A method for determining a maintenance strategy of an eVTOL power battery pack and related equipment

Abstract

The application discloses a maintenance strategy determination method for an eVTOL power battery pack and related equipment, and relates to the technical field of equipment maintenance. The method comprises the following steps: acquiring operation data of each single battery in the power battery pack; extracting working condition statistical data of each single battery based on the operation data of each single battery; inputting the working condition statistical data of each single battery into a life prediction network to determine residual life data of each single battery; and determining a single battery to be replaced from all single batteries of the power battery pack based on the residual life data of each single battery. The working condition statistical data of each single battery is extracted from the operation data, the residual life data of each single battery is obtained by using the life prediction network, and then a differentiated maintenance decision at the single battery level is realized, so that resource waste caused by the replacement of the entire battery pack due to a few short single batteries is avoided.

Description

Technical Field

[0001] This application relates to the field of equipment maintenance technology, and in particular to a method for determining the maintenance strategy of an eVTOL power battery pack and related equipment. Background Technology

[0002] The battery packs of eVTOL (Electric Vertical Take-off and Landing) aircraft consist of multiple individual cells connected in series and parallel. Their maintenance strategies impact flight safety and operational economics. Current technologies typically make maintenance decisions based on the overall health status of the battery pack (e.g., the average health status of the entire pack), replacing the entire pack when its performance degrades to a threshold. However, due to inherent differences in temperature distribution, manufacturing consistency, and charge / discharge conditions among the individual cells within the pack, their aging trajectories differ. Often, only a few cells fail first, while the majority still possess a long remaining lifespan. Current technologies, unable to identify the individual health status of each cell, can only replace the entire pack, leading to the premature failure of many healthy cells and wasting resources on battery lifecycle maintenance. Therefore, a method for determining the maintenance strategy of eVTOL battery packs is urgently needed to address the aforementioned technical problems. Summary of the Invention

[0003] The summary section introduces a series of simplified concepts, which will be further explained in detail in the detailed description section. This summary section is not intended to limit the key and essential technical features of the claimed technical solutions, nor is it intended to determine the scope of protection of the claimed technical solutions.

[0004] Firstly, this application proposes a method for determining the maintenance strategy of an eVTOL power battery pack, wherein the power battery pack comprises multiple individual cells, and the method includes: Obtain the operating data of each individual cell in the power battery pack; Based on the operating data of each individual battery cell, extract the operating condition statistics for each individual battery cell; The operating condition statistics of each individual cell are input into the life prediction network to determine the remaining life data of each individual cell. Based on the remaining life data of each individual cell, the individual cells to be replaced are determined from all the individual cells of the power battery pack.

[0005] In some implementations, extracting the operating condition statistics for each individual battery cell based on the operating data of each individual battery cell includes: For each individual battery cell, the following data are extracted from the operating data: the number of discharge depths within multiple preset discharge depth ranges, the number of charge depths within multiple preset charge depth ranges, the operating time within multiple preset temperature ranges, the number of discharge rates within multiple preset discharge rate ranges, and the number of charge rates within multiple preset charge rate ranges. For each individual battery cell, the number of discharge depths, the number of charge depths, the operating time, the number of discharge rate cycles, and the number of charge rate cycles are arranged in a preset order to generate the operating condition statistics of the individual battery cell.

[0006] In some implementations, the lifetime prediction network is a multilayer perceptron network or a convolutional neural network, and the remaining lifetime data characterizes the remaining usability of the individual battery cell from its current state to the end of its lifetime.

[0007] In some implementations, determining which individual cells to be replaced from all individual cells in the power battery pack based on the remaining life data of each individual cell includes: Individual cells whose remaining lifespan data is less than a preset lifespan threshold are identified as first-class individual cells, wherein the preset lifespan threshold is the minimum remaining lifespan requirement for each individual cell. The first power battery pack is obtained by simulating the replacement of the first type of single cell, and it is determined whether each performance parameter of the first power battery pack is not less than its corresponding minimum performance requirement. If so, the first type of single cell is identified as the single cell to be replaced; If not, then from the individual cells whose remaining lifespan data is not less than the preset lifespan threshold, individual cells are sequentially selected as the second type of individual cells in ascending order of remaining lifespan data, until the performance parameters of the second power battery pack obtained by simulating the replacement of the first type of individual cells and the second type of individual cells are not less than their respective minimum performance requirements, then the first type of individual cells and the second type of individual cells are determined as the individual cells to be replaced.

[0008] In some embodiments, the performance parameters include at least one of available capacity, peak power, and remaining energy, wherein the available capacity is the total amount of electricity output by the power battery pack, the peak power is the maximum instantaneous power output by the power battery pack within a preset time window, and the remaining energy is the usable energy of the power battery pack during a preset flight phase.

[0009] In some implementations, it also includes: A replacement list is generated based on the individual cells to be replaced, the replacement list including the physical location of each individual cell to be replaced in the power battery pack; The remaining individual cells, excluding those to be replaced, are extracted from the power battery pack to generate a continued-use list, which includes the remaining lifespan data for each remaining individual cell.

[0010] In some implementations, after determining the individual battery cell to be replaced, the process further includes: Based on the individual battery to be replaced and the remaining individual batteries, the updated remaining life data of each individual battery in the power battery pack after replacement is determined, wherein the updated remaining life data of the replaced individual battery is the preset standard remaining life data, and the updated remaining life data of the remaining individual batteries is its original remaining life data. Calculate the standard deviation and range of the updated remaining lifetime data for all individual cells; The standard deviation and the range are weighted and summed to determine the degradation difference index of the power battery pack after replacement. The degradation difference index is used to characterize the degree of inconsistency in the remaining life data of each individual cell in the power battery pack.

