Spinning frequency estimation algorithm of MEMS gyro’s output signal based on FFT coefficient

Abstract

To obtain accurate demodulation of the spinning frequency of MEMS carrier-driven gyro output signal, a frequency estimation algorithm combining the real and imaginary parts on the basis of fast Fourier transformation (FFT) coefficient was proposed. Both signal simulation and real-time test were performed to verify the advantages of FFT coefficient frequency estimation algorithm. In the signal simulation, CCS software was used in the demodulation and error analysis of the spinning frequency of gyro signal simulated by signal generator. Results showed that the maximum relative error of spectrum detection algorithm is 2.53%, and the maximum relative error of FFT frequency estimation algorithm is 0.1%. In the real-time test, demodulation and error analysis of the spinning frequency of gyro signal were conducted under real-time test through three-axis precise turntable. The maximum relative error of spectrum detection algorithm is 2.33%, and the maximum relative error of FFT frequency estimation algorithm is 0.1%. Experimental results indicated that, compared with spectrum detection algorithm, FFT frequency estimation algorithm exhibits higher demodulation precision, reduced maximum relative error by 2.23---2.43%, and is convenient for hardware implementation.