An online fault detection analysis method, device and system for a new energy battery

Abstract

The present application relates to the technical field of determining battery health state, in particular to an online fault detection analysis method, device and system of new energy battery. The present application firstly acquires the charging current curve in each cycle process, all discharge current curves, all bad driving sub-sections in each discharge current curve and the battery health state after each cycle; then analyzes and acquires the health discharge index and health charging index of the new energy vehicle battery in each cycle process, so as to acquire the health use weight; finally, based on the health use weight and the battery health state after each cycle, the online fault detection of the new energy vehicle battery to be tested is carried out. The present application analyzes the driving behavior and charging behavior of the new energy battery to evaluate the influence of use on the battery loss, adjusts the particle filtering process, so as to improve the battery health state estimation accuracy, and further improve the online fault detection effect of the new energy battery.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of determining battery health state, and in particular to an online fault detection analysis method, device and system for new energy batteries. BACKGROUND

[0002] New energy batteries refer to devices for storing and releasing electric energy using renewable energy, and are particularly widely used in the field of new energy vehicles. In new energy vehicles, the battery is the core component that consumers pay most attention to, and its quality, performance and reliability will affect the vehicle's endurance and user experience. Therefore, it is crucial to detect the fault of new energy batteries. The fault of new energy batteries is mainly caused by abnormal use and battery life attenuation. The state of health (SOH) is an index for describing the degree of performance attenuation of the battery, which can be used to assess whether the battery is faulty.

[0003] At present, the battery health state is usually evaluated based on the number of cycles of the battery, and then a particle filtering algorithm is used to predict the SOH, so as to evaluate whether the battery is faulty online. However, in the actual use process of new energy vehicles, the bad driving behavior and charging behavior of the driver, such as frequent acceleration and deceleration or overcharging, can all cause the battery to be consumed or degraded faster, thereby affecting the accuracy of the evaluation of the battery health state, and further causing the particle filtering method to be unable to accurately predict the SOH of the new energy battery, which will ultimately affect the online fault detection effect of the new energy battery. SUMMARY

[0004] In order to solve the technical problem that the online fault detection effect of the prior art on new energy batteries is not good, the purpose of the present application is to provide an online fault detection analysis method, device and system for new energy batteries, and the technical solution adopted is as follows:

[0005] The present application provides an online fault detection analysis method for new energy batteries, which comprises:

[0006] Obtaining the charging current curve and all discharge current curves of the new energy vehicle battery in each cycle process, obtaining all bad driving sub-sections in each discharge current curve, and obtaining the battery health state of the new energy vehicle battery after each cycle ends;

[0007] In each cycle process, the health discharge index of the new energy vehicle battery is obtained according to all bad driving sub-sections in each discharge current curve and the current change in each bad driving sub-section. The health charging index of the new energy vehicle battery is obtained according to the deviation of the starting current in the charging current curve of all cycle processes from the preset low charging current and the deviation of the ending current from the preset overcharging current, combined with the overcharging time in each cycle process;

[0008] According to the fluctuation of the health discharge index and the health charge index in all cycle processes, combined with the health discharge index and the health charge index of the new energy automobile battery in the current cycle process, the health use weight of the new energy automobile battery is obtained;

[0009] Based on the health use weight of the new energy automobile battery and the battery health state after each cycle, online fault detection is performed on the new energy automobile battery to be tested.

[0010] Further, the method for obtaining the health discharge index comprises:

[0011] In each of the discharge current curves in each cycle process, the time length ratio of the total time length of the bad driving sub-section to the total time length of the discharge current curve is taken as the bad driving discharge time proportion; according to the bad driving discharge time proportion, all the bad discharge current curves are selected from all the discharge current curves in each cycle process;

[0012] The difference between the preset first number and the number of the bad discharge current curves is taken as the health discharge weight of the new energy automobile battery in each cycle process;

[0013] In each cycle process, according to the bad driving discharge time proportion of each discharge current curve, combined with the change rate of the current range in each bad driving sub-section in each discharge current curve, the bad discharge parameter of the new energy automobile battery is obtained;

[0014] The bad discharge parameter is weighted by using the health discharge weight, the weighted result is normalized, and the normalized result is taken as the health discharge index of the new energy automobile battery in each cycle process.

[0015] Further, the method for obtaining the bad discharge parameter comprises:

[0016] The ratio between the current range in each bad driving sub-section and the time interval corresponding to the current range is taken as the change rate of the current range in the corresponding bad driving sub-section; the change rates of the current range in all bad driving sub-sections in each discharge current curve are integrated to obtain the bad driving discharge parameter of each discharge current curve;

[0017] The bad driving discharge time proportion and the bad driving discharge parameter of each discharge current curve are fused to obtain the bad driving discharge parameter of each discharge current curve; the bad driving discharge parameters of all discharge current curves in each cycle process are integrated to obtain the bad discharge parameter of the new energy automobile battery in each cycle process.

[0018] Further, the method for obtaining the health charge index comprises:

[0019] In each of the charging current curves, a low charge duration is obtained, a low charge reference current is determined according to a size relationship between the starting current and a preset low charge current, and it is determined whether the charging current curve is a low charge current curve; an overcharge reference current is determined according to a size relationship between the ending current and a preset overcharge current, and it is determined whether the charging current curve is an overcharge current curve;

[0020] A difference value between a preset second quantity and a total quantity of the low charge current curves is taken as a first health charging weight value; a sum of absolute values of differences between the low charge reference currents in all the charging current curves and the preset low charge current is taken as a poor low charge parameter; the poor low charge parameter is weighted by using the first health charging weight value, and a weighting result is taken as a first health charging index;

[0021] A difference value between a preset second quantity and a total quantity of the overcharge current curves is taken as a second health charging weight value; a sum of absolute values of differences between the overcharge reference currents in all the charging current curves and the preset overcharge current is calculated, and a product of the sum of absolute values of differences and the overcharge duration is taken as a poor low charge parameter; the poor low charge parameter is weighted by using the second health charging weight value, and a weighting result is taken as a second health charging index;

[0022] The first health charging index and the second health charging index are fused to obtain a health charging index of a new energy automobile battery.