[0011] In some embodiments, the power battery pack is composed of multiple parallel units connected in series, each parallel unit including at least two individual batteries connected in parallel, and further comprising: Based on the updated remaining lifetime data of all individual cells, individual cells whose differences between the updated remaining lifetime data are within a preset range are assigned to the same parallel unit to generate a reconfiguration pairing scheme.

[0012] Secondly, this application proposes a maintenance strategy determination device for an eVTOL power battery pack, wherein the power battery pack includes multiple individual cells, and the device includes: The data acquisition unit is used to acquire the operating data of each individual battery cell in the power battery pack. The data statistics unit is used to extract the operating condition statistics of each individual battery cell based on the operating data of each individual battery cell. The life prediction unit is used to input the operating condition statistics of each individual cell into the life prediction network to determine the remaining life data of each individual cell. A replacement determination unit is used to determine the individual cells to be replaced from all the individual cells of the power battery pack based on the remaining life data of each individual cell.

[0013] Thirdly, this application proposes an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program stored in the memory to implement the steps of the maintenance strategy determination method for the eVTOL power battery pack described in any of the first aspects.

[0014] Fourthly, this application also proposes a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the steps of the method for determining the maintenance strategy of the eVTOL power battery pack according to any one of the first aspects.

[0015] In summary, the maintenance strategy determination method for eVTOL power battery packs provided in this application transforms raw data into a quantitative description of the actual operating conditions of each cell by statistically analyzing the operating data of each cell. This captures the differentiated degradation characteristics of each cell due to variations in location, temperature, and charge / discharge history. The operating data of each cell is input into a lifespan prediction network to determine the remaining lifespan data of each cell, enabling an assessment of the remaining usability of each cell from its current state to the end of its lifespan. This overcomes the shortcomings of traditional methods that cannot determine the remaining lifespan at the cell level. Based on the remaining lifespan data of each cell, the method identifies the cells that need to be replaced from all cells. This allows maintenance decisions to target the weakest cells with performance degradation rather than the entire battery pack, maximizing the continued usability of healthy cells while ensuring safety and reducing maintenance resource consumption. Attached Figure Description

[0016] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit this specification. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A schematic flowchart of a method for determining the maintenance strategy of an eVTOL power battery pack provided in an embodiment of this application; Figure 2 A schematic diagram of a maintenance strategy determination device for an eVTOL power battery pack provided in this application embodiment; Figure 3 This is a schematic diagram of an electronic device structure for determining the maintenance strategy of an eVTOL power battery pack, provided in an embodiment of this application. Detailed Implementation

[0017] The terms "first," "second," "third," "fourth," etc. (if present) 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 data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus. 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 them.

[0018] This application relates to the field of power battery maintenance technology for electric vertical takeoff and landing (eVTOL) aircraft, particularly to power battery packs composed of multiple individual cells (i.e., battery cells) connected in series and parallel. In actual eVTOL operation, due to differences in the physical location of each individual cell within the battery pack (e.g., different heat dissipation conditions) and variations in charge / discharge history (e.g., depth of discharge, temperature, rate of discharge), the aging trajectory of each cell exhibits non-uniformity. Existing maintenance strategies typically only monitor the overall health status of the battery pack (e.g., average health). Once the overall performance of the pack degrades to a threshold, the entire pack is replaced, resulting in the premature scrapping of many healthy cells with a relatively long remaining lifespan, increasing maintenance resource consumption throughout the battery's life cycle. To address this issue, this application provides a method for determining maintenance strategies at the cell level. Application scenarios include periodic inspections, health status assessments, and predictive maintenance of eVTOLs, and it is applicable to various configurations of electric aircraft, such as multi-rotor and tiltrotor aircraft.

[0019] To facilitate understanding of the technical solution of this application, the following explanations are provided for specific terms involved in this application.

[0020] In this application, eVTOL refers to an electric vertical takeoff and landing aircraft, which is an aircraft that uses electricity as its power source and has the ability to take off and land vertically.

[0021] The power battery pack in this application refers to an energy storage device that provides driving power for eVTOL, which is composed of multiple individual batteries connected in series and parallel, and integrates a battery management system.

[0022] The single cell in this application refers to the smallest energy storage unit that constitutes a power battery pack, which is usually a lithium-ion cell. Each single cell has independent operating parameters such as voltage, current, and temperature.

[0023] The operational data in this application refers to the raw data of each individual battery cell collected by the battery management system (BMS) throughout its entire life cycle, including but not limited to cell voltage, cell current, cell temperature, state of charge, number of charge-discharge cycles, depth of discharge for each cycle, and charge-discharge rate.

[0024] The operating condition statistics in this application refer to the features obtained after statistically analyzing the operating data of each individual battery cell. Specifically, these include the number of discharge depths within multiple preset discharge depth intervals, the number of charge depths within multiple preset charge depth intervals, the operating time within multiple preset temperature intervals, the number of discharge rate intervals within multiple preset discharge rate intervals, and the number of charge rate intervals within multiple preset charge rate intervals. These statistics are arranged in a preset order and then input into the lifetime prediction network.

[0025] The lifetime prediction network in this application refers to a pre-trained multilayer perceptron network or convolutional neural network, which is used to receive the operating condition statistics of a single battery cell and output the remaining lifetime data of that single battery cell.

[0026] The remaining lifetime data in this application refers to the degree of remaining usability of a single cell from its current state to the end of its lifetime. The value range is usually from 0 to 1, and the larger the value, the higher the degree of remaining usability.

[0027] The preset lifespan threshold in this application refers to the minimum remaining lifespan requirement set for a single cell. Single cells that are below this threshold are considered to be the first type of single cells that must be replaced.

[0028] The performance parameters of this application refer to the technical requirements that the power battery pack should meet after some individual cells are replaced, including at least one of available capacity, peak power and remaining energy; wherein, available capacity is the total amount of electricity that the power battery pack can output, peak power is the maximum instantaneous power output by the power battery pack within a preset time window, and remaining energy is the usable energy provided by the power battery pack in a preset flight phase.

[0029] The degradation difference index in this application refers to an indicator used to quantify the degree of inconsistency in the remaining life data of individual cells within a power battery pack. It is obtained by calculating the weighted sum of the standard deviation and range of the remaining life data of all cells after the update. The higher the index, the more significant the difference between cells within the pack.

[0030] The reconfiguration and pairing scheme of this application refers to a specific configuration strategy that, after determining the cells to be replaced, reallocates the remaining healthy cells to each parallel unit according to the principle of similar remaining life data, aiming to improve the consistency within the battery pack after reconfiguration.

[0031] The technical solution of this application can be applied to battery health assessment and maintenance decision-making scenarios in the daily operation and maintenance of eVTOL (eVTOL) systems, especially suitable for electric aircraft with multi-rotor or tiltrotor configurations. After each flight mission or during regular maintenance, the operating data of each individual battery cell is obtained through the battery management system, and the method steps of this application are executed to automatically generate a replacement list, a continued use list, and a reconfiguration and pairing scheme. Optionally, a degradation difference index can be calculated to assess maintenance quality. This method can also be integrated into ground maintenance stations or cloud-based analysis platforms to achieve intelligent management of the entire battery lifecycle.

[0032] Please see Figure 1 This is a schematic flowchart illustrating a maintenance strategy determination method for an eVTOL power battery pack provided in an embodiment of this application. The power battery pack includes multiple individual cells, specifically including: S110. Obtain the operating data of each individual cell in the power battery pack; For example, the battery management system collects operational data such as voltage, current, temperature, state of charge, number of charge-discharge cycles, depth of discharge in each cycle, and charge-discharge rate of each individual battery cell throughout its entire life cycle, and records the operational trajectory of each cell from its initial use to the present, so as to distinguish the differentiated degradation information of each cell due to different locations and operating conditions.

[0033] S120. Based on the operating data of each individual battery cell, extract the operating condition statistics of each individual battery cell. For example, for each individual battery cell, the number of discharges in multiple preset depth of discharge ranges, the number of charges in multiple preset depth of charge ranges, the operating time in multiple preset temperature ranges, the number of discharges in multiple preset discharge rate ranges, and the number of charges in multiple preset charge rate ranges are statistically analyzed from the operating data. These statistical quantities are then arranged in a fixed order to form the operating condition statistics of the individual battery cell.

[0034] S130. Input the operating condition statistics of each individual cell into the life prediction network to determine the remaining life data of each individual cell. For example, the operating statistics of each individual cell are input into a pre-trained lifetime prediction network. The network processes the input data and outputs a normalized value, which is used to quantify the remaining usability of the individual cell from the current moment until the end of its lifetime.

[0035] S140. Based on the remaining life data of each individual cell, determine the individual cells to be replaced from all the individual cells in the power battery pack.

[0036] For example, based on the remaining life data of each individual battery cell and a preset life threshold, and combined with the overall performance requirements of the power battery pack after replacement, individual batteries that need to be replaced are selected from all individual batteries so that they meet the safety constraints after replacement.

[0037] It should be noted that safety constraints include constraints on individual battery cells and constraints on the entire battery pack. Constraints on individual battery cells mean that the remaining lifespan of each usable cell in the battery pack after replacement must not be lower than a preset lifespan threshold. This preset lifespan threshold is a minimum remaining lifespan limit pre-defined according to flight safety requirements. Cells below this threshold are considered to no longer meet the conditions for safe use and are replaced.

[0038] The constraint on the entire power battery pack means that after replacing some individual cells, all performance parameters of the entire power battery pack must not be lower than their respective minimum performance requirements. These performance parameters include at least one of the following: usable capacity, peak power, and remaining energy. Usable capacity refers to the total amount of electricity output by the power battery pack when fully charged; peak power refers to the maximum instantaneous power output by the power battery pack within a preset time window; and remaining energy refers to the usable energy provided by the power battery pack during a preset flight phase.

[0039] If replacing only the individual cells with remaining lifespan data below the preset lifespan threshold ensures that all performance parameters of the power battery pack meet the corresponding minimum performance requirements, then no additional cells need to be replaced. However, if replacing only these cells leaves one or more performance parameters below the corresponding minimum performance requirements, then cells with remaining lifespan data not below the preset lifespan threshold must be selected and replaced sequentially from those with the smallest remaining lifespan data, in ascending order, until all performance parameters of the replaced power battery pack meet the corresponding minimum performance requirements. This ensures both the safety of individual cells (no severely aged cells remain) and the functionality of the entire battery pack (the overall performance meets flight mission requirements).

[0040] In summary, the eVTOL power battery pack maintenance strategy determination method provided in this application transforms raw data into a quantitative description of the actual operating conditions of each cell by statistically analyzing the operating data of each cell. This captures the differentiated degradation characteristics of each cell due to differences in location, temperature, and charge / discharge history. The operating data of each cell is input into a life prediction network to determine the remaining lifespan of each cell, enabling an assessment of the remaining usability of each cell from its current state to the end of its lifespan. This overcomes the shortcomings of traditional methods that cannot determine the remaining lifespan at the cell level. Based on the remaining lifespan data of each cell, the method identifies the cells that need to be replaced from all cells. This allows maintenance decisions to target the weakest cells with performance degradation rather than the entire battery pack, maximizing the continued usability of healthy cells while ensuring safety and reducing maintenance resource consumption.

