0.1-5.0m3 / h; the temperature safety constraint condition is: 22℃≤Ti≤60℃; Tset is the required control temperature;

[0036] Step S403, initializing all temperatures T i 0 The air volume u is initialized as the real-time temperature collected at the time, which is initialized as 1m3 / h; i 0

[0037] Step S404, using the alternating direction multiplier method to iterate until convergence:

[0038]

[0039] Step S405, outputting the optimal air volume distribution scheme according to the number of areas N and the required control temperature.

[0040] The technical means adopted by the application to solve the technical problems is: a temperature control device of a formation and capacity distribution system, the device comprises:

[0041] A data acquisition module for acquiring real-time data during the charging and discharging process of the battery module;

[0042] A temperature prediction module for constructing a dynamic thermal model based on a neural network, pre-processing the real-time data as input, and predicting the temperature prediction value after 3 seconds through the dynamic thermal model;

[0043] A scheme generation module for determining the current working mode of the formation and capacity distribution system based on the temperature prediction value, generating a PID control signal, and calculating an optimal air volume scheme by using an alternating direction multiplier method

[0044] ​A temperature control module is configured to output the PID control signal to the thermoelectric refrigerator to segmentally adjust the refrigeration power, and to adjust the rotating speeds of the independent air fans according to the optimal air volume scheme.

[0045] The technical means adopted by the present application to solve the technical problems is: an electronic device, comprising: at least one processor, at least one memory, wherein,

[0046] The memory stores program instructions or codes;

[0047] The program instructions or codes are loaded and executed by the processor, so that the electronic device implements the temperature control method of the formation and component distribution system as claimed in any one of the above.

[0048] The technical means adopted by the present application to solve the technical problems is: a storage medium, which stores program instructions or codes, the program instructions or codes are loaded and executed by the processor to implement the temperature control method of the formation and component distribution system as claimed in any one of the above.

[0049] The present application has the beneficial effects that:

[0050] The neural network model is used to predict the battery temperature in the next 3 seconds, combined with historical data and real-time data, which significantly improves the accuracy of temperature prediction, and automatically switches three working modes (emergency mode, economic mode, and balanced mode) according to the prediction error size, realizes dynamic response and resource optimization allocation, combines the alternating direction multiplier method (ADMM) to calculate the optimal air volume distribution scheme, ensures that the temperature difference of each region does not exceed 0.5℃, avoids local overheating, improves battery consistency and life, and has excellent performance of fast dynamic response, high precision, strong robustness, excellent temperature uniformity and high energy efficiency ratio. BRIEF DESCRIPTION OF DRAWINGS

[0051] Figure 1 A temperature control method flow chart of a formation and component distribution system is shown for the embodiments of the present application;

[0052] Figure 2 A flow chart of step S2 is shown for the embodiments of the present application;

[0053] Figure 3 A flow chart of step S3 is shown for the embodiments of the present application;

[0054] Figure 4 A flow chart of the alternating direction multiplier method for calculating the optimal air volume scheme is shown for the embodiments of the present application;

[0055] Figure 5 A structure block diagram of a temperature control device of a formation and component distribution system is shown for the embodiments of the present application;

[0056] Figure 6 A structural diagram of an electronic device is shown for an embodiment of the present application. DETAILED DESCRIPTION

[0057] The present application is further illustrated below with reference to the accompanying drawings and embodiments.

[0058] The concept, specific structure and technical effects of the present application will be described clearly and completely below in combination with embodiments and drawings, so as to fully understand the purpose, features and effects of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments. Based on the embodiments of the present application, other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application. In addition, all the coupling / connection relations involved in the patent do not mean that the components are directly connected, but means that a better coupling structure can be composed by adding or reducing coupling accessories according to the specific implementation situation. The technical features in the present application can be combined interactively without conflict.

[0059] As shown in Figure 1 The present application provides a temperature control method of a formation and capacity grading system, which comprises:

[0060] Step S1, acquiring real-time data in the process of charging and discharging of a battery module.

