Automobile motor fault prediction method and device and electronic equipment

Abstract

The embodiment of the invention discloses an automobile motor fault prediction method and device and electronic equipment. The method comprises the steps of obtaining at least two pieces of historical operation data of an automobile motor, preprocessing each piece of historical operation data to obtain historical standard data, then training an initial fault prediction model according to each piece of historical standard data to obtain a trained fault prediction model, and performing fault prediction according to the trained fault prediction model. And finally, predicting the obtained real-time operation data based on the trained fault prediction model to obtain a real-time fault prediction result. The requirement for predicting the automobile motor fault in real time is met through the constructed fault prediction model, the accuracy and reliability of fault prediction are improved by introducing environmental noise-resistant data, the application range of the model is widened, more comprehensive and earlier fault early warning is achieved, and safe and stable operation of the automobile motor is effectively guaranteed.

Description

TECHNICAL FIELD

[0001] One or more embodiments of the present specification relate to the technical field of failure prediction, and in particular to a motor fault prediction method for a vehicle. BACKGROUND

[0002] With the wide application of new energy vehicles, the motor of a vehicle as one of its core components, its running state is directly related to the safety, reliability and overall performance of the vehicle. Various faults may occur in the long-term operation of the motor of a vehicle, such as overheating, insulation aging, winding short circuit, bearing wear, etc. If these faults cannot be found and handled in time, it will lead to vehicle breakdown and even cause safety accidents. At present, the traditional motor fault prediction method for a vehicle is mostly periodic maintenance inspection or experience-based judgment, which has the disadvantages of poor timeliness, low accuracy, etc., and it is difficult to meet the high-precision and real-time requirements of new energy vehicles for motor fault prediction. Therefore, an effective motor fault prediction method is urgently needed. SUMMARY

[0003] The embodiments of the present specification provide a motor fault prediction method, device and electronic equipment for a vehicle, and the technical solutions are as follows: In a first aspect, the embodiments of the present specification provide a motor fault prediction method for a vehicle, and the method comprises: obtaining at least two historical running data of a motor of a vehicle, and pre-processing each of the historical running data to obtain historical standard data, wherein each of the historical running data comprises historical motor data and historical environmental data; training an initial fault prediction model according to each of the historical standard data to obtain a trained fault prediction model; predicting real-time running data obtained based on the trained fault prediction model to obtain a real-time fault prediction result.

[0004] In a second aspect, a motor fault prediction device for a vehicle is provided, and the device comprises: an obtaining module configured to obtain at least two historical running data of a motor of a vehicle, and pre-process each of the historical running data to obtain historical standard data, wherein each of the historical running data comprises historical motor data and historical environmental data; a training module configured to train an initial fault prediction model according to each of the historical standard data to obtain a trained fault prediction model; a prediction module configured to predict real-time running data obtained based on the trained fault prediction model to obtain a real-time fault prediction result.

[0005] In a third aspect, an electronic equipment is provided, comprising a device processor and a memory. The device processor is connected with the memory. the memory, for storing executable program code; The device processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, to execute the steps of the method provided in the first aspect or any possible implementation manner of the first aspect.

[0006] In a fourth aspect, a computer readable storage medium is provided, which stores a computer program, and the computer readable storage medium stores instructions, which, when executed on a computer or device processor, cause the computer or device processor to perform the method provided in the first aspect or any possible implementation manner of the first aspect.

[0007] The technical solutions provided by some embodiments of the present specification have at least the following beneficial effects: In one or more embodiments of the present specification, at least two historical running data of the automobile motor are obtained, and each historical running data is preprocessed to obtain historical standard data, then the initial fault prediction model is trained according to each historical standard data to obtain a trained fault prediction model, and finally the real-time running data obtained is predicted based on the trained fault prediction model to obtain a real-time fault prediction result. Through the constructed fault prediction model, the requirement of real-time prediction of automobile motor fault is met, the accuracy and reliability of fault prediction are improved by introducing environmental anti-noise data, the application range of the model is widened, more comprehensive and earlier fault warning is realized, and the safe and stable operation of the automobile motor is effectively guaranteed. BRIEF DESCRIPTION OF DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0009] Figure 1 A flow chart of an automobile motor fault prediction method provided by an embodiment of the present specification; Figure 2 A structural schematic diagram of an automobile motor fault prediction device provided by an embodiment of the present specification; Figure 3 A structural schematic diagram of an electronic device provided by an embodiment of the present specification. DETAILED DESCRIPTION

[0010] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application.

