A method for constructing a power battery health index and related devices

Abstract

The application discloses a power battery health index construction method and related device, relates to the new energy locomotive power battery health management technical field, obtains each multi-source state data in the power battery operation process and each multi-source state data respectively corresponding sampling time, carries out time continuity discrimination to each multi-source state data according to the sampling time, determines the effective data section of the sampling time continuity, in the effective data section, the dynamic characteristics of the multi-source state data changing with the sampling time are extracted respectively for each sampling time, the multi-dimensional health factors corresponding to each sampling time are constructed according to the dynamic characteristics, and the multi-dimensional health factors corresponding to each sampling time are fused processing time by time, and the health index corresponding to each sampling time is generated. The application realizes the continuous, accurate quantitative representation of the health state of the power battery by running continuity perception and dynamic health characteristic fusion, and eliminates the interference introduced by data discontinuity.

Description

Technical Field

[0001] This application relates to the field of health management technology for power batteries of new energy locomotives, and in particular to a method for constructing a health index for power batteries and related devices. Background Technology

[0002] With the rapid development of new energy locomotives, the operating status of power batteries, as the core power component of vehicles, directly affects the operational safety, reliability, and intelligent maintenance level of locomotives. Due to the complex operating environment of new energy locomotives, power batteries are subject to the combined effects of various factors such as changes in traction load, charge-discharge cycles, ambient temperature variations, and differences in internal individual cells, leading to a gradual degradation in their performance. Therefore, accurately acquiring and quantifying the health status of power batteries is a key technological foundation for achieving intelligent health management of power batteries.

[0003] Currently, methods for assessing the health status of power batteries typically extract a segment of data from the historical data stored in the vehicle's Battery Management System (BMS), taking the current moment as the endpoint, and directly using this extracted BMS data for health analysis to obtain the battery health status assessment result at the current moment. For example, methods based on supervised learning models use the extracted BMS data as model input and utilize laboratory tests or manually labeled health status tags as training objectives to train a supervised learning model that can be used for battery health analysis.

[0004] However, the historical data stored in the BMS may not be continuous over time. Due to factors such as vehicle downtime, communication interruptions, and changes in sampling periods during vehicle operation, the extracted data segments may contain data points from different operating stages. Directly using such data for health analysis can easily correlate data changes between different operating stages with the battery's own performance degradation trend, causing the health analysis results to deviate from the battery's true degradation trend. Summary of the Invention

[0005] In view of the above problems, this application provides a method and related apparatus for constructing a health index for a power battery. By judging the sampling continuity of multi-source state data, it determines the effective data segments with continuous sampling time, extracts dynamic features within the effective data segments to construct multi-dimensional health factors, and fuses them to generate a health index. This eliminates the interference introduced by data discontinuity, enabling the health analysis results to truly reflect the battery's performance degradation trend. The specific solution is as follows:

[0006] The first aspect of this application provides a method for constructing a power battery health index, including:

[0007] Acquire the multi-source state data during the operation of the power battery and the sampling time corresponding to each multi-source state data;

[0008] Based on the sampling time, the time continuity of each multi-source state data is determined to identify a valid data segment with continuous sampling time. The valid data segment contains multi-source state data with multiple sampling times.

[0009] Within the effective data segment, dynamic features of the multi-source state data changing with sampling time are extracted for each sampling time. These dynamic features are used to characterize the health degradation trend of the power battery.

[0010] Based on the dynamic characteristics, multidimensional health factors corresponding to each sampling time are constructed, and the multidimensional health factors corresponding to each sampling time are fused over a sampling time-by-sampling time to generate a health index corresponding to each sampling time. The health index is used to characterize the health status of the power battery.

[0011] In one possible implementation, the step of determining the temporal continuity of the multi-source state data based on the sampling time to identify valid data segments with continuous sampling time includes:

[0012] Calculate the time interval between adjacent sampling times in the sampling time corresponding to each of the multi-source state data;

[0013] If the time interval is greater than the preset interval threshold, then the multi-source state data corresponding to the adjacent sampling times are determined to belong to different valid data segments.

[0014] If the time interval is less than or equal to the interval threshold, then the multi-source state data corresponding to the adjacent sampling times are determined to belong to the same valid data segment.

[0015] To obtain the effective data segment with continuous sampling time.

[0016] In one possible implementation, extracting the dynamic features of the multi-source state data changing with sampling time for each sampling time within the effective data segment includes:

[0017] For each sampling time within the valid data segment:

[0018] Calculate the rate of change characteristic between the multi-source state data at the current sampling time and the multi-source state data at the previous sampling time. If the current sampling time is the first sampling time within the valid data segment, the rate of change characteristic is omitted.

[0019] A preset time window is extracted forward from the sampling time, and the variance or standard deviation of the multi-source state data within the preset time window is calculated as the fluctuation feature corresponding to the sampling time. If the time length of the multi-source state data forward from the sampling time is insufficient to constitute the preset time window, the fluctuation feature is omitted.

[0020] Calculate the voltage difference between the highest and lowest voltages of individual cells in the multi-source state data at that sampling time, and use it as the voltage consistency feature corresponding to that sampling time.

[0021] Calculate the temperature difference between the highest and lowest temperatures of individual cells in the multi-source state data at that sampling time, and use it as the temperature consistency feature corresponding to that sampling time.

[0022] Calculate the power of the total battery pack voltage and total battery pack current in the multi-source state data at that sampling time, and use it as the power feature corresponding to that sampling time.

[0023] In one possible implementation, the rate of change feature includes at least one of the following: total battery pack voltage rate of change, total battery pack current rate of change, state of charge rate of change, maximum temperature rate of change of individual cells, and minimum temperature rate of change of individual cells.

[0024] The fluctuation characteristics include at least one of the following: total battery pack voltage fluctuation, total battery pack current fluctuation, state of charge fluctuation, maximum temperature fluctuation of individual cells, and minimum temperature fluctuation of individual cells.

[0025] In one possible implementation, the multidimensional health factors include at least two of the following: voltage health factor, temperature health factor, current health factor, state of charge health factor, and consistency health factor.

[0026] The voltage health factor is constructed based on the total voltage change rate and the total voltage fluctuation of the battery pack, and is used to evaluate the voltage stability and voltage degradation trend of the power battery.

[0027] The temperature health factor is constructed based on the highest temperature change rate of the single cell, the lowest temperature change rate of the single cell, the highest temperature fluctuation of the single cell, and the lowest temperature fluctuation of the single cell, and is used to evaluate the thermal stability and temperature anomaly degree of the power battery.

[0028] The current health factor is constructed based on the total current change rate and the total current fluctuation of the battery pack, and is used to evaluate the load response capability and current stability of the power battery.

[0029] The state of charge health factor is constructed based on the state of charge change rate, the state of charge fluctuation and the power characteristics, and is used to evaluate the energy change stability and state of charge estimation reliability of the power battery.

