CN115936201A - Motor fault early warning method and device, vehicle and storage medium - Google Patents
Motor fault early warning method and device, vehicle and storage medium Download PDFInfo
- Publication number
- CN115936201A CN115936201A CN202211493004.9A CN202211493004A CN115936201A CN 115936201 A CN115936201 A CN 115936201A CN 202211493004 A CN202211493004 A CN 202211493004A CN 115936201 A CN115936201 A CN 115936201A
- Authority
- CN
- China
- Prior art keywords
- data
- motor
- driving behavior
- vehicle operation
- probability
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Abstract
The application relates to the technical field of motor service life prediction, in particular to a motor fault early warning method, a motor fault early warning device, a vehicle and a storage medium, wherein the method comprises the following steps: acquiring driving behavior data, vehicle operation data and motor delivery data of a current vehicle; inputting driving behavior data, vehicle operation data and motor delivery data into a motor life prediction model trained in advance, and outputting the probability of temperature faults, the probability of voltage faults and the probability of high-level faults; if any probability of the temperature fault, the probability of the voltage fault and the probability of the high-grade fault is larger than a first preset threshold value, a vehicle-mounted machine system of the current vehicle sends out early warning reminding of motor faults to a user. Therefore, the problems that the motor faults are predicted according to the driving behavior data of the user, the vehicle operation data, the motor maintenance data and the like, early warning is carried out before the faults occur and the like are solved, and the service life of the motor is prolonged.
Description
Technical Field
The application relates to the technical field of motor service life prediction, in particular to a motor fault early warning method, a motor fault early warning device, a vehicle and a storage medium.
Background
Energy and environmental problems are becoming more severe, new energy automobiles are receiving much attention, and electric automobiles are becoming important to research by virtue of the advantages of zero emission, low noise, wide power source and the like. Electric vehicles and fuel-powered vehicles have many similarities in vehicle body structure, but significant differences exist in structural form and working environment of the power transmission system. Torsion vibration reduction elements such as a torque converter and a clutch are omitted in the electric automobile transmission system, and the system is represented as an under-damping system; meanwhile, the transmission system adopts a structural form of multi-stage speed reduction and less-gear speed change, the power transmission path is shorter, and the cycle number is greatly increased. These new features presented by electric vehicle transmission systems bring new theoretical and technical problems, wherein transmission system life prediction and system reliability are bottlenecks that limit further performance improvements. With the development of the field of car networking and the field of big data, the capabilities of detecting driving behaviors and car functions and returning data back and entering data in real time are provided. Some clustering-like methods were also used to score these data, but the results were not further processed and utilized.
According to the related art, a fault prediction method (with the publication number of CN 110865628A) of a new energy automobile electric control system based on working condition data is characterized in that a fault prediction model is built by obtaining fault data of the automobile electric control system as a sample, a probability matrix is obtained, the fault with the maximum probability is selected as a prediction result, only the fault data of an automobile is adopted as training model data, and the prediction result is not accurate enough.
Disclosure of Invention
The application provides a motor fault early warning method, a motor fault early warning device, a motor and a storage medium, and solves the problems that the motor fault is predicted according to user driving behavior data, vehicle operation data, motor maintenance data and the like, early warning is carried out before the fault occurs, and the service life of the motor is prolonged.
An embodiment of a first aspect of the present application provides a motor fault early warning method, including the following steps: acquiring driving behavior data, vehicle operation data and motor delivery data of a current vehicle; inputting the driving behavior data, the vehicle operation data and the motor delivery data into a motor service life prediction model trained in advance, and outputting the probability of temperature faults, the probability of voltage faults and the probability of high-level faults; and if any probability of the temperature fault, the probability of the voltage fault and the probability of the high-level fault is greater than a first preset threshold value, sending out early warning prompt of motor fault to a user through a vehicle machine system of the current vehicle.
According to the technical means, the problems that the fault of the motor is predicted according to the driving behavior data of the user, the vehicle operation data, the motor maintenance data and the like, early warning is carried out before the fault occurs and the like are solved, and the service life of the motor is prolonged.
Further, before inputting the driving behavior data, the vehicle operation data, and the motor factory data into the pre-trained motor life prediction model, the method further includes: acquiring driving behavior data, vehicle operation data and motor delivery data of a plurality of target vehicles; normalizing the driving behavior data, the vehicle operation data and the motor delivery data of the target vehicles, performing factor correlation analysis on the driving behavior data, the vehicle operation data and the motor delivery data of the target vehicles after the normalization processing, and screening out data to be trained for motor life prediction according to an analysis result; and training a preset Support Vector Machine (SVM) training model by using the data to be trained to obtain the pre-trained motor life prediction model.
According to the technical means, by carrying out normalization processing and factor correlation analysis on the driving behavior data, the operation data and the motor delivery data of the vehicle, the adverse effect caused by singular data is eliminated, the number of analysis variables is reduced, and the complexity of problems is reduced.
