Training method of battery state detection model, and battery state detection method and device

Abstract

The invention discloses a training method of a battery state detection model and a battery state detection method and device, and belongs to the field of vehicles. The method comprises the following steps: acquiring a data signal of a vehicle, wherein the data signal comprises at least one of a voltage signal, a current signal or a temperature signal of a power battery of the vehicle; training a battery state detection model based on the data signal to obtain data fitting loss and physical constraint loss; optimizing model parameters of the battery state detection model by using data fitting loss and physical constraint loss; and determining a trained battery state detection model based on the optimized model parameters, wherein the trained battery state detection model is used for detecting the battery state of the power battery of the vehicle. According to the method, the data fitting loss and the physical constraint loss are utilized, so that the battery state detection model follows the physical law in the training process, the dependence on a large amount of data is reduced, the consumption of computing resources is reduced, and the detection accuracy is relatively high.

Description

Technical Field

[0001] Embodiments of the present application relate to the field of vehicles, and in particular to a battery status detection model training method, a battery status detection method and a battery status detection device. Background Art

[0002] During the driving process of the vehicle, the power battery provides energy for the vehicle, and its performance and life directly determine the vehicle's endurance and reliability. As the number of charge and discharge cycles of the power battery increases, the active materials inside the battery will gradually undergo irreversible chemical reactions. Therefore, it is necessary to monitor the battery health state (SOH) of the power battery.

[0003] SOH is an important indicator to measure the degree of battery performance degradation, reflecting the current power of the power battery relative to the maximum storage capacity of the power battery. Through the SOH value, you can intuitively understand the current health status and performance level of the battery. Summary of the invention

[0004] The embodiment of the present application provides a battery status detection model training method, a battery status detection method and a device, which improves the accuracy and efficiency of battery status detection. The technical solution is as follows:

[0005] In a first aspect, an embodiment of the present application provides a method for training a battery status detection model, the method comprising:

[0006] Acquiring a data signal of the vehicle, the data signal comprising at least one of a voltage signal, a current signal or a temperature signal of a power battery of the vehicle;

[0007] Training a battery state detection model based on the data signal to obtain a data fitting loss and a physical constraint loss, wherein the data fitting loss represents a difference between a predicted value of the battery state detection model and an actual value of the battery state, and the physical constraint loss represents a difference between a predicted value of the battery state detection model and a theoretical value of a physical law of a battery system, wherein the physical constraint loss includes at least one of an electrochemical loss, a thermal conduction loss, or a battery capacity attenuation loss;

[0008] Optimizing the model parameters of the battery state detection model by using the data fitting loss and the physical constraint loss to obtain optimized model parameters;

[0009] A trained battery status detection model is determined based on the optimized model parameters, and the trained battery status detection model is used to detect the battery status of the power battery of the vehicle.

[0010] In a second aspect, an embodiment of the present application provides a battery status detection method, the method comprising:

[0011] Acquiring a data signal of the vehicle, wherein the data signal includes at least one of a voltage signal, a current signal, or a temperature signal;

[0012] The battery status of the vehicle is detected based on a battery status detection model and the data signal to obtain a detection result of the battery status, wherein the battery status detection model is trained based on any one of the above methods.

[0013] In a third aspect, an embodiment of the present application provides a training device for a battery status detection model, the device comprising:

[0014] A first acquisition module, used to acquire a data signal of the vehicle, wherein the data signal includes at least one of a voltage signal, a current signal or a temperature signal of a power battery of the vehicle;

[0015] a training module, configured to train a battery state detection model based on the data signal to obtain a data fitting loss and a physical constraint loss, wherein the data fitting loss represents a difference between a predicted value of the battery state detection model and an actual value of the battery state, and the physical constraint loss represents a difference between a predicted value of the battery state detection model and a theoretical value of a physical law of a battery system, wherein the physical constraint loss includes at least one of an electrochemical loss, a thermal conduction loss, or a battery capacity attenuation loss;

[0016] The training module is further used to optimize the model parameters of the battery state detection model by using the data fitting loss and the physical constraint loss to obtain optimized model parameters;

[0017] The training module is also used to determine a trained battery status detection model based on the optimized model parameters, and the trained battery status detection model is used to detect the battery status of the vehicle's power battery.

[0018] In one possible implementation, the training module is used to determine a signal residual using the battery state detection model and the data signal, wherein the signal residual represents the difference between a detection value of the battery state detection model and a theoretical value, wherein the theoretical value is determined based on the data signal and a physical equation; calculate at least one of the electrochemical loss, the thermal conduction loss, or the battery capacity attenuation loss based on the number of the data signal and the signal residual; and determine the physical constraint loss based on at least one of the electrochemical loss, the thermal conduction loss, or the battery capacity attenuation loss and a corresponding reference weight.

[0019] In a possible implementation, the theoretical value is obtained based on a physical equation of a battery system and a data signal, and the theoretical value includes at least one of a theoretical voltage, a theoretical current, a theoretical temperature change, or a theoretical capacity decay rate; wherein the theoretical voltage and the theoretical current are calculated based on Ohm's law and the voltage signal or the current signal, and the theoretical temperature change is calculated based on the temperature signal and the Joule heating effect; and the theoretical value of the battery capacity decay is obtained based on the data signal and an aging equation.

[0020] In one possible implementation, the training module is used to initialize the battery status detection model to obtain an initialized battery status detection model; determine the data fitting loss and the physical constraint loss based on the data signal and the initialized battery status detection model; determine the total loss using the data fitting loss, the first weight corresponding to the data fitting loss, the physical constraint loss, and the second weight corresponding to the physical constraint loss; and optimize the battery status detection model using the total loss to obtain the optimized model parameters.

[0021] In one possible implementation, the training module is used to determine the gradient corresponding to the total loss based on the model parameters of the battery status detection model; calculate the first-order moment estimate and the second-order moment estimate based on the gradient, the first-order moment estimate is the weighted average of the exponent of the gradient, and the second-order moment estimate is the weighted average of the exponent of the square of the gradient; use the first-order moment estimate and the second-order moment estimate to update the model parameters of the battery status detection model to obtain the optimized model parameters.

[0022] In one possible implementation, the device also includes a processing module, which is used to preprocess the data signal to obtain a preprocessed data signal, wherein the preprocessing includes at least one of outlier removal, denoising, normalization, time alignment or signal enhancement; the training module is used to train the battery state detection model using the preprocessed data signal to obtain the data fitting loss and the physical constraint loss.

