Temperature control method for new energy automobile battery

Abstract

The invention provides a battery pack temperature equalization control method. According to the method, multi-zone temperature and heat dissipation distribution data acquisition is carried out, a temperature distribution model is established in combination with K-means clustering and a heat conduction equation, heat dissipation parameters are optimized by using a least square method, differential temperature control instructions are generated by using a PID control algorithm, and Q-learning reinforcement learning is introduced to dynamically adjust control parameters. According to the method, accurate monitoring and intelligent regulation and control of the internal temperature of the battery pack are achieved, the problems that a traditional temperature control system is lagged in response and poor in equalization effect are effectively solved, the temperature equalization performance of the battery pack is remarkably improved through multi-level data analysis and a self-adaptive control strategy, the service life of the battery is prolonged, and the service life of the battery pack is prolonged. And reliable guarantee is provided for safe and efficient operation of the new energy automobile power battery.

Description

Technical Field

[0001] The present invention belongs to the field of new energy batteries, and more specifically relates to a temperature control method for a new energy vehicle battery. Background Art

[0002] Temperature control for new energy vehicle batteries is a complex technical challenge. Temperature distribution within the battery pack often varies significantly, primarily due to factors such as varying heat generation and inconsistent heat dissipation conditions within the battery cells. For example, cells in the center of the pack typically experience higher temperatures than those at the edges due to poor heat dissipation, while cells in the lower portion of the pack experience lower temperatures due to the influence of cooling air from the underbody. This uneven temperature distribution, if not effectively controlled, can severely impact battery performance and lifespan. Traditional integrated temperature control solutions struggle to account for temperature differences across these areas, making precise temperature control impossible. Summary of the Invention

[0003] The present invention proposes a temperature control method for a new energy vehicle battery, which implements a technical solution of differentiated temperature control for different areas of a battery pack, thereby improving the adaptability and reliability of the battery temperature control system.

[0004] The technical solution of the present invention is achieved as follows: A temperature control method for a new energy vehicle battery, the method comprising: The first step is to collect temperature values ​​in the center, edge, lower cells, and upper areas of the battery pack. The heat dissipation wind speed distribution and coolant flow distribution are also recorded to generate a multi-region temperature and heat dissipation distribution dataset containing temperature, wind speed, and flow. The second step is to partition and classify the temperature values ​​based on the multi-region temperature and heat dissipation distribution dataset, extract the temperature statistical characteristics and heat dissipation wind speed distribution characteristics of the central area, edge area, lower battery cell, and upper area, and generate a partitioned temperature and wind speed feature dataset; The third step is to calculate the expected temperature distribution of each area based on the zoned temperature and wind speed characteristic datasets, using the pre-established cell heating model, inputting temperature statistical characteristics and wind speed distribution characteristics, and generating a theoretical temperature distribution dataset. The fourth step is to calculate the temperature deviation of each region based on the theoretical temperature distribution dataset and the multi-region temperature and heat dissipation distribution dataset. If the deviation exceeds the preset threshold, the cooling wind speed distribution, coolant flow distribution, and heat dissipation power distribution are adjusted using the least squares method. The iteration step size is set, and the deviation weight matrix is ​​defined based on the regional thermal sensitivity to generate an optimized heat dissipation parameter dataset. Step 5: Generate a differentiated temperature control instruction dataset based on the optimized heat dissipation parameter dataset; Step 6: Based on the differentiated temperature control instruction data set, the heat dissipation device is driven to operate. By collecting real-time temperature data, it is determined whether the temperatures of the center area, edge area, lower battery cell, and upper area have reached the preset equilibrium threshold, and a real-time temperature equilibrium status data set is generated. In the seventh step, based on the real-time temperature equilibrium state data set, the temperature deviation of each area is detected. If the deviation exceeds the preset threshold, the proportional-integral-derivative control parameters are adjusted through the Q-learning reinforcement learning algorithm to generate an updated temperature control parameter data set; Step 8: Based on the updated temperature control parameter dataset, the proportional-integral-derivative control parameter range is adjusted to regenerate a differentiated temperature control instruction dataset. The cooling wind speed distribution and coolant flow distribution in the center area, edge area, lower battery cells, and upper area are adjusted to generate a new real-time temperature equilibrium state dataset. In the ninth step, based on the new real-time temperature equilibrium status data set, combined with the ambient temperature and boundary heat dissipation conditions, three consecutive real-time temperature equilibrium status data sets are compared to determine whether the temperature distribution inside the battery pack is stable within the preset equilibrium range, and generate the final temperature control system operation status data set.

