Vehicle fault monitoring method and device and nonvolatile storage medium

Abstract

The invention discloses a vehicle fault monitoring method and device and a nonvolatile storage medium. The new energy vehicle alarm prediction method comprises the steps of obtaining target vehicle data of a new energy vehicle in a target time period; based on the target vehicle data, a vehicle fault prediction model is adopted to obtain the fault prediction probability of the new energy vehicle after the target time period, the vehicle fault prediction model is obtained based on a vehicle data set through machine learning, and the vehicle data set comprises multiple groups of vehicle data collected by multiple fault vehicles before the fault time period; the fault time period is the time period when the plurality of fault vehicles fail. The fault prediction probabilities respectively correspond to the plurality of groups of vehicle data. The technical problem that the safety of personnel in the vehicle cannot be effectively guaranteed due to the fact that whether the vehicle is abnormal or not is detected in real time in the related technology and an alarm is given according to the abnormal result is solved.

Description

Technical Field

[0001] The present invention relates to the field of vehicle alarm prediction, and in particular to a vehicle fault monitoring method, device and non-volatile storage medium. Background Art

[0002] New energy vehicles are equipped with power batteries. During use, if the power battery temperature is too high, or the power battery is charged or discharged quickly, it may cause safety problems for the vehicle, so new energy vehicles need to be monitored in real time.

[0003] In the related art, the detection of new energy vehicles is based on the real-time data collected to determine whether there is an abnormality, wherein the real-time data is data collected within a short time interval with seconds or minutes as the boundary, and an alarm is issued when an abnormality exists. However, this method is to remind the people in the vehicle whether there is an abnormality. When the test result obtained is that there is an abnormality, the vehicle has already failed, which may seriously affect the safety of the people in the vehicle.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a vehicle fault monitoring method, device and non-volatile storage medium to at least solve the technical problem that the safety of people in the vehicle cannot be effectively guaranteed due to the real-time detection of whether there is an abnormality in the vehicle in the related technology and the alarm when the abnormality exists.

[0006] According to one aspect of an embodiment of the present invention, a vehicle fault monitoring method is provided, comprising: acquiring target vehicle data of a new energy vehicle within a target time period; based on the target vehicle data, using a vehicle fault prediction model to obtain a fault prediction probability of the new energy vehicle after the target time period, wherein the vehicle fault prediction model is based on a vehicle data set and is obtained through machine learning, the vehicle data set comprising multiple groups of vehicle data collected from multiple faulty vehicles before the fault time period, and the fault prediction probabilities corresponding to the multiple groups of vehicle data, respectively, and the fault time period is a time period when faults occur in the multiple faulty vehicles.

[0007] Optionally, before the target vehicle data is based on and a vehicle fault prediction model is used to obtain the fault prediction probability of the new energy vehicle after the target time period, the method also includes: obtaining a predetermined number of groups of vehicle data among the multiple groups of vehicle data in the following manner: dividing the historical vehicle data of any faulty vehicle among the multiple faulty vehicles into a predetermined number of parts according to the data generation period to obtain the predetermined number of vehicle data; obtaining the multiple groups of vehicle data in the manner of obtaining the predetermined number of groups of vehicle data, wherein the number of groups of the multiple groups of vehicle data is less than the predetermined number of groups.

[0008] Through the above method, the historical data of the faulty vehicle is divided into multiple segments by time, and a predetermined number of groups of data are selected as training samples. This can achieve data dimensionality reduction, help reduce the computational burden of model training, and at the same time ensure the amount of data required for model training, so that the model can effectively learn and predict faults under limited computing resources.

[0009] Optionally, before the failure prediction probability of the new energy vehicle after the target time period is obtained by using a vehicle failure prediction model based on the target vehicle data, the method also includes: obtaining the failure prediction probabilities corresponding to a predetermined number of groups of vehicle data among the multiple groups of vehicle data in the following manner: obtaining the data generation time periods corresponding to the predetermined number of groups of vehicle data; determining the failure prediction probabilities corresponding to the predetermined number of groups of vehicle data based on the data generation time periods corresponding to the predetermined number of groups of vehicle data; obtaining the failure prediction probabilities corresponding to the multiple groups of vehicle data by using the method of obtaining the failure prediction probabilities corresponding to the predetermined number of groups of vehicle data.

[0010] By obtaining the time label of each set of data in the above way, the predicted probability of each set of data before the failure occurs can be calculated based on the historical failure data and time information. This can provide the subsequent model training process with information about the correlation between data and the probability of failure, so that the model can learn to predict the trend from normal to failure.

[0011] Optionally, when the vehicle data set also includes the fault types corresponding to the multiple groups of vehicle data, the method also includes: based on the target vehicle data, using the vehicle fault prediction model to obtain the fault prediction probability and fault type of the new energy vehicle after the target time period.

[0012] Through the above methods, fault prediction not only has the fault prediction probability of possible faults, but also has the possible fault types of corresponding faults. In this way, real-time monitoring of the vehicle can be achieved, thereby reminding the vehicle driver to deal with possible problems of the vehicle in time and effectively ensuring the safety of the people in the vehicle.