[0023] Further, the overcharge duration obtaining method, the low charge reference current and the overcharge reference current obtaining method, and the low charge current curve and the overcharge current curve determining method comprise:

[0024] In each of the charging current curves, if the starting current is greater than or equal to the preset low charge current, the preset low charge current is taken as the low charge reference current; if the starting current is less than the preset low charge current, the starting current is taken as the low charge reference current; if the ending current is greater than the preset overcharge current, the ending current is taken as the overcharge reference current, and if the ending current is less than or equal to the preset overcharge current, the ending current is taken as the overcharge reference current;

[0025] If the starting current in the charging current curve is less than the preset low charge current, the charging current curve is taken as a low charge current curve; if the starting current in the charging current curve is greater than the preset overcharge current, the charging current curve is taken as an overcharge current curve;

[0026] A time point at which the preset overcharge current is first reached in the charging current curve is taken as an overcharge time point, and a duration between the overcharge time point and a time point corresponding to the ending current is taken as an overcharge duration.

[0027] Further, the health use weight obtaining method comprises:

[0028] According to the range of the health discharge index in all cycles and the deviation of each health discharge index from the average level, a health discharge fluctuation parameter is obtained;

[0029] According to the range of the health charge index in all cycles and the deviation of each health charge index from the average level, a health charge fluctuation parameter is obtained;

[0030] The health discharge fluctuation parameter and the health charge fluctuation parameter are fused, and the fusion result is negatively correlated to obtain a cycle stability coefficient of the new energy vehicle battery;

[0031] According to the health discharge index, the health charge index and the cycle stability coefficient, a health use weight of the new energy vehicle battery is obtained; the health discharge index, the health charge index and the cycle stability coefficient are positively correlated to the health use weight.

[0032] Further, the method for online fault detection of the new energy vehicle battery under test comprises:

[0033] In the current cycle, based on the particle filtering algorithm and the battery health state of the new energy vehicle battery after each cycle, the future battery health state is predicted; if the future battery health state is lower than a preset health threshold, a fault warning is performed;

[0034] The method for predicting the future battery health state comprises:

[0035] In the prediction process based on the particle filtering algorithm, the health use weight of the new energy vehicle battery is used to weight the battery health state of each particle, and the corresponding weighted result is used as the corrected battery health state of the corresponding particle; any particle is taken as a target particle, the particle with the minimum difference between the battery health state and the corrected battery health state of the target particle is taken as a reference particle of the target particle, and the weight of the reference particle is taken as the corrected weight of the target particle; based on the corrected weight and the corrected battery health state of each particle after resampling, the future battery health state is obtained.

[0036] Further, the method for obtaining the battery health state comprises:

[0037] Based on a double exponential degradation model, the battery health state of the new energy vehicle battery after each cycle is evaluated.

[0038] The application further provides a new energy battery online fault detection analysis device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor.

[0039] The application further provides a new energy battery online fault detection analysis system, which comprises:

[0040] The battery cycle data acquisition module is used for acquiring the charging current curve and all the discharge current curves of the new energy automobile battery in each cycle process, acquiring all the bad driving sub-sections in each discharge current curve and the battery health state of the new energy automobile battery after each cycle ends.

[0041] The battery cycle data analysis module is used for acquiring the health discharge index of the new energy automobile battery according to the total time length of the bad driving sub-sections in each discharge current curve and the current change in each bad driving sub-section in each cycle process, acquiring the health charging index of the new energy automobile battery according to the deviation of the starting current in the charging current curve of all the cycle processes from the preset low charging current and the deviation of the ending current from the preset overcharging current, combining the overcharging time length in each cycle process, acquiring the health use weight of the new energy automobile battery according to the fluctuation change of the health discharge index and the health charging index in all the cycle processes, and combining the health discharge index and the health charging index of the new energy automobile battery in the current cycle process.

[0042] The battery fault detection module is used for performing online fault detection on the new energy automobile battery to be tested based on the health use weight of the new energy automobile battery and the battery health state in the current cycle process.

[0043] The application has the following beneficial effects:

[0044] The application firstly acquires the charging current curve and all discharge current curves of the new energy automobile battery in each cycle process, acquires all bad driving sub-sections in each discharge current curve and the battery health state of the new energy automobile battery after each cycle, so as to analyze and evaluate the health charging and discharging of the new energy automobile battery in each cycle process, and further predict and evaluate whether the new energy automobile battery is faulty, thereby improving the fault evaluation accuracy; in each cycle process, the health discharge index of the new energy automobile battery is acquired according to all bad driving sub-sections in each discharge current curve and the current change in each bad driving sub-section; the health charging index of the new energy automobile battery is acquired according to the deviation of the starting current in the charging current curve of all cycle processes from the preset low charging current and the deviation of the ending current from the preset overcharging current, combined with the overcharging time in each cycle process; the health use weight of the new energy automobile battery is acquired according to the fluctuation change of the health discharge index and the health charging index in all cycle processes, combined with the health discharge index and the health charging index of the new energy automobile battery in the current cycle process, the health use weight reflects the persistence of the bad driving behavior or the bad charging behavior on the damage of the new energy battery, and further reflects the influence of the bad behavior on the health state of the new energy automobile battery, thereby facilitating the adjustment of the particle filtering process; the online fault detection of the new energy automobile battery to be tested is performed based on the health use weight of the new energy automobile battery and the battery health state after each cycle. The application analyzes the driving behavior and charging behavior of the new energy battery to evaluate the influence of the use on the battery damage, adjusts the particle filtering process, improves the battery health state estimation accuracy, and further improves the online fault detection effect of the new energy battery. BRIEF DESCRIPTION OF DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, and the advantages thereof, a brief introduction will be given to the drawings needed in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor based on these drawings.

[0046] Figure 1 A flowchart of a new energy battery online fault detection analysis method provided by an embodiment of the present application;

[0047] Figure 2 A flowchart of a health discharge index acquisition method provided by an embodiment of the present application;

[0048] Figure 3 A flowchart of a health charging index acquisition method provided by an embodiment of the present application;

[0049] Figure 4 A flow chart of a method for obtaining a healthy use weight is provided for an embodiment of the present application.

[0050] Figure 5 A system block diagram of an online fault detection and analysis system for a new energy battery is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0051] In order to further illustrate the technical means and effects taken by the present application to achieve the predetermined purposes, the following describes in detail the specific embodiments, structures, features and effects of the online fault detection and analysis method, device and system for a new energy battery according to the present application, with reference to the accompanying drawings and preferred embodiments. Different “one embodiment” or “another embodiment” in the following description do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0053] The following specifically describes the specific scheme of the online fault detection and analysis method, device and system for a new energy battery according to the present application, with reference to the accompanying drawings.

[0054] Please refer to Figure 1 which shows a flow chart of an online fault detection and analysis method for a new energy battery according to an embodiment of the present application, specifically including:

[0055] Step S1, obtaining the charging current curve and all discharge current curves of the new energy vehicle battery in each cycle process; obtaining all bad driving sub-sections in each discharge current curve and the battery health state of the new energy vehicle battery after each cycle ends.

