Battery temperature early warning method and device for new energy vehicle charging process

Abstract

The application discloses a battery temperature early warning method and device for a new energy vehicle charging process, and relates to the technical field of vehicle charging, and comprises the following steps: acquiring a charging current value, a direct current internal resistance value and a cooling system operation parameter at a current time in a charging process; calculating a charging current rate factor, calculating a battery internal resistance aging offset coefficient through the deviation of the direct current internal resistance value and a factory standard direct current internal resistance value; acquiring a cooling efficiency attenuation degree through intelligent analysis; calculating and acquiring a battery comprehensive thermal risk index based on the charging current rate factor, the battery internal resistance aging offset coefficient and the cooling efficiency attenuation degree; dynamically adjusting a warning temperature threshold value of the current charging process to acquire a dynamic warning temperature threshold value; and triggering battery temperature warning when the current temperature value of the battery is greater than or equal to the dynamic warning temperature threshold value. The technical problem that the prior art triggers warning according to a preset fixed temperature threshold, resulting in low warning accuracy, high false alarm rate and poor user experience, is solved.

Description

Technical Field

[0001] This invention relates to the field of automotive charging technology, and more specifically to a battery temperature early warning method and device for the charging process of new energy vehicles. Background Technology

[0002] In the new energy vehicle industry, power batteries are the core energy source, and their charging safety is receiving increasing attention. During charging, lithium-ion power batteries experience charging and discharging currents of up to hundreds of amperes, which causes a rapid rise in internal battery temperature. If battery temperature cannot be effectively monitored, it will accelerate battery performance degradation and may even lead to safety accidents such as thermal runaway. Statistics show that thermal runaway accidents caused by lithium-ion battery charging account for about one-third of all accidents involving new energy vehicles. Therefore, real-time monitoring and risk warning of battery temperature during charging are crucial for ensuring the safe operation of new energy vehicles.

[0003] Currently, existing technologies primarily rely on preset fixed temperature thresholds for warning triggering. However, during battery use, the impact of battery internal resistance aging on thermal risks is not considered. When the battery temperature briefly exceeds the threshold due to interference factors but actually shows a downward trend, existing methods still trigger unnecessary warnings, leading to frequent false alarms and affecting user experience and warning reliability. Summary of the Invention

[0004] This invention provides a battery temperature early warning method and device for the charging process of new energy vehicles, which solves the technical problems of low early warning accuracy, high false alarm rate and poor user experience caused by early warning triggering based on a preset fixed temperature threshold in the prior art.

[0005] In a first aspect, the present invention provides a battery temperature early warning method for the charging process of new energy vehicles, the method comprising:

[0006] Obtain the charging current value, DC internal resistance value, and cooling system operating parameters of the battery at the current moment during the charging process;

[0007] Based on the ratio of the charging current value to the battery rated capacity value, the charging current rate factor is calculated, and based on the deviation of the DC internal resistance value from the factory standard DC internal resistance value, the battery internal resistance aging offset coefficient is calculated.

[0008] The operating parameters of the cooling system are intelligently analyzed to obtain the cooling efficiency decay rate;

[0009] Based on the charging current rate factor, the battery internal resistance aging offset coefficient, and the cooling efficiency decay, the battery comprehensive thermal risk index is calculated and obtained.

[0010] Based on the battery comprehensive thermal risk index, the warning temperature threshold of the current charging process is dynamically adjusted to obtain the dynamic warning temperature threshold.

[0011] When the current battery temperature is greater than or equal to the dynamic warning temperature threshold, a battery temperature warning is triggered.

[0012] Secondly, the present invention provides a battery temperature warning device for use in the charging process of new energy vehicles, the device comprising:

[0013] The charging parameter acquisition module is used to acquire the charging current value, DC internal resistance value and cooling system operating parameters of the battery at the current moment during the charging process;

[0014] The battery internal resistance aging offset coefficient calculation module is used to calculate the charging current rate factor based on the ratio of the charging current value to the battery rated capacity value, and to calculate the battery internal resistance aging offset coefficient based on the deviation between the DC internal resistance value and the factory standard DC internal resistance value.

[0015] The intelligent analysis module for cooling system operating parameters is used to intelligently analyze the operating parameters of the cooling system and obtain the cooling efficiency decay rate.

[0016] The battery comprehensive thermal risk index acquisition module is used to calculate and acquire the battery comprehensive thermal risk index based on the charging current rate factor, the battery internal resistance aging offset coefficient, and the cooling efficiency decay.

[0017] The warning temperature threshold dynamic adjustment module is used to dynamically adjust the warning temperature threshold of the current charging process according to the battery comprehensive thermal risk index, and obtain the dynamic warning temperature threshold.

[0018] The battery temperature warning module is used to trigger a battery temperature warning when the current battery temperature value is greater than or equal to the dynamic warning temperature threshold.

[0019] One or more technical solutions provided in this invention have at least the following technical effects or advantages:

[0020] The present invention provides a battery temperature early warning method and device for the charging process of new energy vehicles. First, by acquiring charging current values, DC internal resistance values, and cooling system operating parameters, a multi-dimensional data foundation covering heat generation intensity, battery aging status, and heat dissipation capacity is provided for battery thermal risk assessment. Second, a charging current rate factor is calculated based on the ratio of the charging current value to the rated capacity value, and a battery internal resistance aging offset coefficient is calculated based on the deviation of the DC internal resistance value from the factory standard DC internal resistance value, achieving a quantitative characterization of charging rate and battery aging degree. Third, the cooling efficiency decay degree is obtained through intelligent analysis of cooling system operating parameters, incorporating the impact of cooling system performance degradation on thermal safety into the assessment system. Furthermore, a comprehensive battery thermal risk index is calculated based on the charging current rate factor, battery internal resistance aging offset coefficient, and cooling efficiency decay degree, achieving a multi-factor comprehensive quantification of battery thermal safety status. Further, the early warning temperature threshold is dynamically adjusted according to the comprehensive battery thermal risk index, enabling the early warning sensitivity to adaptively adjust with changes in thermal risk. Finally, an early warning is triggered when the current battery temperature reaches or exceeds the dynamic early warning temperature threshold, achieving an accurate and timely response to battery thermal risks. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the battery temperature early warning method for the charging process of new energy vehicles provided by the present invention.

[0022] Figure 2 This is a logical schematic diagram of the battery temperature early warning method for the charging process of new energy vehicles provided by the present invention.

[0023] Figure 3 This is a schematic diagram of the battery temperature warning device for the charging process of new energy vehicles provided by the present invention.

[0024] In the attached diagram, the components represented by each number are as follows:

[0025] Module 11 for acquiring charging parameters; Module 12 for calculating battery internal resistance aging offset coefficient; Module 13 for intelligent analysis of cooling system operating parameters; Module 14 for acquiring battery comprehensive thermal risk index; Module 15 for dynamic adjustment of warning temperature threshold; Module 16 for battery temperature warning. Detailed Implementation

[0026] This invention provides a battery temperature early warning method and device for the charging process of new energy vehicles, which solves the technical problems of low early warning accuracy, high false alarm rate and poor user experience caused by early warning triggering based on a preset fixed temperature threshold in the prior art.

[0027] The present invention will now be described in detail with reference to the accompanying drawings.

[0028] Example 1, as Figure 1 , Figure 2 As shown, the present invention provides a battery temperature early warning method for the charging process of new energy vehicles, the method comprising:

[0029] S100: Acquires the charging current value, DC internal resistance value, and cooling system operating parameters of the battery at the current moment during the charging process;

[0030] In this embodiment of the invention, the charging current value is the instantaneous amplitude of the DC current output by the charging equipment to the battery during the charging process of the new energy vehicle power battery; the DC internal resistance value is the internal equivalent resistance exhibited by the battery under DC charging and discharging conditions; and the cooling system operating parameters are the operating state variables related to the cooling function in the thermal management system of the new energy vehicle battery.

[0031] Specifically, the system acquires the DC current value in the charging circuit at the current moment through a current sensor. Simultaneously, the system obtains the DC internal resistance value at the current moment through an internal resistance estimation module. The DC internal resistance value can be obtained using an online estimation method, utilizing the terminal voltage and charging current values ​​from multiple consecutive sampling moments within a preset time window, and employing a parameter identification algorithm for real-time estimation. Meanwhile, temperature sensors installed on the coolant inlet and outlet pipes respectively acquire the current coolant inlet and outlet temperatures, and a flow meter installed in the coolant circulation pipe acquires the current coolant flow rate. These charging current values, DC internal resistance values, and cooling system operating parameters collectively constitute the input data basis for subsequent thermal risk assessment.

[0032] Step S100 in the method of this embodiment of the invention includes:

[0033] The charging current value at the current moment is obtained as the charging current value, and the estimated DC internal resistance value at the current moment is obtained as the DC internal resistance value. The estimated DC internal resistance value is obtained by using a recursive least squares method including a forgetting factor to estimate the terminal voltage value and charging current value of the battery at the current moment and within a preset time window before the current moment, based on the terminal voltage value and charging current value of the battery at multiple consecutive sampling moments within the current moment and the preset time window before the current moment.

[0034] The current coolant inlet temperature, coolant outlet temperature, and coolant flow rate are obtained and used as the operating parameters of the cooling system.

[0035] In this embodiment of the invention, firstly, the charging current value at the current moment is acquired as the charging current value. This charging current value can be directly acquired in real-time by the current sensor of the battery management system. The battery management system is typically equipped with a high-precision Hall effect current sensor or a shunt resistor current sensor, capable of continuously acquiring the real-time charging current of the battery during the charging process with a millisecond-level sampling period. Subsequently, the estimated DC internal resistance value at the current moment is acquired as the DC internal resistance value. This estimated DC internal resistance value is based on the battery's terminal voltage and charging current values ​​at multiple consecutive sampling moments within a preset time window prior to the current moment, and is estimated using a recursive least squares method including a forgetting factor.

[0036] Recursive least squares (RLS) is a recursive parameter estimation algorithm used to track time-varying parameters online. The forgetting factor is a crucial parameter in RLS, typically ranging from 0.95 to 0.99. It applies an exponential weighted decay to historical data, ensuring that data from further back contributes less to the parameter estimation and enabling timely tracking of the time-varying characteristics of battery parameters.

[0037] Specifically, the system acquires the terminal voltage and charging current values ​​at the current moment and multiple consecutive sampling moments within a preset time window prior to the current moment. The terminal voltage is the voltage between the positive and negative terminals of the battery, which can be obtained through a voltage sensor in the battery management system; the charging current is obtained through a current sensor. The preset time window can be set according to the actual application scenario; for example, it can be set to 60 seconds, with a sampling interval of 1 second, thus containing 60 sampling moments of terminal voltage and charging current values. Based on the acquired terminal voltage and charging current values, an equivalent circuit model of the battery is constructed. For example, a first-order RC equivalent circuit model or a second-order RC equivalent circuit model can be constructed. Then, a recursive least squares method including a forgetting factor is used to identify the parameters in the equivalent circuit model online, and the DC internal resistance value of the battery is extracted from the identified parameters.

[0038] In one feasible embodiment, the battery's rated voltage is 3.7V, and the current charging current is 100A. Within a 60-second time window prior to the current moment, the terminal voltage and charging current values ​​are continuously collected at 1-second sampling intervals for 60 sampling moments. These 60 sets of terminal voltage and charging current values ​​are input into a pre-established second-order RC equivalent circuit model, and parameter identification is performed using a recursive least squares method with a forgetting factor λ=0.98. At each sampling moment, the recursive least squares method constructs an observation vector containing the current charging current value and its historical values; calculates the gain vector; updates the parameter estimation vector; and updates the covariance matrix. After 60 recursive iterations, the parameter estimation vector converges to a stable value. The parameter term corresponding to the DC internal resistance is extracted from the parameter estimation vector to obtain the estimated DC internal resistance value at the current moment. For example, if the identified DC internal resistance value is 0.015Ω, this value is used as the DC internal resistance value at the current moment.

[0039] Secondly, the current coolant inlet temperature, coolant outlet temperature, and coolant flow rate are obtained as operating parameters of the cooling system.

[0040] Specifically, the power batteries of new energy vehicles typically employ liquid cooling for thermal management. Coolant circulates within cooling pipes, carrying away the heat generated during battery charging. The coolant inlet temperature refers to the temperature of the coolant before it enters the battery pack, while the coolant outlet temperature refers to the temperature of the coolant after it flows out of the battery pack. Both temperatures can be obtained in real-time using temperature sensors installed at the inlet and outlet of the coolant piping. The coolant flow rate refers to the volume or mass of coolant flowing through the cooling pipes per unit time, which can also be obtained in real-time using flow sensors installed within the coolant piping.

[0041] In this embodiment of the invention, the existing sensors and data acquisition capabilities of the battery management system are utilized to comprehensively acquire charging current values, DC internal resistance values, and cooling system operating parameters. This provides complete data input from both electrical and thermal management characteristics to subsequent calculations of the battery's comprehensive thermal risk index, establishing a data foundation for dynamic adjustment of the early warning temperature threshold.

[0042] S200: Calculate the charging current rate factor based on the ratio of the charging current value to the battery rated capacity value, and calculate the battery internal resistance aging offset coefficient based on the deviation between the DC internal resistance value and the factory standard DC internal resistance value.

[0043] In this embodiment of the invention, the rated capacity of the battery refers to the amount of electricity that the battery can release under standard test conditions as specified by the battery manufacturer; the charging current rate factor is the ratio of the charging current value to the rated capacity value of the battery; the factory standard DC internal resistance value refers to the DC internal resistance value measured under standard test conditions when the battery is in a healthy state at the time of manufacture; the internal resistance deviation is the increment of the current DC internal resistance value relative to the factory standard DC internal resistance value; and the battery internal resistance aging offset coefficient is the ratio of the internal resistance deviation to the factory standard DC internal resistance value.

[0044] Specifically, firstly, the battery's rated capacity is read from its storage parameters. The current charging current is then divided by the rated capacity to calculate the charging current rate factor. Simultaneously, the factory standard DC internal resistance is read from the battery's factory parameter record. The current DC internal resistance is subtracted from the factory standard DC internal resistance to obtain the internal resistance deviation. Finally, the internal resistance deviation is divided by the factory standard DC internal resistance to calculate the battery's internal resistance aging offset coefficient.

[0045] Step S200 in the method of this embodiment of the invention includes:

[0046] Obtain the rated capacity value of the battery, divide the charging current value by the rated capacity value, and obtain the charging current rate factor;

[0047] Obtain the factory standard DC internal resistance value of the battery at the time of manufacture, and subtract the factory standard DC internal resistance value from the DC internal resistance value to obtain the internal resistance deviation.

[0048] Divide the internal resistance deviation by the factory standard DC internal resistance value to obtain the battery internal resistance aging offset coefficient.

[0049] In this embodiment of the invention, firstly, the rated capacity of the battery is obtained. Then, the charging current value is divided by the rated capacity value to obtain the charging current rate factor. The rated capacity of the battery is an indicator characterizing the battery's ability to release or store electrical energy under standard conditions. It is typically determined by the battery manufacturer through standard charge-discharge tests and marked on the battery nameplate at the time of shipment. The charging current rate factor is a dimensionless index that measures the relative magnitude of the battery's charging rate to its rated capacity. It is used to characterize the relative level of the current intensity under the current charging condition relative to the battery's rated capacity. Its value ranges from 0 to 1; the closer it is to 1, the higher the rate of the current charging current relative to the battery's rated capacity, and the greater the ohmic heat generation power of the battery per unit time.

[0050] Specifically, the formula for calculating the charging current rate factor is as follows:

[0051] Charging current rate factor = current charging current value / battery rated capacity value. The dimension of the charging current rate factor is the hourly rate, usually represented by C. The larger the value, the higher the charging rate and the greater the heat generation rate of the battery per unit time.

[0052] Assuming a new energy vehicle's power battery has a rated capacity of 150Ah, and the battery management system currently detects a charging current of 225A, then the charging current rate factor = 225 / 150 = 1.5, indicating that the current charging rate is a relatively low 1.5C.

[0053] Secondly, obtain the battery's factory-standard DC internal resistance value. This factory-standard DC internal resistance value is the DC internal resistance value measured under standard testing conditions when the battery is in its factory healthy state. It is obtained by the battery manufacturer through factory testing and recorded in the battery's technical specifications. Standard testing conditions are typically room temperature (25°C) and 50% state of charge. Internal resistance deviation is the difference between the currently estimated DC internal resistance value and the factory-standard DC internal resistance value. Its dimensions are consistent with the DC internal resistance value, and it is used to characterize the absolute deviation of the battery's internal resistance relative to its factory healthy state. The internal resistance deviation can be obtained by subtracting the factory-standard DC internal resistance value from the current DC internal resistance value: Internal resistance deviation = DC internal resistance value - Factory-standard DC internal resistance value.

[0054] When the internal resistance deviation is positive, it indicates that the battery's internal resistance has increased relative to its factory condition, indicating that the battery has aged; the larger the internal resistance deviation value, the more severe the aging.

[0055] Finally, the internal resistance deviation is divided by the factory standard DC internal resistance value to obtain the battery internal resistance aging deviation coefficient. The battery internal resistance aging deviation coefficient is a dimensionless relative index obtained by normalizing the internal resistance deviation to the factory standard DC internal resistance value. It is used to characterize the degree of relative deviation of the battery internal resistance from its factory health state.

[0056] Specifically, the battery internal resistance aging deviation coefficient = internal resistance deviation / factory standard DC internal resistance value. The value range of the battery internal resistance aging deviation coefficient is [0,1]. The closer it is to 1, the more severe the increase in battery internal resistance compared to its factory state, and the greater the heat generated by the battery under the same charging current; the closer it is to 0, the lower the degree of battery internal resistance aging. Since the battery internal resistance generally shows a monotonically increasing trend throughout the battery's lifespan, the battery internal resistance aging deviation coefficient is always positive under normal operating conditions. If the battery internal resistance aging deviation coefficient shows a slight negative value due to measurement noise or estimation error, it can be truncated to 0 to conform to physical reality.

[0057] In this embodiment of the invention, the charging current value is compared with the rated capacity value to obtain the charging current rate factor, thereby achieving a normalized characterization of the charging rate between batteries of different capacities; the ratio of the internal resistance deviation to the factory standard DC internal resistance value is calculated to obtain the battery internal resistance aging offset coefficient, quantifying the degree of aging growth of the battery internal resistance relative to the factory condition, and providing a data basis for subsequent comprehensive thermal risk assessment.

[0058] S300: Intelligently analyze the operating parameters of the cooling system to obtain the cooling efficiency decay rate;

[0059] In this embodiment of the invention, the cooling performance decay is the degree of decline of the actual heat dissipation capacity of the cooling system under the current operating state relative to its rated heat dissipation capacity; intelligent analysis is the process of using a pre-built machine learning model or deep learning model to extract features and recognize patterns from the time-series data of the cooling system operating parameters, thereby quantitatively predicting the cooling performance decay.

[0060] Over long-term use, the actual heat dissipation capacity of a cooling system will gradually fall below its factory rated capacity due to factors such as coolant performance degradation, scaling or blockage inside the cooling pipes, reduced flow rate caused by coolant pump wear, and radiator surface contamination. Therefore, intelligent analysis of the cooling system's operating parameters is necessary to determine the degree of cooling performance degradation.

[0061] Specifically, the coolant inlet temperature, coolant outlet temperature, and coolant flow rate values ​​from multiple consecutive sampling times within a preset time window prior to the current time are each used to construct three time series. These multivariate time series input samples are then fed into a pre-built cooling performance degradation prediction model. The model ultimately outputs a predicted value for the degree of cooling performance degradation.

