1. A method of calculating a rotor temperature, comprising:

obtaining vehicle data of an ith time period, wherein the vehicle data of the ith time period comprises the data of the ith time period₁Time to ith_nThe method comprises the steps that vehicle data at a moment comprise electrical parameters, physical parameters and environmental parameters, the electrical parameters are data related to electrical performance of a motor equipped on a target vehicle, the physical parameters are data related to working conditions of the motor, the environmental parameters are data related to the environment where the motor is located, i and n are natural numbers, i is more than or equal to 1, and n is more than or equal to 2;

carrying out feature construction on the vehicle data of the ith time period to obtain target feature variables, wherein the target feature variables comprise the feature variables of the vehicle data of the ith time period and the feature variables of the vehicle data before the ith time period;

calling rotor temperature prediction model pair i₁The rotor temperature at the moment and the target characteristic variable are processed to obtain the ith_nRotor temperature at time of day, the rotor temperature predictionThe model comprises a machine learning model employing a deep learning algorithm, the rotor temperature being a temperature of a rotor of the electric motor;

and repeating the steps until i is equal to N, wherein N is the number of the time periods, N is a natural number, and N is more than or equal to 2.

2. The method according to claim 1, wherein the performing feature construction on the vehicle data of the ith time period to obtain a target feature variable comprises:

carrying out feature extraction on the vehicle data of the ith time period to obtain a feature variable of the ith time period;

adding an auxiliary characteristic variable in the characteristic variable of the ith time period to obtain an expanded characteristic variable, wherein the auxiliary characteristic variable comprises the characteristic variable of the vehicle data before the ith time period;

and carrying out dimension scaling on the expanded characteristic variables to obtain the target characteristic variables.

3. The method of claim 2, wherein the auxiliary variables comprise electrical, physical and environmental parameters in a predetermined number of time periods prior to the i-th time period.

4. The method of claim 3, wherein the auxiliary variables further comprise averages of the electrical, physical and environmental parameters over a predetermined number of time periods prior to the i-th time period.

5. The method of claim 4, wherein the auxiliary variables further comprise variances of the electrical, physical and environmental parameters in a predetermined number of time periods prior to the i-th time period.

6. A method according to any of claims 1 to 5, wherein the electrical parameter comprises current data of the rotor and/or voltage data of the rotor.

7. The method of claim 6, wherein the electrical parameters further comprise bus current data of the motor and/or bus voltage data of the motor.

8. Method according to any of claims 1 to 5, wherein said physical parameter comprises the rotational speed of said electric motor and/or the torque of said electric motor.

9. The method of any one of claims 1 to 5, wherein the environmental parameter comprises a temperature of an environment in which the motor is located.

10. The method according to any one of claims 1 to 5, wherein the environmental parameter further comprises a temperature of cooling water of the motor.

11. A computing device, comprising:

an acquisition module for acquiring vehicle data of an ith time period, wherein the vehicle data of the ith time period comprises data of the ith time period₁Time to ith_nThe method comprises the steps that vehicle data at a moment comprise electrical parameters, physical parameters and environmental parameters, the electrical parameters are data related to electrical performance of a motor equipped on a target vehicle, the physical parameters are data related to working conditions of the motor, the environmental parameters are data related to the environment where the motor is located, i and n are natural numbers, i is more than or equal to 1, and n is more than or equal to 2;

the processing module is used for carrying out feature construction on the vehicle data in the ith time period to obtain target feature variables, wherein the target feature variables comprise the feature variables of the vehicle data in the ith time period and the feature variables of the vehicle data before the ith time period; calling rotor temperature prediction model pair i₁The rotor temperature at the moment and the target characteristic variable are processed to obtain the ith_nA rotor temperature at a time, the rotor temperature prediction model including using deep learningA machine learning model of an algorithm, the rotor temperature being a temperature of a rotor of the electric motor; and repeating the steps until i is equal to N, wherein N is the number of the time periods, N is a natural number, and N is more than or equal to 2.