[0056] It should be noted that one complete cycle process of the new energy vehicle battery refers to the battery power consumption from 100% to 0% and then full charging to 100%. Considering that in actual use, users usually do not charge until the power is consumed to 0%, and usually do not charge when the power is high. Therefore, the present embodiment defines the end of the current charging to the end of the next charging as one complete cycle.

[0057] In an embodiment of the present application, the charging current curve and all discharge current curves of the new energy vehicle battery in each cycle process are first obtained through the battery management system of the new energy vehicle. One cycle process includes one charging process and several driving processes, so the charging current curve of the new energy vehicle battery after connecting to the charging pile and the discharge current curve of the new energy vehicle battery in several driving processes can be obtained.

[0058] Wherein, there may be multiple emergency braking, frequent emergency acceleration or emergency deceleration and other bad driving behaviors during driving, all bad driving sub-sections in the discharge current curve corresponding to each driving process are obtained through the battery management system, so as to analyze and evaluate the health charging condition and the health discharging condition of the new energy vehicle battery in each cycle process, and then to predict and evaluate whether the new energy vehicle battery is faulty, thereby improving the fault evaluation accuracy.

[0059] The embodiment of the application further evaluates the battery health state of the new energy vehicle battery after each cycle, and provides a data basis for subsequent prediction of future battery health state. In a preferred embodiment of the application, the battery health state of the new energy vehicle battery after each cycle is evaluated based on a double exponential degradation model. The battery health state is a known technical index, which is a quantitative parameter of the battery state. The larger the battery health state, the healthier the battery, and the less likely the battery is to fail.

[0060] It should be noted that in other embodiments, the implementer can also evaluate the battery health state through the remaining capacity and internal impedance of the battery, which is prior art based on the battery management system to obtain the charging current curve and the discharge current curve, to obtain the current curve sub-section corresponding to the bad driving behavior, i.e. the bad driving sub-section, and to evaluate the battery health state based on the double exponential degradation model, which is not described here.

[0061] In another embodiment of the application, the implementer can also connect current sensors and voltage sensors to the new energy vehicle battery, and collect the voltage change curve and the current change curve of the battery in each cycle through the sensors; when the battery is charging, the current will instantaneously increase and then maintain constant current charging, then enter the constant voltage charging state, and finally the current slowly decreases; when the battery is discharging, both the current and the current will decrease; therefore, the charging and discharging stages of the battery can be distinguished by the change characteristics of the current and the voltage, so as to obtain the charging current curve and all the discharge current curves; then considering that when the vehicle accelerates, the battery discharge current increases and the voltage decreases accordingly; when the vehicle brakes and decelerates, the kinetic energy is recovered to charge the battery, the current increases sharply, and the voltage increases, and then returns to normal; based on these change characteristics, all bad driving sub-sections in each discharge current curve can be further screened out;

[0062] To improve the recognition efficiency, the implementer can also manually label the corresponding feature sub-section, build a sample set to train a classification model, input the current change curve and the voltage change curve collected by the sensor into the trained classification model, and the classification model automatically labels the charging current curve and all the discharge current curves in the current change curve, as well as all the bad driving sub-sections in each discharge current curve. It should be noted that the application of the classification model is a known technology, which is not described here.

[0063] Step S2, in each cycle process, the health discharge index of the new energy automobile battery is obtained according to all the bad driving sub-sections in each discharge current curve and the current change in each bad driving sub-section; the health charging index of the new energy automobile battery is obtained according to the deviation of the starting current in the charging current curve of all cycle processes from the preset low charging current and the deviation of the ending current from the preset overcharging current, combined with the overcharging time in each cycle process.

[0064] It is considered that in the driving process of the new energy automobile, the bad driving behavior of the driver will accelerate the degradation of the new energy automobile battery, for example, when the driver needs to provide higher current output of the battery during the rapid acceleration, the electrochemical reaction in the battery is intense, which further causes the battery to heat and accelerate the aging of the battery; for example, the frequent braking such as rapid braking will cause the battery to be unable to effectively recover energy, increase the burden of the battery, and further accelerate the aging of the battery.

[0065] It is also considered that the more the bad driving behavior or the longer the duration, the greater the damage to the battery, so the total duration of the bad driving sub-section in each discharge current curve can help to evaluate the health discharge index; at the same time, the bad degree of the bad driving behavior also has different influences on the damage to the battery, for example, the damage to the battery caused by the too fast deceleration speed is relatively higher than that caused by the slightly slow deceleration speed; the bad degree of the bad driving behavior can be evaluated by the current change, the greater the bad degree, the faster the current change, so as to evaluate the health discharge index.

[0066] Therefore, in each cycle process, the health discharge index of the new energy automobile battery is obtained according to the total duration of the bad driving sub-section in each discharge current curve and the current change in each bad driving sub-section; the health discharge index comprehensively evaluates all the bad driving behaviors in each cycle process to evaluate the discharge health condition of the new energy automobile battery, and the greater the health discharge index, the smaller the influence of the driving behavior on the damage to the new energy automobile battery in the cycle process.

[0067] Preferably, in one embodiment of the present application, the method for obtaining the health discharge index comprises:

[0068] Please refer to Figure 2 which shows a flow chart of a method for obtaining a health discharge index provided by one embodiment of the present application, and specifically comprises:

[0069] Step S201, in each discharge current curve of each cycle process, the time ratio of the total duration of the bad driving sub-section to the total duration of the discharge current curve is taken as the bad driving discharge time proportion; all the bad discharge current curves are selected from all the discharge current curves in each cycle process according to the bad driving discharge time proportion.

[0070] As an example, first, in each discharge current curve in each cycle process, the proportion of poor driving discharge time is obtained. The greater the proportion of poor driving discharge time, the more frequent or longer the duration of poor driving behavior in the driving process corresponding to each discharge current curve, and the greater the impact on the new energy battery.

[0071] Then, when the proportion of poor driving discharge time is greater than a preset proportion threshold such as 1 / 4, the discharge current curve corresponding to the proportion is taken as a poor discharge current curve. The poor discharge current curve corresponds to a driving process in which the poor driving behavior is more frequent or longer in duration. The more poor driving processes, the greater the impact of driving behavior on the new energy vehicle battery in the cycle process.

