Observer based stochastic trajectory tracking control of mechanical systems

This paper concerns an observer based stochastic trajectory tracking control of mechanical systems. We consider mechanical systems in the presence of noise as stochastic systems and derive a condition for a stabilizing and tracking controller to achieve each control objective in probability. Here we assume that only position measurements can be available. Velocity information is reconstructed by the position information. The condition for the combined controller-observer is derived. Since our method is based on bounded stability, the norm of tracking error with respect to a given desired trajectory remains hounded in probability and the margin of error is assignable.