[0013] Optionally, before obtaining the fault prediction probability and fault type of the new energy vehicle after the target time period by using the vehicle fault prediction model based on the target vehicle data, the method also includes: acquiring fault vehicle data corresponding to the multiple faulty vehicles collected within the fault time period; determining the fault types and vehicle codes corresponding to the multiple faulty vehicles based on the fault vehicle data corresponding to the multiple faulty vehicles; determining the fault types corresponding to the multiple faulty vehicles within the historical time period based on the fault types and vehicle codes corresponding to the multiple faulty vehicles; and determining the fault types corresponding to the multiple groups of vehicle data based on the fault types corresponding to the multiple faulty vehicles.

[0014] Through the above method, the characteristic values ​​of vehicle data in multiple first time periods are effectively extracted, and the above characteristic values ​​are matched with the vehicle fault types, and the fault prediction probabilities are matched accordingly, so as to effectively predict the target vehicle data in the target time period, obtain the fault prediction probability and fault type that may occur after the target time period, and improve the accuracy of the prediction.

[0015] Optionally, the method of determining the fault types respectively corresponding to the multiple groups of vehicle data based on the fault types respectively corresponding to the multiple faulty vehicles includes: obtaining the fault types respectively corresponding to a predetermined number of groups of vehicle data among the multiple groups of vehicle data in the following manner: dividing the historical vehicle data of any faulty vehicle among the multiple faulty vehicles into a predetermined number of portions according to the data generation period to obtain the predetermined number of vehicle data; determining the fault type corresponding to any faulty vehicle as the fault type respectively corresponding to the predetermined number of vehicle data; and obtaining the fault types respectively corresponding to the multiple groups of vehicle data by adopting the method of obtaining the fault types respectively corresponding to the predetermined number of groups of vehicle data.

[0016] Through the above methods, the model can not only predict the probability of fault occurrence, but also identify the type of fault, thereby providing more specific and useful warning information. Through time series analysis and fault type labeling, the model's learning ability and prediction accuracy can be enhanced, and the practical value of fault warning can be improved.

[0017] Optionally, after the failure prediction probability of the new energy vehicle in the target time period is obtained by using a vehicle failure prediction model based on the target vehicle data, the method further includes: based on the failure prediction probability, when the failure prediction probability exceeds a first prediction probability, issuing a fault prompt message; or when the failure prediction probability exceeds a second prediction probability, restricting the vehicle startup behavior. According to another aspect of an embodiment of the present invention, a vehicle failure monitoring device is also provided, including: an acquisition module for acquiring target vehicle data of new energy vehicles in a target time period; a failure prediction module for obtaining the failure prediction probability of the new energy vehicle after the target time period by using a vehicle failure prediction model based on the target vehicle data, wherein the vehicle failure prediction model is based on a vehicle data set and obtained by machine learning, the vehicle data set including multiple groups of vehicle data collected before the failure time period from multiple faulty vehicles, and the failure prediction probabilities corresponding to the multiple groups of vehicle data, respectively, and the failure time period is the time period when the multiple faulty vehicles fail.

[0018] Through the above methods, early warning and active control can effectively improve the safety of the vehicle, thereby ensuring the life safety and property safety of the people in the vehicle.

[0019] According to another aspect of an embodiment of the present invention, a non-volatile storage medium is provided, wherein the non-volatile storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executing any one of the vehicle fault monitoring methods.

[0020] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps of any one of the vehicle fault monitoring methods are implemented.

[0021] In an embodiment of the present invention, target vehicle data of new energy vehicles within a target time period is obtained; based on the target vehicle data, a vehicle fault prediction model is used to obtain a fault prediction probability of the new energy vehicle after the target time period, wherein the vehicle fault prediction model is based on a vehicle data set and obtained through machine learning, wherein the vehicle data set includes multiple groups of vehicle data collected from multiple faulty vehicles before the fault time period, and the fault prediction probabilities corresponding to the multiple groups of vehicle data, respectively, and the fault time period is a time period when the multiple faulty vehicles fail, thereby achieving the purpose of obtaining the fault prediction probability based on the vehicle fault prediction model, thereby achieving an early warning based on the fault prediction probability, enabling the personnel in the vehicle to perform adaptive operations before the fault occurs, and effectively ensuring the technical effect of the safety of the personnel in the vehicle, thereby solving the technical problem that the safety of the personnel in the vehicle cannot be effectively guaranteed due to the real-time detection of whether the vehicle has an abnormality in the related technology and the warning of the abnormal result. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0023] Figure 1 is a flow chart of a vehicle fault monitoring method according to an embodiment of the present invention;

[0024] Figure 2 is a flow chart of an optional vehicle fault monitoring method according to an embodiment of the present invention;

[0025] Figure 3 is a schematic diagram of a vehicle fault monitoring device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0028] First, to facilitate understanding of the embodiments of the present invention, some terms or nouns involved in the present invention are explained below:

[0029] Vehicle Identification Number (VIN) is a unique 17-digit number and letter combination used to identify and distinguish a car. VIN is usually located on the hood, door, engine cover and other parts of the car, and can be used to query the car's manufacturer, model, production year, engine type and other information.

[0030] Single cell voltage refers to the voltage of a single cell in a battery or battery pack. It refers to the voltage of each individual battery cell in a battery pack, usually in volts (V). When using a battery pack or battery, the stability of the single cell voltage can be monitored and maintained to protect the normal operation and safety of the system.

[0031] Machine learning refers to a branch of artificial intelligence that uses data and statistical techniques to allow computer systems to automatically learn and improve their performance to achieve specific task goals. Machine learning is a way for computer systems to continuously optimize algorithms and models through large amounts of data input and pattern recognition, so that they can automatically identify, classify and predict data.

[0032] According to an embodiment of the present invention, a method embodiment of vehicle fault monitoring is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0033] Figure 1 is a flow chart of a vehicle fault monitoring method according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0034] Step S102, obtaining target vehicle data of new energy vehicles within a target period;

[0035] Optionally, the target vehicle data of the new energy vehicle within the target period is obtained to perform real-time detection of the new energy vehicle, determine the predicted probability of the fault, and effectively remind the personnel in the vehicle to perform adaptive operations in advance. Among them, the target period can be a time period with a time interval of one month, a time period with a time interval of one day, or a time period with a time interval of several minutes. The target vehicle data is the vehicle data of the new energy vehicle collected within the target period, and the specific duration setting of the target period can be set according to the time fault detection needs (such as according to the fault type). For example, if the real-time requirement is high, such as a fault alarm of the thermal runaway type, the target period can be set to a period of 1-5 minutes forward from the current moment, and the statistics of 1-5 minutes forward from the current moment, or nearly 10-20 frames of data can be selected as the target vehicle data. If the real-time requirement is not high for the fault alarm, the corresponding target period can be a day or a month before the current moment, etc., and the corresponding target vehicle data can be taken from one day or one month of data for offline prediction. This type of fault alarm is based on the overall change trend over a period of time for fault prediction, and this trend change is not visible in a short time, so statistical information of data with a longer span time is required. Specifically, by using the target vehicle data collected over a longer period of time as the detection data, more comprehensive data can be obtained than the vehicle data corresponding to a time period with intervals of seconds. More vehicle status characteristics can be determined from the above-mentioned longer time period to achieve the effect of predicting the vehicle status that may occur in the future.

[0036] Step S104, based on the target vehicle data, a vehicle fault prediction model is used to obtain a fault prediction probability of the new energy vehicle after the target time period, wherein the vehicle fault prediction model is based on a vehicle data set and is obtained through machine learning, the vehicle data set includes multiple groups of vehicle data collected from multiple faulty vehicles before the fault time period, and the fault prediction probabilities corresponding to the multiple groups of vehicle data, respectively, and the fault time period is a time period when multiple faulty vehicles fail.

[0037] Optionally, based on the target vehicle data during the target time period, the probability of a failure that may occur after the target time period is determined, which can effectively achieve the prediction effect of vehicle failure. Based on the probability of a failure that may occur after the target time period, early warning can be given to occupants of the vehicle, and effective adaptive operations can be performed in the event of a warning to ensure the safety of occupants of the vehicle.

[0038] Optionally, multiple groups of vehicle data corresponding to historical time periods before the vehicle failure period are used as input samples of a vehicle failure prediction model. The vehicle failure period indicates that a failure has occurred at this time, and the failure prediction probability is 100%. When the failure period is divided into time intervals of one month, the historical time period before the failure period is divided into multiple first time periods. The time intervals corresponding to the multiple first time periods are the same as the time interval of the failure period. The multiple first time periods correspond to multiple different failure prediction probabilities, respectively. Vehicle state features in the vehicle data corresponding to the multiple first time periods are extracted, and the failure prediction probabilities corresponding to the multiple vehicle state features are determined. The vehicle state features of the multiple first time periods and the corresponding failure prediction probabilities are used as input samples to obtain a vehicle failure prediction model through machine learning.

[0039] In an optional embodiment, based on the target vehicle data, a vehicle fault prediction model is adopted to obtain the fault prediction probability of the new energy vehicle after the target time period. The method also includes: obtaining a predetermined number of groups of vehicle data among multiple groups of vehicle data in the following manner: dividing the historical vehicle data of any faulty vehicle among multiple faulty vehicles into a predetermined number of parts according to the data generation period to obtain a predetermined number of vehicle data; obtaining multiple groups of vehicle data by adopting the method of obtaining the predetermined number of groups of vehicle data, wherein the number of the multiple groups of vehicle data is less than the predetermined number of groups.

[0040] Optionally, by dividing the historical data of the faulty vehicle into multiple segments according to time, and selecting a predetermined number of groups of data as training samples, the diversity and timeliness of the model training data can be ensured. In this way, the model can learn the data characteristics of the vehicle at different time points before the fault, thereby improving the accuracy of fault prediction. In addition, since the number of the selected multiple groups of vehicle data is less than the predetermined number of groups, this is actually a data dimension reduction, which helps to reduce the computational burden of model training, and also ensures the amount of data required for model training, so that the model can effectively learn and predict faults under limited computing resources. Optionally, according to the data generation period (i.e., the first period), the historical vehicle data of any faulty vehicle is divided into a predetermined number of copies, wherein the number of the first period corresponds to the number of the predetermined number of copies one by one, and the predetermined number of vehicle data is obtained at this time. Based on each vehicle data, at least one of the following is included: vehicle identification number (VIN), single cell voltage array, temperature array, current, mileage, charge and discharge status, remaining power state (System Development Life Cycle, SOC), time, factory number, engineering code. A primary statistical table is formed based on the above predetermined number of vehicle data.