[0061] The real-time data comprises charging and discharging current I(t), ambient temperature Tenv(t) and battery SOC value.

[0062] Specifically, 24 NTC temperature sensors (interval ≤ 50 mm) are uniformly arranged on the surface of the battery module, the sampling frequency is 10 Hz, the real-time temperature value is Ti(t) (i = 1-24), (Ti(t) represents the temperature value of the i th sensor at time t), and the charging and discharging current (accuracy ± 0.1 A) is synchronously collected, and the above-mentioned multi-source data is synchronously collected through software time stamp alignment protocol (NTP).

[0063] In another possible implementation, the NTC temperature sensor can also be replaced by an infrared thermal imager, but the cost will also increase correspondingly and the anti-interference ability will decrease. The NTC temperature sensor described in the present application is the optimal choice.

[0064] Step S2, constructing a dynamic thermal model based on a neural network, pre-processing the real-time data as input, and predicting the temperature prediction value after 3 seconds through the dynamic thermal model.

[0065] In a possible implementation, as shown in Figure 2 The step S2 comprises:

[0066] Step S201, using historical charging and discharging data as sample data, dividing the sample data into a training set, a validation set and a test set according to 7:2:1; the historical charging and discharging data includes charging and discharging current I(t), ambient temperature Tenv(t) and battery SOC value.

[0067] Among them, the charging and discharging current range: 1000mA-120000mA; the ambient temperature range is 20℃-60℃; the SOC value range is 0%-100%.

[0068] Step S202, constructing a 5-dimensional vector X(t)=[I(t),Tavg(t),dI / dt,SOC,Tenv(t)] according to the historical charging and discharging data in the training set, and using a 5-dimensional vector time sequence with a length of 10 seconds and a step of 100ms as input after normalizing the 5-dimensional vector X(t); Tavg(t) is the average temperature, and dI / dt is the current change rate.

[0069] Among them, the current change rate dI / dt is obtained by difference calculation.

[0070] Optionally, according to the obtained charging and discharging current I(t), the central difference method is used to calculate: dI / dt≈[I(t+Δt)-I(t-Δt)] / (2Δt)(Δt=0.1s).

[0071] Then, the 5-dimensional vector X(t) adopts Min-Max normalization (scaling to [0, 1] range) and the specific formula is as follows:

[0072] X_normalized=(X-X_min) / (X_max-X_min);

[0073] Among them, X is the original data value, X_min is the minimum value of the feature (column) in the entire training data set; X_max: the maximum value of the feature (column) in the entire training data set; X_normalized: the normalized value, the range is between [0, 1].

[0074] After the above pretreatment, the format of the sample data is unified to data between [0, 1].

[0075] Step S203, constructing an LSTM neural network model, obtaining the temperature prediction value of the 5-dimensional vector time sequence in the training set after 3 seconds, and adjusting the parameters of the LSTM neural network model according to the data in the validation set until the training termination condition is reached.

[0076] In a possible implementation, an LSTM neural network model is constructed, which includes two stacked LSTM layers, each layer having 128 neurons; the two stacked LSTM layers are followed by a fully connected output layer; the LSTM layers use RELU as the activation function, and the output layer uses a linear activation function; and the mean square error is used as the loss function, and the Adam optimizer is used for training.

[0077] The preprocessed 5-dimensional vector time series is input into the LSTM neural network model, and a predicted value Tpred(t+3s) is obtained according to the LSTM neural network model.

[0078] An error |ΔT| between the temperature predicted value Tpred(t+3s) and the actual measured value Treal(t+3s) in the test set is calculated.

[0079] If the error |ΔT|>0.5℃, the LSTM neural network model weights are updated using online incremental learning until the proportion of the error |ΔT|≤0.5℃>92%, and the training of the LSTM neural network model is completed.

[0080] Step S204, the trained LSTM neural network model is used to predict the temperature predicted value after 3 seconds.

[0081] The real-time data obtained in step S1 is input into the trained LSTM neural network, which predicts the temperature predicted value after 3 seconds.