[0041] In some instances, based on the operational data of each individual cell, operating condition statistics for each individual cell are extracted, including: For each individual battery cell, extract from the operating data the number of discharge depths within multiple preset discharge depth ranges, the number of charge depths within multiple preset charge depth ranges, the operating time within multiple preset temperature ranges, the number of discharge rates within multiple preset discharge rate ranges, and the number of charge rates within multiple preset charge rate ranges. For each individual battery cell, the depth of discharge count, depth of charge count, operating time, discharge rate count, and charge rate count are arranged in a preset order to generate statistical data on the operating conditions of the individual battery cell.

[0042] For example, for each individual battery cell, the voltage, current, temperature, and state of charge (SOC) of that cell throughout its entire lifecycle are obtained from the operational data of the battery management system. Voltage is used to assist in estimating the SOC and verifying the validity of charge / discharge event boundary points. Based on voltage, current, and temperature, the SOC at each moment is estimated. According to the direction of the current or the time series of the SOC, the start and end times of each discharge and charge event are determined, and the depth of discharge or depth of charge for that event is calculated based on the difference between the start and end SOCs. The discharge rate or charge rate for that event is obtained by dividing the current value by the cell's rated capacity. Temperature data is recorded at each sampling time interval.

[0043] Within multiple preset discharge depth ranges, charge depth ranges, temperature ranges, discharge rate ranges, and charge rate ranges, for each discharge event, the discharge depth count for that range is incremented by 1 based on the range its discharge depth falls into; for each charge event, the charge depth count for that range is incremented by 1 based on the range its charge depth falls into. For temperature data, the operating duration of each temperature range is incremented by the duration of a sampling interval based on the temperature value at each sampling time. For each discharge event, the discharge rate count for that range is incremented by 1 based on the discharge rate range its discharge rate falls into; for each charge event, the charge rate count for that range is incremented by 1 based on the charge rate range its charge rate falls into.

[0044] By accumulating all events and sampling moments throughout the entire lifecycle, the number of discharge depths, the number of charge depths, the operating time within multiple preset discharge depth ranges, the number of discharge rates within multiple preset charge depth ranges, and the number of charge rates within multiple preset charge rate ranges for each individual battery cell are obtained. The boundary values ​​for the preset discharge depth range, preset charge depth range, preset temperature range, preset discharge rate range, and preset charge rate range are all preset based on battery aging mechanisms and engineering experience.

[0045] After obtaining the statistics across the five dimensions mentioned above, the values ​​of each interval within these statistics are arranged in a predetermined fixed order to generate the operating condition statistics for the individual battery. First, the number of discharge depths corresponding to multiple preset discharge depth intervals is arranged; second, the number of charge depths corresponding to multiple preset charge depth intervals is arranged; third, the operating time corresponding to multiple preset temperature intervals is arranged; then, the number of discharge rates corresponding to multiple preset discharge rate intervals is arranged; and finally, the number of charge rates corresponding to multiple preset charge rate intervals is arranged. The values ​​from all the statistics are concatenated into a one-dimensional array in this order, which is the operating condition statistics. Since each individual battery uses the same preset interval division and the same arrangement order, the operating condition statistics for all individual batteries have the same dimensions and data structure, facilitating input into the lifetime prediction network.

[0046] In summary, this application extracts operating condition statistics using the above methods. It independently counts the number of cycles in each discharge and charge depth range for each individual cell, quantifying the differentiated impact of different discharge and charging habits on the aging rate. This avoids indiscriminately ignoring the additional damage caused by deep charge-discharge cycles. The operating time within each temperature range is statistically analyzed to quantify the effectiveness of temperature-related aging mechanisms such as high-temperature accelerated aging and low-temperature charging lithium plating. Since temperature is a key external stress affecting battery life, the cumulative duration in different temperature ranges directly reflects the contribution of thermal stress to battery aging. The number of discharge and charge rate cycles are statistically analyzed separately, allowing for the separate quantification of stress caused by high-current shocks (such as high-rate discharge during takeoff) and fast charging. This avoids smoothing out the significant impact of extreme operating conditions on lifespan using average rates. The above multidimensional statistics are concatenated in a fixed order to form operating condition statistics, preserving multi-factor aging information. This allows the lifespan prediction network to use this information for regression mapping, improving the accuracy of remaining lifespan prediction.

[0047] In some instances, the lifetime prediction network is a multilayer perceptron network or a convolutional neural network, and the remaining lifetime data characterizes the remaining usability of a single cell from its current state to the end of its life.

[0048] For example, the operating condition statistics of each individual battery cell are input into a pre-trained lifespan prediction network. This lifespan prediction network employs a multilayer perceptron network or a convolutional neural network (CNN) structure. The multilayer perceptron network performs nonlinear transformations on the input operating condition statistics through multiple fully connected layers, while the CNN extracts local correlation patterns between adjacent features in the operating condition statistics through convolutional and pooling layers. The network internally includes an input layer, multiple hidden layers, and an output layer. The number of nodes in the input layer matches the dimensionality of the operating condition statistics, and the output layer has one node to output the remaining lifespan data. The network is trained using supervised learning, with training samples including a large amount of historical operating condition statistics for individual batteries and their corresponding actual lifespan labels. By learning a nonlinear mapping relationship from operating condition statistics to remaining lifespan, the network can predict the corresponding remaining lifespan for any input operating condition statistics after training. The network output layer uses a sigmoid activation function, restricting the output value to the range of 0 to 1.

[0049] It should be noted that when a convolutional neural network (CNN) is used in the lifespan prediction network, it is a one-dimensional CNN. Operating condition statistics are a one-dimensional vector, where each position corresponds to a statistical value within a preset interval (e.g., the number of discharge depths in the first interval, the number of discharge depths in the second interval, etc.). The statistics between adjacent positions are physically related; for example, there may be an order relationship between different intervals within the same dimension, or they may be adjacent in vector arrangement across different dimensions. The one-dimensional CNN extracts the correlation patterns between locally adjacent features in this vector by sliding a one-dimensional convolutional kernel, then reduces the dimensionality and retains salient features through pooling layers, and finally maps to a single output value through a fully connected layer. Compared to two-dimensional CNNs, one-dimensional CNNs are more suitable for processing sequential or vector-like one-dimensional input data.

[0050] The remaining lifetime data output by the lifetime prediction network is a normalized value between 0 and 1. This value represents the remaining usability of the individual battery cell from its current state to the end of its lifespan. The closer the value is to 1, the higher the remaining usability and the further away from the end of the lifespan; the closer the value is to 0, the lower the remaining usability and the closer to the end of the lifespan. Specifically, a remaining lifetime data of 1 indicates that the individual battery cell is in a brand-new state, while a remaining lifetime data of 0 indicates that the individual battery cell can no longer provide effective electrical energy and its lifespan has completely ended. Unlike traditional methods that only count the cumulative number of cycles, this remaining lifetime data is predicted based on multi-dimensional operating condition statistics such as the actual discharge depth distribution, charge depth distribution, temperature stress distribution, and charge / discharge rate distribution experienced by the individual battery cell. It reflects the differentiated contribution of different aging stress modes to lifespan consumption, thus more accurately describing the aging state of the individual cell.

[0051] In summary, this embodiment of the application uses a lifespan prediction network to determine remaining lifespan data in the manner described above. Multilayer perceptron networks or convolutional neural networks possess strong nonlinear fitting capabilities, learning the high-dimensional mapping relationship between operating condition statistics and remaining lifespan, overcoming the shortcomings of traditional linear models or empirical formulas in accurately modeling multi-factor coupled aging effects. Using operating condition statistics as network input instead of the original time series reduces the input dimensionality and network complexity, making network training easier to converge and inference faster, suitable for real-time deployment in eVTOL ground maintenance systems or cloud platforms. The output layer uses the Sigmoid activation function to normalize the results to the 0-1 range, giving the remaining lifespan data a uniform numerical range, facilitating comparison with preset lifespan thresholds and horizontal comparisons between different units. This lifespan prediction network independently predicts the remaining lifespan data for each unit, achieving unit-level aging assessment.

[0052] In some instances, based on the remaining lifespan data of each individual cell, the cells to be replaced are determined from all the cells in the power battery pack, including: Individual cells with remaining life data less than a preset life threshold are identified as Class I individual cells, where the preset life threshold is the minimum remaining life requirement for each individual cell. The first power battery pack is obtained by simulating the replacement of the first type of single cell, and it is determined whether the various performance parameters of the first power battery pack are not less than their respective minimum performance requirements. If so, the first type of single cell will be identified as the single cell to be replaced. If not, then from the individual cells whose remaining lifespan data is not less than the preset lifespan threshold, individual cells are selected in ascending order of remaining lifespan data as the second type of individual cells, until the simulated replacement of the first type of individual cells and the second type of individual cells shows that all performance parameters of the second power battery pack are not less than their respective minimum performance requirements. At this point, the first type of individual cells and the second type of individual cells are determined as the individual cells to be replaced.

[0053] For example, all individual cells with remaining life data less than a preset life threshold are marked as Class I individual cells. This preset life threshold is the minimum remaining life requirement allowed for each individual cell; cells below this value are considered no longer safe for continued use. A simulated replacement operation is performed on the power battery pack, assuming that all Class I individual cells are replaced with brand new individual cells (with remaining life data at a preset standard value, typically 1), while the remaining individual cells remain unchanged, thus forming the first power battery pack. Based on the remaining life data of all individual cells in the first power battery pack and the series-parallel topology of the battery pack, various performance parameters of the first power battery pack are calculated, including at least one of available capacity, peak power, and remaining energy, and each calculated performance parameter is compared with its corresponding minimum performance requirement. If all performance parameters are not less than the corresponding minimum performance requirement, then the Class I individual cells are directly identified as the individual cells that ultimately need to be replaced.

[0054] If one or more performance parameters of the first power battery pack are lower than the corresponding minimum performance requirements, it indicates that replacing only the first-type individual cells is insufficient to restore the overall performance of the pack to a safe level. In this case, from the individual cells whose remaining lifespan data is not less than a preset lifespan threshold (i.e., cells not marked as first-type), individual cells are selected as second-type individual cells in ascending order of remaining lifespan data. For each newly selected cell, the first-type individual cells and all previously selected second-type individual cells are replaced with brand new cells in a simulation environment, and the performance parameters of the new power battery pack are recalculated. This selection and calculation process is repeated until all performance parameters of the second power battery pack obtained after a certain selection are not less than their respective minimum performance requirements. At this point, the selection stops, and the first-type individual cells and all second-type individual cells included in that selection are determined as the final individual cells to be replaced. During the above process, the smaller the remaining lifespan data, the more severe the aging of the individual cell; therefore, it is prioritized for replacement, and the number of replacements is controlled as much as possible while meeting the overall performance requirements of the pack.