[0041] Specifically, the vehicle code refers to the identification code included on the vehicle battery, which is used to identify and track the manufacturing date, model, production batch and other information of the battery. It can be composed of a string of numbers and letters, which can be marked on the label on the battery. Each vehicle corresponds to a vehicle code, and the vehicle codes on multiple vehicles are different. New energy vehicles have multiple battery packs to form a battery pack. Each battery pack can output a single cell voltage. The single cell voltage array is a data set of multiple single cell voltages in the first time period. The temperature array refers to a data set of temperature values ​​near the battery pack. The charge and discharge status is used to indicate the charge state, discharge state and static state that exist in the first time period. Among them, the static state is a state of not charging and not discharging.

[0042] Optionally, based on the vehicle data obtained in the above-mentioned multiple first time periods, the vehicle data in any first time period is divided as follows: the vehicle data is sliced ​​and divided according to the charging state, discharging state, and static state, that is, the above-mentioned vehicle data is classified according to the charging state, discharging state, and static state, so that the detailed data can be divided into multiple states. When the first time period is one month, each state includes multiple vehicle data in time order. Starting from the starting record, the charging time of the first time period is counted to determine whether the vehicle is shallowly charged, shallowly discharged, overcharged or over-discharged. The primary statistical table is divided and sorted based on the above-mentioned charging state, discharging state and static state to form a first secondary statistical table.

[0043] Optionally, based on the primary statistical table, vehicle data is not only divided according to charging and discharging status, but also sliced ​​according to preset mileage thresholds according to mileage. Vehicle data exceeding the preset mileage threshold in the first time period and vehicle data not exceeding the preset mileage threshold are classified to form a second secondary statistical table, and the above-mentioned first secondary statistical table and the second secondary statistical table are merged to form a secondary statistical table.

[0044] Optionally, based on the secondary statistical table, filter the vehicle data that is overcharged, over-discharged, and exceeds the preset mileage threshold, and form the above filtered vehicle data into an intermediate statistical table. Based on the vehicle code included in the corresponding vehicle data in the intermediate statistical table, focus on the vehicle data after the target period corresponding to the vehicle code. It should be noted that the probability of failure of vehicles that are overcharged, over-discharged, and exceed the preset mileage threshold is much greater than that of vehicles that are not overcharged, over-discharged, and exceed the preset mileage threshold. It is necessary to focus on the above vehicles based on the intermediate statistical table.

[0045] Optionally, the preset mileage threshold may be that the mileage in the first time period exceeds 3,000 km, or the total mileage exceeds 100,000 kilometers. Overcharging and over-discharging are determined based on the remaining power value and the number of charging times during the charging or discharging process. It is also possible to count the number of fast charges based on the ratio between SOC and charging time to obtain the fast charging ratio. Fast charging will reduce battery performance and usage time. Based on mileage, overcharging, over-discharging and fast charging ratio, the fault prediction probability of the vehicle is comprehensively considered, and the accuracy of the fault prediction probability judgment is effectively achieved.

[0046] In an optional embodiment, based on the target vehicle data, a vehicle fault prediction model is used to obtain the fault prediction probability of the new energy vehicle after the target time period. The method also includes: obtaining the data generation time periods corresponding to a predetermined number of groups of vehicle data; based on the data generation time periods corresponding to the predetermined number of groups of vehicle data, determining the fault prediction probabilities corresponding to the predetermined number of groups of vehicle data; and obtaining the fault prediction probabilities corresponding to multiple groups of vehicle data by obtaining the fault prediction probabilities corresponding to the predetermined number of groups of vehicle data.

[0047] Optionally, each set of data in the vehicle data set corresponds to a fault prediction probability, which is determined by the proximity between the data generation period and the fault period. By obtaining the time label of each set of data, the predicted probability of each set of data before the fault occurs can be calculated based on the historical fault data and time information. The above method provides the subsequent model training process with information about the correlation between data and the probability of fault occurrence, so that the model can learn to predict the trend from normal to fault. It enables the model to more accurately identify the risk level of vehicle failures, thereby providing more specific and timely information during fault warnings, which helps to improve the accuracy and lead time of warnings and ensure vehicle safety.

[0048] Optionally, slice division is performed based on the vehicle data collected in multiple first time periods, focusing on vehicles that are overcharged, over-discharged, and exceed a preset mileage threshold, extracting characteristic values ​​from the vehicle data corresponding to the vehicles that are overcharged, over-discharged, and exceed a preset mileage threshold in the above-mentioned multiple first time periods, and determining the fault prediction probabilities corresponding to the multiple first time periods respectively based on the above-mentioned characteristic values.

