A lithium ion battery thermal management control method based on joint state estimation

Abstract

A lithium-ion battery thermal management control method based on joint state estimation includes: establishing a second-order RC equivalent electrical model of the lithium-ion battery and identifying the electrical model parameters using an improved robust adaptive multi-forgetting factor recursive least squares method; establishing a two-state lumped parameter thermal model of the lithium-ion battery and identifying the thermal model parameters using an improved goat optimization algorithm; constructing an electrothermal coupling model based on the identified electrical model and the two-state lumped parameter thermal model parameters, and then constructing a minimum entropy adaptive Kalman filter based on a near-end policy optimization algorithm to jointly estimate the battery's SOC and SOT; based on the joint estimation results, constructing a deep reinforcement learning-optimized adaptive model predictive control algorithm, and using the deep reinforcement learning-optimized adaptive model predictive control algorithm to perform thermal management control of the lithium-ion battery. This invention can achieve more accurate and effective thermal management control of lithium-ion batteries.

Description

Technical Field

[0001] This invention relates to the field of lithium-ion battery technology, and more specifically to a lithium-ion battery thermal management control method based on joint state estimation. Background Technology

[0002] With the widespread application of high-energy-density lithium-ion batteries in electric vehicles, their thermal safety and state estimation accuracy have become core bottlenecks restricting overall vehicle performance. Especially in distributed electric drive scenarios such as hub motor drive, batteries frequently experience high-current surges and complex operating condition switching, leading to three challenges for traditional single-state estimation methods: First, the electrical model parameters are highly time-varying, and conventional recursive least squares methods struggle to track dynamic characteristics due to fixed forgetting factors; second, the accuracy of thermal model parameter identification is insufficient, and existing optimization algorithms are prone to getting trapped in local optima, affecting the reliability of temperature field prediction; third, the strong coupling characteristics between SOC (State of Charge) and SOT (Seat Temperature) are not fully modeled, and independent estimation introduces cross-error.

[0003] To address the aforementioned issues, it is urgent to construct an electro-thermal joint estimation and coordinated control system to achieve more accurate and effective thermal management control of lithium-ion batteries. Summary of the Invention

[0004] In view of this, the present invention provides a lithium-ion battery thermal management control method based on joint state estimation to construct an electro-thermal joint estimation and coordinated control system, thereby achieving more accurate and effective lithium-ion battery thermal management control.

[0005] A thermal management control method for lithium-ion batteries based on joint state estimation includes:

[0006] Step S1: Establish a second-order RC equivalent electrical model for a lithium-ion battery, and use an improved robust adaptive multi-forgetting factor recursive least squares method to identify the electrical model parameters.

[0007] Step S2: Establish a dual-state lumped parameter thermal model of the lithium-ion battery, and use an improved goat optimization algorithm to identify the thermal model parameters.

[0008] Step S3: Based on the electrical model and dual-state lumped parameter thermal model parameters identified in steps S1 and S2, an electrothermal coupling model is constructed. Based on the electrothermal coupling model, a minimum entropy adaptive Kalman filter based on the near-end strategy optimization algorithm is constructed to jointly estimate the SOC and SOT of the battery.

[0009] Step S4: Based on the joint estimation results in step S3, construct an adaptive model predictive control algorithm optimized by deep reinforcement learning, and use the adaptive model predictive control algorithm optimized by deep reinforcement learning to perform thermal management control of the lithium-ion battery.

[0010] The lithium-ion battery thermal management control method based on joint state estimation provided by the present invention has the following beneficial effects:

[0011] 1. This invention comprehensively considers the lithium-ion battery electrical model and the dual-state lumped parameter thermal model. It characterizes the battery dynamic characteristics through the second-order RC equivalent electrical model and the dual-state lumped parameter thermal model, respectively. It also adopts an improved robust adaptive multi-forgetting factor recursive least squares method and an improved goat optimization algorithm to improve the parameter identification accuracy, thus providing a parameter and model basis for the joint estimation of SOT and SOC.

[0012] 2. Based on the electrical and thermal models of lithium-ion batteries, a novel electrothermal coupling model is constructed. Based on the identified electrical model and dual-state lumped parameter thermal model parameters, a minimum entropy adaptive Kalman filter based on the near-end policy optimization algorithm (PPO) is constructed to jointly estimate SOC and SOT, thereby solving state coupling interference.

[0013] 3. This invention uses a minimum entropy adaptive Kalman filter based on a near-end policy optimization algorithm to estimate the SOC and battery temperature in real time. It then constructs a deep reinforcement learning-optimized adaptive model predictive control algorithm to perform thermal management control on lithium-ion batteries. This ultimately achieves precise and effective thermal management control of lithium-ion batteries. This method overcomes the shortcomings of traditional thermal management response lag and overly conservative approaches through a complete innovation of parameter identification, joint estimation, and dynamic control, providing effective protection for battery safety under high dynamic conditions. Attached Figure Description

[0014] Figure 1 A flowchart of a lithium-ion battery thermal management control method based on joint state estimation provided in an embodiment of the present invention;

[0015] Figure 2 This is a comparison chart of the SOC estimation method proposed in this invention with the traditional extended Kalman algorithm and the actual values.

[0016] Figure 3 This is a comparison chart of the SOT estimation method proposed in this invention with the traditional extended Kalman algorithm and the actual value;

[0017] Figure 4 This is a comparison chart of the temperature control method proposed in this invention with that of the traditional PID algorithm and the ideal value. Detailed Implementation

[0018] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain embodiments of the present invention, and should not be construed as limiting the present invention.

[0019] Please see Figure 1 The embodiments of the present invention provide a lithium-ion battery thermal management control method based on joint state estimation, including steps S1-S4:

[0020] Step S1: Establish a second-order RC equivalent electrical model for a lithium-ion battery, and use an improved robust adaptive multi-forgetting factor recursive least squares method to identify the electrical model parameters.

[0021] According to Kirchhoff's voltage and current laws in circuit theory, and assuming the battery discharge direction is positive, the relationship between voltage and current in the second-order RC equivalent circuit of a lithium-ion battery is expressed as follows:

[0022]

[0023]

[0024]

[0025] in, This is the operating current. and These are the electrochemical polarization voltage and concentration polarization voltage of a lithium-ion battery, respectively. and These are the electrochemical polarization resistance and concentration polarization resistance of a lithium-ion battery, respectively. and These are the electrochemical polarization capacitance and concentration polarization capacitance of a lithium-ion battery, respectively. for The differential, for The differential, For time The differential, Open circuit voltage, For resistance parameters, This refers to the output voltage.

[0026] The SOC of a lithium-ion battery is expressed as:

[0027]

[0028] in, for The SOC value of the battery at any given time. For the initial The SOC value of the battery at any given time. For Coulomb efficiency, For maximum available capacity, for arrive In between, In time step Current at that time for The differential;

[0029] After discretization, we get:

[0030]

[0031]

[0032]

[0033] in, and They are respectively Electrochemical polarization voltage and concentration polarization voltage of lithium-ion batteries at any given time. and They are respectively Electrochemical polarization voltage and concentration polarization voltage of lithium-ion batteries at any given time. for Current at any moment for Current at any moment for Output voltage at any given time for Open-circuit voltage at any given time Sampling time, This represents the electrochemical polarization time constant of a lithium-ion battery. Concentration polarization time constant of lithium-ion batteries;

[0034] The state space is represented as:

[0035]

[0036]

[0037] in, and for The two internal state variables at time 1, for The SOC value of the battery at any given time. and for The two internal state variables at time 1, and These are resistance parameters;

[0038] set up , Then the parameters to be identified in the electric model for:

[0039] .

[0040] Specifically, the following sampling steps yield an improved robust adaptive multi-forgetting factor recursive least squares method:

[0041] Data preprocessing, the expression is:

[0042]

[0043]

[0044] in, and These are the normalized input and output data, respectively. for The original input variables at time 1. for The original output variable at time step, and These are the mean values ​​of the input and output data, respectively. and These are the standard deviations of the input and output data, respectively.

[0045] Experimental data were collected, SOC-OCV curves were fitted, and the relationship between the SOC-OCV curves and sampling time was used to obtain the target temperature. Open-circuit voltage at any given moment;

[0046] The robust adaptive forgetting factor recursive least squares filter parameters are updated as follows:

[0047]

[0048]

[0049] in, for Gain vector at time step for The error covariance matrix at time t. An adaptive forgetting factor. Indicates transpose. for The parameter estimation vector at time step [time]. for The parameter estimation vector at time step [time]. This is the gain matrix;

[0050] The covariance matrix is ​​updated as follows:

[0051]

[0052] in, for The error covariance matrix at time t. The regularization coefficient is . It is the identity matrix;

[0053] The adaptive forgetting factor is updated to:

[0054]

[0055] in, and These are the minimum and maximum values ​​of the forgetting factor, respectively. For adjustment coefficients, It is the square of the Euclidean norm.

[0056] Based on the above-mentioned improved robust adaptive multi-forgetting factor recursive least squares method, the internal resistance, capacitance and other parameters of the second-order RC equivalent electrical model can be identified.

[0057] Step S2: Establish a dual-state lumped parameter thermal model of the lithium-ion battery, and use an improved goat optimization algorithm to identify the thermal model parameters.

[0058] The expression for the two-state lumped-parameter thermal model of a lithium-ion battery is as follows:

[0059]

[0060]

[0061] in, and These are the core heat capacity and surface heat capacity of the battery, respectively. For output temperature The differential, , This refers to the core temperature of the battery. For ambient temperature, For temperature difference The differential, , The surface temperature of the battery. For time The differential, The rate at which the battery generates heat. For the core thermal resistance of the battery, The surface thermal resistance of the battery;

[0062] The Laplace transform yields:

[0063]

[0064]

[0065] in, Let be the complex frequency variable in the Laplace transform; according to the above equation, when , , , and Once determined, the heat generation rate will be... As input to the model, it can output as well as Then the parameters to be identified in the thermal model for:

[0066] .

[0067] Specifically, the improved goat optimization algorithm is used for thermal model parameter identification, including:

[0068] The Laplace transform yields and By performing difference processing, we obtain the discretized difference equation:

[0069]

[0070]

[0071]

[0072]

[0073]

[0074] in, for Output temperature at any time for Output temperature at any time for Output temperature at any time , , , These are unknowns to be identified related to thermophysical parameters. for real-time heat flow for real-time heat flow To use time intervals;

[0075] The expression for initializing the goat population location is:

[0076]

[0077] in, For the first Initial values ​​of individuals, and These are the lower and upper limits of the search space, respectively. It is a random number;

[0078] Calculate the fitness value of each goat to determine the initial global optimum. ;

[0079] For each iteration, the adaptive weights and dynamic search radius are calculated using the following expressions:

[0080]

[0081]

[0082] in, For the first Adaptive weights for the next iteration and These are the maximum and minimum values ​​of the adaptive weights, respectively. The maximum number of iterations, For the first The dynamic search radius of the next iteration. The maximum search radius. It is the attenuation constant;

[0083] The elite guide location is:

[0084]

[0085] in, For the first The position of elite individuals in the next iteration The total number of elite individuals, For the first In the nth iteration An elite solution;

[0086] The expression to update the position of each goat is:

[0087]

[0088] in, For the first The goats in the first Position in the next iteration For the first The goats in the first Position in the next iteration , It is a random number. For the first The global optimal position in the next iteration. For the first The position of a goat randomly selected in the next iteration Random numbers that are distributed according to a standard normal distribution. For elite guidance coefficient;

[0089] When the maximum number of iterations is reached When the convergence condition is met, the algorithm terminates, outputs the global optimal solution, and completes the construction of the improved goat optimization algorithm. Then, the improved goat optimization algorithm is used to identify the thermal model parameters. Specifically, voltage, current, and temperature data under lithium-ion battery discharge conditions are used as identification data inputs. Combined with the heat generation formula and the already identified entropy coefficient, the thermal property parameters are identified.

[0090] Step S3: Based on the electrical model and dual-state lumped parameter thermal model parameters identified in steps S1 and S2, an electrothermal coupling model is constructed. Based on the electrothermal coupling model, a minimum entropy adaptive Kalman filter based on the near-end strategy optimization algorithm is constructed to jointly estimate the SOC and SOT of the battery.

[0091] Specifically, step S3 includes:

[0092] The state-space equations of a second-order RC equivalent electrical model of a lithium-ion battery are constructed and discretized, and the expression is as follows:

[0093]

[0094]

[0095]

[0096]

[0097]

[0098]

[0099]

[0100] in, for The SOC value of the battery at any given time. For the battery's rated capacity, for Time-based process noise, and for The two internal state variables at time 1, , Polarized capacitor, For the voltage at the observation terminal, for The open-circuit voltage that changes with temperature. for Observation noise at any given moment for The state variables of the electric model at time t, for The input vector of the electric model at time step, for The observed variables of the time-varying electric model;

[0101] The state-space equations of the two-state lumped-parameter thermal model are constructed and discretized, and the expression is as follows:

[0102]

[0103]

[0104]

[0105]

[0106]

[0107]

[0108]

[0109] in, for The state variables of the time-limited thermal model for The state variables of the time-limited thermal model Here is the state transition matrix. For the input matrix, For the observation matrix, for The input vector of the time-limited thermal model, for Observed variables of the time-limited thermal model;

[0110] By leveraging the bidirectional coupling between the electrical and thermal models, an electrothermal coupled model is constructed, and the temperature-dependent parameters of the electrical model are updated synchronously. The expression is as follows:

[0111]

[0112] in, for The total internal heat generation power of the battery at any given time. The ohmic resistance varies with battery temperature. This represents the current absolute temperature of the battery. This is the partial derivative of the open-circuit voltage with respect to temperature;

[0113] Then, a minimum entropy adaptive Kalman filter based on the near-end policy optimization algorithm is constructed. The process is as follows:

[0114] Define the entropy of the filter as:

[0115]

[0116]

[0117] in, for Entropy at time, It is the natural logarithm Logarithmic function with base 0. Let be the covariance matrix of the filter. Let be the observation matrix of the filter. for Error covariance matrix of state estimation at time step. The covariance matrix of the measurement noise;

[0118] The optimization objective is to minimize the cumulative entropy, expressed as:

[0119]

[0120] in, It is to optimize the objective function. It is the total time;

[0121] The minimum entropy adaptive Kalman filter is optimized using a near-end policy optimization algorithm, which models the adjustment of the filter's hyperparameters as a policy. , It is in state Select action The probability, include Filters and observation information, For parameter adjustment amount, reward The expression is:

[0122]

[0123] in, It is the smoothing coefficient. and They are Time and State estimate at time 1;

[0124] Objective function of the proximal policy optimization algorithm for:

[0125]

[0126]

[0127]

[0128] in, Let be the expected function. This indicates taking the minimum value. Importance sampling ratio, For the dominant function, For the clipping function, It is a hyperparameter. For the old strategy, As a discount factor, For generalized advantage estimation (GAE) parameters, for Entropy at any given moment;

[0129] Loss of the value function for:

[0130]

[0131] in, For value function, From Time's up Accumulated rewards from discounts over time;

[0132] Adding an entropy regularization term to strategy optimization The expression is:

[0133]