[0062] Step S300 in the method of this embodiment of the invention includes:

[0063] The time series consisting of the coolant inlet temperature, coolant outlet temperature, and coolant flow rate values ​​of multiple consecutive sampling times within the current time and the preset time window before the current time is input into the pre-constructed cooling performance decay prediction model to obtain the cooling performance decay degree output by the cooling performance decay prediction model.

[0064] The pre-construction of the cooling performance degradation prediction model includes:

[0065] Collect multiple sets of historical cooling system operation data. Each set of data includes the start time of a historical period and the historical coolant inlet temperature, historical coolant outlet temperature, and historical coolant flow rate values ​​of multiple consecutive sampling times within a preset time window before the start time of the historical period, as well as the corresponding sample cooling efficiency decay.

[0066] A cooling performance degradation prediction model is constructed, which includes a temporal convolutional layer and a trend prediction layer. The temporal convolutional layer is used to perform multi-layer convolutional encoding on the temporal sequences of coolant inlet temperature, coolant outlet temperature, and coolant flow rate to extract the temporal variation features of temperature difference and flow rate. The trend prediction layer is used to receive the output of the temporal convolutional layer and output the cooling performance degradation degree after fully connected mapping.

[0067] The cooling performance attenuation prediction model is trained using the historical cooling system operation data, with the goal of minimizing the deviation between the output cooling performance attenuation degree and the sample cooling performance attenuation degree, until the deviation meets the preset convergence condition, thus obtaining the pre-constructed cooling performance attenuation prediction model.

[0068] In this embodiment of the invention, a time series consisting of the coolant inlet temperature, coolant outlet temperature, and coolant flow rate values ​​collected at multiple consecutive sampling times within a preset time window prior to the current time is input into a pre-constructed cooling performance attenuation prediction model to obtain the cooling performance attenuation degree output by the model. The cooling performance attenuation degree is a dimensionless index characterizing the degree of attenuation of the cooling system relative to its rated heat dissipation capacity under the current state. Its value is between 0 and 1, with a larger value indicating more severe cooling performance attenuation. The preset time window can be set according to the actual application scenario; for example, it can be set to 120 seconds, and the sampling interval can be set to 1 second. Therefore, the preset time window contains 120 sampling times of coolant inlet temperature, coolant outlet temperature, and coolant flow rate values.

[0069] For example, assuming the current time is t0, the preset time window is 120 seconds, and the sampling interval is 1 second, then within the 120-second time window before the current time t0, a total of 120 sets of cooling system operating parameters are collected. Each set of parameters includes the coolant inlet temperature, coolant outlet temperature, and coolant flow rate at a sampling time, obtaining 120 sampling times from the earliest time to the current time. These 120 sets of data are arranged in chronological order, forming a 120×3 multivariate time-series sequence, which serves as the input to the cooling performance degradation prediction model. After forward computation, the cooling performance degradation prediction model outputs a floating-point number between 0 and 1, representing the cooling performance degradation degree at the current time.

[0070] The pre-construction of the cooling performance degradation prediction model includes:

[0071] First, multiple sets of historical cooling system operation data are collected. Each set of data includes the starting time of a historical period and the historical coolant inlet temperature, historical coolant outlet temperature, and historical coolant flow rate values ​​at multiple consecutive sampling times within a preset time window before the starting time of the historical period, as well as the corresponding sample cooling efficiency decay.

[0072] Specifically, a large amount of historical charging process data can be extracted from the cloud data platform or on-board data recording system of new energy vehicles. For example, cooling system operation data from 500 complete charging processes over the past year can be collected. For each charging process, a specific moment is selected as the starting point of the historical time period. Time-series data of coolant inlet temperature, coolant outlet temperature, and coolant flow rate within a preset time window before that starting point, such as 120 seconds, are extracted. Based on data within a preset future time period after that starting point, such as 60 seconds, the corresponding sample cooling efficiency decay is calculated. This results in a dataset containing 500 training samples. Each sample includes historical coolant inlet temperature, historical coolant outlet temperature, and historical coolant flow rate, along with a label indicating the corresponding sample cooling efficiency decay.

[0073] Secondly, a cooling performance degradation prediction model is constructed, which includes a temporal convolutional layer and a trend prediction layer. The temporal convolutional layer is a feature extraction layer built based on a temporal convolutional network. By performing convolution operations in the time dimension, it can capture local temporal patterns and long-term dependencies in time-series data. The temporal convolutional layer is used to perform multi-layer convolutional encoding on the temporal sequences of coolant inlet temperature, coolant outlet temperature, and coolant flow rate. By stacking multiple convolutional layers, features at different time scales are extracted layer by layer. Lower-level convolutional kernels focus on short-term local change patterns, while higher-level convolutional kernels focus on long-term trend change patterns, extracting temporal variation features of temperature difference and flow rate.

[0074] The time-series variation characteristic of temperature difference refers to the pattern of change between the outlet and inlet temperatures of the coolant over time. The temperature difference can be calculated from the difference between the outlet and inlet temperatures. This characteristic reflects the changing trend of the cooling system's ability to remove heat. The time-series variation characteristic of flow rate refers to the pattern of change of the coolant flow rate over time. This characteristic reflects the changing trend of the cooling system's circulation efficiency.

[0075] Additionally, the trend prediction layer contains one or more fully connected network modules that receive the output of the temporal convolutional layer and output the cooling performance decay after fully connected mapping. The high-dimensional feature vectors extracted by the temporal convolutional layer are mapped to a one-dimensional output space through linear transformation and nonlinear activation, outputting the predicted value of the cooling performance decay.

[0076] Specifically, assume the input temporal sequence has a dimension of 120×3. The first temporal convolutional layer contains 64 kernels of size 3 with a stride of 1, outputting a feature map of dimension 120×64 after ReLU activation. The second temporal convolutional layer contains 128 kernels of size 3 with a stride of 1, outputting a feature map of dimension 120×128 after ReLU activation. The third temporal convolutional layer contains 256 kernels of size 3 with a stride of 1, outputting a feature map of dimension 120×256 after ReLU activation. Subsequently, the feature map output from the third temporal convolutional layer is subjected to global average pooling along the temporal dimension, resulting in a global feature vector of dimension 1×256. The trend prediction layer receives the 256-dimensional global feature vector as input and contains two fully connected layers. The first layer contains 128 neurons, and the second layer contains 1 neuron. After mapping by the two fully connected layers, it outputs a scalar value, namely the cooling efficiency decay. The output cooling efficiency decay is constrained to the [0,1] interval by the Sigmoid activation function to ensure the reasonableness of the output value.

[0077] Finally, the cooling performance degradation prediction model is trained using historical cooling system operation data. The goal is to minimize the deviation between the output cooling performance degradation degree and the sample cooling performance degradation degree until the deviation meets the preset convergence condition, thus obtaining the pre-constructed cooling performance degradation prediction model.

[0078] Specifically, the 500 collected training samples were divided into training and validation sets in an 8:2 ratio. The training hyperparameters were set as follows: batch size of 32, learning rate of 0.001, maximum number of training epochs of 200, and patience value of 20 for the early stopping strategy. In each training epoch, 32 samples were sequentially taken from the training set. The input time series sequence of each sample was input into the cooling performance decay prediction model for forward propagation calculation to obtain the predicted cooling performance decay. The mean squared error loss between the predicted value and the sample cooling performance decay was calculated. The gradient of the loss with respect to all trainable parameters in the model was calculated using the backpropagation algorithm. The Adam optimizer was used to update all trainable parameters based on the calculated gradient and the preset learning rate.

[0079] After each training round, the average loss is calculated on the validation set. In the first 30 training rounds, the average loss on the validation set rapidly decreases from 0.25 to 0.05; it decreases to 0.03 by round 50; and to 0.018 by round 80. Thereafter, the validation set loss stabilizes around 0.018 between rounds 80 and 100, without further significant decreases, triggering an early stopping strategy and halting training. At this point, the model parameters are saved, resulting in a pre-built prediction model for cooling performance degradation.

[0080] The acquired input data is then fed into the cooling performance degradation prediction model, which outputs the cooling performance degradation degree. The cooling performance degradation degree is a dimensionless parameter normalized to the interval [0,1], used to characterize the degree of degradation of the cooling system relative to its rated heat dissipation capacity under actual operating conditions. The closer the cooling performance degradation degree is to 1, the more severe the degradation of the cooling system's actual heat dissipation capacity relative to its rated heat dissipation capacity under current operating conditions, indicating a significant deterioration in the cooling system's heat dissipation performance. Conversely, the closer the cooling performance degradation degree is to 0, the closer the actual heat dissipation capacity of the cooling system is to its rated heat dissipation capacity, indicating a slight degradation in the cooling system's heat dissipation performance, essentially maintaining its original factory condition.

[0081] In step S300 of the method of this embodiment of the invention, obtaining the sample cooling efficiency attenuation degree includes:

[0082] The difference between the historical coolant outlet temperature value and the historical coolant inlet temperature value at each sampling moment in the cooling system operating parameters within a preset future time period after obtaining the historical time period start time is used as the inlet and outlet temperature difference.

[0083] Calculate the product of the inlet and outlet temperature difference at each sampling time, the historical coolant flow rate at the corresponding sampling time, and the specific heat capacity of the coolant used, to obtain the actual heat dissipation power value at each sampling time.

[0084] The arithmetic mean of the actual heat dissipation power values ​​at all sampling times within the preset future time period is taken as the future average heat dissipation power value;

[0085] Based on the future average heat dissipation power value and the rated heat dissipation power value of the cooling system, the sample cooling efficiency decay rate is calculated.