[0049] Specifically, taking the first period as one month as an example, the data of one vehicle shows a fault during the fault period, and the fault prediction probability is 100%. Based on the vehicle code corresponding to the faulty vehicle and the intermediate statistical table, the vehicle data of the faulty vehicle in multiple first periods are determined. Based on the vehicle data analysis, it is determined that the fault prediction probability corresponding to the vehicle data in the first period closest to the fault period is 90%. Starting from the first period closest to the fault period, and moving forward in chronological order, the fault prediction probability changes to 80%, 70%, and 60% step by step. Based on the characteristic value of the vehicle data in the first period corresponding to the fault prediction probability, it is determined that when the target vehicle data input in the target period meets the above characteristic value, the probability of a fault at this time is the corresponding fault prediction probability of the characteristic value.

[0050] In an optional embodiment, when the vehicle data set also includes fault types corresponding to multiple groups of vehicle data, the method also includes: based on the target vehicle data, using a vehicle fault prediction model to obtain the fault prediction probability and fault type of the new energy vehicle after the target time period.

[0051] Optionally, the fault prediction not only includes the fault prediction probability of possible faults, but also the possible fault types of the corresponding faults. For example, undervoltage, high temperature, impedance and other fault types. The target vehicle data is used as the input of the vehicle fault prediction model, and the possible fault types of the vehicle and the corresponding fault prediction probabilities of the fault types are output. In this way, real-time monitoring of the vehicle can be achieved, thereby reminding the vehicle driver to respond to possible problems of the vehicle in a timely manner and effectively ensuring the safety of the people in the vehicle.

[0052] In an optional embodiment, before obtaining the fault prediction probability and fault type of the new energy vehicle after the target time period by using a vehicle fault prediction model based on the target vehicle data, the method also includes: obtaining fault vehicle data corresponding to multiple faulty vehicles within the fault time period, wherein the fault time period is a time period in which multiple faulty vehicles fail; determining the fault types and vehicle codes corresponding to the multiple faulty vehicles; determining the fault types corresponding to the multiple faulty vehicles within the historical time period based on the fault types and vehicle codes corresponding to the multiple faulty vehicles; and determining the fault types corresponding to multiple groups of vehicle data based on the fault types corresponding to the multiple faulty vehicles.

[0053] Optionally, faulty vehicle data corresponding to multiple faulty vehicles within the fault period are obtained, and the fault data include the fault type, that is, the reasons for the faults of the multiple faulty vehicles. The fault may be caused by the battery, or the engine may fail due to excessive mileage, or the battery may have reduced working efficiency due to excessive temperature.

[0054] Optionally, based on the above-mentioned fault type, the vehicle code included in the fault vehicle data is labeled, that is, a fault type label is set for the vehicle code of the fault vehicle. Based on the above-mentioned vehicle code carrying the fault type label, the vehicle data corresponding to the vehicle code is searched in the intermediate statistical table, and the vehicle data within multiple first time periods are determined. In this way, the vehicle will have a corresponding vehicle fault type. The characteristic values ​​of the vehicle data within multiple first time periods are effectively extracted, and the above-mentioned characteristic values ​​are matched with the vehicle fault types, and the fault prediction probabilities are matched accordingly, so as to effectively predict the target vehicle data within the target time period, obtain the fault prediction probability and fault type that may occur after the target time period, and improve the accuracy of the prediction.

[0055] In an optional embodiment, based on the fault types corresponding to multiple faulty vehicles, the fault types corresponding to multiple groups of vehicle data are determined, including: obtaining the fault types corresponding to a predetermined number of groups of vehicle data among the multiple groups of vehicle data in the following manner: dividing the historical vehicle data of any faulty vehicle among the multiple faulty vehicles into a predetermined number of portions according to the data generation period to obtain a predetermined number of vehicle data; determining the fault type corresponding to any faulty vehicle as the fault type corresponding to the predetermined number of vehicle data; obtaining the fault types corresponding to the multiple groups of vehicle data by adopting the method of obtaining the fault types corresponding to the predetermined number of groups of vehicle data.

[0056] Optionally, by dividing the historical vehicle data by time and assigning fault type labels to the data in the corresponding time period, a dataset labeled with fault types can be created for training the vehicle fault prediction model. The above method enables the model to not only predict the probability of a fault, but also identify the type of fault, thereby providing more specific and useful warning information. For example, the model can predict specific types such as battery undervoltage, battery overheating, or brake system failure, which will greatly help drivers and maintenance personnel to quickly locate the problem and take measures. Through time series analysis and fault type labeling, the learning ability and prediction accuracy of the model can be enhanced, and the practical value of fault warning can be improved.

[0057] Optionally, the historical vehicle data of any faulty vehicle among the multiple faulty vehicles is divided into a predetermined number of copies according to the data generation period, that is, the historical vehicle data of any faulty vehicle is divided into multiple vehicle data corresponding to the first time period according to the time interval of the first time period, wherein the predetermined number of copies corresponds to the first time period one by one. The fault type corresponding to the vehicle having a fault in the fault time period is marked on the vehicle data corresponding to the multiple first time periods at the same time, that is, the historical vehicle data of the faulty vehicle are all determined to be a fault condition that may cause the corresponding fault type.