[0005] The beneficial effects of the present invention are as follows: through the acquisition of multi-region temperature and heat dissipation distribution data, a temperature distribution model is established in combination with K-means clustering and the heat conduction equation, the least squares method is used to optimize the heat dissipation parameters, a PID control algorithm is used to generate differentiated temperature control instructions, and Q-learning reinforcement learning is introduced to dynamically adjust the control parameters. This method realizes the precise monitoring and intelligent regulation of the internal temperature of the battery pack, effectively solving the problems of delayed response and poor balancing effect of traditional temperature control systems. Through multi-level data analysis and adaptive control strategies, the present invention significantly improves the temperature balancing performance of the battery pack, extends the battery life, and provides reliable protection for the safe and efficient operation of power batteries for new energy vehicles. DETAILED DESCRIPTION

[0006] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the embodiments of the present invention. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0007] A temperature control method for a new energy vehicle battery, the method comprising: The first step is to collect the temperature values ​​of the center area, edge area, lower cell, and upper area inside the battery pack, and record the heat dissipation wind speed distribution and coolant flow distribution at the same time to generate a multi-region temperature and heat dissipation distribution data set containing temperature, wind speed, and flow. Specifically, it includes: The sensor array collects the temperature of the central area, edge area, lower battery cell temperature, upper area temperature, cooling wind speed distribution, and coolant flow distribution to generate a multi-region data set. The K-means clustering algorithm is used to analyze the multi-region data set and determine the thermal state classification based on the temperature distribution characteristics, wind speed distribution characteristics, and flow distribution characteristics. If the thermal state classification exceeds the preset threshold, the multi-region data set is analyzed through the convolutional neural network model to determine the temperature distribution characteristics of the abnormal area, and the support vector machine algorithm is used to predict the changing trends of the wind speed distribution characteristics and the flow distribution characteristics. If the predicted wind speed distribution characteristic changing trend is lower than the preset threshold, the cooling wind speed distribution is adjusted by the linear interpolation method to obtain the optimized wind speed distribution parameters. If the predicted flow distribution characteristic changing trend is lower than the preset threshold, the coolant flow distribution is adjusted by the gradient descent method to obtain the optimized flow distribution parameters. The multi-region data set is updated through the optimized wind speed distribution parameters and flow distribution parameters to generate dynamic distribution data of the battery pack thermal status.

[0008] The second step is to partition and classify the temperature values ​​based on the multi-region temperature and heat dissipation distribution dataset, extract the temperature statistical characteristics and heat dissipation wind speed distribution characteristics of the center area, edge area, lower battery cell, and upper area, and generate a partition temperature and wind speed feature dataset. Specifically, it includes: The sensor array acquires the temperature of the center area, edge area, lower battery cell temperature, upper area, and cooling wind speed. The K-means clustering algorithm is used to partition and classify the temperature values, extract temperature statistical features and wind speed distribution features, and obtain partition temperature and wind speed feature datasets. Based on the partitioned temperature and wind speed characteristic data sets, the principal component analysis method is used to extract the main components of the temperature statistical characteristics and wind speed distribution characteristics to obtain the feature data set after dimensionality reduction. If the temperature statistical characteristics in the feature data set after dimensionality reduction exceed the preset threshold, the feature data set is analyzed using the random forest algorithm to determine the temperature distribution characteristics of the abnormal area and obtain the abnormal area distribution data. Based on the abnormal area distribution data, the linear interpolation method is used to adjust the heat dissipation wind speed distribution, and the optimized wind speed distribution parameters are obtained to update the partition temperature and wind speed feature data sets, generate an updated feature data set, and use the mean shift algorithm to re-partition and classify the temperature and wind speed values ​​to obtain the optimized partition classification results, and generate the partition temperature and wind speed distribution data of the battery pack.

