A new energy automobile battery pack temperature control system

Abstract

This invention relates to the field of battery pack temperature control technology, specifically a temperature control system for a new energy vehicle battery pack. The system includes a state acquisition module, a thermal state assessment module, a control sequence generation module, and a control decision module. The state acquisition module acquires the battery pack temperature vector, coolant flow rate, and battery pack current parameters, and calculates the predicted temperature vector. The thermal state assessment module calculates thermal evolution intensity metrics and model mismatch metrics. The control sequence generation module generates an extended coolant flow rate sequence and a re-optimized coolant flow rate sequence. The control decision module determines the target coolant flow rate and outputs it to the liquid cooling execution unit. This invention improves the accuracy of battery pack temperature control and reduces the computational complexity of control calculations.

Description

Technical Field

[0001] This invention relates to the field of battery pack temperature control technology, specifically a temperature control system for new energy vehicle battery packs. Background Technology

[0002] During the charging and discharging process, the power batteries of new energy vehicles continuously generate heat. If the battery temperature is too high or the temperature distribution between batteries is uneven, it can easily lead to accelerated battery performance degradation, reduced cycle life, and in severe cases, even safety risks. Therefore, new energy vehicles are usually equipped with a battery thermal management system, which regulates the battery pack temperature through a liquid cooling circuit to keep the battery within a suitable operating temperature range.

[0003] Existing technologies typically achieve temperature regulation by collecting battery temperature information and controlling coolant flow. Common methods include threshold-based rule control, proportional-integral-derivative (PID) control, and predictive control based on thermal models. Among these, rule-based control is simple in structure but has limited control accuracy, PID struggles to simultaneously ensure battery temperature stability and temperature consistency, while predictive control based on thermal models can improve control accuracy to some extent, but relies on relatively complex thermal model calculations.

[0004] In actual vehicle operation, the thermal characteristics of the battery pack will change with the ambient temperature, the degree of battery aging and the operating conditions. The pre-established thermal model is prone to prediction bias. Summary of the Invention

[0005] The purpose of this invention is to provide a temperature control system for new energy vehicle battery packs, in order to solve the technical problem that existing battery temperature control methods based on thermal models are prone to prediction deviations when the battery thermal characteristics change with operating conditions, thereby affecting the accuracy of temperature control.

[0006] To achieve the above objectives, the present invention provides a temperature control system for a new energy vehicle battery pack, the system comprising: The status acquisition module is used to acquire the battery pack temperature vector, coolant flow rate, and battery pack current parameters at the control moment; and calculate the corresponding predicted temperature vector based on the battery pack temperature vector, coolant flow rate, and battery pack current parameters through a pre-established discretized thermal resistance network model.

[0007] The thermal state assessment module is used to calculate the thermal evolution intensity metric based on the battery pack temperature vector, coolant flow rate, and battery pack temperature vector and coolant flow rate at historical control times; and to calculate the model mismatch metric based on the battery pack temperature vector, predicted temperature vector, and thermal evolution intensity metric.

[0008] The control sequence generation module is used to obtain the coolant flow rate sequence calculated at the previous control time; perform time-shift extension on the coolant flow rate sequence to obtain the extended coolant flow rate sequence; recursively correct the model parameters of the discretized thermal resistance network model according to the model mismatch metric to obtain the updated discretized thermal resistance network model; and perform constraint solving based on the updated discretized thermal resistance network model to obtain the re-optimized coolant flow rate sequence.

[0009] The control decision module is used to input the extended coolant flow rate sequence and the re-optimized coolant flow rate sequence into the updated discretized thermal resistance network model to predict the temperature, thereby obtaining the extended temperature trajectory and the re-optimized temperature trajectory; calculate the extended comprehensive performance index and the re-optimized comprehensive performance index based on the extended temperature trajectory, the re-optimized temperature trajectory, and the thermal evolution intensity measure; determine the target coolant flow rate based on the extended comprehensive performance index and the re-optimized comprehensive performance index, and output it to the liquid cooling execution unit.

[0010] Furthermore, the system also includes: The model building module is used to divide the battery pack into multiple hot nodes and establish corresponding cooling channel hot nodes; establish thermal resistance between each battery hot node; establish convective thermal resistance between the battery hot nodes and the cooling channel hot nodes; and set heat capacity parameters for each battery hot node.

[0011] Based on the preset battery structure parameters and cooling channel heat transfer parameters, the corresponding thermal resistance, convection resistance and heat capacity parameters are determined, and a discrete thermal resistance network model of the battery pack is constructed.