[0086] In this embodiment of the invention, firstly, the difference between the historical coolant outlet temperature and the historical coolant inlet temperature at each sampling moment in the cooling system operating parameters within a preset future time period after the historical time period start time is obtained. This difference is used as the inlet-outlet temperature difference, i.e., inlet-outlet temperature difference = historical coolant outlet temperature - historical coolant inlet temperature. The historical time period start time is a selected reference moment from the historical cooling system operating data, serving as the starting point of the preset future time period. The preset future time period is a continuous time interval after the historical time period start time, and its duration can be set according to the actual application scenario, for example, it can be set to 60 seconds. Within this preset future time period, the coolant inlet temperature and coolant outlet temperature are collected at a preset sampling interval, for example, 1 second. Each sampling moment corresponds to a set of temperature measurements. The inlet-outlet temperature difference refers to the difference between the coolant outlet temperature and the coolant inlet temperature at the same sampling moment, measured in degrees Celsius. It reflects the degree of temperature rise caused by the heat carried away from the battery system by the coolant as it flows through the battery pack. The larger the inlet-outlet temperature difference, the more heat the coolant carries away from the battery system per unit time, and the better the heat dissipation effect of the cooling system.

[0087] Secondly, the actual heat dissipation power value at each sampling time is obtained by multiplying the inlet and outlet temperature difference, the historical coolant flow rate at the corresponding sampling time, and the specific heat capacity of the coolant used. That is, the actual heat dissipation power value = inlet and outlet temperature difference × historical coolant flow rate at the corresponding sampling time × specific heat capacity of the coolant used. The obtained actual heat dissipation power value is the rate at which the cooling system removes heat from the battery system at a certain sampling time.

[0088] Historical coolant flow rate values ​​are the volumetric or mass flow rates of coolant flowing through the cooling pipes at the corresponding sampling time. When using volumetric flow rate, it needs to be multiplied by the coolant density to convert it to mass flow rate. Specific heat capacity of coolant is the amount of heat required to raise the temperature of a unit mass of coolant by a unit temperature. Different types of coolant have different specific heat capacities, which can be obtained from technical data provided by the coolant manufacturer or measured under laboratory conditions.

[0089] Specifically, for the Lth sampling moment within a preset future time period, the historical coolant flow rate, the inlet and outlet temperature difference, and the specific heat capacity of the coolant used are obtained. These three parameters are multiplied to obtain the actual heat dissipation power value for that sampling moment. This calculation is performed sequentially for each of the L sampling moments within the preset future time period to obtain a sequence of actual heat dissipation power values.

[0090] Assume that at a certain sampling moment, the coolant flow rate is 0.5 kg / s, and the coolant used is a 50% (v / v) ethylene glycol-water mixture with a specific heat capacity of 3.3 × 10³ J / (kg·K). At this moment, the coolant inlet temperature is 25℃, and the coolant outlet temperature is 32℃. Therefore, the inlet-outlet temperature difference is 32 - 25 = 7℃. The actual heat dissipation power at this sampling moment is 0.5 × 3.3 × 10³ × 7 = 11.55 kW, indicating that at this sampling moment, the cooling system removes heat from the battery system at a power of 11.55 kW.

[0091] Next, the arithmetic mean of the actual heat dissipation power values ​​at all sampling times within the preset future time period is taken as the future average heat dissipation power value. That is, the sum of the actual heat dissipation power values ​​at all sampling times is divided by the number of sampling times to obtain the arithmetic mean of the actual heat dissipation power values.

[0092] Finally, based on the future average heat dissipation power and the rated heat dissipation power of the cooling system, where the rated heat dissipation power is the standard heat dissipation capacity of the cooling system under factory-safe, healthy conditions—that is, the maximum heat dissipation power that the cooling system can achieve under rated operating conditions when all components are in brand new or healthy condition—the sample cooling performance degradation degree is calculated. This can be obtained by subtracting the quotient of the future average heat dissipation power divided by the rated heat dissipation power of the cooling system at the time of manufacture from 1. The difference is taken as the sample cooling performance degradation degree. Specifically, the sample cooling performance degradation degree = 1 - (future average heat dissipation power / rated heat dissipation power). The sample cooling performance degradation degree is a dimensionless index characterizing the degree of degradation of the cooling system's heat dissipation capacity relative to the rated heat dissipation power over a historical period. Its value is between 0 and 1; a larger value indicates a more severe degradation in cooling performance.

[0093] Specifically, when the cooling system is in good condition at the factory, the future average heat dissipation power is close to the rated heat dissipation power, and the sample cooling efficiency decay rate is close to 0. As the cooling system is used for longer periods, factors such as the decline in coolant performance, the accumulation of dirt on the radiator surface, and the reduction in cooling pump efficiency lead to a decrease in actual heat dissipation capacity. The future average heat dissipation power gradually decreases, and the sample cooling efficiency decay rate gradually increases and approaches 1.

[0094] In this embodiment of the invention, through the accumulation of historical data samples and model training, the cooling performance degradation prediction model can learn the complex mapping relationship between the time series of cooling system operating parameters and the degree of cooling performance degradation. During actual charging, it accurately predicts the degree of cooling performance degradation based on the current time series of cooling system operating parameters. Using the coolant inlet / outlet temperature difference, coolant flow rate, and coolant specific heat capacity from historical cooling system operating parameters, the actual heat dissipation power value at each sampling time is calculated, and the average value over future periods is obtained as the future average heat dissipation power value. Finally, the sample cooling performance degradation degree is obtained by comparing it with the rated heat dissipation power value, reflecting the relative degree of degradation of the cooling system's heat dissipation capacity.

[0095] S400: Calculate and obtain the comprehensive thermal risk index of the battery based on the charging current rate factor, the battery internal resistance aging offset coefficient and the cooling efficiency decay.

[0096] In this embodiment of the invention, the battery comprehensive thermal risk index is a dimensionless comprehensive index used to quantitatively characterize the degree of thermal safety risk faced by the battery during charging. The higher the index value, the greater the current thermal safety risk of the battery.

[0097] Specifically, the battery internal resistance aging offset coefficient and cooling efficiency degradation are added together to obtain a sum that serves as the battery risk factor. Then, preset first and second risk weight coefficients are obtained, and the battery risk factor is multiplied by the first risk weight coefficient to obtain the first risk component; the charging current rate factor is multiplied by the second risk weight coefficient to obtain the second risk component. The first and second risk components are then added together to obtain the battery comprehensive thermal risk index.

[0098] Step S400 in the method of this embodiment of the invention includes:

[0099] The battery internal resistance aging offset coefficient and the cooling efficiency decay rate are added together, and the sum is used as the battery risk factor.

[0100] Obtain a preset first risk weight coefficient and a preset second risk weight coefficient, wherein the sum of the first risk weight coefficient and the second risk weight coefficient is 1;

[0101] Multiply the battery risk factor by the first risk weight coefficient to obtain the first risk component, and multiply the charging current rate factor by the second risk weight coefficient. The resulting product is recorded as the second risk component.

[0102] The first risk component and the second risk component are added together, and the sum is used as the battery comprehensive thermal risk index.

[0103] In this embodiment of the invention, firstly, the battery internal resistance aging offset coefficient and the cooling efficiency degradation degree are added together, and the sum is used as the battery risk factor. That is, battery risk factor = battery internal resistance aging offset coefficient + cooling efficiency degradation degree. The battery risk factor comprehensively reflects the degree of battery thermal risk under the combined effect of increased heat generation and decreased heat dissipation.

[0104] Secondly, obtain the preset first risk weight coefficient and the preset second risk weight coefficient. The weight coefficient is a numerical value used to measure the relative importance of different factors in the comprehensive evaluation, and its value ranges between 0 and 1. The first risk weight coefficient is the weight assigned to the battery risk factor, reflecting the long-term state factors of battery aging and cooling degradation; the second risk weight coefficient is the weight assigned to the charging current rate factor, reflecting the real-time charging condition factor. The sum of the first risk weight coefficient and the second risk weight coefficient equals one.

[0105] Specifically, the values ​​of the weighting coefficients can be dynamically determined based on the actual state of the battery. For example, the first risk weighting coefficient can be set to 0.6, and the second risk weighting coefficient to 0.4. These weighting coefficients allow the battery's comprehensive thermal risk index to flexibly adjust the contribution of two different types of factors to the final comprehensive risk.

[0106] Next, the battery risk factor is multiplied by the first risk weight coefficient to obtain the first risk component, i.e., first risk component = battery risk factor × first risk weight coefficient. Then, the charging current rate factor is multiplied by the second risk weight coefficient, and the product is recorded as the second risk component, i.e., second risk component = charging current rate factor × second risk weight coefficient. By calculating these weights separately, the two types of factors with unified dimensions are scaled according to their respective weight coefficients, allowing factors of different natures to be fairly compared and integrated within a unified comprehensive index framework.

[0107] Finally, the first risk component and the second risk component are added together, and the sum is used as the battery's comprehensive thermal risk index. The first risk weight coefficient can be obtained based on the relative magnitude of the cooling performance degradation and the battery's internal resistance aging deviation coefficient. The quotient of the cooling performance degradation divided by the sum of the battery's internal resistance aging deviation coefficient and the cooling performance degradation is used as the first risk weight coefficient; when the sum of the battery's internal resistance aging deviation coefficient and the cooling performance degradation is zero, the first risk weight coefficient is set to zero.

[0108] The battery comprehensive thermal risk index = first risk component + second risk component. The battery comprehensive thermal risk index is a dimensionless parameter normalized to the interval [0,1]. The closer its value is to 1, the higher the overheating risk faced by the battery under the combined effects of the current charging current rate, internal resistance aging deviation, and cooling efficiency degradation; the closer its value is to 0, the lower the overheating risk of the battery under the combined effects of the above multiple factors.

[0109] In step S400 of the method of this embodiment of the invention, obtaining the preset first risk weight coefficient and the preset second risk weight coefficient includes:

[0110] The cumulative usage time of the battery from the first use to the current time is obtained as the battery service life;

[0111] Obtain the preset design total service time of the battery, and divide the battery service time by the preset design total service time to obtain the battery aging index;

[0112] The battery aging index is used as the first risk weight coefficient, and 1 is subtracted from the first risk weight coefficient to obtain the second risk weight coefficient.

[0113] In this embodiment of the invention, firstly, the cumulative usage time of the battery from the first use time to the current time is obtained as the battery service time, and the preset design total service time of the battery is obtained.

[0114] Specifically, battery service life is the total time elapsed from the first use of the battery to the present moment, measured in hours or days. The pre-designed total service life is the expected lifespan of the battery under normal usage conditions as specified by the battery manufacturer, and its dimension is consistent with that of battery service life. The battery aging index is the proportion of battery service life to the pre-designed total service life, with a value between 0 and 1, used to characterize the aging stage of the battery in its life cycle.

[0115] When the battery is brand new, its service life is 0, the first risk weight coefficient is 0, and the second risk weight coefficient is 1. At this time, the battery's comprehensive thermal risk index is completely determined by the charging current rate factor.

[0116] Secondly, the preset total service life of the battery is obtained, and the battery service life is divided by the preset total service life to obtain the battery aging index. The battery aging index is a dimensionless parameter normalized to the interval [0,1]. The larger the value, the higher the proportion of the battery's cumulative service life from the first use to the current time to its preset total service life, that is, the greater the degree to which the battery is approaching the end of its life.

[0117] Specifically, in the calculation of the battery comprehensive thermal risk index, the battery aging degree index serves as the first risk weight coefficient, used to characterize the weight proportion of battery risk factors in the comprehensive assessment. When the battery aging degree index value is large, the first risk weight coefficient is correspondingly large, and the battery risk factors constituted by the battery internal resistance aging offset coefficient and cooling efficiency decay degree occupy a higher weight in the battery comprehensive thermal risk index; when the battery aging degree index value is small, the first risk weight coefficient is correspondingly small, and the charging current rate factor occupies a higher weight in the battery comprehensive thermal risk index.

[0118] Finally, the battery aging level index is used as the first risk weight coefficient, and the second risk weight coefficient is obtained by subtracting the first risk weight coefficient from 1. That is, the second risk weight coefficient = 1 - the first risk weight coefficient.

[0119] As service life increases, the weighting coefficient for the first risk gradually increases, while the weighting coefficient for the second risk gradually decreases. Long-term factors such as battery aging and cooling system degradation contribute more significantly to the overall thermal risk index. Specifically, when battery aging is low, changes in battery internal resistance and cooling system degradation have a relatively small impact on thermal risk, and the real-time charging current rate factor is the primary determinant of thermal risk, thus requiring a higher weighting. Conversely, when battery aging is high, the impact of increased battery internal resistance and cooling system degradation on thermal risk significantly increases, requiring even higher weighting.

[0120] For example, assuming a power battery's designed total service life is 8000 hours, if the battery's service life is 2000 hours, then the battery aging index = 2000 / 8000 = 0.25, the first risk weight coefficient is 0.25, and the second risk weight coefficient is 0.75, indicating that the battery is still in the early stages of its lifespan, and the contribution of real-time charging conditions to thermal risk is dominant. Assuming the battery risk factor = 0.5 and the charging current rate factor = 0.75, then the first risk component = 0.5 × 0.25 = 0.125, the second risk component = 0.75 × 0.75 = 0.5625, and the battery comprehensive thermal risk index = 0.125 + 0.5625 = 0.6875.

[0121] In this embodiment of the invention, the battery internal resistance aging offset coefficient and the cooling efficiency decay degree are first added together to form the battery risk factor, and the charging current rate factor is calculated. Then, the first risk weight coefficient and the second risk weight coefficient are dynamically calculated through the battery aging degree index, so that the contribution ratio of each factor to the comprehensive thermal risk index at different battery life stages can be adaptively adjusted, thereby improving the assessment accuracy at different aging stages. After multiplying the first risk weight coefficient and the second risk weight coefficient with the battery risk factor and the charging current rate factor respectively, the sum is obtained to obtain the comprehensive thermal risk index of the battery, which reflects the comprehensive thermal risk level of the battery in the current state.

[0122] S500: Based on the battery comprehensive thermal risk index, dynamically adjust the warning temperature threshold of the current charging process to obtain the dynamic warning temperature threshold;

[0123] In this embodiment of the invention, the warning temperature threshold is a critical temperature value used to determine whether the battery is in a thermal safety hazard state. When the current battery temperature reaches or exceeds this threshold, the system determines that the battery has a thermal safety risk and triggers a warning.

[0124] Specifically, a preset basic warning temperature threshold is obtained, and this threshold is corrected based on the battery's comprehensive thermal risk index to obtain a preset minimum safe temperature limit. When the dynamic warning temperature threshold calculated according to the correction formula is lower than the minimum safe temperature limit, the dynamic warning temperature threshold is taken as the minimum safe temperature limit.

[0125] Step S500 in the method of this embodiment of the invention includes:

[0126] Obtain a preset basic warning temperature threshold, which is the highest safe temperature value allowed by the battery under factory health conditions and rated heat dissipation conditions.

[0127] The basic warning temperature threshold is corrected based on the battery comprehensive thermal risk index to obtain a dynamic warning temperature threshold, wherein the dynamic warning temperature threshold and the battery comprehensive thermal risk index are negatively correlated.

[0128] Obtain a preset minimum safe temperature limit. When the dynamic warning temperature threshold is lower than the minimum safe temperature limit, take the dynamic warning temperature threshold as the minimum safe temperature limit.

[0129] In this embodiment of the invention, firstly, a preset basic warning temperature threshold is obtained. The basic warning temperature threshold is the highest safe temperature value allowed by the battery under factory health conditions and rated heat dissipation conditions. Specifically, the basic warning temperature threshold is the highest safe temperature value allowed by the battery manufacturer through extensive safety testing, determined by the battery manufacturer when the battery is in a brand-new factory condition and the cooling system is at its rated heat dissipation capacity. It represents the upper limit of the temperature at which the battery can safely operate during charging when the battery has not aged and the cooling system's heat dissipation capacity is at its rated level.

[0130] Specifically, a preset baseline warning temperature threshold can be read from the battery management system's storage unit or the battery's technical specifications file. For example, the baseline warning temperature threshold for a certain model of ternary lithium-ion power battery is 55°C, while the baseline warning temperature threshold for lithium iron phosphate power battery may be 60°C.

[0131] Secondly, the basic warning temperature threshold is corrected based on the battery comprehensive thermal risk index to obtain the dynamic warning temperature threshold. This can be achieved by multiplying the basic warning temperature threshold by one and subtracting the difference in the battery comprehensive thermal risk index; the resulting product is the dynamic warning temperature threshold, i.e., dynamic warning temperature threshold = basic warning temperature threshold × (1 - battery comprehensive thermal risk index). The dynamic warning temperature threshold and the battery comprehensive thermal risk index are negatively correlated.

[0132] Specifically, the dynamic warning temperature threshold is negatively correlated with the battery's comprehensive thermal risk index. The dynamic warning temperature threshold decreases as the battery's comprehensive thermal risk index increases. In other words, a higher comprehensive thermal risk index indicates a greater thermal risk to the battery, necessitating a lower warning temperature threshold to trigger the warning at a lower temperature, allowing for earlier protective measures. Therefore, the larger the reduction in the dynamic warning temperature threshold compared to the basic warning temperature threshold, the lower the warning trigger temperature threshold when the battery faces a high thermal risk, enabling the warning signal to be issued earlier. Conversely, when the battery's comprehensive thermal risk index is low, it indicates a lower current thermal risk, allowing for a higher warning temperature threshold to avoid excessively frequent false alarms.

[0133] Finally, a preset minimum safe temperature limit is obtained. When the dynamic warning temperature threshold is lower than the minimum safe temperature limit, the dynamic warning temperature threshold is taken as the minimum safe temperature limit; when the dynamic warning temperature threshold is higher than the minimum safe temperature limit, the calculated dynamic warning temperature threshold is used. The minimum safe temperature limit is the lower limit of the battery temperature warning threshold, and its dimensions are consistent with the basic warning temperature threshold.

[0134] Specifically, the warning temperature threshold should not be lowered indefinitely. Even when the battery's overall thermal risk index is extremely high, the warning temperature threshold should not fall below a certain reasonable minimum value. If the warning temperature threshold is too low, the battery will frequently trigger warnings during normal charging, causing unnecessary charging interruptions or user inconvenience and affecting the charging experience. Therefore, a preset minimum safe temperature limit needs to be set.

[0135] The setting of the minimum safe temperature limit needs to balance safety and usability. It is typically determined by subtracting a preset maximum allowable reduction from the base warning temperature threshold. For example, if the maximum allowable reduction is 15°C, then the minimum safe temperature limit = base warning temperature threshold - 15°C. The minimum safe temperature limit is also determined by the battery manufacturer based on the battery's safety characteristics and actual usage scenarios, and is preset in the battery management system.

[0136] Specifically, assuming the minimum safe temperature limit is 40℃, the warning temperature threshold can be lowered to a maximum of 40℃ and must not be lower than this value. Assuming the basic warning temperature threshold is 55℃ and the battery's comprehensive thermal risk index is 0.6, the dynamic warning temperature threshold is 55 × (1 - 0.4) = 33℃. Since the dynamic warning temperature threshold is less than the minimum safe temperature limit, the dynamic warning temperature threshold is set at 40℃.

[0137] In this embodiment of the invention, the basic warning temperature threshold is negatively correlated with the battery comprehensive thermal risk index, so that the warning temperature threshold is negatively correlated with the battery thermal risk, thereby achieving dynamic adjustment of the warning sensitivity. At the same time, by setting a minimum safe temperature limit, the dynamic warning temperature threshold is constrained to avoid frequent false alarms caused by excessively low thresholds, thereby improving the timeliness of warnings while ensuring system stability.

[0138] S600: When the current battery temperature value is greater than or equal to the dynamic warning temperature threshold, a battery temperature warning is triggered.

[0139] In this embodiment of the invention, the current battery temperature is the instantaneous value of the battery temperature collected in real time by a temperature sensor installed on the surface of the battery cell or inside the battery module. For the battery pack, temperature values ​​from multiple measuring points can be collected, and the highest value among them is taken as the current battery temperature to ensure safety under the most unfavorable operating conditions.

[0140] Specifically, a temperature sensor installed inside the battery pack collects the current battery temperature in real time. This current temperature is compared to a dynamic warning temperature threshold. When the current battery temperature is greater than or equal to the dynamic warning temperature threshold, it is determined that the current battery temperature has reached or exceeded the dynamically set safety threshold, indicating a risk of overheating and triggering a battery temperature warning. Upon triggering the warning, relevant warning information is simultaneously output.

[0141] Step S600 in the method of this embodiment of the invention includes:

[0142] The current temperature values ​​of the battery are obtained from multiple consecutive sampling times within the most recent preset time window before the current time, forming a short-time sequence of battery temperature.

[0143] Calculate the slope of the linear regression of the short-time series of battery temperature;

[0144] Obtain the current battery temperature value at the current moment, and compare the current battery temperature value with the dynamic warning temperature threshold.

[0145] When the current battery temperature is greater than or equal to the dynamic warning temperature threshold and the linear regression slope is positive, a battery temperature warning is triggered. The battery temperature warning includes the current battery temperature, the dynamic warning temperature threshold, the battery comprehensive thermal risk index, and the linear regression slope.

[0146] When the current temperature value of the battery is greater than or equal to the dynamic warning temperature threshold, but the linear regression slope is negative or zero, it is determined that the current temperature value of the battery is falling or remaining stable, and the battery temperature warning is not triggered. The charging current value of the current charging process is corrected and adjusted using the battery comprehensive thermal risk index.

[0147] If the current temperature of the battery is lower than the dynamic warning temperature threshold, the battery temperature warning will not be triggered.

[0148] In this embodiment of the invention, firstly, the current battery temperature values ​​are acquired from multiple consecutive sampling times within the most recent preset time window before the current moment, forming a short-time series of battery temperature. The preset time window is used to construct the input data for the cooling performance degradation prediction model, and its duration is primarily set based on the thermal dynamic response characteristics of the cooling system and the measured values ​​of the coolant circulation cycle. The thermal dynamic response characteristics of the cooling system refer to the time required for the coolant to flow from the inlet to the outlet and complete one cycle in the cooling pipe. This time is directly related to the length of the cooling pipe, the coolant flow rate, and the response delay of the cooling pump. If the window duration is less than a complete coolant circulation cycle, the time-series data of the inlet and outlet temperature difference and flow rate within the window will not fully reflect the actual heat dissipation state of the cooling system at the current moment. Therefore, the preset time window can be set to no less than 1.5 times the coolant circulation cycle to ensure that the collected time-series data can cover the complete change process of the cooling system's thermal state. For example, if the coolant circulation cycle in the cooling pipe is 80 seconds, the preset time window can be set to 120 seconds.

[0149] Specifically, the battery temperature is measured in real time within a preset time window by a temperature sensor installed inside or on the surface of the battery pack. Then, the battery temperature values ​​at all sampling times within the preset time window are arranged in chronological order to form a short-time sequence of battery temperature.

[0150] Secondly, the slope of the linear regression for the short-time series of battery temperature is calculated. Linear regression is a statistical analysis method that fits the linear relationship between the independent and dependent variables using the least squares method.

[0151] Specifically, using the sampling time sequence as the independent variable and the battery temperature value as the dependent variable, a linear regression is performed on the short-time series of battery temperature to obtain the best-fit line. The slope of this line is the linear regression slope, and its sign and magnitude reflect the trend and rate of change of battery temperature within a preset time window. A positive slope indicates that the current battery temperature value is increasing within the preset time window; a negative slope indicates that the current battery temperature value is decreasing within the preset time window; and a slope of zero or close to zero indicates that the current battery temperature value remains stable within the preset time window.

[0152] Preferably, the short-time series of battery temperature includes data points at n sampling times (t). i ,T i ), where i represents the time value at the i-th sampling time, T i Let represent the battery temperature value at the i-th sampling time. The formula for calculating the linear regression slope k is: , where t i T represents the time value at the i-th sampling moment. i Let represent the battery temperature value at the i-th sampling time, and n represent the number of sampling times.

[0153] For example, assume a preset time window of 120 seconds and a sampling interval of 1 second, which includes 121 sampling moments. Assume the battery temperature values ​​at these 121 sampling moments show a continuous upward trend. Substituting these 121 data points into the linear regression slope calculation formula above, we assume a slope k ≈ 0.025℃ / s.

[0154] Next, obtain the current battery temperature value at the current moment and compare it with the dynamic warning temperature threshold.

[0155] When the current battery temperature is greater than or equal to the dynamic warning temperature threshold, and the linear regression slope is positive, it indicates that the current battery temperature is equal to or exceeds the dynamic warning temperature threshold, is at a high level, and the temperature is still rising. This triggers a battery temperature warning, which includes the current battery temperature, the dynamic warning temperature threshold, the battery comprehensive thermal risk index, and the linear regression slope.

[0156] When the current battery temperature is greater than or equal to the dynamic warning temperature threshold, but the linear regression slope is negative or zero, it indicates that the current battery temperature has reached or exceeded the dynamic warning temperature threshold, but the temperature is decreasing or stabilizing, and the battery thermal state is improving towards safety or at least not deteriorating further. Therefore, it is determined that the current battery temperature is decreasing or stabilizing, and the battery temperature warning is not triggered. The charging current value during the current charging process is then adjusted using the battery comprehensive thermal risk index.

[0157] Specifically, the charging current correction adjustment method is as follows: Based on the battery's comprehensive thermal risk index, the corrected charging current value for the current charging process is determined. The corrected charging current value is calculated as: Current charging current value × (1 - Battery comprehensive thermal risk index). This proactively reduces the charging power when the battery's current temperature approaches the warning boundary. Simultaneously, a lower limit for current correction adjustment is set, such as 20% of the rated charging current, to avoid over-adjustment. When the battery's comprehensive thermal risk index is high, the corrected charging current value is smaller; when the battery's comprehensive thermal risk index is low, the corrected charging current value is larger. The corrected charging current value should not be lower than the minimum charging current value allowed by the battery management system to ensure normal charging. Furthermore, if the battery's comprehensive thermal risk index remains above the safety threshold, even if the temperature has stabilized, the charging current can still be proactively reduced to prevent the temperature from rising again.

[0158] When the current battery temperature is below the dynamic warning temperature threshold, it indicates that the battery's thermal state is within a safe range, and the battery temperature warning will not be triggered. At this time, the battery continues to charge according to the normal charging strategy, while continuously monitoring temperature changes and various parameters for further assessment.

[0159] In this embodiment of the invention, the linear regression slope of the short-time series of battery temperature is used as one of the criteria for triggering an early warning. Whether the current battery temperature value exceeds the dynamic early warning temperature threshold and the sign of the linear regression slope are used as two judgment conditions to obtain different judgment results and determine whether to trigger a battery temperature early warning, effectively avoiding false alarms caused by short-term temperature fluctuations. At the same time, the battery temperature early warning includes multi-dimensional data such as the current battery temperature value, the dynamic early warning temperature threshold, the battery comprehensive thermal risk index, and the linear regression slope, providing maintenance personnel with comprehensive judgment information.

[0160] The embodiments of this application, through the specific implementation methods described above, achieve the following technical effects:

[0161] In this embodiment, firstly, the existing sensors and data acquisition capabilities of the battery management system are utilized to comprehensively acquire charging current values, DC internal resistance values, and cooling system operating parameters. This provides complete data input from both electrical and thermal management characteristics perspectives for subsequent calculation of the battery's comprehensive thermal risk index, establishing a data foundation for dynamic adjustment of the early warning temperature threshold.

[0162] Secondly, by comparing the charging current value with the rated capacity value, the charging current rate factor is obtained to achieve a normalized characterization of the charging rate between batteries of different capacities; the ratio of the internal resistance deviation to the factory standard DC internal resistance value is calculated to obtain the battery internal resistance aging offset coefficient, quantifying the degree of aging growth of the battery internal resistance relative to the factory condition, and providing a data basis for subsequent comprehensive thermal risk assessment.

[0163] Furthermore, through the accumulation of historical data samples and model training, the cooling performance degradation prediction model can learn the complex mapping relationship between the time series of cooling system operating parameters and the degree of cooling performance degradation. During actual charging, it accurately predicts the degree of cooling performance degradation based on the current time series of cooling system operating parameters. Using the coolant inlet / outlet temperature difference, coolant flow rate, and coolant specific heat capacity from historical cooling system operating parameters, the actual heat dissipation power value at each sampling moment is calculated, and the average value over future periods is obtained as the future average heat dissipation power value. Finally, the sample cooling performance degradation degree is obtained by comparing it with the rated heat dissipation power value, reflecting the relative degree of degradation of the cooling system's heat dissipation capacity.

[0164] Furthermore, the battery internal resistance aging offset coefficient and cooling efficiency decay are first added together to form the battery risk factor, and the charging current rate factor is calculated. Subsequently, the first risk weight coefficient and the second risk weight coefficient are dynamically calculated through the battery aging degree index, so that the contribution ratio of each factor to the comprehensive thermal risk index at different battery life stages can be adaptively adjusted, thereby improving the assessment accuracy at different aging stages. After multiplying the first risk weight coefficient and the second risk weight coefficient with the battery risk factor and the charging current rate factor respectively, the summation yields the comprehensive thermal risk index of the battery, which reflects the comprehensive thermal risk level of the battery in the current state.