[0009] The third step is to calculate the expected temperature distribution of each area based on the zone temperature and wind speed characteristic data set, through the pre-established battery cell heating model, input temperature statistical characteristics and wind speed distribution characteristics, and generate a theoretical temperature distribution data set. Specifically, it includes: Obtain temperature statistical features and wind speed distribution features from the partition temperature and wind speed feature dataset, input the pre-built calorific value model, calculate the unit volume calorific value Q, and obtain the unit calorific value dataset. Indicates the heat output per unit volume. represents the temperature value of the i-th measuring point, n represents the total number of temperature measuring points, α represents the temperature influence coefficient, represents the function of wind speed changing with time, β represents the wind speed influence coefficient, and t represents time; According to the unit heat data set, combined with the preset thermal diffusivity and boundary heat dissipation conditions, the heat conduction equation is used. T / t=γ( ²T / x²+ ²T / y²+ ²T / z²)+Q, where T represents temperature, t represents time, x, y, and z represent spatial coordinates, and γ represents the thermal diffusion coefficient. The theoretical temperature distribution of each region is calculated to obtain a theoretical temperature distribution data set; For the theoretical temperature distribution data set, if the partition temperature exceeds the preset threshold, the abnormal area is determined by comparing the theoretical temperature distribution with the actual partition temperature, and the abnormal area distribution data is obtained. The wind speed distribution characteristics are adjusted using the linear interpolation method to generate the optimized wind speed distribution parameters; The partitioned temperature and wind speed characteristic data sets are updated using the optimized wind speed distribution parameters to obtain an updated characteristic data set. The calorific value model is re-entered to calculate the adjusted calorific value per unit volume Q to obtain an adjusted calorific value data set. Combined with the thermal diffusion coefficient γ and boundary heat dissipation conditions, the heat conduction equation is used to recalculate the temperature distribution of each region to obtain the optimized temperature distribution data set.

[0010] The fourth step is to calculate the temperature deviation of each region based on the theoretical temperature distribution data set and the multi-region temperature and heat dissipation distribution data set. If the deviation exceeds the preset threshold, the least squares method is used to adjust the heat dissipation wind speed distribution, coolant flow distribution, and heat dissipation power distribution, set the iteration step size, and define the deviation weight matrix based on the regional thermal sensitivity to generate the optimized heat dissipation parameter data set. Specifically, it includes: Based on the theoretical temperature distribution dataset and the multi-region temperature and heat dissipation distribution dataset, the temperature deviation of each region is calculated. If the deviation exceeds the preset threshold, the cooling wind speed distribution, coolant flow distribution, and heat dissipation power distribution are adjusted using the least squares method. The iteration step size is set to 0.01, and the deviation weight matrix is ​​defined based on the regional thermal sensitivity. The optimized heat dissipation parameter dataset is generated, including: From the theoretical temperature distribution data set and the multi-region temperature and heat dissipation distribution data set, the temperature value of each region is obtained, and the temperature deviation is calculated by point-by-point comparison to obtain a temperature deviation data set; if the deviation value in the temperature deviation data set exceeds the preset threshold, a deviation weight matrix is ​​generated according to the regional thermal sensitivity to obtain a deviation weight data set. represents the mean temperature deviation, represents the temperature value of the i-th measurement point, represents the reference temperature value, n represents the number of measurement points, θ represents the preset threshold value, represents the regional thermal sensitivity weight matrix element, and Represents the spatial position coordinates, represents the weight coefficient, represents the thermal sensitivity attenuation parameter, represents the kth element in the bias weight dataset, represents the weight matrix element, represents the temperature deviation value, and m represents the number of regional grid divisions; The least squares method is used to optimize the heat dissipation wind speed distribution through the deviation weighted data set. The preset iteration step size is set and the adjusted wind speed parameters are calculated to obtain the optimized wind speed distribution data set. Combined with the deviation weighted data set, the least squares method is used to adjust the coolant flow distribution and the flow optimization parameters are calculated to obtain the optimized flow distribution data set. The least squares method is used to adjust the heat dissipation power distribution and the power optimization parameters are calculated to obtain the optimized power distribution data set. A comprehensive heat dissipation parameter dataset is generated by integrating the optimized wind speed distribution dataset, the optimized flow distribution dataset, and the optimized power distribution dataset to obtain an optimized heat dissipation parameter dataset, and a multi-region temperature and heat dissipation distribution dataset is updated to obtain an updated temperature and heat dissipation distribution dataset.