[0134] in, The entropy coefficient, For strategy In state Entropy below;

[0135] The filter parameters, the policy network, and the value network of the near-end policy optimization algorithm are initialized, and then the filter is predicted and updated. The expression is:

[0136]

[0137]

[0138]

[0139]

[0140]

[0141] in, for The predicted state value at time 10:00. Here is the state transition matrix. For the prediction error covariance matrix, for Error covariance matrix of state estimation at time step. The process noise covariance matrix is... For Kalman gain, for The actual measured value at time, It is the identity matrix;

[0142] Then calculate the entropy and reward, and then apply the strategy. Select parameter adjustment action ,renew and Next, the data is stored, and finally, the near-end policy optimization algorithm is executed to update and optimize the objective function. and the loss of the value function Thus, a minimum entropy adaptive Kalman filter based on the near-end strategy optimization algorithm is constructed, which is then used to jointly estimate the SOC and SOT of the battery.

[0143] It should be noted that the joint estimation proposed in this invention is mainly divided into two parts: SOC estimation and SOT estimation. The SOC estimation part utilizes the identified open-circuit voltage and electrical model parameters that vary with temperature and SOC. The two state estimation modules are coupled through the transfer of external characteristic parameters and estimation results. By combining the electrical model of the second-order RC equivalent circuit, the state equation of the two-state lumped parameter thermal model, and the minimum entropy adaptive Kalman filter based on the near-end strategy optimization algorithm, the joint estimation of SOC and SOT is realized.

[0144] Step S4: Based on the joint estimation results in step S3, construct an adaptive model predictive control algorithm optimized by deep reinforcement learning, and use the adaptive model predictive control algorithm optimized by deep reinforcement learning to perform thermal management control of the lithium-ion battery.

[0145] Specifically, step S4 includes:

[0146] Define the state variables of the electrothermal coupling model The input vector of the electrothermal coupling model , for Battery temperature at any time, for Cooling power at any given time;

[0147] Objective function of adaptive model predictive control algorithm for:

[0148]

[0149] in, Is Time prediction Output at any moment This is a reference value for the output. , This is the battery's reference SOC value. This is a battery temperature reference value. It predicts the total step size. Is Time prediction Time-based control input, Is Time prediction The control state at any time, This is a status reference value. , , These are the cost weight matrices for output, control input, and terminal state, respectively.

[0150] The goal of adaptive model predictive control algorithms is to minimize the objective function. The solution employs quadratic programming for minimization, with deep reinforcement learning used for dynamic adjustment. , , and , To optimize control performance, a deep reinforcement learning-optimized adaptive model predictive control algorithm was constructed, which was then used for thermal management control of lithium-ion batteries, realizing a closed-loop thermal management strategy of real-time state estimation and adaptive optimization control.

[0151] Figure 2 and Figure 3 To compare the joint estimation of SOC and SOT proposed in this invention with the traditional Extended Kalman algorithm and the true values, a software-in-the-loop test was built using AMESIM software. Figure 2 It can be seen that the method proposed in this invention significantly improves the estimation accuracy of SOC, with an accuracy improvement of approximately 10%. Figure 3 It can be seen that the method proposed in this invention significantly improves the estimation accuracy of SOT, with an accuracy improvement of about 10%. Figure 4 To compare the temperature using the control method of this invention with that of traditional PID control, by Figure 4 It can be seen that, compared with the traditional PID control algorithm, the temperature control stability of the present invention is higher.

[0152] In summary, the lithium-ion battery thermal management control method based on joint state estimation according to the above embodiments has the following beneficial effects:

[0153] 1. This invention comprehensively considers the lithium-ion battery electrical model and the dual-state lumped parameter thermal model. It characterizes the battery dynamic characteristics through the second-order RC equivalent electrical model and the dual-state lumped parameter thermal model, respectively. It also adopts an improved robust adaptive multi-forgetting factor recursive least squares method and an improved goat optimization algorithm to improve the parameter identification accuracy, thus providing a parameter and model basis for the joint estimation of SOT and SOC.

[0154] 2. Based on the electrical and thermal models of lithium-ion batteries, a novel electrothermal coupling model is constructed. Based on the identified electrical model and dual-state lumped parameter thermal model parameters, a minimum entropy adaptive Kalman filter based on the near-end policy optimization algorithm (PPO) is constructed to jointly estimate SOC and SOT, thereby solving state coupling interference.

[0155] 3. This invention uses a minimum entropy adaptive Kalman filter based on a near-end policy optimization algorithm to estimate the SOC and battery temperature in real time. It then constructs a deep reinforcement learning-optimized adaptive model predictive control algorithm to perform thermal management control on lithium-ion batteries. This ultimately achieves precise and effective thermal management control of lithium-ion batteries. This method overcomes the shortcomings of traditional thermal management response lag and overly conservative approaches through a complete innovation of parameter identification, joint estimation, and dynamic control, providing effective protection for battery safety under high dynamic conditions.

[0156] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A thermal management control method for lithium-ion batteries based on joint state estimation, characterized in that, include: Step S1: Establish a second-order RC equivalent electrical model for a lithium-ion battery, and use an improved robust adaptive multi-forgetting factor recursive least squares method to identify the electrical model parameters. Step S2: Establish a dual-state lumped parameter thermal model of the lithium-ion battery, and use an improved goat optimization algorithm to identify the thermal model parameters. Step S3: Based on the electrical model and dual-state lumped parameter thermal model parameters identified in steps S1 and S2, an electrothermal coupling model is constructed. Based on the electrothermal coupling model, a minimum entropy adaptive Kalman filter based on the near-end strategy optimization algorithm is constructed to jointly estimate the SOC and SOT of the battery. Step S4: Based on the joint estimation results in step S3, construct an adaptive model predictive control algorithm optimized by deep reinforcement learning, and use the adaptive model predictive control algorithm optimized by deep reinforcement learning to perform thermal management control of the lithium-ion battery. Specifically, step S3 includes: The state-space equations of a second-order RC equivalent electrical model of a lithium-ion battery are constructed and discretized. Construct and discretize the state-space equations of a two-state lumped-parameter thermal model; By leveraging the bidirectional coupling between the electrical and thermal models, an electrothermal coupled model is constructed, and the temperature-dependent parameters of the electrical model are updated synchronously. The expression is as follows: in, for The total internal heat generation power of the battery at any given time. The ohmic resistance varies with battery temperature. This represents the current absolute temperature of the battery. Let be the partial derivative of the open-circuit voltage with respect to temperature. for The open-circuit voltage that changes with temperature. Then, a minimum entropy adaptive Kalman filter based on the near-end policy optimization algorithm is constructed. The process is as follows: Define the entropy of the filter as: in, for Entropy at time, It is the natural logarithm Logarithmic function with base 0. Let be the covariance matrix of the filter. Let be the observation matrix of the filter. for Error covariance matrix of state estimation at time step. The covariance matrix of the measurement noise; The optimization objective is to minimize the cumulative entropy, expressed as: in, It is to optimize the objective function. It is the total time; The minimum entropy adaptive Kalman filter is optimized using a near-end policy optimization algorithm, which models the adjustment of the filter's hyperparameters as a policy. , It is in state Select action The probability, reward The expression is: in, It is the smoothing coefficient. and They are Time and State estimate at time 1; Objective function of the proximal policy optimization algorithm for: in, Let be the expected function. This indicates taking the minimum value. Importance sampling ratio, For the dominant function, For the clipping function, It is a hyperparameter. For the old strategy, As a discount factor, For parameters of generalized dominance estimation, for Entropy at any given moment; Loss of the value function for: in, For value function, From Time's up Accumulated rewards from discounts over time; Adding an entropy regularization term to strategy optimization The expression is: in, The entropy coefficient, For strategy In state Entropy below; The filter parameters, the policy network, and the value network of the near-end policy optimization algorithm are initialized, and then the filter is predicted and updated. The expression is: in, for The predicted state value at time 10:

00. Here is the state transition matrix. For the prediction error covariance matrix, for Error covariance matrix of state estimation at time step. The process noise covariance matrix is... For Kalman gain, for The actual measured value at time, It is the identity matrix; Then calculate the entropy and reward, and then apply the strategy. Select parameter adjustment action ,renew and Next, the data is stored, and finally, the near-end policy optimization algorithm is executed to update and optimize the objective function. and the loss of the value function Thus, a minimum entropy adaptive Kalman filter based on the near-end strategy optimization algorithm is constructed, which is then used to jointly estimate the SOC and SOT of the battery.

2. The lithium-ion battery thermal management control method based on joint state estimation according to claim 1, characterized in that, In step S1, a second-order RC equivalent electrical model of a lithium-ion battery is established, specifically including: The relationship between voltage and current in the second-order RC equivalent circuit of a lithium-ion battery is expressed as: in, This is the operating current. and These are the electrochemical polarization voltage and concentration polarization voltage of a lithium-ion battery, respectively. and These are the electrochemical polarization resistance and concentration polarization resistance of a lithium-ion battery, respectively. and These are the electrochemical polarization capacitance and concentration polarization capacitance of a lithium-ion battery, respectively. for The differential, for The differential, For time The differential, Open circuit voltage, For resistance parameters, This refers to the output voltage. The SOC of a lithium-ion battery is expressed as: in, for The SOC value of the battery at any given time. For the initial The SOC value of the battery at any given time. For Coulomb efficiency, For maximum available capacity, for arrive In between, In time step Current at that time for The differential; After discretization, we get: in, and They are respectively Electrochemical polarization voltage and concentration polarization voltage of lithium-ion batteries at any given time. and They are respectively Electrochemical polarization voltage and concentration polarization voltage of lithium-ion batteries at any given time. for Current at any moment for Current at any moment for Output voltage at any given time for Open-circuit voltage at any given time Sampling time, This represents the electrochemical polarization time constant of a lithium-ion battery. Concentration polarization time constant of lithium-ion batteries; The state space is represented as: in, and for The two internal state variables at time 1, for The SOC value of the battery at any given time. and for The two internal state variables at time 1, and These are resistance parameters; set up , Then the parameters to be identified in the electric model for: 。 3. The lithium-ion battery thermal management control method based on joint state estimation according to claim 2, characterized in that, In step S1, the following steps are sampled to obtain the improved robust adaptive multi-forgetting factor recursive least squares method: Data preprocessing, the expression is: in, and These are the normalized input and output data, respectively. for The original input variables at time 1. for The original output variable at time step, and These are the mean values ​​of the input and output data, respectively. and These are the standard deviations of the input and output data, respectively. The relationship between the SOC-OCV curve and sampling time was used to obtain the target temperature. Open-circuit voltage at any given moment; The robust adaptive forgetting factor recursive least squares filter parameters are updated as follows: in, for Gain vector at time step for The error covariance matrix at time t. An adaptive forgetting factor. Indicates transpose. for The parameter estimation vector at time step [time]. for The parameter estimation vector at time step [time]. This is the gain matrix; The covariance matrix is ​​updated as follows: in, for The error covariance matrix at time t. The regularization coefficient is . It is the identity matrix; The adaptive forgetting factor is updated to: in, and These are the minimum and maximum values ​​of the forgetting factor, respectively. For adjustment coefficients, It is the square of the Euclidean norm.

4. The lithium-ion battery thermal management control method based on joint state estimation according to claim 3, characterized in that, In step S2, the expression for the two-state lumped-parameter thermal model of the lithium-ion battery is: in, and These are the core heat capacity and surface heat capacity of the battery, respectively. For output temperature The differential, , This refers to the core temperature of the battery. For ambient temperature, For temperature difference The differential, , The surface temperature of the battery. For time The differential, The rate at which the battery generates heat. For the core thermal resistance of the battery, The surface thermal resistance of the battery; The Laplace transform yields: in, For the complex frequency variable in the Laplace transform; Then the parameters to be identified in the thermal model for: 。 5. The lithium-ion battery thermal management control method based on joint state estimation according to claim 4, characterized in that, In step S2, an improved goat optimization algorithm is used to identify the parameters of the thermal model, specifically including: The Laplace transform yields and By performing difference processing, we obtain the discretized difference equation: in, for Output temperature at any time for Output temperature at any time for Output temperature at any time , , , These are unknowns to be identified related to thermophysical parameters. for real-time heat flow for real-time heat flow To use time intervals; The initialization expression for the goat population location is: in, For the first Initial values ​​of individuals, and These are the lower and upper limits of the search space, respectively. It is a random number; Calculate the fitness value of each goat to determine the initial global optimum. ; For each iteration, the adaptive weights and dynamic search radius are calculated using the following expressions: in, For the first Adaptive weights for the next iteration and These are the maximum and minimum values ​​of the adaptive weights, respectively. The maximum number of iterations, For the first The dynamic search radius of the next iteration. The maximum search radius. It is the attenuation constant; The elite guide location is: in, For the first The position of elite individuals in the next iteration The total number of elite individuals, For the first In the nth iteration An elite solution; The expression to update the position of each goat is: in, For the first The goats in the first Position in the next iteration For the first The goats in the first Position in the next iteration , It is a random number. For the first The global optimal position in the next iteration. For the first The position of a goat randomly selected in the next iteration Random numbers that are distributed according to a standard normal distribution. For elite guidance coefficient; When the maximum number of iterations is reached When the convergence condition is met, the algorithm terminates, outputs the global optimal solution, and completes the construction of the improved goat optimization algorithm. Then, the improved goat optimization algorithm is used to identify the parameters of the thermal model.

6. The lithium-ion battery thermal management control method based on joint state estimation according to claim 5, characterized in that, In step S3, the state-space equations of the second-order RC equivalent electrical model of the lithium-ion battery are constructed and discretized, and the expression is: in, for The SOC value of the battery at any given time. For the battery's rated capacity, for Time-based process noise, and for The two internal state variables at time 1, , Polarized capacitor, For the voltage at the observation terminal, for Observation noise at any given moment for The state variables of the electric model at time t, for The input vector of the electric model at time step, for The observed variables of the time-varying electric model; The state-space equations of the two-state lumped-parameter thermal model are constructed and discretized, and the expression is as follows: in, for The state variables of the time-limited thermal model for The state variables of the time-limited thermal model Here is the state transition matrix. For the input matrix, For the observation matrix, for The input vector of the time-limited thermal model, for The observed variables of the time-limited thermal model.

7. The lithium-ion battery thermal management control method based on joint state estimation according to claim 6, characterized in that, Step S4 specifically includes: Define the state variables of the electrothermal coupling model The input vector of the electrothermal coupling model , for Battery temperature at any time, for Cooling power at any given time; Objective function of adaptive model predictive control algorithm for: in, Is Time prediction Output at any moment This is a reference value for the output. , This is the battery's reference SOC value. This is a battery temperature reference value. It predicts the total step size. Is Time prediction Time-based control input, Is Time prediction Control status at all times, This is a status reference value. , , These are the cost weight matrices for output, control input, and terminal state, respectively. The goal of adaptive model predictive control algorithms is to minimize the objective function. The solution employs quadratic programming for minimization, with deep reinforcement learning used for dynamic adjustment. , , and , To optimize control performance, a deep reinforcement learning-optimized adaptive model predictive control algorithm is constructed, which is then used for thermal management control of lithium-ion batteries.

Images