Convergence of Nonconvergent IRK Discretizations of Optimal Control Problems with State Inequality Constraints

It has been observed that optimization codes are sometimes able to solve inequality state constrained optimal control problems with discretizations which do not converge when used as integrators on the constrained dynamics. Understanding this phenomenon could lead to a more robust design for direct transcription codes as well as better test problems. This paper examines how this phenomenon can occur. The difference between solving index 3 differential algebraic equations (DAEs) using the trapezoid method in the context of direct transcription for optimal control problems and a straightforward implicit Runge--Kutta (IRK) formulation of the same trapezoidal discretization is analyzed. It is shown through numerical experience and theory that the two can differ as much as O(1/h3) in the control. The optimization can use a small sacrifice in the accuracy of the states in the early stages of the trapezoidal method to produce better accuracy in the control, whereas more precise solutions converge to an incorrect solution. Convergence independent of the index of the constraints is then proven for one class of problems. The theoretical results are used to explain computational observations.

[1]  Elijah Polak,et al.  Optimization: Algorithms and Consistent Approximations , 1997 .

[2]  J. Betts,et al.  A sparse nonlinear optimization algorithm , 1994 .

[3]  Suresh P. Sethi,et al.  A Survey of the Maximum Principles for Optimal Control Problems with State Constraints , 1995, SIAM Rev..

[4]  Stephen L. Campbell,et al.  Some comments on DAE theory for IRK methods and trajectory optimization , 2000 .

[5]  E. Polak,et al.  Consistent Approximations for Optimal Control Problems Based on Runge--Kutta Integration , 1996 .

[6]  Linda R. Petzold,et al.  Numerical solution of initial-value problems in differential-algebraic equations , 1996, Classics in applied mathematics.

[7]  R. Pytlak.  Numerical Methods for Optimal Control Problems with State Constraints , 1999 .

[8]  A. Dontchev.  Error Estimates for a Discrete Approximation to Constrained Control Problems , 1981 .

[9]  W. Hager,et al.  A new approach to Lipschitz continuity in state constrained optimal control 1 1 This research was su , 1998 .

[10]  M. N. Vrahatis,et al.  Ordinary Differential Equations In Theory and Practice , 1997, IEEE Computational Science and Engineering.

[11]  J. Betts,et al.  Compensating for order variation in mesh refinement for direct transcription methods , 2000 .

[12]  Bruce A. Conway,et al.  Discrete approximations to optimal trajectories using direct transcription and nonlinear programming , 1992 .

[13]  J. Betts,et al.  MESH REFINEMENT IN DIRECT TRANSCRIPTION METHODS FOR OPTIMAL CONTROL , 1998 .

[14]  Elijah Polak,et al.  On the use of consistent approximations in the solution of semi-infinite optimization and optimal control problems , 1993, Math. Program..

[15]  William W. Hager,et al.  Second-Order Runge-Kutta Approximations in Control Constrained Optimal Control , 2000, SIAM J. Numer. Anal..

[16]  R. Pytlak,et al.  Runge–Kutta Based Procedure for the Optimal Control of Differential-Algebraic Equations , 1998 .

[17]  J. Betts,et al.  Application of sparse nonlinear programming to trajectory optimization , 1992 .

[18]  William W. Hager,et al.  The Euler approximation in state constrained optimal control , 2001, Math. Comput..

[19]  J. Betts,et al.  Path-constrained trajectory optimization using sparse sequential quadratic programming , 1991 .

[20]  J. Cullum.  Finite-Dimensional Approximations of State-Constrained Continuous Optimal Control Problems , 1972 .