[0072] In other examples, the implementer can also set other preset proportion thresholds, which are not described here. The implementer can also directly filter the poor discharge current curve according to the total number of poor driving sub-sections. When the total number of poor driving sub-sections in each discharge current curve exceeds a preset number threshold such as 4, the discharge current curve is taken as a poor discharge current curve.

[0073] Step S202, the difference between the preset first number and the number of poor discharge current curves is taken as the health discharge weight of the new energy vehicle battery in each cycle process.

[0074] As an example, in each cycle process, the total number of discharge current curves in a preset proportion is taken as the preset first number, where the preset proportion is 0.25, and the implementer can also set it by himself. The smaller the preset proportion, the more strict the judgment of the poor driving behavior in the current cycle process, and the smaller the health discharge weight. When the difference between the preset first number and the number of poor discharge current curves is less than 0 and the smaller the difference, the more serious the poor driving behavior in the current cycle process, and the corresponding health discharge weight is negative. On the contrary, the new energy vehicle battery is less affected by the poor driving behavior in the current cycle process, and the health discharge weight is positive.

[0075] Step S203, in each cycle process, the poor driving discharge parameter of the new energy vehicle battery is obtained according to the proportion of poor driving discharge time of each discharge current curve and the change rate of current range in each poor driving sub-section of each discharge current curve.

[0076] The rate of change of the current range in each poor driving sub-section in each discharge current curve can reflect the degree of poor driving behavior, and the greater the rate of change, the faster the current changes and the greater the degree of poor driving behavior. The poor driving discharge time proportion of all discharge current curves in each cycle process is further integrated to evaluate the poor discharge parameter of the new energy vehicle battery. The poor discharge parameter reflects the degree of poor driving behavior in each cycle process, and the greater the degree of poor driving behavior, the greater the impact on the battery. The health discharge parameter is further evaluated in combination with the health discharge weight.

[0077] In a preferred embodiment of the present application, the method for obtaining the poor discharge parameter comprises:

[0078] The ratio between the current range in each poor driving sub-section and the corresponding time interval of the current range is taken as the rate of change of the current range in the corresponding poor driving sub-section. The rate of change of the current range in all poor driving sub-sections in each discharge current curve is integrated to obtain the poor driving discharge parameter of each discharge current curve.

[0079] The poor driving discharge time proportion and the poor driving discharge parameter of each discharge current curve are fused to obtain the poor discharge parameter of each discharge current curve. The poor discharge parameters of all discharge current curves in each cycle process are integrated to obtain the poor discharge parameter of the new energy vehicle battery in each cycle process.

[0080] As an example, the calculation formula of the poor discharge parameter is as follows for any cycle process:

[0081] wherein D is the poor discharge parameter of the new energy vehicle battery in the cycle process; j is the serial number of the discharge current curve in the cycle process; J is the total number of the discharge current curves in the cycle process; v is the serial number of the poor driving sub-section in each discharge current curve; V j is the total number of the poor driving sub-sections in the jth discharge current curve in the cycle process; T j ′ is the total duration of the jth discharge current curve in the cycle process; T j is the total duration of the poor driving sub-section in the jth discharge current curve in the cycle process; is the poor driving discharge time proportion of the jth discharge current curve in the cycle process; A v is the current range in the vth poor driving sub-section of the jth discharge current curve in the cycle process; T is the corresponding time interval of the current range in the vth poor driving sub-section of the jth discharge current curve in the cycle process; is the rate of change of the current range in the vth poor driving sub-section of the jth discharge current curve in the cycle process; a bad driving discharge parameter of the jth discharge current curve in the cycle process; a bad driving discharge parameter of the jth discharge current curve in the cycle process.

[0082] In the example, the greater the current range change rate in the bad driving sub-section, the more intense the current change, the greater the badness of the current bad driving behavior, and the greater the impact on the battery. The sum of the current range change rates in all bad driving sub-sections in each discharge current curve is evaluated as a bad driving discharge parameter, which is then multiplied and combined with the bad driving discharge time proportion to obtain a bad driving discharge parameter of each discharge current curve. The greater the bad driving discharge parameter, the greater the impact of the bad driving behavior corresponding to the driving process on the battery. The sum of the bad driving discharge parameters of all discharge current curves is further evaluated to assess the impact of the bad driving behavior in the corresponding cycle process on the battery. The greater the sum, the greater the impact of the bad driving parameter on the battery.

[0083] It should be noted that there is a current change in the bad driving sub-section, so the current range is not 0, and the corresponding time interval is also not 0. However, there may be no bad driving sub-section in each discharge current curve, so the bad driving discharge time proportion may be 0.

[0084] In other examples, the implementer can also evaluate the current change by using the sum of the current range and the current variance in each bad driving sub-section. The greater the sum, the more intense the current change, and the greater the bad impact of the bad driving behavior. The sum is used instead of the change rate of the current range to evaluate the bad driving discharge parameter in each cycle process.

[0085] In step S204, the bad driving discharge parameter is weighted using the healthy discharge weight, the weighted result is normalized, and the normalized result is used as the healthy discharge index of the new energy vehicle battery in each cycle process.

[0086] As an example, in any cycle process, the healthy discharge weight is multiplied and combined with the bad driving discharge parameter to obtain a weighted result, the weighted result is linearly normalized, and the normalized result is used as the healthy discharge index. When the healthy discharge weight is negative and smaller, and the bad driving discharge parameter is greater, the weighted result is smaller, the normalized result is smaller, and the healthy discharge index is smaller. Conversely, when the healthy discharge weight is positive and greater, and the bad driving discharge parameter is relatively smaller, the weighted result is relatively greater, and the healthy discharge index is relatively greater.

[0087] In another embodiment of the present application, the implementer can also directly linearly normalize the health discharge weight value obtained in step S202, and take the normalized value as the health discharge parameter; when the health discharge weight value is negative, the smaller the normalized value, the smaller the health discharge parameter; then perform negative correlation mapping on the unhealthy discharge parameter obtained in step S203, such as taking it as x in the exponential function exp(-x) with the natural constant e as the base number, adjusting the logic so that the smaller the unhealthy discharge parameter, the larger the negative correlation mapping value; then multiply and combine the health discharge parameter and the negative correlation mapping value of the unhealthy discharge parameter to obtain the health discharge index of the new energy vehicle battery in each cycle process.