12. A computer device, characterized in that it comprises a processor and a memory, in which at least one instruction or program is stored, which is loaded and executed by the processor to implement the method of calculating the rotor temperature according to any one of claims 1 to 9.

13. A computer-readable storage medium having stored thereon at least one instruction, which is loaded and executed by a processor, to implement the method of calculating a rotor temperature according to any one of claims 1 to 9.

Background

The motor is widely used in the field of driving of vehicles due to its high reliability. An electric motor generally comprises a stator, a rotor (which is usually a permanent magnet), an end cover and other components, and for a vehicle equipped with the electric motor (for example, a permanent magnet synchronous motor), when the temperature of the rotor is too high, irreversible demagnetization can occur, so that the monitoring of the temperature of the rotor is an important index in vehicle control.

In the related art, the temperature of the rotor may be acquired through wireless telemetry by equipping a vehicle with a motor having wireless telemetry. However, the motor with wireless telemetry is expensive, difficult to install in a vehicle, and has a narrow applicability.

Disclosure of Invention

The application provides a method, a device, equipment and a storage medium for calculating the rotor temperature, which can solve the problems of higher cost and narrower applicability of the method for acquiring the rotor temperature through a wireless telemetry technology in the related technology.

In one aspect, an embodiment of the present application provides a method for calculating a rotor temperature, including:

obtaining vehicle data of an ith time period, wherein the vehicle data of the ith time period comprises the data of the ith time period₁Time to ith_nThe method comprises the steps that vehicle data at a moment comprise electrical parameters, physical parameters and environmental parameters, the electrical parameters are data related to electrical performance of a motor equipped on a target vehicle, the physical parameters are data related to working conditions of the motor, the environmental parameters are data related to the environment where the motor is located, i and n are natural numbers, i is more than or equal to 1, and n is more than or equal to 2;

carrying out feature construction on the vehicle data of the ith time period to obtain target feature variables, wherein the target feature variables comprise the feature variables of the vehicle data of the ith time period and the feature variables of the vehicle data before the ith time period;

calling rotor temperature prediction model pair i₁The rotor temperature at the moment and the target characteristic variable are processed to obtain the ith_nRotor of timeA temperature, the rotor temperature prediction model comprising a machine learning model employing a deep learning algorithm, the rotor temperature being a temperature of a rotor of the electric motor;

and repeating the steps until i is equal to N, wherein N is the number of the time periods, N is a natural number, and N is more than or equal to 2.

Optionally, the performing feature construction on the vehicle data in the ith time period to obtain a target feature variable includes:

carrying out feature extraction on the vehicle data of the ith time period to obtain a feature variable of the ith time period;

adding auxiliary characteristic variables in the characteristic variables of the ith time period to obtain original target characteristic variables, wherein the auxiliary characteristic variables comprise the characteristic variables of the vehicle data before the ith time period;

and carrying out dimension scaling on the original target characteristic variable to obtain the target characteristic variable.

Optionally, the auxiliary variables include electrical parameters, physical parameters and environmental parameters in a predetermined number of time periods before the ith time period.

Optionally, the auxiliary variable further includes an average value of the electrical parameter, the physical parameter, and the environmental parameter in a predetermined number of time periods before the ith time period.

Optionally, the auxiliary variable further includes a variance of the electrical parameter, the physical parameter, and the environmental parameter in a predetermined number of time periods before the ith time period.

Optionally, the electrical parameter comprises current data of the rotor and/or voltage data of the rotor.

Optionally, the electrical parameter further includes bus current data of the motor and/or bus voltage data of the motor.

Optionally, the physical parameter comprises a rotational speed of the electric motor and/or a torque of the electric motor.

Optionally, the environmental parameter includes a temperature of an environment in which the motor is located.

Optionally, the environmental parameter further includes a temperature of cooling water of the motor.

