Battery pack state detection method, device, and storage medium

Abstract

This application discloses a battery pack state detection method, device, and storage medium, applied to a target charging device used to charge the battery pack. The battery pack state detection method includes: emitting an electromagnetic wave signal through a signal transmitter; acquiring target battery propagation data of the electromagnetic wave signal through a signal receiver; wherein the target battery propagation data includes signal parameters of several target signals, the target signals being electromagnetic wave signals propagated from individual battery cells inside the battery pack to the signal receiver, the propagation paths of different target signals are different, and the signal parameters of the target signals include at least one of the signal propagation time and signal amplitude; and determining the battery pack state data using the target battery propagation data. This method enables accurate automatic detection of the battery pack state data by a third-party charging device each time the battery pack is charged.

Description

Technical Field

[0001] This application relates to the field of equipment monitoring technology, and in particular to a method, device and storage medium for detecting the state of a battery pack. Background Technology

[0002] Taking new energy vehicles as an example, new energy vehicles are usually equipped with battery packs. Since the battery packs will continue to age during the user's use of the vehicle, it is necessary to test the status of the battery pack so that the user can understand the health status of the battery pack. Summary of the Invention

[0003] The main technical problem addressed by this application is to provide a method, device, and computer-readable storage medium for detecting the state of a battery pack, which can automatically and accurately detect the state data of the battery pack through a third-party charging device each time the battery pack is charged.

[0004] To address the aforementioned technical problems, this application provides a method for detecting the state of a battery pack, applied to a target charging device used to charge the battery pack. The method includes: transmitting an electromagnetic wave signal via a signal transmitter; acquiring target battery propagation data related to the electromagnetic wave signal via a signal receiver; wherein the target battery propagation data includes signal parameters of several target signals, the target signals being electromagnetic wave signals propagated from individual battery cells within the battery pack to the signal receiver, the propagation paths of different target signals being different, and the signal parameters of the target signals including at least one of the signal propagation time and signal amplitude; and determining the state data of the battery pack using the target battery propagation data.

[0005] In one embodiment, determining the state data of the battery pack using target battery propagation data includes: comparing the target battery propagation data with historical battery propagation data to obtain a deviation result of the target battery propagation data; and using the deviation result to determine the state data of the battery pack.

[0006] In one embodiment, comparing target battery propagation data with historical battery propagation data to obtain deviation results of the target battery propagation data includes: acquiring the overall deviation of the target battery propagation data relative to historical battery propagation data and at least one local deviation, wherein the overall deviation is obtained using the signal parameters of all target signals in the target battery propagation data and the signal parameters of all historical signals in the historical battery propagation data, and each local deviation is obtained using the signal parameters of different local abnormal signal sets, each local abnormal signal set including multiple target signals and multiple historical signals; using the deviation results to determine the state data of the battery pack includes: using the overall deviation and at least one local deviation to determine the state data of the battery pack.

[0007] In one embodiment, the battery pack status data includes a target health value of the battery pack. Determining the battery pack status data using the overall deviation and at least one local deviation includes: determining whether there is a target local deviation that meets the deviation conditions among the at least one local deviation; if there is no target local deviation, then determining the target health value using the overall deviation; if there is a target local deviation, then determining the target health value using the overall deviation and each target local deviation.

[0008] In one embodiment, when there is a target local deviation, the target health value is determined using the overall deviation and each target local deviation, including: determining a first health value using the overall deviation; determining a second health value using each target local deviation; and determining the target health value using the product of the first health value and each second health value.

[0009] In one embodiment, determining the state data of the battery pack using target battery propagation data includes: obtaining target vehicle attribute information of the vehicle where the battery pack is located; obtaining a target dataset corresponding to the target vehicle attribute information from a database, wherein the database includes several preset datasets corresponding to preset vehicle attribute information, the preset datasets include several preset battery propagation data and preset state data corresponding to each preset battery propagation data, and the target dataset is one of the preset datasets; obtaining the preset state data corresponding to the preset battery propagation data that matches the target battery propagation data in the target dataset, and using it as the state data of the battery pack.

[0010] In one embodiment, determining battery status data using target battery propagation data includes: determining multiple target extraction features corresponding to the target battery propagation data; for each target extraction feature, determining a first evaluation value of the target extraction feature using the feature value of the target extraction feature and the health benchmark value of the target extraction feature; combining the first evaluation values ​​of each target extraction feature to determine a second evaluation value; and using the second evaluation value to determine the battery pack status data.

[0011] In one embodiment, determining a first evaluation value of a target extracted feature using the feature value of the target extracted feature and the benchmark value of the target extracted feature includes: determining the deviation rate of the feature value of the target extracted feature relative to the benchmark value of the target extracted feature; and determining the product of the weight coefficient of the target extracted feature and the deviation rate of the target extracted feature as the first evaluation value of the target extracted feature.

[0012] And / or, by combining the first evaluation values ​​of the extracted features of each target, a second evaluation value is determined, including: taking the sum of the first evaluation values ​​as the second evaluation value.

[0013] And / or, the battery pack status data includes the target fault level of the battery pack. Using a second evaluation value, the battery pack status data is determined, including: determining the target interval to which the second evaluation value belongs; determining the fault level corresponding to the target interval from a preset mapping relationship as the target fault level, wherein the preset mapping relationship includes several preset intervals and preset fault levels corresponding to each preset interval.

[0014] And / or, the battery pack status data includes the target fault level of the battery pack, and the method further includes: determining the highest evaluation value among the first evaluation values; and using the target extracted features corresponding to the highest evaluation value and the target fault level to determine the target fault cause of the battery pack.

[0015] In one embodiment, acquiring target battery propagation data about electromagnetic wave signals through a signal receiving end includes: acquiring first propagation data about electromagnetic wave signals through the signal receiving end, wherein the first propagation data includes signal parameters of a plurality of first signals; dividing the first propagation data into a plurality of second propagation data with different propagation medium types; and identifying propagation data with a propagation medium type of battery cell from the plurality of second propagation data as target battery propagation data.

[0016] In one embodiment, for each second propagation data, the propagation medium type of the second propagation data is determined using at least one of the first result of the first method and the second result of the second method.

[0017] In the first approach: first extracted feature data of the second propagation data is determined, the first extracted feature data includes multiple first extracted features; multiple second sample extracted feature data similar to the first extracted feature data are determined from the first sample dataset, wherein the first sample dataset includes several first sample extracted feature data and sample type labels corresponding to each first sample extracted feature data, the sample type labels being battery cell or non-battery cell; at least the sample type labels of each second sample extracted feature data are taken as the first result.

[0018] In the second approach: determine the second extracted feature data of the second propagation data, the second extracted feature data includes multiple second extracted features; use a prediction model to predict the second extracted feature data to obtain a second result, the second result includes the prediction probability corresponding to the prediction type label of the second propagation data, the prediction type label is battery cell or non-battery cell.

[0019] In one embodiment, determining the propagation medium type of the second propagation data using the first result and the second result includes: using the first result to determine whether a first condition is met; wherein the first condition is any of the following: the proportion of samples labeled as battery cells in the first result is greater than or equal to a proportion threshold; the number of samples labeled as battery cells in the first result is greater than or equal to a number threshold; a first weight is greater than a second weight, the first weight is obtained by using the similarity value between the extracted feature data of each second sample labeled as battery cells and the first extracted feature data, and the second weight is obtained by using the similarity value between the extracted feature data of each second sample labeled as non-battery cells and the first extracted feature data; using the second result to determine whether a second condition is met; wherein the second condition includes: the predicted type label is battery cells and the predicted probability is greater than or equal to a first probability threshold; in response to meeting the first condition and the second condition, determining that the propagation medium type of the second propagation data is battery cells.

[0020] In one embodiment, the electromagnetic wave signal carries first identification information. Acquiring target battery propagation data of the electromagnetic wave signal via a signal receiver includes: acquiring initial propagation data of the electromagnetic wave signal via the signal receiver, wherein the initial propagation data includes signal parameters of several initial signals; verifying whether the second identification information carried in each initial signal is consistent with the first identification information; removing the signal parameters of initial signals whose second identification information is inconsistent with the first identification information from the initial propagation data to obtain first propagation data; and determining the target battery propagation data from the first propagation data. And / or, the battery pack status data includes the target health value of the battery pack, and the method further includes: using multiple historical health values ​​of the battery pack and the historical time corresponding to each historical health value to fit a health prediction curve; using the health prediction curve to determine the time information corresponding to the preset health value.

[0021] To solve the above-mentioned technical problems, another technical solution adopted in this application is: to provide an electronic device, including a memory and a processor coupled to each other, wherein the memory stores program instructions; and the processor is used to execute the program instructions stored in the memory to implement the above-mentioned battery pack state detection method.

[0022] To solve the above-mentioned technical problems, another technical solution adopted in this application is to provide a computer-readable storage medium for storing program instructions that can be executed by a processor to implement the above-mentioned battery pack state detection method.

[0023] In the above scheme, when the battery pack is charged using a target charging device, the target charging device can emit electromagnetic wave signals through a signal transmitter and acquire target battery propagation data of the electromagnetic wave signals through a signal receiver. The target battery propagation data includes signal parameters of several target signals, including at least one of the signal propagation time and signal amplitude. Since each target signal is an electromagnetic wave signal that propagates through different battery cells within the battery pack to the signal receiver, the signal propagation time and amplitude of the target signals will change accordingly when the battery pack ages. Therefore, the target battery propagation data can be used to accurately detect the battery pack's status data. This method allows for the automatic and accurate detection of the battery pack's status data by a third-party charging device each time the battery pack is charged. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the structure of an embodiment of the charging system provided in this application; Figure 2 This is a schematic diagram of the internal structure of the battery pack provided in this application; Figure 3 This is a flowchart illustrating an embodiment of the battery pack state detection method provided in this application; Figure 4 This is a schematic diagram of the propagation path of a target signal in a single battery cell provided in this application; Figure 5 This is a schematic diagram of the propagation path of a single battery cell for another target signal provided in this application; Figure 6 This is a flowchart illustrating an embodiment of the method for acquiring target battery propagation data provided in this application; Figure 7 This is a schematic diagram of the first propagation data provided in this application; Figure 8 This is a schematic diagram of the propagation data obtained by squaring the signal amplitudes of each first signal in the first propagation data multiple times, as provided in this application. Figure 9 This is a schematic diagram of the propagation data obtained by squaring the propagation time of each signal in the Type III propagation data multiple times, as provided in this application. Figure 10 This is a flowchart illustrating an embodiment of the method for determining the status data of a battery pack provided in this application; Figure 11 This is a schematic diagram of lithium dendrites appearing in a single battery cell provided in this application; Figure 12 This is a flowchart illustrating another embodiment of the method for determining the status data of a battery pack provided in this application; Figure 13This is a flowchart illustrating yet another embodiment of the method for determining the state data of a battery pack provided in this application; Figure 14 This is a flowchart illustrating yet another embodiment of the method for determining the state data of a battery pack provided in this application; Figure 15 This is a trend graph showing the change in the health value of the battery pack provided in this application over time; Figure 16 This is a schematic diagram of the framework of an embodiment of the battery pack state detection device provided in this application; Figure 17 This is a schematic diagram of the framework of an embodiment of the electronic device provided in this application; Figure 18 This is a schematic diagram of a framework of an embodiment of the computer-readable storage medium provided in this application. Detailed Implementation

[0025] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0026] It should be noted that the term "and / or" in this application is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The terms "first" and "second" in this application 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. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. The term "multiple" in this application means at least two, such as two, three, etc. The term "several" in this application means at least two. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0027] Please see Figure 1 , Figure 1 This is a schematic diagram of the structure of an embodiment of the charging system provided in this application. Figure 1 As shown, the charging system includes a battery pack and a charging device. The charging device is used to charge the battery pack and also to detect the status of the battery pack.

[0028] The charging equipment includes a charging pile, a DC charging gun, an RF transmitting unit, an RF receiving unit, a detection unit, and a data processing unit.

[0029] The battery pack is a lithium battery pack. Figure 2 This is a schematic diagram of the internal structure of the battery pack provided in this application, such as... Figure 2 As shown, the battery pack contains several battery cells, which are connected in series or in parallel. The battery pack includes a total positive terminal at one end and a total negative terminal at the other end.

[0030] The DC charging gun includes a positive DC power supply terminal (DC+) and a negative DC power supply terminal (DC-). When the DC charging gun is connected to the vehicle, the positive DC power supply terminal of the DC charging gun is connected to the total positive terminal of the battery pack, and the negative DC power supply terminal of the DC charging gun is connected to the total negative terminal of the battery pack.

[0031] The radio frequency (RF) transmitter unit is located at the positive terminal of the DC power supply of the DC charging gun and is used to transmit electromagnetic wave signals (or beacon signals). The signal transmitting end of the RF transmitter unit is the signal transmitting end of the charging device.

[0032] The radio frequency (RF) receiving unit is located at the negative terminal of the DC power supply of the DC charging gun. It is used to receive several initial signals of electromagnetic wave signals propagating through different propagation paths, and to analyze these initial signals. The signal receiving end of the RF receiving unit is the signal receiving end of the charging device.

[0033] The detection unit is located at the negative terminal of the DC power supply of the DC charging gun. The detection unit is connected to the RF receiving unit and is used to detect several initial signals parsed by the RF receiving unit to obtain initial propagation data. The initial propagation data includes signal parameters for each initial signal. Signal parameters include at least one of signal propagation time and signal amplitude. The signal propagation time refers to the time it takes for the signal to travel from the RF transmitting unit to the RF receiving unit. The signal propagation time reflects the propagation time of the electromagnetic wave signal in the propagation path and is directly related to the propagation path length and the characteristics of the propagation medium. The signal amplitude reflects the energy loss of the electromagnetic wave signal in the propagation path and is directly related to the conductivity and dielectric constant of the propagation medium.

[0034] The data processing unit is communicatively connected to the RF transmitting unit, RF receiving unit, and detection unit to coordinate the transmission, reception, and detection of electromagnetic wave signals. Since some of the initial signals obtained may be useless for battery pack status detection, the data processing unit also determines the target battery propagation data from the initial propagation data output by the detection unit, and then uses the target battery propagation data to perform battery pack status detection. The target battery propagation data includes signal parameters of several target signals. These target signals are electromagnetic wave signals propagated through the individual battery cells inside the battery pack to the signal receiving end; the propagation paths of different target signals are different for each battery cell.

[0035] The data processing unit is also used to store the target battery propagation data and the corresponding battery pack status data obtained from each charging of the battery pack in a local storage module (such as a FLASH memory, memory card, hard disk, etc.) or upload it to a cloud server. That is, the data processing unit can communicate with the cloud server to exchange data. Figure 1 The storage module and cloud server are not shown in the diagram. The data processing unit can also communicate with the vehicle controller of the vehicle where the battery pack is located to exchange data. Figure 1 The vehicle controller is not shown. The data processing unit may be a controller installed inside the DC charging gun or the charging pile. Figure 1 The example shown is based solely on the configuration of a data processing unit within a DC charging gun.

[0036] For example, the data processing unit is an MCU (Microcontroller Unit), an FPGA (Field-Programmable Gate Array), etc.

[0037] The following explains the principle of the charging device emitting electromagnetic wave signals to detect the state of the battery in this application.

[0038] During vehicle use, battery packs gradually age, leading to lithium deposition within the pack. This deposition process alters the distance between the positive and negative electrodes and changes the impedance characteristics. The deeper the deposition, the smaller the distance between the electrodes, resulting in differences in propagation time and amplitude between electromagnetic signals from an aged battery pack and those from a healthy one. Lithium dendrites, a type of lithium deposition, are needle-like or dendritic metallic lithium deposits formed during lithium ion reduction during charging. Dendrite growth is a fundamental factor affecting the safety and stability of lithium-ion battery packs: it causes instability at the electrode-electrolyte interface during cycling, damaging the solid electrolyte interphase (SEI) film; it continuously consumes electrolyte and leads to irreversible lithium deposition, resulting in dead lithium and low coulombic efficiency; and its sharp crystals can even pierce the separator, causing internal short circuits, thermal runaway, combustion, and explosion, ultimately leading to the ignition of the entire vehicle. The formation of lithium dendrites significantly reduces the distance between the positive and negative electrodes of the battery, resulting in a greater difference in the propagation time and amplitude of electromagnetic wave signals from a battery pack with lithium dendrites compared to those from a battery pack without lithium dendrites.

[0039] The core idea of ​​this application is to indirectly detect the degree of lithium plating and lithium dendrite formation inside the battery pack by transmitting electromagnetic wave signals and obtaining the target battery propagation data obtained by the electromagnetic wave signals propagating through the propagation path of the battery cells inside the battery pack to the signal receiving end, thereby detecting the state of the battery pack.

[0040] It should be noted that the electromagnetic wave signal in this application is implemented using UWB (Ultra Wide Band) technology. UWB technology is a wireless carrier communication technology that does not use a sinusoidal carrier but instead uses nanosecond-level non-sinusoidal narrow pulses to transmit data, thus occupying a wide frequency spectrum. The electromagnetic wave signal transmitted using UWB technology is a single-cycle pulse with an extremely short duration and a very small duty cycle. It has strong multipath resolution capability, and multipath signals (i.e., signals obtained by electromagnetic wave signals propagating through different propagation paths) can be separated in time.

[0041] Please see Figure 3 , Figure 3 This is a flowchart illustrating an embodiment of the battery pack state detection method provided in this application. The method is applied to a target charging device, which is the charging device used for the current charging of the battery pack. It should be noted that the battery pack state detection method in this embodiment can be executed during the process of the target charging device charging the battery pack, or it can be executed when the target charging device is connected to the vehicle but has not yet started charging the battery pack.

[0042] like Figure 3As shown, the method includes the following steps: S31: Transmits electromagnetic wave signals through the signal transmitting end.

[0043] S32: Obtain target battery propagation data about electromagnetic wave signals through the signal receiving end.

[0044] The target battery propagation data includes signal parameters of several target signals. The target signals are electromagnetic wave signals propagated from the individual battery cells within the battery pack to the signal receiver. The signal parameters of the target signals include at least one of the signal propagation time and signal amplitude. For example, the signal parameters may only include the signal propagation time. Another example is that the signal parameters may only include the signal amplitude. Yet another example is that the signal parameters may include both the signal propagation time and the signal amplitude.

[0045] Because a battery pack contains several individual battery cells connected in series or parallel, electromagnetic waves propagate from the signal transmitter to the battery pack and reach the signal receiver through different propagation paths within the battery cells. These individual cell propagation paths can be understood as the paths taken by the electromagnetic wave signal from the overall positive terminal to the overall negative terminal of the battery pack, traversing multiple battery cells. Figure 4 and Figure 5 As shown, Figure 4 This is a schematic diagram of the propagation path of a target signal in a single battery cell, as provided in this application. Figure 5 This is a schematic diagram of the propagation path of a single battery cell for another target signal provided in this application.

[0046] Because the propagation paths of different target signals from individual battery cells are different, the propagation times of each target signal are also different. For example... Figure 2 , Figure 4 or Figure 5 The internal structure of the battery pack shown will contain 2048 By analyzing the propagation paths of individual battery cells, 2048 target signals can be obtained. The propagation times of these 2048 target signals can be denoted as T1, T2, ..., T. 2048 Furthermore, the signal amplitudes of these 2048 target signals can be denoted as A1, A2, ... A 2048 .

[0047] S33: Use the data transmitted from the target battery to determine the status data of the battery pack.

[0048] In some implementations, the battery pack status data may include data such as the battery pack's target health value and target fault level.

[0049] The target health value reflects the health level of the battery pack; a higher target health value indicates a healthier battery pack, while a lower target health value indicates a less healthy battery pack. For example, the target health value can be a health score or health level of the battery pack.

[0050] The target fault level is one of several preset fault levels. The lower the target fault level, the healthier the battery pack is, and the higher the target fault level, the less healthy the battery pack is.

[0051] In this embodiment, when the battery pack is charged using a target charging device, the target charging device can emit electromagnetic wave signals through a signal transmitter and acquire target battery propagation data of the electromagnetic wave signals through a signal receiver. The target battery propagation data includes signal parameters of several target signals, including at least one of the signal propagation time and signal amplitude. Since each target signal is an electromagnetic wave signal that propagates through different battery cells within the battery pack to the signal receiver, the signal propagation time and signal amplitude of the target signals will change accordingly when the battery pack ages. Therefore, the target battery propagation data can be used to accurately detect the battery pack's status data. This method allows for the automatic and accurate detection of the battery pack's status data by a third-party charging device each time the battery pack is charged.

[0052] Please see Figure 6 , Figure 6 This is a flowchart illustrating an embodiment of the method for acquiring target battery propagation data provided in this application, which is applied to a target charging device. Figure 6 As shown, the method includes the following steps: S61: Obtain initial propagation data of electromagnetic wave signals through the signal receiving end.

[0053] The target charging device can first obtain the first identification information, modulate the first identification information into an electromagnetic wave signal, and then transmit the electromagnetic wave signal carrying the first identification information through the signal transmitting end.

[0054] In some examples, the first identification information is the identification information of the battery pack, such as the ID of the battery pack, or the first identification information may also be the identification information of the vehicle in which the battery pack is located, such as the ID of the vehicle in which the battery pack is located.

[0055] In one example, the target charging device can communicate with the vehicle containing the battery pack via CAN (Controller Area Network) to obtain first identification information.

[0056] Furthermore, the target charging device acquires initial propagation data of the electromagnetic wave signal through the signal receiver. The initial propagation data includes signal parameters of several initial signals.

[0057] S62: Verify whether the second identification information carried in each initial signal is consistent with the first identification information carried in the electromagnetic wave signal.

[0058] After acquiring the initial propagation data, the target charging device further demodulates the second identification information carried in each initial signal and compares it with the first identification information to verify whether they match. It is understandable that in practical applications, different charging devices may simultaneously charge the corresponding battery packs. If the second identification information carried by the initial signal does not match the first identification information, it indicates that the initial signal may be generated by electromagnetic wave signals emitted by other nearby charging devices; that is, the initial signal is an interference signal and needs to be eliminated. If the second identification information carried by the initial signal matches the first identification information, it indicates that the initial signal is a valid signal.

[0059] S63: Remove the signal parameters of the initial signal whose second identification information is inconsistent with the first identification information from the initial propagation data to obtain the first propagation data.

[0060] The first propagation data includes signal parameters of several first signals.

[0061] In steps S61 to S63, by removing the signal parameters of the initial signal whose second identification information is inconsistent with the first identification information from the initial propagation data, electromagnetic interference from other charging devices can be avoided, which can further improve the accuracy of subsequent battery pack status detection.

[0062] After obtaining the first propagation data, steps S64 and S65 need to be further executed to determine the target battery propagation data from the first propagation data.

[0063] S64: Divide the first propagation data into multiple second propagation data with different propagation medium types.

