Optimal supervisory control under partial observation for service robots

In this paper, we propose an optimal supervisory control algorithm for discrete event system (DES) to minimize the cost function under partial observation. Under the partial observation, it is generally impossible to know the current state. Moreover, it can cause the cost overestimation which prevents the optimal supervisory control. Specifically, we introduce the notions of multiple cost events and state indices. We propose a new cost-calculating algorithm based on the notions to resolve the overestimation by identifying the current state. To validate the proposed algorithm, we present an example of the task planning of the practical service robot.