CN116653817A

CN116653817A - Fault early warning method, model training method, device, vehicle and storage medium

Info

Publication number: CN116653817A
Application number: CN202310597004.1A
Authority: CN
Inventors: 江婉榕; 李昌; 范有才; 施暄宣; 罗浩田
Original assignee: Chongqing Changan Automobile Co Ltd
Current assignee: Chongqing Changan Automobile Co Ltd
Priority date: 2023-05-24
Filing date: 2023-05-24
Publication date: 2023-08-29

Abstract

The application relates to the field of intelligent automobiles, in particular to a fault early warning method, a model training method, a device, a vehicle and a storage medium. The method is applied to a vehicle terminal; the vehicle terminal comprises a fault early warning model; the method comprises the following steps: acquiring state data of a vehicle in a first preset time period; according to the change trend of the current of the storage battery, M starting and stopping time intervals of the vehicle in a first preset time period are determined; dividing state data in a first preset time period according to M starting and stopping time intervals to obtain M state data sets; aggregating the state data in the M state data sets to obtain a target state data set; and determining a fault early-warning result according to the target state data set and the fault early-warning model. Therefore, the problem that the accuracy of fault early warning is low if the representativeness of the division of the state data of the vehicle is poor when the fault early warning is carried out on the parts of the automobile can be solved.

Description

Fault early warning method, model training method, device, vehicle and storage medium

Technical Field

The application relates to the technical field of data processing, in particular to a fault early warning method, a model training device, a vehicle and a storage medium.

Background

In recent years, with the rapid development of the automobile industry, the integration level and the automation degree of automobile parts are higher and higher, and the structural complexity of the automobile parts is also increasing, and under such circumstances, the reliability, the safety and the maintainability of the automobile parts are also more and more important, so that the fault early warning method based on the automobile parts becomes a big research hot spot in the automobile industry. In the related art, the fault early warning method of the automobile parts mainly depends on whether the signal value acquired by the vehicle-mounted terminal reaches a set early warning threshold value or not. However, if the vehicles judge that the signal value collected by the vehicle-mounted terminal reaches the set early warning threshold value and then send out the warning signal, the problem that the response time of the driver is insufficient or the vehicle has faults exists, and poor use experience is brought to users.

Aiming at the problems, the related technology provides an intelligent part fault early warning algorithm based on cloud historical data and real-time data of a vehicle, but the algorithm uses a fixed time threshold to determine a starting and stopping time interval of the vehicle, so that the state data of the vehicle is divided according to the starting and stopping time interval, the problem that the fixed time threshold is difficult to determine or the state data of the vehicle in the starting and stopping time interval is incomplete exists, and the problem that the representativeness of the state data is poor when the state data of the vehicle is divided, and the accuracy of fault early warning is low is caused.

Disclosure of Invention

The application provides a fault early warning method, a model training device, a vehicle and a storage medium, which at least solve the problem that the accuracy of fault early warning is low if the representativeness of the division of state data of the vehicle is poor when the fault early warning is carried out on parts of the vehicle. The technical scheme of the application is as follows:

according to a first aspect of the present application, a fault early warning method is provided, which is applied to a vehicle terminal; the vehicle terminal comprises a fault early warning model; the method comprises the following steps: acquiring state data of a vehicle in a first preset time period; the state data includes a trend of change in battery current of the vehicle; according to the change trend of the current of the storage battery, M starting and stopping time intervals of the vehicle in a first preset time period are determined; m is an integer greater than or equal to 1; dividing state data in a first preset time period according to M starting and stopping time intervals to obtain M state data sets; wherein one state data set is the state data of the vehicle in one starting and stopping time interval; aggregating the state data in the M state data sets to obtain a target state data set; and determining a fault early-warning result according to the target state data set and the fault early-warning model.

According to the technical means, compared with the method for determining the starting and stopping time interval of the vehicle by using the fixed time threshold in the related art, the method for determining the starting and stopping time interval of the vehicle and dividing the state data of the vehicle according to the starting and stopping time interval has the advantages that the fixed time threshold is difficult to determine or the state data of the vehicle in one starting and stopping time interval is incomplete, the method for determining the starting and stopping time interval of the vehicle and dividing the state data of the vehicle according to the change trend of the current of the storage battery of the vehicle can avoid dividing the state data originally in one starting and stopping time interval into different starting and stopping time intervals, the accuracy of dividing the state data is improved, the accuracy of fault early warning is further improved, and the use experience of a user is improved.

Meanwhile, compared with the prior art that state data in a single starting and stopping time interval are input into a fault early warning model to perform fault early warning on a vehicle, the problem that the state data in the historical starting and stopping time interval of the vehicle affect the fault early warning of the vehicle is not considered, and the method provided by the application obtains the target state data set by aggregating the state data in a plurality of state data sets, and inputs the target state data set into the fault early warning model to obtain a fault early warning result, so that the fault early warning result of the vehicle is more accurate and reliable.

In one possible implementation manner, determining M start-stop time intervals of the vehicle in the first preset time period according to the variation trend of the battery current includes: according to the change trend of the storage battery current, M+1 time nodes with abrupt change of the storage battery current in a first preset time period are determined; and dividing the first preset time period into M starting and stopping time intervals according to M+1 time nodes where the current of the storage battery is suddenly changed.

According to the technical means, compared with the prior art, the method has the advantages that the starting and stopping time intervals of the vehicle are divided according to the starting and stopping time of the vehicle, or the starting and stopping time intervals of the vehicle are divided through the fixed time threshold, so that the problem of inaccurate division of the state data of the vehicle is caused.

In another possible implementation manner, aggregating state data in the M state data sets to obtain a target state data set includes: respectively aggregating each type of state data in the M state data sets to obtain target state data of each type of state data in the M state data sets; and obtaining a target state data set according to the target state data of each type of state data in the M state data sets.

According to the technical means, it can be understood that the fault state of the parts of the vehicle is not only dependent on the state data in the current starting and stopping time interval of the vehicle, but also is related to the state data in the historical starting and stopping time interval of the vehicle, if the single starting and stopping time interval of the state data is input into the fault early warning model to perform fault early warning on the vehicle, the problem that the fault early warning accuracy is low can be caused by ignoring the influence of the state data in the historical starting and stopping time interval on the fault state of the parts of the vehicle is solved, and therefore the method provided by the application considers the influence of the state data in the historical starting and stopping time interval on the fault state of the parts of the vehicle, the fault characteristics of the obtained target state data are more obvious by aggregating the state data in a plurality of state data sets, the current state of the vehicle can be reflected, the input state data can be more representative by taking the target state data as the input of the fault early warning model, and the accuracy of the fault early warning is improved.

According to a second aspect of the present application, there is provided a model training method, the method comprising: acquiring state data of the vehicle in a second preset time period; the second preset time period comprises a period of time before vehicle maintenance and a period of time after vehicle maintenance; the state data includes a trend of change in battery current of the vehicle; according to the change trend of the current of the storage battery, N starting and stopping time intervals of the vehicle in a second preset time period are determined; n is an integer greater than or equal to 1; dividing the state data in a second preset time period according to the N starting and stopping time intervals to obtain N state data sets; wherein one state data set is the state data of the vehicle in one starting and stopping time interval; constructing training samples according to the N state data sets; training a fault early-warning model according to the training sample to obtain a trained fault early-warning model; the fault early warning model is used for sending a fault early warning result to a user.