[0011] The terms "first", "second", "third", and the like in the description and the claims of the present specification and the above-described drawings are used to distinguish different objects, and are not used to describe a particular order. In addition, the terms "comprise" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but can optionally include steps or units not listed, or can optionally include other steps or units inherent to such processes, methods, products, or devices.

[0012] The following description provides examples and does not limit the scope, applicability, or examples set forth in the claims. Changes can be made in the function and arrangement of elements described without departing from the scope of the present specification. Various examples can omit, substitute, or add various procedures or components as appropriate. For example, the described methods can be performed in a different order than described, and various steps can be added, omitted, or combined. Furthermore, features described with respect to some examples can be combined in other examples.

[0013] See Figure 1 , Figure 1 The overall flowchart of an automobile motor fault prediction method provided by an embodiment of the present specification is shown.

[0014] As Figure 1 shown, the automobile motor fault prediction method can at least include the following steps: Step 101, obtaining at least two historical running data of an automobile motor, and pre-processing each of the historical running data to obtain historical standard data.

[0015] Each of the historical running data includes historical motor data and historical environment data.

[0016] In the embodiment of the present specification, the automobile motor fault prediction method can be applied to an automobile motor fault prediction system which can include a vehicle-mounted server, a cloud server and a collection sensor. In order to improve the prediction rate and reduce the delay effect to achieve fast response, the historical operation data corresponding to the automobile motor can be collected by the collection sensor, then transmitted to the cloud server to train the prediction model and deployed to the vehicle-mounted server in real time, and finally the real-time operation data collected by the collection sensor is predicted by the vehicle-mounted server. In order to consider various automobile motor operation parameters and environmental factors, the running state of the automobile motor is comprehensively reflected from different angles, so that the fault prediction result is more comprehensive and objective, the potential fault characteristics of the automobile motor can be more accurately captured, the identification ability of early faults and complex faults is improved, and the historical operation data obtained includes historical motor data and historical environmental data. Specifically, the motor data includes but is not limited to current, voltage, speed, temperature, vibration acceleration, vibration frequency, noise intensity, torque, shaft eccentricity, bearing clearance, power factor, active power, reactive power, stator winding resistance, starting current, starting time, insulation resistance, motor efficiency, etc., and the environmental data includes but is not limited to the environmental parameters of the motor, such as environmental temperature, humidity, dust concentration, altitude, etc.

[0017] Then, since the historical operation data obtained can include abnormal, missing and other multi-format data, subsequent pre-processing of each historical operation data is required to obtain historical standard data for subsequent model training.

[0018] In an implementation manner, the pre-processing of each historical operation data to obtain historical standard data includes: abnormal correction of each historical operation data based on a spline interpolation method to obtain historical interpolation data; normalization processing of each historical interpolation data according to a Z-score standard method to obtain historical standard data.

[0019] In the embodiment of the present specification, since the sensor can have abnormal conditions such as data packet loss during actual data collection and transmission, resulting in missing and other abnormal states in the historical operation data, the spline interpolation method is required to correct the abnormality of each historical operation data to obtain historical interpolation data. Specifically, taking the temperature data in the historical operation data as an example, when using the spline interpolation method, the data points of each temperature data can be sorted first, then a cubic spline function is constructed through a cubic polynomial, and the target data point is ensured to be passed through and the curve is ensured to be smooth at each data node. Then, the mean value Ut and the standard deviation Vt in the sliding window are calculated, and if , it is determined that an exception occurs, the exception value is replaced by Ut to perform exception correction to obtain historical interpolation data. Further, each historical interpolation data of arbitrary distribution is normalized by using a normalization formula according to a Z-score standard method, so as to eliminate the difference in format and dimension, and historical standard data is obtained, so as to improve the convergence speed of a subsequent neural network model.

[0020] In an implementable manner, after the spline interpolation method is used to correct each historical operation data to obtain historical interpolation data, the method further includes: generating environmental adversarial noise data corresponding to the historical environmental data based on an FGSM algorithm; merging each historical interpolation data and the environmental adversarial noise data to obtain historical complete data; The normalization processing of each historical interpolation data according to the Z-score standard method includes: The normalization processing of each historical complete data according to the Z-score standard method includes:

[0021] In the embodiments of the present disclosure, since the physical boundary of environmental data (such as PM2.5 concentration) is wider than that of motor data, and the abnormal mode caused by environmental mutation (such as heavy rain) is different from equipment failure, in order to maintain the coupling relationship between environmental data and motor data, resist false alarms caused by sensor abnormal fluctuations, and reduce the false alarm rate, after the historical interpolation data is obtained, the data model gradient is calculated by using a fast gradient sign attack algorithm (FGSM), the gradient direction is corrected according to the environmental parameter physical correlation rule, and finally the environmental adversarial noise data corresponding to the historical environmental data is generated within the preset environmental parameter boundary. As an example, in the high-altitude motor failure prediction scene, the original environmental data is [altitude 2830m, temperature 12℃, humidity 45%, dust 85 ], and the data model gradient is calculated as follows: , which respectively correspond to altitude / temperature / humidity / dust. Then, the gradient is corrected. Since the altitude increases and the temperature decreases, the temperature gradient sign is corrected to , so as to generate the environmental adversarial noise data (perturbation intensity coefficient =0.1) as , and the corresponding physical boundary constraint is the temperature change limit value: , wherein certain environmental adversarial noise data X1=[altitude 2930m, 13℃, 46%, 86 ]. Further, each historical interpolation data and the generated environmental adversarial noise data are merged to obtain historical complete data, and the historical complete data is directly processed when the normalization processing is performed according to the Z-score standard method.

[0022] Step 102, training the initial fault prediction model according to each of the historical standard data to obtain a trained fault prediction model.

[0023] In the embodiments of the present specification, after obtaining each of the historical standard data, in order to perform fault prediction on the subsequently obtained real-time running data, an initial fault prediction model can be constructed first, taking the standard data or the corresponding data features as the model input, and taking the corresponding fault prediction label value as the model input. Then, the initial fault prediction model is trained according to each of the historical standard data to obtain a trained fault prediction model.

[0024] Wherein, when constructing the initial fault prediction model, a deep learning model such as BP neural network or 1D-CNN can be used.

[0025] In one implementation manner, the training of the initial fault prediction model according to each of the historical standard data to obtain a trained fault prediction model comprises: performing feature extraction on each of the historical standard data based on the PCA algorithm to obtain historical data features; integrating the historical data features and the historical fault label values corresponding to each of the historical standard data to obtain a historical training set; training the initial fault prediction model according to the historical training set based on the back propagation method to obtain a trained fault prediction model.

[0026] In the embodiments of the present specification, in order to facilitate the training of the constructed initial fault prediction model, after obtaining each of the historical standard data, the covariance matrix of the historical standard data needs to be decomposed for eigenvalues first through the PCA algorithm, and the eigenvector corresponding to the maximum eigenvalue is selected to construct a projection matrix, so as to map the original high-dimensional data to a low-dimensional orthogonal space, realize the de-redundancy dimension reduction, and obtain the historical data features. Then, the historical fault label values actually corresponding to each of the historical standard data are determined, and each of the historical data features and the historical fault label values are one-to-one mapped and integrated to obtain a historical training set. Finally, the initial fault prediction model is trained by using the back propagation (BP) neural network to obtain a trained fault prediction model, and the data features are taken as the model input when the initial fault prediction model is constructed, and the corresponding fault prediction label value is taken as the model input.

[0027] Step 103, predicting the obtained real-time running data based on the trained fault prediction model to obtain a real-time fault prediction result.

[0028] In the embodiments of the present application, in order to find potential faults in advance during the operation of the automobile motor, provide sufficient time for the prevention and treatment of faults, enable the vehicle to be maintained and serviced in time, avoid vehicle downtime and safety accidents caused by sudden failures, and ensure the normal operation of the automobile and the safety of the life and property of the user, after obtaining the trained fault prediction model, the fault prediction model can be deployed to the vehicle server in real time through the cloud server, and finally the real-time operation data collected by the sensor is predicted for faults through the vehicle server. Specifically, after obtaining the real-time operation data, it also needs to be converted into its corresponding real-time data features, and transmitted to the trained fault prediction model as a model input to obtain a real-time fault prediction result. The real-time fault prediction result can be represented as the probability or state prediction result of various faults of the automobile motor.

[0029] In an implementation manner, the real-time fault prediction result is obtained by predicting the obtained real-time operation data based on the trained fault prediction model, including: a real-time fault label prediction value is obtained by predicting the obtained real-time operation data based on the trained fault prediction model; a real-time fault prediction result is determined according to a comparison result of the real-time fault label prediction value and a preset fault threshold.

[0030] In the embodiments of the present application, since the model output of the fault prediction model is generally a fault label prediction value, and the fault prediction result corresponding to the same fault label prediction value under different fault standards may be different. Therefore, the trained fault prediction model can be used to predict the obtained real-time operation data to obtain a real-time fault label prediction value. Then, the real-time fault label prediction value and the preset fault threshold are compared in value size. When the comparison result represents that the real-time fault label prediction value is greater than the preset fault threshold, it is determined that the real-time fault prediction result at this time is that there is a fault risk. When the comparison result represents that the real-time fault label prediction value is not greater than the preset fault threshold, it is determined that the real-time fault prediction result at this time is that there is no fault risk. The fault threshold can also be determined according to the ratio between the real-time fault label prediction value and the preset fault threshold. The fault threshold can be adaptively adjusted according to the actual needs of the user.

[0031] In an implementation manner, after the real-time fault prediction result is obtained by predicting the obtained real-time operation data based on the trained fault prediction model, including: a fault feature parameter corresponding to the real-time fault prediction result is analyzed based on a LIME interpretation algorithm; the fault feature parameter is weighted and corrected according to the power physical topology relationship to obtain a fault prediction report.

[0032] In the embodiments of the present specification, since the real-time fault prediction result can be represented as the probability of various faults of the automobile motor or the state prediction result, in order to assist maintenance personnel or target users to quickly locate the root cause of the fault and shorten the subsequent maintenance time, after obtaining the real-time fault prediction result, the local interpretability principle in the LIME explanation algorithm can be used to construct a ridge regression linear proxy model by sampling in the neighborhood of the real-time data point, simulate the local decision boundary, and analyze the corresponding fault feature parameters of each fault feature in the prediction result. Then, the power physical topology relationship is introduced as domain knowledge constraint to realize the physical rationality calibration of fault attribution, and the node connection weight (such as the energy transmission path of motor winding-bearing-rotor) is used to correct the fault feature parameters to obtain a fault prediction report. Specifically, when the power physical topology relationship is introduced, an automobile motor topology graph can be constructed, wherein {winding, bearing, rotor, cooling system...} and the like are taken as nodes of the automobile motor topology graph, and the physical connection relationship such as "winding-bearing" is taken as an edge of the automobile motor topology graph.

[0033] As an example, a certain kind of fault prediction report can be "fault type: bearing wear", "root cause: [{feature: "axial vibration acceleration", contribution: 35.2%, physical association: "rotor eccentricity leads to increased bearing load"}, {feature: "winding temperature gradient", contribution: 28.7%, physical association: "insulation aging leads to decreased heat dissipation efficiency"}]".

[0034] In an implementable manner, the method further comprises: determining the predicted fault level corresponding to the fault prediction report based on a fault-level database; executing an active control strategy according to the predicted fault level, and uploading the predicted fault level.

[0035] In the embodiments of the present specification, in order to convert the fault prediction result into an active control instruction and delay the fault deterioration progress under certain special fault conditions, the predicted fault level corresponding to the fault prediction report can be determined according to the pre-established fault-level database. Then, an active control strategy is executed according to the predicted fault level, and the predicted fault level is uploaded. As an example: when the predicted bearing fault level exceeds three levels, the maximum speed of the automobile motor is automatically limited, and the BMS is notified to adjust the power distribution.

[0036] The above describes particular embodiments of the present specification. Other embodiments are within the scope of the appended claims. In some cases, the acts or steps recited in the claims can be performed in a different order than those in the embodiments and still achieve desirable results. In addition, the processes depicted in the figures do not necessarily require the particular order shown or sequential order to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous or possible.

[0037] Next, please refer to Figure 2 , Figure 2 The structure of an automobile motor fault prediction device provided by the embodiments of the present specification is shown. It should be noted that Figure 2 The automobile motor fault prediction device shown is used to execute the method of the embodiments of the present application Figure 1 For ease of illustration, only parts related to the embodiments of the present application are shown, and specific technical details are not disclosed, please refer to the embodiments shown in the present application Figure 1 .

[0038] As Figure 2 shown, the automobile motor fault prediction device can at least include: The acquisition module 201 is configured to acquire at least two historical running data of an automobile motor, and pre-process each of the historical running data to obtain historical standard data, wherein each of the historical running data includes historical motor data and historical environment data; The training module 202 is configured to train an initial fault prediction model according to each of the historical standard data to obtain a trained fault prediction model; The prediction module 203 is configured to predict real-time running data based on the trained fault prediction model to obtain real-time fault prediction results.

[0039] In an implementable manner, the acquisition module 201 is specifically configured to: correct each of the historical running data based on a spline interpolation method to obtain historical interpolation data; normalize each of the historical interpolation data according to a Z-score standard method to obtain historical standard data.

[0040] In an implementable manner, the acquisition module 201 is specifically further configured to: generate environment adversarial noise data corresponding to the historical environment data based on an FGSM algorithm; merge each of the historical interpolation data and the environment adversarial noise data to obtain historical complete data; The normalization of each of the historical interpolation data according to the Z-score standard method includes: According to the Z-score standard method, the historical complete data is normalized.

[0041] In an implementation, the training module 202 is specifically configured to: According to the PCA algorithm, the historical standard data is extracted to obtain historical data features; The historical data features and the historical fault label values corresponding to the historical standard data are integrated to obtain a historical training set; According to the back propagation method, the historical training set is used to train an initial fault prediction model to obtain a trained fault prediction model.

[0042] In an implementation, the prediction module 203 is specifically configured to: Based on the trained fault prediction model, the real-time operation data obtained is predicted to obtain a real-time fault label prediction value; According to the comparison result of the real-time fault label prediction value and a preset fault threshold, a real-time fault prediction result is determined.

[0043] In an implementation, the prediction module 203 is specifically configured to: Based on the LIME interpretation algorithm, a fault feature parameter corresponding to the real-time fault prediction result is analyzed; According to the power physical topology relationship, the fault feature parameter is weighted and corrected to obtain a fault prediction report.

[0044] In an implementation, the prediction module 203 is specifically configured to: Based on the fault-grade database, a prediction fault grade corresponding to the fault prediction report is determined; According to the prediction fault grade, an active control strategy is executed, and the prediction fault grade is uploaded.

[0045] Those skilled in the art can clearly understand that the technical solutions of the embodiments of the present application can be implemented by means of software and / or hardware. The "unit" and "module" in the specification refer to software and / or hardware that can independently complete or cooperate with other components to complete a specific function, and the hardware may, for example, be a field programmable gate array (FPGA), an integrated circuit (IC), etc.

[0046] The processing units and / or modules of the embodiments of the present application can be implemented by analog circuits that implement the functions of the embodiments of the present application, or can be implemented by software that executes the functions of the embodiments of the present application.

[0047] Next, please refer to Figure 3 ,Figure 3 A structural schematic diagram of an electronic device is shown.

[0048] As Figure 3 shown, the electronic device 300 can include at least one device processor 301, at least one network interface 303, a user interface 303, a memory 305, and at least one communication bus 302.

[0049] The communication bus 302 can be used to realize the connection and communication of the above-mentioned components.

[0050] The user interface 303 can include a key, and the optional user interface can also include a standard wired interface, a wireless interface.

[0051] The network interface 304 can include, but is not limited to, a Bluetooth module, an NFC module, a Wi-Fi module, etc.

[0052] The device processor 301 can include one or more processing cores. The device processor 301 connects various parts in the entire electronic device 300 through various interfaces and lines, executes various functions of the electronic device 300 and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 305, and calling data stored in the memory 305. Optionally, the device processor 301 can be implemented in at least one of the hardware forms of DSP, FPGA, and PLA. The device processor 301 can integrate one or a combination of CPU, GPU, and modem, etc. Among them, the CPU mainly processes the operating system, user interface, and application program, etc.; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; and the modem is used for processing wireless communication. It can be understood that the above-mentioned modem can also not be integrated into the device processor 301, but be realized by a separate chip.

[0053] The memory 305 can include RAM and can also include ROM. Optionally, the memory 305 includes a non-transitory computer readable medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 can include a program storage area and a data storage area, wherein the program storage area can store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area can store data involved in the above-mentioned various method embodiments, etc. The memory 305 can also be at least one storage device located away from the above-mentioned device processor 301. Figure 3As shown, the memory 305 as a computer storage medium can include an operating system, a network communication module, a user interface module, and program instructions.

[0054] Specifically, the device processor 301 can be configured to invoke the automobile motor fault prediction application stored in the memory 305, and specifically perform the following operations: Obtain at least two historical running data of the automobile motor, and pre-process each of the historical running data to obtain historical standard data, each of the historical running data comprising historical motor data and historical environment data; Train an initial fault prediction model according to each of the historical standard data to obtain a trained fault prediction model; Predict real-time running data obtained based on the trained fault prediction model to obtain real-time fault prediction results.

[0055] As an option of an embodiment of the present specification, the pre-processing of each of the historical running data to obtain historical standard data comprises: Abnormal correction of each of the historical running data based on a spline interpolation method to obtain historical interpolation data; Normalization processing of each of the historical interpolation data according to a Z-score standard method to obtain historical standard data.

[0056] As an option of an embodiment of the present specification, after the abnormal correction of each of the historical running data based on the spline interpolation method to obtain historical interpolation data, the method further comprises: Generating environment adversarial noise data corresponding to the historical environment data based on an FGSM algorithm; Merging each of the historical interpolation data and the environment adversarial noise data to obtain historical complete data; The normalization processing of each of the historical interpolation data according to the Z-score standard method comprises: Normalization processing of each of the historical complete data according to the Z-score standard method.

[0057] As an option of an embodiment of the present specification, the training of an initial fault prediction model according to each of the historical standard data to obtain a trained fault prediction model comprises: Feature extraction of each of the historical standard data based on a PCA algorithm to obtain historical data features; Integrating the historical data features and historical fault label values corresponding to each of the historical standard data to obtain a historical training set; Training the initial fault prediction model according to the historical training set based on a back propagation method to obtain a trained fault prediction model.

[0058] As an option of the embodiment of the present specification, the predicting, based on the trained fault prediction model, the obtained real-time operation data to obtain a real-time fault prediction result, comprises: predicting, based on the trained fault prediction model, the obtained real-time operation data to obtain a real-time fault label prediction value; determining a real-time fault prediction result according to a comparison result of the real-time fault label prediction value and a preset fault threshold.

[0059] As an option of the embodiment of the present specification, after the predicting, based on the trained fault prediction model, the obtained real-time operation data to obtain a real-time fault prediction result, comprises: analyzing a fault feature parameter corresponding to the real-time fault prediction result based on a LIME interpretation algorithm; weighting and correcting the fault feature parameter according to an electric power physical topology relationship to obtain a fault prediction report.

[0060] As an option of the embodiment of the present specification, the method further comprises: determining a predicted fault level corresponding to the fault prediction report based on a fault-level database; executing an active control strategy according to the predicted fault level, and uploading the predicted fault level.

[0061] The embodiment of the present specification also provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to implement the steps of the above method. The computer readable storage medium can include but is not limited to any type of disk, including a floppy disk, an optical disk, a DVD, a CD-ROM, a micro drive, and a magneto-optical disk, a ROM, a RAM, an EPROM, an EEPROM, a DRAM, a VRAM, a flash memory device, a magnetic card or an optical card, a nano system (including a molecular memory IC), or any type of medium or device suitable for storing instructions and / or data.

[0062] It should be noted that, for the foregoing method embodiments, in order to simply describe, they are all described as a series of action combinations, but those skilled in the art should know that the present application is not limited to the action sequence described, because according to the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions and modules involved are not necessarily necessary for the present application.

