Uncertainty Analysis Using the Gap Metric Approach for a Single Leg of a Quadruped Robot

Abstract

Robots are built to interact with humans and tools in different types of environment. One of the big issues for the control system design of robots is the existence of uncertainties that can move the operating point to an unexpected place causing instability and lose of performance. In special, legged robots experiences large variability in the mass of the legs, as well as other parameters, during the walk. Therefore, robust and adaptive control strategies are developed to handle the uncertainties problems such as the H∞ control and gain scheduling, however there are drawbacks in each technique. The robust control for instance tends to generate high order controllers and to be very conservative besides the complex mathematical background involved in the robust control design. Based on that, this paper tries to overcome some of the issues of robust control using a relatively new mathematical tool in the control literature and applications called gap metric . [...]