Method for detecting fault of electric machine

Abstract

The invention discloses a method for detecting a motor fault, and the method comprises the steps: training a motor fault detection model of a motor of a model for a whole vehicle platform, and the motor fault detection model comprises the corresponding relation between the vibration signal data of the motor and the fault information; and acquiring actual vibration signal data in the operation process of the motors of the same model of the same whole vehicle platform, and inputting the actual vibration signal data into the motor fault detection model so as to obtain fault information corresponding to the actual vibration signal data. According to the motor fault detection method and device, the fault information of the motor can be obtained by means of the motor fault detection model and the actual vibration signal data of the motor in operation, and therefore motor fault detection is achieved in a simple mode.

Description

Technical Field

[0001] The present invention relates to the technical fields of vehicles and electronic control, and in particular, to a method for detecting motor faults. Background Art

[0002] In the field of motors, for example, for three-phase motors such as permanent magnet synchronous motors or induction asynchronous motors, various motor faults may occur during motor operation, such as: stator faults, rotor faults, and mechanical matching faults between the rotor and the stator. These motor faults will affect the normal operation of the motor. For example, transient interturn short circuits are a common stator winding or rotor winding fault, which is characterized by a short circuit between two or more turns of coils. Transient interturn short circuits include short circuits caused by moisture, high temperature, or extreme operating conditions. When the motor is used in a vehicle, the higher the vehicle system voltage (for example, 800V or higher), the greater the risk of transient interturn short circuits in the motor. The development of interturn short circuits will cause the current of the motor to increase, thereby exacerbating the insulation degradation of the winding, and ultimately leading to the complete failure of the motor due to the accumulation of insulation deterioration of the stator winding or rotor winding. Therefore, early detection of interturn short circuits can maximize the prevention of motor faults and reduce maintenance costs.

[0003] However, the existing motor fault detection methods are relatively complex. Therefore, it is necessary to design a new type of motor fault detection method. Summary of the Invention

[0004] The starting point of the present invention is to provide a method for detecting motor faults, thereby solving the above problems existing in the prior art.

[0005] An embodiment of the present invention provides a method for detecting motor faults, characterized in that the method includes:

[0006] Training a motor fault detection model for a motor of a certain model on a certain vehicle platform, where the motor fault detection model includes the correspondence between the vibration signal data of the motor and the fault information;

[0007] Obtaining the actual vibration signal data during the operation of the same model motor on the same vehicle platform, and inputting the actual vibration signal data into the motor fault detection model to obtain the fault information corresponding to the actual vibration signal data.

[0008] Optionally, training a motor fault diagnosis model for a motor of a certain model on a certain vehicle platform includes:

[0009] Obtaining the sample vibration signal data of the sample motor of the said model on the said vehicle platform under different motor fault conditions;

[0010] Input the sample vibration signal data and its corresponding fault information into the motor fault diagnosis model to be trained, and obtain the motor fault diagnosis model through training.

[0011] Optionally, obtain the sample vibration signal data and the actual vibration signal data by means of a sensor device provided on the stator core and / or the outer casing of the motor.

[0012] Optionally, the sensor device includes at least one of the following: a piezoelectric sensor, a piezoresistive sensor, a capacitive sensor, an acceleration sensor.

[0013] Optionally, in the process of training and using the motor fault detection model, at least one of the following methods is adopted: a time-domain analysis method, a frequency-domain analysis method, a time-domain and frequency-domain combined analysis method, an AI algorithm, a low-pass or band-pass filtering method.

[0014] Optionally, the fault information includes at least one of the following: fault information related to motor stator faults, fault information related to motor rotor faults, fault information related to mechanical matching faults between the rotor and the stator.

[0015] Optionally, the fault information related to motor stator faults includes transient inter-turn short circuits in the stator windings; the fault information related to motor rotor faults includes at least one of the following: transient inter-turn short circuits in the rotor windings, demagnetization of the rotor permanent magnets, fracture of the rotor shaft, rotor bearing faults; the fault information related to mechanical matching faults between the rotor and the stator includes at least one of the following: an eccentric fault with uneven air gap distribution between the rotor and the stator, an unbalance fault with uneven mass distribution of the rotor relative to the rotation axis.