Further, the normalizing the driving behavior data, the vehicle operation data and the motor factory data of the plurality of target vehicles includes: deleting the driving behavior data, the vehicle operation data and the motor factory data of the target vehicles, wherein the variable loss is larger than a second preset threshold, based on a preset filling strategy, carrying out mean value missing value filling on numerical variables of which the variable loss is smaller than or equal to the second preset threshold, and carrying out mode filling on classification variables of which the variable loss is smaller than or equal to the second preset threshold to obtain the driving behavior data, the vehicle operation data and the motor factory data of the target vehicles after being preprocessed; and processing numerical variables in the preprocessed driving behavior data, the preprocessed vehicle operation data and the preprocessed motor delivery data of the plurality of target vehicles based on a preset data normalization algorithm, and processing classification variables in the preprocessed driving behavior data, the preprocessed vehicle operation data and the preprocessed motor delivery data of the plurality of target vehicles based on a preset dummy variable generation algorithm to obtain the normalized driving behavior data, the normalized vehicle operation data and the normalized motor delivery data of the plurality of target vehicles.
According to the technical means, the driving behavior data, the running data and the motor delivery data of the vehicle are subjected to normalization processing, so that adverse effects caused by singular data are eliminated, and the data are limited in a certain range.
Further, the performing factor correlation analysis on the normalized driving behavior data of the target vehicles, the normalized vehicle operation data and the normalized motor delivery data, and screening out data to be trained for motor life prediction according to an analysis result includes: respectively establishing logistic regression with dependent variables one by the independent variables in the driving behavior data, the vehicle running data and the motor delivery data of the plurality of target vehicles after the normalization processing, and calculating a first variable coefficient and a second variable coefficient; and determining a screening variable according to the second variable coefficient and the second variable coefficient, and screening the data to be trained for predicting the service life of the motor from the screening variable according to a preset division ratio.
According to the technical means, logistic regression is used for variable screening, and relevant influences among independent variables are removed.
Further, before the training of the preset SVM training model by using the data to be trained, the method further includes: constructing a label variable of the service life of the motor based on a preset oneVSone strategy; and determining a plurality of label dummy variables based on the values of the label variables so as to perform multi-classification prediction according to the label dummy variables.
According to the technical means, a preset oneVSone strategy is used for classification prediction.
The embodiment of the second aspect of the application provides a motor fault early warning method, which comprises the following steps: the acquisition module is used for acquiring driving behavior data, vehicle operation data and motor delivery data of the current vehicle; the output module is used for inputting the driving behavior data, the vehicle operation data and the motor delivery data into a motor service life prediction model trained in advance and outputting the probability of temperature faults, the probability of voltage faults and the probability of high-level faults; and the early warning module is used for sending out early warning prompt of motor fault to a user through a vehicle-mounted machine system of the current vehicle if any one of the probability of the temperature fault, the probability of the voltage fault and the probability of the high-grade fault is greater than a first preset threshold value.
Further, before inputting the driving behavior data, the vehicle operation data, and the motor factory data into the pre-trained motor life prediction model, the output module is further configured to: acquiring driving behavior data, vehicle operation data and motor delivery data of a plurality of target vehicles; normalizing the driving behavior data, the vehicle operation data and the motor delivery data of the target vehicles, performing factor correlation analysis on the driving behavior data, the vehicle operation data and the motor delivery data of the target vehicles after the normalization processing, and screening out data to be trained for motor life prediction according to an analysis result; and training a preset SVM training model by using the data to be trained to obtain the pre-trained motor life prediction model.
Further, the output module is further configured to perform normalization processing on the driving behavior data, the vehicle operation data, and the motor factory data of the plurality of target vehicles, and is further configured to: deleting the driving behavior data, the vehicle operation data and the motor factory data of the target vehicles, wherein the variable loss is larger than a second preset threshold, based on a preset filling strategy, carrying out mean value missing value filling on numerical variables of which the variable loss is smaller than or equal to the second preset threshold, and carrying out mode filling on classification variables of which the variable loss is smaller than or equal to the second preset threshold to obtain the driving behavior data, the vehicle operation data and the motor factory data of the target vehicles after being preprocessed; and processing numerical variables in the preprocessed driving behavior data, the preprocessed vehicle operation data and the preprocessed motor delivery data of the plurality of target vehicles based on a preset data normalization algorithm, and processing classification variables in the preprocessed driving behavior data, the preprocessed vehicle operation data and the preprocessed motor delivery data of the plurality of target vehicles based on a preset dummy variable generation algorithm to obtain the normalized driving behavior data, the normalized vehicle operation data and the normalized motor delivery data of the plurality of target vehicles.
Further, the output module is further configured to: respectively establishing logistic regression with dependent variables one by one for the independent variables in the normalized driving behavior data, vehicle operation data and motor delivery data of the target vehicles, and calculating a first variable coefficient and a second variable coefficient; and determining a screening variable according to the second variable coefficient and the second variable coefficient, and screening the data to be trained for predicting the service life of the motor from the screening variable according to a preset division ratio.
Further, before the preset SVM training model is trained by using the data to be trained, the output module is further configured to: constructing a label variable of the service life of the motor based on a preset oneVSone strategy; and determining a plurality of label dummy variables based on the values of the label variables so as to perform multi-classification prediction according to the label dummy variables.
An embodiment of a third aspect of the present application provides a vehicle, comprising: the motor fault early warning system comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program to realize the motor fault early warning method according to the embodiment.
