Sensor fault diagnosis methods, devices, equipment and storage media

Abstract

This application discloses a sensor fault diagnosis method, apparatus, device, and storage medium. The method includes: acquiring the current detection signal of a current sensor and the single-cell voltage detection signal of each voltage sensor; importing the sum of the single-cell voltage detection signals and the current detection signal into a preset observer to obtain an estimated state of charge (SOC) value of the target battery pack; if the SOC value indicates a sensor fault in the target battery pack, calculating the correlation coefficient between the single-cell voltage detection signals of every two adjacent battery cells in the target battery pack; and performing fault diagnosis based on the correlation coefficients. This application significantly reduces the computational load and improves computational efficiency and fault diagnosis sensitivity by constructing a sensor fault diagnosis method that combines an observer with correlation coefficients, thereby ensuring the reliability of the battery management system.

Description

Technical Field

[0001] This application relates to the field of battery management technology, specifically to a sensor fault diagnosis method, apparatus, device, and storage medium. Background Technology

[0002] Power batteries are the primary energy source for electric vehicles, electric bicycles, and other similar devices, and their performance directly affects the operation of these devices. Currently, battery systems consist of battery modules and a Battery Management System (BMS). Collecting the necessary data is fundamental to the BMS's functionality; all algorithms and effective control implemented within the BMS rely on data acquisition by sensors. The sampling rate and accuracy of sensors are crucial indicators of sensor quality and, consequently, significant factors affecting battery system performance. Therefore, sensors and their performance hold an extremely important position within battery systems.

[0003] In related technologies, fault diagnosis of sensors is mainly achieved by constructing observers. For example, an observer is constructed for each battery cell in a battery module. The input of each observer is the current sensor measurement and the voltage sensor measurement of the corresponding battery cell. The output of each observer is the residual formed by the voltage estimate and the voltage sensor measurement of the corresponding battery cell. It is evident that the residual is affected by both the current and voltage sensor measurements. Therefore, when the current sensor fails, the residual generated by each observer will be affected. However, when the voltage sensor of a particular battery cell fails, only the residual output by the observer corresponding to the failed cell will be affected. Fault diagnosis of the sensor can be achieved by analyzing the combined response of the residuals.

[0004] However, since an observer is built for each battery cell in the relevant technology, the computational load is very large if there are many battery cells, resulting in low computational efficiency and thus reducing the sensitivity of fault diagnosis. Summary of the Invention

[0005] This application provides a sensor fault diagnosis method, apparatus, device, and storage medium, aiming to solve the problems of low computational efficiency and low fault diagnosis sensitivity in related technologies that use the method of building an observer for each battery cell to diagnose sensor faults in a battery system.

[0006] In a first aspect, this application provides a sensor fault diagnosis method. This method is used to diagnose faults in a current sensor in a target battery pack and in a voltage sensor corresponding to each battery cell connected in series in the target battery pack. The current sensor is used to sample the total current of the target battery pack, and the voltage sensor is used to sample the individual voltage of the corresponding battery cell. The method includes:

[0007] The current detection signal obtained by the current sensor sampling the total current of the target battery pack and the single cell voltage detection signal obtained by the voltage sensor sampling the single cell voltage of the corresponding battery cell are acquired.

[0008] The sum of the voltage detection signals of each individual cell and the current detection signal are imported into a preset observer to obtain the estimated state of charge of the target battery pack.

[0009] If the state of charge estimate indicates that there is a sensor fault in the target battery pack, then calculate the correlation coefficient between the single cell voltage detection signals of every two adjacent battery cells in the target battery pack.

[0010] Fault diagnosis is performed on the current sensor and each voltage sensor based on the correlation coefficients.

[0011] In one possible implementation of this application, fault diagnosis is performed on the current sensor and each voltage sensor based on various correlation coefficients, including:

[0012] If there are no abnormalities in the correlation coefficients, then it is determined that the voltage sensors are not faulty and the current sensors are faulty.

[0013] If two adjacent correlation coefficients are abnormal, it is determined that the voltage sensor corresponding to the battery cell associated with the two adjacent correlation coefficients is faulty, while the current sensor and other voltage sensors in the target battery pack are not faulty.

[0014] In one possible implementation of this application, calculating the correlation coefficient between the individual cell voltage detection signals of every two adjacent cells in the target battery pack includes:

[0015] The voltage detection signals of each pair of adjacent battery cells are filtered using a preset sliding window, and the correlation coefficient between the voltage detection signals of the two cells within the sliding window is calculated.

[0016] In one possible implementation of this application, the sum of the voltage detection signals of each individual cell and the current detection signal are imported into a preset observer to obtain an estimated value of the state of charge of the target battery pack. Prior to this, the method further includes:

[0017] Construct an equivalent model of the target battery pack based on its performance parameters;

[0018] The open-circuit voltage of the target battery pack is tested to obtain the open-circuit voltage of the target battery pack, and the relationship between the open-circuit voltage and the state of charge of the target battery pack is established.

[0019] The target battery pack was subjected to simulated dynamic stress test to obtain the voltage and current correspondence under the simulated dynamic stress test. Based on the voltage and current correspondence, the model parameters in the equivalent model were identified to determine the quantitative relationship between the model parameters and the state of charge.

[0020] An observer is established based on the relationship between open-circuit voltage and the state of charge of the target battery pack, as well as the quantitative relationship between model parameters and state of charge.

[0021] In one possible implementation of this application, an open-circuit voltage test is performed on the target battery pack to obtain the open-circuit voltage of the target battery pack, including:

[0022] A trickle discharge test was performed on each cell in the target battery pack to obtain the open-circuit voltage of each cell.

[0023] The open-circuit voltage of the target battery pack is obtained by summing the open-circuit voltages of each individual cell.

[0024] In one possible implementation of this application, the model parameters in the equivalent model are identified based on the voltage-current correspondence, including:

[0025] Based on the voltage-current correspondence, one or more algorithms, including recursive least squares with genetic factors, particle swarm optimization, and genetic algorithm, are used to identify the model parameters in the equivalent model.

[0026] In one possible implementation of this application, the observer estimates the state of charge of the target battery pack based on one or more of the extended Kalman filter algorithm, the state estimation algorithm, and the data-driven algorithm to obtain the estimated state of charge of the target battery pack.

[0027] Secondly, this application also provides a sensor fault diagnosis device, which is used to diagnose faults in a current sensor in a target battery pack and a voltage sensor corresponding to each battery cell connected in series in the target battery pack. The current sensor is used to sample the total current of the target battery pack, and the voltage sensor is used to sample the individual voltage of the corresponding battery cell. The device includes:

[0028] The acquisition module is used to acquire the current detection signal obtained by the current sensor sampling the total current of the target battery pack and the single cell voltage detection signal obtained by the voltage sensor sampling the single cell voltage of the corresponding battery cell.

[0029] The state of charge estimation module is used to import the sum of the voltage detection signals of each cell and the current detection signal into a preset observer to obtain the estimated state of charge of the target battery pack.

[0030] The correlation coefficient calculation module is used to calculate the correlation coefficient between the single-cell voltage detection signals of every two adjacent battery cells in the target battery pack if the state of charge estimate indicates that there is a sensor fault in the target battery pack.

[0031] The diagnostic module is used to diagnose faults in the current sensor and each voltage sensor based on various correlation coefficients.

[0032] In one possible implementation of this application, the diagnostic module is specifically used for:

[0033] If there are no abnormalities in the correlation coefficients, then it is determined that the voltage sensors are not faulty and the current sensors are faulty.

[0034] If two adjacent correlation coefficients are abnormal, it is determined that the voltage sensor corresponding to the battery cell associated with the two adjacent correlation coefficients is faulty, while the current sensor and other voltage sensors in the target battery pack are not faulty.

[0035] In one possible implementation of this application, the correlation coefficient calculation module is specifically used for:

[0036] The voltage detection signals of each pair of adjacent battery cells are filtered using a preset sliding window, and the correlation coefficient between the voltage detection signals of the two cells within the sliding window is calculated.

[0037] In one possible implementation of this application, the sensor fault diagnosis device further includes a construction module. Before the state of charge estimation module imports the sum of the voltage detection signals of each individual cell and the current detection signal into a preset observer to obtain the estimated state of charge of the target battery pack, the construction module is used for:

[0038] Construct an equivalent model of the target battery pack based on its performance parameters;

[0039] The open-circuit voltage of the target battery pack is tested to obtain the open-circuit voltage of the target battery pack, and the relationship between the open-circuit voltage and the state of charge of the target battery pack is established.

[0040] The target battery pack was subjected to simulated dynamic stress test to obtain the voltage and current correspondence under the simulated dynamic stress test. Based on the voltage and current correspondence, the model parameters in the equivalent model were identified to determine the quantitative relationship between the model parameters and the state of charge.

[0041] An observer is established based on the relationship between open-circuit voltage and the state of charge of the target battery pack, as well as the quantitative relationship between model parameters and state of charge.

[0042] In one possible implementation of this application, the construction module is specifically used for:

[0043] A trickle discharge test was performed on each cell in the target battery pack to obtain the open-circuit voltage of each cell.

[0044] The open-circuit voltage of the target battery pack is obtained by summing the open-circuit voltages of each individual cell.

[0045] In one possible implementation of this application, the construction module is specifically used for:

[0046] Based on the voltage-current correspondence, one or more algorithms, including recursive least squares with genetic factors, particle swarm optimization, and genetic algorithm, are used to identify the model parameters in the equivalent model.

[0047] Thirdly, this application also provides a sensor fault diagnosis device, which includes a memory and a processor. The memory is used to store a computer program, and when the computer program is executed by the processor, it is used to implement the sensor fault diagnosis method in the first aspect or any possible implementation of the first aspect.

[0048] Fourthly, this application also provides a computer-readable storage medium storing a computer program that is executed by a processor to implement the steps of the sensor fault diagnosis method of the first aspect or any possible implementation of the first aspect.

[0049] From the above, it can be concluded that this application has the following beneficial effects:

[0050] In this application, the observer estimates the state of charge (SOC) of the target battery pack based on the current detection signal from the current sensor and the single-cell voltage detection signal from the voltage sensor corresponding to each battery cell. Based on the SOC estimate, it determines whether a sensor fault exists in the target battery pack. When a sensor fault is determined, for each battery cell in the target battery pack, the correlation coefficient between the single-cell voltage detection signals of every two adjacent battery cells is calculated. Based on the calculated correlation coefficient, fault diagnosis is performed on the current sensor and each voltage sensor. Compared to the prior art, which constructs an observer for each battery cell to diagnose sensor faults, this significantly reduces the computational load, thereby improving computational efficiency and fault diagnosis sensitivity, and ensuring the reliability of the battery management system. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in this application, the accompanying drawings used in the description of this application will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0052] Figure 1 This is a schematic diagram of an architecture for a sensor fault diagnosis strategy in existing technology;

[0053] Figure 2 This is a flowchart illustrating a sensor fault diagnosis method provided in an embodiment of this application;

[0054] Figure 3 This is a schematic diagram of the equivalent model of the target battery pack provided in the embodiments of this application;

[0055] Figure 4 This is a schematic diagram of a model for fault diagnosis of a voltage sensor based on the correlation coefficient provided in an embodiment of this application;

[0056] Figure 5 This is a schematic diagram of an architecture for a sensor fault diagnosis strategy provided in an embodiment of this application;

[0057] Figure 6 This is a schematic diagram of the sensor fault diagnosis device provided in the embodiments of this application;

[0058] Figure 7 This is a schematic diagram of a sensor fault diagnosis device provided in the embodiments of this application.

[0059] Figure label:

[0060] 600 - Sensor fault diagnosis device; 601 - Acquisition module; 602 - State of charge estimation module; 603 - Correlation coefficient calculation module; 604 - Diagnosis module; 605 - Construction module;

[0061] 701 - Processor; 702 - Memory; 703 - Power supply; 704 - Input unit; 705 - Output unit. Detailed Implementation

[0062] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0063] In the description of this application, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are used only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more features. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified.

[0064] In this application, the term "exemplary" is used to mean "used as an example, illustration, or description." Any embodiment described as "exemplary" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of this application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0065] Before introducing the sensor fault diagnosis method, apparatus, device and storage medium of this application, we will first introduce the sensor fault diagnosis method in the prior art.

[0066] In battery system sensors, current sensors directly impact the estimation of the battery's state of charge (SOC) and energy state. Since battery model parameters are affected by SOC and temperature, data collected by current and voltage sensors can also be used to correct these parameters. When sensors malfunction, the battery management system (BMS) is unable to accurately capture the battery module's state, potentially leading to incorrect decisions and affecting the module's operation. Therefore, timely detection, rapid diagnosis, and troubleshooting of sensor faults are of paramount importance.

[0067] In related technologies, an observer-based approach is used for sensor fault diagnosis. For example, an observer is constructed for each cell in a battery module. The input to each observer is the current sensor measurement and the corresponding cell's voltage sensor measurement. The output of each observer is the residual formed by the voltage estimate and the voltage sensor measurement for the corresponding cell. It is evident that the residual is influenced by both the current and voltage sensor measurements. Therefore, when the current sensor malfunctions, the residual generated by each observer will be affected. However, when the voltage sensor of a particular cell malfunctions, only the residual output by the observer corresponding to the faulty cell will be affected. By analyzing the combined response of the residuals, sensor fault diagnosis can be achieved.

[0068] like Figure 1 The diagram illustrates an architecture of a sensor fault diagnosis strategy in the prior art. Taking the fault diagnosis of a sensor in a series-connected battery pack as an example, assume that the series-connected battery pack includes four battery cells connected in series (Cell1 / Cell2 / Cell3 / Cell4). The sensor for this series-connected battery pack includes a current sensor (I) and voltage sensors (V1 / V2 / V3 / V4) corresponding to each battery cell (Cell1 / Cell2 / Cell3 / Cell4). The specific sensor fault diagnosis strategy involves constructing four observers, each corresponding to a battery cell. The input of each observer is the current sensor measurement value and the voltage sensor measurement value of the corresponding battery cell. The observer outputs the residual formed by the voltage estimate of the corresponding battery cell and the voltage sensor measurement value. The residual is affected by both the current sensor measurement value and the voltage sensor measurement value. Therefore, when the current sensor fails, the residuals generated by all four observers will be affected. However, when the voltage of a battery cell fails, only the residual output by the observer corresponding to the failed cell will be affected. In this way, fault diagnosis of the series-connected battery pack sensor can be achieved based on the combined response of the residuals.

[0069] However, when using the pure observer fault diagnosis method, it relies on establishing an accurate battery model, which results in a large amount of computation. Furthermore, an observer is built for each battery cell. If there are many battery cells, this will greatly increase the computational load, reduce computational efficiency, and also reduce the sensitivity of fault diagnosis to some extent.

[0070] Based on this, this application provides a sensor fault diagnosis method, apparatus, device, and storage medium, which will be described in detail below.

[0071] First, this application provides a sensor fault diagnosis method. The subject executing this method can be a sensor fault diagnosis device, or a server device, physical host, or user equipment (UE) that integrates the sensor fault diagnosis device. Specifically, the UE can be a terminal device such as a smartphone, tablet computer, laptop computer, handheld computer, or desktop computer.

[0072] This sensor fault diagnosis method can be used to detect faults in battery packs, specifically to diagnose faults in the sensors of a target battery pack. The sensors in the target battery pack can include current sensors and voltage sensors corresponding to each individual battery cell connected in series within the target battery pack.

[0073] Please see Figure 2 , Figure 2 This is a flowchart illustrating a sensor fault diagnosis method provided in an embodiment of this application. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that shown here.

[0074] In this embodiment of the application, the sensor fault diagnosis method may include the following steps:

[0075] Step S201: Obtain the current detection signal obtained by the current sensor sampling the total current of the target battery pack and the single cell voltage detection signal obtained by the voltage sensor sampling the single cell voltage of the corresponding battery cell.

[0076] In this embodiment, the target battery pack may include multiple battery cells connected in series. A current sensor may be connected in series in the battery cell circuit of the target battery pack to detect the total current in the circuit and obtain a current detection signal. Each battery cell may also have a voltage sensor connected in parallel, and each voltage sensor may be used to sample the cell voltage of the corresponding battery cell to obtain a cell voltage detection signal.

[0077] In order to diagnose faults in the current sensor and voltage sensors of the target battery pack, in this embodiment of the application, when the current sensor and voltage sensors detect the current and voltage of the target battery pack in real time, the detected current detection signal and the voltage detection signal of each individual cell are input to the sensor fault diagnosis device, so that the sensor fault diagnosis device can obtain the current detection signal obtained by the current sensor sampling the total current of the target battery pack and the individual cell voltage detection signal obtained by each voltage sensor sampling the individual cell voltage of the corresponding battery cell.

[0078] Understandably, the current detection signal and the individual cell voltage detection signal acquired in real time by the sensor fault diagnosis device can be the current detection value and the individual cell voltage detection value corresponding to each sampling moment.

[0079] Step S202: The sum of the voltage detection signals of each cell and the current detection signal are imported into a preset observer to obtain the estimated state of charge of the target battery pack.

[0080] In this embodiment of the application, before performing sensor fault diagnosis, an observer can first be constructed for the target battery pack. After the sum of the voltage detection signals of each cell and the current detection signal are imported into the observer, the observer can estimate the state of charge of the target battery pack based on the current detection signal (i.e., the current detection value) and the voltage detection signals of each cell (i.e., the cell voltage detection value), and obtain the estimated state of charge value. The residual is formed by the estimated state of charge value and the sum of the voltage detection values ​​of each cell.

[0081] Understandably, when a current sensor or any voltage sensor malfunctions, it will affect the acquisition of the current detection signal or the single-cell voltage detection signal, thereby affecting the estimation accuracy of the target battery pack's state of charge. Therefore, based on the estimated state of charge value, it can be determined whether the target battery pack has a sensor malfunction. If the estimated state of charge value deviates from the reference value / detection value, it can be determined that there is a current sensor malfunction or a voltage sensor malfunction.

[0082] Since the residual is constructed from the sum of the estimated state of charge and the detected voltage values ​​of each individual cell, sensor fault diagnosis of the target battery pack can also be achieved based on the combined response of the residuals. For example, if the residual exceeds a preset residual threshold, it can be determined that there is a fault in either the current sensor or the voltage sensor.

[0083] Step S203: If the state of charge estimate indicates that there is a sensor fault in the target battery pack, then calculate the correlation coefficient between the single cell voltage detection signals of every two adjacent battery cells in the target battery pack.

[0084] In this embodiment of the application, if it is determined from the state of charge estimate that the target battery pack has a sensor fault, i.e., a current sensor fault or a voltage sensor fault, the correlation coefficient between the individual cell voltage detection signals of every two adjacent battery cells can be calculated.

[0085] The correlation coefficient is a measure of the degree of linear correlation between variables. It reflects the closeness of the relationship between two variables. Therefore, the correlation coefficient measures whether the trends of two curves match, rather than their shapes. In an ideal working environment, the correlation coefficient between the voltages of two series-connected battery cells should be close to 1. When the voltage of either battery cell becomes abnormal, the correlation coefficient will decrease significantly, indicating a voltage anomaly. Based on the combined response to the voltage anomaly, the faulty sensor can be identified.

[0086] Step S204: Perform fault diagnosis on the current sensor and each voltage sensor based on the correlation coefficients.

[0087] As can be seen from step S203, when the voltage of any one of two adjacent battery cells is abnormal, the correlation coefficient between those two battery cells will significantly decrease, indicating that one of the battery cells has a voltage abnormality. This further suggests that the voltage sensor connected in parallel with that battery cell is faulty. Therefore, if the correlation coefficient between the voltage detection signals of any two adjacent battery cells is normal, it can be determined that the target battery pack does not have a voltage abnormality, thus confirming that the voltage sensors are not faulty. Since step S203 determines that the target battery pack has a sensor fault based on the state of charge estimation, if a sensor fault exists but the voltage sensors are not faulty, it can be determined that the current sensor is faulty.

[0088] Conversely, if there is an abnormal correlation coefficient, it can be understood that if the voltage sensor corresponding to a certain battery cell is faulty, then the two correlation coefficients associated with the single cell voltage detection signal of that battery cell will both be abnormal. Therefore, if there are two adjacent abnormal correlation coefficients, it can be determined that the voltage sensor corresponding to the battery cell associated with the two adjacent correlation coefficients is faulty. At the same time, the other voltage sensors and current sensors in the target battery pack are not faulty.

[0089] This enables fault diagnosis of current and voltage sensors in the target battery pack and quick identification of the faulty sensor.

[0090] In this embodiment, the observer estimates the state of charge (SOC) of the target battery pack based on the current detection signal from the current sensor and the single-cell voltage detection signal from the voltage sensor corresponding to each battery cell. Based on the SOC estimate, it determines whether a sensor fault exists in the target battery pack. When a sensor fault is determined, for each battery cell in the target battery pack, the correlation coefficient between the single-cell voltage detection signals of every two adjacent battery cells is calculated. The calculated correlation coefficient is then used to diagnose the faults in the current sensor and each voltage sensor. Compared to the prior art where an observer is built for each battery cell for sensor fault diagnosis, this embodiment significantly reduces the computational load by constructing an observer combined with correlation coefficients, thereby improving computational efficiency and fault diagnosis sensitivity, and ensuring the reliability of the battery management system.

[0091] In some embodiments of this application, the sum of the voltage detection signals of each individual cell and the current detection signal are imported into a preset observer to obtain an estimated value of the state of charge of the target battery pack. Prior to this, the method may further include:

[0092] Construct an equivalent model of the target battery pack based on its performance parameters;

[0093] The open-circuit voltage of the target battery pack is tested to obtain the open-circuit voltage of the target battery pack, and the relationship between the open-circuit voltage and the state of charge of the target battery pack is established.

[0094] The target battery pack was subjected to simulated dynamic stress test to obtain the voltage and current correspondence under the simulated dynamic stress test. Based on the voltage and current correspondence, the model parameters in the equivalent model were identified to determine the quantitative relationship between the model parameters and the state of charge.

[0095] An observer is established based on the relationship between open-circuit voltage and the state of charge of the target battery pack, as well as the quantitative relationship between model parameters and state of charge.

[0096] In this embodiment of the application, the target battery pack can be a lithium-ion battery pack composed of lithium-ion battery cells, and the performance parameters of the target battery pack can be the type and model of the lithium-ion battery and the corresponding technical parameters of the model battery, such as the available capacity of the battery.

[0097] Specifically, for example, if the target battery pack is a series battery pack consisting of four NCM811 square cells connected in series, in the embodiments of this application, the series battery pack can have various modeling forms, which can realize functions such as battery pack state estimation. Existing modeling methods are divided into the following categories: large single-cell model, multi-cell model, maximum-minimum model, and mean-deviation model, etc.

[0098] Without considering the inconsistencies within the series-connected battery pack, the series-connected battery pack can be regarded as a large battery cell with a large voltage and a large capacity, thus enabling the construction of a large cell model.

[0099] The voltage of this large battery cell is the sum of the voltages of its individual smaller battery cells, and the current is the current of the series-connected battery pack. Based on this, various cell modeling methods can be applied to model the series-connected battery pack.

[0100] Considering that sensor fault diagnosis for series-connected battery packs requires a comprehensive consideration of computational load and accuracy, this embodiment selects a first-order equivalent circuit model. For example... Figure 3 As shown, the equivalent model mainly includes an open-circuit voltage source U. oc Given an internal resistance R and an RC network including a first resistor R1 and a first capacitor C1, the following mathematical expression can be derived from Kirchhoff's laws:

[0101]

[0102] U t =Uoc -U1-I·R

[0103] In the formula, I represents the input current, with charging current being positive and discharging current being negative; u1 represents the voltage across the first capacitor C1, called the polarization voltage; U t This indicates the battery terminal voltage. The first capacitor, C1, is also known as the polarization capacitor.

[0104] The following two equations also hold true:

[0105]

[0106] U oc =spline(S soc )

[0107] Among them, U oc S represents the open-circuit voltage of a series-connected battery pack, which is a nonlinear function of the state of charge; soc The state of charge of the series-connected battery pack is represented by η, and the coulombic efficiency is represented by C. bat This indicates the available capacity of the series-connected battery pack.

[0108] Discretizing the above equation yields the following expression:

[0109] U1(k+1)=exp(-Δt / (R1C1))·U1(k)+R1·(1-exp(-Δt / (R1C1)))·I(k)

[0110] U t (k)=spline(S soc (k))-U1(k)-U2(k)-R·I(k)

[0111]

[0112] In the formula, k is the sampling time and Δt is the sampling interval.

[0113] After the equivalent model is constructed, trickle discharge tests can be performed on each cell in the target battery pack to obtain the open-circuit voltage of each cell. The open-circuit voltage of the target battery pack can then be obtained by summing the open-circuit voltages of each cell.

[0114] Specifically, in this embodiment, a trickle discharge test can be performed on the series-connected battery pack at room temperature to obtain the open circuit voltage (OCV) of each of the four individual cells. Adding the four OCVs together yields the equivalent open circuit voltage of the large cell, which is the open circuit voltage of the target battery pack. Based on the open circuit voltage of the target battery pack, the relationship curve between the open circuit voltage and the state of charge of the large cell can be obtained, and this curve can be fitted using a polynomial or other empirical formula.

[0115] In addition, a simulated dynamic stress test (DST) can be performed on the target battery pack to obtain the voltage and current correspondence under the simulated dynamic stress test. Based on the obtained voltage and current correspondence, the model parameters in the equivalent model can be identified to determine the quantitative relationship between the model parameters and the state of charge.

[0116] In this embodiment of the application, the method for identifying model parameters can be one or more of the following: recursive least squares with genetic factors, particle swarm optimization (PSO) algorithm, and genetic algorithm (GA). The specific method can be determined according to the actual application scenario.

[0117] Based on the aforementioned results, namely the relationship between open-circuit voltage and the state of charge of the target battery pack, as well as the quantitative relationship between model parameters and state of charge, an observer can be established and initialized.

[0118] In this embodiment, the observer can estimate the state of charge of the series battery pack based on one or more of the extended Kalman filter (EKF), state estimation algorithm and data-driven algorithm to generate residuals. Finally, the fault diagnosis and separation of the series battery pack sensor is realized based on the response of the residuals and the response of the correlation coefficient.

[0119] The main idea of ​​the observer-based fault diagnosis method is to reconstruct a system based on a more accurate mathematical model of the system, with the original system's input measurement value and output measurement value as the total input. This system will output an estimated value of the original system's output measurement value, and the difference between the estimated value and the measured value will generate a residual. When the sensor fails, the original system's input measurement value and output measurement value will change, and the residual will also change. The sensor fault diagnosis can be achieved by detecting the change in the residual.

[0120] Based on this, this application embodiment establishes an EKF observer. The specific calculation process of the EKF algorithm is as follows, for the nonlinear system with the following expression:

[0121] x k+1 =f(x) k ,u k )+w k

[0122] y k =g(x k ,u k )+v k

[0123] This expression represents the state equation and measurement equation for a discrete-time nonlinear time-varying system, where x k y represents the system state variable. k u represents the system's measured output quantity. k Indicates system input; w k For system state noise, v k For measuring noise, both are independent Gaussian noise with zero mean; f() is the state equation of the nonlinear system, and g() is the measurement equation.

[0124] The covariance of system state noise and measurement noise is represented by Q. k and R k The expression is:

[0125] Q k =E(w k w k T )

[0126] R k =E(v) k v k T )

[0127] Among them, w k Let E() represent the system state noise, and E() be the covariance equation.

[0128] The initial state of the system is determined as follows:

[0129]

[0130]

[0131] in, This represents the state estimate measurement update value at the initial time. This represents the updated value of the error covariance measurement at the initial time.

[0132] Set the initial covariance Q0 and R0 between the system's state noise and measurement noise.

[0133] Assumption

[0134] The specific calculation process of the Kalman filter algorithm based on this system is as follows.

[0135]

[0136] P k - =A k-1 P k-1 +A k-1 T +Q k-1

[0137] G k =P k - C k (C k P k - C k T +R k ) -1

[0138]

[0139] P k + =(IG) k C k )P k -

[0140] in, This represents the state estimate time update value at time k. This represents the state estimate measurement update value at time k. This represents the time-updated value of the error covariance at time k. G represents the updated value of the error covariance measurement at time k. k Let K represent the Kalman gain matrix at time k.

[0141] After constructing the EKF observer, the sum of the current sensor measurement data and the four voltage sensor measurement data can be imported into the EKF observer to perform state estimation and obtain the estimated state of charge (SOC) value of the series battery pack. The residual is obtained by subtracting the estimated value from the reference value. Without considering faults in the temperature sensor or the battery itself within the target battery pack, a fault in the current sensor or any voltage sensor will affect the acquisition of the current or voltage signal, thus affecting the accuracy of the SOC estimation for the series battery pack. Based on this, fault detection of the current or voltage sensor can be achieved by estimating the SOC of the battery pack. If the estimated SOC deviates from the reference value, it can be determined that there is a fault in either the current or voltage sensor. It should be noted that this embodiment does not consider the case of multiple sensors failing simultaneously; that is, in this embodiment, there is only one faulty sensor.

[0142] In related technologies, based on the state of charge estimation results of series battery packs based on extended Kalman filter observers, it is possible to diagnose faults of current sensors or voltage sensors. However, it can only determine that a fault has occurred, but cannot determine whether the fault is caused by the current sensor or the voltage sensor, nor can it determine which voltage sensor has failed.

[0143] Based on this, this application proposes a sensor fault diagnosis method based on the improved correlation coefficient method.

[0144] The correlation coefficient between two variables can be calculated using the following formula:

[0145]

[0146] In the formula, r x,y σ is the correlation coefficient between variables x and y; cov(x,y) is the covariance of variables x and y; x ,σ y Let μ be the standard deviation of variables x and y. x ,μ y Let x and y be the means of variables x and y, and n be the number of samples.

[0147] An important property of the correlation coefficient can be expressed by the following formula:

[0148] r αx+β,y =r x,y

[0149] In the formula, α and β are constants. This property can be derived from the definition of the correlation coefficient. From this property, we know that the correlation coefficient measures whether the trends of two curves match, rather than the shape of the curves. Therefore, ideally, during operation, the correlation coefficient of the voltages of two series-connected battery cells should be close to 1. When the voltage of either cell becomes abnormal, the correlation coefficient will decrease significantly, thus reflecting the voltage anomaly.

[0150] Therefore, in this embodiment, the correlation coefficient method is used to detect voltage sensor faults. In order to apply the correlation coefficient method online, a recursive form of the correlation coefficient calculation expression can be obtained:

[0151]

[0152]

[0153] The above formula can be simplified to:

[0154]

[0155] The recursive formula for the correlation coefficient above can be used to obtain the similarity of the time-domain trends of the voltage curves of two individual cells at the beginning of the measurement. However, as online operating condition data accumulates, the historical data has a high degree of similarity, and the calculation results of the correlation coefficient may fail to reflect voltage anomalies. At the same time, the continuous accumulation of historical data will put pressure on storage capacity. Therefore, based on this, the embodiments of this application have improved the correlation coefficient. Specifically, a preset sliding window is used to filter the individual cell voltage detection signals of every two adjacent battery cells, and the correlation coefficient between the two individual cell voltage detection signals within the sliding window is calculated.

[0156] In this embodiment, only the correlation coefficient of the data in the sliding window is calculated at each instant. The improved formula for calculating the correlation coefficient is as follows:

[0157]

[0158]

[0159] In the formula, N represents the size of the sliding window. It's worth noting that the window size can be determined based on the specific application scenario. If a large amount of data is used in the calculation, the impact of abnormal voltage changes on the correlation coefficient can be negligible. To maintain the sensitivity of fault detection, a small sliding window is preferred. On the other hand, when the sliding window size is too small, noise will be considered abnormal fluctuations, and measurement noise will also affect the calculation. Therefore, the appropriate sliding window size should be selected based on the application.

[0160] Suppose we introduce two additional signals X and Y into two random variables x and y, respectively. The correlation coefficient between the two variables can then be calculated as follows:

[0161]

[0162] Assuming that x, y are independent of X, Y, the above equation can be transformed into:

[0163]

[0164] In the above equation, when the battery is at rest, the first term in the numerator is zero because the voltage is very close to its open-circuit voltage. When X and Y are independent, identically distributed white noise, the second term is also zero. This indicates that in this case, the correlation coefficient is close to zero, leading to a sudden drop in the calculation, which could result in misdiagnosis of the fault.

[0165] If we further extend this to include two signals as variables, and simultaneously set the mean of the included signals to 0, then the correlation coefficient can be calculated as follows:

[0166]

[0167] In the formula, A and B are signals newly added to x and y, and are assumed to be independent of x, y, X, and Y. When A and B are not independent, the above formula provides a solution that avoids zero correlation coefficients. In this case, when the battery is in a state of zero current, the above formula simplifies to:

[0168]

[0169] The above formula shows that if the variance of the noise is negligible compared to A and B, then the correlation coefficient between signals x+X+A and y+Y+B is equal to the correlation coefficient between A and B. Using this property, the same signal can be added to two voltage signals, which means r A,B When the signal is continuously input, the additional signal can be ignored. At the same time, the variance of the two signals should be greater than the variance of the noise. In this way, when the battery current is 0, the correlation coefficient of the two voltages will be close to 1.

[0170] A simple design involves adding a square wave with an amplitude three times the noise standard deviation, or nine times the noise variance, as shown in the following expression:

[0171]

[0172] In the formula, S is an additive square wave. Therefore, when the battery current is 0, the correlation coefficient is close to 0.9. Obviously, when the amplitude of the square wave is larger, the correlation coefficient is closer to 1. However, the increase in amplitude also reduces the sensitivity of the detection to the actual voltage drop. Therefore, in the design of the additive square wave, considering a threshold of 0.5, 0.9 is a reasonable target. The period of the square wave should be smaller than the sliding window size, and the period can be selected as two samples.

[0173] The improved correlation coefficient method described above can be used to monitor the voltage sensor faults of four battery cells. Specifically, the correlation coefficient between the voltage signals measured by the voltage sensors of each pair of adjacent battery cells is calculated, including the correlation coefficient between the first and last cells. Figure 4 As shown, when the voltage sensor of a battery cell fails, both correlation coefficients associated with that battery cell will decrease. Therefore, the location of the fault can be isolated by the number of overlapping cells.

[0174] For example, when and When all readings show a sudden drop, it indicates that the voltage sensor on battery cell 2 is faulty. If there is no abnormal correlation coefficient, it can be assumed that all four voltage sensors are functioning correctly, but the current sensor is faulty.

[0175] like Figure 5As shown, by estimating the state of charge of a series battery pack based on an observer, it is possible to determine the presence of a sensor fault, but it is impossible to determine whether the fault is in the current sensor or the voltage sensor, let alone which voltage sensor has failed. By improving the correlation coefficient method, if a voltage sensor fault exists, it is possible to determine which individual voltage sensor has failed. Combining the two approaches results in the series battery pack sensor fault diagnosis strategy using the observer-integrated improved correlation coefficient method proposed in this application. Compared with the prior art, the method in this application significantly reduces the computational load and improves computational efficiency and fault diagnosis sensitivity, thus ensuring the reliability of the battery management system.

[0176] To better implement the sensor fault diagnosis method of this application, an embodiment of this application also provides a sensor fault diagnosis device. For example... Figure 6 As shown, Figure 6 This is a functional module diagram of the sensor fault diagnosis device provided in this application embodiment. The sensor fault diagnosis device 600 is used to diagnose faults in the current sensor in the target battery pack and the voltage sensor corresponding to each battery cell connected in series in the target battery pack. The current sensor is used to sample the total current of the target battery pack, and the voltage sensor is used to sample the individual voltage of the corresponding battery cell. The sensor fault diagnosis device 600 may include:

[0177] The acquisition module 601 is used to acquire the current detection signal obtained by the current sensor sampling the total current of the target battery pack and the single cell voltage detection signal obtained by the voltage sensor sampling the single cell voltage of the corresponding battery cell.

[0178] The state of charge estimation module 602 is used to import the sum of the voltage detection signals of each cell and the current detection signal into a preset observer to obtain the estimated state of charge of the target battery pack.

[0179] The correlation coefficient calculation module 603 is used to calculate the correlation coefficient between the single cell voltage detection signals of every two adjacent battery cells in the target battery pack if the state of charge estimate indicates that there is a sensor fault in the target battery pack.

[0180] The diagnostic module 604 is used to perform fault diagnosis on the current sensor and each voltage sensor based on the correlation coefficients.

[0181] In some embodiments of this application, the diagnostic module 604 may specifically be used for:

[0182] If there are no abnormalities in the correlation coefficients, then it is determined that the voltage sensors are not faulty and the current sensors are faulty.

[0183] If two adjacent correlation coefficients are abnormal, it is determined that the voltage sensor corresponding to the battery cell associated with the two adjacent correlation coefficients is faulty, while the current sensor and other voltage sensors in the target battery pack are not faulty.

[0184] In some embodiments of this application, the correlation coefficient calculation module 603 can specifically be used for:

[0185] The voltage detection signals of each pair of adjacent battery cells are filtered using a preset sliding window, and the correlation coefficient between the voltage detection signals of the two cells within the sliding window is calculated.

[0186] In some embodiments of this application, the sensor fault diagnosis device 600 further includes a construction module 605, which, before the state of charge estimation module 602 imports the sum of the voltage detection signals of each cell and the current detection signal into a preset observer to obtain the state of charge estimate of the target battery pack, can specifically be used for:

[0187] Construct an equivalent model of the target battery pack based on its performance parameters;

[0188] The open-circuit voltage of the target battery pack is tested to obtain the open-circuit voltage of the target battery pack, and the relationship between the open-circuit voltage and the state of charge of the target battery pack is established.

[0189] The target battery pack was subjected to simulated dynamic stress test to obtain the voltage and current correspondence under the simulated dynamic stress test. Based on the voltage and current correspondence, the model parameters in the equivalent model were identified to determine the quantitative relationship between the model parameters and the state of charge.

[0190] An observer is established based on the relationship between open-circuit voltage and the state of charge of the target battery pack, as well as the quantitative relationship between model parameters and state of charge.

[0191] In some embodiments of this application, the construction module 605 may also be used for:

[0192] A trickle discharge test was performed on each cell in the target battery pack to obtain the open-circuit voltage of each cell.

[0193] The open-circuit voltage of the target battery pack is obtained by summing the open-circuit voltages of each individual cell.

[0194] In some embodiments of this application, the construction module 605 may also be used for:

[0195] Based on the voltage-current correspondence, one or more algorithms, including recursive least squares with genetic factors, particle swarm optimization, and genetic algorithm, are used to identify the model parameters in the equivalent model.

[0196] It should be noted that in this application, the contents of the acquisition module 601, the state of charge estimation module 602, the correlation coefficient calculation module 603, the diagnosis module 604, and the construction module 605 correspond one-to-one with those described above. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the sensor fault diagnosis device and its corresponding unit modules described above can be referred to as follows: Figure 2 The description of the sensor fault diagnosis method corresponding to any embodiment will not be repeated here.

[0197] To better implement the sensor fault diagnosis method of this application, this application also provides a sensor fault diagnosis device, which may include a processor 701 and a memory 702. The memory 702 can be used to store a computer program, which, when executed by the processor 701, can be used to implement the following functions:

[0198] The current detection signal obtained by the current sensor sampling the total current of the target battery pack and the single cell voltage detection signal obtained by the voltage sensor sampling the single cell voltage of the corresponding battery cell are acquired.

[0199] The sum of the voltage detection signals of each individual cell and the current detection signal are imported into a preset observer to obtain the estimated state of charge of the target battery pack.

[0200] If the state of charge estimate indicates that there is a sensor fault in the target battery pack, then calculate the correlation coefficient between the single cell voltage detection signals of every two adjacent battery cells in the target battery pack.

[0201] Fault diagnosis is performed on the current sensor and each voltage sensor based on the correlation coefficients.

[0202] like Figure 7 As shown, it illustrates a structural schematic diagram of the sensor fault diagnosis device involved in this application, specifically:

[0203] The sensor fault diagnosis device may include components such as a processor 701 with one or more processing cores, a memory 702 with one or more computer-readable storage media, a power supply 703, and an input unit 704. Those skilled in the art will understand that... Figure 7 The structure shown does not constitute a limitation on the sensor fault diagnosis device. The sensor fault diagnosis device may also include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein:

[0204] The processor 701 is the control center of the device, connecting various parts of the device through various interfaces and lines. It executes software programs and / or unit modules stored in the memory 702, and calls data stored in the memory 702, to perform various functions and process data of the sensor fault diagnosis device, thereby providing overall monitoring of the interactive device. Optionally, the processor 701 may include one or more processing cores; the processor 701 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor. Preferably, the processor 701 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and application programs, and the modem processor mainly handles wireless communication. It is understood that the aforementioned modem processor may not be integrated into the processor 701.

[0205] The memory 702 can be used to store software programs and modules. The processor 701 executes various functional applications and data processing by running the software programs and modules stored in the memory 702. The memory 702 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function, etc.; the data storage area may store data created based on the use of the sensor fault diagnosis device, etc. In addition, the memory 702 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 702 may also include a memory controller to provide the processor 701 with access to the memory 702.

[0206] The sensor fault diagnosis device may also include a power supply 703 that supplies power to the various components. Preferably, the power supply 703 can be logically connected to the processor 701 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. The power supply 703 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0207] The sensor fault diagnosis device may also include an input unit 704 and an output unit 705. The input unit 704 can be used to receive input digital or character information, and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0208] Although not shown, the sensor fault diagnosis device may also include a display unit, etc., which will not be described in detail here. Specifically, in this application, the processor 701 in the interactive device loads the executable files corresponding to the processes of one or more application programs into the memory 702 according to the following instructions, and the processor 701 runs the application programs stored in the memory 702 to realize various functions, as follows:

[0209] The current detection signal obtained by the current sensor sampling the total current of the target battery pack and the single cell voltage detection signal obtained by the voltage sensor sampling the single cell voltage of the corresponding battery cell are acquired.

[0210] The sum of the voltage detection signals of each individual cell and the current detection signal are imported into a preset observer to obtain the estimated state of charge of the target battery pack.

[0211] If the state of charge estimate indicates that there is a sensor fault in the target battery pack, then calculate the correlation coefficient between the single cell voltage detection signals of every two adjacent battery cells in the target battery pack.

[0212] Fault diagnosis is performed on the current sensor and each voltage sensor based on the correlation coefficients.

[0213] Those skilled in the art will understand that all or part of the steps in the various methods described above can be accomplished by instructions, or by controlling related hardware with instructions. These instructions can be stored in a computer-readable storage medium and loaded and executed by the processor 701.

[0214] Therefore, this application provides a computer-readable storage medium, which may include: read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, etc. Computer instructions are stored thereon, and these computer instructions are loaded by processor 701 to execute the steps in any of the sensor fault diagnosis methods provided in this application. For example, when the computer instructions are executed by processor 701, they perform the following functions:

[0215] The current detection signal obtained by the current sensor sampling the total current of the target battery pack and the single cell voltage detection signal obtained by the voltage sensor sampling the single cell voltage of the corresponding battery cell are acquired.

[0216] The sum of the voltage detection signals of each individual cell and the current detection signal are imported into a preset observer to obtain the estimated state of charge of the target battery pack.

[0217] If the state of charge estimate indicates that there is a sensor fault in the target battery pack, then calculate the correlation coefficient between the single cell voltage detection signals of every two adjacent battery cells in the target battery pack.

[0218] Fault diagnosis is performed on the current sensor and each voltage sensor based on the correlation coefficients.

[0219] The computer instructions stored in the computer-readable storage medium can execute the present application as follows. Figure 2 Corresponding to the steps in the sensor fault diagnosis method in any embodiment, the present application can be implemented as described above. Figure 2 For details on the beneficial effects that the sensor fault diagnosis method can achieve in any embodiment, please refer to the preceding description, which will not be repeated here.

[0220] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the detailed descriptions of other embodiments above, which will not be repeated here.

[0221] In practice, each of the above units or structures can be implemented as an independent entity or can be arbitrarily combined to be implemented as the same or several entities. For specific implementation of each of the above units or structures, please refer to the previous embodiments, which will not be repeated here.

[0222] The above provides a detailed description of a sensor fault diagnosis method, apparatus, device, and storage medium provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The above description is only for the purpose of helping to understand the method and its core ideas. At the same time, those skilled in the art will recognize that there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A sensor fault diagnosis method, characterized in that, The method is used to diagnose faults in a current sensor in a target battery pack and a voltage sensor corresponding to each battery cell connected in series in the target battery pack, wherein the current sensor is used to sample the total current of the target battery pack, and the voltage sensor is used to sample the individual voltage of the corresponding battery cell; the method includes: The current detection signal obtained by the current sensor sampling the total current of the target battery pack and the single cell voltage detection signal obtained by each voltage sensor sampling the single cell voltage of the corresponding battery cell are acquired. The sum of the individual cell voltage detection signals and the current detection signal are imported into a preset observer to obtain the estimated state of charge of the target battery pack. If the state of charge estimate indicates that there is a sensor fault in the target battery pack, then calculate the correlation coefficient between the individual cell voltage detection signals of every two adjacent battery cells in the target battery pack. Fault diagnosis is performed on the current sensor and each of the voltage sensors based on the respective correlation coefficients. The fault diagnosis of the current sensor and each of the voltage sensors based on the correlation coefficients includes: If none of the correlation coefficients are abnormal, then it is determined that none of the voltage sensors are faulty and the current sensor is faulty. If two adjacent correlation coefficients are abnormal, it is determined that the voltage sensor corresponding to the battery cell associated with the two adjacent correlation coefficients is faulty, while the current sensor and other voltage sensors in the target battery pack are not faulty.

2. The method according to claim 1, characterized in that, The calculation of the correlation coefficient between the individual cell voltage detection signals of every two adjacent cells in the target battery pack includes: The voltage detection signals of each pair of adjacent battery cells are filtered using a preset sliding window, and the correlation coefficient between the voltage detection signals of the two cells within the sliding window is calculated.

3. The method according to claim 1, characterized in that, Before the step of inputting the sum of the individual cell voltage detection signals and the current detection signal into a preset observer to obtain the estimated state of charge of the target battery pack, the method further includes: Construct an equivalent model of the target battery pack based on its performance parameters; An open-circuit voltage test is performed on the target battery pack to obtain the open-circuit voltage of the target battery pack, and the relationship between the open-circuit voltage and the state of charge of the target battery pack is established. The target battery pack is subjected to simulated dynamic stress test to obtain the voltage and current correspondence under simulated dynamic stress test. Based on the voltage and current correspondence, the model parameters in the equivalent model are identified to determine the quantitative relationship between the model parameters and the state of charge. The observer is established based on the relationship between the open-circuit voltage and the state of charge of the target battery pack, as well as the quantitative relationship between the model parameters and the state of charge.

4. The method according to claim 3, characterized in that, The step of performing an open-circuit voltage test on the target battery pack to obtain the open-circuit voltage of the target battery pack includes: A trickle discharge test was performed on each individual cell in the target battery pack to obtain the open-circuit voltage of each individual cell. The open-circuit voltage of the target battery pack is obtained by summing the open-circuit voltages of each individual cell.

5. The method according to claim 3, characterized in that, The step of identifying the model parameters in the equivalent model based on the voltage-current correspondence includes: Based on the voltage-current correspondence, the model parameters in the equivalent model are identified using one or more algorithms, including recursive least squares with genetic factors, particle swarm optimization, and genetic algorithm.

6. The method according to any one of claims 1-5, characterized in that, The observer estimates the state of charge of the target battery pack based on one or more algorithms, including the extended Kalman filter algorithm, the state estimation algorithm, and the data-driven algorithm, to obtain the estimated state of charge of the target battery pack.

7. A sensor fault diagnosis device, characterized in that, The device is used for fault diagnosis of a current sensor in a target battery pack and a voltage sensor corresponding to each battery cell connected in series in the target battery pack. The current sensor is used to sample the total current of the target battery pack, and the voltage sensor is used to sample the individual voltage of the corresponding battery cell. The device includes: The acquisition module is used to acquire the current detection signal obtained by the current sensor sampling the total current of the target battery pack and the single cell voltage detection signal obtained by each voltage sensor sampling the single cell voltage of the corresponding battery cell. The state of charge estimation module is used to input the sum of the voltage detection signals of each of the individual cells and the current detection signal into a preset observer to obtain the estimated state of charge value of the target battery pack. The correlation coefficient calculation module is used to calculate the correlation coefficient between the individual cell voltage detection signals of every two adjacent battery cells in the target battery pack if the estimated state of charge indicates that there is a sensor fault in the target battery pack. A diagnostic module is used to perform fault diagnosis on the current sensor and each of the voltage sensors based on the respective correlation coefficients. The diagnostic module is specifically used for: If none of the correlation coefficients are abnormal, then it is determined that none of the voltage sensors are faulty and the current sensor is faulty. If two adjacent correlation coefficients are abnormal, it is determined that the voltage sensor corresponding to the battery cell associated with the two adjacent correlation coefficients is faulty, while the current sensor and other voltage sensors in the target battery pack are not faulty.

8. A sensor fault diagnosis device, characterized in that, The sensor fault diagnosis device includes a memory and a processor. The memory is used to store a computer program. When the computer program is executed by the processor, it is used to implement the sensor fault diagnosis method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that is executed by a processor to implement the steps of the sensor fault diagnosis method according to any one of claims 1 to 6.

Images