Sensitivity analysis of optimal control problems for switched systems

This paper investigates the problem of sensitivity analysis for switched autonomous systems involving uncertain parameters. The aim is to find switching instants which minimizes the cost function, and meanwhile makes it robust to the variation of the original objective with respect to unknown parameters. We focus on the fixed switching mode sequences and convert the sensitivity problem into a conventional optimal switched control problem, which can be solved directly using the gradient descent algorithm. The efficacy of the proposed method is illustrated via numerical example.