关键词: 卷簧式月壤取样器, 振动控制, 小波多分辨分析, 功率谱, 支持向量机

Abstract:

 It will lead to failure of planet soil sampling if the sampling head is in contact with a big-size rock,So identifying this situation is significant in the process of planet sampling. In this paper, a vibration sampling method is proposed to solve this problem. The vibration signals are acquired and analyzed in the process of sampling experiments, in order to achieve the sampling head-rock contact (SHRC) identification by taking advantage of the special structure of the mini coiling-type sampler, which takes a flexible coiling spring as its sampling arm. First of all, the vibration waves are generated by a vibration motor and the vibration signals are acquired by an acceleration sensor.Then, the multi-resolution wavelet is used to filter the vibration signals.The power spectrum estimation method Burg is introduced to extract the SHRC-related features of vibration signals.At last, the status of whether the sampling head touching the rock is identified by the C-SVM (Clustering Support Vector Machine). Experimental results show that the SHRC identification problem can be solved successfully by the analysis of the vibration signals.

Key words: Coiling-type sampler, Vibration control, Multi-resolution wavelet, Power spectrum estimation, SVM