| Multiphase induction motor has been widely used in marine integrated electric propulsion system because of its advantages of large capacity,simple structure and high reliability.Compared with the three-phase motor,the amplitude of the low-order harmonics of the air-gap magnetic field of the multi-phase induction motor is effectively suppressed,but the rotor tooth harmonics caused by the cogging effect is still widespread.On the one hand,the rotor tooth harmonics increase the loss and torque ripple,and on the other hand,the rotor tooth harmonics contain the motor slip information,which can be fully utilized to realize the online estimation of the motor speed.This paper mainly studies the characteristics and detection methods of rotor tooth harmonic of multi-phase induction motors,and then realizes the reliable and accurate motor speed estimation.High-performance control of multi-phase induction motors relies on reliable and fast speed signal feedback.Although traditional speed sensors such as photoelectric encoders and rotary transformer can achieve high-performance speed detection,they have strict requirements on installation accuracy,ambient temperature,humidity,as well as eccentricity and vibration.Especially under some extreme working conditions,such as when the marine power compartment is flooded or impacted,the traditional speed sensor may fail,thereby causing the marine power system to fail.Compared with the model-based method and the method with high-frequency signal injection,the rotor tooth harmonic method has the advantages that the estimation accuracy is not affected by the variation of motor parameters and does not need additional signal injection.However,the traditional rotor tooth harmonic method usually obtains the tooth harmonic signal from the stator current,which has some limitations in versatility and signal to noise ratio(SNR).On the basis of the existing methods.This paper carries out the research on the speed estimation method of induction motor based on the detection coil,and further improves the signal to noise ratio and estimation accuracy of the rotor tooth harmonic method.The main research work is as follows:This paper firstly uses the magnetic potential permeability method and the finite element method to model and analyze the harmonic magnetic field of the rotor teeth of the ninephase induction motor.Then,a method for detecting the rotor tooth harmonic based on detection coils is proposed,which solves the problem that insufficient SNR of rotor tooth harmonic in traditional stator current method.The time domain and frequency domain characteristics of the induced voltage of the detection coil are analyzed,and the optimal design method of the detection coil span and the combination of multiple detection coils is given,which effectively improves the SNR of the signal of the rotor tooth harmonic in the detection coil,and the finite element verification is carried out.A scheme of first-order harmonic signal processing is designed,which consists of signal demodulation,frequency shifting,bandpass filtering,short-time Fourier calculation and speed calculation,and the real-time estimation of speed is realized.Finally,the real-time speed estimation experiment based on detecting coil is carried out on an 8k W nine-phase induction motor prototype.Using the 8k W squirrel-cage nine-phase induction motor experimental platform,the obtained output voltage waveform of the detection coil is subjected to fast Fourier decomposition,and the amplitude is corrected after windowing,and the components of the rotor tooth harmonic are obtained after demodulation and filtering.Through the relationship between the rotor tooth harmonic frequency and the motor speed,the speed estimation is completed,compared with the actual measured speed,and the cause of the error is analyzed.The experimental results show that the size of the signal truncation window width in the short-time Fourier calculation and the interference of other harmonics in the output voltage signal of the detection coil are the main reasons that affect the speed estimation accuracy under the condition of variable speed. |
PDF Full Text Request