[0011] The fifth step is to generate a differentiated temperature control instruction dataset based on the optimized heat dissipation parameter dataset. This specifically includes: According to the optimized heat dissipation parameter data set and the differentiated temperature control instruction data set, the temperature sensor data of each area is obtained and the actual temperature distribution data set is calculated; if the deviation between the actual temperature distribution data set and the theoretical temperature distribution data set exceeds the preset threshold, the proportional-integral-differential control algorithm u(t) = Kpe(t) + Ki is used. e(t)dt + Kdde(t) / dt, where u(t) represents the control output, e(t) represents the deviation between the theoretical temperature distribution data set and the actual temperature, Kp represents the proportional coefficient, Ki represents the integral coefficient, and Kd represents the differential coefficient, to generate a differentiated temperature control instruction data set.

[0012] The sixth step is to drive the heat dissipation device to operate according to the differentiated temperature control instruction data set, and to collect real-time temperature data to determine whether the temperature of the center area, edge area, lower battery cell, and upper area reaches the preset equilibrium threshold, and generate a real-time temperature equilibrium status data set.

[0013] Based on the differentiated temperature control instruction data set, the heat dissipation device is driven to operate. The sensor collects real-time temperature data at a frequency of 1Hz to determine whether the temperature of the center area, edge area, lower battery cell, and upper area reaches the preset equilibrium threshold. The real-time temperature equilibrium status data set is generated, including: The sensor collects real-time temperature data of the center area, edge area, lower battery cell, and upper battery cell at a frequency of 1Hz. The heat dissipation device is driven to operate according to the differentiated temperature control instruction data set, and it is determined whether the temperature of each area meets the preset balance threshold to obtain a real-time temperature balance status data set; Based on the real-time temperature equilibrium status dataset, the deviation value between the temperature of each area and the preset equilibrium threshold is obtained. If the deviation value exceeds the preset deviation threshold, the k-nearest neighbor algorithm is used to calculate the temperature adjustment priority of each area to obtain the temperature adjustment priority dataset; Using the temperature adjustment priority dataset, linear interpolation is used to calculate the operating parameter adjustment values ​​of the heat sink for the center area and upper cells, resulting in an optimized operating parameter dataset for the center area and upper cells. Combined with the real-time temperature data for the edge area and lower cells, if the temperature of the edge area or lower cells falls below a preset lower threshold, the operating intensity of the heat sink in the corresponding area is reduced, resulting in an optimized operating intensity dataset for the edge area and lower cells. By integrating the optimized operating parameter data sets of the central area and the upper battery cells and the optimized operating intensity data sets of the edge area and the lower battery cells, a comprehensive optimized operating parameter data set is generated, the operating instructions of the heat dissipation device are updated, and an updated temperature control instruction data set is generated; The updated temperature control instruction data set is used to drive the heat dissipation device to operate, collect the real-time temperature data of the next cycle, and obtain the real-time temperature equilibrium state data set of the next cycle.

[0014] The seventh step is to detect the temperature deviation in each area based on the real-time temperature equilibrium state data set. If the deviation exceeds the preset threshold, the proportional-integral-derivative control parameters are adjusted using the Q-learning reinforcement learning algorithm to generate an updated temperature control parameter data set. This includes: Based on the real-time temperature equilibrium state dataset, the temperature deviation in each area is detected. If the deviation exceeds the preset threshold, the proportional-integral-derivative control parameters are adjusted using the Q-learning reinforcement learning algorithm. The state space is defined as the temperature deviation and wind speed, and the action space is the adjustment range of Kp, Ki, and Kd. The reward function is set to the negative value of the temperature deviation, the learning rate is 0.1, and the discount factor is 0.9. The update formula is Q(s,a) ← Q(s,a) + α[r + γmaxQ(s',a') - Q(s,a)], where Q(s,a) represents the state-action value function, s represents the current state, a represents the action, r represents the reward, α represents the learning rate, γ represents the discount factor, s' represents the next state, and a' represents the next action. This generates an updated temperature control parameter dataset.

