Mobile website
Volume 50,Issue 2,2023 Table of Contents
2023, 50(2):1-6,12.
DOI: 10.12177/emca.2022.140
Abstract:
The permanent magnet motor for electric submersible pump works in a special environment of high temperature and high pressure. The lubricating oil is used as the lubricating and cooling material of the submersible motor and lubricating oil moves with the rotor in the air gap. In previous studies, only the oil friction loss of the rotor is considered. However, compared with cooling methods such as air cooling and water cooling, the higher viscosity of the lubricating oil is a factor that cannot be ignored when oil cooling. And the motor speed of the electric submersible pump is high. Whether the oil friction loss caused by lubricating oil on the stator side can be ignored and its influence on the temperature rise of the motor need to be further explored. Therefore, taking a 15 kW submersible permanent magnet motor as an example, the approximate analytical formula of the stator oil friction loss is derived, and the relationship between the stator oil friction loss and the rotor oil friction loss is proposed. Based on computational fluid dynamics and fluid-structure coupling heat transfer theory, the fluid flow characteristics at the air gap are studied. The accuracy of the approximate analytical formula for stator oil friction loss is proved. The temperature rise distribution of each part of the motor is obtained, and the influence of the stator oil friction loss on the temperature rise of the motor is analyzed.
The permanent magnet motor for electric submersible pump works in a special environment of high temperature and high pressure. The lubricating oil is used as the lubricating and cooling material of the submersible motor and lubricating oil moves with the rotor in the air gap. In previous studies, only the oil friction loss of the rotor is considered. However, compared with cooling methods such as air cooling and water cooling, the higher viscosity of the lubricating oil is a factor that cannot be ignored when oil cooling. And the motor speed of the electric submersible pump is high. Whether the oil friction loss caused by lubricating oil on the stator side can be ignored and its influence on the temperature rise of the motor need to be further explored. Therefore, taking a 15 kW submersible permanent magnet motor as an example, the approximate analytical formula of the stator oil friction loss is derived, and the relationship between the stator oil friction loss and the rotor oil friction loss is proposed. Based on computational fluid dynamics and fluid-structure coupling heat transfer theory, the fluid flow characteristics at the air gap are studied. The accuracy of the approximate analytical formula for stator oil friction loss is proved. The temperature rise distribution of each part of the motor is obtained, and the influence of the stator oil friction loss on the temperature rise of the motor is analyzed.
2023, 50(2):7-12.
DOI: 10.12177/emca.2022.164
Abstract:
In view of the problems of stator electrically excited brushless synchronous motor（SEEBSM） with reluctance rotor, such as slow starting and low load capacity, the effect of adding cage bars inside the reluctance rotor on the electromagnetic performance and temperature is studied. The working principle and operation mode of the reluctance rotor with cage bars are described. The effect of the reluctance rotor with and without cage bars and different combination forms of cage bars on the rotation speed, and the effect of cage bars on the load capacity are studied by finite element method respectively. The threedimensional steadystate temperature field physical model is established, and the steadystate temperature rise of reluctance rotor with and without cage bars is simulated. The results show that the reluctance rotor with cage bars can greatly improve the starting performance and load capacity. Although the steadystate temperature rise of main structure increases significantly during load operation, it still meets the requirement of insulation class.
In view of the problems of stator electrically excited brushless synchronous motor（SEEBSM） with reluctance rotor, such as slow starting and low load capacity, the effect of adding cage bars inside the reluctance rotor on the electromagnetic performance and temperature is studied. The working principle and operation mode of the reluctance rotor with cage bars are described. The effect of the reluctance rotor with and without cage bars and different combination forms of cage bars on the rotation speed, and the effect of cage bars on the load capacity are studied by finite element method respectively. The threedimensional steadystate temperature field physical model is established, and the steadystate temperature rise of reluctance rotor with and without cage bars is simulated. The results show that the reluctance rotor with cage bars can greatly improve the starting performance and load capacity. Although the steadystate temperature rise of main structure increases significantly during load operation, it still meets the requirement of insulation class.
3 Research on Cooling Fans Performance and Aerodynamic Noise of Metro Permanent Magnet Traction Motor
2023, 50(2):13-19.
DOI: 10.12177/emca.2022.169
Abstract:
According to the balance design requirements of self-fan air-cooling metro permanent magnet traction motor for cooling effect and aerodynamic noise suppression, the cooling effect and aerodynamic noise of a metro permanent magnet traction motor with two cooling fans are numerically analyzed and experimentally studied. The main difference between the two fans is the outer diameter of the fan. The outer diameter of fan B is about 23% smaller than that of fan A. The research results show that compared with fan A, the ventilation quantity, static pressure and power consumption caused by fan B in the motor air duct are reduced by 29%, 47.4% and 53% respectively. Under the rated operating condition of the motor, the temperature rise of the motor caused by the flow reduction increases by about 9 K, which meets the design requirements of the motor temperature rise, and the corresponding aerodynamic noise is reduced by 3.3 dB(A), which has a good noise reduction effect.
According to the balance design requirements of self-fan air-cooling metro permanent magnet traction motor for cooling effect and aerodynamic noise suppression, the cooling effect and aerodynamic noise of a metro permanent magnet traction motor with two cooling fans are numerically analyzed and experimentally studied. The main difference between the two fans is the outer diameter of the fan. The outer diameter of fan B is about 23% smaller than that of fan A. The research results show that compared with fan A, the ventilation quantity, static pressure and power consumption caused by fan B in the motor air duct are reduced by 29%, 47.4% and 53% respectively. Under the rated operating condition of the motor, the temperature rise of the motor caused by the flow reduction increases by about 9 K, which meets the design requirements of the motor temperature rise, and the corresponding aerodynamic noise is reduced by 3.3 dB(A), which has a good noise reduction effect.
4 Strength and Interference Analysis of Carbon Fiber Sleeve Rotor for Interior Permanent Magnet Motor
2023, 50(2):20-23,35.
DOI: 10.12177/emca.2022.173
Abstract:
In order to solve the strength problem of interior high-speed permanent magnet motor at high speed, the rotor punch structure and the external carbon fiber banding of interior permanent magnet motor are comprehensively designed and studied. Finite element method is used to analyze the overall stiffness coordination deformation of the punch and carbon fiber sleeve, and the different effects of the carbon fiber binding preload on the initial binding state and highspeed operation are mainly considered. The relationship between the initial interference and the thickness of the binding layer and the static and dynamic stress distribution is analyzed, and the rules of the carbon fiber sleeve strength and the punch strength of the interior permanent magnet motor rotor under dynamic and static conditions are obtained.
In order to solve the strength problem of interior high-speed permanent magnet motor at high speed, the rotor punch structure and the external carbon fiber banding of interior permanent magnet motor are comprehensively designed and studied. Finite element method is used to analyze the overall stiffness coordination deformation of the punch and carbon fiber sleeve, and the different effects of the carbon fiber binding preload on the initial binding state and highspeed operation are mainly considered. The relationship between the initial interference and the thickness of the binding layer and the static and dynamic stress distribution is analyzed, and the rules of the carbon fiber sleeve strength and the punch strength of the interior permanent magnet motor rotor under dynamic and static conditions are obtained.
2023, 50(2):24-35.