In another aspect, an embodiment of the present application provides a computing apparatus, including:

an acquisition module for acquiring vehicle data of an ith time period, wherein the vehicle data of the ith time period comprises data of the ith time period₁Time to ith_nThe method comprises the steps that vehicle data at a moment comprise electrical parameters, physical parameters and environmental parameters, the electrical parameters are data related to electrical performance of a motor equipped on a target vehicle, the physical parameters are data related to working conditions of the motor, the environmental parameters are data related to the environment where the motor is located, i and n are natural numbers, i is more than or equal to 1, and n is more than or equal to 2;

the processing module is used for carrying out feature construction on the vehicle data in the ith time period to obtain target feature variables, wherein the target feature variables comprise the feature variables of the vehicle data in the ith time period and the feature variables of the vehicle data before the ith time period; calling rotor temperature prediction model pair i₁The rotor temperature at the moment and the target characteristic variable are processed to obtain the ith_nA rotor temperature at a time, the rotor temperature prediction model comprising a machine learning model employing a deep learning algorithm, the rotor temperature being a temperature of a rotor of the electric motor; and repeating the steps until i is equal to N, wherein N is the number of the time periods, N is a natural number, and N is more than or equal to 2.

In another aspect, the present application provides a computer device, where the device includes a processor and a memory, where the memory stores at least one instruction or program, and the instruction or program is loaded and executed by the processor to implement the method for calculating the rotor temperature as described in any one of the above.

In another aspect, an embodiment of the present application provides a computer-readable storage medium, where at least one instruction is stored in the storage medium, and the instruction is loaded and executed by a processor to implement the method for calculating the rotor temperature as described in any one of the above.

The technical scheme at least comprises the following advantages:

the rotor temperature is obtained by processing the vehicle data by calling the machine learning model containing the deep learning algorithm, the problems of high cost and narrow applicability of measuring the rotor temperature through wireless remote measurement are solved, and meanwhile, the predicted rotor temperature has high accuracy and accuracy because the vehicle data contains electrical parameters, physical parameters and environmental parameters highly related to the rotor temperature.

Drawings

In order to more clearly illustrate the detailed description of the present application or the technical solutions in the prior art, the drawings needed to be used in the detailed description of the present application or the prior art description will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative efforts.

FIG. 1 is a schematic illustration of an environment in which a method for monitoring rotor temperature is implemented according to an exemplary embodiment of the present application;

FIG. 2 is a flow chart of a method of calculating a rotor temperature provided by an exemplary embodiment of the present application;

FIG. 3 is a flow chart of a feature construction method provided by an exemplary embodiment of the present application;

FIG. 4 is a flow chart of a method of model optimization provided by an exemplary embodiment of the present application;

FIG. 5 is a schematic diagram of a training architecture for a rotor temperature prediction model provided by an exemplary embodiment of the present application;

FIG. 6 is a block diagram of a computing device provided in an exemplary embodiment of the present application;

FIG. 7 is a block diagram of a computer device provided in an exemplary embodiment of the present application.

Detailed Description

The technical solutions in the present application will be described clearly and completely with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

In the description of the present application, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of description and simplicity of description, and do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be construed as limiting the present application. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.

In the description of the present application, it is to be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; the connection can be mechanical connection or electrical connection; the two elements may be directly connected or indirectly connected through an intermediate medium, or may be communicated with each other inside the two elements, or may be wirelessly connected or wired connected. The specific meaning of the above terms in the present application can be understood in a specific case by those of ordinary skill in the art.

In addition, the technical features mentioned in the different embodiments of the present application described below may be combined with each other as long as they do not conflict with each other.

Referring to fig. 1, which shows a schematic diagram of an application environment of a monitoring method of a rotor temperature provided in an exemplary embodiment of the present application, as shown in fig. 1, it includes a target vehicle 110 and a server 120, the target vehicle 110 is equipped with an electric motor (not shown in fig. 1), a control device 111 (the control device 111 may be an Electronic Control Unit (ECU)), and a communication device 112. Wherein:

the control device 111 and the communication device 112 can establish communication connection through a bus, a hard wire or other wired or wireless communication modes; the control device 111 may establish a communication connection with the server 120 via the communication device 112.

