Sampling head–rock contact identification for regolith sampling in space

Abstract The success rate of planet sampling can be greatly increased if rocks at the sampling spot can be identified when the sampling head touches them and instead another place is chosen for sampling. In this paper a vibration sampling method is proposed to identify such sampling head–rock contact situation. Based on a flexible coiling-type sampling arm, the vibration sampling method is developed to achieve the rock contact identification for the sampling head. Firstly, the vibration signals are generated by a vibration motor and collected via an acceleration sensor; secondly, the multi-resolution wavelet is used to filter the vibration signals, and two of the power spectrum estimation methods, Welch and Burg, are introduced to extract the features of the vibration signals; at last, whether the sampling head is touching the rock is determined by the K Nearest Neighbors. The experiments show that the sampling head–rock identification problem can be achieved successfully by adopting the vibration sampling method, and in addition, the Burg method is verified to be the better method for the identification. The technique described in the paper is very useful for achieving a better sampling mission in space.