[0015] The eighth step is to adjust the proportional-integral-derivative control parameter range based on the updated temperature control parameter data set, regenerate the differentiated temperature control instruction data set, adjust the heat dissipation wind speed distribution and coolant flow distribution in the center area, edge area, lower battery cell, and upper area, and generate a new real-time temperature equilibrium state data set. Specifically including: The updated temperature control parameter data set is used to adjust the boundary range of the proportional-integral-derivative control parameters to generate a control parameter data set with adjusted boundaries. Differentiated temperature control instruction data sets are generated for the center area, edge area, lower cell, and upper cell. By using differentiated temperature control instruction datasets, the cooling wind speed increments and coolant flow increments are allocated to the central area, edge area, lower battery cells, and upper battery cells to obtain a regional cooling allocation dataset. If the cooling wind speed increment or coolant flow increment in the regional cooling allocation dataset exceeds a preset threshold, the increment is smoothed using linear interpolation to obtain a smoothed cooling allocation dataset. The operating instructions of the cooling device and coolant pump are updated to generate an updated temperature control instruction dataset. The updated temperature control instruction data set is used to drive the heat dissipation device and the coolant pump to operate, collect the temperature data of the next cycle, and obtain the real-time temperature equilibrium state data set of the next cycle. Based on the real-time temperature equilibrium state data set of the next cycle, the temperature deviation of each area is calculated to generate a temperature deviation distribution data set.

[0016] The ninth step is to compare three consecutive real-time temperature equilibrium state data sets based on the new real-time temperature equilibrium state data set, combined with the ambient temperature and boundary heat dissipation conditions, to determine whether the temperature distribution inside the battery pack is stable within the preset equilibrium range, and generate the final temperature control system operation status data set. Specifically, it includes: The real-time temperature equilibrium state dataset is integrated with the ambient temperature and boundary heat dissipation conditions to calculate the temperature mean and variance of each area within the battery pack to obtain a temperature distribution feature dataset. If the variance in the temperature distribution feature dataset exceeds the preset equilibrium range, a sliding window method is used to perform time series analysis on three consecutive real-time temperature equilibrium state datasets to obtain a temperature change trend dataset. Based on the temperature change trend dataset, the fluctuation amplitude and change rate of the temperature in each area are calculated to generate a temperature fluctuation feature dataset. The K-means clustering algorithm is then used to classify the internal areas of the battery pack using the temperature fluctuation feature dataset to obtain a temperature distribution partition dataset. If the number of partitions in the temperature distribution partition dataset is less than the preset balanced partition threshold, linear interpolation is performed on the partition boundaries to generate an optimized temperature distribution partition dataset. According to the optimized temperature distribution partition data set, the temperature control system operating parameters are updated, the temperature control system operating status data set is generated, the heat dissipation device and the coolant pump are driven to adjust the operating mode, and the real-time temperature data of the next cycle is collected to obtain the updated real-time temperature balance status data set.

[0017] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A temperature control method for a new energy vehicle battery, characterized in that: The method comprises: The first step is to collect temperature values ​​in the center, edge, lower cells, and upper areas of the battery pack. The heat dissipation wind speed distribution and coolant flow distribution are also recorded to generate a multi-region temperature and heat dissipation distribution dataset containing temperature, wind speed, and flow. The second step is to partition and classify the temperature values ​​based on the multi-region temperature and heat dissipation distribution dataset, extract the temperature statistical characteristics and heat dissipation wind speed distribution characteristics of the central area, edge area, lower battery cell, and upper area, and generate a partitioned temperature and wind speed feature dataset; The third step is to calculate the expected temperature distribution of each area based on the zoned temperature and wind speed characteristic datasets, using the pre-established cell heating model, inputting temperature statistical characteristics and wind speed distribution characteristics, and generating a theoretical temperature distribution dataset. The fourth step is to calculate the temperature deviation of each region based on the theoretical temperature distribution dataset and the multi-region temperature and heat dissipation distribution dataset. If the deviation exceeds the preset threshold, the cooling wind speed distribution, coolant flow distribution, and heat dissipation power distribution are adjusted using the least squares method. The iteration step size is set, and the deviation weight matrix is ​​defined based on the regional thermal sensitivity to generate an optimized heat dissipation parameter dataset. Step 5: Generate a differentiated temperature control instruction dataset based on the optimized heat dissipation parameter dataset; Step 6: Based on the differentiated temperature control instruction data set, the heat dissipation device is driven to operate. By collecting real-time temperature data, it is determined whether the temperatures of the center area, edge area, lower battery cell, and upper area have reached the preset equilibrium threshold, and a real-time temperature equilibrium status data set is generated. In the seventh step, based on the real-time temperature equilibrium state data set, the temperature deviation of each area is detected. If the deviation exceeds the preset threshold, the proportional-integral-derivative control parameters are adjusted through the Q-learning reinforcement learning algorithm to generate an updated temperature control parameter data set; Step 8: Based on the updated temperature control parameter dataset, the proportional-integral-derivative control parameter range is adjusted to regenerate a differentiated temperature control instruction dataset. The cooling wind speed distribution and coolant flow distribution in the center area, edge area, lower battery cells, and upper area are adjusted to generate a new real-time temperature equilibrium state dataset. In the ninth step, based on the new real-time temperature equilibrium status data set, combined with the ambient temperature and boundary heat dissipation conditions, three consecutive real-time temperature equilibrium status data sets are compared to determine whether the temperature distribution inside the battery pack is stable within the preset equilibrium range, and generate the final temperature control system operation status data set.