[0082] In another possible implementation, the LSTM neural network model can also be a 1D-CNN+GRU combined model, but the corresponding prediction delay will also increase, and the model accuracy will also decrease. The LSTM neural network model described in the present application is the optimal embodiment.

[0083] Step S3, based on the temperature predicted value, the current working mode of the formation and equalization system is determined, a PID control signal is generated, and the optimal air volume scheme is calculated using the alternating direction multiplier method.

[0084] The working mode includes emergency mode, economic mode and balanced mode.

[0085] In a possible implementation, as shown in Figure 3 The step S3 includes:

[0086] Step S301, when the error |ΔT|>=1℃, it is the emergency mode, all thermoelectric coolers TEC are started to the maximum cooling power, a PID control signal is generated according to the on-off state of the thermoelectric cooler TEC, and all fans are operated to the highest speed;

[0087] Step S302, when the error |ΔT| is between 0.3-1℃, it is the economic mode, start part or all of the thermoelectric cooler TEC, generate PID control signal according to the thermoelectric cooler TEC on-off state, the fan runs at the basic speed calculated by the alternating direction multiplier method distribution scheme;

[0088] Step S303, when the error |ΔT| <=0.3℃, it is the equilibrium mode, close the thermoelectric cooler TEC, generate PID control signal according to the thermoelectric cooler TEC on-off state, the fan completely runs according to the optimal air volume distribution scheme (Qi) calculated by the alternating direction multiplier method.

[0089] In another possible implementation, step S3 further comprises: adaptively adjusting the PID dynamic optimization algorithm parameters according to the current rate of change dI / dt, including:

[0090] Kp=Kp0*(1+α|dI / dt|); Ki=Ki0 / (1+β|dI / dt|); Kd=Kd0*e-γ|dI / dt|; Wherein α=0.02, β=0.05, γ=0.03; When |dI / dt|>5A / s, switch to the reinforced proportional mode (Kp is increased by 20%); When |dI / dt|>10A / s, the differential term weight index attenuation suppresses the oscillation.

[0091] Wherein α=0.02, β=0.05, γ=0.03; When |dI / dt|>5A / s, switch to the reinforced proportional mode (Kp is increased by 20%); When |dI / dt|>10A / s, the differential term weight index attenuation suppresses the oscillation.

[0092] Wherein, as shown in Figure 4 The optimal air volume scheme calculated by the alternating direction multiplier method includes:

[0093] Step S401, obtain the temperature T i of each cabinet i , and the air volume u of each air conditioner outlet , Wherein i=1,…,N;

[0094] Step S402, minimize the sum of the squared deviations of all area temperatures from the set value according to the following objective function, while constraining the temperature difference between any two areas to be no more than 0.5℃:

[0095]

[0096] Wherein the control quantity constraint condition is: the total air volume is constant (15m3 / h), and the single area air volume range is 0.1-5.0m3 / h; The temperature safety constraint condition is: 22℃<=Ti<=60℃; Tset is the required control temperature;

[0097] Step S403, initialize all temperatures T i 0 to the real-time temperature collected at that time, initialize the air volume u i 0= 1 m3 / h;

[0098] Step S404, using the alternating direction multiplier method to iterate until convergence:

[0099]

[0100] Step S405, according to the number of regions N and the required control temperature, output the optimal air volume distribution scheme.

[0101] In an exemplary embodiment, taking the number of regions N = 10 and the required control temperature of 25℃ as an example, the thermoelectric cooler TEC temperature and the fan output air volume in each region are as follows:

[0102]

[0103] Step S4, output the PID control signal to the thermoelectric cooler for segmented adjustment of refrigeration power, while adjusting the speed of several independent fans according to the optimal air volume scheme.

[0104] It is worth mentioning that after the above-mentioned "real-time current monitoring → LSTM dynamic prediction → multi-mode actuator cooperation → regional air volume optimization" process is completed, real-time data after control is collected as feedback and added to step S1, thereby completing the calibration of the LSTM neural network model.

[0105] Through the above embodiments, the neural network model is used to predict the battery temperature for the next 3 seconds, combined with historical data and real-time data, which significantly improves the accuracy of temperature prediction, and automatically switches three working modes (emergency mode, economic mode, and balanced mode) according to the prediction error size, realizes dynamic response and resource optimization allocation, combined with the alternating direction multiplier method (ADMM) to calculate the optimal air volume distribution scheme, ensures that the temperature difference of each region does not exceed 0.5℃, avoids local overheating, improves battery consistency and life, and has excellent performance of fast dynamic response, high precision, strong robustness, excellent temperature uniformity and high energy efficiency ratio. The specific indicators are as follows:

[0106]

[0107] It should be understood that, although each step in the above flowchart is shown in sequence according to the direction of the arrow, these steps are not necessarily executed in the order indicated by the arrow. Unless explicitly stated herein, there is no strict order limitation for the execution of these steps, and these steps can be executed in other orders. Moreover, at least part of the steps in the above flowchart can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily sequential, but can be alternately executed with at least part of other steps or sub-steps or stages of other steps.

[0108] The following is an apparatus embodiment of the present application, which can be used to execute the temperature control method of the formation and batching system involved in the present application. For details not disclosed in the apparatus embodiment of the present application, please refer to the method embodiment of the temperature control method of the formation and batching system involved in the present application.

[0109] Please refer to Figure 5 In the embodiment of the present application, a temperature control device 50 of a formation and batching system is provided, which comprises a data marking module 501, a jamming marking module 502, a data storage module 503, and a data reading module 504.

[0110] The data acquisition module 501 is configured to acquire real-time data in the process of charging and discharging of the battery module.

[0111] The temperature prediction module 502 is configured to construct a dynamic thermal model based on a neural network, pre-process the real-time data as input, and predict a temperature prediction value after 3 seconds through the dynamic thermal model.

[0112] The scheme generation module 503 is configured to determine the current working mode of the formation and batching system based on the temperature prediction value, generate a PID control signal, and calculate an optimal air volume scheme by using an alternating direction multiplier method.

[0113] The temperature control module 504 is configured to output the PID control signal to a thermoelectric refrigerator for segmented adjustment of refrigeration power, and adjust the rotating speed of a plurality of independent air fans according to the optimal air volume scheme.

[0114] It should be noted that the temperature control device of the formation and component storage system provided in the above embodiment is only exemplified by the above division of the functional modules when performing temperature control, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the temperature control device of the formation and component storage system is divided into different functional modules to complete all or part of the functions described above. Each module can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so that the processor calls and executes the operations corresponding to each module.

[0115] In addition, the temperature control device of the formation and component storage system and the method embodiment of the temperature control method of the formation and component storage system provided in the above embodiment belong to the same concept, and the specific way in which each module performs operations has been described in detail in the method embodiment, which will not be described here.

[0116] Please refer to Figure 6 , an electronic device 4000 is provided in the embodiments of the present application.

[0117] In Figure 6 , the data interaction between the processor 4001 and the memory 4003 can be realized through at least one communication bus 4002. The communication bus 4002 can include a channel for transmitting data between the processor 4001 and the memory 4003. The communication bus 4002 can be a PCI (Peripheral Component Interconnect, Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture, Extended Industry Standard Architecture) bus, etc. The communication bus 4002 can be divided into an address bus, a data bus, a control bus, etc. For the sake of representation, Figure 6 In the above embodiment, only one thick line is used to represent the communication bus 4002, but it does not mean that there is only one bus or only one type of bus.

[0118] Optionally, the electronic device 4000 can also include a transceiver 4004, which can be used for data interaction between the electronic device and other electronic devices, such as data transmission and / or data reception, etc. It should be noted that the transceiver 4004 is not limited to one in actual application, and the structure of the electronic device 4000 does not constitute a limitation on the embodiments of the present application.

[0119] The processor 4001 can be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic device, a transistor logic device, a hardware component, or any combination thereof. It can implement or execute various exemplary logical blocks, modules and circuits described in conjunction with the disclosure. The processor 4001 can also be a combination of computing functions, such as one or more microprocessor combinations, combinations of DSP and microprocessor, etc. The memory 4003 can be a ROM (Read Only Memory) or other type of static storage device that can store static information and instructions, a RAM (Random Access Memory) or other type of dynamic storage device that can store information and instructions, an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, an optical disk storage (including a compact disk, a laser disk, an optical disk, a digital versatile disk, a Blu-ray disk, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium capable of carrying or storing desired program instructions or code in the form of instructions or data structures and accessible by the electronic device 400, but not limited to.

[0120] The memory 4003 stores program instructions or code, which can be read by the processor 4001 through the communication bus 4002.

[0121] The program instructions or code stored in the memory 4003 are executed by the processor 4001 to implement the temperature control method of the chemical composition system in the above embodiments.

[0122] In addition, the present application provides a storage medium, which stores program instructions or code, and the program instructions or code are loaded and executed by the processor to implement the temperature control method of the chemical composition system as described above.

[0123] The embodiment of the present application provides a computer program product, the computer program product includes program instructions or codes, the program instructions or codes are stored in a storage medium, a processor of an electronic device reads the program instructions or codes from the storage medium, loads and executes the program instructions or codes, so that the electronic device implements the temperature control method of the formation and component distribution system as described above.

[0124] The above is a specific description of the preferred embodiment of the application, but the application is not limited to the described embodiments. Those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the application. These equivalent modifications or replacements are all included in the scope defined by the claims of the present application.

Claims

1. A temperature control method for a composition and volumetric system, characterized in that, The method includes: Step S1: Obtain real-time data during the charging and discharging process of the battery module; Step S2: Construct a dynamic thermal model based on a neural network. After preprocessing the real-time data, use it as input to predict the temperature value 3 seconds later using the dynamic thermal model. Step S3: Based on the predicted temperature value, determine the current working mode of the formation and capacity system, generate a PID control signal, and simultaneously use the alternating direction multiplier method to calculate the optimal air volume scheme; Step S4: Output the PID control signal to the thermoelectric cooler to adjust the cooling power in segments, and at the same time adjust the speed of several independent fans according to the optimal air volume scheme.

2. The temperature control method for a composition and capacity system according to claim 1, characterized in that, Step S2 includes: Step S201: Using historical charge and discharge data as sample data, the sample data is divided into a training set, a validation set, and a test set in a ratio of 7:2:1; the historical charge and discharge data includes charge and discharge current I(t), ambient temperature Tenv(t), and battery SOC value; Step S202: Construct a 5-dimensional vector X(t) = [I(t),Tavg(t),dI / dt,SOC,Tenv(t)] based on the historical charge and discharge data in the training set. After normalizing the 5-dimensional vector X(t), use a 5-dimensional vector time series with a length of 10 seconds and a step size of 100ms as input; where Tavg(t) is the average temperature and dI / dt is the rate of change of current. Step S203: Construct an LSTM neural network model to obtain the temperature prediction value of the 5-dimensional vector time series in the training set 3 seconds later. Train and adjust the parameters of the LSTM neural network model according to the data in the validation set until the training termination condition is reached. Step S204: Use the trained LSTM neural network model to predict the temperature value 3 seconds from now.

3. The temperature control method for a composition and capacity system according to claim 2, characterized in that, The training termination conditions in step S203 include: Calculate the error |ΔT| between the predicted temperature value Tpred(t+3s) 3 seconds later and the actual measured value Treal(t+3s) in the test set; If the error |ΔT| > 0.5℃, online incremental learning is used to update the weights of the LSTM neural network model until the proportion of errors |ΔT| ≤ 0.5℃ is > 92%, thus completing the training of the LSTM neural network model.