[0064] Understandably, each individual battery cell within a battery pack is typically connected in parallel with a resistor for voltage sampling. Therefore, the electromagnetic wave signal emitted by the target charging device's transmitter will propagate to the receiver through this resistor within the battery pack. Since the vehicle interior contains a significant amount of metal, the electromagnetic wave signal emitted by the target charging device will also propagate to the receiver through reflection and refraction by this metal. Furthermore, the electromagnetic wave signal emitted by the target charging device's transmitter will also propagate to the receiver through the air. Therefore, the propagation media for electromagnetic waves include four types: individual battery cells, battery resistors, metal, and air.

[0065] However, in this application, the state detection of the battery pack only requires data whose propagation medium type is a single battery cell. This can also be understood as the state detection of the battery pack only requiring data whose propagation path is the propagation path of a single battery cell, i.e., only the aforementioned target battery propagation data. Considering that the signal parameters of signals with different propagation medium types will differ, and that signal parameters of the same propagation medium type have similar patterns, step S64 first divides the acquired first propagation data into multiple second propagation data with different propagation medium types. Subsequently, the required target battery propagation data can be identified from these multiple second propagation data.

[0066] In one embodiment, the signal parameters of each first signal in the first propagation data are squared at least once to divide the first propagation data into multiple second propagation data of different propagation medium types. These multiple second propagation data include the aforementioned second propagation data of a single battery cell, second propagation data of a battery resistor, second propagation data of a metal, and second propagation data of control. Each second propagation data includes signal parameters of several second signals of the same propagation medium type.

[0067] In this embodiment, the signal propagation time of each first signal in the first propagation data may be squared at least once; or, the signal amplitude of each first signal in the first propagation data may be squared at least once; or, the signal propagation time of each first signal in the first propagation data may be squared at least once, and the signal amplitude of each first signal in the first propagation data may also be squared at least once. The specific number of squared operations is based on the second propagation data that can distinguish the aforementioned four types of propagation media in actual operation. In this embodiment, the specific number of squared operations is not specifically limited.

[0068] It is understandable that squaring can differentiate propagation data from different propagation media types, and therefore squaring can be used to divide the data. It should be noted that this embodiment only uses squaring as an example; in other embodiments, other methods can be used to divide similar data to obtain multiple second propagation data sets.

[0069] Figure 7 This is a schematic diagram of the first propagation data provided in this application. For example... Figure 7 As shown, the signal parameters of each first signal in the first propagation data are mostly similar, making it difficult to directly distinguish the target battery propagation data.

[0070] Figure 8 This is a schematic diagram of the propagation data obtained by squaring the signal amplitudes of each first signal in the first propagation data multiple times, as provided in this application. For example... Figure 8As shown, after squaring the signal amplitudes of each first signal in the first propagation data multiple times, it is divided into three categories: I, II, and III. However, this does not complete the classification of the aforementioned four categories of propagation data. Furthermore, since there are an unusually large number of propagation data in category III, the propagation data in category III can be squared multiple times further.

[0071] Figure 9 This is a schematic diagram of the propagation data obtained by squaring the propagation time of each signal in the Type III propagation data multiple times, as provided in this application. Figure 8 After processing, the Class III data was divided into Figure 9 The data includes Category III and Category IV propagation data. This completes the classification of the second propagation data into four categories.

[0072] Figure 7 , Figure 8 and Figure 9 In this context, A represents the signal amplitude and T represents the signal propagation time.

[0073] S65: Identify propagation data from multiple second propagation data where the propagation medium type is a battery cell, and use it as the target battery propagation data.

[0074] After dividing the first propagation data into multiple second propagation data with different propagation medium types, since the propagation medium type of each second propagation data is unknown at this time, it is necessary to further identify the propagation data with the propagation medium type of battery cell from the multiple second propagation data as the target battery propagation data.

[0075] In step S65, for each piece of second propagation data, the propagation medium type of the second propagation data is determined using at least one of the first result of the first method and the second result of the second method. The first result is obtained by processing the second propagation data using the first method, and the second result is obtained by processing the second propagation data using the second method. The second propagation data includes signal parameters of several second signals.

[0076] In one embodiment, the second result is obtained using the first method through the KNN (K-Nearest Neighbor) algorithm. This embodiment further includes the following steps: Step 1: Determine the first extracted feature data of the second propagation data.

[0077] Before determining the first extracted feature data of the second propagation data, the signal parameters of several second signals in the second propagation data can be denoised, such as by wavelet transform. For example, the signal parameters include signal propagation time and signal amplitude, and wavelet transform can be performed on both the signal propagation time and the signal amplitude of several second signals.

[0078] When the target charging device charges the battery pack, fluctuations in the charging current introduce high-frequency noise (such as current ripple and electromagnetic interference), which affects the propagation time and amplitude of each second signal in the acquired second propagation data. This causes jitter in the propagation time of each second signal; for example, a second signal with a propagation time of 100ns may become 96ns or 102ns due to noise. Noise also superimposes on the amplitude of each second signal, causing glitches; for example, a second signal with an amplitude of 0.5V may become 0.46V or 0.52V due to noise. Therefore, it is necessary to denoise the signal parameters of several second signals in the second propagation data to restore their true signal parameters.

[0079] In step one, the first extracted feature data of the second propagation data includes multiple first extracted features extracted using the second propagation data.

[0080] In one example, several first extracted features include: the maximum signal amplitude of several second signals in the second propagation data, the minimum signal amplitude of several second signals, the mean signal amplitude of several second signals, the standard deviation of the signal amplitude of several second signals, the mean signal amplitude attenuation rate of several second signals, the mean signal propagation time of several second signals, the standard deviation of the signal propagation time of several second signals, the median of the signal propagation time of several second signals, the Pearson correlation coefficient corresponding to the signal propagation time and signal amplitude of several second signals, the 90th quantile of the signal propagation time difference corresponding to several second signals, the number of effective signals in several second signals, the rate of change of the number of effective signals in several second signals, the distribution entropy of the effective signal propagation time difference corresponding to several second signals, the ratio of the total voltage of the battery pack to the mean signal propagation time of several second signals, the ratio of the maximum internal resistance of the battery cell in the battery pack to the peak signal amplitude of several second signals, and the signal propagation time and signal amplitude of several second signals in the second propagation data, etc.

[0081] The average signal amplitude attenuation rate of several second signals is the average of the signal amplitude attenuation rates of each second signal. The signal amplitude attenuation rate measures the degree of signal energy loss during propagation; a higher attenuation rate indicates greater signal energy loss. Furthermore, the signal amplitude attenuation rate reflects the attenuation effect of the propagation medium. For example, the signal amplitude attenuation rate of a propagation path using a battery cell as the propagation medium is much higher than that of a propagation path using air as the propagation medium. For each second signal, the difference between the second signal's amplitude and the ideal signal amplitude is determined. Dividing this difference by the ideal signal amplitude yields the signal amplitude attenuation rate of the second signal. The ideal signal amplitude refers to the amplitude of an electromagnetic wave signal propagating in a vacuum / air (i.e., without medium loss), affected only by the propagation distance.

[0082] The Pearson correlation coefficient, corresponding to the propagation time and amplitude of several second signals, is used to reflect the degree of linear correlation between the propagation time and amplitude of the second signals. The calculation of the Pearson correlation coefficient can be based on known techniques and will not be elaborated upon here.

[0083] The steps for determining the 90th quantile of the signal propagation time difference corresponding to several second signals include: for each second signal, determining the difference between the signal propagation time of the second signal and the reference propagation time, as the signal propagation time difference of the second signal; sorting the signal propagation time differences of the second signals in ascending order, and taking the value at the 90th position (i.e., 90% of the signal propagation time differences are less than or equal to this value) as the 90th quantile of the signal propagation time difference corresponding to several second signals. The reference propagation time can be the minimum signal propagation time among the signal propagation times of the several second signals.

[0084] A number of interference signals exist among several second signals. The number of effective signals among these second signals refers to the number of effective signals remaining after removing interference signals. An effective signal is defined as a signal whose amplitude is greater than or equal to the minimum detectable threshold and whose propagation time is greater than or equal to the multipath time-resolved threshold. The minimum detectable threshold and the multipath time-resolved threshold are empirical values. It is understandable that in a healthy battery pack with a regular internal structure (neatly arranged battery cells), the number of effective signals is relatively stable. In a battery pack with lithium dendrites, the internal structure is disordered (separator puncture, uneven electrolyte distribution), and the number of effective signals will increase significantly.

[0085] The rate of change (denoted by ΔNeff) of the number of valid signals among several second signals reflects the trend of change in the number of valid signals. The steps for determining the rate of change of the number of valid signals among several second signals include: determining the difference between the number of valid signals among several second signals and the number of reference signals; dividing the difference between the number of valid signals among several second signals and the number of reference signals by the number of reference signals to obtain the rate of change of the number of valid signals among several second signals. The ΔNeff of a healthy battery pack is typically stable within ±5%, while the ΔNeff of a faulty battery pack (e.g., with lithium dendrites) increases with the number of charging cycles (e.g., ΔNeff = 10% at the 20th charge and 30% at the 200th charge).

[0086] The distribution entropy of the effective signal propagation time difference among several second signals reflects the uniformity of the distribution of the signal propagation time difference (denoted by Δti) of each effective signal among several second signals. The propagation time differences of each signal in a healthy battery pack are uniformly distributed within the time interval, corresponding to a higher entropy value, such as Hcount≈3.0 bit; the propagation time differences of each signal in a faulty battery pack are concentrated in a few intervals, corresponding to a lower entropy value, such as Hcount≈1.5 bit. The steps for determining the distribution entropy of the effective signal propagation time difference among several second signals include: for each effective signal among several second signals, determining the difference between the effective signal's propagation time and the reference propagation time, as the effective signal's signal propagation time difference; dividing the effective signal into multiple intervals based on the minimum and maximum time differences among the effective signal propagation time differences; counting the number of signal propagation time differences contained in each interval; for each interval, determining the ratio of the number of signal propagation time differences contained in the interval to the total number of signal propagation time differences, as the interval probability; substituting the interval probabilities into the Shannon entropy formula to calculate the distribution entropy of the effective signal propagation time difference. The reference propagation time can be the minimum propagation time among several second signals. The Shannon entropy formula can be found using known techniques and will not be explained further here.

[0087] Furthermore, after determining the first extracted feature data of the second propagation data, the first extracted feature data can be normalized to map multiple first extracted features of the first extracted feature data to the [0,1] interval, thereby eliminating magnitude differences.

[0088] Step 2: Identify multiple second samples from the first sample dataset that are similar to the first extracted feature data and extract feature data.

[0089] The first sample dataset includes several first sample extracted feature data and corresponding sample type labels for each first sample extracted feature data. The sample type labels are either battery cell or non-battery cell. Each first sample extracted feature data in the first sample dataset includes multiple first sample extracted features. The acquisition method of the first sample extracted feature data is the same as that of the first extracted feature data in the second propagation data, and will not be repeated here. The sample propagation data corresponding to the first sample extracted feature data undergoes the same denoising processing as the second propagation data, and the first sample extracted feature data undergoes the same normalization processing.

[0090] The process of identifying multiple first-sample extracted feature data that are similar to the first extracted feature data from the first sample dataset further includes the following sub-steps: Sub-step one: For each first sample in the first sample dataset, extract feature data and determine the similarity value between the extracted feature data of the first sample and the first extracted feature data.

[0091] In some examples, the standard Euclidean distance or weighted Euclidean distance between multiple first-sample extracted features of the first-sample extracted feature data and multiple first-extracted features of the first-extracted feature data is used as the similarity value between the first-sample extracted feature data and the first-extracted feature data.

[0092] When using standard Euclidean distance, the extracted features across all dimensions are not differentiated by design weights; for example, all weights are set to 1. Standard Euclidean distance can be calculated using the following formula:

[0093] In the above formula, The standard Euclidean distance between multiple first-sample extracted features and multiple first-extracted features of the first-sample extracted feature data is represented. This indicates that the i-th sample in the first sample has extracted features. This represents the i-th first extracted feature. and Extracted features corresponding to the same dimension, such as the signal propagation time difference distribution entropy.

[0094] When using weighted Euclidean distance, the weights of the extracted features in each dimension need to be designed according to actual requirements. For example, the weights of important extracted features (such as the signal propagation time difference distribution entropy) are designed to be higher, while the weights of unimportant extracted features are designed to be lower. The weighted Euclidean distance can be calculated using the following formula:

[0095] In the above formula, The weighted Euclidean distance between multiple first-sample extracted features and multiple first-extracted features of the first-sample extracted feature data is represented. This indicates that the i-th sample in the first sample has extracted features. This represents the i-th first extracted feature; This represents the weighted weight corresponding to the extracted feature of the i-th dimension.

[0096] Sub-step two: Select multiple first samples from the first sample dataset that meet the similarity requirements and extract feature data from them, and use them as multiple second samples to extract feature data.

[0097] In one example, the number of feature data extracted from multiple second samples, i.e., the nearest neighbor number K, is determined through a network search. For example, the nearest neighbor number K is 5.

[0098] Step 3: At least the sample type label of the feature data extracted from each second sample is taken as the second result.

[0099] In step three, in addition to the sample type label of each second sample's extracted feature data, the similarity value between each second sample's extracted feature data and the first extracted feature data can also be used as a second result.

[0100] In one embodiment, the second result of obtaining the second propagation data using the second method is achieved through a prediction model. This embodiment further includes the following steps: Step 1: Determine the second extracted feature data of the second propagation data.

[0101] Before determining the second extracted feature data of the second propagation data, the signal parameters of several second signals in the second propagation data can be denoised, such as by wavelet transform. The relevant content will not be elaborated here.

[0102] In step one, the second extracted feature data of the second propagation data includes multiple second extracted features extracted using the second propagation data.

[0103] In one example, multiple second extracted features include: the maximum signal amplitude of several second signals in the second propagation data, the minimum signal amplitude of several second signals, the average signal amplitude of several second signals, the multiple squared values ​​of the signal amplitude of several second signals, the average signal amplitude attenuation rate of several second signals, the average signal propagation time of several second signals, the variance of the signal propagation time of several second signals, the entropy of the signal propagation time difference corresponding to several second signals, and the signal propagation time and signal amplitude of several second signals in the second propagation data, etc.

[0104] Among them, the multiple squared values ​​of the signal amplitudes of several second signals refer to the values ​​obtained after performing multiple squared operations on the signal amplitudes of several second signals.

[0105] The entropy of the signal propagation time difference corresponding to several second signals reflects the dispersion of the distribution of the signal propagation time difference (denoted by Δti) of each second signal. The Δti distribution is more dispersed along the propagation path of a single battery cell, resulting in a larger entropy value; the Δti distribution is more concentrated along the propagation path of non-battery cells (such as propagation through air), resulting in a smaller entropy value. The steps for determining the entropy of the signal propagation time difference corresponding to several second signals include: for each second signal, determining the difference between the signal propagation time of the second signal and the reference propagation time, as the signal propagation time difference of the second signal; summing the signal propagation time differences of each second signal to obtain the sum of signal propagation time differences; for each second signal, determining the ratio of the signal propagation time difference of the second signal to the sum of signal propagation time differences, as the probability of the signal propagation time difference of the second signal; substituting the probability of the signal propagation time difference of each second signal into the Shannon entropy formula to calculate the entropy of the signal propagation time difference corresponding to several second signals. The reference propagation time can be the minimum signal propagation time among the signal propagation times of several second signals. The Shannon entropy formula can be found in known techniques and will not be explained further here.

[0106] Furthermore, after determining the second extracted feature data of the second propagation data, the second extracted feature data can be normalized to map multiple second extracted features of the second extracted feature data to the [0,1] interval, thereby eliminating magnitude differences.

[0107] Step two: Use the prediction model to predict the second extracted feature data to obtain the second result.

[0108] The second result includes the prediction probability corresponding to the prediction type label of the second propagation data, where the prediction type label is either a battery cell or a non-battery cell. For example, the second result is the prediction probability when the prediction type label of the second propagation data is a battery cell.

[0109] In one example, the prediction model is a DNN (Deep Neural Network) model. The network structure of the DNN model includes an input layer, two hidden layers, and an output layer. Each hidden layer contains 64 neurons, and the activation function is ReLU. The output layer contains two neurons, and the activation function is Softmax. The output layer outputs the prediction probability when the prediction type label is "battery cell" or the prediction probability when the prediction type label is "non-battery cell".

[0110] The prediction model is trained using a constructed second sample dataset, which includes several third sample extracted feature data sets and corresponding sample type labels for each third sample extracted feature data set. The method for obtaining the third sample extracted feature data is the same as that for obtaining the second extracted feature data in the second propagation data, and will not be repeated here. The sample propagation data corresponding to the third sample extracted feature data undergoes the same denoising process as the second propagation data, and the third sample extracted feature data undergoes the same normalization process.

[0111] In one example, the feature data of several third samples in the second sample dataset are obtained by processing the propagation data of several samples acquired under a relatively ideal test environment. The relatively ideal test environment refers to an environment without propagation paths of resistance within the battery pack, air propagation paths, or reflection and refraction propagation paths from metal within the vehicle; only propagation paths through the individual battery cells within the battery pack exist. Furthermore, when acquiring the propagation data of several samples under the relatively ideal test environment, a battery pack with a known state (such as the presence of lithium dendrite growth, capacity decay, or abnormal internal resistance) can be selected to simulate the current and voltage conditions during actual charging, ensuring that the acquired propagation data of several samples is only related to the internal structure and state of the battery pack.

[0112] Feature data from a portion (e.g., 60%) of the third samples in the second sample dataset is used as a training set to train the prediction model, while feature data from another portion (e.g., 40%) of the third samples is used as a validation set to validate the prediction model.

[0113] In one example, during the training of the prediction model using the training set: the prediction model predicts based on the feature data extracted from the third sample, obtaining the sample prediction result corresponding to the feature data extracted from the third sample. The sample prediction result includes the sample prediction probability corresponding to the sample prediction type label; the training loss is calculated based on the sample prediction result corresponding to the feature data extracted from the third sample and the sample labeling result corresponding to the feature data extracted from the third sample, for example, by calculating the training loss function through cross-entropy; the model parameters of the prediction model are adjusted using the training loss, such as the connection weights, biases, activation function configurations, etc. of each layer of the prediction model.

[0114] In one example, the Adam optimizer can be used to adjust the model parameters of the prediction model, with an initial learning rate of 0.001 that decays with each training epoch. Details about the Adam optimizer can be found in known techniques and will not be elaborated upon here.

[0115] In one example, training of the prediction model is stopped when the prediction accuracy of the prediction model is consistently above 95% using a validation set.

[0116] In one embodiment, to improve the accuracy of determining the propagation medium type of the second propagation data, thereby improving the accuracy of identifying the target battery propagation data, after obtaining the aforementioned first and second results, the propagation medium type of the second propagation data is determined by combining the first and second results. This embodiment further includes the following sub-steps: Sub-step one: Use the first result to determine whether the first condition is met.

[0117] In one example, the first result includes sample type labels for the feature data extracted from each second sample. In this example, the first condition includes: the proportion of samples in the first result whose sample type label is "battery cell" is greater than or equal to a proportion threshold. The proportion threshold can be set according to actual needs, for example, a proportion threshold of 80%. When the proportion of samples in the first result whose sample type label is "battery cell" is greater than or equal to the proportion threshold, the first condition is determined to be met; when the proportion of samples in the first result whose sample type label is "battery cell" is less than the proportion threshold, the first condition is determined not to be met.

[0118] In another example, the first result includes sample type labels for the feature data extracted from each second sample. In this example, the first condition includes: the number of labels in the first result whose sample type label is "battery cell" is greater than or equal to a quantity threshold. The quantity threshold can be set according to actual needs, for example, a quantity threshold of 4. When the number of labels in the first result whose sample type label is "battery cell" is greater than or equal to the quantity threshold, the first condition is determined to be met; when the number of labels in the first result whose sample type label is "battery cell" is less than the quantity threshold, the first condition is determined not to be met.

[0119] In yet another example, the first result includes the sample type label of each second sample's extracted feature data and the similarity value between each second sample's extracted feature data and the first extracted feature data. In this example, the first condition includes: the first weight is greater than the second weight.

[0120] The first weight is obtained by using the similarity value between the feature data extracted from each second sample with the sample type label of battery cell and the first extracted feature data.

[0121] The calculation process of the first weight includes: determining the reciprocal of the similarity value of the extracted feature data of each second sample (i.e., the reciprocal of the similarity value); summing the reciprocals of the similarity values ​​of the extracted feature data of all second samples to obtain the first sum; summing the reciprocals of the similarity values ​​of the extracted feature data of each second sample with the sample type label of battery cell to obtain the second sum; and using the ratio of the second sum to the first sum as the first weight.

[0122] The second weight is obtained by using the similarity value between the feature data extracted from each second sample whose sample type label is non-battery cell and the first extracted feature data.

[0123] The calculation process for the second weight includes: determining the reciprocal of the similarity value of the extracted feature data of each second sample; summing the reciprocals of the similarity values ​​of the extracted feature data of all second samples to obtain the first sum; summing the reciprocals of the similarity values ​​of the extracted feature data of each second sample whose sample type label is non-battery cell to obtain the third sum; and using the ratio of the third sum to the first sum as the second weight.

[0124] In this example, when the first weight is greater than the second weight, the first condition is determined to be met; when the first weight is less than or equal to the second weight, the first condition is determined not to be met.

[0125] Sub-step two: Use the second result to determine whether the second condition is met.

[0126] In one example, the second result includes the predicted probability that the predicted type label of the second propagation data is a battery cell. The second condition includes: the predicted probability is greater than or equal to a first probability threshold. The first probability threshold is set according to actual needs, such as 0.95.

[0127] In this example, the second condition is determined to be met when the predicted probability is greater than or equal to the first probability threshold; the second condition is determined not to be met when the predicted probability is less than the first probability threshold.

[0128] In another example, if the second result includes the predicted probability that the predicted type label of the second propagation data is a non-battery cell, then the second condition includes: the predicted probability is less than or equal to a second probability threshold. The second probability threshold is set according to actual needs.

[0129] In this example, when the predicted probability is less than or equal to the second probability threshold, the second condition is determined to be met; when the predicted probability is greater than the second probability threshold, the second condition is determined not to be met.

[0130] Sub-step three: In response to the satisfaction of the first condition and the first condition, determine that the propagation medium type of the second propagation data is a battery cell.

[0131] In sub-step three, when both the first and second conditions are met, the propagation medium type of the second propagation data can be determined to be a battery cell, that is, the second propagation data is the target battery propagation data.

[0132] In one example, if neither the first nor the second condition is met, the propagation medium of the second propagation data is considered to be non-cell, such as propagation through air, resistance within the battery pack, or metal within the vehicle containing the battery pack.

[0133] In one example, to further improve the accuracy of determining the propagation medium type, when at least one of the first and second conditions is not met (i.e., the first condition is met but the second condition is not, or the second condition is met but the first condition is not, or neither the first nor the second condition is met), the propagation medium type of the second propagation data is not directly assumed to be a battery cell. Instead, the following processing procedure is performed: The first step is to check whether there are any abnormalities in the acquisition process of the first and second extracted feature data of the second propagation data. For example, whether denoising processing was performed; whether the minimum detectable threshold and multipath time resolution threshold used to determine the number of effective signals in several second signals are reasonable, such as if the minimum detectable threshold is too low, which may lead to the mixing of noise signals.