[0088] It is considered that unhealthy charging behaviors also accelerate the degradation of new energy vehicle batteries, such as overcharging, i.e., charging the battery to 100% for a long time and keeping it for too long, which may cause instability of the internal chemical substances of the battery, thereby accelerating the aging of the battery; and charging the battery after discharging it to a very low power, hereinafter referred to as low charging, will also cause damage to the battery, and charging after excessive discharging will affect the capacity and cycle life of the battery;

[0089] It is also considered that when the new energy battery is overcharged or low-charged, the charging pile will intelligently adjust the charging current according to the power of the battery and the demand of the battery management system; when the battery power is low, the charging current is large; and when the battery is close to full power, the charging current will gradually decrease; therefore, whether there is an unhealthy charging behavior such as low charging or overcharging in the current charging process can be judged based on the starting charging current and the ending charging current in the charging current curve, and the health charging index is evaluated;

[0090] Therefore, in the embodiment of the present application, the health charging index of the new energy vehicle battery is obtained according to the deviation of the starting current relative to the preset low charging current and the deviation of the ending current relative to the preset overcharging current in the charging current curve of all cycle processes, combined with the overcharging time in each cycle process; the health charging index comprehensively evaluates the charging health of the new energy vehicle battery, and the larger the health charging index, the smaller the impact of the charging behavior on the degradation of the new energy vehicle battery.

[0091] Preferably, in one embodiment of the present application, the method for obtaining the health charging index comprises:

[0092] Please refer to Figure 3 which shows a flow chart of a method for obtaining a health charging index provided by an embodiment of the present application, and specifically comprises:

[0093] Step S211, in each charging current curve, the overcharge duration is obtained, the low charge reference current is determined according to the size relationship between the starting current and the preset low charging current, and it is judged whether the charging current curve is a low charging current curve, the overcharge reference current is determined according to the size relationship between the ending current and the preset overcharge current, and it is judged whether the charging current curve is an overcharge current curve.

[0094] In a preferred embodiment of the present application, the method for obtaining the overcharge duration comprises:

[0095] The time point at which the preset overcharge current is first reached in the charging current curve is taken as the overcharge time point, and the duration between the overcharge time point and the time point corresponding to the ending current is taken as the overcharge duration.

[0096] In a preferred embodiment of the present application, the method for obtaining the low charge reference current and the overcharge reference current comprises:

[0097] In each charging current curve, if the starting current is greater than or equal to the preset low charging current, the preset low charging current is taken as the low charge reference current; if the starting current is less than the preset low charging current, the starting current is taken as the low charge reference current; if the ending current is greater than the preset overcharge current, the ending current is taken as the overcharge reference current, and if the ending current is less than or equal to the preset overcharge current, the ending current is taken as the overcharge reference current.

[0098] As an example, the judgment formula of the low charge reference current is: Wherein, i is the serial number of the charging current curve; L i is the low charge reference current of the i th charging current curve; A0 is the preset low charging current; A i is the starting current of the i th charging current curve.

[0099] As an example, the judgment formula of the overcharge reference current is: Wherein, i is the serial number of the charging current curve; H i is the overcharge reference current of the i th charging current curve; A1 is the preset overcharge current; A ′ i is the ending current of the i th charging current curve.

[0100] In a preferred embodiment of the present application, the method for judging the low charging current curve and the overcharge current curve comprises:

[0101] If the starting current in the charging current curve is less than the preset low charging current, the charging current curve is taken as the low charging current curve; if the starting current in the charging current curve is greater than the preset overcharge current, the charging current curve is taken as the overcharge current curve;

[0102] It should be noted that the preset low charging current is the charging current corresponding to 20% of the new energy battery power, and the preset overcharging current is the charging current corresponding to 90% of the new energy battery power. Different types of charging piles such as fast charging or slow charging, different battery capacities, and the design of the battery management system will all affect the size of the preset low charging current and the preset overcharging current. Therefore, the specific charging current value needs to be determined according to the actual situation, and will not be described here.

[0103] It should be noted that in other embodiments, the implementer can also directly use the battery management system to determine the low charging and overcharging behaviors, and then determine the overcharging duration, the low charging current curve, and the overcharging current curve. Since there may be both low charging behavior and overcharging behavior in one charging process, the low charging current curve and the overcharging current curve may correspond to the same charging current curve.

[0104] In step S212, the difference between the preset second number and the total number of low charging current curves is taken as a first health charging weight; the sum of the absolute values of the differences between the low charging reference currents in all charging current curves and the preset low charging current is taken as a bad low charging parameter; and the first health charging weight is used to weight the bad low charging parameter, and the weighted result is taken as a first health charging index.

[0105] As an example, in each cycle, a preset proportion of the total number of charging current curves is taken as the preset second number, where the preset proportion is 0.5, and the implementer can also set it by himself. The smaller the preset proportion, the stricter the judgment of low charging bad behavior, and the smaller the first health charging weight may be. When the difference between the preset second number and the total number of low charging current curves is less than 0 and the smaller it is, the more serious the low charging bad behavior is considered to be, and the corresponding first health charging weight is negative. On the contrary, the new energy vehicle battery is considered to be less affected by the loss of low charging behavior, and the first health charging weight is positive.

[0106] The calculation formula of the first health charging index is: Wherein, F1 is the first health charging index; i is the serial number of the charging current curve; I is the total number of charging current curves, which is also the number of cycles up to the current; 0.5xI is the preset second number; I0 is the total number of low charging current curves; A0 is the preset low charging current; A i is the starting current of the i-th charging current curve; is the bad low charging index.

[0107] In the example, when the starting current is less than the preset low charging current, it is considered as low charging behavior, and the starting current is taken as the low charging reference current, so that the absolute value of the difference between the low charging reference current and the preset low charging current in each charging current curve is greater than 0, otherwise the absolute value of the difference is 0, and then the bad low charging index is obtained; the greater the bad low charging index, the more likely the low charging behavior has a greater impact on battery loss, and the first health charging weight is further used to weight the bad low charging parameter; when the first health charging weight is negative and the bad low charging parameter is greater, it indicates that the first health charging index is smaller.

[0108] In another example, the implementer can also directly linearly normalize the first health charging weight, so that the value range is adjusted to 0 to 1, and when the first health charging weight is negative, the smaller the normalized value is; then the bad low charging parameter is negatively correlated, such as taking it as x in the exponential function exp(-x) with natural constant e as the base number, adjusting the logic so that the smaller the bad low charging parameter is, the greater the negatively correlated mapping value is; then the normalized value of the first health charging weight and the negatively correlated mapping value of the bad low charging parameter are multiplied and combined to obtain the first health charging index.