[0058] Optionally, based on the fault types corresponding to the above-mentioned multiple first time periods, the possible fault prediction probability of the fault type is determined, and the method of determining the fault prediction probability is the same as the above-mentioned method, starting from the first time period closest to the fault time period, and moving forward in chronological order, the fault prediction probability is gradually changed to 80%, 70%, and 60%. The fault prediction probability corresponds to the characteristic value of the vehicle data in the first time period, and it is determined that when the target vehicle data input in the target time period meets the above-mentioned characteristic value, the probability of a fault occurring at this time is the corresponding fault prediction probability of the characteristic value, and the type of the fault occurring at this time is the fault type corresponding to the characteristic value.

[0059] In an optional embodiment, based on the target vehicle data, a vehicle fault prediction model is used to obtain the fault prediction probability of the new energy vehicle in the target time period. The method also includes: based on the fault prediction probability, when the fault prediction probability exceeds the first prediction probability, issuing a fault prompt message; or when the fault prediction probability exceeds the second prediction probability, restricting the vehicle startup behavior.

[0060] Optionally, based on the above-mentioned fault prediction probability and fault type, different response measures are made for a variety of different fault prediction profiles. When the fault prediction probability does not exceed the first prediction probability (40%), the possibility of a fault is low at this time. It may be due to the deviation of the fault prediction caused by different driving habits. At this time, the vehicle will not be warned. When the fault prediction probability exceeds the first prediction probability but is lower than the second prediction probability (75%), there is a trend of a fault at this time. Based on the above-mentioned prediction probability, a fault prompt information is sent to the vehicle, and the fault prompt information includes at least the fault type and the fault prediction probability. When the fault prediction probability exceeds the second prediction probability, the vehicle is about to fail. While sending a fault prompt information to the vehicle, the vehicle driver is reminded to repair the vehicle. If the fault prediction probability exceeds the second prediction probability for a long time, the vehicle may be a hazard that affects the safety of the people in the vehicle. In order to protect the safety of the people in the vehicle, after the vehicle stops and turns off, the vehicle is restricted from restarting again, and the driver is not allowed to drive the vehicle on the road again, which effectively ensures the safety of the people in the vehicle. In the above manner, through early warning and active control, the safety of the vehicle can be effectively improved, thereby ensuring the life safety and property safety of the people in the vehicle.

[0061] Through the above steps S102 to S104, the purpose of obtaining the fault prediction probability based on the vehicle fault prediction model can be achieved, thereby realizing early warning based on the fault prediction probability, enabling the personnel in the vehicle to make adaptive operations before the fault occurs, and effectively ensuring the technical effect of the safety of the personnel in the vehicle, thereby solving the technical problem that the safety of the personnel in the vehicle cannot be effectively guaranteed due to the real-time detection of whether the vehicle has an abnormality in the related technology and the warning based on the existence of the abnormality.

[0062] Based on the above embodiments and optional embodiments, the present invention proposes an optional implementation mode: Figure 2 is a flow chart of an optional vehicle fault monitoring method according to an embodiment of the present invention, such as Figure 2 As shown, the method includes:

[0063] Step S1, cleaning the battery detailed data of the vehicle on the big data monitoring platform;

[0064] Specifically, the big data cluster computing system (spark) is used to read the cached data in the distributed stream processing platform (kafka), clean the abnormal data, and store the statistics in the data warehouse (hive data warehouse). Then the cluster computing system computing engine is used to read the vehicle data of the past year, perform dimensional index calculations (i.e., slice and divide the vehicle data), generate intermediate statistical tables (primary statistical tables and secondary statistical tables), and then perform secondary processing to generate a vehicle information table of focus (intermediate statistical table).

[0065] Step S2, slicing the vehicle data according to the charging state, discharging state and static state;

[0066] Specifically, vehicles with long total mileage (total mileage exceeding 100,000 kilometers), frequently used vehicles (vehicles with charging or discharging times exceeding a threshold), and vehicles with battery performance degradation are screened, and the monthly mileage, charge amount, remaining power (SOC), fast charging ratio, slow charging ratio, voltage, pressure difference, temperature, temperature difference and other information corresponding to the above vehicles are counted.

[0067] Step S3, statistical dimension indicators (i.e., vehicle data after slicing and dividing the vehicle data) are respectively generated into a primary statistical table, a secondary statistical table, and an intermediate statistical table;

[0068] Step S4, extracting characteristic values ​​of the above-mentioned dimension indicators;

[0069] Step S5, obtaining the vehicle information of focus in the alarm information table (intermediate statistical table);

[0070] Step S6, performing sample grouping and model training on the intermediate statistical table;

[0071] The vehicle data in the intermediate statistical table is characterized to obtain characteristic values, and the machine learning model is trained to obtain a vehicle fault prediction model. That is, vehicles that are predicted to have abnormal alarm risks are monitored in a focused manner and fed back to the maintenance department in a timely manner for follow-up processing to eliminate potential safety hazards.

[0072] Step S7: model alarm prediction and verification.

[0073] Through the above steps S1 to S7, this embodiment focuses on vehicles with a large total mileage (total mileage exceeds 100,000 kilometers), frequently used vehicles (vehicles with charging or discharging times exceeding a threshold), and vehicles with battery performance degradation. Since new energy vehicles are more flammable than fuel vehicles, drivers are more concerned about safety, and the risk issues in daily use are also higher than other similar vehicles. Focus on the battery information of such vehicles and analyze the relationship between voltage, temperature, mileage, and SOC under different states. After a period of data analysis, the health and abnormal information of the battery can be predicted. Early detection and early prevention can be achieved, and unsafe hidden dangers can be predicted in advance.

[0074] In this embodiment, a vehicle fault monitoring device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the terms "module" and "device" can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0075] According to an embodiment of the present invention, there is also provided an embodiment of a device for implementing the above vehicle fault monitoring method. Figure 3 is a schematic diagram of the structure of a vehicle fault monitoring device according to an embodiment of the present invention. Figure 3 As shown, the above vehicle fault monitoring device includes: an acquisition module 301 and a fault prediction module 302, wherein:

[0076] An acquisition module 301 is used to acquire target vehicle data of new energy vehicles within a target period;

[0077] The fault prediction module 302 is connected to the acquisition module 301, and is used to obtain the fault prediction probability of the new energy vehicle after the target time period based on the target vehicle data and using a vehicle fault prediction model, wherein the vehicle fault prediction model is based on a vehicle data set and obtained through machine learning, and the vehicle data set includes multiple groups of vehicle data collected from multiple faulty vehicles before the fault time period, and the fault prediction probabilities corresponding to the multiple groups of vehicle data, respectively, and the fault time period is the time period when the multiple faulty vehicles fail.

[0078] It should be noted that the above modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following ways: the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0079] It should be noted that the acquisition module 301 and the fault prediction module 302 correspond to steps S102 to S104 in the embodiment, and the examples and application scenarios implemented by the modules and the corresponding steps are the same, but are not limited to the contents disclosed in the embodiment. It should be noted that the modules as part of the device can be run in a computer terminal.

[0080] It should be noted that the optional or preferred implementation of this embodiment can refer to the relevant description in the embodiment, which will not be repeated here.

[0081] The above-mentioned vehicle fault monitoring device may also include a processor and a memory. The above-mentioned acquisition module 301 and the fault prediction module 302 are stored in the memory as program modules, and the processor executes the above-mentioned program modules stored in the memory to realize corresponding functions.

[0082] The processor includes a kernel, which retrieves the corresponding program module from the memory. The kernel may be one or more. The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. The memory includes at least one memory chip.

[0083] According to an embodiment of the present application, an embodiment of a non-volatile storage medium is also provided. Optionally, in this embodiment, the non-volatile storage medium includes a stored program, wherein when the program is executed, the device where the non-volatile storage medium is located is controlled to execute any of the vehicle fault monitoring methods.

[0084] Optionally, in this embodiment, the non-volatile storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group, and the non-volatile storage medium includes a stored program.

[0085] Optionally, when the program is running, the device where the non-volatile storage medium is located is controlled to perform the following functions: obtaining target vehicle data of new energy vehicles within a target time period; based on the target vehicle data, using a vehicle fault prediction model to obtain a fault prediction probability of the new energy vehicle after the target time period, wherein the vehicle fault prediction model is based on a vehicle data set and is obtained through machine learning, the vehicle data set includes multiple groups of vehicle data collected from multiple faulty vehicles before the fault time period, and the fault prediction probabilities corresponding to the multiple groups of vehicle data, and the fault time period is a time period when multiple faulty vehicles fail.

[0086] According to an embodiment of the present application, an embodiment of a processor is also provided. Optionally, in this embodiment, the processor is used to run a program, wherein when the program is run, any one of the vehicle fault monitoring methods is executed.

[0087] According to an embodiment of the present application, an embodiment of a computer program product is also provided. Optionally, in this embodiment, the computer program product includes a computer program, and when the computer program is executed by a processor, the computer program implements any one of the steps of the vehicle fault monitoring method described above.

[0088] Optionally, the above-mentioned computer program product, when executed on a data processing device, is suitable for executing a program that initializes the following method steps: obtaining target vehicle data of new energy vehicles within a target time period; based on the target vehicle data, using a vehicle fault prediction model to obtain a fault prediction probability of the new energy vehicle after the target time period, wherein the vehicle fault prediction model is based on a vehicle data set and is obtained through machine learning, and the vehicle data set includes multiple groups of vehicle data collected from multiple faulty vehicles before the fault time period, and the fault prediction probabilities corresponding to the multiple groups of vehicle data, and the fault time period is a time period when multiple faulty vehicles fail.

[0089] An embodiment of the present invention provides an electronic device, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, the following steps are implemented: obtaining target vehicle data of new energy vehicles within a target time period; based on the target vehicle data, using a vehicle fault prediction model to obtain a fault prediction probability of the new energy vehicle after the target time period, wherein the vehicle fault prediction model is based on a vehicle data set and is obtained through machine learning. The vehicle data set includes multiple groups of vehicle data collected from multiple faulty vehicles before the fault time period, and the fault prediction probabilities corresponding to the multiple groups of vehicle data, respectively. The fault time period is a time period when multiple faulty vehicles fail.

[0090] The above sequence of the embodiments of the present invention is for description only and does not represent the superiority or inferiority of the embodiments.

[0091] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0092] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the above modules can be a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, modules or indirect coupling or communication connection of modules, which can be electrical or other forms.