A fourth aspect of the present application provides a computer-readable storage medium, on which a computer program is stored, where the computer program is executed by a processor, so as to implement the method for early warning of a motor fault according to the foregoing embodiments.
Therefore, the driving behavior data, the vehicle running data and the motor delivery data of the current vehicle are obtained, the data are input into the motor service life prediction model trained in advance, the probability of the temperature fault, the probability of the voltage fault and the probability of the high-level fault are output, and if any one of the probability of the temperature fault, the probability of the voltage fault and the probability of the high-level fault is larger than a first preset threshold value, the vehicle-mounted system of the current vehicle sends out early warning prompt of the motor fault to a user. Therefore, the problems that the fault of the motor is predicted according to the driving behavior data of the user, the vehicle operation data, the motor maintenance data and the like, early warning is carried out before the fault occurs and the like are solved, and the service life of the motor is prolonged.
Additional aspects and advantages of the present application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the present application.
Drawings
The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
fig. 1 is a flowchart of a method for early warning of a motor fault according to an embodiment of the present disclosure;
FIG. 2 is a flow chart of a method for early warning of motor failure according to one embodiment of the present application;
fig. 3 is a block diagram illustrating an early warning apparatus for motor failure according to an embodiment of the present disclosure;
fig. 4 is a schematic structural diagram of a vehicle according to an embodiment of the present application.
Description of reference numerals: 10-motor fault early warning device, 100-acquisition module, 200-output module, 300-early warning module, 401-memory, 402-processor and 403-communication interface.
Detailed Description
Reference will now be made in detail to the embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the drawings are exemplary and intended to be used for explaining the present application and should not be construed as limiting the present application.
A motor failure warning method, a motor failure warning device, a vehicle, and a storage medium according to embodiments of the present application are described below with reference to the drawings. In the method, the driving behavior data, the vehicle operation data and the motor delivery data of the current vehicle are obtained, the data are input into a motor service life prediction model trained in advance, the probability of temperature faults, the probability of voltage faults and the probability of high-level faults are output, and if any one of the probability of temperature faults, the probability of voltage faults and the probability of high-level faults is greater than a first preset threshold value, the vehicle-mounted machine system of the current vehicle sends out early warning prompt of the motor faults to the user. Therefore, the problems that the fault of the motor is predicted according to the driving behavior data of the user, the vehicle operation data, the motor maintenance data and the like, early warning is carried out before the fault occurs and the like are solved, and the service life of the motor is prolonged.
Specifically, fig. 1 is a schematic flowchart of a motor fault early warning method provided in an embodiment of the present application.
As shown in fig. 1, the motor fault early warning method includes the following steps:
in step S101, driving behavior data, vehicle operation data, and motor factory data of the current vehicle are acquired.
As shown in fig. 2, step S is data acquisition. Collecting driving behavior data of a vehicle includes: vehicle speed, number of rapid accelerations, number of rapid decelerations, number of rapid turns, etc.; the vehicle operation data includes: average vehicle speed, total driving range, torque, rotating speed, total current, total voltage, motor temperature, battery temperature and the like in a unit stroke; the motor factory data comprises: manufacturer, delivery time, etc.
For example, suppose a vehicle has run for t hours during driving, has run for S km:
(1) The driving behavior data of the vehicle includes:
1) Vehicle speed: instantaneous speed at time t;
2) The number of rapid acceleration times: the number of times of rapid acceleration statistics at 0-t moment;
3) The number of rapid deceleration times: counting the number of times of rapid deceleration at 0-t moment;
4) Sharp turn times: and counting the number of sharp turns at 0-t.
(2) The vehicle operation data includes:
1) Average vehicle speed in unit trip: the value of s/t;
2) Total driving range: the value of s;
3) The maximum torque is the peak torque which can be provided by the driving motor in 0-t hours in the running process of the vehicle and is marked as P;
4) Rotating speed: in the running process of the vehicle, the rotating speed of the driving motor is recorded as N within 0-t hours;
5) Total current: in the running process of the vehicle, the working current I of the motor is 0-t hours;
6) Total voltage: and in the running process of the vehicle, the working voltage V of the motor is 0-t hours.
(3) The motor factory data comprises:
1) The manufacturer: some major car manufacturers, denoted M, categorical variables;
2) Delivery time: the time interval from the time when the mileage of the car is 0 to the present is denoted as T.
In step S102, driving behavior data, vehicle operation data, and motor factory data are input to a motor life prediction model trained in advance, and the probability of a temperature fault, the probability of a voltage fault, and the probability of a high-level fault are output.
It should be understood that the input values of the pre-trained motor life prediction model are driving behavior data, vehicle operation data and motor factory data, and the output values are the probability of temperature fault, the probability of voltage fault and the probability of high-level fault, so that the input values can be input into the pre-trained motor life prediction model after the driving behavior data, the vehicle operation data and the motor factory data are obtained, and the probability of temperature fault, the probability of voltage fault and the probability of high-level fault are output.
Wherein, in some embodiments, before inputting the driving behavior data, the vehicle operation data and the motor factory data into the pre-trained motor life prediction model, the method further comprises: acquiring driving behavior data, vehicle operation data and motor delivery data of a plurality of target vehicles; normalizing the driving behavior data, the vehicle operation data and the motor delivery data of the plurality of target vehicles, performing factor correlation analysis on the driving behavior data, the vehicle operation data and the motor delivery data of the plurality of target vehicles after normalization, and screening out data to be trained for motor service life prediction according to an analysis result; and training a preset SVM training model by using the data to be trained to obtain a pre-trained motor life prediction model.
In some embodiments, the normalizing process is performed on the driving behavior data, the vehicle operation data and the motor factory data of a plurality of target vehicles, and includes: deleting the driving behavior data, the vehicle operation data and the motor factory data of the plurality of target vehicles, wherein the variable loss is greater than a second preset threshold, performing mean value missing value filling on numerical variables of which the variable loss is less than or equal to the second preset threshold based on a preset filling strategy, and performing mode filling on classification variables of which the variable loss is less than or equal to the second preset threshold to obtain the preprocessed driving behavior data, the vehicle operation data and the motor factory data of the plurality of target vehicles; and processing numerical variables in the preprocessed driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles based on a preset data normalization algorithm, and processing classification variables in the preprocessed driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles based on a preset dummy variable generation algorithm to obtain the normalized driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles.
It should be understood that step S2 is data preprocessing, as shown in fig. 2. Deleting variables of which the variable missing number is greater than 30% (namely a second preset threshold) in the driving behavior data, the vehicle operation data and the motor delivery data of the plurality of target vehicles, meanwhile, based on a preset filling strategy, performing mean missing value filling on numerical variables of which the variable missing is less than or equal to 30%, and performing mode filling on classified variables of which the variable missing is less than or equal to 30%, so as to obtain the preprocessed driving behavior data, the vehicle operation data and the motor delivery data of the plurality of target vehicles.
Further, as shown in fig. 2, step S3 is data normalization. And processing numerical variables in the preprocessed driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles based on a preset data normalization algorithm.
The variables of the driving behavior and the vehicle operation data are numerical variables, and the data are preprocessed by using a preset data normalization algorithm, wherein the preset data normalization algorithm is as follows:
assume that the numerical variables before preprocessing are: x = [ X ] 1 ,x 2 ...x t ]Recording the variables after normalization pretreatment as follows: y = [ Y = 1 ,y 2 ...y t ]，
The correlation is:
and further, processing classification variables in the preprocessed driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles based on a preset dummy variable generation algorithm.
The motor related data are classified variables, data preprocessing is performed by using a dummy variable generation algorithm, and the preset dummy variable generation algorithm is as follows:
assume that there are t classes for the classification variable X before preprocessing, and the value of each class is denoted as X (t). T-1 dummy variables can be generated, and the conversion relationship between the new variables and the original variables is:
the driving behavior data, the vehicle operation data and the motor delivery data of the plurality of target vehicles after normalization processing are obtained by normalizing the driving behavior data, the vehicle operation data and the motor delivery data of the plurality of target vehicles.
Further, in some embodiments, performing factor correlation analysis on the normalized driving behavior data of the plurality of target vehicles, the vehicle operation data, and the motor factory data, and screening out data to be trained for motor life prediction according to an analysis result, includes: respectively establishing logistic regression with dependent variables one by independent variables in the driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles after normalization processing, and calculating a first variable coefficient and a second variable coefficient; and determining a screening variable according to the second variable coefficient and the second variable coefficient, and screening the data to be trained for predicting the service life of the motor from the screening variable according to a preset division ratio.
Specifically, as shown in fig. 2, step S4 is a factor correlation analysis. Considering that the dependent variable is a classification variable and the related influence between independent variables needs to be removed, establishing logistic regression respectively between the independent variables and the dependent variable one by one, and calculating a first variable coefficient Beta and a second variable coefficient R 2 (since there is only one argument, R 2 The confidence of the variable can be directly measured).
The logistic regression formula is as follows:
wherein, theta = [ alpha, beta ], alpha is intercept, beta is variable coefficient, y is label variable, and theta is vector combination of the intercept and the variable coefficient.
According to a logistic regression formula, a classification expression of the target variable can be obtained:
P(y＝1|x:θ)＝g(θ T *[1,x])
P(y＝0|x:θ)＝1-g(θ T *[1,x])
and integrating the classification expressions of the target variables to obtain a log-likelihood function:
the minimum value of the first variable coefficient Beta can be obtained.
After the value of the first coefficient of variation Beta is found, the second coefficient of variation R 2 The calculation formula of (c) is as follows:
wherein, y t In order to be a real label, the label,
for a predictive tag, <' >>
Is the average value of the labels. />
Calculating a first variable coefficient Beta and a second variable coefficient R 2 And then, setting a threshold value according to the actual situation, and screening the variables. The first variable coefficient Beta is the influence of each variable on the service life of the motor. Second coefficient of variation R 2 Namely the reliability of whether the influence of each variable on the service life of the motor is established or not.
Finally, as shown in FIG. 2, step S5 is to divide the training and testing set. And screening data to be trained for motor life prediction from the screening variables according to a preset division ratio of 7:3, wherein the data to be trained comprises a division training set and a test set.