2. The temperature control method of a new energy vehicle battery according to claim 1, characterized in that: The first step specifically includes: The sensor array collects the temperature of the central area, edge area, lower battery cell temperature, upper area temperature, cooling wind speed distribution, and coolant flow distribution to generate a multi-region data set. The K-means clustering algorithm is used to analyze the multi-region data set and determine the thermal state classification based on the temperature distribution characteristics, wind speed distribution characteristics, and flow distribution characteristics. If the thermal state classification exceeds the preset threshold, the multi-region data set is analyzed through the convolutional neural network model to determine the temperature distribution characteristics of the abnormal area, and the support vector machine algorithm is used to predict the changing trends of the wind speed distribution characteristics and the flow distribution characteristics. If the predicted wind speed distribution characteristic changing trend is lower than the preset threshold, the cooling wind speed distribution is adjusted by the linear interpolation method to obtain the optimized wind speed distribution parameters. If the predicted flow distribution characteristic changing trend is lower than the preset threshold, the coolant flow distribution is adjusted by the gradient descent method to obtain the optimized flow distribution parameters. The multi-region data set is updated through the optimized wind speed distribution parameters and flow distribution parameters to generate dynamic distribution data of the battery pack thermal status.

3. The temperature control method of a new energy vehicle battery according to claim 1, characterized in that: The second step specifically includes: The sensor array acquires the temperature of the center area, edge area, lower battery cell temperature, upper area, and cooling wind speed. The K-means clustering algorithm is used to partition and classify the temperature values, extract temperature statistical features and wind speed distribution features, and obtain partition temperature and wind speed feature datasets. Based on the partitioned temperature and wind speed characteristic data sets, the principal component analysis method is used to extract the main components of the temperature statistical characteristics and wind speed distribution characteristics to obtain the feature data set after dimensionality reduction. If the temperature statistical characteristics in the feature data set after dimensionality reduction exceed the preset threshold, the feature data set is analyzed using the random forest algorithm to determine the temperature distribution characteristics of the abnormal area and obtain the abnormal area distribution data. Based on the abnormal area distribution data, the linear interpolation method is used to adjust the heat dissipation wind speed distribution, and the optimized wind speed distribution parameters are obtained to update the partition temperature and wind speed feature data sets, generate an updated feature data set, and use the mean shift algorithm to re-partition and classify the temperature and wind speed values ​​to obtain the optimized partition classification results, and generate the partition temperature and wind speed distribution data of the battery pack.

4. The temperature control method of a new energy vehicle battery according to claim 1, characterized in that: The third step specifically includes: Obtain temperature statistical features and wind speed distribution features from the partition temperature and wind speed feature dataset, input the pre-built calorific value model, calculate the unit volume calorific value Q, and obtain the unit calorific value dataset. Indicates the heat output per unit volume. represents the temperature value of the i-th measuring point, n represents the total number of temperature measuring points, α represents the temperature influence coefficient, represents the function of wind speed changing with time, β represents the wind speed influence coefficient, and t represents time; According to the unit heat data set, combined with the preset thermal diffusivity and boundary heat dissipation conditions, the heat conduction equation is used. T / t=γ( ²T / x²+ ²T / y²+ ²T / z²)+Q, where T represents temperature, t represents time, x, y, and z represent spatial coordinates, and γ represents the thermal diffusion coefficient. The theoretical temperature distribution of each region is calculated to obtain a theoretical temperature distribution data set; For the theoretical temperature distribution data set, if the partition temperature exceeds the preset threshold, the abnormal area is determined by comparing the theoretical temperature distribution with the actual partition temperature, and the abnormal area distribution data is obtained. The wind speed distribution characteristics are adjusted using the linear interpolation method to generate the optimized wind speed distribution parameters; The partitioned temperature and wind speed characteristic data sets are updated using the optimized wind speed distribution parameters to obtain an updated characteristic data set. The calorific value model is re-entered to calculate the adjusted calorific value per unit volume Q to obtain an adjusted calorific value data set. Combined with the thermal diffusion coefficient γ and boundary heat dissipation conditions, the heat conduction equation is used to recalculate the temperature distribution of each region to obtain the optimized temperature distribution data set.

