Force Sensing for Multi-Legged Walking Robots: Theory and Experiments Part 1: Overview and Force Sensing

Abstract

This paper is intended to direct the attention of researchers and designers of multi-legged walking robots to the problem of force sensing for such vehicles. The use of force information enables such systems to achieve new mechanical properties, increases the reliability and functional capabilities of walking robots and simplifies many control algorithms. Interest in developing walking machines as vehicles with increased passability has picked up in the last sixty years. Walking machines have a number of principle advantages over wheeled and caterpillar machines: • They can move over terrains with obstacles up to the size of a leg. • The movement of their legs provides comfortable motion of the body with cargo and passengers over uneven terrain. • Ground deformation under their support legs creates less of a mechanical load on the ground than a wheel or caterpillar machine’s continuous track (especially important for tundra, mountain and hill slopes, forests, etc.). • They can work in a complexly-structured environment (sloped, confined work and operation, etc.) and move on consolidating ground and unknown terrain with varying load capacity. • They can use one or more legs as an auxiliary manipulator. The advantages of walking machines determine their range of potential applications (Song & Waldron, 1989; Okhotsimsky et al., 1992; Berns, 2006). In order to achieve foot force distribution by any method, the vehicle legs have to be equipped with force sensors, as shown in Fig. 1.1, and force feedback should be introduced into the control system (Golubev et al., 1979; McGhee et al., 1980; Klein & Briggs,1980; Devjanin et al., 1982). Basic areas of research on motion control for walking robots where the use of force control is necessary can be specified as follows: