A method for quantitative determination of dynamic stability of multi-legged robots based on a foot force estimation algorithm

Abstract

The invention provides a quantitative determination method for the dynamic stability of a multi-legged robot based on a leg force estimation algorithm, which is applicable to real-time detection for the stability of the robot when walking in an unstructured environment. The quantitative determination method is characterized in that motion states of joints and the robot body are acquired by a sensor, and variable values required by calculating the stability margin is solved through obverse and inverse kinematics; then the leg force of each supporting leg is estimated through the leg force estimation algorithm, and a supporting shaft and a landing point at which rollover occurs most easily are acquired according to the leg forces; a resultant force which is effective to rollover of the supporting shaft and the landing point is calculated according to a resultant force of the leg forces at the mass center and the moment of the resultant force, thus stability angles of the supporting shaft and the landing point can be acquired according to a formula, and a minimum stability angle is selected so as to calculate to acquire a normalized force angle stability margin value of the robot according to a formula. The advantages lie in that the method provided by the invention directly judges the stability margin for the supporting shaft and the landing point; stability variations brought about by the height of the mass center can be represented; and a vulnerable multi-dimensional force sensor is not required to be adopted to measure the leg force.

Description

technical field [0001] The invention belongs to the field of robot control, in particular to a method for quantitatively judging the dynamic stability of a robot based on a leg force estimation algorithm, which detects in real time the stability margin during the movement of a multi-legged robot. Background technique [0002] A multi-legged robot is a complex robot with multi-branched, redundant and time-varying topology, which has stronger terrain adaptability than wheeled robots, and has great application potential and broad application prospects in military and civilian fields. It has received extensive attention and attention from robotics researchers at home and abroad. [0003] Multi-legged robots can generally be divided into four-legged, six-legged and eight-legged robots according to the number of feet. Because the walking environment of a multi-legged robot is often complex, it may encounter rough terrain, stairs, obstacles, and even rolling stones. These complex ...

AI Technical Summary

Problems solved by technology

However, FSR can only detect positive pressure and cannot detect forces in other directions; multi-dimensional force sensors are expensive and easily damaged

Images