[0134] The second step is to re-collect and process at least one new second propagation data and re-verify whether the first and second conditions are both met in order to reduce random anomalies caused by interference in a single collection (such as environmental interference, temporary equipment jitter, etc.). If at least one of the first and second conditions is still not met, proceed to the third step.

[0135] The third step is to prompt for manual review.

[0136] In the third step, relevant personnel can conduct a manual review based on the vehicle's operating condition information (such as the charging status of the battery pack, the temperature of the battery pack, etc.).

[0137] Alternatively, in other embodiments, the propagation medium type of the second propagation data may be determined using only a single first result or a single second result.

[0138] For example, after obtaining the aforementioned first result, if the first result determines that the first condition is met, then the propagation medium type of the second propagation data is considered to be a battery cell; if the first result determines that the first condition is not met, then the propagation medium type of the second propagation data is considered to be a non-battery cell.

[0139] For example, after obtaining the second result, if the second result determines that the second condition is met, then the propagation medium type of the second propagation data is considered to be a battery cell; if the second result determines that the second condition is not met, then the propagation medium type of the second propagation data is considered to be a non-battery cell.

[0140] Please see Figure 10 , Figure 10 This is a flowchart illustrating an embodiment of a method for determining the state data of a battery pack provided in this application. This method is applied to a target charging device. Figure 10 As shown, the method includes the following steps: S101: Compare the target battery propagation data with historical battery propagation data to obtain the deviation results of the target battery propagation data.

[0141] S102: Use the deviation results to determine the status data of the battery pack.

[0142] In steps S101 and S102, the target battery propagation data includes battery propagation data (propagation medium type is individual battery cells) acquired by the target charging device during the current charging of the battery pack. For details on how to acquire the target battery propagation data, please refer to the previous embodiments; further details will not be repeated here.

[0143] Historical battery propagation data includes signal parameters from several historical signals. This historical battery propagation data serves as benchmark data for comparison with the target battery propagation data.

[0144] In one embodiment, the historical battery propagation data includes initial historical battery propagation data, which the target charging device uses as a benchmark for comparison to determine the battery pack's state data, such as the battery pack's target health value.

[0145] In one scenario, the initial historical battery propagation data refers to the battery propagation data acquired by the target charging device when the battery pack is first charged. In this scenario, when the battery pack is first charged by the target charging device, the target charging device can store the acquired battery propagation data as initial historical battery propagation data in its local storage module. The target charging device can then retrieve the initial historical battery propagation data from the local storage module and compare it with the initial historical battery propagation data to obtain the deviation result of the target battery propagation data.

[0146] In another scenario, considering that a user's vehicle usage history may involve charging the battery pack using different charging devices, the initial historical battery propagation data refers to the battery propagation data acquired by the target charging device or other charging devices during the initial charging of the battery pack. In this scenario, during the initial charging of the battery pack, the target charging device or other charging device can store the acquired battery propagation data as initial historical battery propagation data on a cloud server. Subsequently, the target charging device can retrieve the initial historical battery propagation data of the battery pack from the cloud server and compare the target battery propagation data with the initial historical battery propagation data to obtain the deviation result of the target battery propagation data.

[0147] In step S101, the target battery propagation data and historical battery propagation data are compared to obtain the deviation result of the target battery propagation data. The step further includes the following steps: obtaining the overall deviation and at least one local deviation of the target battery propagation data compared with the historical battery propagation data as the deviation result of the target battery propagation data; then, using the overall deviation and at least one local deviation, the state data of the battery pack is determined.

[0148] The overall deviation reflects the degree of influence of overall lithium plating on the battery pack's state. A larger overall deviation indicates a higher degree of lithium plating and a greater impact on the battery pack's state. The overall deviation is obtained using the signal parameters of all target signals in the target battery propagation data and the signal parameters of all historical signals in the historical battery propagation data.

[0149] In one embodiment, obtaining the overall deviation of the target battery propagation data from historical battery propagation data further includes the following steps: Step 1: Based on the signal parameters of several target signals in the target battery propagation data and the signal parameters of several historical signals in the historical battery propagation data, several first deviations are obtained.

[0150] Step one may further include the following sub-steps: Sub-step one involves sorting several target signals and several historical signals.

[0151] In one example, the signal parameters include signal propagation time. Several target signals are sorted in descending order of their propagation time, and several historical signals are sorted in descending order of their propagation time. In other examples, several target signals and several historical signals can also be sorted in ascending order of their propagation time.

[0152] In another example, the signal parameters include signal amplitude. Several target signals are sorted in descending order of their signal amplitude, and several historical signals are sorted in descending order of their signal amplitude. In other examples, several target signals and several historical signals may also be sorted in ascending order of their signal amplitude.

[0153] Sub-step two: For the i-th target signal and the i-th historical signal, compare the signal parameters of the i-th target signal with the signal parameters of the i-th historical signal to obtain the i-th first deviation.

[0154] In one example, the signal parameters only include signal propagation time. The signal propagation time of the i-th target signal is subtracted from the signal propagation time of the i-th historical signal to obtain the i-th signal propagation time deviation. The signal propagation time deviation of the i-th signal is divided by the signal propagation time of the i-th historical signal to obtain the i-th signal propagation time ratio. The i-th signal propagation time ratio is used as the i-th first deviation.

[0155] In another example, the signal parameters only include the signal amplitude. The signal amplitude of the i-th target signal is subtracted from the signal amplitude of the i-th historical signal to obtain the i-th signal amplitude deviation. The signal amplitude deviation of the i-th signal is divided by the signal amplitude of the i-th historical signal to obtain the i-th signal amplitude ratio. The i-th signal amplitude ratio is used as the i-th first deviation.

[0156] In another example, the signal parameters include both signal propagation time and signal amplitude. The average of the aforementioned i-th signal propagation time ratio and the aforementioned i-th signal amplitude ratio is determined as the i-th first deviation.

[0157] Step two: Combine several first deviations to obtain the overall deviation.

[0158] In step two, the average of several first deviations can be calculated as the overall deviation.

[0159] Local deviation reflects the degree of influence of local lithium dendrites on the battery pack's state. The greater the local deviation, the higher the degree of local lithium dendrites in the battery pack, and the greater the influence on the battery pack's state. Each local deviation is obtained using signal parameters from different local anomaly signal sets, each of which includes multiple target signals and multiple historical signals.

[0160] In one embodiment, obtaining at least one local deviation of the target battery propagation data compared to historical battery propagation data further includes the following steps: Step 1: Based on the signal parameters of several target signals in the target battery propagation data and the signal parameters of several historical signals in the historical battery propagation data, several first deviations are obtained.

[0161] For details on obtaining information related to the first deviation, please refer to the corresponding content in the implementation method for determining the overall deviation, which will not be repeated here.

[0162] Step 2: Select multiple second deviations that meet the abnormal deviation requirements from a number of first deviations.

[0163] The abnormal deviation requirement includes: the deviation degree is greater than a first deviation threshold. The first deviation threshold can be preset according to actual needs.

[0164] Step 3: Use multiple second deviations to obtain at least one local deviation.

[0165] In one example, the average of multiple second deviations is determined as a local deviation. In this example, all target signals and all historical signals involved in the multiple second deviations constitute a local anomalous signal set. For example, the number of second deviations that meet the anomalous deviation criteria is 64. These 64 second deviations are determined using the signal parameters of 64 target signals and 64 historical signals, and these 64 target signals and 64 historical signals constitute a local anomalous signal set.

[0166] In another example, a clustering algorithm is used to cluster multiple second deviations, resulting in at least one deviation cluster. Each deviation cluster includes multiple third deviations, meaning that second deviations with similar values ​​are grouped into the same deviation cluster. For each deviation cluster, the average of the multiple third deviations within that cluster is determined to obtain the local deviation corresponding to that cluster. For example, the clustering algorithm can be a known algorithm such as K-means, which will not be described in detail here. In this example, one deviation cluster corresponds to one local deviation, and all target signals and all historical signals involved in a deviation cluster constitute a set of local anomalous signals. For example, deviation cluster 1 and deviation cluster 2 are obtained through clustering. Deviation cluster 1 includes 64 third deviations, and deviation cluster 2 also includes 64 third deviations. The 64 third deviations of deviation cluster 1 are determined using the signal parameters of 64 target signals and 64 historical signals. These 64 target signals and 64 historical signals constitute a set of local anomalous signals. The 64 third deviations of deviation cluster 2 are determined using the signal parameters of another 64 target signals and another 64 historical signals. These other 64 target signals and these other 64 historical signals constitute another set of local anomaly signals.

[0167] Understandably, in this example, a local deviation corresponds to a lithium dendrite appearing in the battery pack.

[0168] Figure 11 This is a schematic diagram showing lithium dendrites appearing in a single battery cell provided in this application. Figure 11 As shown, the black battery cell BT x This refers to a single battery cell exhibiting lithium dendrites. For example... Figure 11 The thick black line in the diagram indicates the path from the overall positive terminal of the battery pack through the individual battery cell BT. x There are a total of 1*8*8=64 propagation paths from the individual battery cells to the overall negative terminal of the battery pack, corresponding to 64 target signals, and the individual battery cell BT... x When an anomaly occurs, the signal parameters of the 64 target signals and the battery cell BT are related. xWhen no anomalies are observed, the signal parameters of the 64 target signals are different.

[0169] In step S102, the battery pack status data includes the target health value of the battery pack. The battery pack status data is determined using the overall deviation and at least one local deviation, and further includes the following steps: Step 1: Determine whether there is a target local deviation that meets the deviation conditions among at least one local deviation.

[0170] The deviation condition includes: the deviation degree is greater than a second deviation threshold. The second deviation threshold can be preset according to actual needs. For example, the second deviation threshold is 5%, and the target local deviation degree is greater than 5%.

[0171] If there is no local deviation from the target, proceed to step two. If there is a local deviation from the target, proceed to step three.

[0172] Step two: Determine the target health value using the overall deviation.

[0173] In step two, the overall deviation is inversely proportional to the target health value. That is, the higher the overall deviation, the higher the degree of lithium plating in the battery pack, and the lower the corresponding target health value; the lower the overall deviation, the lower the degree of lithium plating in the battery pack, and the higher the corresponding target health value.

[0174] In one embodiment, the step of determining the target health value using the overall deviation further includes: determining a first health value using the overall deviation; and determining the target health value using the first health value.

[0175] The step of determining the first health value using the overall deviation further includes: subtracting the overall deviation from the preset value to obtain the first difference; and using the product of the first difference and the overall deviation weight as the first health value.

[0176] For example, the default value is 1.

[0177] For example, the overall deviation weight is 1. The overall deviation weight can be set to other values ​​according to actual needs.

[0178] The step of determining the target health value using the first health value further includes: using the first health value as the target health value; or, using the product of the first health value and 100% as the target health value.

[0179] In a specific application, the process of determining the first health value can be represented by the following formula: Hn = (1 - RDn) * Q0 Where Hn represents the first health value, RDn represents the overall deviation, and Q0 represents the overall deviation weight.

[0180] In a specific application, the process of determining the target health value can be represented by the following formula: SOH = Hn * 100% Where SOH represents the target health value and Hn represents the first health value.

[0181] Step 3: Determine the target health value by using the overall deviation and the local deviation of each target.

[0182] In one embodiment, the target health value is determined using the overall deviation and the local deviation of each target, further including the following sub-steps: Sub-step one: Use the overall deviation to determine the first health value.

[0183] The process of determining the first health value in sub-step one can refer to the relevant content in the previous implementation method, and will not be repeated here.

[0184] Sub-step two: For each target local deviation, use the target local deviation to determine the second health value.

[0185] In sub-step two, a target local deviation corresponds to a determined second health value. The step of determining the second health value corresponding to the target local deviation further includes: subtracting the target local deviation from a preset value to obtain a second difference; and using the product of the second difference and the local deviation weight of the target local deviation as the second health value corresponding to the target local deviation.

[0186] Sub-step three: Determine the target health value by multiplying the first health value by each of the second health values.

[0187] The target health value can be obtained by directly multiplying the first health value by each of the second health values. Alternatively, the target health value can be obtained by multiplying the product of the first health value and each of the second health values ​​by 100%.

[0188] In a specific application, the process of determining the second health value corresponding to the local deviation of the target can be represented by the following formula: hn = (1 - rdn) * Qn Where hn represents the second health value corresponding to the target local deviation, rdn represents the target local deviation, and Qn represents the local deviation weight corresponding to the target local deviation. The local deviation weights corresponding to each target local deviation can be the same or different. For example, the local deviation weight for each target local deviation can be 0.5. The local deviation weights corresponding to each target local deviation can be set according to actual needs.

[0189] In a specific application, if there is only one local deviation of the target, the process of determining the target health value can be represented by the following formula: SOH = Hn * hn * 100% Where SOH represents the target health value, Hn represents the first health value, and hn represents the second health value.

[0190] The following example illustrates the process of determining the target health value of the battery pack in steps one through three: On the 10th charge of the 40th day, the overall deviation was detected to be 4%. The target health value of the battery pack on the 40th day was determined by the following formula: SOH = (1-0.04) * 1 * 100% = 96%.

[0191] On the 20th charge of the 80th day, the overall deviation was detected as 5%, and a local deviation of 15% was detected. The target health value of the battery pack on the 80th day was determined by the following formula: SOH = (1-0.05) * 1 * (1-0.15) * 0.5 * 100% = 40%.

[0192] On the 30th charge of the 120th day, the overall deviation was detected to be 6%, and a local deviation of 30% was detected. The target health value of the battery pack on the 120th day was determined by the following formula: SOH = (1-0.06) * 1 * (1-0.3) * 0.5 * 100% = 31%.

[0193] On the 40th charge of the 160th day, the overall deviation was detected as 7%, a local deviation of 40% and another local deviation of 10%. The target health value of the battery pack on the 160th day was determined by the following formula: SOH = (1-0.07) * 1 * (1-0.4) * 0.5 * (1-0.1) * 0.5 * 100% = 13%.

[0194] In steps one through three above, when none of the determined local deviations meet the deviation conditions, it is considered that the battery pack has no lithium dendrites or the degree of lithium dendrites is low. Only the influence of lithium plating degree on the battery pack can be considered, and the influence of lithium dendrite degree on the battery pack is not considered. Therefore, the target health value of the battery pack is determined only using the overall deviation. When there is a target local deviation that meets the deviation conditions among the determined local deviations, it is considered that the lithium dendrites in the battery pack have reached a certain degree. In addition to considering the influence of lithium plating degree on the battery pack, the influence of lithium dendrite degree on the battery pack also needs to be considered. Therefore, the target health value of the battery pack is determined using the overall deviation and each target local deviation, so as to improve the rationality and accuracy of the determined target health value of the battery pack.

[0195] Please see Figure 12 , Figure 12 This is a flowchart illustrating another embodiment of the method for determining the state data of a battery pack provided in this application, which is applied to a target charging device. Figure 12 As shown, the method includes the following steps: S121: Obtain the target vehicle attribute information of the vehicle where the battery pack is located.

[0196] In one embodiment, the target charging device can communicate with the vehicle controller of the vehicle where the battery pack is located via CAN to obtain the target vehicle attribute information of the vehicle where the battery pack is located.

[0197] For example, the target vehicle attribute information includes the manufacturer information of the vehicle to which the battery pack is located and the model information of the vehicle to which the battery pack is located.

[0198] S122: Retrieve the target dataset corresponding to the target vehicle attribute information from the database.

[0199] Figure 13 This is a schematic diagram of the database provided in this application. For example... Figure 13 As shown, the database includes several preset datasets corresponding to preset vehicle attribute information. Each preset dataset includes several preset battery propagation data sets and preset state data corresponding to each preset battery propagation data set.

[0200] In one embodiment, the preset vehicle attribute information includes the manufacturer information of the preset vehicle and the model information of the preset vehicle.

[0201] In one embodiment, for each preset battery propagation data in the preset dataset, the preset state data corresponding to the preset battery propagation data includes health values ​​(health scores or health levels) or fault levels, etc., corresponding to the preset battery propagation data. For example, the preset dataset for a certain model from a certain car manufacturer includes preset battery propagation data corresponding to health levels of 100%, 99%, 98%...1% and 0%. As another example, the preset dataset for a certain model from a certain car manufacturer includes preset battery propagation data corresponding to fault levels of no fault, minor fault, moderate fault, and severe fault.

[0202] In one implementation, the database may reside on a local or cloud server of the target charging device. The data in the database is acquired by the target charging device and other charging devices when charging battery packs from a large number of vehicles from different manufacturers.

[0203] In step S122, the target dataset is one of the preset datasets. The preset dataset corresponding to the preset vehicle attribute information that is the same as the target vehicle attribute information is obtained from the database and used as the target dataset.

[0204] S123: Obtain the preset state data corresponding to the preset battery propagation data that matches the target battery propagation data in the target dataset, and use it as the state data of the battery pack.

[0205] After obtaining the target dataset, the preset battery propagation data in the target dataset are further matched with the target battery propagation data. The preset state data corresponding to the preset battery propagation data that matches the target battery propagation data is used as the battery pack's state data. For example, the health value corresponding to the preset battery propagation data that matches the target battery propagation data is used as the target health value of the battery pack.

[0206] In one embodiment, the target battery propagation data includes signal parameters of several target signals, and each preset battery propagation data includes signal parameters of several preset signals. The step of determining the preset battery propagation data in the target data set that matches the target battery propagation data further includes the following steps: Step 1: For each preset battery propagation data, compare the signal parameters of several preset signals in the preset battery propagation data with the signal parameters of several target signals in the target battery propagation data to obtain several first signal deviation values ​​corresponding to the preset battery propagation data.

[0207] Step one may further include the following sub-steps: Sub-step one involves sorting several preset signals and several target signals.

[0208] In sub-step one, several preset signals and several target signals can be sorted according to the magnitude of signal propagation time or the magnitude of signal amplitude.

[0209] Sub-step two: For the j-th preset signal and the j-th target signal, compare the signal parameters of the j-th preset signal with the signal parameters of the j-th target signal to obtain the j-th first signal deviation value.

[0210] In one example, the signal parameters only include the signal propagation time. The propagation time of the j-th preset signal is subtracted from the propagation time of the j-th target signal to obtain the propagation time deviation of the j-th signal. The propagation time deviation of the j-th signal is used as the deviation value of the j-th first signal.

[0211] In another example, the signal parameters only include the signal amplitude. The signal amplitude of the j-th preset signal is subtracted from the signal amplitude of the j-th target signal to obtain the j-th signal amplitude deviation. The j-th signal amplitude deviation is used as the j-th first signal deviation value.

[0212] In another example, the signal parameters include both signal propagation time and signal amplitude. The average of the aforementioned j-th signal propagation time deviation and the aforementioned j-th signal amplitude deviation is determined as the j-th first signal deviation value.

[0213] Step 2: For each preset battery propagation data, combine several first signal deviation values ​​of the preset battery propagation data to obtain the second signal deviation value of the preset battery propagation data.

[0214] For example, the average of several first signal deviation values ​​of preset battery propagation data can be calculated as the second signal deviation value of preset battery propagation data.

[0215] Step 3: Select the preset battery propagation data whose second signal deviation value meets the requirements, and use it as the preset battery propagation data that matches the target battery propagation data.

[0216] For example, the preset battery propagation data with the smallest second signal deviation value is selected as the preset battery propagation data that matches the target battery propagation data.

[0217] After determining the status data of the battery pack, the status data of the battery pack and the corresponding target battery propagation data can be further added to the corresponding preset dataset in the database to update the data in the database.

[0218] In this embodiment, preset battery propagation data that matches the target battery propagation data of the battery pack can be obtained from the database, and the corresponding preset state data can be obtained as the state data of the battery pack. As the data in the database is continuously updated and improved, the amount of data in the database increases, making the determined state data of the battery pack more accurate.

[0219] Please see Figure 13 , Figure 13 This is a flowchart illustrating another embodiment of the method for determining the state data of a battery pack provided in this application, which is applied to a target charging device. Figure 13 As shown, the method includes the following steps: S131: Determine if the battery pack is being charged for the first time.

[0220] The first charge of the battery pack refers to the first charge of the battery pack made by the user during vehicle use through the target charging device; or, the first charge of the battery pack refers to the first charge of the battery pack made by the user during vehicle use, which may be made through other charging devices.

[0221] In one embodiment, if the first historical battery propagation data of the battery pack is obtained, the current charging of the battery pack is considered not to be the first charging; if the first historical battery propagation data of the battery pack is not obtained, the current charging of the battery pack is considered to be the first charging.

[0222] Furthermore, if the battery pack is being charged for the first time, it is assumed that there is no historical battery propagation data for reference, and steps S132 to S134 are executed. If the battery pack is not being charged for the first time, steps S135 to S136 are executed.

[0223] S132: Obtain the target vehicle attribute information of the vehicle where the battery pack is located.

[0224] S133: Retrieve the target dataset corresponding to the target vehicle attribute information from the database.

[0225] S134: Obtain the preset state data corresponding to the preset battery propagation data that matches the target battery propagation data in the target dataset, and use it as the state data of the battery pack.

[0226] The contents of steps S132 to S134 can be referred to Figure 12 Steps S121 to S123 in the illustrated embodiment will not be repeated here.

[0227] S135: Compare the target battery propagation data with historical battery propagation data to obtain the deviation results of the target battery propagation data.

[0228] S136: Use the deviation results to determine the status data of the battery pack.

[0229] Please see Figure 14 , Figure 14 This is a flowchart illustrating another embodiment of the method for determining the state data of a battery pack provided in this application, which is applied to a target charging device. Figure 14 As shown, the method includes the following steps: S141: Determine multiple target extraction features corresponding to the target battery propagation data.

[0230] In one embodiment, the target battery propagation data includes signal parameters of several target signals, and the multiple target extraction features include the number of effective signals among the several target signals, the signal propagation time difference entropy corresponding to the several target signals, the amplitude attenuation rate of the several target signals, and the distribution entropy of the effective signal propagation time difference corresponding to the several target signals. The specific determination process of the number of effective signals, the signal propagation time difference entropy, the amplitude attenuation rate, and the distribution entropy of the effective signal propagation time difference can be referred to the corresponding content above, and will not be repeated here.

[0231] S142: For each target extraction feature, the first evaluation value of the target extraction feature is determined by using the feature value of the target extraction feature and the health benchmark value of the target extraction feature.

[0232] In one embodiment, for each target extracted feature, determining a first evaluation value for the target extracted feature further includes the following steps: Step 1: Determine the deviation rate of the feature values ​​of the extracted target features relative to the health baseline values ​​of the extracted target features.

[0233] In step one, the difference between the feature value of the target extracted feature (i.e., the specific numerical value corresponding to the target extracted feature) and the health benchmark value of the target extracted feature is determined. The difference between the feature value of the target extracted feature and the health benchmark value of the target extracted feature is divided by the health benchmark value of the target extracted feature to obtain the deviation rate of the target extracted feature. The deviation rate of the target extracted feature reflects the degree of deviation between the target extracted feature and the corresponding health benchmark value.

[0234] The health baseline values ​​for different target extracted features are preset according to the actual situation. For example, the baseline value corresponding to the number of effective signals in several target signals is 500 Neff, and the baseline value corresponding to the signal propagation time difference entropy of several target signals is 2.5 bits.

[0235] Step 2: Determine the product of the weight coefficient of the target extracted feature and the deviation rate of the target extracted feature, and use it as the first evaluation value of the target extracted feature.

[0236] In step two, different target extracted features each have corresponding weight coefficients, which may be the same or different for different target extracted features. Furthermore, the weight coefficients of the target extracted features represent the priority or importance of the target extracted features.

[0237] In one embodiment, the target battery propagation data includes signal parameters of several target signals. The multiple target extraction features include the number of effective signals in the several target signals, the signal propagation time difference entropy corresponding to the several target signals, the amplitude attenuation rate of the several target signals, and the distribution entropy of the effective signal propagation time difference corresponding to the several target signals. Furthermore, the weighting coefficient corresponding to the number of effective signals in the several target signals is 0.35, the weighting coefficient corresponding to the signal propagation time difference entropy corresponding to the several target signals is 0.30, the weighting coefficient corresponding to the amplitude attenuation rate of the several target signals is 0.20, and the weighting coefficient corresponding to the distribution entropy of the effective signal propagation time difference corresponding to the several target signals is 0.15.

[0238] In step two, the first evaluation value of the extracted target features can reflect the degree of anomaly of the extracted target features.

[0239] In one embodiment, the first evaluation value of each target extracted feature can be obtained from the output of the intermediate layer of the aforementioned prediction model (such as a DNN model). For example, the intermediate layer of the prediction model (such as each neuron in the hidden layer) can output the feature activation value corresponding to each target extracted feature, that is, the first evaluation value corresponding to each target extracted feature.

[0240] S143: Determine the second evaluation value by combining the first evaluation value of the extracted features of each target.

[0241] In one embodiment, the sum of the first evaluation values ​​is used as the second evaluation value. The second evaluation value reflects the overall degree of abnormality of the battery pack.

[0242] S144: Use the second evaluation value to determine the status data of the battery pack.

[0243] In one embodiment, the battery pack status data includes a target fault level for the battery pack. Determining the target fault level of the battery pack using a second evaluation value includes: determining the target interval to which the second evaluation value belongs; and determining the fault level corresponding to the target interval from a preset mapping relationship, as the target fault level. The preset mapping relationship includes several preset intervals and preset fault levels corresponding to each preset interval. The target interval is one of the several preset intervals.

[0244] For example, several preset intervals include (0, 0.1), [0.1, 0.3), [0.3, 0.6), and [0.6, 1]. Specifically, the preset fault level corresponding to the preset interval (0, 0.1) is no fault, where the battery pack's health value is greater than 95%; the preset fault level corresponding to the preset interval [0.1, 0.3) is a minor fault, where the battery pack's health value is between 90% and 95%; the preset fault level corresponding to the preset interval [0.3, 0.6) is a moderate fault, where the battery pack's health value is between 80% and 90%; and the preset fault level corresponding to the preset interval [0.6, 1] is a severe fault, where the battery pack's health value is less than 80%.

[0245] In one embodiment, the battery pack status data includes the target fault level of the battery pack, and the method further includes: determining the highest evaluation value among each first evaluation value; and using the target extraction feature corresponding to the highest evaluation value and the target fault level to determine the target fault cause of the battery pack.

[0246] In this implementation, the target cause of the battery pack failure can be located by referring to the table below.

[0247] Table 1:

[0248] For example, if the target fault level of the battery pack is determined to be severe fault, and the target extracted feature corresponding to the highest evaluation value is the number of effective signals among several target signals, it indicates that a lot of disordered paths have been added inside the battery pack (such as non-uniformity of battery dielectric caused by lithium dendrites). The target fault cause of the battery pack can be determined to be "mild disorder of battery internal structure".

[0249] For example, if the target fault level of the battery pack is determined to be a minor fault, and the target extraction feature corresponding to the highest evaluation value is the amplitude attenuation rate of several target signals, it indicates that the battery dielectric loss is aggravated, and the target fault cause of the battery pack can be determined to be "a slight increase in battery internal resistance or electrolyte aging".

[0250] Optionally, in this embodiment, when using Figure 10 , Figure 12 , Figure 13 or Figure 14 After determining the battery pack status data using the method shown, further battery pack status data can be output.

[0251] For example, the target charging device can announce the battery pack's status data via voice.

[0252] For example, the charging station for the target charging device also includes a display screen, on which the target charging device can display the status data of the battery pack.

[0253] For example, the target charging device can push battery pack status data to the user's terminal device. The user's terminal device can be a user's mobile phone, vehicle terminal, etc.

[0254] Optionally, in this embodiment, when using Figure 10 , Figure 12 , Figure 13 or Figure 14 After determining the battery pack status data using the method shown, further maintenance recommendations corresponding to the battery pack status data can be output.

[0255] For example, the battery pack status data is input into a large language model, which is then used to generate maintenance suggestions corresponding to the battery pack status data, and the maintenance suggestions are output.

[0256] For example, maintenance suggestions can be pushed to users' terminal devices.

[0257] Optionally, in this embodiment, multiple historical health values ​​obtained from previous state checks of the battery pack can also be used to predict the subsequent health development of the battery pack. This embodiment may include the following steps: Step 1: Using multiple historical health values ​​of the battery pack and the historical time corresponding to each historical health value, a health prediction curve is obtained by fitting.

[0258] In step one, a known curve fitting method can be used to obtain a health prediction curve, which will not be explained further here. The fitted health prediction curve represents the relationship between the battery pack's historical health value and historical time. Historical time can be in days.

[0259] Step 2: Use the health prediction curve to determine the time information corresponding to the preset health value of the battery pack.

[0260] By substituting a preset health value into the health prediction curve, the target time corresponding to the preset health value can be obtained. This target time is then used as the time information corresponding to the battery pack's health value being the preset health value. For example, the preset health value is 0, but it can also be any other health value.

[0261] In other embodiments, multiple historical health values ​​of the battery pack and the historical time corresponding to each historical health value can be imported into Excel, and the forecast worksheet function in Excel can be used to visualize the health values ​​of the battery pack in terms of time. Figure 15 This is a trend graph showing the change in the health value of the battery pack provided in this application over time. Figure 15 It is foreseeable that the battery pack's health value will be reset to zero within the next 170-220 days.

[0262] Please see Figure 16 , Figure 16 This is a schematic diagram of a framework of an embodiment of the battery pack state detection device provided in this application. In this embodiment, the battery pack state detection device 160 includes: a signal acquisition module 161 and a state detection module 162.

[0263] The signal acquisition module 161 is used to transmit electromagnetic wave signals through a signal transmitter and to acquire target battery propagation data of the electromagnetic wave signals through a signal receiver. The target battery propagation data includes signal parameters of several target signals. The target signals are electromagnetic wave signals propagated from individual battery cells inside the battery pack to the signal receiver. Different target signals have different propagation paths within the battery cells, and the signal parameters of the target signals include at least one of the signal propagation time and signal amplitude.

[0264] The status detection module 162 is used to determine the status data of the battery pack by using data transmitted from the target battery.

[0265] In one embodiment, the state detection module 162 is used to compare the target battery propagation data with historical battery propagation data to obtain the deviation result of the target battery propagation data; and to determine the state data of the battery pack using the deviation result.

[0266] In one embodiment, the state detection module 162 is used to acquire the overall deviation and at least one local deviation of the target battery propagation data compared with the historical battery propagation data. The overall deviation is obtained using the signal parameters of all target signals in the target battery propagation data and the signal parameters of all historical signals in the historical battery propagation data. Each local deviation is obtained using the signal parameters of different local abnormal signal sets. Each local abnormal signal set includes multiple target signals and multiple historical signals. The state data of the battery pack is determined using the overall deviation and at least one local deviation.

[0267] In one embodiment, the battery pack status data includes the target health value of the battery pack. The status detection module 162 is used to determine whether there is a target local deviation that meets the deviation condition among at least one local deviation. If there is no target local deviation, the target health value is determined using the overall deviation. If there is a target local deviation, the target health value is determined using the overall deviation and each target local deviation.

[0268] In one embodiment, when there is a target local deviation, the state detection module 162 is used to determine a first health value using the overall deviation; for each target local deviation, a second health value is determined using the target local deviation; and the target health value is determined by multiplying the first health value by each second health value.

[0269] In one embodiment, the state detection module 162 is used to obtain target vehicle attribute information of the vehicle where the battery pack is located; obtain the target dataset corresponding to the target vehicle attribute information from the database, wherein the database includes several preset datasets corresponding to preset vehicle attribute information, the preset datasets include several preset battery propagation data and preset state data corresponding to each preset battery propagation data, and the target dataset is one of the preset datasets; obtain the preset state data corresponding to the preset battery propagation data that matches the target battery propagation data in the target dataset, and use it as the state data of the battery pack.

[0270] In one embodiment, the state detection module 162 is used to determine multiple target extraction features corresponding to the target battery propagation data; for each target extraction feature, a first evaluation value of the target extraction feature is determined using the feature value of the target extraction feature and the health benchmark value of the target extraction feature; a second evaluation value is determined by combining the first evaluation values ​​of each target extraction feature; and the state data of the battery pack is determined using the second evaluation value.

[0271] In one embodiment, the state detection module 162 is used to determine the deviation rate of the feature value of the target extracted feature relative to the benchmark value of the target extracted feature; and to determine the product of the weight coefficient of the target extracted feature and the deviation rate of the target extracted feature as the first evaluation value of the target extracted feature.

[0272] And / or, the state detection module 162 is used to sum the first evaluation values ​​as the second evaluation value.

[0273] And / or, the battery pack status data includes the target fault level of the battery pack, and the status detection module 162 is used to determine the target interval to which the second evaluation value belongs; determine the fault level corresponding to the target interval from the preset mapping relationship as the target fault level, wherein the preset mapping relationship includes several preset intervals and the preset fault level corresponding to each preset interval.

[0274] And / or, the battery pack status data includes the target fault level of the battery pack, and the status detection module 162 is also used to determine the highest evaluation value among the first evaluation values; using the target extraction features corresponding to the highest evaluation value and the target fault level, the target fault cause of the battery pack is determined.

[0275] In one embodiment, the signal acquisition module 161 is used to acquire first propagation data of electromagnetic wave signals through a signal receiver, wherein the first propagation data includes signal parameters of a plurality of first signals; divide the first propagation data into a plurality of second propagation data of different propagation medium types; and identify the propagation data of a battery cell as the propagation medium type from the plurality of second propagation data as the target battery propagation data.

[0276] In one embodiment, for each second propagation data, the signal acquisition module 161 is used to determine the propagation medium type of the second propagation data using at least one of the first result of the first method and the second result of the second method.

[0277] In the first approach: first extracted feature data of the second propagation data is determined, the first extracted feature data includes multiple first extracted features; multiple second sample extracted feature data similar to the first extracted feature data are determined from the first sample dataset, wherein the first sample dataset includes several first sample extracted feature data and sample type labels corresponding to each first sample extracted feature data, the sample type labels being battery cell or non-battery cell; at least the sample type labels of each second sample extracted feature data are taken as the first result.

[0278] In the second approach: determine the second extracted feature data of the second propagation data, the second extracted feature data includes multiple second extracted features; use a prediction model to predict the second extracted feature data to obtain a second result, the second result includes the prediction probability corresponding to the prediction type label of the second propagation data, the prediction type label is battery cell or non-battery cell.

[0279] In one embodiment, the signal acquisition module 161 is used to determine whether a first condition is met using a first result; wherein the first condition is any of the following: the proportion of samples labeled as battery cells in the first result is greater than or equal to a proportion threshold; the number of samples labeled as battery cells in the first result is greater than or equal to a number threshold; a first weight is greater than a second weight, the first weight is obtained by using the similarity value between the extracted feature data of each second sample labeled as battery cells and the first extracted feature data, and the second weight is obtained by using the similarity value between the extracted feature data of each second sample labeled as non-battery cells and the first extracted feature data; and the second result is used to determine whether a second condition is met; wherein the second condition includes: the predicted type label is battery cells and the predicted probability is greater than or equal to a first probability threshold; in response to meeting the first and second conditions, the propagation medium type of the second propagation data is determined to be battery cells.

[0280] In one embodiment, the electromagnetic wave signal carries first identification information. The signal acquisition module 161 is used to acquire initial propagation data of the electromagnetic wave signal through the signal receiving end. The initial propagation data includes signal parameters of several initial signals. The module verifies whether the second identification information carried in each initial signal is consistent with the first identification information. The module removes the signal parameters of the initial signals whose second identification information is inconsistent with the first identification information from the initial propagation data to obtain the first propagation data. The module determines the propagation data of the target battery from the first propagation data.

[0281] And / or, the battery pack status data includes the target health value of the battery pack, and the battery pack status detection device 160 also includes a prediction module 163. The prediction module 163 is used to fit a health prediction curve using multiple historical health values ​​of the battery pack and the historical time corresponding to each historical health value; and to determine the time information corresponding to the preset health value using the health prediction curve.

[0282] It should be noted that the apparatus of this embodiment can perform the steps in the above method. For detailed descriptions of the relevant content, please refer to the method section above, which will not be repeated here.

[0283] Please see Figure 17 , Figure 17 This is a schematic diagram of a framework of an embodiment of the electronic device provided in this application. In this embodiment, the electronic device 170 includes a memory 171 and a processor 172.

[0284] Processor 172 can also be referred to as CPU (Central Processing Unit). Processor 172 may be an integrated circuit chip with signal processing capabilities. Processor 172 can also be a general-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component. A general-purpose processor can be a microprocessor, or processor 172 can be any conventional processor 172, etc.

[0285] The memory 171 in the electronic device 170 is used to store the program instructions required for the processor 172 to run.

[0286] The processor 172 is used to execute program instructions to implement the battery pack state detection method in this application.

[0287] Please see Figure 18 , Figure 18 This is a schematic diagram of a framework of an embodiment of the computer-readable storage medium provided in this application. The computer-readable storage medium 180 of this application embodiment stores program instructions 181, which, when executed, implement the battery pack state detection method provided in this application. The program instructions 181 can be formed into a program file and stored in the aforementioned computer-readable storage medium 180 in the form of a software product, so that a computer device (which may be a personal computer, server, or network device, etc.) can execute all or part of the steps of the methods of various embodiments of this application. The aforementioned computer-readable storage medium 180 includes various media capable of storing program code, such as a USB flash drive, mobile hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, or terminal devices such as computers, servers, mobile phones, and tablets.

[0288] In some embodiments, the functions or modules of the apparatus provided in this disclosure can be used to perform the methods described in the above method embodiments. The specific implementation can be referred to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.

[0289] The description of the various embodiments above tends to emphasize the differences between the various embodiments. The similarities or similarities between them can be referred to, and for the sake of brevity, they will not be repeated here.

[0290] In the several embodiments provided in this application, it should be understood that the disclosed methods, apparatuses, and systems can be implemented in other ways. For example, the apparatus implementations described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of apparatuses or units may be electrical, mechanical, or other forms.

[0291] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0292] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0293] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods of various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0294] The above description is merely an embodiment of this application and does not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for detecting the state of a battery pack, characterized in that, The method is applied to a target charging device for charging a battery pack, and includes: Electromagnetic wave signals are emitted through the signal transmitting end; The target battery propagation data of the electromagnetic wave signal is obtained through the signal receiving end; wherein, the target battery propagation data includes signal parameters of several target signals, the target signals are the signals of the electromagnetic wave signal propagated from the battery cells inside the battery pack to the signal receiving end, the propagation paths of the battery cells of different target signals are different, and the signal parameters of the target signals include at least one of the signal propagation time and signal amplitude of the target signals. The state data of the battery pack is determined by using the data propagated from the target battery.

2. The method according to claim 1, characterized in that, The step of using the data propagated from the target battery to determine the status data of the battery pack includes: By comparing the target battery propagation data with historical battery propagation data, the deviation result of the target battery propagation data is obtained; The deviation results are used to determine the status data of the battery pack.

3. The method according to claim 2, characterized in that, The step of comparing the target battery propagation data with historical battery propagation data to obtain the deviation result of the target battery propagation data includes: The overall deviation and at least one local deviation of the target battery propagation data compared to the historical battery propagation data are obtained. The overall deviation is obtained using the signal parameters of all target signals in the target battery propagation data and the signal parameters of all historical signals in the historical battery propagation data. Each local deviation is obtained using the signal parameters of different local abnormal signal sets. Each local abnormal signal set includes multiple target signals and multiple historical signals. The step of using the deviation result to determine the status data of the battery pack includes: The state data of the battery pack is determined using the overall deviation and the at least one local deviation.

4. The method according to claim 3, characterized in that, The battery pack status data includes the target health value of the battery pack. Determining the battery pack status data using the overall deviation and the at least one local deviation includes: Determine whether there is a target local deviation among the at least one local deviation that meets the deviation condition; If the target local deviation does not exist, the target health value is determined using the overall deviation. If the target local deviation exists, the target health value is determined by using the overall deviation and each of the target local deviations.

5. The method according to claim 4, characterized in that, In the case of the aforementioned local deviation, determining the target health value using the overall deviation and each of the aforementioned local deviations includes: The first health value is determined using the overall deviation. For each of the aforementioned target local deviations, a second health value is determined using the target local deviations; The target health value is determined by multiplying the first health value by each of the second health values.

6. The method according to claim 1, characterized in that, The step of using the data propagated from the target battery to determine the status data of the battery pack includes: Obtain the target vehicle attribute information of the vehicle where the battery pack is located; The target dataset corresponding to the target vehicle attribute information is obtained from the database, wherein the database includes a number of preset datasets corresponding to preset vehicle attribute information, the preset datasets include a number of preset battery propagation data and preset state data corresponding to each preset battery propagation data, and the target dataset is one of the preset datasets; Obtain the preset state data corresponding to the preset battery propagation data that matches the target battery propagation data in the target dataset, and use it as the state data of the battery pack.

7. The method according to claim 1, characterized in that, The step of using the data propagated from the target battery to determine the battery's state data includes: Determine multiple target extraction features corresponding to the target battery propagation data; For each of the target extraction features, a first evaluation value of the target extraction feature is determined using the feature value of the target extraction feature and the health benchmark value of the target extraction feature; A second evaluation value is determined by combining the first evaluation values ​​of the extracted features of each target. The status data of the battery pack is determined using the second evaluation value.

8. The method according to claim 7, characterized in that, The step of determining a first evaluation value of the target extracted features using the feature values ​​of the target extracted features and the benchmark value of the target extracted features includes: Determine the deviation rate of the feature value of the extracted target feature relative to the benchmark value of the extracted target feature; The product of the weight coefficient of the target extracted feature and the deviation rate of the target extracted feature is determined as the first evaluation value of the target extracted feature; And / or, determining the second evaluation value by integrating the first evaluation value of each of the target extracted features includes: The sum of all the first evaluation values ​​is taken as the second evaluation value; And / or, the battery pack status data includes the target fault level of the battery pack, and determining the battery pack status data using the second evaluation value includes: Determine the target range to which the second evaluation value belongs; The fault level corresponding to the target interval is determined from the preset mapping relationship and used as the target fault level. The preset mapping relationship includes a number of preset intervals and preset fault levels corresponding to each preset interval. And / or, the battery pack status data includes the target fault level of the battery pack, and the method further includes: Determine the highest evaluation value among all the first evaluation values; The target fault cause of the battery pack is determined by using the target extracted features corresponding to the highest evaluation value and the target fault level.

9. The method according to claim 1, characterized in that, The step of acquiring target battery propagation data regarding the electromagnetic wave signal through a signal receiver includes: The signal receiving end acquires first propagation data about the electromagnetic wave signal, wherein the first propagation data includes signal parameters of a plurality of first signals; The first propagation data is divided into multiple second propagation data with different propagation medium types; Propagation data whose propagation medium type is a battery cell is identified from the plurality of second propagation data and used as the target battery propagation data.

10. The method according to claim 9, characterized in that, For each of the second propagated data, the propagation medium type of the second propagated data is determined by using at least one of the first result of the first method and the second result of the second method; In the first method: first extracted feature data of the second propagation data is determined, the first extracted feature data including a plurality of first extracted features; From the first sample dataset, a plurality of second sample extracted feature data similar to the first extracted feature data are determined, wherein the first sample dataset includes a plurality of first sample extracted feature data and a sample type label corresponding to each of the first sample extracted feature data, wherein the sample type label is a battery cell or a non-battery cell; at least the sample type label of each of the second sample extracted feature data is used as the first result; In the second method: determine the second extracted feature data of the second propagation data, the second extracted feature data including multiple second extracted features; use a prediction model to predict the second extracted feature data to obtain the second result, the second result including the prediction probability corresponding to the prediction type label of the second propagation data, the prediction type label being a battery cell or a non-battery cell.

11. The method according to claim 10, characterized in that, Using the first result and the second result, the propagation medium type of the second propagation data is determined, including: Using the first result, determine whether a first condition is met; wherein the first condition is any of the following: the proportion of samples labeled as battery cells in the first result is greater than or equal to a proportion threshold; the number of samples labeled as battery cells in the first result is greater than or equal to a number threshold; a first weight is greater than a second weight, wherein the first weight is obtained by using the similarity value between the extracted feature data of each second sample labeled as battery cell and the first extracted feature data, and the second weight is obtained by using the similarity value between the extracted feature data of each second sample labeled as non-battery cell and the first extracted feature data; Using the second result, determine whether the second condition is met; wherein, the second condition includes: the predicted type label is the battery cell and the predicted probability is greater than or equal to the first probability threshold; In response to the satisfaction of the first condition and the second condition, the propagation medium type of the second propagation data is determined to be the battery cell.

12. The method according to claim 1, characterized in that, The electromagnetic wave signal carries first identification information, and the step of obtaining target battery propagation data about the electromagnetic wave signal through the signal receiving end includes: The initial propagation data of the electromagnetic wave signal is obtained through the signal receiving end, wherein the initial propagation data includes signal parameters of several initial signals; Verify whether the second identification information carried in each of the initial signals is consistent with the first identification information; The first propagation data is obtained by removing the signal parameters of the initial signal whose second identification information is inconsistent with the first identification information from the initial propagation data; The target battery propagation data is determined from the first propagation data; And / or, the battery pack status data includes the target health value of the battery pack, and the method further includes: By using multiple historical health values ​​of the battery pack and the historical time corresponding to each historical health value, a health prediction curve is obtained through fitting. The health prediction curve is used to determine the time information corresponding to the preset health value.

13. An electronic device, characterized in that, Including interconnected memory and processor, The memory stores program instructions; The processor is used to execute program instructions stored in the memory to implement the method according to any one of claims 1-12.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store program instructions that can be executed by a processor to implement the method of any one of claims 1-12.

Images