[0109] In step S213, the difference between the preset second quantity and the total number of overcharging current curves is taken as the second health charging weight; the sum of the absolute values of the differences between the overcharging reference currents in all charging current curves and the preset overcharging current is calculated, and the product of the sum of the absolute values of the differences and the overcharging duration is taken as the bad overcharging parameter; the second health charging weight is used to weight the bad overcharging parameter, and the weighted result is taken as the second health charging index.

[0110] As an example, when the difference between the preset second quantity and the total number of overcharging current curves is less than 0 and smaller, it is considered that the overcharging bad behavior is more serious, and the corresponding second health charging weight is negative; on the contrary, it is considered that the new energy automobile battery is less affected by the loss of overcharging behavior, and the second health charging weight is positive.

[0111] The calculation formula of the second health charging index is: Wherein, F2 is the second health charging index; i is the serial number of the charging current curve; I is the total number of charging current curves, which is also the current cycle number; 0.5xI is the preset second quantity; I1 is the total number of overcharging current curves; A1 is the preset overcharging current; A ′ i is the ending current of the i-th charging current curve; T i is the overcharging duration of the i-th charging current curve. is the bad overcharging index.

[0112] In the example, when the end current is greater than the preset overcharge current, it is considered to belong to overcharge behavior, and the end current is taken as the overcharge reference current, so that the absolute value of the difference between the overcharge reference current in each charging current curve and the preset overcharge current is greater than 0, otherwise the absolute value of the difference is 0, and then the bad overcharge index is obtained in combination with the overcharge duration; the greater the bad overcharge index, the greater the impact of overcharge behavior on battery loss, and the second health charging weight is further used to weight the bad overcharge parameter; when the second health charging weight is negative and the bad overcharge parameter is greater, it means that the second health charging index is smaller.

[0113] In another example, the implementer can also directly linearly normalize the second health charging weight, so that the value range is adjusted to 0 to 1, and when the second health charging weight is negative, the normalized value is smaller; then the bad overcharge parameter is negatively correlated, such as taking it as x in the exponential function exp(-x) with natural constant e as the base number, adjusting the logic so that the smaller the bad overcharge parameter, the greater the negatively correlated mapping value; then the normalized value of the second health charging weight and the negatively correlated mapping value of the bad overcharge parameter are multiplied and combined to obtain the second health charging index.

[0114] Step S214, fusing the first health charging index and the second health charging index to obtain the health charging index of the new energy automobile battery.

[0115] As an example, considering that the first health charging index and the second health charging index can be negative, the first health charging index and the second health charging index are added and combined, and the sum value is linearly normalized to obtain the health charging index; in other examples, the first health charging index and the second health charging index can be normalized first, and then the two normalized values are positively correlated by using other basic mathematical operations, such as addition or weighted summation, and the like, which will not be described here.

[0116] Step S3, according to the fluctuation changes of the health discharge index and the health charging index in all cycles, in combination with the health discharge index and the health charging index of the new energy automobile battery in the current cycle, the health use weight of the new energy automobile battery is obtained.

[0117] In view of the fact that the adverse driving behavior or adverse charging behavior in different cycles may have contingency, that is, the damage effect on the battery is not persistent, the aging speed of the battery can be relatively slow; therefore, the embodiment of the application further combines the fluctuation changes of the healthy discharge index and the healthy charging index in all cycles to evaluate the healthy use weight of the new energy vehicle battery; the healthy use weight reflects the persistence of the damage of the adverse driving behavior or adverse charging behavior on the new energy battery, and further reflects the influence of the adverse behavior on the health status of the new energy vehicle battery; if the damage effect is higher and the persistence is also stronger, the healthy use weight is lower; if the damage effect is lower for a long time, the battery aging speed is slower, and the healthy use weight is higher.

[0118] Preferably, in an embodiment of the application, the method for obtaining the healthy use weight comprises:

[0119] Referring to Figure 4 , a flowchart of a method for obtaining a healthy use weight provided by an embodiment of the application is shown, which specifically comprises:

[0120] In step S301, a healthy discharge fluctuation parameter is obtained according to the range of the healthy discharge index in all cycles and the deviation of each healthy discharge index from the average level, and a healthy charging fluctuation parameter is obtained according to the range of the healthy charging index in all cycles and the deviation of each healthy charging index from the average level.

[0121] In view of the fact that the range can generally reflect the fluctuation change of data, and the deviation of data from the average level can further reflect the fluctuation of data, based on this, the fluctuation of the healthy discharge index and the healthy charging index can be preliminarily evaluated to obtain the healthy discharge fluctuation parameter and the healthy charging fluctuation parameter; the two fluctuation parameters respectively reflect the behavior stability of the corresponding driving behavior or charging behavior, so as to subsequently evaluate the habit of the healthy use behavior and determine the healthy use weight.

[0122] As an example, the calculation formula of the healthy discharge fluctuation parameter is: wherein D1 is the healthy discharge fluctuation parameter; L1 is the range of all healthy discharge indexes; e is the serial number of the healthy discharge index; E is the total number of the healthy discharge indexes; X1 e is the e-th healthy discharge index; is the average of all healthy discharge indexes.

[0123] In the calculation formula of the healthy discharge index, the larger the range is, the greater the fluctuation of the healthy discharge index is, and the lower the possibility of the tendency of using the new energy battery is; by using The greater the sum of the deviation and the value, the greater the fluctuation of the health discharge index, and the lower the possibility of the tendency of the health use of the new energy battery. The health discharge fluctuation parameter is obtained by multiplying the range and the deviation.

[0124] In another example, the implementer can also use the variance or standard deviation of all health discharge indexes as the health discharge fluctuation parameter, wherein the variance, the standard deviation and the range are all prior art, and details are not described herein.

[0125] Similarly, the health charge index can be obtained based on the obtaining method of the health discharge index, and details are not described herein.

[0126] In step S302, the health discharge fluctuation parameter and the health charge fluctuation parameter are fused, the fusion result is negatively correlated, and the negatively correlated mapping result is used as the cycle stability coefficient of the new energy vehicle battery.

[0127] As an example, the health discharge fluctuation parameter and the health charge fluctuation parameter are added, the sum is used as x in the exponential function exp(-x) with the natural constant e as the base number, the logic is adjusted and normalized, and the cycle stability coefficient is obtained; the greater the cycle stability coefficient, the more the cycle use behavior of the new energy vehicle battery tends to be stable.

[0128] In step S303, the health use weight of the new energy vehicle battery is obtained according to the health discharge index, the health charge index and the cycle stability coefficient; the health discharge index, the health charge index and the cycle stability coefficient are positively correlated with the health use weight.

[0129] As an example, the calculation formula of the health use weight is: R=F*norm(X1+X2); wherein R is the health use weight; F is the cycle stability coefficient; X1 is the health discharge index; X2 is the health charge index; and norm() is a normalization function.

[0130] In the calculation formula of the health use weight, the health discharge index and the health charge index are added and combined, and then linearly normalized, and the normalized value and the cycle stability coefficient are multiplied and combined to obtain the health use weight; since the value range of the normalized value and the cycle stability coefficient is 0 to 1, the value range of the health use weight is also 0 to 1, wherein the greater the cycle stability coefficient and the greater the normalized value, the greater the health use weight.

[0131] In step S4, the online fault detection of the new energy vehicle battery to be tested is performed based on the health use weight of the new energy vehicle battery and the battery health state after each cycle ends.

[0132] After the health usage weight of the new energy vehicle battery is obtained, the health state of the battery can be combined to perform online fault detection on the new energy vehicle battery to be tested.

[0133] Preferably, in one embodiment of the present application, considering that particle filtering is a recursive estimation method based on random sampling, which is widely used in state estimation and prediction of dynamic systems, especially in the case of nonlinear and non-Gaussian noise, and performs well, the particle filtering algorithm can be used to perform online fault detection on the new energy vehicle battery to be tested, and the specific method includes:

[0134] In the current cycle, based on the particle filtering algorithm and the health state of the new energy vehicle battery at the end of each cycle, the future battery health state is predicted, and if the future battery health state is lower than the preset health threshold, a fault warning is performed.

[0135] The method for predicting the future battery health state includes:

[0136] In the prediction process based on the particle filtering algorithm, the health usage weight of the new energy vehicle battery is used to weight the battery health state of each particle, and the corresponding weighted result is used as the corrected battery health state of the corresponding particle; any particle in the initial particle set is taken as a target particle, and the particle with the smallest difference between the battery health state and the corrected battery health state of the target particle is taken as the reference particle of the target particle, and the weight of the reference particle is taken as the corrected weight of the target particle; based on the corrected weight and the corresponding corrected battery health state of each particle after resampling, the future battery health state is obtained.

[0137] As an example, in the prediction process based on the particle filtering algorithm, the health usage weight of the new energy vehicle battery is multiplied by the battery health state of each particle to obtain the corrected battery health state of the corresponding particle, and the smaller the health usage weight, the smaller the corrected battery health state, the weight of the particle is adjusted to be closest to the weight of the particle, thereby improving the state estimation accuracy; after adjusting the weights of all particles, the particles are resampled to predict the future battery health state; the preset health threshold is set to 80%, and when the future battery health state is lower than 80%, it is considered that the state of the new energy vehicle battery is not good and the battery is faulty.

[0138] It should be noted that predicting the battery health state based on the particle filtering algorithm is a well-known prior art to those skilled in the art, and will not be described here.

[0139] The present application also provides an online fault detection analysis device for a new energy battery, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the online fault detection analysis method for a new energy battery described in steps S1-S4 when executing the computer program.

[0140] The application further provides an online fault detection analysis system for a new energy battery. Figure 5 The application further provides an online fault detection analysis system for a new energy battery.

[0141] The battery cycle data acquisition module 101 is used to acquire the charging current curve and all the discharging current curves of the new energy vehicle battery in each cycle process, acquire all the bad driving sub-sections in each discharging current curve, and acquire the battery health state of the new energy vehicle battery after each cycle.

[0142] The battery cycle data analysis module 102 is used to acquire the health discharging index of the new energy vehicle battery according to the total time length of the bad driving sub-sections in each discharging current curve and the current change in each bad driving sub-section in each cycle process, acquire the health charging index of the new energy vehicle battery according to the deviation of the initial current from the preset low charging current and the deviation of the end current from the preset overcharging current in the charging current curve of all the cycle processes, and acquire the health use weight of the new energy vehicle battery according to the fluctuation change of the health discharging index and the health charging index in all the cycle processes and the health discharging index and the health charging index of the new energy vehicle battery in the current cycle process.

[0143] The battery fault detection module 103 is used to perform online fault detection on the new energy vehicle battery to be tested based on the health use weight of the new energy vehicle battery and the battery health state in the current cycle process.

[0144] The application first acquires the charging current curve and all the discharging current curves of the new energy vehicle battery in each cycle process, acquires all the bad driving sub-sections in each discharging current curve, and acquires the battery health state of the new energy vehicle battery after each cycle, then analyzes the health discharging index and the health charging index of the new energy vehicle battery in each cycle process to acquire the health use weight of the new energy vehicle battery, and finally performs online fault detection on the new energy vehicle battery to be tested based on the health use weight of the new energy vehicle battery and the battery health state after each cycle.

[0145] It is to be noted that the sequential order of the above-described embodiments of the present application only for the purpose of description, but not the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or can be advantageous.

[0146] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the difference from other embodiments.

Claims

1. An online fault detection and analysis method for new energy batteries, characterized in that, The method includes: Obtain the charging current curve and all discharging current curves of the new energy vehicle battery in each cycle; obtain all bad driving segments in each discharging current curve, and the battery health status of the new energy vehicle battery after each cycle. In each cycle, the healthy discharge index of the new energy vehicle battery is obtained based on all the bad driving sub-segments in each discharge current curve and the current changes in each bad driving sub-segment; the healthy charging index of the new energy vehicle battery is obtained based on the deviation of the starting current from the preset low charging current and the deviation of the ending current from the preset overcharging current in the charging current curves of all cycles, combined with the overcharging time in each cycle. Based on the fluctuations of the healthy discharge index and the healthy charging index during all cycles, and combined with the healthy discharge index and the healthy charging index of the new energy vehicle battery during the current cycle, the healthy usage weight of the new energy vehicle battery is obtained. Based on the health usage weight of the new energy vehicle battery and the battery health status at the end of each cycle, online fault detection is performed on the new energy vehicle battery under test.

2. The online fault detection and analysis method for new energy batteries according to claim 1, characterized in that, The method for obtaining the health discharge index includes: In each discharge current curve of each cycle, the ratio of the total duration of the bad driving segment to the total duration of the discharge current curve is taken as the bad driving discharge time percentage; based on the bad driving discharge time percentage, all bad discharge current curves are selected from all discharge current curves in each cycle. The difference between the preset first quantity and the number of the defective discharge current curves is used as the healthy discharge weight of the new energy vehicle battery in each cycle. In each cycle, based on the proportion of the discharge time of the poor driving in each discharge current curve, and combined with the rate of change of the current range in each poor driving sub-segment of each discharge current curve, the poor discharge parameters of the new energy vehicle battery are obtained. The undesirable discharge parameters are weighted using the healthy discharge weights, and the weighted results are normalized. The normalized results are used as the healthy discharge index of the new energy vehicle battery in each cycle.

