Lithium ion battery preheating method and system in low-temperature environment

Abstract

The invention discloses a lithium ion battery preheating method and system in a low-temperature environment, and the method comprises the steps: carrying out the experimental data analysis of a lithium ion battery, and constructing a lithium ion battery RC equivalent circuit model and a lithium ion battery thermal model; simulating the lithium ion battery RC equivalent circuit model and the lithium ion battery thermal model to obtain a lithium ion battery simulation data set; performing optimization processing on the lithium ion battery simulation data set through a neural network regression prediction model and a multi-objective optimization algorithm to obtain optimal preheating parameters of the lithium ion battery; and performing optimal preheating treatment on the lithium ion battery based on the optimal preheating parameter of the lithium ion battery to obtain the preheated lithium ion battery. According to the method, the optimal preheating parameters enabling the battery to reach the maximum available energy can be found under different low-temperature conditions, the preheating energy consumption is reduced, and the overall energy utilization rate is improved. The lithium ion battery preheating method and system in the low-temperature environment can be widely applied to the technical field of lithium ion battery treatment.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium ion battery processing, and in particular to a lithium ion battery preheating method and system in a low temperature environment. Background Art

[0002] The performance of lithium-ion batteries in low temperature environments is significantly reduced. This phenomenon is mainly reflected in reduced capacity, increased internal resistance, reduced charge and discharge efficiency, and shortened cycle life. In order to overcome these problems, low-temperature preheating has been widely studied and applied to improve the performance of lithium-ion batteries in low temperature conditions. The core goal of low-temperature preheating is to reasonably control the heating process of the battery so that the battery can quickly heat up to a suitable operating temperature in a low-temperature environment, thereby increasing its electrochemical reaction activity and ion migration speed, thereby increasing the available energy.

[0003] Currently, there are two main methods for preheating lithium-ion batteries at low temperatures: internal heating and external heating. Internal heating directly heats the battery by integrating heating elements such as heating films, resistance wires or thermoelectric materials inside the battery. This method has a fast preheating speed, but may accelerate battery aging and pose safety risks, and it also requires high precision in the control circuit. External heating heats the battery as a whole by installing a heating device on the outside of the battery, such as a heating pad, heating box or heating cabin. This method uses an external heat source to conduct heat and increase the battery temperature. It has high safety and reliability, but the preheating efficiency is low under extremely low temperature conditions, resulting in extended preheating time and energy waste. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a lithium-ion battery preheating method and system in a low temperature environment, which can find the optimal preheating parameters for the battery to achieve the maximum available energy under different low temperature conditions, reduce preheating energy consumption, and improve overall energy utilization.

[0005] The first technical solution adopted by the present invention is: a lithium ion battery preheating method in a low temperature environment, comprising the following steps:

[0006] Analyze the experimental data of lithium-ion batteries and construct the RC equivalent circuit model and thermal model of lithium-ion batteries;

[0007] Conduct experimental simulation on the lithium-ion battery RC equivalent circuit model and the lithium-ion battery thermal model to obtain a lithium-ion battery simulation data set;

[0008] The lithium-ion battery simulation data set is optimized through the neural network regression prediction model and multi-objective optimization algorithm to obtain the optimal preheating parameters of the lithium-ion battery;

[0009] The lithium-ion battery is optimally preheated based on the optimal preheating parameters of the lithium-ion battery to obtain a preheated lithium-ion battery.

[0010] Furthermore, the step of analyzing the experimental data of the lithium-ion battery and constructing the RC equivalent circuit model and the thermal model of the lithium-ion battery specifically includes:

[0011] The internal temperature data of the lithium-ion battery is obtained through the thermocouple and the external temperature data of the lithium-ion battery is obtained through the PI heating film;

[0012] Based on the internal temperature data and external temperature data of the lithium-ion battery, the preheating power and preheating time of the PI heating film are adjusted, the lithium-ion battery is externally preheated, and the internal and external temperature values ​​of the lithium-ion battery under different ambient temperatures, preheating powers and preheating times are measured to construct a PI heating film preheating polynomial regression equation;

[0013] Conduct discharge experiments on fully charged lithium-ion batteries at different ambient temperatures and different discharge currents to obtain the value of lithium-ion battery discharge capacity changing with ambient temperature and discharge current;

[0014] Conduct mixed power pulse characteristic experiments on fully charged lithium-ion batteries at different ambient temperatures, different states of charge, and different discharge currents to obtain lithium-ion battery terminal voltage and current data;

[0015] The values ​​of lithium-ion battery discharge capacity changing with ambient temperature and discharge current, as well as the lithium-ion battery terminal voltage and current data are fitted to construct the lithium-ion battery RC equivalent circuit model and lithium-ion battery thermal model.

[0016] Furthermore, the step of fitting the values ​​of the lithium-ion battery discharge capacity changing with the ambient temperature and the discharge current and the lithium-ion battery terminal voltage and current data to construct the lithium-ion battery RC equivalent circuit model and the lithium-ion battery thermal model specifically includes:

[0017] The terminal voltage and current data of lithium-ion batteries are fitted by Matlab's cftool toolbox to construct a polynomial regression equation for the terminal voltage and current data of lithium-ion batteries.

[0018] According to the change of the discharge capacity of the lithium-ion battery with the ambient temperature and the discharge current, the polarization internal resistance value, the polarization capacitance value and the ohmic internal resistance value of the lithium-ion battery are determined;

[0019] Combining the polarization internal resistance value, polarization capacitance value and ohmic internal resistance value of the lithium-ion battery, a polynomial regression equation for the change of the discharge capacity of the lithium-ion battery with the ambient temperature and the discharge current is constructed;

[0020] According to the polynomial regression equation about the terminal voltage and current data of the lithium-ion battery and the polynomial regression equation about the value of the discharge capacity of the lithium-ion battery changing with the ambient temperature and the discharge current, combined with Kirchhoff's voltage and current law, the RC equivalent circuit model of the lithium-ion battery is constructed;

[0021] According to the polynomial regression equation about the terminal voltage and current data of the lithium-ion battery and the polynomial regression equation about the value of the discharge capacity of the lithium-ion battery changing with the ambient temperature and the discharge current, combined with Joule's law, a thermal model of the lithium-ion battery is constructed.

[0022] Further, the step of performing experimental simulation on the lithium-ion battery RC equivalent circuit model and the lithium-ion battery thermal model to obtain a lithium-ion battery simulation data set specifically includes:

[0023] Construct a percentage model of capacity attenuation of lithium-ion batteries based on the battery semi-empirical aging model formula;

[0024] The lithium-ion battery RC equivalent circuit model, lithium-ion battery thermal model, and lithium-ion battery capacity attenuation percentage model are coupled with the PI heating film preheating polynomial regression equation and input into the COMSOL simulation software for experimental simulation to obtain a lithium-ion battery simulation data set.