[0030] The consistency health factor is constructed based on the voltage consistency characteristic and the temperature consistency characteristic, and is used to evaluate the voltage consistency and temperature consistency of the power battery.

[0031] In one possible implementation, the step of performing time-by-time fusion processing on the multidimensional health factors corresponding to each sampling time includes:

[0032] For each sampling time:

[0033] The multidimensional health factors corresponding to the sampling time were normalized to obtain the normalized multidimensional health factors.

[0034] Calculate the change sensitivity corresponding to each of the normalized multidimensional health factors, wherein the change sensitivity is determined based on the change rate of the corresponding normalized multidimensional health factors in each sampling time period;

[0035] Based on the change sensitivity of the normalized multidimensional health factors, the fusion weights of the normalized multidimensional health factors are determined.

[0036] The normalized multidimensional health factors are weighted and fused based on the fusion weights to obtain the health index corresponding to the sampling time.

[0037] One possible implementation also includes:

[0038] The health index corresponding to each sampling time is continuously smoothed and corrected to obtain the corrected health index corresponding to each sampling time.

[0039] Based on the pre-constructed mapping relationship between health indices and health level labels, the corrected health indices corresponding to each sampling time are mapped to the corresponding health level labels; thus obtaining the health level labels corresponding to each sampling time.

[0040] The second aspect of this application provides a computer program product, including computer-readable instructions, which, when executed on an electronic device, cause the electronic device to implement the power battery health index construction method described in the first aspect or any implementation thereof.

[0041] A third aspect of this application provides an electronic device, comprising at least one processor and a memory connected to the processor, wherein:

[0042] The memory is used to store computer programs;

[0043] The processor is used to execute the computer program so that the electronic device can implement the power battery health index construction method of the first aspect or any implementation thereof.

[0044] The fourth aspect of this application provides a computer storage medium carrying one or more computer programs, which, when executed by an electronic device, enable the electronic device to implement the power battery health index construction method described in the first aspect or any implementation thereof.

[0045] By employing the above technical solution, the power battery health index construction method provided in this application acquires multi-source state data and the sampling time corresponding to each multi-source state data during the operation of the power battery. It then determines the effective data segment with continuous sampling time by judging the continuity of the sampling time, thereby eliminating discontinuous data points caused by vehicle shutdown, communication interruption, or changes in the sampling period. This ensures that all multi-source state data within the effective data segment originates from the same continuous operating process, eliminating interference introduced by data mixing across operating stages from the data source. Furthermore, since the sampling times within the effective data segment are continuously arranged on the time axis, when extracting the dynamic features of multi-source state data changing with sampling time for each sampling time within the effective data segment, the extracted dynamic features only reflect the same... The dynamic features accurately represent the real changes in battery state parameters over time during continuous operation, without data jumps across data segments being included in feature calculations. Therefore, the dynamic features can truly characterize the health degradation trend of the power battery, rather than the non-degradation differences between different operating stages. Furthermore, based on the aforementioned dynamic features, multi-dimensional health factors corresponding to each sampling time are constructed, and these multi-dimensional health factors are fused time-by-time. This allows each sampling time point to independently generate a health index that integrates multiple health dimensions, avoiding the problem of one-sided representation by a single health factor. The resulting health indices corresponding to each sampling time can continuously and accurately reflect the trajectory of the power battery's health status changes during operation, effectively improving the accuracy of power battery health status assessment. Attached Figure Description

[0046] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale.

[0047] Figure 1 A schematic diagram of a system architecture provided for this application;

[0048] Figure 2 A flowchart illustrating a method for constructing a power battery health index provided in this application;

[0049] Figure 3 This is a schematic diagram illustrating the process of extracting dynamic features from a single sampling time.

[0050] Figure 4 This is a schematic diagram illustrating the process of fusing multidimensional health factors corresponding to a single sampling time.

[0051] Figure 5 A schematic diagram of a power battery health index construction device provided in this application;

[0052] Figure 6 This is a schematic diagram of the structure of an electronic device provided in this application. Detailed Implementation

[0053] The embodiments of this application are described below with reference to the accompanying drawings. The terminology used in the implementation section of this application is for explaining specific embodiments only and is not intended to limit the scope of this application.

[0054] As will be known to those skilled in the art, with the development of technology and the emergence of new scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0055] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate; this is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of elements is not necessarily limited to those elements, but may include other elements not explicitly listed or inherent to those processes, methods, products, or apparatuses.

[0056] This application provides a method and related device for constructing a power battery health index, which can be applied to scenarios such as online health status assessment, battery anomaly detection, remaining life prediction, and predictive maintenance decision-making for power batteries in new energy locomotives under actual operating conditions.

[0057] It should be noted that the above scenarios are merely examples and are not intended to limit this application.

[0058] Optionally, the method for constructing the power battery health index provided in this application can be applied to, for example... Figure 1The system architecture shown includes a terminal 100 and a server 200. The server 200 may include one or more servers (…). Figure 1 (This example uses a server as an illustration).

[0059] Either terminal 100 or server 200 can be used independently to execute the power battery health index construction method provided in this application embodiment. Alternatively, terminal 100 and server 200 can also work together to execute the power battery health index construction method provided in this application embodiment. For example, terminal 100 can acquire multi-source state data and corresponding sampling times during power battery operation, and determine the effective data segment based on the time continuity of the sampling time; server 200 can further extract dynamic features, construct multi-dimensional health factors, and fuse them to generate a health index, thereby reducing the computational burden on terminal 100.

[0060] The following description Figure 1 The product form of the terminal 100; In this application embodiment, the terminal 100 can be a vehicle terminal, a battery management system (BMS), a vehicle T-BOX (Telematics BOX), a train control unit, or a locomotive on-board intelligent monitoring device, etc., which have data acquisition and computing capabilities. It can also be a mobile phone, tablet computer, wearable device, vehicle device, augmented reality (AR) / virtual reality (VR) device, laptop computer, ultra-mobile personal computer (UMPC), netbook, personal digital assistant (PDA), etc. This application embodiment does not impose any restrictions on this.

[0061] To enable those skilled in the art to better understand this application, the method for constructing the power battery health index according to embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0062] Reference Figure 2 , Figure 2 This is a flowchart illustrating a method for constructing a power battery health index according to an embodiment of this application, as shown below. Figure 2 As shown, the method for constructing the power battery health index may include:

[0063] Step S201: Obtain the multi-source state data and the sampling time corresponding to each multi-source state data during the operation of the power battery.

[0064] Here, the power battery consists of a group of cells; multi-source state data refers to monitoring parameters collected and recorded by the BMS during the operation of the power battery, reflecting the battery's operating status in various dimensions. In the actual operation scenario of new energy vehicles, the power battery is subjected to the coupled effects of multiple factors such as changes in traction load, charge-discharge cycles, changes in ambient temperature, and differences between individual cells. A single parameter is insufficient to comprehensively reflect the health status of the power battery. Therefore, the multi-source state data obtained in this embodiment includes, but is not limited to: the highest voltage of a single cell (i.e., the maximum voltage among all the individual cells contained in the power battery). The minimum voltage of a single cell (i.e., the minimum voltage among all the individual cells in a power battery). Total voltage of battery pack Total current of battery pack Battery pack state of charge (SOC) and maximum temperature of individual cells (i.e., the maximum temperature among all individual cells in the power battery). And the lowest temperature of a single cell (i.e., the minimum temperature among all the individual cells in a power battery). There are seven categories of BMS data.

[0065] Sampling time refers to the timestamp information corresponding to each set of multi-source state data collected by the BMS (in this embodiment, a set of multi-source state data may include the seven types of BMS data mentioned above). Specifically, the BMS periodically collects battery state data according to a certain sampling period, and each set of multi-source state data carries a unique timestamp to identify the collection time of that set of multi-source state data.

[0066] For example, the current moment can be taken as the endpoint, and a segment of historical multi-source state data can be extracted from the BMS storage history according to a preset time window or a preset number of multi-source state data groups, which can be used as the multi-source state data obtained in this embodiment.

[0067] Step S202: Determine the time continuity of each multi-source state data according to the sampling time, and determine the effective data segment with continuous sampling time. The effective data segment contains multi-source state data with multiple sampling times.

[0068] Those skilled in the art should understand that new energy vehicles may experience vehicle shutdowns, communication interruptions, and changes in sampling periods during actual operation. This can lead to discontinuities in the historical multi-source status data stored in the BMS over time. If data spanning multiple interruption points is mixed together, the voltage state differences before and after shutdown may be misinterpreted as battery degradation trends. To avoid this problem, this embodiment performs time continuity determination on the multi-source status data obtained above based on the sampling time, obtaining valid data segments with continuous sampling time. Here, continuous sampling time means that the sampling time interval between two adjacent sets of multi-source status data does not exceed a preset interval threshold; a valid data segment refers to a group of multi-source status data with continuous adjacent sampling times and no data interruptions within the segment.

[0069] Based on this, the optional process of "determining the effective data segments with continuous sampling time by judging the temporal continuity of each multi-source state data according to the sampling time" may include: calculating the time interval between adjacent sampling times in the sampling time corresponding to each multi-source state data; if the time interval is greater than a preset interval threshold, then the multi-source state data corresponding to adjacent sampling times are determined to belong to different effective data segments; conversely, if the time interval is less than or equal to the interval threshold, then the multi-source state data corresponding to adjacent sampling times belong to the same effective data segment; by judging all time intervals, the effective data segments with continuous sampling time can be obtained.

[0070] For example, suppose the normal sampling period of the BMS is 1 second, and the preset interval threshold is 3 seconds. Ten sets of multi-source state data were acquired sequentially, with adjacent sampling time intervals of 1.1s, 0.9s, 1.0s, 2.5s, 120s, 1.0s, 0.8s, 1.2s, and 0.9s, respectively. It can be seen that the first four sampling time intervals are all no greater than 3 seconds, indicating that the sampling time of the first to fifth sets of multi-source state data is continuous and belongs to the same valid data segment. The fifth sampling time interval is 120s, which is greater than 3s, indicating that a data interruption occurred here. Therefore, the fifth set of multi-source state data is the end of the previous segment, and the sixth set is the beginning of the next segment. All sampling time intervals after the fifth sampling time interval are no greater than 3 seconds, indicating that the sampling time of the sixth to tenth sets of data is continuous and belongs to another valid data segment. Thus, the ten sets of multi-source state data are divided into two valid data segments.

[0071] In this way, the multi-source state data in the effective data segment all come from the same running process with continuous sampling time. Data mixing across running stages is excluded from the effective data segment, eliminating the interference introduced by cross-interruption point calculations during subsequent feature extraction from the data source.

[0072] Step S203: Within the effective data segment, extract the dynamic features of the multi-source state data as a function of sampling time for each sampling time. The dynamic features are used to characterize the health degradation trend of the power battery.

[0073] The dynamic characteristics in this embodiment differ from static statistical indicators such as cumulative charge-discharge cycles and total runtime. They refer to the temporal characteristics of multi-source state data changing with sampling time in a continuous time series, including the rate of change of parameters, the degree of fluctuation, and the consistency between individual data.

[0074] Since the data within the valid data segment all originate from the same continuous operating process with consecutive sampling times, this step only extracts dynamic features within the valid data segment. Specifically, a set of dynamic features is extracted for each sampling time within the valid data segment. This set of dynamic features is calculated using only adjacent multi-source state data within the same valid data segment; multi-source state data across data segments are not included in the calculation. Because cross-segment data is excluded, the calculated features do not incorporate non-degradation differences between different operating stages. Therefore, the extracted dynamic features can accurately reflect the health degradation trend of the power battery.

[0075] Step S204: Construct multidimensional health factors corresponding to each sampling time based on dynamic characteristics, and perform time-by-time fusion processing on the multidimensional health factors corresponding to each sampling time to generate health indices corresponding to each sampling time. The health indices are used to characterize the health status of the power battery.

[0076] It should be understood that battery health degradation is the result of the coupling effect of multiple factors. Therefore, it is necessary to construct health factors from multiple dimensions based on dynamic characteristics. These multidimensional health factors refer to comprehensive indicators constructed based on different types of dynamic characteristics to evaluate the battery health status from different dimensions.

[0077] Optionally, the multidimensional health factors in this embodiment include at least two of the following: voltage health factor, temperature health factor, current health factor, state of charge (SOC) health factor, and consistency health factor. Each health factor characterizes the health status of the power battery from different dimensions. Specifically, the voltage health factor is used to evaluate the voltage stability and voltage degradation trend of the power battery; the temperature health factor is used to evaluate the thermal stability and temperature anomaly degree of the power battery; the current health factor is used to evaluate the load response capability and current stability of the power battery; the SOC health factor is used to evaluate the energy change stability and SOC estimation reliability of the power battery; and the consistency health factor is used to evaluate the voltage consistency and temperature consistency of the power battery.

[0078] After constructing the multidimensional health factors for each sampling time, this embodiment can perform time-by-time fusion processing on the multidimensional health factors corresponding to each sampling time to generate a health index corresponding to each sampling time. Here, time-by-time fusion processing means... Health index at any time is determined by Health is determined by multiple factors at any given moment.

[0079] The aforementioned health index is a dimensionless quantitative indicator; a higher value indicates a better battery health, while a lower value indicates more severe battery degradation. Since the health index is generated over time, the resulting health index sequence continuously reflects the trajectory of the battery's health status throughout its operation.

[0080] In summary, the method for constructing a power battery health index provided in this application acquires multi-source state data and the corresponding sampling time for each multi-source state data during the operation of the power battery. It then determines the effective data segments with continuous sampling time by judging the continuity of the sampling time, thereby eliminating discontinuous data points caused by vehicle shutdown, communication interruption, or changes in the sampling period. This ensures that the multi-source state data within the effective data segment all originate from the same continuous operating process, eliminating interference introduced by data mixing across operating stages from the data source. Furthermore, since the sampling times within the effective data segment are arranged continuously on the time axis, when extracting the dynamic features of the multi-source state data changing with sampling time for each sampling time within the effective data segment, the extracted dynamic features only reflect the same continuous... The actual changes in battery state parameters over time during operation are captured, and data jumps across data segments are not included in feature calculations. Therefore, dynamic features can accurately represent the health degradation trend of the power battery, rather than the non-degradation differences between different operating stages. Furthermore, based on the above dynamic features, multi-dimensional health factors corresponding to each sampling time are constructed, and the multi-dimensional health factors corresponding to each sampling time are fused time-by-time. This allows each sampling time point to independently generate a health index that integrates multiple health dimensions, avoiding the problem of one-sided representation by a single health factor. The resulting health index corresponding to each sampling time can continuously and accurately reflect the trajectory of the power battery's health status changes during operation, effectively improving the accuracy of power battery health status assessment.

[0081] In some embodiments of this application, the process of "extracting dynamic features of multi-source state data as a function of sampling time for each sampling time within the effective data segment" in step S203 above is described.

[0082] For ease of description, the following section uses a single sampling time within the valid data segment as an example to illustrate the process of extracting dynamic features for that sampling time. The extraction process for dynamic features at other sampling times is the same and will not be repeated here. See also Figure 3 This diagram illustrates the process of extracting dynamic features from a single sampling time.

[0083] Step S301: Calculate the rate of change characteristics between the multi-source state data at the current sampling time and the multi-source state data at the previous sampling time.

[0084] Wherein, if the sampling time is the first sampling time within the effective data segment, the rate of change feature is omitted.

[0085] Here, the rate of change feature is used to characterize the rate of change of multi-source state data between adjacent sampling times. Specifically, for this sampling time... Multi-source state data Its time with the previous sampling time The rate of change characteristics among the multi-source state data are as follows:

[0086] Formula (1);

[0087] in, Indicates sampling time The corresponding rate of change characteristics, Indicates the previous sampling time The corresponding multi-source state data, Indicates sampling time Compared with the previous sampling time The difference between them (i.e., the sampling time interval).

[0088] Since the sampling times within the valid data segment are continuous, the difference between adjacent sampling times reflects the continuous change in battery state during the same operation. Therefore, the calculated rate of change feature can accurately reflect the rate of change and trend of multi-source state data. If the sampling time is the first sampling time within the valid data segment, there is no previous sampling time. To avoid using default values ​​to replace the multi-source state data of the previous sampling time, which would lead to inaccurate rate of change features, this embodiment allows the rate of change feature of this sampling time to be omitted. For sampling times with default dynamic features, the health index can be omitted.

[0089] Optionally, the rate of change features in this embodiment include at least one of the following: total battery pack voltage change rate, total battery pack current change rate, state of charge change rate, maximum temperature change rate of individual cells, and minimum temperature change rate of individual cells. Specifically, the total battery pack voltage change rate is the difference between the total battery pack voltage at the current sampling time and the previous sampling time divided by the sampling time interval, as detailed in formula (2); the total battery pack current change rate is the difference between the total battery pack current at the current sampling time and the previous sampling time divided by the sampling time interval, as detailed in formula (3); the state of charge change rate is the difference between the state of charge at the current sampling time and the previous sampling time divided by the sampling time interval, as detailed in formula (4); the maximum temperature change rate of individual cells is the difference between the maximum temperature of individual cells at the current sampling time and the previous sampling time divided by the sampling time interval, as detailed in formula (5); and the minimum temperature change rate of individual cells is the difference between the minimum temperature of individual cells at the current sampling time and the previous sampling time divided by the sampling time interval, as detailed in formula (6).

[0090] Formula (2);

[0091] Formula (3);

[0092] Formula (4);

[0093] Formula (5);

[0094] Formula (6);

[0095] in, Indicates sampling time The corresponding total voltage change rate of the battery pack, Indicates sampling time The corresponding total voltage of the battery pack, Indicates the previous sampling time The corresponding total voltage of the battery pack; Indicates sampling time The corresponding rate of change of total battery pack current. Indicates sampling time The corresponding total current of the battery pack, Indicates the previous sampling time The corresponding total current of the battery pack; Indicates sampling time The corresponding rate of change of state of charge, Indicates sampling time The corresponding state of charge of the battery pack, Indicates the previous sampling time The corresponding state of charge of the battery pack; Indicates sampling time The corresponding highest temperature change rate of a single cell. Indicates sampling time The corresponding highest temperature of a single cell, Indicates the previous sampling time The corresponding highest temperature of a single cell; Indicates sampling time The corresponding lowest temperature change rate of a single cell, Indicates sampling time The corresponding lowest temperature of a single cell, Indicates the previous sampling time The corresponding lowest temperature of a single cell.

[0096] Step S302: Take the sampling time as the end point and extract a preset time window forward. Calculate the variance or standard deviation of the multi-source state data within the preset time window as the fluctuation characteristics corresponding to the sampling time.

[0097] If the duration of the multi-source state data (which must be multi-source state data within the same valid data segment) preceding the sampling time is insufficient to constitute a preset time window, then the fluctuation characteristics will be omitted.

[0098] Here, fluctuation characteristics are used to characterize the degree of fluctuation in multi-source state data over a continuous period of time. Specifically, this sampling time can be... As the endpoint, cut forward by a preset length. A preset time window is defined, and then the variance or standard deviation of the multi-source state data within the preset time window is calculated as the fluctuation characteristic corresponding to that sampling time. Taking the standard deviation as an example, the calculation formula for this fluctuation characteristic is as follows:

[0099] Formula (7);

[0100] in, Indicates sampling time The corresponding fluctuation characteristics, Indicates sampling time The corresponding multi-source state data, This represents the function for calculating standard deviation.

[0101] Since the window capture range is limited to the same valid data segment, all data points within the window come from the same continuous running process. Therefore, the calculated fluctuation characteristics reflect the stability of multi-source state data during the same running process.