3. The online fault detection and analysis method for new energy batteries according to claim 2, characterized in that, The method for obtaining the adverse discharge parameters includes: The ratio between the current range within each bad driving segment and the corresponding time interval of the current range is taken as the rate of change of the current range within the corresponding bad driving segment; by combining the rate of change of the current range within all bad driving segments in each discharge current curve, the bad driving discharge parameters of each discharge current curve are obtained. By integrating the proportion of the discharge time due to poor driving and the discharge parameters due to poor driving of each discharge current curve, the poor discharge parameters of each discharge current curve are obtained; by combining the poor discharge parameters of all discharge current curves in each cycle, the poor discharge parameters of the new energy vehicle battery in each cycle are obtained.

4. The online fault detection and analysis method for new energy batteries according to claim 1, characterized in that, The method for obtaining the health charging index includes: In each of the charging current curves, the overcharge duration is obtained, the low-charge reference current is determined based on the relationship between the starting current and the preset low-charge current, and it is determined whether the charging current curve is a low-charge current curve. The overcharge reference current is determined based on the relationship between the ending current and the preset overcharge current, and it is determined whether the charging current curve is an overcharge current curve. The difference between the preset second quantity and the total number of low charging current curves is used as the first healthy charging weight; the sum of the absolute values ​​of the differences between the low charging reference current and the preset low charging current in all the charging current curves is used as the poor low charging parameter; the poor low charging parameter is weighted using the first healthy charging weight, and the weighted result is used as the first healthy charging index. The difference between the preset second quantity and the total number of the overcharge current curves is used as the second healthy charging weight; the sum of the absolute values ​​of the differences between the overcharge reference current and the preset overcharge current in all the charging current curves is calculated, and the product of the sum of the absolute values ​​of the differences and the overcharge duration is used as the poor low-charge parameter; the poor low-charge parameter is weighted using the second healthy charging weight, and the weighted result is used as the second healthy charging index; By combining the first healthy charging index and the second healthy charging index, a healthy charging index for new energy vehicle batteries is obtained.

5. The online fault detection and analysis method for a new energy battery according to claim 4, characterized in that, The method for obtaining the overcharge duration, the method for obtaining the low-charge reference current and the overcharge reference current, and the method for determining the low-charge current curve and the overcharge current curve include: In each of the charging current curves, if the starting current is greater than or equal to the preset low charging current, the preset low charging current is used as the low charging reference current; if the starting current is less than the preset low charging current, the starting current is used as the low charging reference current; if the ending current is greater than the preset overcharging current, the ending current is used as the overcharging reference current; if the ending current is less than or equal to the preset overcharging current, the ending current is used as the overcharging reference current. If the starting current in the charging current curve is less than the preset low charging current, the charging current curve is used as the low charging current curve; if the starting current in the charging current curve is greater than the preset overcharging current, the charging current curve is used as the overcharging current curve. The time point at which the preset overcharge current is first reached in the charging current curve is taken as the overcharge time point, and the time between the overcharge time point and the time point corresponding to the end current is taken as the overcharge duration.

6. The online fault detection and analysis method for new energy batteries according to claim 1, characterized in that, The method for obtaining the health usage weight includes: Based on the range of the healthy discharge index in all cycles and the deviation of each healthy discharge index from the average level, the healthy discharge fluctuation parameter is obtained. Based on the range of the health charging index in all cycles and the deviation of each health charging index from the average level, the health charging fluctuation parameter is obtained. The healthy discharge fluctuation parameters and the healthy charging fluctuation parameters are fused together, and the fusion result is negatively correlated and mapped. The negative correlation mapping result is used as the cycle stability coefficient of the new energy vehicle battery. The healthy use weight of the new energy vehicle battery is obtained based on the healthy discharge index, the healthy charging index, and the cycle stability coefficient; the healthy discharge index, the healthy charging index, and the cycle stability coefficient are all positively correlated with the healthy use weight.

7. The online fault detection and analysis method for new energy batteries according to claim 1, characterized in that, The method for online fault detection of the battery of the new energy vehicle under test includes: During the current cycle, based on the particle filter algorithm and the battery health status of the new energy vehicle battery after each cycle, the future battery health status is predicted; if the future battery health status is lower than the preset health threshold, a fault warning is issued. The method for predicting future battery health includes: In the prediction process based on the particle filter algorithm, the battery health state of each particle is weighted using the health usage weight of the new energy vehicle battery, and the corresponding weighted result is used as the corrected battery health state of the corresponding particle. Taking any particle as the target particle, the particle whose battery health state is the smallest difference from the corrected battery health state of the target particle is used as the reference particle of the target particle, and the weight of the reference particle is used as the corrected weight of the target particle. Based on the corrected weight of each particle after resampling and the corresponding corrected battery health state, the future battery health state is obtained.

8. The online fault detection and analysis method for new energy batteries according to claim 1, characterized in that, The method for obtaining the battery health status includes: The battery health status of new energy vehicle batteries is assessed at the end of each cycle based on a bi-exponential degradation model.

9. An online fault detection and analysis device for new energy batteries, characterized in that, The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the online fault detection and analysis method for a new energy battery as described in any one of claims 1 to 8.

10. An online fault detection and analysis system for new energy batteries, characterized in that, The system includes: Battery Cycle Data Acquisition Module: Used to acquire the charging current curve and all discharging current curves of the new energy vehicle battery in each cycle; acquire all bad driving segments in each discharging current curve, and the battery health status of the new energy vehicle battery after each cycle. Battery Cycle Data Analysis Module: In each cycle, based on the total duration of the undesirable driving segment in each discharge current curve and the current change within each undesirable driving segment, a healthy discharge index of the new energy vehicle battery is obtained; based on the deviation of the starting current from the preset low charging current and the deviation of the ending current from the preset overcharging current in the charging current curves of all cycles, combined with the overcharging duration in each cycle, a healthy charging index of the new energy vehicle battery is obtained; based on the fluctuations of the healthy discharge index and the healthy charging index in all cycles, combined with the healthy discharge index and the healthy charging index of the new energy vehicle battery in the current cycle, a healthy usage weight of the new energy vehicle battery is obtained. Battery Fault Detection Module: Based on the health usage weight of the new energy vehicle battery and the battery health status in the current cycle, the module performs online fault detection on the new energy vehicle battery under test.

Images