4. The temperature control method for a composition and capacity system according to claim 2, characterized in that, The structure of the LSTM neural network model includes two stacked LSTM layers, each with 128 neurons; the two stacked LSTM layers are followed by a fully connected output layer; the LSTM layers use ReLU as the activation function, and the output layer uses a linear activation function; mean squared error is used as the loss function, and the Adam optimizer is used for training.

5. The temperature control method for a composition and capacity system according to claim 3, characterized in that, Step S3 includes: Step S301: When the error |ΔT|>=1℃, it is an emergency mode. All thermoelectric coolers (TECs) are started to the maximum cooling power. A PID control signal is generated according to the on / off state of the thermoelectric coolers (TECs). At the same time, all fans run to the highest speed. Step S302: When the error |ΔT| is between 0.3℃ and 1℃, it is in economic mode. Part or all of the thermoelectric coolers (TEC) are started. A PID control signal is generated according to the on / off state of the thermoelectric coolers (TEC). The fan runs at the base speed calculated by the alternating direction multiplier method. Step S303: When the error |ΔT| <= 0.3℃, it is in equilibrium mode. The thermoelectric cooler TEC is turned off. A PID control signal is generated according to the on / off state of the thermoelectric cooler TEC. The fan operates completely according to the optimal air volume distribution scheme (Qi) calculated by the alternating direction multiplier method.

6. The temperature control method for a composition and capacity system according to claim 5, characterized in that, The method further includes: adaptively adjusting the parameters of the PID dynamic optimization algorithm based on the current change rate dI / dt, including: Kp=Kp0*(1+α|dI / dt|); Ki=Ki0 / (1+β|dI / dt|); Kd=Kd0*e - γ|dI / dt|; where α=0.02, β=0.05, γ=0.03; when |dI / dt|>5A / s, switch to enhanced proportional mode (Kp increased by 20%); when |dI / dt|>10A / s, the differential term weights decay exponentially to suppress oscillations.

7. The temperature control method for a composition and capacity system according to claim 5, characterized in that, The optimal airflow calculation scheme using the alternating direction multiplier method includes: Step S401: Obtain the temperature T of each rack. i And the air volume u of each air conditioner vent i where i = 1, ..., N; Step S402: Minimize the sum of the squared deviations of the temperatures of all regions from the set values ​​according to the following objective function, while constraining the temperature difference between any two regions to not exceed 0.5℃: The control constraints are as follows: total air volume is constant (15 m3 / h), and the air volume range of a single area is 0.1-5.0 m3 / h; the temperature safety constraints are: 22℃≤Ti≤60℃; Tset is the required control temperature. Step S403: Initialize all temperatures T i 0 Initialize the airflow u based on the real-time temperature collected at that time. i 0 = 1 m³ / h; Step S404: Iterate using the alternating direction multiplier method until convergence: Partial update: Global update: Dual Update: Step S405: Based on the number of areas N and the required control temperature, output the optimal air volume distribution scheme.

8. A temperature control device for a chemical composition and capacity system, characterized in that, The device includes: The data acquisition module is used to acquire real-time data during the charging and discharging process of the battery module; The temperature prediction module is used to construct a dynamic thermal model based on a neural network. The real-time data is preprocessed and used as input to predict the temperature value 3 seconds later through the dynamic thermal model. The scheme generation module is used to determine the current working mode of the formation and capacity system based on the temperature prediction value, generate a PID control signal, and calculate the optimal air volume scheme using the alternating direction multiplier method. The temperature control module is used to output the PID control signal to the thermoelectric cooler to adjust the cooling power in stages, and at the same time adjust the speed of several independent fans according to the optimal air volume scheme.

9. An electronic device, characterized in that, include: At least one processor, at least one memory, wherein, The memory stores program instructions or code; The program instructions or code are loaded and executed by the processor, causing the electronic device to implement the temperature control method of the formation and capacity system as described in any one of claims 1 to 7.

10. A storage medium storing program instructions or code thereon, characterized in that, The program instructions or code are loaded and executed by the processor to implement the temperature control method for the decomposition and capacity system as described in any one of claims 1 to 7.

Images