Numerical procedure for optimal control of hybrid systems with sliding modes, Part I

TL;DR

This paper introduces a numerical method for solving hybrid optimal control problems with sliding modes, utilizing DAEs, adjoint equations, and Radau IIA Runge--Kutta discretization to improve accuracy and convergence.

Contribution

It presents a novel numerical procedure that handles sliding modes with DAEs and computes gradients via adjoint equations, ensuring accurate tracking and convergence.

Findings

Sliding mode handled with DAEs for accuracy

Gradients computed using adjoint equations with jump conditions

Discrete adjoint trajectories converge to continuous solutions

Abstract

This paper concerns the numerical procedure for solving hybrid optimal control problems with sliding modes. The proposed procedure has several features which distinguishes it from the other procedures for the problem. First of all a sliding mode is coped with differential-algebraic equations (DAEs) and that guarantees accurate tracking of the sliding motion surface. The second important feature is the calculation of cost and constraints functions gradients with the help of adjoint equations. The adjoint equations presented in the paper take into account sliding motion and exhibit jump conditions at transition instants. The procedure uses the discretization of system equations by Radau IIA Runge--Kutta scheme and the evaluation of optimization functions gradients with the help of the adjoint equations stated for discretized system equations. In the first part of the paper we demonstrate the correspondence between the discrete adjoint equations and the discretized version of the continuous adjoint equations in the case of system equations described by ODEs. We show that the discrete adjoint state trajectories converge to their continuous counterparts in the case of ODEs.