According to the technical means, unlike the prior art, when the fault early warning model is trained, a fixed time threshold is generally used for determining a historical starting and stopping time interval of the vehicle, and the historical state data of the vehicle is divided according to the historical starting and stopping time interval of the vehicle, so that the problem that the fixed time threshold is difficult to determine or the historical state data of the vehicle in one starting and stopping time interval is incomplete can exist.

In one possible implementation manner, determining N start-stop time intervals of the vehicle in the second preset time period according to the variation trend of the battery current includes: according to the change trend of the storage battery current, determining N+1 time nodes with abrupt change of the storage battery current in a second preset time period; and dividing the second preset time period into N starting and stopping time intervals according to the N+1 time nodes with abrupt change of the current of the storage battery.

According to the technical means, compared with the prior art, the method has the advantages that the historical starting and stopping time intervals of the vehicle are divided according to the starting and stopping time of the vehicle, or the historical starting and stopping time intervals of the vehicle are divided through the fixed time threshold, so that the problem that the historical state data of the vehicle are not accurately divided is solved.

In another possible embodiment, constructing a training sample from the N state data sets includes: dividing N state data sets into a plurality of sample data according to a preset window length by adopting a sliding window algorithm; the state data set in the preset window length is aggregated into one piece of sample data; the length of the preset window is the length of K starting and stopping time intervals; determining a label of each piece of sample data according to the fault condition of the vehicle in each preset window length; the label includes: failure and no failure; and constructing a training sample according to the plurality of pieces of sample data and the label of each piece of sample data.

According to the technical means, compared with the prior art, the method has the advantages that the historical state data and the labels thereof in one historical starting and stopping time interval are used as one sample data, the state data in N state data sets are aggregated each time by adopting a sliding window algorithm, and the aggregated state data and the labels thereof are used as one sample data, so that the representativeness of the sample data can be improved, the sample data can be more embodied in the state of a vehicle running at the moment, and the training sample is more accurate and reliable.

According to a third aspect of the present application, there is provided a malfunction early warning apparatus applied to a vehicle terminal; the vehicle terminal comprises a fault early warning model; the fault early warning device comprises: the acquisition module is used for acquiring state data of the vehicle in a first preset time period; the state data includes a trend of change in battery current of the vehicle; the determining module is used for determining M starting and stopping time intervals of the vehicle in a first preset time period according to the change trend of the storage battery current; m is an integer greater than or equal to 1; the dividing module is used for dividing the state data in the first preset time period according to the M starting and stopping time intervals to obtain M state data sets; wherein one state data set is the state data of the vehicle in one starting and stopping time interval; the aggregation module is used for aggregating the state data in the M state data sets to obtain a target state data set; and the determining module is also used for determining a fault early-warning result according to the target state data set and the fault early-warning model.

In one possible implementation manner, the determining module is specifically configured to determine m+1 time nodes where the battery current is suddenly changed in a first preset time period according to a change trend of the battery current; and dividing the first preset time period into M starting and stopping time intervals according to M+1 time nodes where the current of the storage battery is suddenly changed.

In a possible implementation manner, the aggregation module is specifically configured to aggregate each type of state data in the M state data sets respectively to obtain target state data of each type of state data in the M state data sets; and obtaining a target state data set according to the target state data of each type of state data in the M state data sets.

According to a fourth aspect of the present application, there is provided a model training apparatus comprising: the acquisition module is used for acquiring state data of the vehicle in a second preset time period; the second preset time period comprises a period of time before vehicle maintenance and a period of time after vehicle maintenance; the state data includes a trend of change in battery current of the vehicle; the determining module is used for determining N starting and stopping time intervals of the vehicle in a second preset time period according to the change trend of the storage battery current; n is an integer greater than or equal to 1; the dividing module is used for dividing the state data in the second preset time period according to the N starting and stopping time intervals to obtain N state data sets; wherein one state data set is the state data of the vehicle in one starting and stopping time interval; the construction module is used for constructing training samples according to the N state data sets; the training module is used for training the fault early-warning model according to the training sample to obtain a trained fault early-warning model; the fault early warning model is used for sending a fault early warning result to a user.

In one possible implementation manner, the determining module is specifically configured to determine, according to a trend of change in the battery current, n+1 time nodes where the battery current is suddenly changed in a second preset time period; and dividing the second preset time period into N starting and stopping time intervals according to the N+1 time nodes with abrupt change of the current of the storage battery.

In a possible implementation manner, the construction module is specifically configured to divide the N state data sets into a plurality of sample data according to a preset window length by adopting a sliding window algorithm; the state data set in the preset window length is aggregated into one piece of sample data; the length of the preset window is the length of K starting and stopping time intervals; determining a label of each piece of sample data according to the fault condition of the vehicle in each preset window length; the label includes: failure and no failure; and constructing a training sample according to the plurality of pieces of sample data and the label of each piece of sample data.

According to a fifth aspect of the present application, there is provided a vehicle comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor executing the program to implement the method of any one of the above first or second aspects and any one of its possible implementation manners.

According to a sixth aspect of the present application there is provided a computer readable storage medium, which when executed by a processor of an electronic device, enables the electronic device to perform the method of the first or second aspect and any one of its possible implementation manners.

Therefore, the technical characteristics of the application have the following beneficial effects:

(1) Compared with the prior art that a fixed time threshold is used for determining the starting and stopping time interval of the vehicle, and then the state data of the vehicle is divided according to the starting and stopping time interval, the problem that the fixed time threshold is difficult to determine or the state data of the vehicle in one starting and stopping time interval is incomplete exists. Meanwhile, compared with the prior art that state data in a single starting and stopping time interval are input into a fault early warning model to perform fault early warning on a vehicle, the problem that the state data in the historical starting and stopping time interval of the vehicle affect the fault early warning of the vehicle is not considered, and the method provided by the application obtains the target state data set by aggregating the state data in a plurality of state data sets, and inputs the target state data set into the fault early warning model to obtain a fault early warning result, so that the fault early warning result of the vehicle is more accurate and reliable.

(2) It can be understood that the fault state of the parts of the vehicle not only depends on the state data in the current starting and stopping time interval of the vehicle, but also is related to the state data in the historical starting and stopping time interval of the vehicle, if the state data in the single starting and stopping time interval is input into the fault early warning model to perform fault early warning on the vehicle, the problem that the fault state of the parts of the vehicle is affected by the state data in the historical starting and stopping time interval is possibly caused to be lower in accuracy of the fault early warning is ignored, so that the method provided by the application considers the influence of the state data in the historical starting and stopping time interval on the fault state of the parts of the vehicle, the fault characteristics of the obtained target state data are more obvious by aggregating the state data in a plurality of state data sets, the current state of the vehicle can be reflected, the input state data is more representative by taking the target state data as the input of the fault early warning model, and the accuracy of the fault early warning is improved.

(3) In the method, the historical start-stop section of the vehicle is determined and the historical state data of the vehicle is divided into different start-stop time sections according to the current trend of a storage battery of the vehicle, so that the historical state data originally belonging to the first start-stop time section can be prevented from being divided into different start-stop time sections, the accuracy of the historical state data division is improved, the state data in each training sample is more accurate and representative, and the fault early-warning model obtained by training according to the training sample is more reliable, and the accuracy of a fault early-warning result is improved.

(4) Compared with the prior art, the method has the advantages that the historical starting and stopping time intervals of the vehicle are divided according to the starting and stopping time of the vehicle, or the historical starting and stopping time intervals of the vehicle are divided through the fixed time threshold, so that the problem of inaccurate division of the historical state data of the vehicle is caused.

(5) Compared with the prior art, the method has the advantages that the historical state data and the labels thereof in one historical starting and stopping time interval are used as one sample data, the state data in N state data sets are aggregated each time by adopting a sliding window algorithm, and the aggregated state data and the labels thereof are used as one sample data, so that the representativeness of the sample data can be improved, the sample data can better represent the state of a vehicle running at the moment, and the training sample is more accurate and reliable.

It should be noted that, the technical effects caused by any implementation manner of the third aspect to the sixth aspect may refer to the technical effects caused by the corresponding implementation manner of the first aspect or the second aspect, and are not described herein.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application as claimed.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and together with the description, serve to explain the principles of the application and do not constitute a undue limitation on the application.

FIG. 1 is a block diagram of a fault early warning system, shown in accordance with an exemplary embodiment;

FIG. 2 is a flow chart illustrating a fault pre-warning method according to an exemplary embodiment;

FIG. 3 is a flow chart illustrating another fault pre-warning method according to an exemplary embodiment;

FIG. 4 is a flow chart illustrating another method of fault early warning according to an exemplary embodiment;

FIG. 5 is a flow chart illustrating another fault warning method according to an exemplary embodiment;

FIG. 6 is a flow chart illustrating another fault pre-warning method according to an exemplary embodiment;

FIG. 7 is a flowchart illustrating a model training method according to an exemplary embodiment;

FIG. 8 is a flowchart illustrating another model training method, according to an example embodiment;

FIG. 9 is a flowchart illustrating another model training method, according to an example embodiment;

FIG. 10 is a flowchart illustrating another model training method, according to an example embodiment;

FIG. 11 is a flowchart illustrating a failure training method and a model training method, according to an example embodiment;

FIG. 12 is a block diagram of a fault early warning device according to an exemplary embodiment;

FIG. 13 is a block diagram of a model training apparatus, according to an example embodiment;

fig. 14 is a block diagram of a vehicle according to an exemplary embodiment.

The system comprises a sensor 100, a processor 200, a training device 300, an execution unit 400, a fault early warning device 500, an acquisition module 501, a determination module 502, a division module 503, an aggregation module 504, a model training device 600, an acquisition module 601, a determination module 602, a construction module 603, a training module 604, a vehicle 700, a processor 701 and a memory 702.

Detailed Description

In order to enable a person skilled in the art to better understand the technical solutions of the present application, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

It should be noted that the terms "first," "second," and the like in the description and the claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the application described herein may be implemented in sequences other than those illustrated or otherwise described herein. The implementations described in the following exemplary examples do not represent all implementations consistent with the application. Rather, they are merely examples of apparatus and methods consistent with aspects of the application as detailed in the accompanying claims.

The following describes a fault early warning method, a model training method, a device, a vehicle and a storage medium according to the embodiments of the present application with reference to the accompanying drawings. With the development of artificial intelligence, machine learning and deep learning technologies have made great progress, state data of vehicles are obtained based on cloud, and an intelligent part fault early warning algorithm is constructed according to the state data of the vehicles, so that early warning can be carried out on the vehicles before the vehicles break down, and the precision and speed of fault early warning can be effectively improved. However, when the state data of the vehicle in the cloud is extracted, the state data of the vehicle is generally divided by using a fixed window, but the relevance between the state data of the vehicle and each time the vehicle starts and stops is strong, if the fixed window is used, the data crossing the start and stop interval is split, the size of the fixed window is difficult to determine, and therefore the representativeness of the division of the state data is poor, and the accuracy of fault early warning is affected.

In view of the above problems, a related art proposes a data aggregation method for dividing the status data of each vehicle according to the stopping time threshold of the vehicle, which solves the problem that the data crossing the starting and stopping intervals is split to a certain extent, but after the vehicle is flameout, there may be a phenomenon that the vehicle is not dormant or wakes up abnormally, and after the vehicle is flameout and stopped for a period of time, the status data is still transmitted to the cloud, if the starting and stopping time interval of the vehicle is determined according to the stopping time threshold of the vehicle, and further the status data of the vehicle is divided into the starting and stopping time interval of the next time, the accuracy of dividing the status data is not high, and the accuracy of fault early warning is affected.

Therefore, the application provides a fault early warning method, which determines the starting and stopping time interval of the vehicle and divides the state data of the vehicle according to the change trend of the current of the storage battery of the vehicle, so that the state data originally belonging to the primary starting and stopping time interval can be prevented from being divided into different starting and stopping time intervals, the accuracy of the state data division is improved, the accuracy of fault early warning is further improved, and the use experience of a user is improved.

For ease of understanding, embodiments of the present application are described in detail below with reference to the accompanying drawings.

Fig. 1 is a schematic diagram of a fault early warning system according to an embodiment of the present application, where the fault early warning system includes: sensor 100, processor 200, training device 300, and execution unit 400. Wherein the sensor 100, the processor 200, the execution unit 400 are communicatively connected; the processor 200 is communicatively coupled to the exercise device 300.

In some embodiments, the processor 200 may be integrated into the sensor 100 as a sub-module of the sensor 100; alternatively, the sensor 100 and the processor 200 may be separate two devices.

A sensor 100 for collecting status data of the vehicle.

Wherein the status data of the vehicle includes a trend of change in battery current of the vehicle.

In some embodiments, the categories of status data include: voltage of the storage battery, current of the storage battery, temperature outside the vehicle, temperature inside the vehicle, power state feedback, engine speed, engine temperature and sensor temperature.

Wherein, the battery sensor is used for collecting the state data of the voltage of the battery, the current of the battery and the temperature of the battery of the vehicle; the power supply sensor is used for collecting state data fed back by a power supply of the vehicle; a temperature sensor for collecting state data of an in-vehicle temperature, an out-of-vehicle temperature, an engine temperature, and a sensor temperature of the vehicle; the rotational speed sensor is used for collecting state data of the engine rotational speed of the vehicle.

It will be appreciated that the number of sensors installed in different vehicles may be different, and in the method provided in the embodiment of the present application, one or more sensors 100 may be provided, and those skilled in the art may choose the method according to actual needs, which is not limited in the embodiment of the present application.

The processor 200 is configured to acquire the vehicle status data collected by the sensor 100.

In some embodiments, the processor 200 is further configured to send the vehicle status data collected from the sensor 100 to the training device 300.

Training means 300 for training a fault early warning model; the training apparatus 300 may be a server, for example, a single server, or a server cluster including a plurality of servers. In some implementations, the server cluster may also be a distributed cluster.

Still another example, the exercise device 300 may be a cell phone, tablet, desktop, laptop, handheld computer, notebook, ultra-mobile personal computer (UMPC), netbook, cell phone, personal digital assistant (personal digital assistant, PDA), augmented reality (augmented reality, AR) \virtual reality (VR) device, or the like. The specific form of the training device 300 is not particularly limited in the present application. The training apparatus 300 may interact with a user by one or more of a keyboard, a touch pad, a touch screen, a remote control, a voice interaction, or a handwriting device.

In some embodiments, training device 300 may send the fault pre-warning results of the fault pre-warning model to processor 200.

The execution unit 400 is configured to receive the air conditioner operation parameters sent by the processor 200, and remind a user according to the fault early warning result.

Fig. 2 is a schematic diagram of a fault early warning method applied to a vehicle terminal, where the vehicle terminal includes a fault early warning model.

The application will take a storage battery in a vehicle part as an example, and a specific implementation mode of the fault early warning method will be described. The method comprises the following steps:

s101, acquiring state data of the vehicle in a first preset time period.

The first preset time period is a period of time before the current moment. Illustratively, the first preset time period may be 2 days or 3 days. Optionally, the first preset time period includes a current time.

Wherein the status data includes a trend of change in battery current of the vehicle.

In some embodiments, the state data of the vehicle is uploaded to the cloud end in real time, so that the state data of the vehicle in the first preset period of time can be extracted from the cloud end server, as shown in fig. 3, the above step S101 may be implemented as the following steps S1011-S1012.

S1011, determining the category of the state data to be extracted.

In some embodiments, the cloud server determines a type of state data to be extracted according to a state data list uploaded by a controller area network (Controller Area Network, CAN) and state data with higher fault correlation with the storage battery, which is part of the past state data. Alternatively, the category of the state data to be extracted may be determined according to the advice of the battery expert.

Wherein the categories of the status data include: voltage of the storage battery, current of the storage battery, temperature outside the vehicle, temperature inside the vehicle, power state feedback, engine speed, engine temperature and sensor temperature.

For example, in this embodiment, the cloud server extracts 13 types of status data according to the status data list uploaded by the CAN and the status data having a higher correlation with the battery fault. Wherein the categories of the status data include: the voltage of the storage battery, the current of the storage battery, the temperature outside the vehicle, the power state feedback, the engine speed and the like.

S1012, extracting corresponding state data from the cloud server according to the type of the state data required to be extracted.

In some embodiments, after the corresponding state data is extracted from the cloud server, the state data needs to be processed first, and it CAN be understood that, because the state data of the vehicle is transmitted to the cloud server through the CAN network in the vehicle, and is processed and stored by the cloud server, and the state data is relatively complex, the state data of the vehicle CAN cause problems such as data deletion, duplication, abnormality, and the like in the process of storing the state data of the vehicle to the cloud server, and therefore, the vehicle data needs to be preprocessed.

Illustratively, the extracted state data is checked for the presence of duplicate entries, null values, and outliers. If the repeated item exists, only the first piece of data is reserved in the repeated pieces of data; if the null value exists, selecting the next data or the last data to fill, and avoiding the condition that the first data and the last data are null values; if the abnormal value exists, deleting the abnormal value whole strip data.

S102, determining M starting and stopping time intervals of the vehicle in a first preset time period according to the change trend of the storage battery current.

Wherein M is an integer greater than or equal to 1.

It can be understood that when the vehicle is started, the storage battery can provide current for the starter, and when the vehicle is started, the current provided by the storage battery is very high, and after the vehicle is started, the current provided by the storage battery tends to be in a stable state, so that the starting and stopping time interval of the vehicle can be divided according to the current change trend of the storage battery.

In some embodiments, as shown in fig. 4, the step S102 may be specifically implemented as: steps S1021-S1022.

S1021, determining M+1 time nodes where the storage battery current is suddenly changed in a first preset time period according to the change trend of the storage battery current.

It can be understood that when the vehicle is started, the storage battery provides higher current for the starter, the change trend of the current of the storage battery is that the current is suddenly changed, and after the starter is started, the current provided by the storage battery for the starter tends to be stable, and the current of the storage battery is smaller and the current change trend is more stable. Therefore, m+1 time nodes at which the battery current is suddenly changed in the first preset time period can be determined according to the change trend of the battery current.

And S1022, dividing the first preset time period into M starting and stopping time intervals according to M+1 time nodes with abrupt change of the current of the storage battery.

In some embodiments, the time interval between two adjacent time nodes of the m+1 time nodes is taken as a start-stop time interval. Specifically, a time node of abrupt change of the current of the storage battery is used as a starting point of a starting and stopping time interval; and taking the time node of the current mutation of the next storage battery as the end point of the starting and stopping time interval.

Alternatively, the battery typically provides 150A-200A of current to the starter at vehicle start-up. Therefore, the current level at which the battery current suddenly changes can be set to 150A, and when the battery current reaches 150A, it is regarded that the battery current suddenly changes.

For example, the time nodes of the abrupt change of the battery current are arranged in time sequence, the first time node when the battery current reaches 150A is used as the starting point of the first start-stop time interval, the second time node when the battery current reaches 150A is used as the end point of the first start-stop time interval, and the second time node is also used as the starting point of the second start-stop time interval, and the first preset time period can be divided into M start-stop time intervals by the same method.

It can be understood that when the first preset time period is actually divided, since the time intervals of each start and stop of the vehicle are not consistent, the time length of each start and stop time interval is different, and the number of the start and stop time intervals is also different according to the different time lengths of the first preset time period and the actual start and stop conditions of the vehicle, which is not limited by the embodiment of the present application.

It can be understood that, compared with the prior art, the method provided by the application can be used for dividing the vehicle start-stop time interval according to the time of starting and extinguishing the vehicle or dividing the vehicle start-stop time interval through a fixed time threshold value, which can cause the problem of inaccurate division of the vehicle state data.

S103, dividing the state data in a first preset time period according to M start-stop time intervals to obtain M state data sets.

Wherein one set of status data is the status data of the vehicle during one start-stop time interval.

In some embodiments, as shown in fig. 5, the step S103 may be specifically implemented as: steps S1031 to S1033.

S1031, dividing the state data in the first preset time period according to M starting and stopping time intervals.

Specifically, the M start-stop time intervals are arranged in time sequence, and state data in a first preset time period are divided into different start-stop intervals in time. Wherein, in each start-stop interval, the category of the contained state data is consistent.

S1032, extracting the characteristics of the state data in each start-stop section, and carrying out normalization processing on the characteristics of the state data.

In some embodiments, the step S1032 may be specifically implemented as the following steps:

and a1, extracting the characteristics of each type of state data in a time domain and a frequency domain in each start-stop interval.

For example, the salient features of the state data may be analyzed by analyzing those features of the state data that change over time in each of the start-stop time intervals.

Alternatively, if the feature of the state data does not change significantly in the time domain or the extraction effect is not good, the feature of the state data may be extracted in the frequency domain by fourier transform.

For example, taking the class of battery voltages in the state data as an example, the extracted features in the time domain of the battery voltages include: maximum, minimum, mean, variance, kurtosis, skewness, and waveform factor; the extracted features in the frequency domain of the battery voltage include: maximum, minimum, mean, center of gravity frequency, average frequency, root mean square frequency.