[0012] Furthermore, the thermal state assessment module includes: The thermal evolution intensity measurement calculation unit is used to obtain the current control time. Battery pack temperature vector Previous control time Battery pack temperature vector First two control moments Battery pack temperature vector Coolant flow rate at the current control moment Coolant flow rate at the previous control moment .

[0013] The acceleration vector of temperature change was calculated. ;in To control the cycle; the flow regulation rate is calculated. .

[0014] Based on the temperature change acceleration vector Calculate the scalar of average temperature acceleration and temperature distribution acceleration variance .

[0015] in This represents the number of temperature measurement points for the battery pack. , The acceleration vector of the temperature change The Each component.

[0016] According to the average temperature acceleration scalar Temperature distribution acceleration variance and flow regulation rate The thermal evolution intensity metric is calculated using a preset nonlinear mapping function. .

[0017] Furthermore, the thermal state assessment module includes: The model mismatch metric calculation unit is used to obtain the measured battery pack temperature vector and the predicted battery pack temperature vector at each historical control moment within a preset sliding time window; and to calculate the difference between the measured battery pack temperature vector and the predicted battery pack temperature vector at each historical control moment within the sliding time window to construct the corresponding instantaneous mismatch vector. The number of backtracking steps , The length of the preset sliding time window.

[0018] According to the thermal evolution intensity measurement Calculate the dynamic forgetting factor ;in This is a reference value for thermal evolution intensity. This is the preset time decay constant.

[0019] According to the dynamic forgetting factor For each instantaneous mismatch vector within the preset sliding time window Weighted fusion is performed to obtain the weighted mismatch baseline vector. ; for the weighted mismatch reference vector The model mismatch metric is obtained by performing L2 norm calculation. .

[0020] Furthermore, the control sequence generation module includes: The model update unit is used to measure the model mismatch at the current control moment. Model mismatch measurement with the previous control time step Calculate the rate of change of model mismatch .

[0021] when and At that time, the thermal resistance parameters and thermal capacity parameters in the discretized thermal resistance network model are recursively updated to obtain the updated discretized thermal resistance network model.

[0022] when and When the prediction bias compensation term of the discretized thermal resistance network model is updated, the updated discretized thermal resistance network model is obtained.

[0023] when At the same time, the model parameters of the discretized thermal resistance network model remain unchanged; where To preset the mismatch measurement threshold, This is a preset threshold for the rate of change of mismatch.

[0024] The coolant flow optimization unit is used to construct a temperature tracking error function based on the updated discretized thermal resistance network model; and to solve the temperature tracking error function under the constraints of coolant flow upper and lower limits and coolant flow rate change to obtain a re-optimized coolant flow sequence.

[0025] Furthermore, the model update unit includes: The parameter recursive update unit is used to update the parameters according to the preset long-term memory correction factor. The updated discretized thermal resistance network model is obtained by recursively updating the thermal resistance parameter R and the heat capacity parameter C of the discretized thermal resistance network model; wherein, the new thermal resistance parameter in the updated discretized thermal resistance network model is... New heat capacity parameters in the updated discretized thermal resistance network model ;in and This is the preset sensitivity coefficient.

[0026] Furthermore, the coolant flow optimization unit includes: The error function construction unit is used to predict the battery pack temperature in the prediction time domain based on the updated discretized thermal resistance network model, and obtain the predicted temperature sequence. A temperature tracking error function is constructed based on the predicted temperature sequence. The temperature tracking error function is as follows: .

[0027] in, The target temperature for the battery pack; To predict the first in the time domain The battery pack temperature vector of the step; To predict the coolant flow rate in the time domain; N is the length of the prediction time domain; and These are the weighting coefficients; This represents the variance of the battery pack's temperature distribution.

[0028] Furthermore, the control decision module includes: The extended comprehensive performance index calculation unit is used to calculate the temperature deviation sequence and temperature distribution variance sequence of the battery pack in the predicted time domain based on the extended temperature trajectory; calculate the temperature deviation index based on the temperature deviation sequence; calculate the temperature uniformity index based on the temperature distribution variance sequence; calculate the coolant flow rate change sequence based on the extended coolant flow rate change sequence; calculate the control energy consumption index based on the coolant flow rate change sequence; calculate the index weighting coefficient based on the thermal evolution intensity measure; and perform a weighted combination calculation of the temperature deviation index, temperature uniformity index, and control energy consumption index based on the index weighting coefficient to obtain the extended comprehensive performance index.

[0029] The re-optimized comprehensive performance index calculation unit is used to calculate the temperature deviation sequence and temperature distribution variance sequence based on the re-optimized temperature trajectory; calculate the temperature deviation index based on the temperature deviation sequence; calculate the temperature uniformity index based on the temperature distribution variance sequence; calculate the control energy consumption index based on the re-optimized coolant flow sequence; and perform weighted combination calculation of the temperature deviation index, temperature uniformity index, and control energy consumption index based on the index weighting coefficients to obtain the re-optimized comprehensive performance index.

