Optimal Real-Time Scheduling of Control Tasks With State Feedback Resource Allocation

摘要

This paper proposes a new approach for the optimal integrated control and real-time scheduling of control tasks. First, the problem of the optimal integrated control and nonpreemptive off-line scheduling of control tasks in the sense of the ${cal H}_2$ performance criterion is addressed. It is shown that this problem may be decomposed into two subproblems. The first subproblem aims at finding the optimal nonpreemptive off-line schedule and may be solved using the branch and bound method. The second subproblem uses the lifting technique to determine the optimal control gains, based on the solution of the first subproblem. Second, an efficient online scheduling algorithm is proposed. This algorithm, called the reactive pointer placement (RPP) scheduling algorithm, uses the plant state information to dispatch the computational resources in a way that improves control performance. Control performance improvements as well as stability guarantees are formally proven. Finally, simulations as well as experimental results are presented in order to illustrate the effectiveness of the proposed approach.