A wireless communication connection may be established between the communication device 112 and the server 120 through a mobile communication network (e.g., a third generation mobile network (3G) technology, a Long Term Evolution (LTE) technology, or a fifth generation mobile network (5G) technology).

In an alternative embodiment of the present application, a rotor temperature prediction model is deployed in the server 120, which includes a machine learning model employing a deep learning algorithm for predicting the rotor temperature of the motor of the target vehicle.

A control device 111 for acquiring vehicle data at predetermined time intervals, and transmitting the vehicle data to the server 120 through the communication device 112 after each acquisition of the vehicle data.

For example, the predetermined time interval is Δ T, and at an initial time, vehicle data is acquired at a first time, and after Δ T, vehicle data is acquired at a second time, and so on. In the target driving cycle of the target vehicle 110, it may generate vehicle data from a first time to an M-th time, where a time interval between each time is Δ T, may set a window and a step length of the window, and intercept the vehicle data from the first time to the M-th time to obtain vehicle data for N time periods. Wherein M, N is a natural number, and M is more than N and is more than or equal to 2.

For example, if the size of the window is 3 and the step length is 1, the vehicle data of the first time zone is the vehicle data of the first time, the vehicle data of the second time and the vehicle data of the third time, the vehicle data of the second time zone is the vehicle data of the second time, the vehicle data of the third time and the vehicle data of the fourth time, … …, the vehicle data of the i-th time zone is the vehicle data of the i-th time, the vehicle data of the i +1 th time and the vehicle data of the i +2 th time, … …, and the vehicle data of the N-th time zone is the vehicle data of the M-2 th time, the vehicle data of the M-1 th time and the vehicle data of the M-th time. Wherein i is a natural number, and i is more than or equal to 1.

Wherein the vehicle data includes electrical, physical and environmental parameters. The electrical parameter is data relating to the electrical performance of an electric motor equipped on a target vehicle, the physical parameter is data relating to the operating condition of the electric motor, and the environmental parameter is data relating to the environment in which the electric motor is located.

Optionally, in this embodiment of the present application, the electrical parameter includes current data of the rotor and/or voltage data of the rotor; optionally, the electrical parameters further include bus current data of the motor and/or bus voltage data of the motor; optionally, the physical parameter comprises rotational speed data of the motor and/or torque data of the motor; optionally, the environmental parameter includes a temperature of an environment in which the motor is located; optionally, the environmental parameter further comprises a temperature of cooling water of the motor.

And the server 120 is configured to process the vehicle data by calling the rotor temperature prediction model to obtain a rotor temperature, and send the rotor temperature to the control device 111.

For example, the process of the server 120 obtaining the rotor temperature by processing the vehicle data by invoking the rotor temperature prediction model includes, but is not limited to: performing feature construction on the vehicle data of the ith time period in the N time periods to obtain target feature variables, wherein the target feature variables comprise the feature variables of the vehicle data of the ith time period and the feature variables of the vehicle data before the ith time period; calling rotor temperature prediction model pair i₁The rotor temperature and the target characteristic variable at the moment are processed to obtain the ith_nThe rotor temperature at that moment; and (5) enabling i to be i +1, and repeating the steps until i is N. Through the above steps, the rotor temperature for each of the N periods can be obtained.

Wherein the vehicle data of the ith time period includes the ith time period₁Time to ith_nVehicle data at time, n is window size, i_nThe time represents the nth time in the ith time period, n is a natural number and is more than or equal to 2. For example, as described above, if the ith slot is the second slot and n is 3, the ith slot is set to be the second slot₁The vehicle data at the time is the vehicle data at the second time, i₂The vehicle data at the time is the number of vehicles at the third timeAccording to the i₃The vehicle data at the time is the vehicle data at the fourth time.

The control device 111 is further configured to receive the rotor temperature via the communication device 112, and obtain a rotor temperature curve from the rotor temperature at each time, where the rotor temperature curve is a time-temperature curve.

In another alternative embodiment of the present application, the rotor temperature prediction model may be deployed in the control device 111 or other device in the target vehicle 110 (e.g., a domain controller).

A control device 111 for acquiring vehicle data of an i-th time period; carrying out feature construction on the vehicle data in the ith time period to obtain a target feature variable; calling rotor temperature prediction model pair i₁The rotor temperature and the target characteristic variable at the moment are processed to obtain the ith_nThe rotor temperature at that moment; and (5) enabling i to be i +1, and repeating the steps until i is N.

Optionally, in this embodiment of the application, after vehicle data is acquired at predetermined time intervals, the vehicle data is sent to the server 120 through the communication device 112, where the vehicle data is used to enable the server 120 to train and iterate the rotor temperature prediction model according to the vehicle data to obtain an updated rotor temperature prediction model, enable the control device 111 to receive an upgrade file through the communication device 112, and enable the rotor temperature prediction model to be upgraded into the updated rotor temperature prediction model according to the upgrade file.

Referring to fig. 2, a flow chart of a method for calculating a rotor temperature provided in an exemplary embodiment of the present application is shown, the method being applicable to the implementation environment provided in the embodiment of fig. 1, and the method including:

in step 201, vehicle data of the ith time slot is acquired.

As described above, the vehicle data includes electrical parameters, physical parameters, and environmental parameters. Because the correlation between the electrical parameters, the physical parameters and the environmental parameters and the rotor temperature is high, the calculation accuracy of the rotor temperature can be improved by acquiring vehicle data containing the electrical parameters, the physical parameters and the environmental parameters.

For example, when step 201 is executed by the control device, the control device may obtain the vehicle data by direct measurement and/or indirect measurement (data calculated by data that can be directly measured). For example, if the physical parameters in the motor data include the rotation speed data of the motor and the torque data of the motor, the rotation speed data and the torque data can be acquired by a sensor equipped in the vehicle; when step 201 is performed by the server, the vehicle data transmitted thereto after the control apparatus acquisition may be received by the communication apparatus.

The control device may acquire vehicle data according to a predetermined time interval Δ T, acquire vehicle data at a first time at an initial time, acquire vehicle data at a second time after Δ T, and so on. Optionally, after acquiring the new vehicle data, the control device sends the new vehicle data to the server through the communication device, and the server receives the vehicle data and may predict the rotor temperature of the target vehicle through the vehicle data, or train and iterate a rotor temperature prediction model.

For example, the vehicle data of N time periods can be obtained by intercepting the vehicle data according to a predetermined step length through a window with the size of N, the vehicle data of the ith time period is selected from the vehicle data of the N time periods, and the vehicle data of the ith time period comprises the ith time period₁Time to ith_nVehicle data of the time of day.

And 202, carrying out feature construction on the vehicle data in the ith time period to obtain a target feature variable.

For example, when the feature of the vehicle data in the ith time period is constructed, the feature variable of the vehicle data before the ith time period can be increased, and the target feature variable is obtained through processing.

Step 203, calling a rotor temperature prediction model pair ith₁The rotor temperature and the target characteristic variable of the time period are processed to obtain the ith_nThe rotor temperature at that moment.

Optionally, the rotor temperature prediction model may include at least one of a regression model, a recurrent neural network model, and a convolutional neural network model.

As mentioned above, when the rotor temperature is highWhen the prediction model is deployed in the server, the control equipment sends vehicle data to the server through the communication equipment, the server conducts feature construction on the vehicle data in the ith time period to obtain a target feature variable, and the rotor temperature prediction model is called to conduct feature construction on the ith time period₁The rotor temperature and the target characteristic variable at the moment are processed to obtain the ith_nThe rotor temperature at the time point is obtained from the first time point to the mth time point by repeating the above steps while making i equal to i + 1.