[0030] Compared with the prior art, the beneficial effects of the present invention are: 1. By constructing a discretized thermal resistance network model of the battery pack, the battery pack temperature vector and predicted temperature vector are obtained. The thermal state is evaluated by combining historical control time data. Based on this, a coolant flow control sequence is generated and control decisions are made to achieve predictive control of battery pack temperature changes, thereby improving the accuracy and stability of battery pack temperature control.

[0031] 2. By calculating the battery pack temperature vector and historical control time data, a thermal evolution intensity metric is obtained. Combined with the predicted temperature vector, a model mismatch metric is constructed. When the model mismatch metric reaches a preset condition, the discretized thermal resistance network model is recursively updated, enabling the thermal model to adaptively correct itself as the thermal characteristics of the battery pack change, thereby reducing model prediction bias and improving the reliability of temperature prediction.

[0032] 3. By time-shifting and extending the coolant flow rate sequence of the previous control moment, the extended coolant flow rate sequence is obtained. After temperature prediction is performed on the extended coolant flow rate sequence and the optimized coolant flow rate sequence, the comprehensive performance index is compared to determine the target coolant flow rate. Under the premise of ensuring temperature control performance, the scale of real-time optimization calculation is reduced, thereby reducing the control calculation complexity and improving the real-time performance of the control algorithm. Attached Figure Description

[0033] Figure 1 This is a block diagram of a temperature control system for a new energy vehicle battery pack according to the present invention. Detailed Implementation

[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] Before providing examples, it is necessary to describe the application scenarios of the present invention. The present invention is applicable to the liquid cooling and thermal management control scenario of power batteries in new energy vehicles. Under the operating conditions where the battery management unit periodically collects battery pack temperature information, battery pack current parameters, and coolant flow information, the coolant flow is controlled to achieve battery pack temperature regulation.

[0036] Example 1: As Figure 1 As shown in the figure, this embodiment provides a temperature control system for a new energy vehicle battery pack, the system comprising: A status acquisition module is used to acquire the battery pack temperature vector, coolant flow rate, and battery pack current parameters at control moments; and to calculate the corresponding predicted temperature vector based on the battery pack temperature vector, coolant flow rate, and battery pack current parameters using a pre-established discretized thermal resistance network model; the system also includes: The model building module is used to divide the battery pack into multiple hot nodes and establish corresponding cooling channel hot nodes; establish thermal resistance between each battery hot node; establish convective thermal resistance between the battery hot nodes and the cooling channel hot nodes; and set heat capacity parameters for each battery hot node.

[0037] Based on the preset battery structure parameters and cooling channel heat transfer parameters, the corresponding thermal resistance, convection resistance and heat capacity parameters are determined, and a discrete thermal resistance network model of the battery pack is constructed.

[0038] For example, taking a new energy vehicle battery pack consisting of 12 battery modules as an example, each battery module contains 16 individual battery cells. An aluminum alloy cooling plate is arranged inside the battery pack, and a serpentine cooling channel is formed inside the cooling plate. The cooling medium is a 50% (by volume) ethylene glycol aqueous solution. The battery modules are fixed together by a metal connection structure, and the cooling plate is in close contact with the bottom of the battery module.

[0039] Each battery module has four temperature sensor mounting locations, corresponding to four battery thermal nodes, resulting in 48 thermal nodes for 12 battery modules. The cooling plate is divided into 12 cooling channel thermal nodes according to the coolant flow direction, with each cooling channel thermal node corresponding to one battery module location. Conductive thermal resistance is established between the battery thermal nodes, with the heat conduction path including the cell housing, module structural components, and connecting plates between modules. Convective thermal resistance is established between the battery thermal nodes and their corresponding cooling channel thermal nodes, with convective heat transfer formed by the flow of coolant in the cooling channels. Battery structural parameters are determined based on cell parameters and battery module structural parameters. The mass of a single battery cell is... Specific heat capacity The thermal capacity of the battery module is obtained by summing the values ​​of the battery cells. The thermal conductivity of the cell casing is taken as... The thermal conductivity of the module structural components is taken as The length of the heat conduction path and the cross-sectional area of ​​the heat conduction are determined based on the battery structure dimensions, using the thermal resistance calculation formula. Calculate the thermal resistance between battery hot nodes, where, Indicates thermal resistance. Indicates the length of the heat conduction path. Indicates thermal conductivity. This represents the thermal conductivity cross-sectional area. The heat transfer parameters of the cooling channel are determined based on the coolant flow rate and channel dimensions; the coolant inlet flow rate is taken as... The hydraulic diameter of the cooling channel is taken as The convective heat transfer coefficient is taken as Through heat transfer relationship Calculate the convective thermal resistance between the battery hot node and the cooling channel hot node, where, Indicates convective thermal resistance. Indicates the convective heat transfer coefficient. This indicates the heat exchange area. The heat capacity of a single battery cell is... The battery module contains 16 cells, and the battery module's heat capacity is... The battery module's heat capacity is evenly distributed among the four corresponding battery thermal nodes, with each battery thermal node having a heat capacity of [missing information]. Substituting the parameters of thermal conductivity, convection, and thermal capacity into the thermal resistance network structure forms a discretized thermal resistance network model.

[0040] Control cycle set to The control unit reads temperature sensor data from the battery management unit, and the battery pack temperature vector includes... One temperature component. Coolant flow rate is acquired by a flow sensor in the cooling circuit, with a sampling resolution of [missing information]. The battery pack current parameters are obtained through the battery management unit's current sampling module, and the current sensor's range is [missing information]. The sampling frequency is In this example, the coolant flow rate measured at the control time is... The battery pack current is .

[0041] The discretized thermal resistance network model uses nodal temperature state equations for prediction. The temperature variation of battery hot nodes is jointly determined by the node heat capacity, the thermal conductivity between adjacent battery hot nodes, and the convective thermal resistance between battery hot nodes and cooling channel hot nodes. The battery heating power is calculated using battery pack current parameters and added to the discretized thermal resistance network model through an equivalent heat source term. The predicted temperature vector is obtained by substituting the battery pack temperature vector, coolant flow rate, and battery pack current parameters into the discretized thermal resistance network model to update the nodal temperature values.

[0042] A thermal state assessment module is used to calculate a thermal evolution intensity metric based on the battery pack temperature vector, coolant flow rate, and the battery pack temperature vector and coolant flow rate at historical control times; and to calculate a model mismatch metric based on the battery pack temperature vector, predicted temperature vector, and thermal evolution intensity metric. The thermal state assessment module includes: The thermal evolution intensity measurement calculation unit is used to obtain the current control time. Battery pack temperature vector Previous control time Battery pack temperature vector First two control moments Battery pack temperature vector Coolant flow rate at the current control moment Coolant flow rate at the previous control moment .

[0043] The acceleration vector of temperature change was calculated. ;in To control the cycle; the flow regulation rate is calculated. .

[0044] Based on the temperature change acceleration vector Calculate the scalar of average temperature acceleration and temperature distribution acceleration variance .

[0045] in This represents the number of temperature measurement points for the battery pack. , The acceleration vector of the temperature change The Each component.

[0046] According to the average temperature acceleration scalar Temperature distribution acceleration variance and flow regulation rate The thermal evolution intensity metric is calculated using a preset nonlinear mapping function. .

[0047] The thermal state assessment module includes: The model mismatch metric calculation unit is used to obtain the measured battery pack temperature vector and the predicted battery pack temperature vector at each historical control moment within a preset sliding time window; and to calculate the difference between the measured battery pack temperature vector and the predicted battery pack temperature vector at each historical control moment within the sliding time window to construct the corresponding instantaneous mismatch vector. The number of backtracking steps , The length of the preset sliding time window.

[0048] According to the thermal evolution intensity measurement Calculate the dynamic forgetting factor ;in This is a reference value for thermal evolution intensity. This is the preset time decay constant.

[0049] According to the dynamic forgetting factor For each instantaneous mismatch vector within the preset sliding time window Weighted fusion is performed to obtain the weighted mismatch baseline vector. ; for the weighted mismatch reference vector The model mismatch metric is obtained by performing L2 norm calculation. .

[0050] For example, in a control cycle of Under operating conditions, the battery pack temperature vector and coolant flow rate data are continuously recorded. Control time. The battery pack temperature vector is The battery pack temperature vector contains 48 temperature components, corresponding to the temperatures of 48 battery thermal nodes. The coolant flow rate is acquired by a cooling loop flow sensor; the current coolant flow rate at the control moment is... The coolant flow rate at the previous control moment was .

[0051] The temperatures of the second battery thermal node at the three control times were 32.5℃, 32.3℃, and 32.2℃, respectively. Substituting these values ​​into... The calculation formula yields the acceleration component of temperature change. The temperature change acceleration vectors of 48 battery thermal nodes were calculated in the same manner. .

[0052] The change in coolant flow rate over two consecutive control cycles is passed through The calculation formula is used to calculate the result; in this example, substituting the data yields the result. The acceleration vector A(k) for temperature change contains 48 components, and the number of temperature measurement points is denoted as... The 48 temperature change acceleration components were substituted into the calculation to obtain... , The average temperature acceleration scalar, the variance of temperature distribution acceleration, and the flow rate regulation rate are substituted into a nonlinear mapping function to calculate the thermal evolution intensity measure. The nonlinear mapping function is... Substitute the data to calculate The coefficients in the nonlinear mapping function are obtained by parameter identification from historical data under different operating conditions.

[0053] The sliding time window length is A sliding time window records the measured and predicted battery pack temperature vectors for the most recent five control moments. Control moments The instantaneous mismatch vector is obtained by reading the measured temperature vector and the predicted temperature vector of the battery pack and calculating the difference. Previous control moment Read the measured temperature vector and the predicted temperature vector of the battery pack and calculate the difference to obtain Calculated in the same way , and .

[0054] The reference value for thermal evolution intensity is obtained based on the thermal response characteristics of the battery pack. The time decay constant is . Substituting into the formula for calculating the dynamic forgetting factor, we obtain... , , , , .

[0055] The dynamic forgetting factor and the corresponding instantaneous mismatch vector are weighted and fused, and the weighted mismatch reference vector is calculated using the formula for calculating the weighted mismatch reference vector. The weighted mismatch baseline vector contains 48 temperature components, each corresponding to a prediction error for a battery thermal node. The weighted mismatch baseline vector... The model mismatch metric is obtained by performing L2 norm calculations and then taking the square root of the sum of the squares of the 48 temperature components. .

[0056] A control sequence generation module is used to obtain the coolant flow rate sequence calculated at the previous control time; perform time-shift extension on the coolant flow rate sequence to obtain an extended coolant flow rate sequence; recursively correct the model parameters of the discretized thermal resistance network model according to the model mismatch metric to obtain an updated discretized thermal resistance network model; and perform constraint solving based on the updated discretized thermal resistance network model to obtain a re-optimized coolant flow rate sequence; the control sequence generation module includes: The model update unit is used to measure the model mismatch at the current control moment. Model mismatch measurement with the previous control time step Calculate the rate of change of model mismatch .

[0057] when and At that time, the thermal resistance parameters and thermal capacity parameters in the discretized thermal resistance network model are recursively updated to obtain the updated discretized thermal resistance network model.

[0058] when and When the prediction bias compensation term of the discretized thermal resistance network model is updated, the updated discretized thermal resistance network model is obtained.

[0059] when At the same time, the model parameters of the discretized thermal resistance network model remain unchanged; where To preset the mismatch measurement threshold, This is a preset threshold for the rate of change of mismatch.

[0060] The coolant flow optimization unit is used to construct a temperature tracking error function based on the updated discretized thermal resistance network model; and to solve the temperature tracking error function under the constraints of coolant flow upper and lower limits and coolant flow rate change to obtain a re-optimized coolant flow sequence.

[0061] The model update unit includes: The parameter recursive update unit is used to update the parameters according to the preset long-term memory correction factor. The updated discretized thermal resistance network model is obtained by recursively updating the thermal resistance parameter R and the heat capacity parameter C of the discretized thermal resistance network model; wherein, the new thermal resistance parameter in the updated discretized thermal resistance network model is... New heat capacity parameters in the updated discretized thermal resistance network model ;in and This is the preset sensitivity coefficient.

[0062] The coolant flow optimization unit includes: The error function construction unit is used to predict the battery pack temperature in the prediction time domain based on the updated discretized thermal resistance network model, and obtain the predicted temperature sequence. A temperature tracking error function is constructed based on the predicted temperature sequence. The temperature tracking error function is as follows: .

[0063] in, The target temperature for the battery pack; To predict the first in the time domain The battery pack temperature vector of the step; To predict the coolant flow rate in the time domain; N is the length of the prediction time domain; and These are the weighting coefficients; This represents the variance of the battery pack's temperature distribution.

[0064] For example, during continuous control cycle operation, the coolant flow rate sequence calculated at the previous control moment is recorded. The prediction time domain length for a given operating moment is taken as... The coolant flow rate sequence calculated at the previous control time step is: Time-shifted extrapolation is processed by sequence translation, retaining the last four flow rates and adding the final flow rate value at the end to obtain the extrapolated coolant flow rate sequence. By leveraging the continuous change in operating conditions between adjacent control moments, the extended coolant flow rate sequence can be used as an initial candidate sequence for subsequent optimization calculations at the current control moment. The model mismatch metric at the current control moment... Model mismatch measurement at the previous control time step Substitute The model mismatch change rate was calculated. The mismatch metric threshold was determined based on historical operational data. The threshold for the rate of change of mismatch is .because and The thermal resistance and thermal capacity parameters are recursively updated on the discretized thermal resistance network model.