5. The temperature control method of a new energy vehicle battery according to claim 1, characterized in that: The fourth step specifically includes: Based on the theoretical temperature distribution dataset and the multi-region temperature and heat dissipation distribution dataset, the temperature deviation of each region is calculated. If the deviation exceeds the preset threshold, the cooling wind speed distribution, coolant flow distribution, and heat dissipation power distribution are adjusted using the least squares method. The iteration step size is set to 0.01, and the deviation weight matrix is ​​defined based on the regional thermal sensitivity. The optimized heat dissipation parameter dataset is generated, including: From the theoretical temperature distribution data set and the multi-region temperature and heat dissipation distribution data set, the temperature value of each region is obtained, and the temperature deviation is calculated by point-by-point comparison to obtain a temperature deviation data set; if the deviation value in the temperature deviation data set exceeds the preset threshold, a deviation weight matrix is ​​generated according to the regional thermal sensitivity to obtain a deviation weight data set. represents the mean temperature deviation, represents the temperature value of the i-th measurement point, represents the reference temperature value, n represents the number of measurement points, θ represents the preset threshold value, represents the regional thermal sensitivity weight matrix element, and Represents the spatial position coordinates, represents the weight coefficient, represents the thermal sensitivity attenuation parameter, represents the kth element in the bias weight dataset, represents the weight matrix element, represents the temperature deviation value, and m represents the number of regional grid divisions; The least squares method is used to optimize the heat dissipation wind speed distribution through the deviation weighted data set. The preset iteration step size is set and the adjusted wind speed parameters are calculated to obtain the optimized wind speed distribution data set. Combined with the deviation weighted data set, the least squares method is used to adjust the coolant flow distribution and the flow optimization parameters are calculated to obtain the optimized flow distribution data set. The least squares method is used to adjust the heat dissipation power distribution and the power optimization parameters are calculated to obtain the optimized power distribution data set. A comprehensive heat dissipation parameter dataset is generated by integrating the optimized wind speed distribution dataset, the optimized flow distribution dataset, and the optimized power distribution dataset to obtain an optimized heat dissipation parameter dataset, and a multi-region temperature and heat dissipation distribution dataset is updated to obtain an updated temperature and heat dissipation distribution dataset.

6. The temperature control method of a new energy vehicle battery according to claim 1, characterized in that: The fifth step specifically includes: According to the optimized heat dissipation parameter data set and the differentiated temperature control instruction data set, the temperature sensor data of each area is obtained and the actual temperature distribution data set is calculated; if the deviation between the actual temperature distribution data set and the theoretical temperature distribution data set exceeds the preset threshold, the proportional-integral-differential control algorithm u(t) = Kpe(t) + Ki is used. e(t)dt + Kdde(t) / dt, where u(t) represents the control output, e(t) represents the deviation between the theoretical temperature distribution data set and the actual temperature, Kp represents the proportional coefficient, Ki represents the integral coefficient, and Kd represents the differential coefficient, to generate a differentiated temperature control instruction data set.

