Abstract: With the continuous improvement of the operation speed of high-speed EMVS, the problem of low-frequency periodic hunting occurs from time to time. It is necessary and urgent to quickly and effectively identify the low frequency periodic shaking section. According to the causes of low-frequency periodic hunting, the time-frequency characteristics of vehicle response data are obtained by using short-time Fourier transform: the main frequency of lateral vibration of vehicle body is very close to the rolling pendulum frequency of upper center of vehicle body, with strong periodicity and high energy. On the basis of calculating the frequency marginal spectrum of short-time Fourier transform analysis, the diagnostic indexes which can describe the characteristics of low-frequency periodic car hunting are proposed: vibration dominant frequency, vibration degree factor and energy concentration rate. In order to improve the computational efficiency, a continuous multi wave fast search algorithm is introduced, and the dynamic response data set including hunting and a small number of non-hunting sections is obtained by taking the continuous 6-wave lateral acceleration of the car body more than 0.02g as the evaluation condition. Combined with the diagnosis index, a fast diagnosis method of low-frequency periodic hunting is formed, which can meet the needs of engineering online application. The results show that the method can effectively identify the low frequency periodic hunting.

Key words: vehicle body hunting diagnosis, vehicle acceleration, hunting motion, continuous multi-wave