[0102] Optionally, the aforementioned fluctuation characteristics include at least one of the following: total battery pack voltage fluctuation, total battery pack current fluctuation, state of charge fluctuation, maximum temperature fluctuation of individual cells, and minimum temperature fluctuation of individual cells. The specific calculation process can be referred to formula (7) above; only the standard deviation calculation function needs to be adjusted. Simply replace it with the corresponding multi-source state data. Specifically, when calculating the total voltage fluctuation of the battery pack, use... Replace it with the total battery pack voltage when calculating the total battery pack current fluctuation. Replace it with the total battery pack current when calculating state-of-charge fluctuations. Replace it with the battery pack's state of charge, and when calculating the maximum temperature fluctuation of a single cell, use... Replace it with the highest temperature of the individual cell; when calculating the minimum temperature fluctuation of the individual cell, use... Replace with the lowest temperature of a single cell.

[0103] Step S303: Calculate the voltage difference between the highest voltage and the lowest voltage of a single cell in the multi-source state data at that sampling time, and use it as the voltage consistency feature corresponding to that sampling time.

[0104] Here, the voltage consistency characteristic is used to characterize the degree of voltage difference between individual cells in a power battery. The larger the value of the voltage consistency characteristic, the greater the voltage difference between individual cells in the battery pack, indicating that some individual cells in the power battery may have problems such as aging, increased internal resistance, or insufficient equalization capability.

[0105] Optionally, the voltage consistency characteristics can be calculated using the following formula: ,in, Indicates sampling time The corresponding voltage consistency characteristics, Indicates sampling time The corresponding highest voltage of a single cell, Indicates sampling time The corresponding minimum voltage of a single cell.

[0106] Step S304: Calculate the temperature difference between the highest and lowest temperatures of individual cells in the multi-source state data at the sampling time, and use it as the temperature consistency feature corresponding to the sampling time.

[0107] Here, the temperature uniformity characteristic is used to characterize the degree of temperature difference between individual cells in the power battery pack. The larger the value of the temperature uniformity characteristic, the greater the temperature difference between individual cells in the battery pack, indicating that some individual cells in the power battery may have uneven heat dissipation or excessive internal resistance leading to severe overheating.

[0108] Optionally, the temperature uniformity characteristic can be calculated using the following formula: ,in, Indicates sampling time The corresponding temperature uniformity characteristics, Indicates sampling time The corresponding highest temperature of a single cell, Indicates sampling time The corresponding lowest temperature of a single cell.

[0109] Step S305: Calculate the power of the total battery pack voltage and total battery pack current in the multi-source state data at the sampling time, and use it as the power feature corresponding to the sampling time.

[0110] Among them, the power characteristic is used to characterize the instantaneous output power of the power battery during operation, reflecting the load output capability of the power battery. Its changing trend can indirectly reflect the changes in battery internal resistance and the degradation of output capability.

[0111] Optionally, the power characteristics can be calculated using the following formula: ,in, Indicates sampling time The corresponding power characteristics, Indicates sampling time The corresponding total voltage of the battery pack, Indicates sampling time The corresponding total current of the battery pack.

[0112] In one possible implementation, in order to better store the above dynamic features and facilitate subsequent processing, this embodiment can combine the above 13 dynamic features in chronological order into the following dynamic health feature vector.

[0113] Formula (8);

[0114] Where H represents the dynamic health feature vector, m takes the value of 13, and n is the number of groups of multi-source state data contained in the effective data segment. Each row in H represents 13 dynamic features for one sampling time, such as... This represents the rate of change of the total battery pack voltage corresponding to the first sampling time within the valid data segment. This represents the rate of change of the total battery pack current corresponding to the first sampling time within the valid data segment, and so on. This represents the power characteristic corresponding to the first sampling time within the valid data segment.

[0115] As can be seen, in the above embodiment, the calculation of the rate of change feature is limited to adjacent sampling times within the effective data segment, the calculation of the fluctuation feature is limited to a preset time window within the effective data segment, and the voltage consistency feature, temperature consistency feature, and power feature are calculated based on multi-source state data at the same sampling time within the effective data segment. Since each step of the dynamic feature extraction is based on the effective data segment as the boundary, the false feature values ​​introduced by calculations across interruption points are effectively avoided. Therefore, the extracted dynamic features can truly reflect the performance changes of the battery during continuous operation.

[0116] As mentioned earlier, dynamic features are used to construct multidimensional health factors corresponding to each sampling time. However, battery health degradation is the result of the coupling effect of multiple factors, and a single dynamic feature is insufficient to comprehensively characterize the battery's health status. Therefore, this embodiment constructs multiple health factors based on different types of dynamic features, with each health factor evaluating the battery's health status from different dimensions.

[0117] In one possible implementation, the voltage health factor is constructed based on the total voltage change rate and total voltage fluctuation of the battery pack; the temperature health factor is constructed based on the highest temperature change rate, lowest temperature change rate, highest temperature fluctuation, and lowest temperature fluctuation of individual cells; the current health factor is constructed based on the total current change rate and total current fluctuation of the battery pack; the state of charge health factor is constructed based on the state of charge change rate, state of charge fluctuation, and power characteristics; and the consistency health factor is constructed based on voltage consistency characteristics and temperature consistency characteristics.

[0118] In one possible implementation, the voltage health factor can be obtained by performing a function calculation on the total voltage change rate and the total voltage fluctuation of the battery pack.

[0119] Considering that the total voltage change rate and the total voltage fluctuation of the battery pack have different dimensions and significantly different value ranges, if direct function calculations (such as weighted summation, multiplication, and division) are performed, the larger dimension will dominate the calculation result, suppressing the characterization effect of the other and making it difficult to reasonably reflect the actual health status of the battery. Therefore, in a preferred implementation, instead of performing function calculations on both, the total voltage change rate and the total voltage fluctuation of the battery pack are treated as two dimensions of the voltage health factor, namely the voltage dynamic health factor and the voltage steady-state health factor, which together constitute the voltage health factor corresponding to the sampling time.

[0120] Similarly, temperature health factors, current health factors, state of charge (SOC) health factors, and consistency health factors can be constructed in the same way, that is, the dynamic characteristics on which they are based are directly used as components of the corresponding health factors, without performing cross-dimensional function mapping, thereby avoiding the bias introduced by dimensional differences. That is, the highest temperature change rate, lowest temperature change rate, highest temperature fluctuation, and lowest temperature fluctuation of a single cell are all used as temperature health factors; the total current change rate and total current fluctuation of the battery pack are all used as current health factors; the rate of change of SOC, SOC fluctuation, and power characteristics are all used as SOC health factors; and voltage consistency characteristics and temperature consistency characteristics are all used as consistency health factors.