Illustratively, there are 126 features in the time and frequency domains from which each category of state data is extracted.

It will be appreciated that the category of state data within each of the start-stop time intervals, as well as the number of features and features extracted for each category of state data, is the same, although the length of time varies for each start-stop time interval.

And a2, screening out the characteristics of the state data to be normalized, and carrying out normalization processing on the characteristics.

In some embodiments, the state data needs to be normalized because its features are classified into dimensionless features and dimensionless features. The normalization processing method comprises the following steps: linear function normalization, maximum and minimum normalization, and neural network normalization.

Alternatively, in the embodiment of the present application, the normalization method may be a maximum minimum normalization method.

S1033, dividing the normalized state data in a first preset time period according to the M start-stop time intervals to obtain M state data sets.

S104, aggregating the state data in the M state data sets to obtain a target state data set.

In some embodiments, as shown in fig. 6, the step S104 may be specifically implemented as: steps S1041 to S1042.

S1041, respectively aggregating each type of state data in the M state data sets to obtain target state data of each type of state data in the M state data sets.

In some embodiments, each feature of each type of state data in the M sets of state data is aggregated to obtain M sets of state data, each type of state data targeting state data.

Wherein each feature of each state data may be aggregated in a mean-taking manner.

Taking the storage battery voltage as an example, carrying out average value aggregation on the characteristic of the maximum value of the storage battery voltage in the M state data sets to obtain the target state characteristic of the maximum value of the storage battery voltage in the M state data sets; the method comprises the steps of carrying out average value aggregation on the characteristic of the peak value of the storage battery voltage in M state data sets to obtain the target state characteristic of the peak value of the storage battery voltage in the M state data sets; similarly, the target state characteristics of all the characteristics of the obtained battery voltage are taken as target state data of the battery voltage.

Taking the temperature of the storage battery as an example, carrying out average value aggregation on the characteristic of the minimum value of the temperature of the storage battery in M state data sets to obtain the target state characteristic of the minimum value of the temperature of the storage battery in M state data sets; carrying out mean value aggregation on the characteristic of the peak value of the temperature of the storage battery in the M state data sets to obtain the target state characteristic of the peak value of the temperature of the storage battery in the M state data sets; similarly, the target state characteristics of all the obtained characteristics of the battery temperature are taken as the target state data of the battery temperature.

S1042, obtaining a target state data set according to the target state data of each type of state data in the M state data sets.

In some embodiments, the target state data of each type of state data obtained in step S1041 described above is collected as a target state data set.

It can be understood that the fault state of the parts of the vehicle not only depends on the state data in the current starting and stopping time interval of the vehicle, but also is related to the state data in the historical starting and stopping time interval of the vehicle, if the state data in the single starting and stopping time interval is input into the fault early warning model to perform fault early warning on the vehicle, the problem that the fault state of the parts of the vehicle is affected by the state data in the historical starting and stopping time interval is possibly caused to be lower in accuracy of the fault early warning is ignored, so that the method provided by the application considers the influence of the state data in the historical starting and stopping time interval on the fault state of the parts of the vehicle, the fault characteristics of the obtained target state data are more obvious by aggregating the state data in a plurality of state data sets, the current state of the vehicle can be reflected, the input state data is more representative by taking the target state data as the input of the fault early warning model, and the accuracy of the fault early warning is improved.

S105, determining a fault early warning result according to the target state data set and the fault early warning model.

In some embodiments, the target state data set is input into a fault early warning model, so that a fault early warning result of the vehicle can be obtained. The method provided by the application is exemplified by a part of a storage battery. Therefore, after the target state data set in the embodiment of the application is input into the fault early-warning model, the fault early-warning result of the part of the storage battery can be obtained. The fault early warning result comprises the following steps: fault and non-fault.

Compared with the prior art that a fixed time threshold is used for determining the starting and stopping time interval of the vehicle, and then the state data of the vehicle is divided according to the starting and stopping time interval, the problem that the fixed time threshold is difficult to determine or the state data of the vehicle in one starting and stopping time interval is incomplete exists.

Fig. 7 is a schematic diagram of a model training method according to an embodiment of the present application, and a specific implementation of the model training method will be described by taking a storage battery in a vehicle component as an example. The method comprises the following steps: S201-S205.

S201, acquiring state data of the vehicle in a second preset time period.

Wherein the second preset time period comprises a period of time before and after vehicle repair; the state data includes a trend of change in battery current of the vehicle.

In some embodiments, the repair time of the vehicle with the battery fault and the frame number of the vehicle are determined according to the after-sale repair record card of the vehicle, and then state data of a period of time before and after the repair of the vehicle are extracted from the cloud server according to the frame number of the vehicle.

Illustratively, in an embodiment of the present application, status data of 500 vehicles in which the storage battery has failed is extracted. Wherein the status data includes status data of three months before and two months after the maintenance.

In some embodiments, the specific implementation of acquiring the state data of the vehicle in the second preset period may refer to steps S1011-S1012 described above.

S202, determining N starting and stopping time intervals of the vehicle in a second preset time period according to the change trend of the storage battery current.

Wherein N is an integer greater than or equal to 1.

In some embodiments, as shown in fig. 8, the step S202 may be specifically implemented as: steps S2021 to S2022.

S2021, determining N+1 time nodes where the storage battery current is suddenly changed in a second preset time period according to the change trend of the storage battery current.

In some embodiments, when the vehicle is started, the storage battery provides higher current for the starter, the change trend of the current of the storage battery is that the current is suddenly changed, and after the starter is started, the current provided by the storage battery for the starter is stable, and the current of the storage battery is smaller and the current change trend is stable. Therefore, according to the change trend of the battery current, n+1 time nodes where the battery current is suddenly changed in the second preset time period can be determined.

S2022, dividing the second preset time period into N starting and stopping time intervals according to N+1 time nodes where the current of the storage battery is suddenly changed.

In some embodiments, the time node of abrupt change of the battery current is used as the starting point of a starting and stopping time interval; and taking the next time node of the abrupt change of the current of the storage battery as the end point of the starting and stopping time interval.

For example, the time nodes of the abrupt change of the battery current are arranged in time sequence, the first time node when the battery current reaches 150A is used as the starting point of the first start-stop time interval, the second time node when the battery current reaches 150A is used as the end point of the first start-stop time interval, and the second time node is also used as the starting point of the second start-stop time interval, and the second preset time period can be divided into N start-stop time intervals by the same method.

It can be understood that when the second preset time period is actually divided, since the time intervals of each start and stop of the vehicle are not consistent, the time length of each start and stop time interval is different, and the number of the start and stop time intervals is also different according to the different time lengths of the second preset time period and the actual start and stop conditions of the vehicle, which is not limited by the embodiment of the present application.

It can be understood that compared with the prior art, the method provided by the application has the advantages that compared with the prior art, the historical starting and stopping time interval of the vehicle is divided according to the starting and stopping time of the vehicle or the historical starting and stopping time interval of the vehicle is divided through the fixed time threshold, the problem of inaccurate division of the historical state data of the vehicle can be caused, the time node of abrupt change of the storage battery current is determined through the change trend of the storage battery current, the historical starting and stopping interval is divided according to the time node of abrupt change of the storage battery current, the division of the historical state data can be more accurate, and the training of a fault early warning model is more accurate and reliable.

S203, dividing the state data in the second preset time period according to the N start-stop time intervals to obtain N state data sets.

In some embodiments, the specific implementation of step S203 may refer to steps S1031-S1033.

In some embodiments, the step S203 further includes: and marking the N state data sets according to the maintenance time.

Illustratively, the N state data sets are divided into a pre-repair state data set and a post-repair state data set according to repair time. Wherein, the state data before the storage battery is failed is more likely to be abnormal than the normal data without failure. For example, the state data before failure may be below or outside of normal. Therefore, the state data sets 15 days before maintenance are marked as 1 based on the maintenance time in the after-sales maintenance record card, namely the state data sets are marked as the data sets in the fault state; the remaining state datasets are marked as 0, i.e. representing that these state datasets are in a non-faulty state.

S204, constructing training samples according to the N state data sets.

In some embodiments, as shown in fig. 9, the step S204 may be specifically implemented as: S2041-S2043.

S2041, dividing the N state data sets into a plurality of sample data according to a preset window length by adopting a sliding window algorithm.

The state data set in the preset window length is aggregated into one piece of sample data; the preset window length is the length of K start-stop time intervals.

In some embodiments, the number M of the start-stop time intervals of the vehicle in the first preset time period in step S102 is the same as the number K of the start-stop time intervals in the preset window length.

In some embodiments, when the sliding window algorithm is used to divide the N state data sets, the state data set before maintenance and the state data set after maintenance may be divided according to maintenance time, so as to obtain sample data before maintenance and sample data after maintenance time.

Taking a state data set before maintenance as an example, N state data sets of a vehicle are listed, and t data sets before maintenance are sorted in time sequence, specifically expressed as: d=d ₁ ,D ₂ ,D ₃ ,…,D _t And (D), wherein D _i Indicating that the ith state data set of the vehicle is started and stopped, and D indicates all the state data extracted by the vehicle.

Then, the step length of each sliding is set to be 1, the start pointers are respectively start and end, and the start pointers point to the boundaries of the preset window respectively. Initially, start point to D ₁ End is directed to D _k D is to ₁ To D _k And (3) aggregating the K state data sets to obtain first sample data. In the second time, start and end slide backward by 1 step respectively, at which time start points to D ₂ End is directed to D _k+1 D is to ₂ To D _k+1 The K state data sets are aggregated to obtain second sample data; push in this way until end points to the last state data set D _t And obtaining the last piece of sample data.

Wherein the state data sets may be aggregated in a mean or maximum manner.

Taking the battery voltage as an example, D ₁ To D _k Aggregation of K state data sets of (1) includes aggregating K statesIn the data set, the characteristic of the maximum value of the storage battery voltage is subjected to mean value aggregation to obtain the mean value of the maximum values of the storage battery voltage in the K state data sets; carrying out average value aggregation on the characteristic of the peak value of the storage battery voltage in the K state data sets to obtain the maximum value of the peak value of the storage battery voltage in the K state data sets; and similarly, taking the average value of all the obtained characteristics of the battery voltage as the sample data of the battery voltage in the first sample data.

Taking the temperature of the storage battery as an example, D ₁ To D _k The aggregation of the K state data sets comprises the steps of carrying out average value aggregation on the characteristic of the minimum value of the temperature of the storage battery in the K state data sets to obtain the target state characteristic of the minimum value of the temperature of the storage battery in the N state data sets; the method comprises the steps of carrying out average value aggregation on the characteristic of the peak value of the temperature of the storage battery in N state data sets to obtain the target state characteristic of the peak value of the temperature of the storage battery in the N state data sets; and similarly, taking the target state characteristics of all the obtained characteristics of the battery temperature as the sample data of the battery temperature in the first sample data.

Alternatively, in the embodiment of the present application, the preset window length may be the length of 7 start-stop time intervals.

S2042, determining the label of each piece of sample data according to the fault condition of the vehicle in each preset window length.

Wherein, the label includes: failure and no failure.

In some embodiments, labeling results for K state data sets each time state data in the K state data sets is aggregated are determined for sample data prior to the repair time. Counting the number count of state data sets marked as 1 in the K state data sets, and marking each piece of sample data by adopting a preset marking rule according to the relation between the count and the K.

Exemplary, the preset labeling rules are: when count is more than or equal to K/2, the label of the sample data is 1, namely if half of the K state data sets are marked as 1, the label of the sample data obtained after aggregation is 1. Wherein, the label of the sample data which does not meet the preset labeling rule is 0.

In some embodiments, for sample data after the repair time, the tag for each piece of sample data is 0.

Where 1 indicates a failure and 0 indicates no failure.

S2043, constructing a training sample according to the plurality of pieces of sample data and the labels of each piece of sample data.

It can be understood that the fault state of the part of the vehicle not only depends on the state data of the vehicle in a single start-stop time interval, but also is related to the state data in a historical start-stop time interval of the vehicle, compared with the prior art, the method has the advantages that the state data in one start-stop time interval and the labels thereof are used as one piece of sample data, the influence of the state data in the historical start-stop time interval on the fault state of the part of the vehicle is ignored, the problem that the accuracy of the obtained fault early-warning model is low is possibly caused, and therefore the influence of the state data in the historical start-stop time interval on the fault state of the part of the vehicle is considered.

S205, training a fault early-warning model according to the training sample to obtain a trained fault early-warning model.

The fault early warning model is used for sending a fault early warning result to a user.

In some embodiments, as shown in fig. 10, the step S205 may be specifically implemented as: steps S2051 to S2052.

S2051, dividing the training sample into a training set and a testing set.

The training set is used for training and parameter optimization of the fault early-warning model, and the testing set is used for evaluating whether the performance of the fault early-warning model reaches an expected target.

Alternatively, the ratio of the number of sample data for the training set and the test set may be 9:1.

S2052, selecting a machine learning model, and performing model training and parameter optimization on the fault early warning model to obtain a trained fault early warning model.

Alternatively, in an embodiment of the application, the machine learning model may be a gradient lifting (extreme gradient boosting, XGBoost) model.

It can be understood that, unlike the prior art, when training the fault early warning model, a fixed time threshold is generally used to determine the historical starting and stopping time interval of the vehicle, and the historical state data of the vehicle is divided according to the historical starting and stopping time interval of the vehicle, so that the problem that the fixed time threshold is difficult to determine, or the historical state data of the vehicle in one starting and stopping time interval is incomplete, is solved.

In order to facilitate understanding, the embodiments of the fault early-warning method and the model training method provided by the application are further described below in an exemplary manner.

Exemplary, as shown in fig. 11, a flowchart of the fault early warning method and the model training method provided by the application is shown. In particular, the method comprises the steps of,

and b1, extracting data before maintenance of the automobile storage battery and state data after maintenance from a cloud server.

And b2, carrying out data cleaning on the extracted state data and dividing the state data into different starting and stopping time intervals according to the current of the automobile storage battery.

And b3, respectively extracting the characteristics of the state data in each starting and stopping time interval in the time domain and the frequency domain, and carrying out normalization processing on the characteristics.

And b4, dividing the processed state data into before-maintenance and after-maintenance according to the maintenance time, and marking the state data.

And b5, aggregating state data in a plurality of starting and stopping time intervals into one sample by adopting a sliding window algorithm, and re-labeling to reconstruct a new training sample.

And b6, selecting a proper machine learning method to establish a storage battery fault early warning model.

And b7, predicting the current storage battery fault result of the vehicle through the state data acquired by the sensor of the vehicle and the historical state data of the vehicle.

The foregoing description of the solution provided by the embodiments of the present application has been mainly presented in terms of a method. In order to achieve the above functions, the positioning device or the electronic device of the vehicle includes a hardware structure and/or a software module that perform the respective functions. Those of skill in the art will readily appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as hardware or combinations of hardware and computer software. Whether a function is implemented as hardware or computer software driven hardware depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

According to the method, the functional modules of the positioning device or the electronic device of the vehicle can be divided, for example, the positioning device or the electronic device of the vehicle can comprise each functional module corresponding to each functional division, and two or more functions can be integrated into one processing module. The integrated modules may be implemented in hardware or in software functional modules. It should be noted that, in the embodiment of the present application, the division of the modules is schematic, which is merely a logic function division, and other division manners may be implemented in actual implementation.

Fig. 12 is a block diagram of a fault early warning apparatus according to an exemplary embodiment, applied to a vehicle terminal; the vehicle terminal includes a fault early warning model. The fault early warning device 500 includes: an acquisition module 501, a determination module 502, a division module 503, and an aggregation module 504.

An obtaining module 501, configured to obtain status data of a vehicle in a first preset period of time; the state data includes a trend of change in battery current of the vehicle.

The determining module 502 is configured to determine M start-stop time intervals of the vehicle in a first preset time period according to a variation trend of the battery current; m is an integer greater than or equal to 1.

A dividing module 503, configured to divide the state data in the first preset time period according to the M start-stop time intervals, to obtain M state data sets; wherein one set of status data is the status data of the vehicle during one start-stop time interval.

The aggregation module 504 is configured to aggregate the state data in the M state data sets to obtain a target state data set.

The determining module 502 is further configured to determine a fault early warning result according to the target state data set and the fault early warning model.

In a possible implementation manner, the determining module 502 is specifically configured to determine m+1 time nodes where the battery current is suddenly changed in the first preset time period according to a change trend of the battery current; and dividing the first preset time period into M starting and stopping time intervals according to M+1 time nodes where the current of the storage battery is suddenly changed.

In a possible implementation manner, the aggregation module 504 is specifically configured to aggregate each type of state data in the M state data sets to obtain target state data of each type of state data in the M state data sets; and obtaining a target state data set according to the target state data of each type of state data in the M state data sets.

It can be understood that compared with the prior art that a fixed time threshold is used for determining the starting and stopping time interval of the vehicle, and then the state data of the vehicle is divided according to the starting and stopping time interval, the problem that the fixed time threshold is difficult to determine or the state data of the vehicle in one starting and stopping time interval is incomplete can exist.

Fig. 13 is a block diagram of a model training apparatus 600 according to an exemplary embodiment, the model training apparatus 600 comprising: an acquisition module 601, a determination module 602, a construction module 603, and a training module 604.

An acquiring module 601, configured to acquire status data of a vehicle in a second preset period of time; the second preset time period comprises a period of time before vehicle maintenance and a period of time after vehicle maintenance; the state data includes a trend of change in battery current of the vehicle.

The determining module 602 is configured to determine N start-stop time intervals of the vehicle in a second preset time period according to a variation trend of the battery current; n is an integer greater than or equal to 1; the dividing module is used for dividing the state data in the second preset time period according to the N starting and stopping time intervals to obtain N state data sets; wherein one set of status data is the status data of the vehicle during one start-stop time interval.

A construction module 603 is configured to construct training samples according to the N state data sets.

The training module 604 is configured to train the fault early-warning model according to the training sample, to obtain a trained fault early-warning model; the fault early warning model is used for sending a fault early warning result to a user.

In a possible implementation manner, the determining module 602 is specifically configured to determine, according to a trend of the battery current, n+1 time nodes where the battery current is suddenly changed in the second preset time period; and dividing the second preset time period into N starting and stopping time intervals according to the N+1 time nodes with abrupt change of the current of the storage battery.

In a possible implementation manner, the construction module 603 is specifically configured to divide the N state data sets into a plurality of sample data according to a preset window length by using a sliding window algorithm; the state data set in the preset window length is aggregated into one piece of sample data; the length of the preset window is the length of K starting and stopping time intervals; determining a label of each piece of sample data according to the fault condition of the vehicle in each preset window length; the label includes: failure and no failure; and constructing a training sample according to the plurality of pieces of sample data and the label of each piece of sample data.

The specific manner in which the various modules perform the operations in the apparatus of the above embodiments have been described in detail in connection with the embodiments of the method, and will not be described in detail herein.

Fig. 14 is a block diagram of a vehicle, according to an exemplary embodiment. As shown in fig. 14, vehicle 700 includes, but is not limited to: a processor 701 and a memory 702.

The memory 702 is configured to store executable instructions of the processor 701. It will be appreciated that the processor 701 is configured to execute instructions to implement the fault pre-warning method and the model training method in the above embodiments.

It should be noted that the vehicle structure shown in fig. 14 is not limiting of the vehicle, and the vehicle may include more or fewer components than shown in fig. 14, or may combine some components, or a different arrangement of components, as will be appreciated by those skilled in the art.

The processor 701 is a control center of the vehicle, connects various parts of the entire vehicle using various interfaces and lines, and performs various functions of the vehicle and processes data by running or executing software programs and/or modules stored in the memory 702, and calling data stored in the memory 702, thereby performing overall monitoring of the vehicle. The processor 701 may include one or more processing units. Alternatively, the processor 701 may integrate an application processor that primarily handles operating systems, user interfaces, applications, etc., with a modem processor that primarily handles wireless communications. It will be appreciated that the modem processor described above may not be integrated into the processor 701.

The memory 702 may be used to store software programs as well as various data. The memory 702 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, application programs (such as a determination unit, a processing unit, etc.) required for at least one functional module, and the like. In addition, the memory 702 may include high-speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other volatile solid-state storage device.

In an exemplary embodiment, a computer readable storage medium is also provided, such as a memory 702, comprising instructions executable by the processor 701 of the vehicle 700 to implement the fault pre-warning method and the model training method of the above embodiments.

In actual implementation, the functions of the acquisition module 501, the determination module 502, the division module 503, and the aggregation module 504 in fig. 12, and the functions of the acquisition module 601, the determination module 602, the construction module 603, and the training module 604 in fig. 13 may be implemented by the processor 701 in fig. 14 calling a computer program stored in the memory 702. For specific execution, reference may be made to the description of the fault early warning method and the model training method in the above embodiments, and details are not repeated here.

Alternatively, the computer-readable storage medium may be a non-transitory computer-readable storage medium, which may be, for example, a read-Only memory (ROM), a random-access memory (Ranbom Accbss Mbmory, RAM), a CB-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

In an exemplary embodiment, embodiments of the application also provide a computer program product comprising one or more instructions executable by the processor 701 of the vehicle to perform the fault pre-warning method and the model training method of the above-described embodiments.

It should be noted that, when the instructions in the computer readable storage medium or one or more instructions in the computer program product are executed by the processor of the vehicle, the respective processes of the embodiments of the foregoing fault early warning method and model training method are implemented, and the technical effects that are the same as those of the foregoing fault early warning method and model training method can be achieved, so that repetition is avoided, and no further description is given here.

From the foregoing description of the embodiments, it will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of functional modules is illustrated, and in practical application, the above-described functional allocation may be performed by different functional modules according to needs, i.e. the internal structure of the apparatus is divided into different functional modules, so as to perform all the classification parts or part of the functions described above.

In the several embodiments provided by the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of modules or units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another apparatus, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.

The units described as separate parts may or may not be physically separate, and the parts shown as units may be one physical unit or a plurality of physical units, may be located in one place, or may be distributed in a plurality of different places. The purpose of the embodiment scheme can be achieved by selecting part or all of the classification part units according to actual needs.

In addition, each functional unit in the embodiments of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a readable storage medium. Based on such understanding, the technical solution of the embodiments of the present application, or the part contributing to the prior art or the whole classification part or part of the technical solution, may be embodied in the form of a software product stored in a storage medium, including several instructions for causing a device (which may be a single-chip microcomputer, a chip or the like) or a processor (procbssor) to perform the whole classification part or part of the steps of the method of the embodiments of the present application. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a ROM, a RAM, a magnetic disk, or an optical disk, etc.

The present application is not limited to the above embodiments, and any changes or substitutions within the technical scope of the present application should be covered by the scope of the present application. Therefore, the protection scope of the application is subject to the protection scope of the claims.

Claims

1. The fault early warning method is characterized by being applied to a vehicle terminal; the vehicle terminal comprises a fault early warning model; the method comprises the following steps:

Acquiring state data of a vehicle in a first preset time period; the state data includes a trend of change in battery current of the vehicle;

according to the change trend of the storage battery current, M starting and stopping time intervals of the vehicle in the first preset time period are determined; m is an integer greater than or equal to 1;

dividing the state data in the first preset time period according to the M start-stop time intervals to obtain M state data sets; wherein one state data set is the state data of the vehicle in a starting and stopping time interval;

aggregating the state data in the M state data sets to obtain a target state data set;

and determining a fault early warning result according to the target state data set and the fault early warning model.

2. The method of claim 1, wherein the determining M start-stop time intervals of the vehicle within the first preset time period according to the trend of the battery current comprises:

according to the change trend of the storage battery current, M+1 time nodes with abrupt change of the storage battery current in the first preset time period are determined;

And dividing the first preset time period into M starting and stopping time intervals according to M+1 time nodes where the current of the storage battery is suddenly changed.

3. The method of claim 1, wherein aggregating the state data in the M state data sets to obtain a target state data set comprises:

respectively aggregating each type of state data in the M state data sets to obtain target state data of each type of state data in the M state data sets;

and obtaining a target state data set according to the target state data of each type of state data in the M state data sets.

4. A method of model training, the method comprising:

acquiring state data of the vehicle in a second preset time period; the second preset time period comprises a period of time before vehicle maintenance and a period of time after vehicle maintenance; the state data includes a trend of change in battery current of the vehicle;

according to the change trend of the storage battery current, N starting and stopping time intervals of the vehicle in the second preset time period are determined; the N is an integer greater than or equal to 1;

dividing the state data in the second preset time period according to the N start-stop time intervals to obtain N state data sets; wherein one state data set is the state data of the vehicle in a starting and stopping time interval;

Constructing training samples according to the N state data sets;

training a fault early warning model according to the training sample to obtain a trained fault early warning model; the fault early warning model is used for sending a fault early warning result to a user.

5. The method of claim 4, wherein determining N start-stop time intervals of the vehicle within the second preset time period according to the trend of the battery current comprises:

according to the change trend of the storage battery current, determining N+1 time nodes with abrupt change of the storage battery current in the second preset time period;

and dividing the second preset time period into N starting and stopping time intervals according to the N+1 time nodes with abrupt change of the storage battery current.

6. The method of claim 4, wherein constructing training samples from the N state data sets comprises:

dividing the N state data sets into a plurality of sample data according to a preset window length by adopting a sliding window algorithm; the state data set in the preset window length is aggregated into one piece of sample data; the length of the preset window is the length of K starting and stopping time intervals;

Determining a label of each sample data according to the fault condition of the vehicle in each preset window length; the tag includes: failure and no failure;

and constructing the training sample according to the plurality of pieces of sample data and the labels of each piece of sample data.

7. The fault early warning device is characterized by being applied to a vehicle terminal; the vehicle terminal comprises a fault early warning model; the fault early warning device comprises:

the acquisition module is used for acquiring state data of the vehicle in a first preset time period; the state data includes a trend of change in battery current of the vehicle;

the determining module is used for determining M starting and stopping time intervals of the vehicle in the first preset time period according to the change trend of the storage battery current; m is an integer greater than or equal to 1;

the dividing module is used for dividing the state data in the first preset time period according to the M starting and stopping time intervals to obtain M state data sets; wherein one state data set is the state data of the vehicle in a starting and stopping time interval;

the aggregation module is used for aggregating the state data in the M state data sets to obtain a target state data set;

And the determining module is also used for determining a fault early warning result according to the target state data set and the fault early warning model.

8. The apparatus of claim 7, wherein the device comprises a plurality of sensors,

the determining module is specifically configured to determine m+1 time nodes where the battery current changes suddenly in the first preset time period according to the change trend of the battery current; and dividing the first preset time period into M starting and stopping time intervals according to M+1 time nodes where the current of the storage battery is suddenly changed.

9. A model training device, comprising:

the acquisition module is used for acquiring state data of the vehicle in a second preset time period; the second preset time period comprises a period of time before vehicle maintenance and a period of time after vehicle maintenance; the state data includes a trend of change in battery current of the vehicle;

the determining module is used for determining N starting and stopping time intervals of the vehicle in the second preset time period according to the change trend of the storage battery current; the N is an integer greater than or equal to 1;

the dividing module is used for dividing the state data in the second preset time period according to the N starting and stopping time intervals to obtain N state data sets; wherein one state data set is the state data of the vehicle in a starting and stopping time interval;

The construction module is used for constructing training samples according to the N state data sets;

the training module is used for training the fault early-warning model according to the training sample to obtain a trained fault early-warning model; the fault early warning model is used for sending a fault early warning result to a user.

10. The apparatus of claim 9, wherein the device comprises a plurality of sensors,

the determining module is specifically configured to determine n+1 time nodes where the battery current changes suddenly in the second preset time period according to the change trend of the battery current; and dividing the second preset time period into N starting and stopping time intervals according to the N+1 time nodes with abrupt change of the storage battery current.

11. A vehicle, characterized by comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor executing the program to implement the method of any one of claims 1 to 6.

12. A computer readable storage medium, characterized in that, when computer-executable instructions stored in the computer readable storage medium are executed by a processor of an electronic device, the electronic device is capable of performing the method of any one of claims 1 to 6.