As described above, when the rotor temperature prediction model is deployed in the target vehicle, the control device performs feature construction on the vehicle data in the ith time period to obtain the target feature variable, and calls the rotor temperature prediction model to perform feature construction on the ith time period₁The rotor temperature and the target vehicle data at the moment are processed to obtain the ith_nThe rotor temperature at the time point is obtained from the first time point to the mth time point by repeating the above steps while making i equal to i + 1.

Wherein, the ith₁The rotor temperature at the moment can be calculated by the previous cycle process if the ith₁The rotor temperature at that moment cannot be calculated by the previous cycle (e.g., ith₁The rotor temperature at the time is the rotor temperature at the first time), the temperature may be a predetermined value.

To sum up, in the embodiment of the application, the rotor temperature is obtained by processing the vehicle data by calling the machine learning model including the deep learning algorithm, so that the problems of high cost and narrow applicability of measuring the rotor temperature through wireless remote measurement are solved, and meanwhile, the predicted rotor temperature has high accuracy and accuracy because the vehicle data includes electrical parameters, physical parameters and environmental parameters highly related to the rotor temperature.

Referring to fig. 3, which shows a flowchart of a feature construction method provided in an exemplary embodiment of the present application, the method may be performed by the control device 111 or the server 120 in the embodiment of fig. 1, and the method may be an optional implementation of step 202 in the embodiment of fig. 2, and the method includes:

step 301, performing feature extraction on the target data of the ith time period to obtain a feature variable of the ith time period.

And step 302, adding an auxiliary characteristic variable in the characteristic variable of the ith time period to obtain an expanded characteristic variable, wherein the auxiliary characteristic variable comprises the characteristic variable in the vehicle data before the ith time period.

Optionally, the auxiliary variable includes an electrical parameter, a physical parameter, and an environmental parameter in a predetermined number (the predetermined number may be 1, or greater than 1) of time periods before the ith time period; optionally, the auxiliary variable further includes an average value of the electrical parameter, the physical parameter, and the environmental parameter in a predetermined number of time periods before the ith time period; optionally, the auxiliary variable further includes a variance of the electrical parameter, the physical parameter, and the environmental parameter in a predetermined number of time periods before the ith time period.

For example, for the characteristic variables of the tenth time period, the electrical, physical and environmental parameters and their mean and variance at the ninth, eighth and seventh time periods may be added as auxiliary variables to obtain the original target characteristic variables.

And step 303, carrying out dimension scaling on the expanded characteristic variables to obtain target characteristic variables.

Referring to fig. 4, which shows a flowchart of a model optimization method provided in an exemplary embodiment of the present application, the method may be a method performed before the embodiment of fig. 2, the training method may be performed by the server 120 in the embodiment of fig. 1, or other server or computer device, and the method includes:

in step 401, historical vehicle data is obtained.

For example, vehicle data of the same vehicle in a plurality of driving cycles or vehicle data of different vehicles in a plurality of driving cycles can be collected as historical vehicle data through a wireless network. The historical vehicle data can be divided according to corresponding time periods, and the vehicle data in each time period further comprises the rotor temperature as a calibration result.

And step 402, cleaning historical vehicle data to obtain cleaned data.

The cleaned data comprises a plurality of driving cycles, and electrical parameters, physical parameters and environmental parameters under different working conditions.

And 403, performing feature construction on the cleaned data to obtain a target feature variable.

For example, for the data in the k-th (k is a natural number, k ≧ 1) driving cycle among the plurality of driving cycles, it includes k_NData of time period, for k_NThe data for each of the time periods may be characterized by the methods provided in the embodiments of fig. 2 and 3.

And step 404, outputting the target characteristic vector to a plurality of algorithm models of different types to obtain an output result.