[0023] In one possible implementation, the preprocessing includes removing outliers, and the processing module is used to verify the data signal based on physical laws, determine a first abnormal data signal, and the abnormal data signal is a data signal in which a voltage signal and a current signal do not match; determine a reference range in the data signal, and determine the data signal exceeding the reference range as a second abnormal data signal; and delete the first abnormal data signal and the second abnormal data signal from the data signal.

[0024] In a fourth aspect, an embodiment of the present application provides a battery status detection device, the device comprising:

[0025] A second acquisition module, used to acquire a data signal of the vehicle, wherein the data signal includes at least one of a voltage signal, a current signal or a temperature signal;

[0026] A detection module is used to detect the battery status of the vehicle based on a battery status detection model and the data signal to obtain a detection result of the battery status, wherein the battery status detection model is trained based on any of the methods described above.

[0027] In a fifth aspect, an embodiment of the present application provides a computer device, comprising a processor and a memory, wherein the memory stores at least one program code, and the at least one program code is loaded and executed by the processor so that the computer device implements any of the above-mentioned battery status detection model training methods or battery status detection methods.

[0028] On the other hand, a computer-readable storage medium is also provided, in which at least one program code is stored, and the at least one program code is loaded and executed by a processor so that a computer implements any of the above-mentioned battery status detection model training methods or battery status detection methods.

[0029] On the other hand, a computer program or a computer program product is also provided, wherein at least one computer instruction is stored in the computer program or the computer program product, and the at least one computer instruction is loaded and executed by a processor so that the computer implements any of the above-mentioned battery status detection model training methods or battery status detection methods.

[0030] The technical solution provided by the embodiments of the present application brings at least the following beneficial effects:

[0031] In the process of training the battery status detection model, the present application uses data fitting loss and physical constraint loss to make the battery status detection model follow physical laws during the training process, reduce dependence on a large amount of data, and converge faster during training, reducing the consumption of computing resources. In addition, the battery status detection model is not trained using a large amount of data, and the number of iterations in the battery status detection model training process is small. The training error will not increase with the number of iterations, so that the trained battery status detection model can have a higher detection accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0033] Figure 1 It is a schematic diagram of an implementation environment of a battery status detection model training method or a battery status detection method provided in an embodiment of the present application;

[0034] Figure 2 It is a flow chart of a method for training a battery status detection model provided in an embodiment of the present application;

[0035] Figure 3 is a flow chart of a battery status detection method provided in an embodiment of the present application;

[0036] Figure 4 It is a structural schematic diagram of a training device for a battery status detection model provided in an embodiment of the present application;

[0037] Figure 5 It is a structural schematic diagram of a battery status detection device provided in an embodiment of the present application;

[0038] Figure 6 It is a structural diagram of a terminal device provided in an embodiment of the present application;

[0039] Figure 7 It is a structural diagram of a server provided in an embodiment of the present application. DETAILED DESCRIPTION

[0040] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0041] It should be noted that the terms "first", "second", etc. in this application are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the terms used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0042] Figure 1 is a schematic diagram of an implementation environment of a battery status detection model training method or a battery status detection method provided in an embodiment of the present application, such as Figure 1As shown, the implementation environment includes: a terminal device 101 and a server 102 .

[0043] The training method of the battery status detection model or the battery status detection method provided in the embodiment of the present application can be executed by the terminal device 101, can be executed by the server 102, or can be jointly executed by the terminal device 101 and the server 102, and the embodiment of the present application does not limit this. For the case where the training method of the battery status detection model or the battery status detection method provided in the embodiment of the present application is jointly executed by the terminal device 101 and the server 102, the server 102 undertakes the main computing work, and the terminal device 101 undertakes the secondary computing work; or, the server 102 bears the secondary computing work, and the terminal device 101 undertakes the main computing work; or, the server 102 and the terminal device 101 adopt a distributed computing architecture for collaborative computing.

[0044] It should be noted that the execution device of the training method of the battery status detection model and the execution device of the battery status detection method may be the same or different, and the embodiments of the present application do not limit this. Exemplarily, the execution device of the training method of the battery status detection model is the terminal device 101, and the execution device of the battery status detection method is the server 102; or, the execution device of the battery status detection method and the training method of the battery status detection model are both the terminal device 101.

[0045] Optionally, the terminal device 101 can be any electronic product that can perform human-computer interaction with a user through one or more methods such as a keyboard, a touchpad, a touch screen, a remote controller, voice interaction, or a handwriting device. The terminal device 101 includes but is not limited to a mobile phone, a computer, an intelligent voice interaction device, a vehicle-mounted terminal, an aircraft, etc. The server 102 is a server, or a server cluster composed of multiple servers, or any one of a cloud computing platform and a virtualization center, which is not limited in the embodiments of the present application. The server 102 is connected to the terminal device 101 through a wired network or a wireless network. The server 102 has a data receiving function, a data processing function, and a data sending function. Of course, the server 102 may also have other functions, which are not limited in the embodiments of the present application.

[0046] Those skilled in the art should understand that the above-mentioned terminal device 101 and server 102 are only for illustration, and other existing or future terminal devices or servers, if applicable to the present application, should also be included in the scope of protection of the present application and are included here by reference.

[0047] The present application embodiment provides a method for training a battery status detection model, which can be applied to the above Figure 1 For example, the method can be implemented by Figure 1The terminal device 101 and the server 102 in the embodiment interact with each other and can also be executed by the terminal device 101. Figure 2 As shown, the method includes the following steps 201 to 204.

[0048] In step 201 , a data signal of a vehicle is acquired, where the data signal includes at least one of a voltage signal, a current signal or a temperature signal of a power battery of the vehicle.

[0049] In an exemplary embodiment of the present application, the voltage signal of the vehicle power battery may include the real-time terminal voltage of the single cell, the overall voltage of the battery pack (PACK), and the dynamic voltage curve of the entire charging and discharging process. For example, in the process of acquiring the voltage signal, a high-precision voltage sensor and a distributed battery management system (BMS) may be used to collect the voltage signal. The current signal of the vehicle power battery may include the real-time current value of the battery during charging and discharging, and distinguish the charging and discharging direction. The temperature signal of the vehicle power battery may include but is not limited to the temperature of the key position on the surface of the single cell, the multi-point distribution temperature inside the battery pack, and the ambient temperature data.

[0050] The data signal of the vehicle may include, but is not limited to, real-time monitoring data of the vehicle's power battery, historical charge and discharge data of the power battery, or laboratory test data. The source of the data signal may be obtained based on actual conditions, and this application does not limit this.

[0051] After acquiring the data signal of the vehicle, the data signal may be preprocessed to obtain a preprocessed data signal, wherein the preprocessing includes at least one of outlier removal, denoising, normalization, time alignment or signal enhancement.

[0052] Exemplarily, when processing vehicle data signals, the data may contain abnormal data signals due to sensor failure, data transmission errors, or environmental factors. The process of removing abnormal values ​​from the data signal may include: verifying the data signal based on physical laws, determining a first abnormal data signal, where the abnormal data signal is a data signal where the voltage signal and the current signal do not match; determining a reference range in the data signal, and determining the data signal exceeding the reference range as a second abnormal data signal; and deleting the first abnormal data signal and the second abnormal data signal from the data signal.

[0053] Exemplarily, the inherent physical laws of the battery system are used to detect whether the voltage signal and the current signal comply with the physical laws. If the voltage signal and the current signal comply with the physical laws, it is considered that the voltage signal and the current signal match; if the voltage signal and the current signal do not comply with the physical laws, it is considered that the voltage signal and the current signal do not match.

[0054] For example, physical laws may include Ohm's law. The voltage signal and current signal in the data signal may be used to determine the theoretical voltage value corresponding to the current signal, and the theoretical voltage value may be compared with the actual voltage value of the collected voltage signal. If the error between the actual voltage value and the theoretical voltage value is less than or equal to the error threshold, it is considered that the voltage signal and the current signal comply with Ohm's law, that is, the voltage signal and the current signal are normal data signals; if the error between the actual voltage value and the theoretical voltage value is greater than the error threshold, it is considered that the voltage signal and the current signal do not comply with Ohm's law, that is, at least one of the voltage signal or the current signal is determined to be the first abnormal data signal.

[0055] For another example, physical laws may include the law of conservation of energy. The law of conservation of energy can be used to verify the rationality of energy during battery charging and discharging. According to the law of conservation of energy, the total energy input or output of a power battery during charging and discharging should be consistent with the sum of the change in energy stored inside the battery and the energy loss. Through the voltage signal and current signal in the data signal, the actual energy value during the charging and discharging process can be calculated and compared with the theoretical energy value. The theoretical energy value is determined by the change in stored energy corresponding to the rated capacity of the battery and the internal resistance loss energy. If the error between the actual energy value and the theoretical energy value is less than or equal to the error threshold, it is considered that the voltage signal and the current signal comply with the law of conservation of energy and are determined to be normal data signals; if the error exceeds the threshold, it indicates that energy is not conserved, and at least one of the voltage signal or the current signal is determined to be a first abnormal data signal.

[0056] Exemplarily, historical data of the vehicle power battery under normal working conditions is obtained, the mean and standard deviation of the voltage signal, current signal and temperature signal are calculated, and the reference range of each signal is determined using the mean and standard deviation. For example, the reference range of the voltage signal can be determined by adding or subtracting a number of times the standard deviation from the mean, and similar methods are used for the current and temperature signals. Then, the voltage signal, current signal and temperature signal are compared with their corresponding reference ranges, and the signal that exceeds its corresponding reference range is determined as the second abnormal data signal.

[0057] After determining the first abnormal data signal and the second abnormal data signal, the first abnormal data signal and the second abnormal data signal may be marked, and the first abnormal data signal and the second abnormal data signal may be deleted from the data signal.

[0058] The embodiment of the present application verifies the matching of voltage signals and current signals according to the laws of physics, and can identify abnormal data caused by sensor failure, circuit errors or data transmission problems, namely, the first abnormal data signal; by comparing the data signal with its corresponding reference range and determining the second abnormal data signal that exceeds the reference range, it is possible to determine extreme data signals that significantly deviate from normal values, thereby ensuring the rationality and consistency of the data signal, improving the quality of the data signal, and providing a reliable data basis for subsequent model training.

[0059] In an exemplary embodiment of the present application, the denoising process may include using sliding window average filtering or wavelet transform on the voltage signal and the current signal to remove high-frequency noise signals; using a low-pass filter on the temperature signal to eliminate noise caused by environmental disturbances to ensure the quality of the data signal. The normalization process may include mapping the voltage (V), current (A), and temperature (℃) of different dimensions corresponding to the voltage signal, the current signal, and the temperature signal to a unified numerical interval, such as [0,1], to eliminate the interference of magnitude differences on the model training process.

[0060] Time alignment processing can include using a unified clock source and hardware trigger signal in the data acquisition device to ensure that the sampling time of the voltage, current and temperature signals is synchronized; when storing data, the current signal, voltage signal and temperature signal are bound and stored at the same timestamp. Signal enhancement processing can include different data enhancement methods, such as time axis translation and scaling, to increase the diversity and quantity of the data set to improve the generalization ability of the model.

[0061] In step 202, the battery state detection model is trained based on the data signal to obtain data fitting loss and physical constraint loss. The data fitting loss represents the difference between the predicted value of the battery state detection model and the actual value of the battery state. The physical constraint loss represents the difference between the predicted value of the battery state detection model and the theoretical value of the physical law of the battery system. The physical constraint loss includes at least one of electrochemical loss, thermal conduction loss or battery capacity attenuation loss.

[0062] In an exemplary embodiment of the present application, the battery status detection model may be a physics-informed neural network (PINN) model, and a deep feedforward neural network (Feedforward Neural Network) may be used as the basic architecture. The deep feedforward neural network may include an input layer, a hidden layer, and an output layer.

[0063] The input layer can be composed of variables corresponding to voltage signals, current signals, and temperature signals; the hidden layer can be set to 3 to 5 layers, each layer is composed of 64 to 256 neurons, and the number of hidden layers and the number of neurons contained in each layer can be appropriately adjusted according to the specific model training effect. The activation function of the hidden layer can choose a leaky linear rectified function (Leaky ReLU). For example, the formula corresponding to the activation function is shown in formula (1):

[0064]

[0065] In the above formula (1), f(x) is the output of the Leaky ReLU function, x is the input value, and a is a constant less than 1, usually called the leakage coefficient. For example, a can be 0.1.

[0066] The activation function of the output layer can be selected as the Sigmoid function. The expression of the Sigmoid function is shown in formula (2):

[0067]

[0068] In the above formula (2), f(x) is the output of the Sigmoid function, and x is the input value. The output value f(x) of the Sigmoid function is limited to between (0,1). The output value can be an estimated value of the battery health status of the power battery, where 1 indicates complete health (capacity retention rate 100%) and 0 indicates complete damage (capacity retention rate 0%).

[0069] After determining the battery status detection model, the battery status detection model can be initialized. Initialize the weights and biases corresponding to each neural network layer. For example, according to the number of neurons n in the input layer and the number of neurons m in the output layer, create a weight matrix W of size n×m, use a random number generator, extract each element of the weight matrix W from a normal distribution with a mean of 0 and a standard deviation of σ, and obtain the initial weight of the application network layer. Set each element of the bias vector b to 0. Then assign the initialized weight matrix W and bias vector b to the corresponding fully connected layer in the neural network, and start training the battery status detection model using the data signal.

[0070] During the training process of the battery status detection model, the preprocessed data signal can be input into the input layer of the battery status detection model, and the data signal passes through the hidden layers in the neural network in turn. In each hidden layer, the output of the neuron is calculated by the activation function. Using the Leaky ReLU activation function, the output of the neuron y = max(0, x), where x is the weighted sum of the input to the neuron, that is, wo is the weight, x i is the output of the previous layer of neurons, b is the bias, i is the index variable used to traverse the output of all previous layer neurons that are input to the current neuron, and k is the number of previous layer neurons involved in the weighted sum calculation. Finally, the data reaches the output layer, which outputs the estimated value of the battery health status of the power battery.

[0071] After determining the estimated value of the battery health state, the data fitting loss and physical constraint loss of the battery state detection model can be determined, where the data fitting loss represents the difference between the predicted value of the battery state detection model and the actual value of the battery state, and the physical constraint loss represents the difference between the predicted value of the battery state detection model and the theoretical value of the physical laws of the battery system.

[0072] In one embodiment, the mean square error (MSE) may be used to determine the data fitting loss of the battery state detection model. data As shown in formula (3):

[0073]

[0074] In the above formula (3), N is the number of data signals, y i The actual value of the battery status of the data signal, It is the predicted value of the battery state output by the battery state detection model. It should be noted that the method of calculating the data fitting loss in this application is an exemplary description, and other methods can also be used to calculate the data fitting loss of the battery state detection model, for example, using the cross entropy error function to calculate the data fitting loss, which is not limited in this application.

[0075] In an exemplary embodiment of the present application, the process of determining the physical constraint loss of the battery status detection model may include steps 2021 to 2023 .

[0076] In step 2021, a signal residual is determined using the battery state detection model and the data signal, where the signal residual represents the difference between a detection value of the battery state detection model and a theoretical value, where the theoretical value is determined based on the data signal and a physical equation.

[0077] Exemplarily, the pre-processed voltage, current, and temperature signals are input into the battery state detection model to obtain the detection value; based on the physical equation of the battery system, the input signal is used to calculate the theoretical value. For example, the electrochemical theoretical value can be the theoretical voltage, which is calculated according to Ohm's law V=Vocv-I×R, where Vocv is the open circuit voltage, R is the internal resistance, and I is the current passing through the internal resistance. The heat conduction theoretical value can be the theoretical temperature, which is calculated according to the Joule heating effect ΔT=k×I 2×R×Δt, calculate the theoretical temperature change, where k is the thermal resistance coefficient, R is the internal resistance, and I is the current passing through the internal resistance. The theoretical value of battery capacity decay is the theoretical capacity decay rate, according to the aging equation SOH1=exp(-a×T×C), where SOH1 is the theoretical capacity decay rate, a is the aging rate coefficient, T is the aging temperature, and C is the number of charge and discharge times of the battery.

[0078] After obtaining the theoretical and predicted values ​​of the voltage signal, current signal, and temperature signal, the electrochemical residual, thermal conduction residual, and capacity decay residual can be calculated respectively. The residual is the difference between the predicted value and the theoretical value.

[0079] In step 2022, at least one of electrochemical loss, thermal conduction loss, or battery capacity attenuation loss is calculated based on the quantity of the data signal and the signal residual.

[0080] Among them, the electrochemical loss represents the difference between the electrical parameters detected during the training of the battery state detection model and the electrical parameters corresponding to the electrode reaction kinetics, and the electrical parameters include at least one of the voltage signal or the current signal. The thermal conduction loss represents the difference between the temperature detected during the training of the battery state detection model and the temperature corresponding to the thermal conduction law. The battery capacity attenuation loss represents the difference between the capacity attenuation detected during the training of the battery state detection model and the capacity attenuation corresponding to the aging mechanism. The electrochemical loss, thermal conduction loss and battery capacity attenuation loss can be the average of the sum of the squares of their corresponding residuals.

[0081] In step 2023, the physical constraint loss is determined based on at least one of the electrochemical loss, the thermal conduction loss or the battery capacity fade loss and the corresponding reference weight.

[0082] For example, the electrochemical loss L e The corresponding reference weight is λ e , heat conduction loss L t The corresponding reference weight is λ t , battery capacity attenuation loss L a The corresponding reference weight is λ a , then the physical constraint loss L phys =λ e ×L e +λ t ×L t +λ a ×L a .

[0083] In step 203, the model parameters of the battery state detection model are optimized using the data fitting loss and the physical constraint loss to obtain optimized model parameters.

[0084] In an exemplary embodiment of the present application, the process of optimizing the battery status detection model may include: initializing the battery status detection model to obtain an initialized battery status detection model; determining the data fitting loss and the physical constraint loss based on the data signal and the initialized battery status detection model; determining the total loss using the data fitting loss, the first weight corresponding to the data fitting loss, the physical constraint loss, and the second weight corresponding to the physical constraint loss; and optimizing the battery status detection model using the total loss to obtain optimized model parameters.

[0085] The process of initializing the battery state detection model and determining the data fitting loss and the physical constraint loss have been described in detail in step 202, and can be found in the relevant content, which will not be described again here. data The corresponding first weight λ data , physical constraint loss L phys The corresponding second weight λ phys , total loss L = λ data ×L data +λ phys ×L phys , where the first weight λ data and the second weight λ phys It can be set based on actual conditions or adjusted dynamically.

[0086] After determining the total loss corresponding to the battery state detection model, the model parameters of the battery state detection model can be updated based on the total loss. The process of updating the model parameters of the battery state detection model based on the total loss can include: determining the gradient corresponding to the total loss based on the model parameters of the battery state detection model; calculating the first-order moment estimate and the second-order moment estimate based on the gradient, the first-order moment estimate is the weighted average of the exponent of the gradient, and the second-order moment estimate is the weighted average of the exponent of the square of the gradient; using the first-order moment estimate and the second-order moment estimate to update the model parameters of the battery state detection model to obtain the optimized model parameters.

[0087] Exemplarily, the gradient is the partial derivative of the total loss function with respect to the model parameters, which is used to indicate the direction of the model parameter update. For example, starting from the output layer, the partial derivative of the total loss with respect to the parameters is calculated layer by layer, that is, the gradient of the total loss with respect to W and the gradient of b.

[0088] The first-order moment estimate of the gradient can be calculated by formula (4), and the second-order moment estimate of the gradient can be calculated by formula (5).

[0089] m t =β1×m t-1 +(1-β1)×g t Formula (4)

[0090]

[0091] In the above formula (4), m t is the first-order moment estimate of the t-th iteration, m t-1 is the first-order moment estimate of the t-1th iteration, β1 is the decay rate, g t is the gradient of the tth iteration, where the initial value m0=0. In the above formula (5), v t is the second-order moment estimate of the t-th iteration, v t-1 is the second-order moment estimate of the t-1th iteration, β2 is the decay rate, g t is the gradient of the tth iteration, where the initial value v0=0.

[0092] In determining the first-order moment estimate m t and the second-order moment estimate v t Afterwards, the updated weight can also be calculated through the first-order moment estimation and the second-order moment estimation, as shown in formula (6).

[0093]

[0094] In the above formula (4), W t+1 is the weight of the model at the t+1th iteration; W t is the weight of the model at the tth iteration; α is the learning rate, a hyperparameter used to control the step size of the weight update; ∈ is a very small number used to prevent the denominator from being zero and increase numerical stability; m t is the first-order moment estimate of the t-th iteration, v t is the second-order moment estimate of the t-th iteration.

[0095] During the training process, the number of iterations or the change in total loss can be used to determine whether the training of the battery status detection model is stopped. For example, if the number of iterations reaches the threshold, the model training stops; if the total loss does not decrease for several consecutive iterations, decreases very little, or starts to increase (overfitting occurs), the model training stops. After the model training stops, the optimized model parameters and the optimized model can be obtained.

[0096] In step 204, a trained battery status detection model is determined based on the optimized model parameters, and the trained battery status detection model is used to detect the battery status of the power battery of the vehicle.

[0097] In an exemplary embodiment of the present application, after obtaining the optimized model parameters, the optimized model parameters are configured to the battery status detection model to obtain a trained battery status detection model.

[0098] In the process of training the battery status detection model, the present application uses data fitting loss and physical constraint loss to make the battery status detection model follow physical laws during the training process, reduce dependence on large amounts of data, and converge faster during training, reducing the consumption of computing resources. In addition, the battery status detection model is not trained using a large amount of data, and the number of iterations in the battery status detection model training process is small, and the training error will not increase with the number of iterations, so that the trained battery status detection model can have a higher detection accuracy.

[0099] The present application also provides a battery status detection method, which can be applied to the above Figure 1 For example, the method can be implemented by Figure 1 The terminal device 101 and the server 102 in the embodiment interact with each other. Figure 3 As shown, the method includes the following steps 301 to 302.

[0100] In step 301, a data signal of a vehicle is acquired, where the data signal includes at least one of a voltage signal, a current signal or a temperature signal.

[0101] The process of acquiring the vehicle data signal may refer to the relevant description of step 201, and will not be elaborated here.

[0102] In step 302, the battery status of the vehicle is detected based on the battery status detection model and the data signal to obtain a detection result of the battery status.

[0103] In an exemplary embodiment of the present application, based on the data signals (battery voltage signal, current signal and temperature signal) collected in real time, accurate detection of the power battery state of health (SOH) is achieved by completing the training of the battery state detection model.

[0104] For example, the vehicle data signal is obtained through the vehicle sensor network (including voltage sensor, current sensor and temperature sensor), and the vehicle data signal can be a voltage signal, current signal and temperature signal collected in real time. The obtained vehicle signal is input into the battery status detection model, and the battery status detection model can process the input data signal and output the battery health status (SOH).

[0105] Among them, the battery status detection model can also accurately detect the state of charge (SOC) and safety status of the power battery. It should be noted that the detection method of the state of charge (SOC) and safety status of the power battery is similar to the detection method of the battery health state (SOH), which will not be repeated here. The battery status detection model can analyze the dynamic changes of voltage and current and temperature distribution during the charging and discharging process in real time, evaluate the battery capacity attenuation, remaining power and potential risks (such as voltage imbalance, local overheating), and output the detection results.

[0106] Take the output test result of the battery health status as an example for explanation. The battery health status can be expressed as a normalized value or percentage. If the battery health status is greater than or equal to the healthy operation threshold of the power battery (for example, 0.8), the battery health status is output; if the battery health status is less than the healthy operation threshold of the power battery, the battery health status and prompt information are output. The prompt information can be used to remind the customer to maintain or replace the power battery in time. Optionally, the output result of the battery health status can be recorded for subsequent management and evaluation of the power battery.

[0107] The exemplary embodiment of the present application detects the battery health status through a battery status detection model, and the user can reasonably arrange the battery use and maintenance plan based on the detection results. For example, when the battery performance begins to decline, appropriate maintenance measures such as balanced charging and temperature control can be taken to extend the battery life, which can not only reduce the frequency of battery replacement, but also reduce the cost of battery use. In addition, when an abnormal battery health status is detected, a reminder can be given in time to effectively prevent the occurrence of battery failure and improve the stability of the power system.

[0108] The embodiment of the present application also provides a training device for a battery status detection model. Figure 4 is a structural diagram of a training device for a battery status detection model provided in an embodiment of the present application, such as Figure 4 As shown, the device comprises:

[0109] A first acquisition module 401 is used to acquire a data signal of a vehicle, where the data signal includes at least one of a voltage signal, a current signal or a temperature signal of a power battery of the vehicle;

[0110] A training module 402 is used to train the battery state detection model based on the data signal to obtain a data fitting loss and a physical constraint loss, wherein the data fitting loss represents the difference between the predicted value of the battery state detection model and the actual value of the battery state, and the physical constraint loss represents the difference between the predicted value of the battery state detection model and the theoretical value of the physical law of the battery system, and the physical constraint loss includes at least one of an electrochemical loss, a thermal conduction loss, or a battery capacity attenuation loss;

[0111] The training module 402 is further used to optimize the model parameters of the battery state detection model by using the data fitting loss and the physical constraint loss to obtain the optimized model parameters;

[0112] The training module 402 is also used to determine a trained battery status detection model based on the optimized model parameters, and the trained battery status detection model is used to detect the battery status of the power battery of the vehicle.

[0113] In one possible implementation, the training module 402 is used to determine the signal residual using the battery state detection model and the data signal, the signal residual characterizing the difference between the detection value of the battery state detection model and the theoretical value, the theoretical value being determined based on the data signal and the physical equation; calculating at least one of the electrochemical loss, thermal conduction loss, or battery capacity attenuation loss based on the number of data signals and the signal residual; and determining the physical constraint loss based on at least one of the electrochemical loss, thermal conduction loss, or battery capacity attenuation loss and the corresponding reference weight.

[0114] In one possible implementation, the theoretical value is obtained based on the physical equations and data signals of the battery system, and the theoretical value includes at least one of theoretical voltage, theoretical current, theoretical temperature change or theoretical capacity decay rate; wherein the theoretical voltage and theoretical current are calculated based on Ohm's law and voltage signal or current signal, and the theoretical temperature change is calculated based on temperature signal and Joule heating effect; the theoretical value of battery capacity decay is obtained based on data signals and aging equations.

[0115] In one possible implementation, the training module 402 is used to initialize the battery status detection model to obtain an initialized battery status detection model; determine the data fitting loss and the physical constraint loss based on the data signal and the initialized battery status detection model; determine the total loss using the data fitting loss, the first weight corresponding to the data fitting loss, the physical constraint loss, and the second weight corresponding to the physical constraint loss; and optimize the battery status detection model using the total loss to obtain optimized model parameters.

[0116] In one possible implementation, the training module 402 is used to determine the gradient corresponding to the total loss based on the model parameters of the battery status detection model; calculate the first-order moment estimate and the second-order moment estimate based on the gradient, the first-order moment estimate is the weighted average of the exponent of the gradient, and the second-order moment estimate is the weighted average of the exponent of the gradient squared; use the first-order moment estimate and the second-order moment estimate to update the model parameters of the battery status detection model to obtain optimized model parameters.

[0117] In one possible implementation, the device also includes a processing module (not shown in the figure), which is used to preprocess the data signal to obtain a preprocessed data signal, wherein the preprocessing includes at least one of outlier removal, denoising, normalization, time alignment or signal enhancement; and a training module 402 is used to train a battery status detection model using the preprocessed data signal to obtain data fitting loss and physical constraint loss.

[0118] In one possible implementation, preprocessing includes outlier removal processing, and processing module 402 is used to verify the data signal based on physical laws, determine a first abnormal data signal, and the abnormal data signal is a data signal where the voltage signal and the current signal do not match; determine a reference range in the data signal, and determine the data signal exceeding the reference range as a second abnormal data signal; and delete the first abnormal data signal and the second abnormal data signal in the data signal.

[0119] In the process of training the battery status detection model, the present application uses data fitting loss and physical constraint loss to make the battery status detection model follow physical laws during the training process, reduce dependence on large amounts of data, and converge faster during training, reducing the consumption of computing resources. In addition, the battery status detection model is not trained using a large amount of data, and the number of iterations in the battery status detection model training process is small, and the training error will not increase with the number of iterations, so that the trained battery status detection model can have a higher detection accuracy.

[0120] The embodiment of the present application also provides a battery status detection device. Figure 5 is a schematic diagram of the structure of a battery status detection device provided in an embodiment of the present application, such as Figure 5 As shown, the device comprises:

[0121] A second acquisition module 501 is used to acquire a data signal of the vehicle, where the data signal includes at least one of a voltage signal, a current signal or a temperature signal;

[0122] The detection module 502 is used to detect the battery status of the vehicle based on the battery status detection model and the data signal to obtain the detection result of the battery status. The battery status detection model is trained based on any of the above methods.

[0123] The exemplary embodiment of the present application detects the battery health status through the battery status detection model, and the user can reasonably arrange the battery usage and maintenance plan based on the detection results. For example, when the battery performance begins to decline, appropriate maintenance measures such as balanced charging and temperature control can be taken to extend the battery life, which can not only reduce the frequency of battery replacement, but also reduce the cost of battery use.

[0124] It should be understood that the above-mentioned device only uses the division of the above-mentioned functional modules as an example to illustrate when implementing its functions. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiments belong to the same concept, and their specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0125] Figure 6 1 is a block diagram of a terminal device provided in an embodiment of the present application. The terminal device 1200 may be any electronic device product that can interact with a user through one or more methods such as a keyboard, a touchpad, a remote control, voice interaction, or a handwriting device. For example, a PC (Personal Computer), a mobile phone, a smart phone, a PDA (Personal Digital Assistant), a wearable device, a PPC (Pocket PC), a tablet computer, etc.

[0126] Typically, the terminal device 1200 includes: a processor 1201 and a memory 1202 .

[0127] The processor 1201 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 1201 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 1201 may also include a main processor and a coprocessor. The main processor is a processor for processing data in an awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in a standby state. In some embodiments, the processor 1201 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 1201 may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0128] The memory 1202 may include one or more computer-readable storage media, which may be non-transitory. The memory 1202 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 1202 is used to store at least one instruction, which is used to be executed by the processor 1201 to implement the training method of the battery status detection model or the battery status detection method provided in the method embodiment of the present application.

[0129] In some embodiments, the terminal device 1200 may further optionally include: a peripheral device interface 1203 and at least one peripheral device. The processor 1201, the memory 1202 and the peripheral device interface 1203 may be connected via a bus or a signal line. Each peripheral device may be connected to the peripheral device interface 1203 via a bus, a signal line or a circuit board. Specifically, the peripheral device includes: at least one of a radio frequency circuit 1204, a display screen 1205, a camera assembly 1206, an audio circuit 1207 and a power supply 1208.

[0130] The peripheral device interface 1203 may be used to connect at least one peripheral device related to I / O (Input / Output) to the processor 1201 and the memory 1202. In some embodiments, the processor 1201, the memory 1202, and the peripheral device interface 1203 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 1201, the memory 1202, and the peripheral device interface 1203 may be implemented on a separate chip or circuit board, which is not limited in this embodiment.

[0131] The radio frequency circuit 1204 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The radio frequency circuit 1204 communicates with the communication network and other communication devices through electromagnetic signals. The radio frequency circuit 1204 converts the electrical signal into an electromagnetic signal for transmission, or converts the received electromagnetic signal into an electrical signal. Optionally, the radio frequency circuit 1204 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a user identity module card, and the like. The radio frequency circuit 1204 can communicate with other terminal devices through at least one wireless communication protocol. The wireless communication protocol includes, but is not limited to: the World Wide Web, a metropolitan area network, an intranet, various generations of mobile communication networks (2G, 3G, 4G and 5G), a wireless local area network and / or a WiFi (Wireless Fidelity) network. In some embodiments, the radio frequency circuit 1204 may also include circuits related to NFC (Near Field Communication), which is not limited in this application.

[0132] The display screen 1205 is used to display a UI (User Interface). The UI may include graphics, text, icons, videos, and any combination thereof. When the display screen 1205 is a touch display screen, the display screen 1205 also has the ability to collect touch signals on the surface or above the surface of the display screen 1205. The touch signal can be input to the processor 1201 as a control signal for processing. At this time, the display screen 1205 can also be used to provide virtual buttons and / or virtual keyboards, also known as soft buttons and / or soft keyboards. In some embodiments, the display screen 1205 can be one, set on the front panel of the terminal device 1200; in other embodiments, the display screen 1205 can be at least two, respectively set on different surfaces of the terminal device 1200 or in a folding design; in other embodiments, the display screen 1205 can be a flexible display screen, set on a curved surface or a folding surface of the terminal device 1200. Even, the display screen 1205 can also be set to a non-rectangular irregular shape, that is, a special-shaped screen. The display screen 1205 can be made of materials such as LCD (Liquid Crystal Display) and OLED (Organic Light-Emitting Diode).

[0133] The camera assembly 1206 is used to capture images or videos. Optionally, the camera assembly 1206 includes a front camera and a rear camera. Typically, the front camera is arranged on the front panel of the terminal device 1200, and the rear camera is arranged on the back of the terminal device 1200. In some embodiments, there are at least two rear cameras, which are any one of a main camera, a depth of field camera, a wide-angle camera, and a telephoto camera, so as to realize the fusion of the main camera and the depth of field camera to realize the background blur function, the fusion of the main camera and the wide-angle camera to realize panoramic shooting and VR (Virtual Reality) shooting function or other fusion shooting functions. In some embodiments, the camera assembly 1206 may also include a flash. The flash can be a monochrome temperature flash or a dual-color temperature flash. A dual-color temperature flash refers to a combination of a warm light flash and a cold light flash, which can be used for light compensation at different color temperatures.

[0134] The audio circuit 1207 may include a microphone and a speaker. The microphone is used to collect sound waves from the user and the environment, and convert the sound waves into electrical signals and input them into the processor 1201 for processing, or input them into the radio frequency circuit 1204 to achieve voice communication. For the purpose of stereo acquisition or noise reduction, there may be multiple microphones, which are respectively arranged at different parts of the terminal device 1200. The microphone may also be an array microphone or an omnidirectional acquisition microphone. The speaker is used to convert the electrical signal from the processor 1201 or the radio frequency circuit 1204 into sound waves. The speaker may be a traditional film speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can not only convert the electrical signal into sound waves audible to humans, but also convert the electrical signal into sound waves inaudible to humans for purposes such as ranging. In some embodiments, the audio circuit 1207 may also include a headphone jack.

[0135] The power supply 1208 is used to power various components in the terminal device 1200. The power supply 1208 can be an alternating current, a direct current, a disposable battery, or a rechargeable battery. When the power supply 1208 includes a rechargeable battery, the rechargeable battery can be a wired rechargeable battery or a wireless rechargeable battery. A wired rechargeable battery is a battery charged through a wired line, and a wireless rechargeable battery is a battery charged through a wireless coil. The rechargeable battery can also be used to support fast charging technology.

[0136] In some embodiments, the terminal device 1200 further includes one or more sensors 1210 , including but not limited to: an acceleration sensor 1211 , a gyroscope sensor 1212 , a pressure sensor 1213 , an optical sensor 1214 , and a proximity sensor 1215 .

[0137] The acceleration sensor 1211 can detect the magnitude of acceleration on the three coordinate axes of the coordinate system established by the terminal device 1200. For example, the acceleration sensor 1211 can be used to detect the components of gravity acceleration on the three coordinate axes. The processor 1201 can control the display screen 1205 to display the user interface in a horizontal view or a vertical view according to the gravity acceleration signal collected by the acceleration sensor 1211. The acceleration sensor 1211 can also be used to collect game or user motion data.

[0138] The gyro sensor 1212 can detect the body direction and rotation angle of the terminal device 1200, and the gyro sensor 1212 can cooperate with the acceleration sensor 1211 to collect the user's 3D actions on the terminal device 1200. The processor 1201 can implement the following functions based on the data collected by the gyro sensor 1212: motion sensing (such as changing the UI according to the user's tilt operation), image stabilization during shooting, game control, and inertial navigation.

[0139] The pressure sensor 1213 can be set on the side frame of the terminal device 1200 and / or the lower layer of the display screen 1205. When the pressure sensor 1213 is set on the side frame of the terminal device 1200, it can detect the user's holding signal of the terminal device 1200, and the processor 1201 performs left and right hand recognition or shortcut operation according to the holding signal collected by the pressure sensor 1213. When the pressure sensor 1213 is set on the lower layer of the display screen 1205, the processor 1201 controls the operability controls on the UI interface according to the user's pressure operation on the display screen 1205. The operability controls include at least one of a button control, a scroll bar control, an icon control, and a menu control.

[0140] The optical sensor 1214 is used to collect the ambient light intensity. In one embodiment, the processor 1201 can control the display brightness of the display screen 1205 according to the ambient light intensity collected by the optical sensor 1214. Specifically, when the ambient light intensity is high, the display brightness of the display screen 1205 is increased; when the ambient light intensity is low, the display brightness of the display screen 1205 is reduced. In another embodiment, the processor 1201 can also dynamically adjust the shooting parameters of the camera assembly 1206 according to the ambient light intensity collected by the optical sensor 1214.

[0141] The proximity sensor 1215, also called a distance sensor, is usually arranged on the front panel of the terminal device 1200. The proximity sensor 1215 is used to collect the distance between the user and the front of the terminal device 1200. In one embodiment, when the proximity sensor 1215 detects that the distance between the user and the front of the terminal device 1200 is gradually decreasing, the processor 1201 controls the display screen 1205 to switch from the screen-on state to the screen-off state; when the proximity sensor 1215 detects that the distance between the user and the front of the terminal device 1200 is gradually increasing, the processor 1201 controls the display screen 1205 to switch from the screen-off state to the screen-on state.

[0142] Those skilled in the art will understand that Figure 6 The structure shown in the figure does not constitute a limitation on the terminal device 1200, and may include more or fewer components than shown in the figure, or combine certain components, or adopt a different component arrangement.

[0143] Figure 7 It is a structural diagram of a server provided in an embodiment of the present application. The server 1300 may have relatively large differences due to different configurations or performances, and may include one or more processors 1301 and one or more memories 1302, wherein the one or more memories 1302 store at least one program code, and the at least one program code is loaded and executed by the one or more processors 1301 to implement the training method of the battery status detection model or the battery status detection method provided in the above-mentioned various method embodiments. Of course, the server 1300 may also have components such as a wired or wireless network interface, a keyboard, and an input and output interface for input and output. The server 1300 may also include other components for implementing device functions, which will not be repeated here.

[0144] In an exemplary embodiment, a computer-readable storage medium is also provided, in which at least one program code is stored. The at least one program code is loaded and executed by a processor to enable a computer to implement any of the above-mentioned battery status detection model training methods or battery status detection methods.

[0145] Optionally, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc (CD-ROM), a magnetic tape, a floppy disk, an optical data storage device, etc.

[0146] In an exemplary embodiment, a computer program or a computer program product is also provided, in which at least one computer instruction is stored, and the at least one computer instruction is loaded and executed by a processor so that a computer implements any of the above-mentioned battery status detection model training methods or battery status detection methods.

[0147] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.) and signals involved in this application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions. For example, the vehicle data signals, optimized model parameters, data fitting losses and physical constraint losses involved in this application are all obtained with full authorization.

[0148] It should be understood that the "plurality" mentioned in this article refers to two or more. "And / or" describes the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects are in an "or" relationship.

[0149] The above description is only an exemplary embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A training method for a battery status detection model, characterized in that: The method comprises: Acquiring a data signal of the vehicle, the data signal comprising at least one of a voltage signal, a current signal or a temperature signal of a power battery of the vehicle; Training a battery state detection model based on the data signal to obtain a data fitting loss and a physical constraint loss, wherein the data fitting loss represents a difference between a predicted value of the battery state detection model and an actual value of the battery state, and the physical constraint loss represents a difference between a predicted value of the battery state detection model and a theoretical value of a physical law of a battery system, wherein the physical constraint loss includes at least one of an electrochemical loss, a thermal conduction loss, or a battery capacity attenuation loss; Optimizing the model parameters of the battery state detection model by using the data fitting loss and the physical constraint loss to obtain optimized model parameters; A trained battery status detection model is determined based on the optimized model parameters, and the trained battery status detection model is used to detect the battery status of the power battery of the vehicle.

2. The method according to claim 1, characterized in that The step of training the battery state detection model based on the data signal to obtain data fitting loss and physical constraint loss includes: Determine a signal residual using the battery state detection model and the data signal, wherein the signal residual represents a difference between a detection value of the battery state detection model and a theoretical value, wherein the theoretical value is determined based on the data signal and a physical equation; Calculating at least one of the electrochemical loss, the thermal conduction loss, or the battery capacity attenuation loss based on the number of the data signals and the signal residual; The physical constraint loss is determined based on at least one of the electrochemical loss, the thermal conduction loss, or the battery capacity fade loss and a corresponding reference weight.

3. The method according to claim 2, characterized in that The theoretical value is obtained based on a physical equation of the battery system and a data signal, and the theoretical value includes at least one of a theoretical voltage, a theoretical current, a theoretical temperature change, or a theoretical capacity decay rate; Among them, the theoretical voltage and the theoretical current are calculated based on Ohm's law and the voltage signal or the current signal, the theoretical temperature change is calculated based on the temperature signal and Joule thermal effect; the theoretical value of the battery capacity attenuation is obtained based on the data signal and the aging equation.

4. The method according to claim 1, characterized in that: The optimizing the model parameters of the battery state detection model by using the data fitting loss and the physical constraint loss to obtain the optimized model parameters includes: Initializing the battery status detection model to obtain an initialized battery status detection model; determining the data fitting loss and the physical constraint loss based on the data signal and the initialized battery state detection model; Determine a total loss using the data fitting loss, a first weight corresponding to the data fitting loss, the physical constraint loss, and a second weight corresponding to the physical constraint loss; The battery state detection model is optimized using the total loss to obtain the optimized model parameters.

5. The method according to claim 4, characterized in that The optimizing the battery status detection model by utilizing the total loss to obtain the optimized model parameters includes: Determining a gradient corresponding to the total loss based on a model parameter of the battery state detection model; Calculating a first-order moment estimate and a second-order moment estimate based on the gradient, the first-order moment estimate being a weighted average of exponents of the gradient, and the second-order moment estimate being a weighted average of exponents of squares of the gradient; The model parameters of the battery state detection model are updated using the first-order moment estimation and the second-order moment estimation to obtain the optimized model parameters.

6. The method according to any one of claims 1 to 5, characterized in that: After obtaining the data signal of the vehicle, the method further includes: Preprocessing the data signal to obtain a preprocessed data signal, wherein the preprocessing includes at least one of outlier removal, denoising, normalization, time alignment, or signal enhancement; The step of training the battery state detection model based on the data signal to obtain data fitting loss and physical constraint loss includes: The battery state detection model is trained using the preprocessed data signal to obtain the data fitting loss and the physical constraint loss.

7. The method according to claim 6, characterized in that The preprocessing includes outlier removal processing, and the outlier removal processing includes: Verify the data signal based on physical laws to determine a first abnormal data signal, where the abnormal data signal is a data signal where a voltage signal and a current signal do not match; determining a reference range in the data signal, and determining a data signal exceeding the reference range as a second abnormal data signal; The first abnormal data signal and the second abnormal data signal are deleted from the data signal.

8. A battery status detection method, characterized in that: The method comprises: Acquire a data signal of the vehicle, wherein the data signal includes at least one of a voltage signal, a current signal or a temperature signal; The battery status of the vehicle is detected based on a battery status detection model and the data signal to obtain a detection result of the battery status, wherein the battery status detection model is trained based on the method described in any one of claims 1 to 7.

9. A training device for a battery status detection model, characterized in that: The device comprises: A first acquisition module, used to acquire a data signal of the vehicle, wherein the data signal includes at least one of a voltage signal, a current signal or a temperature signal of a power battery of the vehicle; a training module, configured to train a battery state detection model based on the data signal to obtain a data fitting loss and a physical constraint loss, wherein the data fitting loss represents a difference between a predicted value of the battery state detection model and an actual value of the battery state, and the physical constraint loss represents a difference between a predicted value of the battery state detection model and a theoretical value of a physical law of a battery system, wherein the physical constraint loss includes at least one of an electrochemical loss, a thermal conduction loss, or a battery capacity attenuation loss; The training module is further used to optimize the model parameters of the battery state detection model by using the data fitting loss and the physical constraint loss to obtain optimized model parameters; The training module is also used to determine a trained battery status detection model based on the optimized model parameters, and the trained battery status detection model is used to detect the battery status of the vehicle's power battery.

10. A battery status detection device, characterized in that: The device comprises: A second acquisition module, used to acquire a data signal of the vehicle, wherein the data signal includes at least one of a voltage signal, a current signal or a temperature signal; A detection module is used to detect the battery status of the vehicle based on a battery status detection model and the data signal to obtain a detection result of the battery status, wherein the battery status detection model is trained based on the method described in any one of claims 1 to 7.

Images