[0065] Long-term memory correction factor This is an empirical value. The initial thermal resistance of a certain battery hot node is... The original heat capacity is The sensitivity coefficient was determined through offline thermal model calibration experiments. 5 and Substituting the values ​​into the recursive update formula, the updated thermal resistance parameters are obtained. Updated heat capacity parameters The thermal resistance and thermal capacity parameters of the 48 battery thermal nodes are updated recursively in the same way. After the update, the parameters are rewritten into the discretized thermal resistance network model to obtain the updated discretized thermal resistance network model.

[0066] Assuming the model mismatch rate of change ,and Perform a forecast bias compensation term update. Forecast bias compensation term Used to correct the prediction output of the discretized thermal resistance network model. The prediction bias compensation term before the update is: Short-term memory correction factor ,according to The updated prediction bias compensation term is calculated. ,in The unit is ℃. Updated prediction bias compensation term. Write it into the discretized thermal resistance network model. Prediction bias compensation term. As a whole temperature offset, it is applied to the predicted temperature vector, and in the subsequent temperature prediction process, the predicted temperature vector... Each temperature component is uniformly compensated, meaning that the predicted temperature of each battery thermal node is superimposed with a compensation term. The compensated predicted temperature is used to correct systematic prediction biases. Thermal resistance and heat capacity parameters remain unchanged.

[0067] The updated discretized thermal resistance network model performs temperature prediction calculations in the prediction time domain to obtain the predicted temperature sequence. Determine the target temperature based on the safe operating temperature range of the battery pack. The weighting coefficient is determined based on the trade-off between temperature control performance and energy consumption. , The updated discretized thermal resistance network model calculates the predicted temperature sequence as follows: The error function value is obtained by substituting the data into the temperature tracking error function. .

[0068] Based on the cooling system design parameters, the upper and lower limits of the coolant flow rate are determined as follows: to The flow rate change rate is constrained to not exceed a certain limit per control cycle. Candidate flow rates that meet the constraints are substituted into the error function for calculation and comparison. The coolant flow rate sequence corresponding to the smallest error function is... This serves as a further optimized coolant flow sequence.

[0069] The control decision module is used to input the extended coolant flow rate sequence and the re-optimized coolant flow rate sequence into the updated discretized thermal resistance network model for temperature prediction, obtaining the extended temperature trajectory and the re-optimized temperature trajectory; calculate the extended comprehensive performance index and the re-optimized comprehensive performance index based on the extended temperature trajectory, the re-optimized temperature trajectory, and the thermal evolution intensity metric; determine the target coolant flow rate based on the extended comprehensive performance index and the re-optimized comprehensive performance index, and output it to the liquid cooling execution unit. The control decision module includes: The extended comprehensive performance index calculation unit is used to calculate the temperature deviation sequence and temperature distribution variance sequence of the battery pack in the predicted time domain based on the extended temperature trajectory; calculate the temperature deviation index based on the temperature deviation sequence; calculate the temperature uniformity index based on the temperature distribution variance sequence; calculate the coolant flow rate change sequence based on the extended coolant flow rate change sequence; calculate the control energy consumption index based on the coolant flow rate change sequence; calculate the index weighting coefficient based on the thermal evolution intensity measure; and perform a weighted combination calculation of the temperature deviation index, temperature uniformity index, and control energy consumption index based on the index weighting coefficient to obtain the extended comprehensive performance index.

[0070] The re-optimized comprehensive performance index calculation unit is used to calculate the temperature deviation sequence and temperature distribution variance sequence based on the re-optimized temperature trajectory; calculate the temperature deviation index based on the temperature deviation sequence; calculate the temperature uniformity index based on the temperature distribution variance sequence; calculate the control energy consumption index based on the re-optimized coolant flow sequence; and perform weighted combination calculation of the temperature deviation index, temperature uniformity index, and control energy consumption index based on the index weighting coefficients to obtain the re-optimized comprehensive performance index.

[0071] For example, extending the coolant flow rate sequence Input the updated discretized thermal resistance network model to perform temperature prediction calculations, and obtain the extended temperature trajectory as follows: The temperature deviation sequence is calculated based on the target temperature. The temperature deviation index is obtained by summing the squares of the temperature deviations. The temperature distribution of the extended temperature trajectory at 48 battery thermal nodes was statistically calculated. The predicted temperature distribution variances at each time point in the time domain were 0.06, 0.07, 0.08, 0.09, and 0.10, respectively. The summation yielded a temperature uniformity index of 0.06 + 0.07 + 0.08 + 0.09 + 0.10 = 0.40.

[0072] Extended coolant flow rate sequence and flow rate at the previous control time The coolant flow rate change sequence was calculated. The energy consumption control index is obtained by summing the squares. .

[0073] The weight coefficients of the indicators are determined based on the thermal evolution intensity measurement using a preset segmented weight mapping rule. At that time, the weighting for temperature deviation was 0.5, the weighting for temperature uniformity was 0.3, and the weighting for control energy consumption was 0.2. The temperature deviation index, temperature uniformity index, and control energy consumption index were each normalized according to preset reference values ​​before being weighted and combined for calculation. The normalized values ​​remained unchanged, and the calculated comprehensive performance index was: .

[0074] Further optimize the coolant flow sequence Input the updated discretized thermal resistance network model to perform temperature prediction calculations and obtain the re-optimized temperature trajectory. The temperature deviation sequence was calculated. The temperature deviation index is obtained by summing the squares. Further optimization of the temperature trajectory yielded temperature distribution variances of 0.05, 0.05, 0.05, 0.05, and 0.04 for the 48 battery thermal nodes. Summing these variances yielded the temperature uniformity index. Further optimization of the coolant flow rate sequence calculation yields a control energy consumption index that remains [value missing]. Temperature deviation, temperature uniformity, and energy consumption control indices are each normalized according to preset reference values ​​and then weighted and combined for calculation. The normalized values ​​remain unchanged, resulting in the optimized comprehensive performance index. .

[0075] The extended comprehensive performance index is 0.595, and the further optimized comprehensive performance index is 0.197. Comparing the two comprehensive performance indices, the coolant flow rate corresponding to the lower comprehensive performance index is selected as the target coolant flow rate. The first control variable in the time domain is predicted. The target coolant flow rate is output to the liquid-cooled actuator.

[0076] Finally, it should be noted that although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A temperature control system for a new energy vehicle battery pack, characterized in that, The system includes: The status acquisition module is used to acquire the battery pack temperature vector, coolant flow rate, and battery pack current parameters at the control moment; and calculate the corresponding predicted temperature vector based on the battery pack temperature vector, coolant flow rate, and battery pack current parameters through a pre-established discretized thermal resistance network model. The thermal state assessment module is used to calculate the thermal evolution intensity metric based on the battery pack temperature vector, coolant flow rate, and battery pack temperature vector and coolant flow rate at historical control times; and to calculate the model mismatch metric based on the battery pack temperature vector, predicted temperature vector, and thermal evolution intensity metric. A control sequence generation module is used to obtain the coolant flow rate sequence calculated at the previous control time; perform time-shift extension on the coolant flow rate sequence to obtain an extended coolant flow rate sequence; recursively correct the model parameters of the discretized thermal resistance network model according to the model mismatch metric to obtain an updated discretized thermal resistance network model; and perform constraint solving based on the updated discretized thermal resistance network model to obtain a re-optimized coolant flow rate sequence. The control decision module is used to input the extended coolant flow rate sequence and the re-optimized coolant flow rate sequence into the updated discretized thermal resistance network model to predict the temperature, thereby obtaining the extended temperature trajectory and the re-optimized temperature trajectory; calculate the extended comprehensive performance index and the re-optimized comprehensive performance index based on the extended temperature trajectory, the re-optimized temperature trajectory, and the thermal evolution intensity measure; determine the target coolant flow rate based on the extended comprehensive performance index and the re-optimized comprehensive performance index, and output it to the liquid cooling execution unit.

2. The temperature control system for a new energy vehicle battery pack according to claim 1, characterized in that, The system also includes: The model building module is used to divide the battery pack into multiple hot nodes and establish corresponding cooling channel hot nodes; establish thermal resistance between each battery hot node; establish convective thermal resistance between the battery hot nodes and the cooling channel hot nodes; and set heat capacity parameters for each battery hot node. Based on the preset battery structure parameters and cooling channel heat transfer parameters, the corresponding thermal resistance, convection resistance and heat capacity parameters are determined, and a discrete thermal resistance network model of the battery pack is constructed.

3. The temperature control system for a new energy vehicle battery pack according to claim 1, characterized in that, The thermal state assessment module includes: The thermal evolution intensity measurement calculation unit is used to obtain the current control time. Battery pack temperature vector Previous control time Battery pack temperature vector First two control moments Battery pack temperature vector Coolant flow rate at the current control moment Coolant flow rate at the previous control moment ; The acceleration vector of temperature change was calculated. ;in To control the cycle; the flow regulation rate is calculated. ; Based on the temperature change acceleration vector Calculate the scalar of average temperature acceleration and temperature distribution acceleration variance ; in This represents the number of temperature measurement points for the battery pack. , The acceleration vector of the temperature change The One component; According to the average temperature acceleration scalar Temperature distribution acceleration variance and flow regulation rate The thermal evolution intensity metric is calculated using a preset nonlinear mapping function. .