7. The temperature control method of a new energy vehicle battery according to claim 1, characterized in that: The sixth step specifically includes: Based on the differentiated temperature control instruction data set, the heat dissipation device is driven to operate. The sensor collects real-time temperature data at a frequency of 1Hz to determine whether the temperature of the center area, edge area, lower battery cell, and upper area reaches the preset equilibrium threshold. The real-time temperature equilibrium status data set is generated, including: The sensor collects real-time temperature data of the center area, edge area, lower battery cell, and upper battery cell at a frequency of 1Hz. The heat dissipation device is driven to operate according to the differentiated temperature control instruction data set, and it is determined whether the temperature of each area meets the preset balance threshold to obtain a real-time temperature balance status data set; Based on the real-time temperature equilibrium status dataset, the deviation value between the temperature of each area and the preset equilibrium threshold is obtained. If the deviation value exceeds the preset deviation threshold, the k-nearest neighbor algorithm is used to calculate the temperature adjustment priority of each area to obtain the temperature adjustment priority dataset; Using the temperature adjustment priority dataset, linear interpolation is used to calculate the operating parameter adjustment values ​​of the heat sink for the center area and upper cells, resulting in an optimized operating parameter dataset for the center area and upper cells. Combined with the real-time temperature data for the edge area and lower cells, if the temperature of the edge area or lower cells falls below a preset lower threshold, the operating intensity of the heat sink in the corresponding area is reduced, resulting in an optimized operating intensity dataset for the edge area and lower cells. By integrating the optimized operating parameter data sets of the central area and the upper battery cells and the optimized operating intensity data sets of the edge area and the lower battery cells, a comprehensive optimized operating parameter data set is generated, the operating instructions of the heat dissipation device are updated, and an updated temperature control instruction data set is generated; The updated temperature control instruction data set is used to drive the heat dissipation device to operate, collect the real-time temperature data of the next cycle, and obtain the real-time temperature equilibrium state data set of the next cycle.

8. The temperature control method of a new energy vehicle battery according to claim 1, characterized in that: The seventh step specifically includes: detecting the temperature deviation of each area according to the real-time temperature equilibrium state dataset; if the deviation exceeds the preset threshold, adjusting the proportional-integral-differential control parameters through the Q-learning reinforcement learning algorithm; defining the state space as the temperature deviation and wind speed, the action space as the adjustment range of Kp, Ki, and Kd, setting the reward function to the negative value of the temperature deviation, the learning rate to 0.1, the discount factor to 0.9, and the update formula to Q(s,a) ← Q(s,a) + α[r + γmaxQ(s',a') - Q(s,a)], where Q(s,a) represents the state-action value function, s represents the current state, a represents the action, r represents the reward, α represents the learning rate, γ represents the discount factor, s' represents the next state, and a' represents the next action, thereby generating an updated temperature control parameter dataset.

9. The temperature control method of a new energy vehicle battery according to claim 1, characterized in that: The eighth step specifically includes: The updated temperature control parameter data set is used to adjust the boundary range of the proportional-integral-derivative control parameters to generate a control parameter data set with adjusted boundaries. Differentiated temperature control instruction data sets are generated for the center area, edge area, lower cell, and upper cell. By using differentiated temperature control instruction datasets, the cooling wind speed increments and coolant flow increments are allocated to the central area, edge area, lower battery cells, and upper battery cells to obtain a regional cooling allocation dataset. If the cooling wind speed increment or coolant flow increment in the regional cooling allocation dataset exceeds a preset threshold, the increment is smoothed using linear interpolation to obtain a smoothed cooling allocation dataset. The operating instructions of the cooling device and coolant pump are updated to generate an updated temperature control instruction dataset. The updated temperature control instruction data set is used to drive the heat dissipation device and the coolant pump to operate, collect the temperature data of the next cycle, and obtain the real-time temperature equilibrium state data set of the next cycle. Based on the real-time temperature equilibrium state data set of the next cycle, the temperature deviation of each area is calculated to generate a temperature deviation distribution data set.

10. The temperature control method of a new energy vehicle battery according to claim 1, characterized in that: The ninth step specifically includes: The real-time temperature equilibrium state dataset is integrated with the ambient temperature and boundary heat dissipation conditions to calculate the temperature mean and variance of each area within the battery pack to obtain a temperature distribution feature dataset. If the variance in the temperature distribution feature dataset exceeds the preset equilibrium range, a sliding window method is used to perform time series analysis on three consecutive real-time temperature equilibrium state datasets to obtain a temperature change trend dataset. Based on the temperature change trend dataset, the fluctuation amplitude and change rate of the temperature in each area are calculated to generate a temperature fluctuation feature dataset. The K-means clustering algorithm is then used to classify the internal areas of the battery pack using the temperature fluctuation feature dataset to obtain a temperature distribution partition dataset. If the number of partitions in the temperature distribution partition dataset is less than the preset balanced partition threshold, linear interpolation is performed on the partition boundaries to generate an optimized temperature distribution partition dataset. According to the optimized temperature distribution partition data set, the temperature control system operating parameters are updated, the temperature control system operating status data set is generated, the heat dissipation device and the coolant pump are driven to adjust the operating mode, and the real-time temperature data of the next cycle is collected to obtain the updated real-time temperature balance status data set.

Images