[0016] Optionally, in the case where the motor is a three-phase motor, the fault information related to motor stator faults includes transient inter-turn short circuits between phases or within phases in the three-phase stator windings.

[0017] Optionally, the fault information further includes the severity of various faults.

[0018] Optionally, after obtaining the fault information corresponding to the actual vibration signal data, at least one of the following measures is adopted: taking fault degradation management for the running motor to make the running motor enter a safe mode.

[0019] The method for detecting motor faults according to the embodiments of the present invention has at least the following advantages:

[0020] In the present invention, the fault information of the motor can be obtained by means of the motor fault detection model and the actual vibration signal data of the running motor, thereby realizing the detection of motor faults in a simple manner.

[0021] In the present invention, during the process of training and using the motor fault detection model, time-domain analysis methods, frequency-domain analysis methods, time-domain - frequency-domain combined analysis methods, AI algorithms, and low-pass or band-pass filtering methods can be adopted, thereby improving the accuracy of motor fault detection.

[0022] In the present invention, sample vibration signal data and actual vibration signal data are acquired by means of a sensor device arranged on the stator core and / or the outer housing of the motor, thereby realizing the acquisition of the vibration signal of the motor in a simple manner.

[0023] In the present invention, in the case where the motor is a three-phase motor, it is possible to detect transient inter-turn short circuits between phases or within phases in the three-phase stator windings, thereby realizing the precise diagnosis of faults in the three-phase motor stator windings.

[0024] In the present invention, the fault information may include the severity of various faults, enabling the user of the motor to take corresponding treatment measures according to the severity of the faults. Description of the Drawings

[0025] Other details and advantages of the present invention will become apparent from the detailed description provided below. It should be understood that the following drawings are merely schematic and not drawn to scale, and thus should not be considered as a limitation to the present application. The following will be described in detail with reference to the drawings, where:

[0026] Figure 1 Schematically shows a flowchart of a method for detecting motor faults according to a specific embodiment of the present invention. Detailed Embodiments

[0027] The embodiments of the present invention will be described below with reference to the drawings. In the following description, many specific details are set forth in order to enable those skilled in the art to more fully understand and implement the present invention. However, it is obvious to those skilled in the art that some of these specific details may not be required for the implementation of the present invention. In addition, it should be understood that the present invention is not limited to the specific embodiments described. On the contrary, the present invention can be implemented by any combination of the features and elements described below, regardless of whether they relate to different embodiments. Therefore, the following aspects, features, embodiments, and advantages are for illustrative purposes only and should not be regarded as elements or limitations of the claims, unless specifically recited in the claims.

[0028] The inventors of the present application have found that for a normally operating motor without faults, the magnetic field distribution in the stator laminations or iron core is uniform, and the magnetic flux is symmetrically distributed. When a fault occurs in the motor, for example, when a transient inter-turn short circuit occurs in the stator winding of the motor, the magnetic flux distribution in the stator laminations or iron core becomes asymmetric, resulting in fluctuations in the rotational speed and torque of the motor rotor shaft, and further causing changes in the vibration signal of the motor. Therefore, in the present invention, first, for a motor of a certain model on a vehicle platform, sample vibration signal data of at least one sample motor in a normal state and various fault states under all working conditions is obtained. Then, a motor fault diagnosis model is obtained based on different sample vibration signal data and their corresponding fault information. After obtaining the motor fault diagnosis model, during the operation of a motor of the same model on the same vehicle platform, the actual vibration signal data of the motor is collected and input into the motor fault detection model, so as to obtain the fault information corresponding to the actual vibration signal data. Thus, the detection of motor faults is achieved in a simple manner, with accurate fault location and high detection sensitivity.

[0029] Now referring to Figure 1 , a flowchart of a method for detecting motor faults according to a specific embodiment of the present invention is schematically shown. This method can be executed by any suitable processing device (such as a vehicle controller or a high-voltage power motor controller), and these variations do not exceed the protection scope of the present invention. As Figure 1 shown, the method includes:

[0030] Step S101, training a motor fault detection model for a motor of a certain model on a vehicle platform, where the motor fault detection model includes the correspondence between the vibration signal data of the motor and the fault information.

[0031] Specifically, the motor fault diagnosis model can be trained in the following manner: First, sample vibration signal data of at least one sample motor under full working conditions in normal operation and fault operation (especially in fault operation) is obtained. For example, by means of fault injection and other methods, different possible fault categories and fault severities can be artificially created. Then, the sample vibration signal data and their corresponding fault information are input into the motor fault diagnosis model to be trained, and the motor fault diagnosis model is obtained through training.

[0032] Those skilled in the art can understand that the correspondence between the vibration signal data and the fault information can be in any suitable form, and these variations do not exceed the protection scope of the present invention. For example, fault diagnosis thresholds corresponding to different fault information can be obtained through the analysis and calculation of the sample vibration signal data and the fault information, and the correspondence between the fault diagnosis thresholds and the corresponding fault information is stored in the motor fault diagnosis model.

[0033] Those skilled in the art can understand that the method for detecting motor faults according to the present invention requires training a motor fault detection model for a specific model of motor on a specific vehicle platform. For example, if a vehicle platform is configured with only one model of high-voltage power motor, a motor fault detection model for the high-voltage power motor of this model on this vehicle platform needs to be trained. If a vehicle platform is configured with multiple models of high-voltage power motors (for example, one model of high-voltage power motor is used as the main drive motor and another model of high-voltage power motor is used as the auxiliary drive motor), then multiple motor fault detection models for each model of high-voltage power motor on this vehicle platform need to be trained separately.

[0034] Those skilled in the art can understand that any suitable method can be used to obtain sample vibration signal data and actual vibration signal data, and these variations do not exceed the protection scope of the present invention. For example, the sample vibration signal data and actual vibration signal data can be obtained by means of a sensor device arranged on the stator core and / or the outer casing of the motor. The sensor device can include, but is not limited to: piezoelectric sensors, piezoresistive sensors, capacitive sensors, acceleration sensors.

[0035] Those skilled in the art can understand that any suitable method can be adopted to train the motor fault diagnosis model according to the sample vibration signal data and its corresponding fault information, and these variations do not exceed the protection scope of the present invention. For example, in the process of training the motor fault diagnosis model, time-domain analysis methods (such as Peak, Root mean square, Crest factor, Principal component analysis, kurtosis coefficient), frequency-domain analysis methods (such as FFT (Fast Fourier Transform), Cepstrum, Envelope, High order spectra), or time-domain and frequency-domain combined analysis methods (such as Short-time Fourier transform, Continuous wavelet transform, Choi-Williams distribution, Discrete wavelet transform, Hilbert-Huang transform, Wigner_Ville distribution) can be used to determine the characteristic parameters in the vibration signal data that have a high degree of correlation with the fault information. In the process of training the motor fault diagnosis model, an AI algorithm (such as a neural network algorithm) can also be used to associate the characteristic parameters in the vibration signal data with the corresponding fault information. In the process of training the motor fault diagnosis model, a low-pass or band-pass filtering method can also be used to filter out the noise interference in the vibration signal, so as to obtain a better signal-to-noise ratio and extract effective vibration signal data. In addition, in the process of training the motor fault diagnosis model, other methods can also be adopted according to the actual situation to improve the accuracy of the motor fault diagnosis model.

[0036] Those skilled in the art can understand that the method of the present invention can be used for any suitable type of motor, such as permanent magnet synchronous motor, asynchronous induction motor, separately excited motor, axial flux motor, and these variations do not exceed the protection scope of the present invention. Those skilled in the art can also understand that the above motors can be used for any suitable type of vehicle platform, such as pure electric vehicle, hybrid vehicle (including plug-in hybrid and range-extended hybrid) and fuel cell vehicle, and these variations do not exceed the protection scope of the present invention.