[0093] The modules described above as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment.

[0094] In addition, each functional module in each embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of software functional modules.

[0095] If the above-mentioned integrated module is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable non-volatile storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a non-volatile storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the various embodiments of the present invention. The aforementioned non-volatile storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0096] The above are only preferred embodiments of the present invention. It should be pointed out that, for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A vehicle fault monitoring method, characterized in that: include: Obtain target vehicle data of new energy vehicles within a target period; Based on the target vehicle data, a vehicle fault prediction model is used to obtain the fault prediction probability of the new energy vehicle after the target time period, wherein the vehicle fault prediction model is based on a vehicle data set and is obtained through machine learning, the vehicle data set includes multiple groups of vehicle data collected from multiple faulty vehicles before the fault time period, and the fault prediction probabilities corresponding to the multiple groups of vehicle data, respectively, and the fault time period is the time period when the multiple faulty vehicles fail.

2. The method according to claim 1, characterized in that The method further comprises: based on the target vehicle data, using a vehicle fault prediction model to obtain a fault prediction probability of the new energy vehicle after the target period. The predetermined number of vehicle data sets among the plurality of vehicle data sets are obtained by: Dividing the historical vehicle data of any faulty vehicle among the multiple faulty vehicles into a predetermined number of copies according to a data generation period, and obtaining the predetermined number of vehicle data; The plurality of groups of vehicle data are obtained in a manner of obtaining the predetermined number of groups of vehicle data, wherein the number of the plurality of groups of vehicle data is less than the predetermined number of groups.

3. The method according to claim 1, characterized in that The method further comprises: based on the target vehicle data, using a vehicle fault prediction model to obtain a fault prediction probability of the new energy vehicle after the target period. The fault prediction probabilities corresponding to the predetermined number of vehicle data in the plurality of vehicle data are obtained by: Obtaining data generation time periods corresponding to the predetermined number of groups of vehicle data respectively; Determining the fault prediction probabilities corresponding to the predetermined number of vehicle data respectively based on the data generation periods corresponding to the predetermined number of vehicle data respectively; The fault prediction probabilities corresponding to the plurality of groups of vehicle data are obtained by obtaining the fault prediction probabilities corresponding to the predetermined number of groups of vehicle data.

4. The method according to claim 1, characterized in that: In the case where the vehicle data set further includes the fault types respectively corresponding to the multiple sets of vehicle data, the method further includes: Based on the target vehicle data, the vehicle fault prediction model is used to obtain the fault prediction probability and fault type of the new energy vehicle after the target time period.

5. The method according to claim 4, characterized in that Before obtaining the predicted fault probability and fault type of the new energy vehicle after the target period based on the target vehicle data and using the vehicle fault prediction model, the method further includes: Acquire faulty vehicle data respectively corresponding to the plurality of faulty vehicles collected during the faulty period; Determining the fault types and vehicle codes respectively corresponding to the multiple faulty vehicles based on the faulty vehicle data respectively corresponding to the multiple faulty vehicles; Determine the fault types corresponding to the multiple faulty vehicles in the historical period based on the fault types and vehicle codes corresponding to the multiple faulty vehicles respectively; Based on the fault types respectively corresponding to the multiple faulty vehicles, the fault types respectively corresponding to the multiple groups of vehicle data are determined.

6. The method according to claim 5, characterized in that The determining, based on the fault types respectively corresponding to the multiple faulty vehicles, the fault types respectively corresponding to the multiple groups of vehicle data comprises: The fault types corresponding to the predetermined number of vehicle data in the plurality of vehicle data are obtained in the following manner: Dividing the historical vehicle data of any faulty vehicle among the multiple faulty vehicles into a predetermined number of copies according to a data generation period, and obtaining the predetermined number of vehicle data; Determine the fault type corresponding to any one of the faulty vehicles as the fault types corresponding to the predetermined number of vehicle data respectively; The fault types corresponding to the plurality of groups of vehicle data are obtained by obtaining the fault types corresponding to the predetermined number of groups of vehicle data.

7. The method according to claim 1, characterized in that After obtaining the failure prediction probability of the new energy vehicle in the target period by using the vehicle failure prediction model based on the target vehicle data, the method further includes: Based on the fault prediction probability, when the fault prediction probability exceeds a first prediction probability, issuing a fault prompt message; When the fault prediction probability exceeds the second prediction probability, the vehicle startup behavior is restricted.

8. A vehicle fault monitoring device, characterized in that: include: An acquisition module, used to acquire target vehicle data of new energy vehicles within a target period; A fault prediction module is used to obtain the fault prediction probability of the new energy vehicle after the target time period based on the target vehicle data and using a vehicle fault prediction model, wherein the vehicle fault prediction model is based on a vehicle data set and obtained through machine learning, the vehicle data set includes multiple groups of vehicle data collected from multiple faulty vehicles before the fault time period, and the fault prediction probabilities corresponding to the multiple groups of vehicle data, respectively, and the fault time period is the time period when the multiple faulty vehicles fail.

9. A non-volatile storage medium, characterized in that: The non-volatile storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the vehicle fault monitoring method described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the vehicle fault monitoring method according to any one of claims 1 to 7 are implemented.

Images