4. The temperature control system for a new energy vehicle battery pack according to claim 3, characterized in that, The thermal state assessment module includes: The model mismatch metric calculation unit is used to obtain the measured battery pack temperature vector and the predicted battery pack temperature vector at each historical control moment within a preset sliding time window; and to calculate the difference between the measured battery pack temperature vector and the predicted battery pack temperature vector at each historical control moment within the sliding time window to construct the corresponding instantaneous mismatch vector. The number of backtracking steps , The length of the preset sliding time window; According to the thermal evolution intensity measurement Calculate the dynamic forgetting factor ;in This is a reference value for thermal evolution intensity. The preset time decay constant; According to the dynamic forgetting factor For each instantaneous mismatch vector within the preset sliding time window Weighted fusion is performed to obtain the weighted mismatch baseline vector. ; for the weighted mismatch reference vector The model mismatch metric is obtained by performing L2 norm calculation. .

5. A temperature control system for a new energy vehicle battery pack according to claim 4, characterized in that, The control sequence generation module includes: The model update unit is used to measure the model mismatch at the current control moment. Model mismatch measurement with the previous control time step Calculate the rate of change of model mismatch ; when and At that time, the thermal resistance parameters and heat capacity parameters in the discretized thermal resistance network model are recursively updated to obtain the updated discretized thermal resistance network model; when and When the prediction bias compensation term of the discretized thermal resistance network model is updated, the updated discretized thermal resistance network model is obtained. when At the same time, the model parameters of the discretized thermal resistance network model remain unchanged; where To preset the mismatch measurement threshold, The preset mismatch change rate threshold is used; The coolant flow optimization unit is used to construct a temperature tracking error function based on the updated discretized thermal resistance network model; and to solve the temperature tracking error function under the constraints of coolant flow upper and lower limits and coolant flow rate change to obtain a re-optimized coolant flow sequence.

6. A temperature control system for a new energy vehicle battery pack according to claim 5, characterized in that, The model update unit includes: The parameter recursive update unit is used to update the parameters according to the preset long-term memory correction factor. The updated discretized thermal resistance network model is obtained by recursively updating the thermal resistance parameter R and the heat capacity parameter C of the discretized thermal resistance network model; wherein, the new thermal resistance parameter in the updated discretized thermal resistance network model is... New heat capacity parameters in the updated discretized thermal resistance network model ;in and This is the preset sensitivity coefficient.

7. A temperature control system for a new energy vehicle battery pack according to claim 5, characterized in that, The coolant flow optimization unit includes: The error function construction unit is used to predict the battery pack temperature in the prediction time domain based on the updated discretized thermal resistance network model, and obtain the predicted temperature sequence. A temperature tracking error function is constructed based on the predicted temperature sequence. The temperature tracking error function is as follows: ； in, The target temperature for the battery pack; To predict the first in the time domain The battery pack temperature vector of the step; To predict the coolant flow rate in the time domain; N is the length of the prediction time domain; and These are the weighting coefficients; This represents the variance of the battery pack's temperature distribution.

8. A temperature control system for a new energy vehicle battery pack according to claim 1, characterized in that, The control decision module includes: The extended comprehensive performance index calculation unit is used to calculate the temperature deviation sequence and temperature distribution variance sequence of the battery pack in the predicted time domain based on the extended temperature trajectory; calculate the temperature deviation index based on the temperature deviation sequence; calculate the temperature uniformity index based on the temperature distribution variance sequence; calculate the coolant flow rate change sequence based on the extended coolant flow rate sequence; calculate the control energy consumption index based on the coolant flow rate change sequence; calculate the index weighting coefficient based on the thermal evolution intensity measure; and perform a weighted combination calculation of the temperature deviation index, temperature uniformity index, and control energy consumption index based on the index weighting coefficient to obtain the extended comprehensive performance index. The re-optimized comprehensive performance index calculation unit is used to calculate the temperature deviation sequence and temperature distribution variance sequence based on the re-optimized temperature trajectory; calculate the temperature deviation index based on the temperature deviation sequence; calculate the temperature uniformity index based on the temperature distribution variance sequence; calculate the control energy consumption index based on the re-optimized coolant flow sequence; and perform weighted combination calculation of the temperature deviation index, temperature uniformity index, and control energy consumption index based on the index weighting coefficients to obtain the re-optimized comprehensive performance index.

Images