[0037] Step S102: Obtain the actual vibration signal data during the operation of the same model motor on the same vehicle platform, and input the actual vibration signal data into the motor fault detection model to obtain the fault information corresponding to the actual vibration signal data.

[0038] Those skilled in the art can understand that any suitable method can be adopted to obtain fault information corresponding to the actual vibration signal data with the aid of the motor fault detection model, and these variations do not exceed the protection scope of the present invention. For example, if the corresponding relationship between the fault diagnosis threshold and the corresponding fault information is stored in the motor fault diagnosis model, the motor can be judged whether a fault occurs and the corresponding fault information can be obtained by comparing the characteristic parameters in the actual vibration signal data with the corresponding fault diagnosis threshold.

[0039] Those skilled in the art can understand that the corresponding relationship between the vibration signal data of the motor and different types of fault information can be established in the motor fault detection model according to the method of the present invention, so that the method of the present invention can detect different types of motor fault information, and these variations do not exceed the protection scope of the present invention. For example, the fault information may include, but is not limited to: fault information related to the stator fault of the motor, fault information related to the rotor fault of the motor, and fault information related to the mechanical matching fault between the rotor and the stator. Among them, the fault information related to the stator fault of the motor may include, but is not limited to: transient inter-turn short circuit of the stator winding. The fault information related to the rotor fault of the motor may include, but is not limited to: transient inter-turn short circuit of the rotor winding, demagnetization of the rotor permanent magnet, fracture of the rotor shaft, and rotor bearing fault. The fault information related to the mechanical matching fault between the rotor and the stator may include, but is not limited to: eccentric fault with uneven air gap distribution between the rotor and the stator, and unbalance fault with uneven mass distribution of the rotor relative to the rotation axis. In addition, the fault information may further include the severity of various faults (for example, general severity level and special severity level), that is to say, the corresponding relationship between the vibration signal data and the severity of different types of motor faults may be further included in the motor fault detection model, whereby the user of the motor can take corresponding treatment measures according to the severity of the motor fault.

[0040] In the case where the motor is a three-phase motor (for example, a permanent magnet synchronous motor or an induction asynchronous motor), the fault information related to the stator fault of the motor may include transient inter-turn short circuit between phases or within a phase in the three-phase stator winding.

[0041] The detection method of the present invention can achieve high-sensitivity fault detection. For example, in the case of early occurrence of a slight transient inter-turn short circuit (non-complete insulation failure state), the transient inter-turn short circuit fault can be detected.

[0042] Those skilled in the art can understand that after obtaining the fault information corresponding to the actual vibration signal data, appropriate measures can be taken according to the actual situation to deal with the motor fault, and these variations do not exceed the protection scope of the present invention. For example, fault degradation management can be adopted for the running motor, or the running motor can be made to enter the safe mode.

[0043] Compared with the prior art, the method for detecting motor faults according to the embodiments of the present invention has at least the following advantages:

[0044] In the present invention, the fault information of the motor can be obtained by means of the motor fault detection model and the actual motor vibration signal data, thus realizing the detection of motor faults in a simple manner.

[0045] In the present invention, during the process of training and using the motor fault detection model, time-domain analysis methods, frequency-domain analysis methods, time-frequency domain combined analysis methods, neural network algorithms, and low-pass or band-pass filtering methods can be adopted, thereby improving the accuracy of motor fault detection.

[0046] In the present invention, the sample vibration signal data and the actual vibration signal data are acquired by means of the sensor device arranged on the stator core and / or the outer housing of the motor, thus realizing the acquisition of vibration signals in a simple manner.

[0047] In the present invention, in the case where the motor is a three-phase motor, it is possible to detect the transient inter-turn short circuit between phases or within a phase in the three-phase stator winding, thus realizing the accurate diagnosis of the stator winding fault of the three-phase motor.