[0063] In the above embodiments, the description of each embodiment is focused on, and the part not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0064] In several embodiments provided in the present application, it should be understood that the disclosed apparatus can be implemented in other manners. For example, the division of the apparatus embodiments described above is merely illustrative, and the division of the units can be changed according to actual needs. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0065] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., can be located in one place, or can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiments.

[0066] In addition, each functional unit in the various embodiments of the present application can be integrated into a processing unit, or each unit can be physically present separately, or two or more units can be integrated into one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0067] When the integrated unit is realized in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0068] A person of ordinary skill in the art can understand that all or part of the steps of the various methods in the above embodiments can be completed by a program instructing relevant hardware, and the program can be stored in a computer readable storage medium, which can include a flash disk, a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk, etc.

[0069] The above described embodiments of the present description have been described. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can be performed in a different order and still achieve desirable results. Additionally, the processes depicted in the figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In certain implementations, multitasking and parallel processing can be advantageous.

Claims

1. A method for predicting automobile motor failure, characterized in that: The method comprises: Acquire at least two historical operating data of the automobile motor, and pre-process each of the historical operating data to obtain historical standard data, wherein each of the historical operating data includes historical motor data and historical environment data; Training the initial fault prediction model according to each of the historical standard data to obtain a trained fault prediction model; The acquired real-time operation data is predicted based on the trained fault prediction model to obtain a real-time fault prediction result.

2. The method according to claim 1, characterized in that The preprocessing of each of the historical operation data to obtain historical standard data includes: Performing abnormal correction on each of the historical operating data based on a spline interpolation method to obtain historical interpolation data; The historical interpolation data are normalized according to the Z-score standard method to obtain historical standard data.

3. The method according to claim 2, characterized in that After performing abnormal correction on each of the historical operation data based on the spline interpolation method to obtain historical interpolation data, the method further includes: Generate environmental anti-noise data corresponding to the historical environmental data based on the FGSM algorithm; Merging the historical interpolation data and the environmental noise-combat data to obtain complete historical data; The normalizing process of each of the historical interpolation data according to the Z-score standard method includes: The historical complete data were normalized according to the Z-score standard method.

4. The method according to claim 1, wherein The initial fault prediction model is trained according to each of the historical standard data to obtain a trained fault prediction model, including: Perform feature extraction on each of the historical standard data based on the PCA algorithm to obtain historical data features; Integrating the historical data features and the historical fault label values ​​corresponding to each of the historical standard data to obtain a historical training set; The historical training set is used to train the initial fault prediction model according to the back propagation method to obtain a trained fault prediction model.

5. The method according to claim 1, wherein The real-time operating data acquired is predicted based on the trained fault prediction model to obtain a real-time fault prediction result, including: Predicting the acquired real-time operating data based on the trained fault prediction model to obtain a real-time fault label prediction value; The real-time fault prediction result is determined based on a comparison result of the real-time fault tag prediction value and a preset fault threshold.

6. The method according to claim 1, characterized in that After predicting the acquired real-time operation data based on the trained fault prediction model to obtain a real-time fault prediction result, the method includes: Analyze the fault characteristic parameters corresponding to the real-time fault prediction result based on the LIME interpretation algorithm; The fault characteristic parameters are weighted and corrected according to the electrical physical topology relationship to obtain a fault prediction report.

7. The method according to claim 6, characterized in that The method further comprises: determining a predicted fault level corresponding to the fault prediction report based on a fault-level database; An active control strategy is executed according to the predicted fault level, and the predicted fault level is uploaded.

8. An automobile motor fault prediction device, characterized in that: The device comprises: An acquisition module, configured to acquire at least two historical operating data of a motor of a vehicle, and preprocess each of the historical operating data to obtain historical standard data, wherein each of the historical operating data includes historical motor data and historical environment data; A training module, configured to train the initial fault prediction model according to each of the historical standard data to obtain a trained fault prediction model; The prediction module is used to predict the acquired real-time operation data based on the trained fault prediction model to obtain a real-time fault prediction result.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, wherein the computer-readable storage medium stores instructions, which, when the instructions are executed on a computer or a processor, cause the computer or processor to execute the steps of the method according to any one of claims 1 to 7.

Images