[0055] In summary, this application's embodiments ensure that severely aged cells remain in the battery pack by including cells with remaining lifespan data below a preset lifespan threshold in the replacement scope. This eliminates safety hazards such as localized thermal runaway or limited capacity in series branches caused by extreme cell degradation. Based on replacing the first type of cells, the overall pack performance parameters are simulated to assess whether the minimum performance requirements are met. Replacement is only stopped when the overall pack performance meets the standards, avoiding the risk of insufficient power or charge shortages during flight missions due to inadequate replacement. When the overall pack performance fails to meet the standards, the second type of cells are selected sequentially in ascending order of remaining lifespan data. This greedy strategy prioritizes replacing cells with relatively severe aging, meeting performance constraints with the fewest replacements, thereby minimizing maintenance resource consumption while ensuring safety.

[0056] In some instances, performance parameters include at least one of available capacity, peak power, and remaining energy, wherein available capacity is the total amount of electricity output by the power battery pack, peak power is the maximum instantaneous power output by the power battery pack within a preset time window, and remaining energy is the energy available to the power battery pack during a preset flight phase.

[0057] For example, usable capacity refers to the total amount of electricity output by the power battery pack when fully charged. Its value is determined by the remaining lifespan data of all individual cells within the battery pack and their series-parallel topology. In a series branch, the usable capacity is limited by the individual cell with the smallest remaining lifespan, because since the series current is the same, the cell with the smallest capacity reaches the discharge cutoff voltage first. In a parallel branch, the usable capacity is the sum of the capacities of all parallel individual cells. Therefore, based on the remaining lifespan data and rated capacity of each individual cell, the current usable capacity of each cell can be calculated. Combined with the series-parallel connection relationship of the battery pack, the usable capacity of the entire power battery pack can be obtained.

[0058] Peak power refers to the maximum instantaneous power output of a power battery pack within a preset time window (e.g., 10 seconds or 30 seconds). Its calculation requires considering the voltage drop and maximum allowable discharge current of individual cells during high-current discharge. Peak power calculation is based on the remaining lifespan data of each individual cell and its corresponding estimated internal resistance. Lower remaining lifespan data typically indicates higher internal resistance and a more significant voltage drop during high-current discharge, thus limiting the maximum instantaneous power output of a single cell. By using a series-parallel topology of the battery pack, the peak power capabilities of all cells are aggregated (in series branches, peak power is limited by the weakest cell; in parallel branches, the peak power of each cell is added together) to obtain the overall peak power of the pack. Alternatively, a mapping relationship can be established beforehand by experimentally calibrating the peak power capabilities of cells with different remaining lifespan data. During calculation, the peak power capability of each cell can be directly obtained by looking up a table based on its remaining lifespan data.

[0059] Remaining energy refers to the usable energy provided by the power battery pack during a preset flight phase (such as the vertical takeoff phase, hovering phase, or the entire flight mission). The calculation of remaining energy incorporates the power demand curve for the preset flight phase. Based on the power required at each moment in the flight phase and the voltage plateau of the battery pack, the corresponding discharge rate at that moment is calculated. According to the battery's energy efficiency at different discharge rates (usually pre-calibrated experimentally, e.g., 0.5C discharge can release 100% of the rated energy, while 5C discharge may only release 90% of the rated energy), the energy output at each moment is calculated, and then integrated over time to obtain the total usable energy for the entire flight phase. Alternatively, based on the average discharge rate and duration of the flight phase, the energy conversion coefficient at that rate is obtained from a table, and the usable capacity is multiplied by this coefficient and then multiplied by the rated voltage to obtain the remaining energy.

[0060] In some instances, it also includes: A replacement list is generated based on the individual cells to be replaced. The replacement list includes the physical location of each individual cell to be replaced in the power battery pack. Extract the remaining individual cells from the power battery pack, excluding those to be replaced, and generate a list of cells to be used. The list of cells to be used includes the remaining lifespan data of each remaining individual cell.

[0061] For example, a replacement list is generated based on the individual battery cell to be replaced. This replacement list includes the physical location of each individual battery cell within the power battery pack. Specifically, after determining the individual battery cell to be replaced through the steps described above, the spatial coordinates or connection identifiers of each individual battery cell within the battery pack are obtained, such as its serial number in a series branch, its number in a parallel unit, and its row and column identifier within the module. This location information is associated with and recorded with the unique identifier of the corresponding individual battery cell to generate the replacement list. This replacement list is used to guide maintenance personnel in quickly locating the cells that need to be removed when disassembling the battery pack, avoiding the accidental removal of healthy cells.

[0062] The remaining cells in the battery pack, excluding the cell to be replaced, are extracted to generate a continued-use list. This list includes the remaining lifespan data for each remaining cell. Specifically, after identifying the cell to be replaced, all cells in the battery pack not marked as replacement are selected as remaining cells. For each remaining cell, its remaining lifespan data is associated with its physical location and unique identifier, forming a continued-use list. This continued-use list informs maintenance personnel which cells can remain in the battery pack for continued service.

[0063] In some instances, after identifying the individual battery cell to be replaced, the process also includes: Based on the individual cells to be replaced and the remaining individual cells, the updated remaining life data of each individual cell in the power battery pack after replacement is determined. The updated remaining life data of the replaced individual cells is the preset standard remaining life data, and the updated remaining life data of the remaining individual cells is their original remaining life data. Calculate the standard deviation and range of the updated remaining lifetime data for all individual cells; The standard deviation and range are weighted and summed to determine the degradation difference index of the power battery pack after replacement. The degradation difference index is used to characterize the degree of inconsistency in the remaining life data of individual cells in the power battery pack.

[0064] For example, after identifying the individual battery cells to be replaced, information on two types of individual batteries is obtained: one type is the individual battery cells to be replaced, and the other type is the remaining individual battery cells that continue to be used. For each individual battery cell to be replaced, after replacing it with a brand new individual battery cell, the remaining life data of the new individual battery cell at that location is set to a preset standard remaining life data, for example, a value of 1, indicating that the new individual cell is in a completely healthy state. For the remaining individual battery cells that continue to be used, they are not replaced before or after the replacement operation, so their updated remaining life data remains the original remaining life data, that is, the value output by the life prediction network in step S130. Thus, the updated remaining life data of each individual battery cell in the power battery pack after replacement (including the newly replaced individual cells and the original remaining individual cells) can be obtained.

[0065] After obtaining the updated remaining lifetime data for all individual cells, the standard deviation and range of these data are calculated. The standard deviation is calculated as follows: first, calculate the arithmetic mean of all updated remaining lifetime data for each cell; then, calculate the sum of squares of the differences between each data point and the mean; divide this sum by the total number of cells and take the square root to obtain the standard deviation. The range is calculated as follows: find the maximum and minimum values ​​among all updated remaining lifetime data; subtract the minimum value from the maximum value to obtain the range. The standard deviation reflects the dispersion of all individual cell remaining lifetime data relative to the mean, while the range reflects the total span of the data distribution. Both describe the inconsistency between cells within the group from different perspectives.

[0066] The standard deviation and range are weighted and summed, and the result is used as the degradation difference index of the replaced power battery pack. During the weighted summation, a first weighting coefficient is assigned to the standard deviation, and a second weighting coefficient is assigned to the range. Both the first and second weighting coefficients are pre-set positive numbers, for example, the first weighting coefficient is 0.6 and the second weighting coefficient is 0.4, or they can be calibrated according to the series and parallel topology of the battery pack and safety requirements. The formula for calculating the weighted summation is: Degradation Difference Index = First Weighting Coefficient × Standard Deviation + Second Weighting Coefficient × Range. The larger the value of the degradation difference index, the more significant the difference in the remaining life data of the individual cells in the replaced power battery pack, i.e., the more uneven the health status; the smaller the value, the closer the remaining life data of the individual cells, i.e., the better the consistency. This index can be used to evaluate the quality of this repair plan or to determine whether re-pairing is necessary.

[0067] In some instances, the power battery pack is composed of multiple parallel units connected in series. Each parallel unit includes at least two individual battery cells connected in parallel with each other, and also includes: Based on the updated remaining life data of all individual cells, individual cells whose differences between the updated remaining life data are within a preset range are assigned to the same parallel unit to generate a reconfiguration pairing scheme.

[0068] For example, for all individual cells in the replaced power battery pack (including newly replaced cells and existing cells), the updated remaining lifespan data of each cell is obtained. A difference threshold is set as a preset range, for example, a preset range of 0.05, indicating that the absolute value of the difference between the updated remaining lifespan data of two cells does not exceed 0.05.

[0069] Iterate through all individual cells and group cells with similar remaining life data into the same group, ensuring that the difference in updated remaining life data between any two cells within the same group does not exceed a preset range. According to the physical topology of the power battery pack, each parallel unit needs to accommodate at least two parallel-connected individual cells. Each group of cells is assigned to a parallel unit, meaning all cells within the same group are configured to be connected in parallel. When the number of individual cells in a group exceeds the capacity of a single parallel unit, the cells in that group are sorted from largest to smallest according to their updated remaining life data and then sequentially assigned to multiple parallel units, ensuring that the number of individual cells in each parallel unit does not exceed the unit's capacity and that the updated remaining life data of the individual cells within each parallel unit remains close. All cells are assigned to their corresponding parallel units, forming a reconfiguration and pairing scheme. This scheme is output in tabular or graphical form to guide maintenance personnel in reassembling the battery pack.

[0070] By generating a reconfiguration pairing scheme, this application, after replacing some individual cells, utilizes the updated remaining lifespan data of all cells to allocate cells with similar lifespans to the same parallel unit, achieving lifespan consistency within the parallel unit. In the parallel unit, the cells are directly connected in parallel, and their terminal voltages are forced to be equal. If the remaining lifespan data of each cell differs significantly, cells with different aging levels will experience uneven current distribution during charging and discharging. Cells with shorter lifespans will bear greater current stress, accelerating their failure and potentially triggering thermal runaway. By controlling the lifespan difference within a preset range, the aging pace of each cell within the parallel unit is ensured to be basically synchronized, suppressing parallel circulating current and uneven heating, and extending the overall lifespan of the reconfigured battery pack. Simultaneously, this scheme considers the topological characteristics of the power battery pack, which connects in parallel first and then in series, avoiding deterioration caused by random assembly. It maximizes the reuse value of healthy cells while ensuring safety redundancy, reducing maintenance resource consumption.

[0071] Please see Figure 2 This is a schematic diagram of a maintenance strategy determination device for an eVTOL power battery pack provided in an embodiment of this application. The power battery pack includes multiple individual cells, including: The data acquisition unit 21 is used to acquire the operating data of each individual cell in the power battery pack; The data statistics unit 22 is used to extract the operating condition statistics of each individual battery cell based on the operating data of each individual battery cell. The life prediction unit 23 is used to input the operating condition statistics of each individual cell into the life prediction network to determine the remaining life data of each individual cell. Replacement determination unit 24 is used to determine the individual cells to be replaced from all individual cells in the power battery pack based on the remaining life data of each individual cell.

[0072] Please see Figure 3 This application also provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor. When the processor 320 executes the computer program 311, it implements the steps of the maintenance strategy determination method for the eVTOL power battery pack.

[0073] Since the electronic device described in this embodiment is the device used to implement the maintenance strategy determination device for the eVTOL power battery pack in this application embodiment, those skilled in the art can understand the specific implementation method and various variations of the electronic device in this embodiment based on the method described in this application embodiment. Therefore, how the electronic device implements the method in this application embodiment will not be described in detail here. Any device used by those skilled in the art to implement the method in this application embodiment is within the scope of protection of this application.

[0074] In practice, when the computer program 311 is executed by the processor, it can implement any of the embodiments corresponding to the first aspect.

[0075] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0076] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

[0077] Although preferred embodiments have been described in this specification, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications beyond the scope of this specification.

[0078] Obviously, those skilled in the art can make various modifications to this specification without departing from its spirit and scope. Therefore, this specification also intends to include any modifications that fall within the scope of the claims and their equivalents.

Claims

1. A method for determining the maintenance strategy of an eVTOL power battery pack, characterized in that, The power battery pack includes multiple individual battery cells, and the method includes: Obtain the operating data of each individual cell in the power battery pack; Based on the operating data of each individual battery cell, extract the operating condition statistics for each individual battery cell; The operating condition statistics of each individual cell are input into the life prediction network to determine the remaining life data of each individual cell. Based on the remaining life data of each individual cell, the individual cells to be replaced are determined from all the individual cells of the power battery pack.

2. The method according to claim 1, characterized in that, The extraction of operating condition statistics for each individual battery cell based on the operating data of each individual battery cell includes: For each individual battery cell, the following data are extracted from the operating data: the number of discharge depths within multiple preset discharge depth ranges, the number of charge depths within multiple preset charge depth ranges, the operating time within multiple preset temperature ranges, the number of discharge rates within multiple preset discharge rate ranges, and the number of charge rates within multiple preset charge rate ranges. For each individual battery cell, the number of discharge depths, the number of charge depths, the operating time, the number of discharge rate cycles, and the number of charge rate cycles are arranged in a preset order to generate the operating condition statistics of the individual battery cell.

3. The method according to claim 1, characterized in that, The lifetime prediction network is a multilayer perceptron network or a convolutional neural network, and the remaining lifetime data represents the remaining usability of the single battery cell from its current state to the end of its lifetime.

4. The method according to claim 1, characterized in that, The process of determining which individual cells to be replaced from all individual cells in the power battery pack, based on the remaining life data of each individual cell, includes: Individual cells whose remaining lifespan data is less than a preset lifespan threshold are identified as first-class individual cells, wherein the preset lifespan threshold is the minimum remaining lifespan requirement for each individual cell. The first power battery pack is obtained by simulating the replacement of the first type of single cell, and it is determined whether each performance parameter of the first power battery pack is not less than its corresponding minimum performance requirement. If so, the first type of single cell is identified as the single cell to be replaced; If not, then from the individual cells whose remaining lifespan data is not less than the preset lifespan threshold, individual cells are sequentially selected as the second type of individual cells in ascending order of remaining lifespan data, until the performance parameters of the second power battery pack obtained by simulating the replacement of the first type of individual cells and the second type of individual cells are not less than their respective minimum performance requirements, then the first type of individual cells and the second type of individual cells are determined as the individual cells to be replaced.

5. The method according to claim 4, characterized in that, The performance parameters include at least one of available capacity, peak power, and remaining energy, wherein the available capacity is the total amount of electricity output by the power battery pack, the peak power is the maximum instantaneous power output by the power battery pack within a preset time window, and the remaining energy is the usable energy of the power battery pack during a preset flight phase.

6. The method according to claim 1, characterized in that, Also includes: A replacement list is generated based on the individual cells to be replaced, the replacement list including the physical location of each individual cell to be replaced in the power battery pack; The remaining individual cells, excluding those to be replaced, are extracted from the power battery pack to generate a continued-use list, which includes the remaining lifespan data for each remaining individual cell.

7. The method according to claim 6, characterized in that, After identifying the individual battery cell to be replaced, the process also includes: Based on the individual battery to be replaced and the remaining individual batteries, the updated remaining life data of each individual battery in the power battery pack after replacement is determined, wherein the updated remaining life data of the replaced individual battery is the preset standard remaining life data, and the updated remaining life data of the remaining individual batteries is its original remaining life data. Calculate the standard deviation and range of the updated remaining lifetime data for all individual cells; The standard deviation and the range are weighted and summed to determine the degradation difference index of the power battery pack after replacement. The degradation difference index is used to characterize the degree of inconsistency in the remaining life data of each individual cell in the power battery pack.

8. The method according to claim 7, characterized in that, The power battery pack is composed of multiple parallel units connected in series. Each parallel unit includes at least two individual batteries connected in parallel with each other, and also includes: Based on the updated remaining lifetime data of all individual cells, individual cells whose differences between the updated remaining lifetime data are within a preset range are assigned to the same parallel unit to generate a reconfiguration pairing scheme.

9. A device for determining the maintenance strategy of an eVTOL power battery pack, characterized in that, The power battery pack includes multiple individual batteries, and the device includes: The data acquisition unit is used to acquire the operating data of each individual battery cell in the power battery pack. The data statistics unit is used to extract the operating condition statistics of each individual battery cell based on the operating data of each individual battery cell. The life prediction unit is used to input the operating condition statistics of each individual cell into the life prediction network to determine the remaining life data of each individual cell. A replacement determination unit is used to determine the individual cells to be replaced from all the individual cells of the power battery pack based on the remaining life data of each individual cell.

10. An electronic device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program stored in the memory to implement the steps of the method for determining the maintenance strategy of the eVTOL power battery pack as described in any one of claims 1 to 8.

Images