DOI: 10.12177/emca.2022.172
Abstract:
The magnitude of vibration and noise is one of the indexes to measure the quality of household variable frequency air conditioning compressor. In order to solve the problem of high frequency noise of the household variable frequency air conditioning compressor, a permanent magnet synchronous motor with 6pole 9slot for household air conditioning compressor and its control system are used as the research objects. Based on the fieldcircuit coupling method, the harmonic current components and high frequency acoustic vibration response characteristics introduced by conventional space vector pulse width modulation (SVPWM) technology and its control strategy are studied. Firstly, the generation mechanism of high frequency noise source of permanent magnet motor is analyzed by analytic method. Secondly, the effect of current harmonics on the order and frequency characteristics of radial electromagnetic force wave is described in principle. Then, the fieldcircuit coupling model of permanent magnet motor is established, and the relationship between high frequency radial electromagnetic force wave frequency and switching frequency under different load conditions is analyzed in detail. Finally, the acoustic vibration characteristic test is used to verify the results. The results show that the fieldcircuit coupling model can consider the high frequency harmonic current caused by nonlinear factors such as motor body, control strategy and switching frequency, and relationship of the high frequency radial electromagnetic force wave frequency and the carrier frequency is obtained, which provides reference for high frequency noise prediction and vibration and noise reduction of permanent magnet motor.
The magnitude of vibration and noise is one of the indexes to measure the quality of household variable frequency air conditioning compressor. In order to solve the problem of high frequency noise of the household variable frequency air conditioning compressor, a permanent magnet synchronous motor with 6pole 9slot for household air conditioning compressor and its control system are used as the research objects. Based on the fieldcircuit coupling method, the harmonic current components and high frequency acoustic vibration response characteristics introduced by conventional space vector pulse width modulation (SVPWM) technology and its control strategy are studied. Firstly, the generation mechanism of high frequency noise source of permanent magnet motor is analyzed by analytic method. Secondly, the effect of current harmonics on the order and frequency characteristics of radial electromagnetic force wave is described in principle. Then, the fieldcircuit coupling model of permanent magnet motor is established, and the relationship between high frequency radial electromagnetic force wave frequency and switching frequency under different load conditions is analyzed in detail. Finally, the acoustic vibration characteristic test is used to verify the results. The results show that the fieldcircuit coupling model can consider the high frequency harmonic current caused by nonlinear factors such as motor body, control strategy and switching frequency, and relationship of the high frequency radial electromagnetic force wave frequency and the carrier frequency is obtained, which provides reference for high frequency noise prediction and vibration and noise reduction of permanent magnet motor.
2023, 50(2):36-40,55.
DOI: 10.12177/emca.2022.163
Abstract:
In order to overcome the normal force of the singlesided permanent magnet eddy current braking and improve the braking performance, the characteristics of two doublesided Halbach arrays permanent magnet eddy current braking are studied. The space magnetic field equations are established by using electromagnetic theory, and the analytical expressions of the two systems braking force are derived. And the corresponding finite element model is built to verify the correctness of the analytical solutions. The braking forces of the two doublesided Halbach arrays permanent magnet eddy current braking systems are compared, and the braking force of the scheme 1 is close to twice of that of the scheme 2 at low speed. Finally, the influence of the permanent magnet length, permanent magnet height, air gap, conductor plate conductivity and thickness on the braking performance of scheme 1 is analyzed. The results show that the ratio of length to height of permanent magnet should be 1.2 to obtain the maximum braking efficiency. To avoid the normal force affecting the stable operation of the system, the upper and lower air gaps should be equal. The change of conductor plate conductivity does not affect the maximum braking force, but its corresponding speed will decrease with conductor plate conductivity increase. The change of conductor plate thickness will change the maximum braking force and its corresponding speed at the same time.
In order to overcome the normal force of the singlesided permanent magnet eddy current braking and improve the braking performance, the characteristics of two doublesided Halbach arrays permanent magnet eddy current braking are studied. The space magnetic field equations are established by using electromagnetic theory, and the analytical expressions of the two systems braking force are derived. And the corresponding finite element model is built to verify the correctness of the analytical solutions. The braking forces of the two doublesided Halbach arrays permanent magnet eddy current braking systems are compared, and the braking force of the scheme 1 is close to twice of that of the scheme 2 at low speed. Finally, the influence of the permanent magnet length, permanent magnet height, air gap, conductor plate conductivity and thickness on the braking performance of scheme 1 is analyzed. The results show that the ratio of length to height of permanent magnet should be 1.2 to obtain the maximum braking efficiency. To avoid the normal force affecting the stable operation of the system, the upper and lower air gaps should be equal. The change of conductor plate conductivity does not affect the maximum braking force, but its corresponding speed will decrease with conductor plate conductivity increase. The change of conductor plate thickness will change the maximum braking force and its corresponding speed at the same time.
LI Yaohua
,
ZHANG Xinquan
,
DONG Guoqing
,
DENG Yizhi
,
LI Maomeng
,
TONG Ruiqi
,
DING Hong
,
RAJIBUL Islam
2023, 50(2):41-48.
DOI: 10.12177/emca.2022.161
Abstract:
Aiming at the the problem of designing and adjusting the weight factors of model predictive torque control (MPTC) for permanent magnet synchronous motor (PMSM), the parallel MPTC is used to intersect voltage vector sets for single-object cost function instead of weighted summing of multi-object cost function, thus eliminating the weight factors. The parallel MPTC for PMSM is built, and the effects of the number of voltage vectors in flux control voltage vector sets and torque control voltage vector sets on the system control performance are studied. The parallel MPTC with the switching times control is further established, and the fuzzy controller is used to dynamically adjust the switching times to control the number of voltage vectors in voltage vector sets. The simulation results show parallel MPTC for PMSM does not require weight factors design and adjustment. The fuzzy parallel MPTC can dynamically adjust the number of voltage vectors in control object sets according to the realtime operation state of the system, so as to optimize the system control performances.
Aiming at the the problem of designing and adjusting the weight factors of model predictive torque control (MPTC) for permanent magnet synchronous motor (PMSM), the parallel MPTC is used to intersect voltage vector sets for single-object cost function instead of weighted summing of multi-object cost function, thus eliminating the weight factors. The parallel MPTC for PMSM is built, and the effects of the number of voltage vectors in flux control voltage vector sets and torque control voltage vector sets on the system control performance are studied. The parallel MPTC with the switching times control is further established, and the fuzzy controller is used to dynamically adjust the switching times to control the number of voltage vectors in voltage vector sets. The simulation results show parallel MPTC for PMSM does not require weight factors design and adjustment. The fuzzy parallel MPTC can dynamically adjust the number of voltage vectors in control object sets according to the realtime operation state of the system, so as to optimize the system control performances.
2023, 50(2):49-55.
DOI: 10.12177/emca.2022.171
Abstract:
The intermittent and fluctuating characteristics of wind energy pose a great challenge to the smooth operation of power grid, which causes grid enterprises to restrict wind power grid connection, resulting in curtailment behavior. Therefore, the real-time and effective prediction of wind power generation is critical for the development of wind power and the smooth operation of power grid. After analyzing several current prediction methods, a short-term wind power prediction model based on kernel principal component analysis-K-means clustering-gated recurrent unit (KPCA-K-means-GRU) is proposed. Multidimensional data can restore the real physical state better, but data with too high dimensions will cause dimension disaster. Therefore, a non-linear KPCA is used to reduce the data dimension while retaining the information of high dimension data. Then based on the idea of similar days for load prediction, unsupervised clustering of reduced dimension data by K-means is used to establish different prediction models to improve prediction accuracy. Finally, the GRU neural network parameters of different kinds of data are trained separately, and then classification prediction is carried out to obtain a more appropriate network model.
The intermittent and fluctuating characteristics of wind energy pose a great challenge to the smooth operation of power grid, which causes grid enterprises to restrict wind power grid connection, resulting in curtailment behavior. Therefore, the real-time and effective prediction of wind power generation is critical for the development of wind power and the smooth operation of power grid. After analyzing several current prediction methods, a short-term wind power prediction model based on kernel principal component analysis-K-means clustering-gated recurrent unit (KPCA-K-means-GRU) is proposed. Multidimensional data can restore the real physical state better, but data with too high dimensions will cause dimension disaster. Therefore, a non-linear KPCA is used to reduce the data dimension while retaining the information of high dimension data. Then based on the idea of similar days for load prediction, unsupervised clustering of reduced dimension data by K-means is used to establish different prediction models to improve prediction accuracy. Finally, the GRU neural network parameters of different kinds of data are trained separately, and then classification prediction is carried out to obtain a more appropriate network model.
2023, 50(2):56-60，66.
DOI: 10.12177/emca.2022.165
Abstract:
Nowadays, a large number of distributed power sources and unmeasurable load branches are connected to the distribution network, and the original protection scheme cannot meet the requirements of the new distribution network. By analyzing the asymmetric fault characteristics of active distribution network with unmeasurable load branches, a negative sequence fault superimposed network equivalent model inside and outside the area is established. By analyzing this equivalent model, the different model expressions of the protected lines under internal and external fault are extracted to design protection criteria based on model identification. A protection scheme based on the negative sequence fault superimposed network model identification is designed. A 10 kV active distribution network model identification is built by using PSCAD simulation software to validate the proposed scheme. The results show that the scheme can accurately identify the fault area.
Nowadays, a large number of distributed power sources and unmeasurable load branches are connected to the distribution network, and the original protection scheme cannot meet the requirements of the new distribution network. By analyzing the asymmetric fault characteristics of active distribution network with unmeasurable load branches, a negative sequence fault superimposed network equivalent model inside and outside the area is established. By analyzing this equivalent model, the different model expressions of the protected lines under internal and external fault are extracted to design protection criteria based on model identification. A protection scheme based on the negative sequence fault superimposed network model identification is designed. A 10 kV active distribution network model identification is built by using PSCAD simulation software to validate the proposed scheme. The results show that the scheme can accurately identify the fault area.
2023, 50(2):61-66.
DOI: 10.12177/emca.2022.159
Abstract:
Aiming at the problems of unstable operation and low efficiency caused by the imbalance of internal output of photovoltaic photothermal complementary generator set, a scheduling method of wind power photovoltaic photothermal complementary generator set based on Kmeans clustering algorithm is proposed. Considering the intermittence, fluctuation and randomness of photovoltaic photothermal generation set, K-means clustering algorithm is used to classify and analyze the scheduled data in advance. The objective functions of four possible combinations of light and wind power are established, and the function value is calculated as the initial condition value of the next scheduling constraint. The scheduling method combines power balance, energy storage balance, photovoltaic photothermal uphill and downhill events, calculates the realtime output value and the optimal scheduling output value, and solves the difference between the two to achieve efficient scheduling. The experimental results show that the proposed method can effectively accomplish the load and power schedule of the generator set, and the problems of operation fluctuation and low efficiency are significantly improved, which plays an important role in the stable operation of power plants.
Aiming at the problems of unstable operation and low efficiency caused by the imbalance of internal output of photovoltaic photothermal complementary generator set, a scheduling method of wind power photovoltaic photothermal complementary generator set based on Kmeans clustering algorithm is proposed. Considering the intermittence, fluctuation and randomness of photovoltaic photothermal generation set, K-means clustering algorithm is used to classify and analyze the scheduled data in advance. The objective functions of four possible combinations of light and wind power are established, and the function value is calculated as the initial condition value of the next scheduling constraint. The scheduling method combines power balance, energy storage balance, photovoltaic photothermal uphill and downhill events, calculates the realtime output value and the optimal scheduling output value, and solves the difference between the two to achieve efficient scheduling. The experimental results show that the proposed method can effectively accomplish the load and power schedule of the generator set, and the problems of operation fluctuation and low efficiency are significantly improved, which plays an important role in the stable operation of power plants.
2023, 50(2):67-72，81.
DOI: 10.12177/emca.2022.160
Abstract:
Massive multivariate heterogeneous smart grid data are compressed and stored without processing, which has the problems of large compression error and long running time, affecting the compression and storage effect. Therefore, a data compression and storage method of massive multivariate heterogeneous smart grid based on state estimation is proposed. The massive multivariate heterogeneous smart grid data are integrated to deduce the accurate smart grid data. And the massive multivariate heterogeneous smart grid data are compressed through the Tucker decomposition smart grid big data compression method. Extensible markup language (XML) technology is used to preprocess the data, and combined with the nonrelational database technology to realize the rapid storage of massive multivariate heterogeneous smart grid data. The experimental results show that this method has high accuracy in estimating the dynamic parameters of line resistance and reactance, low average absolute error and Fnorm error of data compression, short running time and certain practical application performance.
Massive multivariate heterogeneous smart grid data are compressed and stored without processing, which has the problems of large compression error and long running time, affecting the compression and storage effect. Therefore, a data compression and storage method of massive multivariate heterogeneous smart grid based on state estimation is proposed. The massive multivariate heterogeneous smart grid data are integrated to deduce the accurate smart grid data. And the massive multivariate heterogeneous smart grid data are compressed through the Tucker decomposition smart grid big data compression method. Extensible markup language (XML) technology is used to preprocess the data, and combined with the nonrelational database technology to realize the rapid storage of massive multivariate heterogeneous smart grid data. The experimental results show that this method has high accuracy in estimating the dynamic parameters of line resistance and reactance, low average absolute error and Fnorm error of data compression, short running time and certain practical application performance.
2023, 50(2):73-81.
DOI: 10.12177/emca.2022.168
Abstract:
In order to solve the problems of bidirectional DC-DC converter in traditional PI controlled optical storage microgrid system, such as large voltage fluctuation of DC bus, poor efficiency of charge and discharge and weak anti-interference ability, a double closed-loop control strategy based on adaptive particle swarm optimization (APSO) is designed. Firstly, the mathematical model of bidirectional DC-DC converter is established. Secondly, a double closed-loop control system with voltage loop linear active disturbance rejection control (ADRC) and current loop PI control is designed, and feedforward control is added to the voltage loop to enhance the robustness of the control system. Then, a parameter optimization system based on APSO algorithm is proposed to solve the problem that the parameters of active disturbance rejection controller are difficult to tune. The adaptive inertia weight factor is introduced in the algorithm so that the inertia weight can be dynamically adjusted in the process of particle swarm optimization to obtain better optimization results. Finally, a kind of integral of time and absolute error (ITAE) index with penalty function is designed as fitness function, and the control parameters of feedforward linear active disturbance rejection control (FF-LADRC) system are optimized autonomously. The simulation results in MATLAB show that the proposed control strategy can effectively reduce the voltage fluctuation of DC bus, improve the charge-discharge performance of energy storage system, and solve the problem of parameters setting of linear active disturbance rejection controller.
In order to solve the problems of bidirectional DC-DC converter in traditional PI controlled optical storage microgrid system, such as large voltage fluctuation of DC bus, poor efficiency of charge and discharge and weak anti-interference ability, a double closed-loop control strategy based on adaptive particle swarm optimization (APSO) is designed. Firstly, the mathematical model of bidirectional DC-DC converter is established. Secondly, a double closed-loop control system with voltage loop linear active disturbance rejection control (ADRC) and current loop PI control is designed, and feedforward control is added to the voltage loop to enhance the robustness of the control system. Then, a parameter optimization system based on APSO algorithm is proposed to solve the problem that the parameters of active disturbance rejection controller are difficult to tune. The adaptive inertia weight factor is introduced in the algorithm so that the inertia weight can be dynamically adjusted in the process of particle swarm optimization to obtain better optimization results. Finally, a kind of integral of time and absolute error (ITAE) index with penalty function is designed as fitness function, and the control parameters of feedforward linear active disturbance rejection control (FF-LADRC) system are optimized autonomously. The simulation results in MATLAB show that the proposed control strategy can effectively reduce the voltage fluctuation of DC bus, improve the charge-discharge performance of energy storage system, and solve the problem of parameters setting of linear active disturbance rejection controller.
2023, 50(2):82-90.
DOI: 10.12177/emca.2022.167
Abstract:
Under the background of “double carbon”, demand response enabled load (DREL), as a flexible power asset, is gradually actively involved in power grid dispatching. It usually arranges the power consumption according to the time-of-use (TOU) in respond to the calls to create new economic growth points. The energy distribution of DREL is different from the traditional load that is insensitive to price, and as its penetration increases, this difference may bring new challenges to distribution system state estimation (DSSE). Therefore, the historical profile of traditional load cannot accurately describe the DREL behavior, which affects DSSE results and other DSSErelated operations. To solve this problem, an online pseudo-measurement generation algorithm of DSSE based on DREL is presented, and an optimization model for DREL self-adjusting behavior is established, and the uncertainty of DREL is considered. Finally, the effectiveness of the proposed algorithm is verified by IEEE 123 bus test system.
Under the background of “double carbon”, demand response enabled load (DREL), as a flexible power asset, is gradually actively involved in power grid dispatching. It usually arranges the power consumption according to the time-of-use (TOU) in respond to the calls to create new economic growth points. The energy distribution of DREL is different from the traditional load that is insensitive to price, and as its penetration increases, this difference may bring new challenges to distribution system state estimation (DSSE). Therefore, the historical profile of traditional load cannot accurately describe the DREL behavior, which affects DSSE results and other DSSErelated operations. To solve this problem, an online pseudo-measurement generation algorithm of DSSE based on DREL is presented, and an optimization model for DREL self-adjusting behavior is established, and the uncertainty of DREL is considered. Finally, the effectiveness of the proposed algorithm is verified by IEEE 123 bus test system.
2023, 50(2):91-96.
DOI: 10.12177/emca.2022.157
Abstract:
Due to the high cost of motor bearing fatigue experiment and insufficient fault data, the effect of using artificial intelligence algorithms such as machine learning for fault diagnosis is not good. In addition, the accuracy of a single model for motor bearing fault diagnosis is also low. In order to solve these two problems, a fault diagnosis method for motor bearing fault diagnosis is proposed, which combines auxiliary classifier generative adversarial network (ACGAN) with model fusion. Firstly, the vibration data collected are converted into two-dimensional grayscale images, and each grayscale image is labeled and input into ACGAN model to generate a large number of new samples that highly fit the original data. Then the new samples are mixed with the original samples, and the data are dimensionality reduced to input into the model which is a fusion of six base-learners and one meta-learner. Finally, the diagnosis results are output by the fusion model. The test proves that ACGAN and model fusion can effectively improve the accuracy of fault diagnosis of motor bearing.
Due to the high cost of motor bearing fatigue experiment and insufficient fault data, the effect of using artificial intelligence algorithms such as machine learning for fault diagnosis is not good. In addition, the accuracy of a single model for motor bearing fault diagnosis is also low. In order to solve these two problems, a fault diagnosis method for motor bearing fault diagnosis is proposed, which combines auxiliary classifier generative adversarial network (ACGAN) with model fusion. Firstly, the vibration data collected are converted into two-dimensional grayscale images, and each grayscale image is labeled and input into ACGAN model to generate a large number of new samples that highly fit the original data. Then the new samples are mixed with the original samples, and the data are dimensionality reduced to input into the model which is a fusion of six base-learners and one meta-learner. Finally, the diagnosis results are output by the fusion model. The test proves that ACGAN and model fusion can effectively improve the accuracy of fault diagnosis of motor bearing.
You are thevisitor
沪ICP备16038578号-3
Electric Machines & Control Application ® 2025
Supported by:Beijing E-Tiller Technology Development Co., Ltd.
沪ICP备16038578号-3
Electric Machines & Control Application ® 2025
Supported by:Beijing E-Tiller Technology Development Co., Ltd.