[0121] In this embodiment, each health factor retains all the dynamic characteristics upon which it is based, avoiding information loss and calculation bias caused by cross-dimensional function mapping. This allows each health factor to more realistically and comprehensively reflect the health status of its corresponding dimension. For example, the voltage health factor simultaneously includes voltage change trend information (voltage change rate) and voltage stability information (voltage fluctuation). These two aspects characterize the health characteristics of the voltage dimension from different perspectives, without interfering with each other and complementing each other. Furthermore, since each health factor is constructed from multi-source state data of the same continuous operating process within the same valid data segment, the health status information of each dimension remains aligned on the time axis, providing a complete and time-consistent multi-dimensional health factor input for subsequent sampling-time fusion processing. The independent retention of each dimension within each health factor provides finer-grained health information for subsequent fusion processing, ensuring that the information of each dimension is fully preserved before fusion. The resulting health index corresponding to each sampling time can more accurately and precisely reflect the overall health status of the power battery.

[0122] The following embodiment describes the process of "fusing the multidimensional health factors corresponding to each sampling time in step S204 above" in order to "fuse the data time-by-sampling time".

[0123] This embodiment still uses the fusion processing of a single sampling time as an example for description. The fusion processing of other sampling times is the same, and will not be described in detail in this embodiment. See Figure 4 This diagram illustrates the process of fusing multidimensional health factors corresponding to a single sampling time.

[0124] For any sampling time The corresponding multidimensional health factor fusion process includes the following steps:

[0125] Step S401: Normalize the multidimensional health factors corresponding to the sampling time to obtain normalized multidimensional health factors.

[0126] Because different health factors have different dimensions and value ranges (for example, the voltage health factor may range from [0.7, 0.95], while the temperature health factor may range from [0.85, 0.98]), directly performing weighted fusion will result in health factors with larger value ranges dominating the fusion result, while health factors with smaller value ranges will be suppressed, thus failing to reasonably reflect the actual contribution of each health factor to the battery's health status. Therefore, normalization processing of each health factor is required before fusion.

[0127] Optionally, the normalization process can employ max-min normalization. Specifically, for the ... Health Factors Get its maximum value within its valid data segment. and minimum value Then calculate its sampling time according to the following formula. Normalized value:

[0128] Formula (9);

[0129] in, Indicates the first Health factors at sampling time The corresponding normalized value (hereinafter referred to as the first) (Normalized health factors).

[0130] For the preferred method described above, which considers the total voltage change rate and total voltage fluctuation of the battery pack as voltage health factors, the highest temperature change rate, lowest temperature change rate, highest temperature fluctuation, and lowest temperature fluctuation of individual cells as temperature health factors, the total current change rate and total current fluctuation of the battery pack as current health factors, the state of charge (SCC) change rate, SCC fluctuation, and power characteristics as SCC health factors, and voltage consistency characteristics and temperature consistency characteristics as consistency health factors, it is necessary to consider the sampling time. Substitute the corresponding battery pack total voltage change rate, battery pack total voltage fluctuation, single cell maximum temperature change rate, single cell minimum temperature change rate, single cell maximum temperature fluctuation, single cell minimum temperature fluctuation, battery pack total current change rate, battery pack total current fluctuation, state of charge change rate, state of charge fluctuation, power characteristics, voltage consistency characteristics, and temperature consistency characteristics into the following values: Perform normalized calculations.

[0131] After normalization, each health factor is mapped to a unified numerical range [0,1], eliminating dimensional differences and making each health factor comparable and integrable.

[0132] Step S402: Calculate the change sensitivity corresponding to the normalized multidimensional health factors. The change sensitivity is determined based on the change rate of the corresponding normalized health factors in each sampling time period.

[0133] Understandably, the rates of change for different health factors during battery degradation often vary. Some dimensions may show significant changes in the early stages of degradation, while others may only exhibit significant degradation in later stages. The more drastic the change in a dimension, the more significant the change in the corresponding battery state is. Therefore, dimensions with more drastic changes are more sensitive to changes in the overall battery health and should be given higher weight during fusion to enable the health index to more accurately capture the degradation signals of the power battery.

[0134] Based on this, this embodiment can calculate the change sensitivity corresponding to each of the normalized multidimensional health factors, so as to determine the fusion weights corresponding to each of the normalized multidimensional health factors in the next step according to the change sensitivity corresponding to each of the normalized multidimensional health factors.

[0135] Optionally, the sensitivity to change can be determined based on the rate of change of the normalized corresponding health factor at each sampling time. For the first... Normalized health factors Its sensitivity to change Calculate using the following formula:

[0136] Formula (10);

[0137] in, Indicates sampling time The corresponding number Normalized health factors Indicates the previous sampling time The corresponding number Normalized health factors Indicates the first The number of data points included in the normalized health factor.

[0138] For example, for a voltage health factor that includes the total voltage change rate and the total voltage fluctuation of the battery pack, both the total voltage change rate and the total voltage fluctuation of the battery pack need to be substituted into formula (10). Add them together, then use... Take the average value, at this time The value is twice the total number of sampling times included in the valid data segment; for temperature health factors that include the highest temperature change rate of a single cell, the lowest temperature change rate of a single cell, the highest temperature fluctuation of a single cell, and the lowest temperature fluctuation of a single cell, the highest temperature change rate of a single cell, the lowest temperature change rate of a single cell, the highest temperature fluctuation of a single cell, and the lowest temperature fluctuation of a single cell all need to be substituted into formula (10). Add them together, then use... Take the average value, at this time The value is four times the total number of sampling times contained in the valid data segment; and so on. Therefore, for the 5-dimensional health factor in this embodiment, regardless of how the 5-dimensional health factor is constructed, only 5 sensitivity values ​​can be calculated.

[0139] As shown in formula (10), the change sensitivity is the average of the absolute values ​​of the rate of change of the normalized health factor in the corresponding dimension within the effective data segment. The larger the value, the more drastic the change of the normalized health factor in the corresponding dimension over time, and the more sensitive it is to changes in battery state.

[0140] Step S403: Determine the fusion weights of the normalized multidimensional health factors based on their respective change sensitivities.

[0141] In this embodiment, the fusion weights are weighted coefficients determined based on the sensitivity of each health factor to changes. The fusion weights are positively correlated with the sensitivity to changes; the higher the sensitivity of a health factor, the greater its weight during fusion, and vice versa. This ensures that health dimensions that are more sensitive to changes in battery state play a greater role in the health index.

[0142] Optional, taking up to 5 health factors as an example, the first... Normalized fusion weights of health factors The calculation process is as follows:

[0143] Formula (11).

[0144] The sum of all fusion weights is 1 when calculated using formula (11).

[0145] Step S404: Based on the fusion weight, the normalized multidimensional health factors are weighted and fused to obtain the health index corresponding to the sampling time.

[0146] For example, the weighted summation of a 5-dimensional health factor constructed through function mapping can be performed in the following way (12):

[0147] Formula (12);

[0148] in, Indicates sampling time The corresponding health index Indicates sampling time The corresponding normalized voltage health factor, Indicates sampling time The corresponding normalized temperature health factor, Indicates sampling time The corresponding normalized current health factor, Indicates sampling time The corresponding normalized state of charge health factor, Indicates sampling time The corresponding normalized consistent health factors.

[0149] For the aforementioned optimal method that directly uses each dynamic feature as a health factor, the sampling time is considered during weighted summation. The corresponding normalized total voltage change rate and total voltage fluctuation of the battery pack are substituted into Sampling time The corresponding normalized rates of change of maximum temperature, minimum temperature, maximum temperature fluctuation, and minimum temperature fluctuation of individual cells are all substituted into... Sampling time The corresponding normalized rate of change of total battery pack current and the fluctuation of total battery pack current are substituted into Sampling time The corresponding normalized rate of change of state of charge, state of charge fluctuation, and power characteristics are all substituted into Sampling time The corresponding normalized voltage consistency characteristics and temperature consistency characteristics are substituted into This expands formula (12) into a weighted sum of 13 terms, yielding the sampling time. The corresponding health index.

[0150] This embodiment eliminates the differences in dimensions and value ranges between different health factors by performing maximum and minimum normalization on each dimension, making each dimension comparable and fusionable. On this basis, the fusion weight is adaptively determined according to the sensitivity of each health factor to change within the effective data segment, so that the dimensions with more dramatic changes play a greater role in the fusion, avoiding the deviation of the health index caused by unreasonable fixed weight settings. Finally, through weighted fusion over a sampling time, multi-dimensional health information is integrated, and the trajectory of the health index over time is completely preserved, effectively improving the accuracy and adaptability of the health index in representing the comprehensive health status of the power battery.

[0151] Because new energy locomotives experience sampling fluctuations, communication anomalies, and short-term state changes during operation, directly using instantaneous health indices may lead to misjudgments of health status. Therefore, this embodiment further incorporates continuous operation information to dynamically correct the fused health index, as detailed below.

[0152] First, in order to avoid sudden changes in the assessed battery health status due to instantaneous disturbances during the data acquisition process, this embodiment can perform continuous smooth correction on the health index corresponding to each sampling time to obtain the corrected health index corresponding to each sampling time.

[0153] Optionally, the process of continuously smoothing the health index corresponding to each sampling time can be performed using the following formula (13):

[0154] Formula (13);

[0155] in, Indicates sampling time The corresponding revised health index, Indicates the previous sampling time The corresponding corrected health index (its initial value is 0 or the preset default value). This represents the preset correction coefficient, which ranges from 0 to 1. It is used to control the fusion ratio between the current health index and the historical correction value, and can be dynamically adjusted according to the actual scenario.

[0156] Next, based on the pre-constructed mapping relationship between health indices and health level labels, the corrected health indices corresponding to each sampling time are mapped to the corresponding health level labels; thus obtaining the health level labels corresponding to each sampling time.

[0157] Understandably, the corrected health index is a continuous value. However, in practical operations and maintenance applications, operators often need more intuitive criteria for judging health status, and supervised learning models typically require discrete labels as training targets. Therefore, this embodiment can discretize the continuous health index into a finite number of health level labels, with each level corresponding to a health status description with clear semantics.

[0158] The mapping relationship between the health index and the health level label is pre-built. For example, the mapping relationship can be established as follows: when the health index is adjusted... When the corresponding health level label is 0, it indicates a healthy status; when When the corresponding health level label is 1, it indicates mild deterioration; when When the corresponding health level label is 2, it indicates moderate deterioration; when At this point, the corresponding health level label is 3, indicating severe degradation. The above classification method and number of levels are only examples; in practical applications, the granularity of the mapping relationship and the number of levels can be adjusted according to specific needs.

[0159] Therefore, for any sampling time According to its revised health index The value can be obtained by querying the above mapping relationship to obtain the sampling time. The corresponding health level label.

[0160] This embodiment continuously and smoothly corrects the health index by incorporating the correction result from the previous sampling time as prior information into the correction calculation for the current sampling time. This effectively suppresses instantaneous jumps in the health index caused by sampling fluctuations or communication anomalies, making the trajectory of the corrected health index more stable and continuous, and avoiding frequent misjudgments of health status. Furthermore, the continuously corrected health index is mapped to a finite number of discrete health level labels, making the evaluation results more intuitive and easier to understand. This facilitates maintenance personnel in quickly grasping the battery health status and making maintenance decisions. Simultaneously, the generated health level labels can also serve as training labels for subsequent supervised learning models, solving the problem of insufficient label samples in real-world operating scenarios.

[0161] The above describes a method for constructing a power battery health index according to embodiments of this application. The following describes the apparatus for implementing the above-described method for constructing a power battery health index.

[0162] Please see Figure 5 , Figure 5 This is a schematic diagram of a power battery health index construction device provided in an embodiment of this application. Figure 5 As shown, the power battery health index construction device may include:

[0163] The data acquisition module 501 is used to acquire the multi-source state data and the sampling time corresponding to each multi-source state data during the operation of the power battery.

[0164] The continuity discrimination module 502 is used to perform time continuity discrimination on each multi-source state data according to the sampling time, and determine the valid data segment with continuous sampling time. The valid data segment contains multi-source state data with multiple sampling times.

[0165] The dynamic feature extraction module 503 is used to extract dynamic features of the multi-source state data as a function of sampling time for each sampling time within the effective data segment. The dynamic features are used to characterize the health degradation trend of the power battery.

[0166] The health index generation module 504 is used to construct multi-dimensional health factors corresponding to each sampling time based on dynamic characteristics, and to perform time-by-time fusion processing on the multi-dimensional health factors corresponding to each sampling time to generate a health index corresponding to each sampling time. The health index is used to characterize the health status of the power battery.

[0167] Each module in the aforementioned power battery health index construction device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0168] This application also provides an electronic device, which may include at least one processor and a memory connected to the processor, wherein:

[0169] Memory is used to store computer programs;

[0170] The processor is used to execute computer programs to enable electronic devices to implement any of the power battery health index construction methods provided in the embodiments of this application.

[0171] refer to Figure 6 The diagram illustrates a structural schematic suitable for implementing the electronic device in the embodiments of this application. The electronic device in the embodiments of this application may include, but is not limited to, fixed terminals such as mobile phones, laptops, PDAs (personal digital assistants), PADs (tablet computers), desktop computers, etc. Figure 6 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0172] like Figure 6 As shown, the electronic device may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. When the electronic device is powered on, the RAM 603 also stores various programs and data required for the operation of the electronic device. The processing unit 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0173] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 608 including, for example, memory cards, hard drives, etc.; and communication devices 609. Communication device 609 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 6 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown. More or fewer devices may be implemented or have alternatively.

[0174] This application also provides a computer program product, including computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device causes the electronic device to implement any of the power battery health index construction methods provided in this application. Specific steps of this method can be found in the corresponding content disclosed in the foregoing embodiments, and will not be repeated here.

[0175] This application also provides a computer-readable storage medium carrying one or more computer programs. When these programs are executed by an electronic device, the device can implement any of the power battery health index construction methods provided in this application. Specific steps of this method can be found in the corresponding content disclosed in the foregoing embodiments, and will not be repeated here.

[0176] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0177] It should also be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and 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 modules can be selected to achieve the purpose of this embodiment according to actual needs. In addition, in the device embodiment drawings provided in this application, the connection relationship between modules indicates that they have a communication connection, which can be implemented as one or more communication buses or signal lines.

[0178] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware, or it can be implemented by special-purpose hardware including application-specific integrated circuits, special-purpose CPUs, special-purpose memory, special-purpose components, etc. Generally, any function performed by a computer program can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be diverse, such as analog circuits, digital circuits, or special-purpose circuits. However, for this application, software program implementation is more often the preferred implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium, such as a computer floppy disk, USB flash drive, mobile hard disk, ROM, RAM, magnetic disk, or optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, training equipment, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0179] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product.

[0180] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, training device, or data center to another website, computer, training device, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a training device or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state drives (SSDs)).

Claims

1. A method for constructing a health index for a power battery, characterized in that, include: Acquire the multi-source state data during the operation of the power battery and the sampling time corresponding to each multi-source state data; Based on the sampling time, the time continuity of each multi-source state data is determined to identify a valid data segment with continuous sampling time. The valid data segment contains multi-source state data with multiple sampling times. Within the effective data segment, dynamic features of the multi-source state data changing with sampling time are extracted for each sampling time. These dynamic features are used to characterize the health degradation trend of the power battery. Based on the dynamic characteristics, multidimensional health factors corresponding to each sampling time are constructed, and the multidimensional health factors corresponding to each sampling time are fused over a sampling time-by-sampling time to generate a health index corresponding to each sampling time. The health index is used to characterize the health status of the power battery.

2. The method for constructing a power battery health index according to claim 1, characterized in that, The step of determining the temporal continuity of each multi-source state data based on the sampling time to identify valid data segments with continuous sampling time includes: Calculate the time interval between adjacent sampling times in the sampling time corresponding to each of the multi-source state data; If the time interval is greater than the preset interval threshold, then the multi-source state data corresponding to the adjacent sampling times are determined to belong to different valid data segments. If the time interval is less than or equal to the interval threshold, then the multi-source state data corresponding to the adjacent sampling times are determined to belong to the same valid data segment. To obtain the effective data segment with continuous sampling time.

3. The method for constructing a power battery health index according to claim 1, characterized in that, Within the effective data segment, the dynamic features of the multi-source state data changing with sampling time are extracted for each sampling time, including: For each sampling time within the valid data segment: Calculate the rate of change characteristic between the multi-source state data at the current sampling time and the multi-source state data at the previous sampling time. If the current sampling time is the first sampling time within the valid data segment, the rate of change characteristic is omitted. A preset time window is extracted forward from the sampling time, and the variance or standard deviation of the multi-source state data within the preset time window is calculated as the fluctuation feature corresponding to the sampling time. If the time length of the multi-source state data forward from the sampling time is insufficient to constitute the preset time window, the fluctuation feature is omitted. Calculate the voltage difference between the highest and lowest voltages of individual cells in the multi-source state data at that sampling time, and use it as the voltage consistency feature corresponding to that sampling time. Calculate the temperature difference between the highest and lowest temperatures of individual cells in the multi-source state data at that sampling time, and use it as the temperature consistency feature corresponding to that sampling time. Calculate the power of the total battery pack voltage and total battery pack current in the multi-source state data at that sampling time, and use it as the power feature corresponding to that sampling time.

4. The method for constructing a power battery health index according to claim 3, characterized in that, The rate of change characteristics include at least one of the following: total voltage change rate of battery pack, total current change rate of battery pack, state of charge change rate, maximum temperature change rate of individual cell, and minimum temperature change rate of individual cell. The fluctuation characteristics include at least one of the following: total battery pack voltage fluctuation, total battery pack current fluctuation, state of charge fluctuation, maximum temperature fluctuation of individual cells, and minimum temperature fluctuation of individual cells.

5. The method for constructing a power battery health index according to claim 4, characterized in that, The multidimensional health factors include at least two of the following: voltage health factor, temperature health factor, current health factor, state of charge health factor, and consistency health factor. The voltage health factor is constructed based on the total voltage change rate and the total voltage fluctuation of the battery pack, and is used to evaluate the voltage stability and voltage degradation trend of the power battery. The temperature health factor is constructed based on the highest temperature change rate of the single cell, the lowest temperature change rate of the single cell, the highest temperature fluctuation of the single cell, and the lowest temperature fluctuation of the single cell, and is used to evaluate the thermal stability and temperature anomaly degree of the power battery. The current health factor is constructed based on the total current change rate and the total current fluctuation of the battery pack, and is used to evaluate the load response capability and current stability of the power battery. The state of charge health factor is constructed based on the state of charge change rate, the state of charge fluctuation and the power characteristics, and is used to evaluate the energy change stability and state of charge estimation reliability of the power battery. The consistency health factor is constructed based on the voltage consistency characteristic and the temperature consistency characteristic, and is used to evaluate the voltage consistency and temperature consistency of the power battery.

6. The method for constructing a power battery health index according to claim 1, characterized in that, The step of fusing the multidimensional health factors corresponding to each sampling time on a time-by-time basis includes: For each sampling time: The multidimensional health factors corresponding to the sampling time were normalized to obtain the normalized multidimensional health factors. Calculate the change sensitivity corresponding to each of the normalized multidimensional health factors, wherein the change sensitivity is determined based on the change rate of the corresponding normalized multidimensional health factors in each sampling time period; Based on the change sensitivity of the normalized multidimensional health factors, the fusion weights of the normalized multidimensional health factors are determined. The normalized multidimensional health factors are weighted and fused based on the fusion weights to obtain the health index corresponding to the sampling time.

7. The method for constructing a power battery health index according to any one of claims 1-6, characterized in that, Also includes: The health index corresponding to each sampling time is continuously smoothed and corrected to obtain the corrected health index corresponding to each sampling time. Based on the pre-constructed mapping relationship between health indices and health level labels, the corrected health indices corresponding to each sampling time are mapped to the corresponding health level labels; thus obtaining the health level labels corresponding to each sampling time.

8. A computer program product, characterized in that, It includes computer-readable instructions that, when executed on an electronic device, cause the electronic device to implement the power battery health index construction method as described in any one of claims 1 to 7.

9. An electronic device, characterized in that, It includes at least one processor and a memory connected to the processor, wherein: The memory is used to store computer programs; The processor is used to execute the computer program to enable the electronic device to implement the power battery health index construction method as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, The storage medium carries one or more computer programs, which, when executed by an electronic device, enable the electronic device to implement the power battery health index construction method as described in any one of claims 1 to 7.

Images