For example, the different types of algorithm models include a linear regression model, a random forest model, a fully-connected neural network model and a recurrent neural network model, and the target feature vectors may be respectively input to the linear regression model, the random forest model, the fully-connected neural network model and the recurrent neural network model to obtain an output result of each algorithm model, where the output result is the rotor temperature of each time period.

And 405, training each type of algorithm model for a preset number of times according to the output result to obtain optimized models of different types.

Illustratively, an error can be obtained by comparing the output result and the calibration result of each algorithm model, and the algorithm model of each type is trained through the error to obtain optimized models of different types.

During training, the optimized model can be obtained by adjusting key parameters of different types of algorithm models (such as the number and depth of trees in the random forest model, the number of layers of a neural network and an activation function).

And 406, taking the model with the minimum error in the optimized models of different types as the model to be trained.

Step 407, training the model to be trained through the target characteristic variables to obtain a rotor temperature prediction model.

Referring to fig. 5, a schematic diagram of a training architecture of a rotor temperature prediction model provided in an exemplary embodiment of the present application is shown, which includes:

acquisition module510：

The present module may be deployed in a server (hereinafter referred to as "cloud") or a control device (e.g., ECU) of a target vehicle. If the module is deployed at the cloud end, vehicle data is transmitted to the cloud end in real time through communication equipment (such as a Tbox) and can be called by a script program compiled by the cloud end at any time; if the vehicle data reading device is arranged on a control device, the required vehicle data can be directly read in the controller.

Pre-processing module520：

The module is configured to extract data required for training from the vehicle data acquired by the acquisition module 410, and obtain a data set for training a machine learning model and a training set for testing the machine learning model through feature construction.

The data required for training may include variables that affect the temperature of the rotor of the motor, such as, for example, current data of the rotor, voltage data of the rotor, bus current data of the motor, bus voltage data of the motor, rotational speed data of the motor, torque data of the motor, temperature of the environment in which the motor is located, and temperature of cooling water of the motor.

And constructing auxiliary variables for representing local running conditions on the basis of the extracted data, for example, adding values of parameters in a preset time period before the time and average values and variances of the parameters in the corresponding time period as auxiliary characteristics for target data of each time period so as to add characteristics containing the preamble operation information of the motor to the current data row. Further, the initial temperature at the time of start of the rotor temperature is also set as the characteristic amount. After the characteristic variables are constructed, the characteristic variables are zoomed to the same dimension through a data standardization method so as to facilitate the subsequent calculation of the machine learning algorithm model.

Optimization module530：

The module mainly performs parameter optimization and performance comparison on machine learning models such as a linear regression model, a random forest model, a full-connection neural network model and a cyclic neural network model.

In order to improve the accuracy and generalization effect of the model catenary, vehicle data acquired in the same driving cycle is preprocessed and spliced together through a Python script in the module. Then, the script in the module divides vehicle data generated by different driving cycles into a training set and a testing set according to proportion, inputs the machine learning model for preliminary training, compares the performance of different machine learning models, adjusts key parameters (such as the number and depth of trees in random forests, the number of layers of neural networks, an activation function and the like) of different models through the script in the module, evaluates the optimal parameters of each type of model, and trains the optimized model. And then testing the performance and generalization capability of the optimized model by using a test set containing different working conditions, and judging the performance of different models in rotor temperature prediction through error results. And finally, selecting the model with the minimum error and the best overall performance as the model to be trained for training and online deployment.

Training module540：

The module can input the training set to the model to be trained for training to obtain the trained model, and the generalization capability of the trained model is tested based on the test set until the model meets the requirement of monitoring performance indexes.

Deployment module550：

The trained rotor temperature prediction model can be deployed in a server (also called a cloud computing platform or a cloud), so that the function of monitoring the stator temperature of the motor on line is realized.

Referring to fig. 6, a block diagram of a computing device provided in an exemplary embodiment of the present application, which may be implemented as a control device or a server in the above embodiments through software, hardware or a combination of the two, is shown, and the computing device includes an obtaining module 610 and a processing module 620.

The obtaining module 610 is configured to obtain vehicle data of an ith time period.

Processing module 620 forCarrying out feature construction on the vehicle data in the ith time period to obtain target feature variables, wherein the target feature variables comprise the feature variables of the vehicle data in the ith time period and the feature variables of the vehicle data before the ith time period; calling rotor temperature prediction model pair i₁The rotor temperature and the target characteristic variable at the moment are processed to obtain the ith_nThe rotor temperature at that moment; and (5) enabling i to be i +1, and repeating the steps until i is N.

Optionally, the processing module 620 is further configured to perform feature extraction on the vehicle data in the ith time period to obtain a feature variable in the ith time period; adding auxiliary characteristic variables in the characteristic variables of the ith time period to obtain expanded characteristic variables, wherein the auxiliary characteristic variables comprise the characteristic variables of the vehicle data before the ith time period; and carrying out dimension scaling on the expanded characteristic variables to obtain target characteristic variables.

Optionally, the auxiliary variables include electrical parameters, physical parameters and environmental parameters in a predetermined number of time periods before the ith time period.

Optionally, the auxiliary variable further includes an average value of the electrical parameter, the physical parameter, and the environmental parameter in a predetermined number of time periods before the ith time period.

Optionally, the auxiliary variable further includes a variance of the electrical parameter, the physical parameter, and the environmental parameter in a predetermined number of time periods before the ith time period.

Optionally, the electrical parameter comprises current data of the rotor and/or voltage data of the rotor.

Optionally, the electrical parameters further comprise bus current data of the motor and/or bus voltage data of the motor.

Optionally, the physical parameter comprises the speed of the motor and/or the torque of the motor.

Optionally, the environmental parameter comprises a temperature of an environment in which the motor is located.

Optionally, the environmental parameter further comprises a temperature of cooling water of the motor.

Referring to FIG. 7, a block diagram of a computer device provided by an exemplary embodiment of the present application is shown. The computer device may be the control device or the server provided in any of the above embodiments, and includes: a processor 710, and a memory 720.

Processor 710 may be a Central Processing Unit (CPU), a Network Processor (NP), or a combination of a CPU and an NP. The processor 710 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a Programmable Logic Device (PLD), or a combination thereof. The PLD may be a Complex Programmable Logic Device (CPLD), a field-programmable gate array (FPGA), a General Array Logic (GAL), or any combination thereof.

The memory 720 is connected to the processor 710 via a bus or other means, and at least one instruction, at least one program, set of codes, or set of instructions is stored in the memory 720, and loaded and executed by the processor 710 to implement the method executed by the control device or server as in any of the above embodiments. The memory 720 may be a volatile memory (volatile memory), a non-volatile memory (non-volatile memory), or a combination thereof. The volatile memory may be a random-access memory (RAM), such as a Static Random Access Memory (SRAM) or a Dynamic Random Access Memory (DRAM). The nonvolatile memory may be a Read Only Memory (ROM), such as a Programmable Read Only Memory (PROM), an Erasable Programmable Read Only Memory (EPROM), and an electrically erasable programmable read-only memory (EEPROM). The nonvolatile memory may also be a flash memory (flash memory), a magnetic memory such as a magnetic tape (magnetic tape), a floppy disk (floppy disk), and a hard disk. The non-volatile memory may also be an optical disc.

The present application further provides a computer-readable storage medium having at least one instruction, at least one program, a set of codes, or a set of instructions stored therein, which is loaded and executed by the processor to implement the method for calculating the rotor temperature according to any of the above embodiments.

The present application also provides a computer program product, which when run on a computer causes the computer to execute the method for calculating the rotor temperature provided by the above-mentioned method embodiments.

It should be understood that the above examples are only for clarity of illustration and are not intended to limit the embodiments. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. And obvious variations or modifications of this invention are intended to be covered by the scope of the invention as expressed herein.