Optionally, in some embodiments, before training the preset SVM training model using the data to be trained, the method further includes: constructing a label variable of the service life of the motor based on a preset oneVSone strategy; and determining a plurality of label dummy variables based on the values of the label variables so as to perform multi-classification prediction according to the label dummy variables.
Specifically, as shown in fig. 2, step S6 is training the model. The model is trained using SVM. The SVM is a supervised classification algorithm, and the idea is as follows: assuming that there are two types of points in the sample space, a partition hyperplane is found to separate the two types of samples, and the partition hyperplane should be selected to have the best generalization capability, i.e., to maximize the distance between the closest sample points of the two types of samples.
Before the life of the motor is predicted by using the SVM, a preset oneVSone strategy is used for constructing a motor life label variable y, then the multi-classification SVM is converted into a second classification, and the label variable y is classified as follows:
according to the value of Y, 11 label dummy variables are set as follows:
then, using SVM to carry out multi-classification prediction on the 11 conditions, wherein the algorithm basis of the multi-classification prediction is two classifications, and the algorithm principle of the two classification SVM is as follows:
assuming a training data set on a given feature space
T＝{(x 1 ,y 1 ),(x 2 ,y 2 )L(x N ,y N )}；
Wherein x is i ∈R n ，y i ∈{+1,-1},i＝1,2...，N，x i Is the ith feature vector, y i Is a class flag, positive when it equals + 1; the negative example is given when the value is-1.
Assuming that the training data set is linearly separable, the geometric interval is calculated, for a given data set T and hyperplane w x + b =0, defining a hyperplane about the sample point (x + b) N ,y N ) The geometrical interval of (A) is:
the minimum of the geometrical spacing of the hyperplane with respect to all sample points is: γ = min γ i In practice, this distance is the support vector to hyperplane distance.
According to the above definitions, the solving of the maximum segmentation hyperplane problem for the SVM model can be represented as the following constrained optimization problem:
dividing both sides of the constraint by gamma to obtain:
since | | w |, γ are both scalars, for the sake of expression simplicity, let:
obtaining:
y i (w*x i +b)≥1,i＝1,2,3K,N
and because of maximizing gamma, it is equivalent to maximizing
Is also equivalent to minimize->
(/>
For the following derivation, the form is simple, and the result is not affected), so the solution of the maximum segmentation hyperplane problem of the SVM model can be expressed as the following constrained optimization problem:
s.t.y i (w*x i +b)≥1,i＝1,2,3K,N
this is a convex quadratic programming problem with inequality constraints, for which the dual problem (dual problem) can be obtained using the lagrange multiplier method.
First, the constrained original objective function is converted to an unconstrained newly constructed lagrangian objective function:
wherein alpha is i Is a Lagrange multiplier, and alpha i ≥0。
Order to
When the sample point does not satisfy the constraint, i.e. outside the feasible solution area:
y i (w*x i +b)<1
at this time, α is adjusted i Set to infinity, θ (w) is also infinity.
When the full-cost point meets the constraint condition, namely in the feasible solution area:
y i (w*x i +b)≥1
in this case, θ (w) is the primitive function itself. Thus, combining the two cases results in a new objective function:
thus, the original constraint problem is equivalent to:
for a new objective function, firstly solving the maximum value, then solving the minimum value, and exchanging the positions of the minimum value and the maximum value by using the duality of the Lagrange function to obtain:
to have p * ＝d * Two conditions need to be met:
(1) the optimization problem is a convex optimization problem
(2) Satisfy the KKT condition
First, the optimization problem is obviously a convex optimization problem, so the condition (2) is satisfied when the condition is satisfied
To obtain a specific form of solving the dual problem, let L (w, b, α) have a partial derivative of w and b of 0, one can obtain:
substituting the two equations into a Lagrange objective function, and eliminating w and b to obtain:
namely, it is
To find
Maximum for alpha, i.e. a dual problem->
α i ≥0i＝1,2,K,N
Adding a negative sign to the target formula, and converting the solving maximum into solving minimum to obtain:
α i ≥0i＝1,2,K,N
the existing optimization problem is changed into the form, and on the basis of two classifications, a softmax principle is adopted to further calculate to obtain a multi-classification result.
Further, as shown in fig. 2, step S7 is a test model. After the model training is finished, whether the model accuracy rate reaches 70% is judged, vehicle real-time data are input into the model, the fault of the motor is predicted through the model, and the probability of temperature fault, the probability of voltage fault and the probability of high-level fault are calculated.
Wherein, temperature failure: the motor over-temperature is recorded as 1, the motor temperature is normally recorded as 0, and the variable name is C; voltage failure: the motor overvoltage is recorded as 1, the motor undervoltage is recorded as 2, the motor temperature is normally recorded as 0, and the variable name is V; high-level failure: the motor part damage is marked as 1, the part normal or wear is marked as 0 in a controllable range, and the variable is marked as K.
In step S103, if any one of the probability of the temperature fault, the probability of the voltage fault, and the probability of the high-level fault is greater than a first preset threshold, an early warning prompt of the motor fault is sent to the user through the vehicle-mounted system of the current vehicle. Wherein, the first preset threshold may be 0.5.
It can be understood that if the probability of any motor failure is greater than 0.5, which indicates that the motor is about to fail, the voice prompt is performed on the user through the vehicle-mounted machine system of the vehicle. For example, if the motor is over-temperature and the probability of the temperature fault is greater than 0.5, it is indicated that the motor is about to have the temperature fault, the vehicle machine system of the vehicle sends out voice to remind a user that the motor is over-temperature and the temperature fault is about to occur, if the motor is over-temperature and the motor is over-voltage, the probability of the temperature fault is greater than 0.5 and the probability of the voltage fault is greater than 0.5, it is indicated that the motor is about to have the temperature fault and the voltage fault, and at this time, the vehicle machine system of the vehicle sends out voice to remind the user that the motor is about to have the fault.
According to the motor fault early warning method provided by the embodiment of the application, the driving behavior data, the vehicle running data and the motor delivery data of the current vehicle are obtained, the data are input into the motor service life prediction model trained in advance, the probability of the temperature fault, the probability of the voltage fault and the probability of the high-level fault are output, and if any one of the probability of the temperature fault, the probability of the voltage fault and the probability of the high-level fault is larger than a first preset threshold value, the motor fault early warning prompt is sent to a user through a vehicle-mounted machine system of the current vehicle. Therefore, the problems that the motor faults are predicted according to the driving behavior data of the user, the vehicle operation data and the motor maintenance data, early warning is carried out before the faults occur and the like are solved, and the service life of the motor is prolonged.
Next, a motor failure warning device according to an embodiment of the present application will be described with reference to the drawings.
Fig. 3 is a block diagram schematically illustrating an early warning apparatus for motor failure according to an embodiment of the present disclosure.
As shown in fig. 3, the motor failure warning apparatus 10 includes: the system comprises an acquisition module 100, an output module 200 and an early warning module 300.
The acquiring module 100 is configured to acquire driving behavior data of a current vehicle, vehicle operation data, and motor delivery data; the output module 200 is used for inputting driving behavior data, vehicle operation data and motor delivery data into a motor life prediction model trained in advance, and outputting the probability of temperature faults, the probability of voltage faults and the probability of high-level faults; the early warning module 300 is configured to send an early warning prompt of a motor fault to a user through a vehicle-mounted device system of a current vehicle if any one of the probability of the temperature fault, the probability of the voltage fault, and the probability of the high-level fault is greater than a first preset threshold.
Further, in some embodiments, before inputting the driving behavior data, the vehicle operation data, and the motor factory data into the pre-trained motor life prediction model, the output module 200 is further configured to: acquiring driving behavior data, vehicle operation data and motor delivery data of a plurality of target vehicles; normalizing the driving behavior data, the vehicle operation data and the motor delivery data of the plurality of target vehicles, performing factor correlation analysis on the driving behavior data, the vehicle operation data and the motor delivery data of the plurality of target vehicles after normalization, and screening out data to be trained for motor service life prediction according to an analysis result; and training a preset SVM training model by using the data to be trained to obtain a pre-trained motor life prediction model.
Further, in some embodiments, the output module 200 is further configured to perform normalization processing on the driving behavior data, the vehicle operation data, and the motor factory data of the multiple target vehicles: deleting the driving behavior data, the vehicle operation data and the motor delivery data of the target vehicles, wherein the variable loss is greater than a second preset threshold, based on a preset filling strategy, performing mean value missing value filling on numerical variables of which the variable loss is less than or equal to the second preset threshold, and performing mode filling on classified variables of which the variable loss is less than or equal to the second preset threshold to obtain the driving behavior data, the vehicle operation data and the motor delivery data of the target vehicles after pretreatment; and processing numerical variables in the preprocessed driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles based on a preset data normalization algorithm, and processing classification variables in the preprocessed driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles based on a preset dummy variable generation algorithm to obtain the normalized driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles.
Further, in some embodiments, the output module 200 is further configured to: respectively establishing logistic regression with the independent variables in the driving behavior data, the vehicle running data and the motor factory data of the plurality of target vehicles after normalization processing, and calculating a first variable coefficient and a second variable coefficient; and determining a screening variable according to the second variable coefficient and the second variable coefficient, and screening the data to be trained for predicting the service life of the motor from the screening variable according to a preset division ratio.
Further, in some embodiments, before training the predetermined SVM training model using the data to be trained, the output module 200 is further configured to: constructing a label variable of the service life of the motor based on a preset oneVSone strategy; and determining a plurality of label dummy variables based on the values of the label variables so as to perform multi-classification prediction according to the label dummy variables.
It should be noted that the explanation of the embodiment of the motor fault warning method is also applicable to the motor fault warning device of the embodiment, and details are not repeated here.
According to the motor fault early warning device provided by the embodiment of the application, the driving behavior data, the vehicle running data and the motor delivery data of the current vehicle are obtained, the data are input into the motor service life prediction model trained in advance, the probability of temperature fault, the probability of voltage fault and the probability of high-level fault are output, and if any probability of the probability of temperature fault, the probability of voltage fault and the probability of high-level fault is larger than a first preset threshold value, the motor fault early warning prompt is sent to a user through a vehicle machine system of the current vehicle. Therefore, the problems that the motor faults are predicted according to the driving behavior data of the user, the vehicle operation data, the motor maintenance data and the like, early warning is carried out before the faults occur and the like are solved, and the service life of the motor is prolonged.
Fig. 4 is a schematic structural diagram of a vehicle according to an embodiment of the present application. The vehicle may include:
The processor 402 executes the program to implement the motor failure warning method provided in the above-described embodiment.
Further, the vehicle further includes:
a communication interface 403 for communication between the memory 401 and the processor 402.
A memory 401 for storing computer programs operable on the processor 402.
The Memory 401 may include a high-speed RAM (Random Access Memory) Memory, and may also include a non-volatile Memory, such as at least one disk Memory.
If the memory 401, the processor 402 and the communication interface 403 are implemented independently, the communication interface 403, the memory 401 and the processor 402 may be connected to each other through a bus and perform communication with each other. The bus may be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown in FIG. 4, but that does not indicate only one bus or one type of bus.
Optionally, in a specific implementation, if the memory 401, the processor 402, and the communication interface 403 are integrated on a chip, the memory 401, the processor 402, and the communication interface 403 may complete mutual communication through an internal interface.
An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method for early warning of a motor fault as above is implemented.
In the description of the present specification, reference to the description of "one embodiment," "some embodiments," "an example," "a specific example," or "some examples" or the like means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or N embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or to implicitly indicate the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present application, "N" means at least two, e.g., two, three, etc., unless explicitly defined otherwise.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more N executable instructions for implementing steps of a custom logic function or process, and alternate implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present application.
It should be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a programmable gate array, a field programmable gate array, or the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.
While embodiments of the present application have been shown and described above, it will be understood that the above embodiments are exemplary and should not be construed as limiting the present application and that changes, modifications, substitutions and alterations in the above embodiments may be made by those of ordinary skill in the art within the scope of the present application.
Claims (10)
1. The motor fault early warning method is characterized by comprising the following steps of:
acquiring driving behavior data, vehicle operation data and motor delivery data of a current vehicle;
inputting the driving behavior data, the vehicle operation data and the motor delivery data into a motor service life prediction model trained in advance, and outputting the probability of temperature faults, the probability of voltage faults and the probability of high-level faults; and
and if any probability of the temperature fault, the probability of the voltage fault and the probability of the high-level fault is greater than a first preset threshold value, sending out early warning prompt of motor fault to a user through a vehicle machine system of the current vehicle.
2. The method of claim 1, further comprising, prior to inputting the driving behavior data, the vehicle operation data, and the motor factory data to the pre-trained motor life prediction model:
acquiring driving behavior data, vehicle operation data and motor delivery data of a plurality of target vehicles;
normalizing the driving behavior data, the vehicle operation data and the motor delivery data of the plurality of target vehicles, performing factor correlation analysis on the driving behavior data, the vehicle operation data and the motor delivery data of the plurality of target vehicles after normalization, and screening out data to be trained for motor service life prediction according to an analysis result;
and training a preset SVM training model by using the data to be trained to obtain the pre-trained motor life prediction model.
3. The method of claim 2, wherein normalizing the driving behavior data, the vehicle operation data, and the motor delivery data of the plurality of target vehicles comprises:
deleting the driving behavior data, the vehicle operation data and the motor factory data of the target vehicles, wherein the variable loss is greater than a second preset threshold, performing mean value missing value filling on numerical variables of which the variable loss is less than or equal to the second preset threshold based on a preset filling strategy, and performing mode filling on classification variables of which the variable loss is less than or equal to the second preset threshold to obtain the preprocessed driving behavior data, the preprocessed vehicle operation data and the preprocessed motor factory data of the target vehicles;
and processing numerical variables in the preprocessed driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles based on a preset data normalization algorithm, and processing classification variables in the preprocessed driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles based on a preset dummy variable generation algorithm to obtain the driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles after normalization processing.
4. The method according to claim 2 or 3, wherein the performing factor correlation analysis on the normalized driving behavior data, vehicle operation data and motor factory data of the plurality of target vehicles, and screening out data to be trained for motor life prediction according to an analysis result comprises:
respectively establishing logistic regression with dependent variables one by one for the independent variables in the normalized driving behavior data, vehicle operation data and motor delivery data of the target vehicles, and calculating a first variable coefficient and a second variable coefficient;
and determining screening variables according to the second variable coefficients and the second variable coefficients, and screening the data to be trained for predicting the service life of the motor from the screening variables according to a preset division ratio.
5. The method according to claim 2, wherein before training the preset SVM training model by using the data to be trained, the method further comprises:
constructing a label variable of the service life of the motor based on a preset oneVSone strategy;
and determining a plurality of label dummy variables based on the values of the label variables so as to perform multi-classification prediction according to the label dummy variables.
6. A motor failure early warning device, comprising:
the acquisition module is used for acquiring driving behavior data, vehicle operation data and motor delivery data of the current vehicle;
the output module is used for inputting the driving behavior data, the vehicle operation data and the motor delivery data into a motor service life prediction model trained in advance and outputting the probability of temperature faults, the probability of voltage faults and the probability of high-level faults; and
and the early warning module is used for sending out early warning prompt of motor fault to a user through a vehicle-mounted machine system of the current vehicle if any one of the probability of the temperature fault, the probability of the voltage fault and the probability of the high-grade fault is greater than a first preset threshold value.
7. The apparatus of claim 6, wherein the output module, prior to inputting the driving behavior data, the vehicle operation data, and the motor factory data into the pre-trained motor life prediction model, is further to:
acquiring driving behavior data, vehicle operation data and motor delivery data of a plurality of target vehicles;
normalizing the driving behavior data, the vehicle operation data and the motor delivery data of the plurality of target vehicles, performing factor correlation analysis on the driving behavior data, the vehicle operation data and the motor delivery data of the plurality of target vehicles after normalization, and screening out data to be trained for motor service life prediction according to an analysis result;
and training a preset SVM training model by using the data to be trained to obtain the pre-trained motor life prediction model.
8. The apparatus of claim 7, wherein the output module normalizes the driving behavior data, the vehicle operation data, and the motor delivery data of the target vehicles, and further:
deleting the driving behavior data, the vehicle operation data and the motor factory data of the target vehicles, wherein the variable loss is larger than a second preset threshold, based on a preset filling strategy, carrying out mean value missing value filling on numerical variables of which the variable loss is smaller than or equal to the second preset threshold, and carrying out mode filling on classification variables of which the variable loss is smaller than or equal to the second preset threshold to obtain the driving behavior data, the vehicle operation data and the motor factory data of the target vehicles after being preprocessed;
and processing numerical variables in the preprocessed driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles based on a preset data normalization algorithm, and processing classification variables in the preprocessed driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles based on a preset dummy variable generation algorithm to obtain the driving behavior data, vehicle operation data and motor delivery data of the plurality of target vehicles after normalization processing.
9. 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 warning of a motor fault according to any one of claims 1 to 5.
10. A computer-readable storage medium, on which a computer program is stored, which program is executed by a processor for implementing the method for early warning of a motor failure as claimed in any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211493004.9A CN115936201A (en) | 2022-11-25 | 2022-11-25 | Motor fault early warning method and device, vehicle and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211493004.9A CN115936201A (en) | 2022-11-25 | 2022-11-25 | Motor fault early warning method and device, vehicle and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115936201A true CN115936201A (en) | 2023-04-07 |
Family
ID=86555039
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211493004.9A Pending CN115936201A (en) | 2022-11-25 | 2022-11-25 | Motor fault early warning method and device, vehicle and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115936201A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117114352A (en) * | 2023-09-15 | 2023-11-24 | 北京阿帕科蓝科技有限公司 | Vehicle maintenance method, device, computer equipment and storage medium |
-
2022
- 2022-11-25 CN CN202211493004.9A patent/CN115936201A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117114352A (en) * | 2023-09-15 | 2023-11-24 | 北京阿帕科蓝科技有限公司 | Vehicle maintenance method, device, computer equipment and storage medium |
| CN117114352B (en) * | 2023-09-15 | 2024-04-09 | 北京阿帕科蓝科技有限公司 | Vehicle maintenance method, device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110091751B (en) | Electric automobile endurance mileage prediction method, device and medium based on deep learning | |
| CN108303264B (en) | Cloud-based vehicle fault diagnosis method, device and system | |
| US10657737B2 (en) | Vehicle error identification system | |
| CN112883648B (en) | Training method and device for automobile fuel consumption prediction model and computer equipment | |
| KR102414819B1 (en) | Method for financial service based on battery remaining value and apparatus for performing the method | |
| US7640148B2 (en) | Method of modeling vehicle parameter cycles | |
| CN114585983A (en) | Method, device and system for detecting abnormal operation state of equipment | |
| CN115936201A (en) | Motor fault early warning method and device, vehicle and storage medium | |
| CN112345956A (en) | Battery pack charge state detection method and device | |
| CN109720207B (en) | Vehicle energy consumption analysis method and device and computer readable medium | |
| US20230305065A1 (en) | Apparatus and method for diagnosing state of battery | |
| CN111007410A (en) | Vehicle battery diagnostic method and apparatus | |
| CN111191824A (en) | Power battery capacity attenuation prediction method and system | |
| Pau et al. | Microcontroller architectures for battery state of charge prediction with tiny neural networks | |
| CN116945856A (en) | Pure electric vehicle integrated thermal management system and automobile based on edge calculation | |
| US20230152387A1 (en) | Method for prediction of battery life and apparatus for performing the method | |
| CN116008815A (en) | Method, device, equipment, storage medium and vehicle for detecting short circuit in battery cell | |
| CN113538735B (en) | Vehicle use identification method and device | |
| CN115837862A (en) | Abnormal data detection method and device, electronic equipment and storage medium | |
| JP2021124420A (en) | Battery deterioration determination device, battery deterioration determination method, and battery deterioration determination program | |
| US20230278567A1 (en) | Autonomous driving control apparatus and method thereof | |
| CN116224091B (en) | Battery cell fault detection method and device, electronic equipment and storage medium | |
| US20230358814A1 (en) | Method for monitoring e-mobility battery value and apparatus for performing the method | |
| US20230368591A1 (en) | Method for monitoring failure of motor in a car based on clustering algorithm and system using the same | |
| US20240001939A1 (en) | Method for monitoring failure of motor in a car based on scaling algorithm and system using the same |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |