Hard and Soft Sub-Time-Optimal Robust Controllers

In many applicational tasks of motion control – fundamental for research in robotics – problems associated with uncertain and/or varying load (a mass or moment of inertia) can present a substantial difficulty during the synthesis of practical controlling systems. The random concept, where the load has been treated as a stochastic process, is presented in this paper. As a result, through a generalization of the classic switching curve occurring in the time-optimal approach, two control structures have been investigated: the hard, defined on the basis of the rules of the statistical decision theory, and also the soft, which additionally allows the elimination of rapid changes in control values.
The methodology proposed here may be easily adopted for other elements commonly found in mechanical systems, e.g. parameters of drive or motion resistance, giving the sub-time-optimal controlling structures that provide many advantages, especially with respect to robustness.