[0165] Furthermore, the basic warning temperature threshold is negatively correlated with the battery comprehensive thermal risk index, so that the warning temperature threshold is negatively correlated with the battery thermal risk, thereby achieving dynamic adjustment of the warning sensitivity. At the same time, by setting a minimum safe temperature limit, the dynamic warning temperature threshold is constrained to avoid frequent false alarms caused by excessively low thresholds, thereby improving the timeliness of warnings while ensuring system stability.

[0166] Ultimately, the linear regression slope of the short-time series of battery temperature is used as one of the criteria for triggering an early warning. Whether the current battery temperature exceeds the dynamic early warning temperature threshold and the sign of the linear regression slope are used as two judgment conditions to obtain different judgment results and determine whether to trigger a battery temperature early warning, effectively avoiding false alarms caused by short-term temperature fluctuations. At the same time, the battery temperature early warning includes multi-dimensional data such as the current battery temperature value, the dynamic early warning temperature threshold, the battery comprehensive thermal risk index, and the linear regression slope, providing maintenance personnel with comprehensive judgment information.

[0167] Example 2, as Figure 3 As shown, based on the same inventive concept as the battery temperature early warning method for the charging process of new energy vehicles provided in Embodiment 1, this embodiment of the invention also provides a battery temperature early warning device for the charging process of new energy vehicles, the device comprising:

[0168] The charging parameter acquisition module 11 is used to acquire the charging current value, DC internal resistance value and cooling system operating parameters of the battery at the current moment during the charging process;

[0169] The battery internal resistance aging offset coefficient calculation module 12 is used to calculate the charging current rate factor based on the ratio of the charging current value to the battery rated capacity value, and to calculate the battery internal resistance aging offset coefficient based on the deviation between the DC internal resistance value and the factory standard DC internal resistance value.

[0170] The intelligent analysis module 13 for cooling system operating parameters is used to intelligently analyze the operating parameters of the cooling system and obtain the cooling efficiency decay rate.

[0171] The battery comprehensive thermal risk index acquisition module 14 is used to calculate and acquire the battery comprehensive thermal risk index based on the charging current rate factor, the battery internal resistance aging offset coefficient and the cooling efficiency decay.

[0172] The warning temperature threshold dynamic adjustment module 15 is used to dynamically adjust the warning temperature threshold of the current charging process according to the battery comprehensive thermal risk index, and obtain the dynamic warning temperature threshold.

[0173] The battery temperature warning module 16 is used to trigger a battery temperature warning when the current battery temperature value is greater than or equal to the dynamic warning temperature threshold.

[0174] In one embodiment, the charging parameter acquisition module 11 is used for:

[0175] The charging current value at the current moment is obtained as the charging current value, and the estimated DC internal resistance value at the current moment is obtained as the DC internal resistance value. The estimated DC internal resistance value is obtained by using a recursive least squares method including a forgetting factor to estimate the terminal voltage value and charging current value of the battery at the current moment and within a preset time window before the current moment, based on the terminal voltage value and charging current value of the battery at multiple consecutive sampling moments within the current moment and the preset time window before the current moment.

[0176] The current coolant inlet temperature, coolant outlet temperature, and coolant flow rate are obtained and used as the operating parameters of the cooling system.

[0177] In one embodiment, the battery internal resistance aging offset coefficient calculation module 12 is used for:

[0178] Obtain the rated capacity value of the battery, divide the charging current value by the rated capacity value, and obtain the charging current rate factor;

[0179] Obtain the factory standard DC internal resistance value of the battery at the time of manufacture, and subtract the factory standard DC internal resistance value from the DC internal resistance value to obtain the internal resistance deviation.

[0180] Divide the internal resistance deviation by the factory standard DC internal resistance value to obtain the battery internal resistance aging offset coefficient.

[0181] In one embodiment, the intelligent analysis module 13 for cooling system operating parameters is used for:

[0182] The time series consisting of the coolant inlet temperature, coolant outlet temperature, and coolant flow rate values ​​of multiple consecutive sampling times within the current time and the preset time window before the current time is input into the pre-constructed cooling performance decay prediction model to obtain the cooling performance decay degree output by the cooling performance decay prediction model.

[0183] The pre-construction of the cooling performance degradation prediction model includes:

[0184] Collect multiple sets of historical cooling system operation data. Each set of data includes the start time of a historical period and the historical coolant inlet temperature, historical coolant outlet temperature, and historical coolant flow rate values ​​of multiple consecutive sampling times within a preset time window before the start time of the historical period, as well as the corresponding sample cooling efficiency decay.

[0185] A cooling performance degradation prediction model is constructed, which includes a temporal convolutional layer and a trend prediction layer. The temporal convolutional layer is used to perform multi-layer convolutional encoding on the temporal sequences of coolant inlet temperature, coolant outlet temperature, and coolant flow rate to extract the temporal variation features of temperature difference and flow rate. The trend prediction layer is used to receive the output of the temporal convolutional layer and output the cooling performance degradation degree after fully connected mapping.

[0186] The cooling performance attenuation prediction model is trained using the historical cooling system operation data, with the goal of minimizing the deviation between the output cooling performance attenuation degree and the sample cooling performance attenuation degree, until the deviation meets the preset convergence condition, thus obtaining the pre-constructed cooling performance attenuation prediction model.

[0187] The acquisition of the sample cooling efficiency attenuation includes:

[0188] The difference between the historical coolant outlet temperature value and the historical coolant inlet temperature value at each sampling moment in the cooling system operating parameters within a preset future time period after obtaining the historical time period start time is used as the inlet and outlet temperature difference.

[0189] Calculate the product of the inlet and outlet temperature difference at each sampling time, the historical coolant flow rate at the corresponding sampling time, and the specific heat capacity of the coolant used, to obtain the actual heat dissipation power value at each sampling time.

[0190] The arithmetic mean of the actual heat dissipation power values ​​at all sampling times within the preset future time period is taken as the future average heat dissipation power value;

[0191] Based on the future average heat dissipation power value and the rated heat dissipation power value of the cooling system, the sample cooling efficiency decay rate is calculated.

[0192] In one embodiment, the battery comprehensive thermal risk index acquisition module 14 is used for:

[0193] The battery internal resistance aging offset coefficient and the cooling efficiency decay rate are added together, and the sum is used as the battery risk factor.

[0194] Obtain a preset first risk weight coefficient and a preset second risk weight coefficient, wherein the sum of the first risk weight coefficient and the second risk weight coefficient is 1;

[0195] Multiply the battery risk factor by the first risk weight coefficient to obtain the first risk component, and multiply the charging current rate factor by the second risk weight coefficient. The resulting product is recorded as the second risk component.

[0196] The first risk component and the second risk component are added together, and the sum is used as the battery comprehensive thermal risk index.

[0197] The step of obtaining the preset first risk weight coefficient and the preset second risk weight coefficient includes:

[0198] The cumulative usage time of the battery from the first use to the current time is obtained as the battery service life;

[0199] Obtain the preset design total service time of the battery, and divide the battery service time by the preset design total service time to obtain the battery aging index;

[0200] The battery aging index is used as the first risk weight coefficient, and 1 is subtracted from the first risk weight coefficient to obtain the second risk weight coefficient.

[0201] In one embodiment, the warning temperature threshold dynamic adjustment module 15 is used for:

[0202] Obtain a preset basic warning temperature threshold, which is the highest safe temperature value allowed by the battery under factory health conditions and rated heat dissipation conditions.

[0203] The basic warning temperature threshold is corrected based on the battery comprehensive thermal risk index to obtain a dynamic warning temperature threshold, wherein the dynamic warning temperature threshold and the battery comprehensive thermal risk index are negatively correlated.

[0204] Obtain a preset minimum safe temperature limit. When the dynamic warning temperature threshold is lower than the minimum safe temperature limit, take the dynamic warning temperature threshold as the minimum safe temperature limit.

[0205] In one embodiment, the battery temperature warning module 16 is used for:

[0206] The current temperature values ​​of the battery are obtained from multiple consecutive sampling times within the most recent preset time window before the current time, forming a short-time sequence of battery temperature.

[0207] Calculate the slope of the linear regression of the short-time series of battery temperature;

[0208] Obtain the current battery temperature value at the current moment, and compare the current battery temperature value with the dynamic warning temperature threshold.

[0209] When the current battery temperature is greater than or equal to the dynamic warning temperature threshold and the linear regression slope is positive, a battery temperature warning is triggered. The battery temperature warning includes the current battery temperature, the dynamic warning temperature threshold, the battery comprehensive thermal risk index, and the linear regression slope.

[0210] When the current temperature value of the battery is greater than or equal to the dynamic warning temperature threshold, but the linear regression slope is negative or zero, it is determined that the current temperature value of the battery is falling or remaining stable, and the battery temperature warning is not triggered. The charging current value of the current charging process is corrected and adjusted using the battery comprehensive thermal risk index.

[0211] If the current temperature of the battery is lower than the dynamic warning temperature threshold, the battery temperature warning will not be triggered.

[0212] Compared to existing technologies, this invention first provides a multi-dimensional data foundation for battery thermal risk assessment, encompassing heat generation intensity, battery aging status, and heat dissipation capacity, by acquiring charging current values, DC internal resistance values, and cooling system operating parameters. Secondly, it calculates a charging current rate factor based on the ratio of the charging current value to the rated capacity value, and calculates a battery internal resistance aging offset coefficient based on the deviation of the DC internal resistance value from the factory standard DC internal resistance value, achieving a quantitative characterization of charging rate and battery aging degree. Thirdly, it obtains the cooling efficiency degradation degree through intelligent analysis of cooling system operating parameters, incorporating the impact of cooling system performance degradation on thermal safety into the assessment system. Furthermore, it calculates a comprehensive battery thermal risk index based on the charging current rate factor, battery internal resistance aging offset coefficient, and cooling efficiency degradation degree, achieving a multi-factor comprehensive quantification of battery thermal safety status. Further, it dynamically adjusts the warning temperature threshold according to the comprehensive battery thermal risk index, enabling the warning sensitivity to adaptively adjust with changes in thermal risk. Finally, it triggers a warning when the current battery temperature reaches or exceeds the dynamic warning temperature threshold, achieving an accurate and timely response to battery thermal risks.

Claims

1. A battery temperature early warning method for use in the charging process of new energy vehicles, characterized in that, include: Obtain the charging current value, DC internal resistance value, and cooling system operating parameters of the battery at the current moment during the charging process; Based on the ratio of the charging current value to the battery rated capacity value, the charging current rate factor is calculated, and based on the deviation of the DC internal resistance value from the factory standard DC internal resistance value, the battery internal resistance aging offset coefficient is calculated. The operating parameters of the cooling system are intelligently analyzed to obtain the cooling efficiency decay rate; Based on the charging current rate factor, the battery internal resistance aging offset coefficient, and the cooling efficiency decay, the battery comprehensive thermal risk index is calculated and obtained. Based on the battery comprehensive thermal risk index, the warning temperature threshold of the current charging process is dynamically adjusted to obtain the dynamic warning temperature threshold. When the current battery temperature is greater than or equal to the dynamic warning temperature threshold, a battery temperature warning is triggered.

2. The battery temperature early warning method for the charging process of new energy vehicles according to claim 1, characterized in that, The acquisition of the charging current value, DC internal resistance value, and cooling system operating parameters of the battery at the current moment during the charging process includes: The charging current value at the current moment is obtained as the charging current value, and the estimated DC internal resistance value at the current moment is obtained as the DC internal resistance value. The estimated DC internal resistance value is obtained by using a recursive least squares method including a forgetting factor to estimate the terminal voltage value and charging current value of the battery at the current moment and within a preset time window before the current moment, based on the terminal voltage value and charging current value of the battery at multiple consecutive sampling moments within the current moment and the preset time window before the current moment. The current coolant inlet temperature, coolant outlet temperature, and coolant flow rate are obtained and used as the operating parameters of the cooling system.

3. The battery temperature early warning method for the charging process of new energy vehicles according to claim 1, characterized in that, The calculation of the charging current rate factor based on the ratio of the charging current value to the battery's rated capacity value, and the calculation of the battery internal resistance aging offset coefficient based on the deviation of the DC internal resistance value from the factory standard DC internal resistance value, including: Obtain the rated capacity value of the battery, divide the charging current value by the rated capacity value, and obtain the charging current rate factor; Obtain the factory standard DC internal resistance value of the battery at the time of manufacture, and subtract the factory standard DC internal resistance value from the DC internal resistance value to obtain the internal resistance deviation. Divide the internal resistance deviation by the factory standard DC internal resistance value to obtain the battery internal resistance aging offset coefficient.

4. The battery temperature early warning method for the charging process of new energy vehicles according to claim 1, characterized in that, The intelligent analysis of the operating parameters of the cooling system to obtain the cooling efficiency attenuation includes: The time series consisting of the coolant inlet temperature, coolant outlet temperature, and coolant flow rate values ​​of multiple consecutive sampling times within the current time and the preset time window before the current time is input into the pre-constructed cooling performance decay prediction model to obtain the cooling performance decay degree output by the cooling performance decay prediction model. The pre-construction of the cooling performance degradation prediction model includes: Collect multiple sets of historical cooling system operation data. Each set of data includes the start time of a historical period and the historical coolant inlet temperature, historical coolant outlet temperature, and historical coolant flow rate values ​​of multiple consecutive sampling times within a preset time window before the start time of the historical period, as well as the corresponding sample cooling efficiency decay. A cooling performance degradation prediction model is constructed, which includes a temporal convolutional layer and a trend prediction layer. The temporal convolutional layer is used to perform multi-layer convolutional encoding on the temporal sequences of coolant inlet temperature, coolant outlet temperature, and coolant flow rate to extract the temporal variation features of temperature difference and flow rate. The trend prediction layer is used to receive the output of the temporal convolutional layer and output the cooling performance degradation degree after fully connected mapping. The cooling performance attenuation prediction model is trained using the historical cooling system operation data, with the goal of minimizing the deviation between the output cooling performance attenuation degree and the sample cooling performance attenuation degree, until the deviation meets the preset convergence condition, thus obtaining the pre-constructed cooling performance attenuation prediction model.

5. The battery temperature early warning method for the charging process of new energy vehicles according to claim 4, characterized in that, The acquisition of the sample cooling efficiency attenuation includes: The difference between the historical coolant outlet temperature value and the historical coolant inlet temperature value at each sampling moment in the cooling system operating parameters within a preset future time period after obtaining the historical time period start time is used as the inlet and outlet temperature difference. Calculate the product of the inlet and outlet temperature difference at each sampling time, the historical coolant flow rate at the corresponding sampling time, and the specific heat capacity of the coolant used, to obtain the actual heat dissipation power value at each sampling time. The arithmetic mean of the actual heat dissipation power values ​​at all sampling times within the preset future time period is taken as the future average heat dissipation power value; Based on the future average heat dissipation power value and the rated heat dissipation power value of the cooling system, the sample cooling efficiency decay rate is calculated.

6. The battery temperature early warning method for the charging process of new energy vehicles according to claim 1, characterized in that, The comprehensive thermal risk index of the battery is calculated based on the charging current rate factor, the battery internal resistance aging offset coefficient, and the cooling efficiency degradation, including: The battery internal resistance aging offset coefficient and the cooling efficiency decay rate are added together, and the sum is used as the battery risk factor. Obtain a preset first risk weight coefficient and a preset second risk weight coefficient, wherein the sum of the first risk weight coefficient and the second risk weight coefficient is 1; Multiply the battery risk factor by the first risk weight coefficient to obtain the first risk component, and multiply the charging current rate factor by the second risk weight coefficient. The resulting product is recorded as the second risk component. The first risk component and the second risk component are added together, and the sum is used as the battery comprehensive thermal risk index.

7. The battery temperature early warning method for the charging process of new energy vehicles according to claim 6, characterized in that, The process of obtaining the preset first risk weight coefficient and the preset second risk weight coefficient includes: The cumulative usage time of the battery from the first use to the current time is obtained as the battery service life; Obtain the preset design total service time of the battery, and divide the battery service time by the preset design total service time to obtain the battery aging index; The battery aging index is used as the first risk weight coefficient, and 1 is subtracted from the first risk weight coefficient to obtain the second risk weight coefficient.

8. The battery temperature early warning method for the charging process of new energy vehicles according to claim 1, characterized in that, The step of dynamically adjusting the warning temperature threshold for the current charging process based on the battery's comprehensive thermal risk index, and obtaining the dynamic warning temperature threshold, includes: Obtain a preset basic warning temperature threshold, which is the highest safe temperature value allowed by the battery under factory health conditions and rated heat dissipation conditions. The basic warning temperature threshold is corrected based on the battery comprehensive thermal risk index to obtain a dynamic warning temperature threshold, wherein the dynamic warning temperature threshold and the battery comprehensive thermal risk index are negatively correlated. Obtain a preset minimum safe temperature limit. When the dynamic warning temperature threshold is lower than the minimum safe temperature limit, take the dynamic warning temperature threshold as the minimum safe temperature limit.

9. The battery temperature early warning method for the charging process of new energy vehicles according to claim 1, characterized in that, The step of triggering a battery temperature warning when the current battery temperature is greater than or equal to the dynamic warning temperature threshold includes: The current temperature values ​​of the battery are obtained from multiple consecutive sampling times within the most recent preset time window before the current time, forming a short-time sequence of battery temperature. Calculate the slope of the linear regression of the short-time series of battery temperature; Obtain the current battery temperature value at the current moment, and compare the current battery temperature value with the dynamic warning temperature threshold. When the current battery temperature is greater than or equal to the dynamic warning temperature threshold and the linear regression slope is positive, a battery temperature warning is triggered. The battery temperature warning includes the current battery temperature, the dynamic warning temperature threshold, the battery comprehensive thermal risk index, and the linear regression slope. When the current temperature value of the battery is greater than or equal to the dynamic warning temperature threshold, but the linear regression slope is negative or zero, it is determined that the current temperature value of the battery is falling or remaining stable, and the battery temperature warning is not triggered. The charging current value of the current charging process is corrected and adjusted using the battery comprehensive thermal risk index. If the current temperature of the battery is lower than the dynamic warning temperature threshold, the battery temperature warning will not be triggered.

10. A battery temperature early warning device for use in the charging process of new energy vehicles, characterized in that, For implementing the battery temperature early warning method for the charging process of new energy vehicles according to any one of claims 1 to 9, the apparatus comprises: The charging parameter acquisition module is used to acquire the charging current value, DC internal resistance value and cooling system operating parameters of the battery at the current moment during the charging process; The battery internal resistance aging offset coefficient calculation module is used to calculate the charging current rate factor based on the ratio of the charging current value to the battery rated capacity value, and to calculate the battery internal resistance aging offset coefficient based on the deviation between the DC internal resistance value and the factory standard DC internal resistance value. The intelligent analysis module for cooling system operating parameters is used to intelligently analyze the operating parameters of the cooling system and obtain the cooling efficiency decay rate. The battery comprehensive thermal risk index acquisition module is used to calculate and acquire the battery comprehensive thermal risk index based on the charging current rate factor, the battery internal resistance aging offset coefficient, and the cooling efficiency decay. The warning temperature threshold dynamic adjustment module is used to dynamically adjust the warning temperature threshold of the current charging process according to the battery comprehensive thermal risk index, and obtain the dynamic warning temperature threshold. The battery temperature warning module is used to trigger a battery temperature warning when the current battery temperature value is greater than or equal to the dynamic warning temperature threshold.

Images