Research On Decoding Method Of Resolver

Rail transportation,aerospace science and technology,military modernization,etc.,are the key fields of national development,motor is an essential important equipment in these fields,and precise measurement of motor data in order to more accurate control of its rotation is a very important link.The resolver can be widely used in various harsh conditions and accurately measure the rotor Angle,but its output signal is not the Angle information that can be read directly.The existing resolver decoding is mainly based on decoding chip,but it has disadvantages such as high cost and complex circuit,so it is a hot research topic to use software to decode resolver accurately and in real time.Based on the idea of software decoding,a decoding method using Hilbert Transform and a decoding method based on Optimal Principal Frequency Component S Transform(OPFCST)are proposed in this paper.In this paper,aiming at the problem that the quality of the envelope of the resolver decreases when the excitation frequency is low by multiplication and peak detection,the Hilbert transform is proposed to obtain the envelope signal with higher resolution and better anti-noise performance,and the improved ATO algorithm can eliminate the initial Angle error to achieve the Angle decoding of the resolver.By setting different speed and SNR,simulation experiments are carried out under various working conditions.The results show that the Angle decoding error of the decoding method using Hilbert transform and improved ATO is less than 0.5° in the SNR environment of 30 d B.Compared with other methods,the decoding accuracy of the proposed method is superior in the noise environment.The validity of the method is verified.In addition,in the existing studies,although S transform has been applied in obtaining the output signal envelope of resolver,its window width coefficient is only determined by experiment and there is no corresponding theoretical support.In this paper,the theoretical expression to determine the optimal window width coefficient is derived,and the relationship between the window width coefficient and decoding accuracy is obtained,and a decoding method based on OPFCST algorithm is proposed,which can match different window width coefficients according to different environments and different noises,and give consideration to both real-time and accuracy.The simulation experiment in MATLAB shows that the decoding error of resolver based on OPFCST can be controlled within 1° even in the strong noise environment with SNR of20 d B.To certify the practicability of the proposed method,a physical platform is also built in this paper.The DSP chip of TMS320F28346 is used as the core of the decoding system,the AD7606 chip of external sampling is used to connect the waveform to the oscilloscope through DA7724,and the waveform is compared with the real position of the motor rotor actually set.The feasibility of the proposed method in practical application is verified.