[0025] Furthermore, the step of optimizing the lithium-ion battery simulation data set by using the neural network regression prediction model and the multi-objective optimization algorithm to obtain the optimal preheating parameters of the lithium-ion battery specifically includes:

[0026] The neural network regression prediction model is trained based on the lithium-ion battery simulation data set to obtain a trained neural network regression prediction model;

[0027] Obtaining lithium-ion battery input data and predicting through the trained neural network regression prediction model to obtain the percentage value of the discharge capacity of the lithium-ion battery and the capacity attenuation of the lithium-ion battery;

[0028] Calculate the preheating energy consumption and available energy of the ion battery, define the available energy weight factor, aging loss rate weight factor and preheating energy consumption weight factor, and construct a weighted objective function;

[0029] The discharge capacity of the lithium-ion battery, the percentage value of the capacity attenuation of the lithium-ion battery and the weighted objective function are optimized by a multi-objective optimization algorithm to obtain the optimal preheating parameters of the lithium-ion battery.

[0030] Further, the step of training the neural network regression prediction model based on the lithium-ion battery simulation data set to obtain the trained neural network regression prediction model specifically includes:

[0031] Inputting a lithium-ion battery simulation data set into a neural network regression prediction model, wherein the neural network regression prediction model includes a sparrow search algorithm optimization model, a convolutional neural network, and a long short-term memory network;

[0032] Based on the sparrow search algorithm optimization model, the parameters of the neural network regression prediction model are optimized to obtain the optimized neural network regression prediction model;

[0033] Based on the convolutional neural network of the optimized neural network regression prediction model, local feature extraction processing is performed on the lithium-ion battery simulation data set to obtain the features of the lithium-ion battery simulation data set;

[0034] Based on the long short-term memory network of the optimized neural network regression prediction model, the time series feature extraction processing is performed on the characteristics of the lithium-ion battery simulation data set to obtain the time series features of the lithium-ion battery simulation data set;

[0035] The time series characteristics of the lithium-ion battery simulation dataset are used as verification dataset indicators to verify the optimized neural network regression prediction model and construct the trained neural network regression prediction model.

[0036] Furthermore, the step of obtaining the lithium-ion battery input data and predicting through the trained neural network regression prediction model to obtain the percentage value of the lithium-ion battery discharge capacity and the lithium-ion battery capacity decay specifically includes:

[0037] Based on the PI heating film preheating polynomial regression equation, the external temperature and internal temperature values ​​of the lithium-ion battery after preheating at different preheating powers, different preheating times, and different ambient temperatures are obtained;

[0038] Construct lithium-ion battery input data by combining the external temperature and internal temperature values ​​of lithium-ion batteries after preheating at different preheating powers, different preheating times, and different ambient temperatures, ambient temperature, state of charge, and lithium-ion battery discharge current values;

[0039] The input data of the lithium-ion battery is predicted by the trained neural network regression prediction model to obtain the percentage value of the discharge capacity of the lithium-ion battery and the capacity attenuation of the lithium-ion battery.

[0040] Further, the step of performing multi-objective optimization processing on the discharge capacity of the lithium-ion battery, the percentage value of the capacity attenuation of the lithium-ion battery and the weighted objective function by a multi-objective optimization algorithm to obtain the optimal preheating parameters of the lithium-ion battery specifically includes:

[0041] The weighted objective function is solved by a multi-objective optimization algorithm to obtain the Pareto optimal solution;

[0042] Determine target values ​​based on the Pareto optimal solution, wherein the target values ​​include available energy of the lithium-ion battery, capacity loss rate of the lithium-ion battery, and preheating energy consumption of the lithium-ion battery;

[0043] Perform a maximum available energy index search process on the target value to obtain the optimal preheating power and preheating time;

[0044] The lithium-ion battery discharge capacity, lithium-ion battery capacity loss rate, lithium-ion battery preheating energy consumption and lithium-ion battery available energy under the optimal preheating power and preheating time are calculated through auxiliary functions. The Pareto frontier diagram is drawn in combination with the discharge capacity of the lithium-ion battery and the percentage value of the lithium-ion battery capacity attenuation to obtain the optimal preheating parameters of the lithium-ion battery.

[0045] The second technical solution adopted by the present invention is: a lithium-ion battery preheating system in a low temperature environment, comprising:

[0046] The first module is used to analyze the experimental data of lithium-ion batteries and build the RC equivalent circuit model and thermal model of lithium-ion batteries;

[0047] The second module is used to perform experimental simulation on the lithium-ion battery RC equivalent circuit model and the lithium-ion battery thermal model to obtain a lithium-ion battery simulation data set;

[0048] The third module is used to optimize the lithium-ion battery simulation data set through a neural network regression prediction model and a multi-objective optimization algorithm to obtain the optimal preheating parameters of the lithium-ion battery;

[0049] The fourth module is used to perform optimal preheating treatment on the lithium-ion battery based on the optimal preheating parameters of the lithium-ion battery to obtain a preheated lithium-ion battery.

[0050] The beneficial effects of the method and system of the present invention are as follows: the present invention constructs an RC equivalent circuit model and a thermal model of the lithium-ion battery by performing experimental data analysis on the lithium-ion battery, performs experimental simulation on the RC equivalent circuit model and the thermal model of the lithium-ion battery, obtains a simulation data set of the lithium-ion battery, further optimizes the simulation data set of the lithium-ion battery by using a neural network regression prediction model and a multi-objective optimization algorithm, finds the optimal preheating parameters that enable the battery to reach the maximum available energy under different low temperature conditions by optimizing the heating power and the heating time, and finally performs optimal preheating on the lithium-ion battery based on the optimal preheating parameters of the lithium-ion battery, so as to cope with complex and changeable usage scenarios; the optimized preheating strategy can reduce the preheating energy consumption as much as possible and improve the overall energy utilization rate while ensuring the battery performance, so that the battery can provide more effective electrical energy in a low temperature environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1It is a flow chart of steps of a lithium ion battery preheating method in a low temperature environment of the present invention;

[0052] Figure 2 It is a structural block diagram of a lithium-ion battery preheating system in a low temperature environment of the present invention;

[0053] Figure 3 It is a schematic flow chart of an optimal preheating strategy for a lithium-ion battery that is self-adaptive in a low-temperature environment provided by a specific embodiment of the present invention;

[0054] Figure 4 It is a schematic diagram of the comparison of HPPC experimental curves when the SOC is 1 and the discharge current is 0.9A and 1.5A respectively provided in a specific embodiment of the present invention;

[0055] Figure 5 It is a schematic diagram of comparison of fitted temperature rise curves under different preheating powers when PI heating is performed at 248.15K provided by a specific embodiment of the present invention;

[0056] Figure 6 It is a coupling diagram of the RC equivalent circuit model, thermal model, aging model, and external heating model provided by a specific embodiment of the present invention;

[0057] Figure 7 It is a schematic diagram of a comparison curve of battery simulation voltage at different temperatures and different discharge currents provided by COMSOL in a specific embodiment of the present invention;

[0058] Figure 8 It is a schematic diagram of a battery simulation temperature comparison curve at different temperatures and different discharge currents provided by COMSOL in a specific embodiment of the present invention;

[0059] Fig. 9 It is a schematic diagram of a comparison curve of the external temperature and the internal temperature of a battery simulation at a discharge current of 1.5A and 3.0A when the preheating power is 9W and the preheating time is 3 minutes provided by COMSOL in a specific embodiment of the present invention;

[0060] Fig.10 It is a schematic diagram of voltage comparison curves of batteries at 0.5C and 1.0C discharge rates at 9W preheating power and different preheating times provided by COMSOL in a specific embodiment of the present invention;

[0061] Fig.11 It is a schematic diagram of the discharge capacity training results of lithium-ion batteries using the SSA-CNN-LSTM algorithm provided by a specific embodiment of the present invention;

[0062] Fig.12 It is a schematic diagram of the training results of the lithium-ion battery capacity loss percentage of the SSA-CNN-LSTM algorithm provided by a specific embodiment of the present invention;

[0063] Fig.13 It is a Pareto frontier diagram composed of available capacity and capacity loss rate solutions found by NSGA-II multi-objective optimization provided by a specific embodiment of the present invention;

[0064] Fig.14 It is a Pareto front diagram composed of available capacity and preheating energy consumption solutions found by NSGA-II multi-objective optimization provided by a specific embodiment of the present invention;

[0065] Fig.15 It is a Pareto front diagram composed of the capacity loss rate and the preheating energy consumption solution found by using NSGA-II multi-objective optimization provided by a specific embodiment of the present invention. DETAILED DESCRIPTION

[0066] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. The step numbers in the following embodiments are only provided for the convenience of explanation and description, and the order between the steps is not limited in any way. The execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0067] First of all, it should be noted that the design of the preheating strategy of lithium-ion batteries needs to consider the following key factors: preheating power, preheating time. In the design of the preheating strategy, the maximum available energy is an important evaluation indicator. Available energy refers to the effective electrical energy that the battery can provide after completing the preheating process. It not only depends on the discharge capacity of the battery, but also is affected by the preheating energy consumption. Therefore, the preheating strategy needs to find a balance between increasing the battery discharge capacity and reducing the preheating energy consumption to achieve the maximum available energy. Methods for optimizing the preheating strategy include multi-objective optimization algorithms, which optimize the heating power and heating time to find the optimal preheating parameters that enable the battery to achieve the maximum available energy under different low temperature conditions. The optimization objectives include maximizing available energy, minimizing preheating energy consumption, and minimizing aging loss rate. By comprehensively considering these objectives, using multi-objective optimization algorithms such as NSGA-II, the Pareto optimal solution can be found, so as to select preheating parameters suitable for specific application requirements and apply them to the actual battery preheating process to achieve optimal performance.

[0068] Based on this, the embodiment of the present invention utilizes the safety and reliability of the external heating method, adopts a multi-objective optimization algorithm (NSGA-II), optimizes the heating power and heating time, and searches for the optimal preheating parameters that enable the battery to achieve the maximum available energy under different low temperature conditions. By balancing the improvement of battery discharge capacity and the reduction of preheating energy consumption, the objectives of maximizing available energy, minimizing preheating energy consumption, and minimizing aging loss rate are comprehensively considered, so as to find the Pareto optimal solution, and finally achieve high efficiency and reliability in the battery preheating process.

[0069] Reference Figure 1The present invention provides a lithium ion battery preheating method under low temperature environment, the method comprising the following steps:

[0070] S100, analyzing the experimental data of lithium-ion batteries, and constructing the RC equivalent circuit model and thermal model of lithium-ion batteries;

[0071] Specifically, the internal temperature data of the lithium-ion battery is obtained through a thermocouple, and the external temperature data of the lithium-ion battery is obtained through a PI heating film; based on the internal temperature data of the lithium-ion battery and the external temperature data of the lithium-ion battery, the preheating power and preheating time of the PI heating film are adjusted, the lithium-ion battery is externally preheated, and the internal and external temperature values ​​of the lithium-ion battery under different ambient temperatures, preheating powers and preheating times are measured, and a PI heating film preheating polynomial regression equation is constructed;

[0072] In this embodiment, a lithium-ion battery is selected, a hole is opened in the negative electrode, and a thermocouple is inserted into the negative electrode to measure the internal temperature; a PI (polyimide film) heating film is wrapped around the outer surface of the lithium-ion battery, and a thermocouple is attached to measure the external temperature of the battery; under different ambient temperatures, the power and time applied to the PI heating film are changed to achieve external preheating of the battery, and the internal and external temperature values ​​of the lithium-ion battery under different ambient temperatures, preheating powers, and preheating times are measured, and a polynomial regression equation between the external temperature, internal temperature and the preheating power, preheating time, and ambient temperature is constructed.

[0073] Among them, it should be noted that the PI heating film preheats the lithium-ion battery at different preheating powers, preheating times, and ambient temperatures, obtains the external and internal temperatures of the lithium-ion battery, obtains the temperature data after the PI heating film is preheated, and establishes polynomial regression equations of the external temperature and internal temperature with power, time, and ambient temperature.

[0074] Conduct discharge tests on fully-charged lithium-ion batteries at different ambient temperatures and different discharge currents to obtain the value of lithium-ion battery discharge capacity changing with ambient temperature and discharge current; conduct mixed power pulse characteristic tests on fully-charged lithium-ion batteries at different ambient temperatures, different states of charge, and different discharge currents to obtain lithium-ion battery terminal voltage and current data;

[0075] In this embodiment, another complete lithium-ion battery is selected, fully charged at 25°C, and discharge experiments are carried out at different ambient temperatures and different discharge currents to obtain the value of the battery discharge capacity changing with ambient temperature and discharge current; then, HPPC (hybrid power pulse characteristic) experiments are carried out at different ambient temperatures, SOC (state of charge) and discharge currents to measure the terminal voltage and current data of the lithium-ion battery.

[0076] Further, combined with the experimental data, the NCR18650BD lithium battery was selected as the experimental object, the negative electrode of the battery was opened, and a thermocouple was inserted to measure the internal temperature value. The outer surface of the battery was wrapped with a PI heating film, and a thermocouple was pasted to measure the external temperature value. The power of the PI heating film was changed, and the lithium-ion battery was preheated at a preheating power of 7W, 9W, 13W, a preheating time of 1, 2, 3, 4, 5 minutes, and an ambient temperature of 248.15K, 263.15K, and 273.15K, respectively, to obtain the external and internal temperatures of the lithium-ion battery.

[0077] The temperature data of the PI heating film at different powers, times, and ambient temperatures were further obtained, and the cftool tool in Matlab was used for fitting to obtain the unknown coefficients a. 1 -a 7 value, establish the external temperature T out and internal temperature T in With preheating power P, preheating time t, ambient temperature T er The polynomial regression equation is as follows: Figure 4 As shown in the figure, the maximum error between the fitted polynomial regression equation and the experimental data is controlled within 2%, R 2 =0.996, the fitting effect is good;

[0078] T out =a 1 +a 2 Pt+a 3 PT er +a 4 T er +a 5 P 2 +a 6 t 2 +a 7 T er 2

[0079] T in =a 1 +a 2 Pt+a 3 PT er +a 4 T er +a 5 P 2 +a 6 t 2 +a 7 T er 2

[0080] Select a new NCR18650BD battery, fully charge the battery at room temperature 298.15K, and discharge it at 0.9A, 1.5A, 2.1A, and 3.0A at 248.15K, 263.15K, 273.15K, 283.15K, and 298.15K, respectively, to obtain the battery discharge capacity, terminal voltage and other data;

[0081] HPPC experiments were carried out at 248.15K, 263.15K, 273.15K, 283.15K, and 298.15K, 0%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, and 100% SOC, and 0.9A, 1.5A, 2.1A, and 3.0A currents. The process is shown in Table 1. Figure 5 The HPPC experiment of the battery is shown at an ambient temperature of 248.15K, SOC=100%, and currents of 0.9A and 1.5A respectively.

[0082] Table 1 HPPC discharge test process

[0083] step Operation content 1 After the battery was kept at 298.15K for 1h, the SOC was set to 100%. 2 Adjust the temperature of the thermostat to 248.15K and let it stand for 2 hours. 3 Discharge at 0.9A for 30s, let stand for 60s, then charge at 0.9A for 30s 4 Adjust the temperature of the thermostat to 298.15K and let it stand for 2 hours. 5 Discharge 0.3Ah at 1.5A current and let stand for 1h 6 Repeat steps 2 to 5 9 times and end

[0084] The values ​​of lithium-ion battery discharge capacity changing with ambient temperature and discharge current, as well as the lithium-ion battery terminal voltage and current data are fitted to construct the lithium-ion battery RC equivalent circuit model and lithium-ion battery thermal model.

[0085] Among them, the process of fitting the values ​​of the lithium-ion battery discharge capacity changing with the ambient temperature and discharge current and the lithium-ion battery terminal voltage and current data is as follows:

[0086] The terminal voltage and current data of the lithium-ion battery are fitted through the cftool toolbox of Matlab to construct a polynomial regression equation for the terminal voltage and current data of the lithium-ion battery; the polarization internal resistance value, polarization capacitance value and ohmic internal resistance value of the lithium-ion battery are determined according to the value of the discharge capacity of the lithium-ion battery changing with the ambient temperature and the discharge current; the polarization internal resistance value, polarization capacitance value and ohmic internal resistance value of the lithium-ion battery are combined to construct a polynomial regression equation for the value of the discharge capacity of the lithium-ion battery changing with the ambient temperature and the discharge current; based on the polynomial regression equation for the terminal voltage and current data of the lithium-ion battery and the polynomial regression equation for the value of the discharge capacity of the lithium-ion battery changing with the ambient temperature and the discharge current, combined with Kirchhoff's voltage and current law, a lithium-ion battery RC equivalent circuit model is constructed; based on the polynomial regression equation for the terminal voltage and current data of the lithium-ion battery and the polynomial regression equation for the value of the discharge capacity of the lithium-ion battery changing with the ambient temperature and the discharge current, combined with Joule's law, a lithium-ion battery thermal model is constructed.

[0087] In this embodiment, the data in the HPPC experiment under different ambient temperatures, SOCs, and discharge currents are identified, and the coefficient k can be fitted by the cftool toolbox in Matlab. 0 , k 1 、b 1 The value of is expressed as:

[0088]

[0089] Further calculate the polarization internal resistance R of the battery under different ambient temperatures, SOC, and discharge current 1 , polarization capacitance C 1 , whose expression is:

[0090]

[0091]

[0092] Calculate the ohmic internal resistance R of the battery at different ambient temperatures, SOCs, and discharge currents 0 , whose expression is:

[0093]

[0094] Among them, the terminal voltage V of the power battery after discharge is completed A Since the internal electrochemical reaction does not stop immediately, the external manifestation is that the terminal voltage rises rapidly and then gradually tends to a stable value V B , which is the "rebound characteristic" of the power battery. At this time, it is a zero input response. The "rebound characteristic" of the terminal voltage is mainly affected by the ohmic internal resistance R 0 Therefore, according to Ohm's law, the ohmic internal resistance R 0 ;

[0095] According to the above, the R of the battery under different ambient temperatures, SOC, and discharge current is obtained. 0 , R 1 , C 1 Parameters, establish the polynomial regression equation under ambient temperature, SOC, and discharge current; combine the RC equivalent circuit model and Kirchhoff's voltage and current law to obtain the state equation of the RC equivalent circuit model; combine with Joule's law to obtain the thermal model of the battery.

[0096] Specifically, the polarization internal resistance R is obtained 1 , polarization capacitance C 1 、Ohmic internal resistance R 0 The ohmic internal resistance R is established at different ambient temperatures, SOCs, and discharge currents. 0 , Polarization internal resistance R 1, polarization capacitance C 1 About ambient temperature T er , SOC, and discharge current I under the polynomial regression equation, the expression is:

[0097] y=a 1 +a 2 I+a 3 SOC+a 4 T er +a 5 I 2 +a 6 SOC 2 +a 7 T er 2 +a 8 I 3 +a 9 SOC 3 +a 10 T er 3

[0098] Combining the first-order RC equivalent circuit model and Kirchhoff's voltage and current laws, the equation of the first-order RC equivalent circuit model can be obtained as shown below:

[0099]

[0100] In the above formula, U L is the battery terminal voltage, U OC is the battery open circuit voltage, I 1 is the polarization current flowing through the polarization internal resistance,

[0101] Combined with Joule's law, the thermal model of the battery can be obtained, and its expression is as follows:

[0102]

[0103] In the above formula, Q z is the total heat of the battery, t p is the discharge time, Q 1 It is expressed as the heat lost in the heat generation process, h is the equivalent convection heat transfer coefficient, A is the battery heat transfer area, T er is the external environment temperature.

[0104] S200, performing experimental simulation on a lithium-ion battery RC equivalent circuit model and a lithium-ion battery thermal model to obtain a lithium-ion battery simulation data set;

[0105] Specifically, a percentage model of capacity decay of lithium-ion batteries is constructed according to the formula of the battery semi-empirical aging model; the RC equivalent circuit model of lithium-ion batteries, the thermal model of lithium-ion batteries, the percentage model of capacity decay of lithium-ion batteries and the PI heating film preheating polynomial regression equation are coupled and input into the COMSOL simulation software for experimental simulation to obtain a lithium-ion battery simulation data set.

[0106] In this embodiment, the constructed equivalent circuit model, thermal model, battery aging loss model and PI heating film preheating polynomial regression equation are input into the COMSOL simulation software, and the preheating time, preheating power and ambient temperature of the PI heating film are changed to realize the simulation of the battery under different ambient temperatures, SOC and discharge currents, and the simulation results are recorded.

[0107] The battery semi-empirical aging model formula is used to obtain the battery capacity attenuation percentage model. The battery semi-empirical aging model formula is expressed as:

[0108]

[0109] In the above formula, C is the battery discharge capacity, and R is the gas constant;

[0110] The battery equivalent circuit model, thermal model, battery aging capacity attenuation model, and external heating model are further coupled, such as Figure 6 As shown, the simulation data is input into COMSOL software for simulation; the simulation data is compared and evaluated with the real data; if the RMSE of voltage is less than 0.05V and the RMSE of temperature is less than 2K, the simulation model is considered to be good. Figure 7 , Figure 8 and Fig. 9 If not satisfied, further adjust the model.

[0111] The data is further input into the COMSOL simulation software, and the simulation is performed by changing the preheating power and preheating time of the PI heating film and changing the ambient temperature, the SOC of the battery, and the discharge current. The simulated battery voltage, discharge capacity, battery external temperature, battery internal temperature, and percentage value of battery capacity attenuation are obtained and constructed into a data set, such as Fig.10 As shown, the battery voltage changes when it is discharged at 1.5A and 3A respectively after being preheated for 1, 3 and 5 minutes at a preheating power of 9W.

[0112] S300, optimizing the lithium-ion battery simulation data set by using a neural network regression prediction model and a multi-objective optimization algorithm to obtain optimal preheating parameters for the lithium-ion battery;

[0113] Specifically, a neural network regression prediction model is trained based on a lithium-ion battery simulation data set to obtain a trained neural network regression prediction model;

[0114] Among them, the lithium-ion battery simulation data set is input into the neural network regression prediction model, and the neural network regression prediction model includes a sparrow search algorithm optimization model, a convolutional neural network and a long short-term memory network; based on the sparrow search algorithm optimization model, the parameters of the neural network regression prediction model are optimized to obtain the optimized neural network regression prediction model; based on the convolutional neural network of the optimized neural network regression prediction model, the local feature extraction process is performed on the lithium-ion battery simulation data set to obtain the lithium-ion battery simulation data set features; based on the long short-term memory network of the optimized neural network regression prediction model, the time series feature extraction process is performed on the lithium-ion battery simulation data set features to obtain the lithium-ion battery simulation data set time series features; the lithium-ion battery simulation data set time series features are used as verification data set indicators to verify the optimized neural network regression prediction model, and a trained neural network regression prediction model is constructed.

[0115] In this embodiment, the simulation results are constructed as a data set, and the SSA-CNN-LSTM model is trained by the data set. Specifically, the external temperature and internal temperature values ​​under different ambient temperatures, SOC, discharge current and different preheating strategies are sorted as input data, and the percentage values ​​of discharge capacity and battery capacity decay are sorted as output data. Then, the SSA-CNN-LSTM model is trained using these data to accurately estimate the percentage values ​​of the discharge capacity and battery capacity decay of the battery under the above different conditions.

[0116] It should be further explained that, first, 70% of the data set is used as a training data set and 30% is used as a test data set; the training data set is used to train the SSA-CNN-LSTM model, and the test data set is used to verify the accuracy of the model.

[0117] The data in the training set include ambient temperature, SOC, discharge current, and external and internal temperature values ​​after battery preheating; the output data include battery discharge capacity and percentage value of battery capacity attenuation; the training process of the SSA-CNN-LSTM model includes three main parts: parameter optimization, local feature extraction, and time series feature processing, as shown below:

[0118] 1) Parameter optimization: Use the sparrow search algorithm (SSA) to optimize the key parameters of the model, including (NumOfUnits) the number of neurons in the LSTM layer, (InitialLearnRate) the initial learning rate during the optimization process, and (L2Regularization) the L2 regularization coefficient; the SSA algorithm sets the population size to 30 and the number of iterations to 8; SSA goes through steps such as initializing the population, calculating the fitness value, and updating the population, and finally finds the optimal parameter combination to ensure that the model has the best training performance;

[0119] 2) Local feature extraction: Convolutional neural network (CNN) is used to extract local features of input data. In this model, the size of the input sequence is set to [5,1,1], indicating that each input sample is a time series of size 5. Two convolutional layers are set. The first convolutional layer uses (convolution2dLayer), the first convolution kernel size is 2x1, the number of output channels is 10, and the step size is [1,1]. The second convolutional layer also uses (convolution2dLayer) and the second convolution kernel size is 1x1, the number of output channels is 10, and the step size is [1,1]. The pooling layer uses (maxPooling2dLayer) to set the pooling window size to 1x3 and the step size to 1. (flattenLayer) flattens the two-dimensional feature map output by the pooling layer into a one-dimensional vector, and (fullyConnectedLayer) sets the output size to 2 to meet the output dimension requirements. Finally, the output layer (regressionLayer) is used for regression tasks to ensure direct numerical comparison between the network output and the target value.

[0120] 3) Time series feature processing: Long short-term memory network (LSTM) is used to process the time series features of data; in this model, three LSTM layers are set. The first LSTM layer (lstmLayer) is set to have (NumOfUnits) neurons, the output mode is sequence (OutputMode, sequence), and a dropout layer (dropoutLayer) is added after it, and the dropout rate is 0.3; the second LSTM layer is also set to have (NumOfUnits) neurons, the output mode is sequence, and a dropout layer is also set after it, and the dropout rate is 0.3; the third LSTM layer has (NumOfUnits) neurons, the output mode is the last element (OutputMode, last), and a dropout layer is added after it, and the dropout rate is 0.3. Among them, NumOfUnits is a parameter optimized by the SSA algorithm. Finally, the output of the LSTM layer is converted to the required output size through a fully connected layer, followed by a nonlinear transformation through a tanh layer, and finally the output of the regression task is realized through a regression layer, ensuring a direct numerical comparison between the network output and the target value;

[0121] The validation set is input into the SSA-CNN-LSTM regression prediction model to predict the battery discharge capacity and the percentage of battery capacity attenuation; Fig.11 The figure shows the effect of battery discharge capacity training; Fig.12 The figure shows the training effect of the percentage of battery capacity decay; the battery capacity RMSE is less than 50mAh, and the battery capacity decay percentage RMSE is less than 0.008‰, which has a high training effect.

[0122] Obtaining lithium-ion battery input data and predicting through the trained neural network regression prediction model to obtain the percentage value of the discharge capacity of the lithium-ion battery and the capacity attenuation of the lithium-ion battery;

[0123] Specifically, based on the PI heating film preheating polynomial regression equation, the external temperature and internal temperature values ​​of the lithium-ion battery after preheating at different preheating powers, different preheating times and different ambient temperatures are obtained; the external temperature and internal temperature values ​​of the lithium-ion battery after preheating at different preheating powers, different preheating times and different ambient temperatures, the ambient temperature, the state of charge, and the lithium-ion battery discharge current value are combined to construct the lithium-ion battery input data; the lithium-ion battery input data is predicted by the trained neural network regression prediction model to obtain the discharge capacity of the lithium-ion battery and the percentage value of the lithium-ion battery capacity attenuation.

[0124] Calculate the preheating energy consumption and available energy of the lithium-ion battery, define the available energy weight factor, aging loss rate weight factor and preheating energy consumption weight factor, and construct a weighted objective function; use a multi-objective optimization algorithm to perform multi-objective optimization processing on the discharge capacity of the lithium-ion battery, the percentage value of the lithium-ion battery capacity attenuation and the weighted objective function to obtain the optimal preheating parameters of the lithium-ion battery.

[0125] The predicted discharge capacity of the battery is converted into the discharge energy of the battery, and its expression is:

[0126] W sum =CV*3600

[0127] The available energy is defined as the discharge energy minus the energy consumed by preheating, and its expression is:

[0128] W act =W sum -W con

[0129] Among them, the weighted objective function is solved by a multi-objective optimization algorithm to obtain a Pareto optimal solution; based on the Pareto optimal solution, the target value is determined, and the target value includes the available energy of the lithium-ion battery, the capacity loss rate of the lithium-ion battery and the preheating energy consumption of the lithium-ion battery; the target value is searched for the maximum available energy index to obtain the optimal preheating power and preheating time; the lithium-ion battery discharge capacity, lithium-ion battery capacity loss rate, lithium-ion battery preheating energy consumption and lithium-ion battery available energy under the optimal preheating power and preheating time are calculated by an auxiliary function, and a Pareto frontier diagram is drawn in combination with the discharge capacity of the lithium-ion battery and the percentage value of the lithium-ion battery capacity attenuation to obtain the optimal preheating parameters of the lithium-ion battery.

[0130] In this embodiment, a PI heating film polynomial regression equation is loaded to calculate the external temperature and internal temperature values ​​of the battery under different preheating strategies; a trained SSA-CNN-LSTM neural network model is loaded to predict the battery discharge capacity and the percentage value of the battery capacity decay; an input network is created, specifically including SOC, discharge current, external temperature, internal temperature, and ambient temperature, and normalized, and the normalized input data is input into the SSA-CNN-LSTM neural network model to predict the battery discharge capacity and the percentage value of the battery capacity decay; the preheating energy consumption and available energy are calculated, and weight factors w1, w2, and w3 are defined to correspond to available energy, aging loss rate, and preheating energy consumption, respectively. Specifically, w1, w2, and w3 are 0.4, 0.3, and 0.3, respectively, to construct a weighted objective function; the NSGA-II algorithm is used for multi-objective optimization, the goal of which is to maximize the available energy while minimizing the percentage of battery capacity decay and preheating energy consumption, and to extract the optimal solution obtained by optimization, including preheating power, preheating time, and corresponding battery discharge capacity, maximum available energy, and percentage performance indicators of battery capacity decay.

[0131] It should be further explained that the PI heating film polynomial regression equation is first loaded to calculate the external temperature and internal temperature values ​​of the battery after preheating at different preheating powers, preheating times, and ambient temperatures;

[0132] The trained SSA-CNN-LSTM neural network model is further loaded to predict the battery discharge capacity and the percentage value of battery capacity attenuation based on the input data including the battery external temperature, internal temperature, ambient temperature, SOC, and discharge current value;

[0133] Construct an optimization objective function (optimizationObjectiveFunction), and the input parameters include preheating power, preheating time, ambient temperature, SOC, discharge current, model, LSTM network and normalization parameters. Set the constraints as the upper and lower limits of preheating power and the upper and lower limits of preheating time.

[0134] The weight factors w1, w2, and w3 are defined to correspond to the available energy, the aging loss rate, and the preheating energy consumption, respectively. The weight factors are used in the optimization objective function to perform weighted calculations on different target values.

[0135] Use NSGA-II for multi-objective optimization, set the population size to 100, the maximum number of iterations to 200, and the optimization variables to be preheating power and preheating time. The multi-objective optimization function (gamultiobj) is used to solve the Pareto optimal solution (paretoSolutions) and the corresponding objective value (paretoObjectives);

[0136] Extract the target values ​​from the Pareto optimal solution, including available energy, capacity loss rate, and preheating energy consumption. Find the index (optimalIndex) corresponding to the maximum available energy and obtain the corresponding optimal preheating power and preheating time;

[0137] Call the auxiliary function (calculateMetrics) to calculate the discharge capacity, capacity loss rate, preheating energy consumption and available energy under the optimal strategy. Display the corresponding preheating power, preheating time, discharge capacity, available energy, aging capacity loss rate, and draw the Pareto frontier diagram of available energy and preheating energy consumption, available energy and aging loss rate, and preheating energy consumption and aging loss rate, as shown in the following figure. Fig.13 , Fig.14 and Fig.15 As shown, including available energy vs. preheating energy consumption, available energy vs. aging loss rate, preheating energy consumption vs. aging loss rate.

[0138] S400, performing optimal preheating treatment on the lithium-ion battery based on optimal preheating parameters of the lithium-ion battery to obtain a preheated lithium-ion battery.

[0139] Specifically, the obtained optimal preheating parameters are applied to lithium-ion batteries, and the effectiveness and stability of the optimal preheating strategy are verified through battery low-temperature experiments; the specific process includes using the best preheating strategy to test battery performance in an actual low-temperature environment, measuring key parameters such as discharge capacity, percentage of battery capacity decay and preheating energy consumption, and making a detailed comparison with the battery performance under the ordinary preheating strategy; the verification experiment results are used to evaluate the feasibility and advantages of the optimization strategy in practical applications; after the experiment is completed, the actual effect of the optimization strategy is analyzed according to the experimental data, and the optimal preheating strategy is applied to larger-scale battery packs. By adjusting the preheating strategy in the battery management system, the performance of the battery pack in a low-temperature environment can be improved, the service life can be extended, and its reliability and safety under different working conditions can be ensured, so as to achieve efficient and safe operation of lithium-ion batteries under various low-temperature conditions.

[0140] The optimal preheating strategy was input into the simulation model for verification and compared with the original data, as shown in Table 2 below, which shows the comparative data at an ambient temperature of 248.15 K, 100% SOC, and 3.0 A discharge current.

[0141] Table 2 Comparison between the best preheating strategy and the common preheating strategy

[0142] Comparative Analysis Best warm-up strategy Common warm-up strategy Common warm-up strategy Common warm-up strategy Preheating power(W) 12.4 7 9 13 Warm-up time(s) 299 300 300 300 Discharge capacity(Ah) 2.181 2.3032 2.3228 2.048 Aging loss rate (%) 0.00578 0.01163 0.01067 0.00624 Preheating energy consumption (J) 3707.6 2100 2700 3900 Available energy (J) 25316.2 28845.0 28239.7 23379.4

[0143] It can be seen from Table 2 that, compared with the case where the optimal preheating strategy uses a preheating power of 13W and a preheating time of 300s, the available energy increases by 8.28%, the preheating energy consumption decreases by 4.93%, and the aging loss rate decreases by 7.37%; compared with the case where the preheating power of 9W and the preheating time of 300s are used, although the available energy decreases by 10.35%, the preheating energy consumption increases by 37.32%, but the aging loss rate decreases by 45.829%; this proves that the proposed optimal preheating strategy has good practicality.

[0144] In addition, it should be noted that the best preheating strategy under different optimization objectives can be obtained by changing the ratio of weight factors w1, w2, and w3.

[0145] In summary, the embodiment of the present invention adopts the method of external preheating of the PI heating film, obtains the external temperature and internal temperature values ​​of the lithium-ion battery under different preheating powers, preheating times and ambient temperatures, and establishes a regression equation related thereto; and conducts HPPC experiments on the lithium-ion battery under different ambient temperatures, SOCs and discharge currents, and performs parameter identification on the experimental data, and establishes a polynomial regression equation between the electrical parameters and the ambient temperature, SOC and discharge current; couples the RC equivalent circuit model, the thermal model, the aging capacity loss model and the external preheating model, and inputs them into the COMSOL software, changes the preheating strategy, ambient temperature, SOC, discharge current and other variables to realize simulation, and obtains simulation data; establishes a data set for the simulation data, and proposes a regression prediction method for the discharge capacity and the battery aging capacity loss percentage based on the SSA-CNN-LSTM neural network model; adopts the multi-objective optimization genetic algorithm NSGA-II, and automatically finds the optimal preheating strategy for the input ambient temperature, battery internal and external temperature, SOC, and discharge current data to maximize the battery discharge capacity, minimize the percentage of battery capacity attenuation, and minimize the preheating energy consumption. As a method for obtaining the best preheating strategy for lithium-ion batteries in low-temperature environments, the present invention has strong adaptability and can cope with complex and changeable usage scenarios; the optimized preheating strategy can reduce preheating energy consumption as much as possible while ensuring battery performance, improve overall energy utilization, and enable the battery to provide more effective electrical energy in low-temperature environments.

[0146] Therefore, the embodiments of the present invention have the following beneficial effects:

[0147] 1) The embodiment of the present invention uses the advantages of reliability and safety of external heating to heat the battery; couples the lithium battery equivalent circuit model, thermal model, external heating model and battery capacity attenuation model; considers the influence of the temperature difference between the internal temperature and the external temperature of the battery on battery aging; adopts the method of physical model plus data-driven model to make full use of physical laws and data characteristics, and improves the accuracy and reliability of the model;

[0148] 2) The embodiments of the present invention can significantly improve the working performance of lithium-ion batteries in low temperature environments by reasonably designing preheating strategies. In the process of optimizing preheating power and preheating time, preheating energy consumption and discharge capacity are comprehensively considered to achieve maximum available energy; reduce capacity loss rate and slow down battery aging, which helps to extend the service life of lithium-ion batteries and reduce replacement frequency and maintenance costs;

[0149] 3) The preheating strategy of the embodiment of the present invention can also be adjusted according to the specific ambient temperature, SOC and discharge current, and is suitable for different low temperature conditions and application scenarios, with strong flexibility and adaptability.

[0150] Reference Figure 2 , a lithium-ion battery preheating system in a low temperature environment, comprising:

[0151] The first module 201 is used to analyze experimental data of lithium-ion batteries and construct an RC equivalent circuit model and a thermal model of lithium-ion batteries;

[0152] The second module 202 is used to perform experimental simulation on the lithium-ion battery RC equivalent circuit model and the lithium-ion battery thermal model to obtain a lithium-ion battery simulation data set;

[0153] The third module 203 is used to optimize the lithium-ion battery simulation data set through a neural network regression prediction model and a multi-objective optimization algorithm to obtain the optimal preheating parameters of the lithium-ion battery;

[0154] The fourth module 204 is used to perform optimal preheating treatment on the lithium-ion battery based on the optimal preheating parameters of the lithium-ion battery to obtain a preheated lithium-ion battery.

[0155] The contents of the above method embodiments are all applicable to the present system embodiments. The functions specifically implemented by the present system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0156] The above is a specific description of the preferred implementation of the present invention, but the invention is not limited to the embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A method for preheating a lithium-ion battery in a low temperature environment, characterized in that: The following steps are involved: Analyze the experimental data of lithium-ion batteries and construct the RC equivalent circuit model and thermal model of lithium-ion batteries; Conduct experimental simulation on the lithium-ion battery RC equivalent circuit model and the lithium-ion battery thermal model to obtain a lithium-ion battery simulation data set; The lithium-ion battery simulation data set is optimized through the neural network regression prediction model and multi-objective optimization algorithm to obtain the optimal preheating parameters of the lithium-ion battery; The lithium-ion battery is optimally preheated based on the optimal preheating parameters of the lithium-ion battery to obtain a preheated lithium-ion battery.

2. The method for preheating a lithium-ion battery in a low temperature environment according to claim 1, characterized in that: The step of analyzing the experimental data of the lithium-ion battery and constructing the RC equivalent circuit model and the thermal model of the lithium-ion battery specifically includes: The internal temperature data of the lithium-ion battery is obtained through the thermocouple and the external temperature data of the lithium-ion battery is obtained through the PI heating film; Based on the internal temperature data and external temperature data of the lithium-ion battery, the preheating power and preheating time of the PI heating film are adjusted, the lithium-ion battery is externally preheated, and the internal and external temperature values ​​of the lithium-ion battery under different ambient temperatures, preheating powers and preheating times are measured to construct a PI heating film preheating polynomial regression equation; Conduct discharge experiments on fully charged lithium-ion batteries at different ambient temperatures and different discharge currents to obtain the value of lithium-ion battery discharge capacity changing with ambient temperature and discharge current; Conduct mixed power pulse characteristic experiments on fully charged lithium-ion batteries at different ambient temperatures, different states of charge, and different discharge currents to obtain lithium-ion battery terminal voltage and current data; The values ​​of lithium-ion battery discharge capacity changing with ambient temperature and discharge current, as well as the lithium-ion battery terminal voltage and current data are fitted to construct the lithium-ion battery RC equivalent circuit model and lithium-ion battery thermal model.

3. The method for preheating a lithium-ion battery in a low temperature environment according to claim 2, characterized in that: The step of fitting the values ​​of the lithium-ion battery discharge capacity changing with the ambient temperature and the discharge current and the lithium-ion battery terminal voltage and current data to construct the lithium-ion battery RC equivalent circuit model and the lithium-ion battery thermal model specifically includes: The terminal voltage and current data of lithium-ion batteries are fitted by Matlab's cftool toolbox to construct a polynomial regression equation for the terminal voltage and current data of lithium-ion batteries. According to the change of the discharge capacity of the lithium-ion battery with the ambient temperature and the discharge current, the polarization internal resistance value, the polarization capacitance value and the ohmic internal resistance value of the lithium-ion battery are determined; Combining the polarization internal resistance value, polarization capacitance value and ohmic internal resistance value of the lithium-ion battery, a polynomial regression equation for the change of the discharge capacity of the lithium-ion battery with the ambient temperature and the discharge current is constructed; According to the polynomial regression equation about the terminal voltage and current data of the lithium-ion battery and the polynomial regression equation about the value of the discharge capacity of the lithium-ion battery changing with the ambient temperature and the discharge current, combined with Kirchhoff's voltage and current law, the RC equivalent circuit model of the lithium-ion battery is constructed; According to the polynomial regression equation about the terminal voltage and current data of the lithium-ion battery and the polynomial regression equation about the value of the discharge capacity of the lithium-ion battery changing with the ambient temperature and the discharge current, combined with Joule's law, a thermal model of the lithium-ion battery is constructed.

4. The method for preheating a lithium-ion battery in a low temperature environment according to claim 3, characterized in that: The step of performing experimental simulation on the lithium-ion battery RC equivalent circuit model and the lithium-ion battery thermal model to obtain a lithium-ion battery simulation data set specifically includes: Construct a percentage model of capacity attenuation of lithium-ion batteries based on the battery semi-empirical aging model formula; The lithium-ion battery RC equivalent circuit model, lithium-ion battery thermal model, and lithium-ion battery capacity attenuation percentage model are coupled with the PI heating film preheating polynomial regression equation and input into the COMSOL simulation software for experimental simulation to obtain a lithium-ion battery simulation data set.

5. A lithium-ion battery preheating method in a low temperature environment according to claim 4, characterized in that: The step of optimizing the lithium-ion battery simulation data set by using a neural network regression prediction model and a multi-objective optimization algorithm to obtain the optimal preheating parameters of the lithium-ion battery specifically includes: The neural network regression prediction model is trained based on the lithium-ion battery simulation data set to obtain a trained neural network regression prediction model; Obtaining lithium-ion battery input data and predicting through the trained neural network regression prediction model to obtain the percentage value of the discharge capacity of the lithium-ion battery and the capacity attenuation of the lithium-ion battery; Calculate the preheating energy consumption and available energy of the ion battery, define the available energy weight factor, aging loss rate weight factor and preheating energy consumption weight factor, and construct a weighted objective function; The discharge capacity of the lithium-ion battery, the percentage value of the capacity attenuation of the lithium-ion battery and the weighted objective function are optimized by a multi-objective optimization algorithm to obtain the optimal preheating parameters of the lithium-ion battery.

6. A lithium-ion battery preheating method in a low temperature environment according to claim 5, characterized in that: The step of training the neural network regression prediction model based on the lithium-ion battery simulation data set to obtain the trained neural network regression prediction model specifically includes: Inputting a lithium-ion battery simulation data set into a neural network regression prediction model, wherein the neural network regression prediction model includes a sparrow search algorithm optimization model, a convolutional neural network, and a long short-term memory network; Based on the sparrow search algorithm optimization model, the parameters of the neural network regression prediction model are optimized to obtain the optimized neural network regression prediction model; Based on the convolutional neural network of the optimized neural network regression prediction model, local feature extraction processing is performed on the lithium-ion battery simulation data set to obtain the features of the lithium-ion battery simulation data set; Based on the long short-term memory network of the optimized neural network regression prediction model, the time series feature extraction processing is performed on the characteristics of the lithium-ion battery simulation data set to obtain the time series features of the lithium-ion battery simulation data set; The time series characteristics of the lithium-ion battery simulation dataset are used as verification dataset indicators to verify the optimized neural network regression prediction model and construct the trained neural network regression prediction model.

7. A lithium-ion battery preheating method in a low temperature environment according to claim 6, characterized in that: The step of obtaining the lithium-ion battery input data and predicting through the trained neural network regression prediction model to obtain the percentage value of the lithium-ion battery discharge capacity and the lithium-ion battery capacity decay specifically includes: Based on the PI heating film preheating polynomial regression equation, the external temperature and internal temperature values ​​of the lithium-ion battery after preheating at different preheating powers, different preheating times, and different ambient temperatures are obtained; Construct lithium-ion battery input data by combining the external temperature and internal temperature values ​​of lithium-ion batteries after preheating at different preheating powers, different preheating times, and different ambient temperatures, ambient temperature, state of charge, and lithium-ion battery discharge current values; The input data of the lithium-ion battery is predicted by the trained neural network regression prediction model to obtain the percentage value of the discharge capacity of the lithium-ion battery and the capacity attenuation of the lithium-ion battery.

8. A lithium-ion battery preheating method in a low temperature environment according to claim 7, characterized in that: The step of performing multi-objective optimization processing on the discharge capacity of the lithium-ion battery, the percentage value of the capacity attenuation of the lithium-ion battery and the weighted objective function by a multi-objective optimization algorithm to obtain the optimal preheating parameters of the lithium-ion battery specifically includes: The weighted objective function is solved by a multi-objective optimization algorithm to obtain the Pareto optimal solution; Determine target values ​​based on the Pareto optimal solution, wherein the target values ​​include available energy of the lithium-ion battery, capacity loss rate of the lithium-ion battery, and preheating energy consumption of the lithium-ion battery; Perform a maximum available energy index search process on the target value to obtain the optimal preheating power and preheating time; The lithium-ion battery discharge capacity, lithium-ion battery capacity loss rate, lithium-ion battery preheating energy consumption and lithium-ion battery available energy under the optimal preheating power and preheating time are calculated through auxiliary functions. The Pareto frontier diagram is drawn in combination with the discharge capacity of the lithium-ion battery and the percentage value of the lithium-ion battery capacity attenuation to obtain the optimal preheating parameters of the lithium-ion battery.

9. A lithium-ion battery preheating system in a low temperature environment, characterized in that: Includes the following modules: The first module is used to analyze the experimental data of lithium-ion batteries and build the RC equivalent circuit model and thermal model of lithium-ion batteries; The second module is used to perform experimental simulation on the lithium-ion battery RC equivalent circuit model and the lithium-ion battery thermal model to obtain a lithium-ion battery simulation data set; The third module is used to optimize the lithium-ion battery simulation data set through a neural network regression prediction model and a multi-objective optimization algorithm to obtain the optimal preheating parameters of the lithium-ion battery; The fourth module is used to perform optimal preheating treatment on the lithium-ion battery based on the optimal preheating parameters of the lithium-ion battery to obtain a preheated lithium-ion battery.

Images