[0048] In the present invention, the fault information may include the severity of various faults, so that the user of the motor can take corresponding treatment measures according to the severity of the faults.

[0049] It should be noted that the above description is only an example and not a limitation of the present invention. In other embodiments of the present invention, the method may have more, fewer, or different steps, and the relationships such as the sequence, inclusion, and function between the steps may be different from those described and illustrated. For example, usually multiple steps can be combined into a single step, and a single step can also be split into multiple steps. For those of ordinary skill in the art, the sequence change of each step is also within the protection scope of the present invention without creative work.

[0050] Essentially, or in terms of the part that contributes to the prior art, or all or part of the technical solution of the present invention can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which can be a vehicle controller or a high-voltage motor controller), a processor, or a microcontroller to execute all or part of the steps of the methods described in various embodiments of the present invention.

[0051] Those of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When this program is executed, it executes the steps including the above method embodiments; and the aforementioned storage medium includes: various media such as ROM, RAM, magnetic disks, or optical discs that can store program codes.

[0052] Although the present invention has been disclosed above with preferred embodiments, the present invention is not limited thereto. Any person skilled in the art, without departing from the spirit and scope of the present invention, makes various changes and modifications, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the scope defined by the claims.

Claims

1. A method for detecting motor faults, characterized in that, The method includes: Training a motor fault detection model for a motor of a certain model on a whole vehicle platform, where the motor fault detection model includes the correspondence between the vibration signal data of the motor and the fault information; Obtaining the actual vibration signal data during the operation of the motor of the same model on the same whole vehicle platform, and inputting the actual vibration signal data into the motor fault detection model to obtain the fault information corresponding to the actual vibration signal data.

2. The method according to claim 1, wherein, Training a motor fault diagnosis model for a motor of a certain model on a whole vehicle platform includes: Obtaining the sample vibration signal data of the sample motor of the said model on the said whole vehicle platform under different motor fault conditions; Inputting the sample vibration signal data and its corresponding fault information into the motor fault diagnosis model to be trained, and obtaining the motor fault diagnosis model through training.

3. The method according to claim 2, wherein Obtaining the sample vibration signal data and the actual vibration signal data by means of a sensor device arranged on the stator core and / or the outer housing of the motor.

4. The method according to claim 3, wherein The sensor device includes at least one of the following: piezoelectric sensor, piezoresistive sensor, capacitive sensor, acceleration sensor.

5. The method according to claim 1, wherein During the process of training and using the motor fault detection model, at least one of the following methods is adopted: time-domain analysis method, frequency-domain analysis method, time-domain - frequency-domain combined analysis method, AI algorithm, low-pass or band-pass filtering method.

6. The method according to claim 1, wherein The fault information includes at least one of the following: fault information related to motor stator faults, fault information related to motor rotor faults, fault information related to mechanical matching faults between the rotor and the stator.

7. According to the method described in claim 6, wherein, The fault information related to motor stator faults includes transient inter-turn short circuit of the stator winding; The fault information related to motor rotor faults includes at least one of the following: transient inter-turn short circuit of the rotor winding, demagnetization of the rotor permanent magnet, fracture of the rotor shaft, rotor bearing fault; The fault information related to mechanical matching faults between the rotor and the stator includes at least one of the following: eccentric fault with uneven air gap distribution between the rotor and the stator, unbalance fault with uneven mass distribution of the rotor relative to the rotation axis.

8. The method according to claim 7, wherein, In the case where the motor is a three-phase motor, the fault information related to motor stator faults includes transient inter-turn short circuit between phases or within a phase in the three-phase stator winding.

9. The method according to claim 6, wherein The fault information also includes the severity of various faults.

10. The method according to claim 1, wherein, After obtaining the fault information corresponding to the actual vibration signal data, at least one of the following measures is adopted: taking fault degradation management for the running motor to make the running motor enter the safe mode.

Images