Effect of Parametrization on Solving Nonlinear Optimal Control Problems

R. Luus (Canada)


Parametrization; Iterative dynamic programming; Global optimization; Final state constraints; Fed batch reactor


In using iterative dynamic programming to de termine the optimal control policy of nonlinear sys tems, the time interval is divided into P time intervals to convert the optimal control problem into an opti mization problem involving P stages. In each time interval the control is kept constant or is allowed to take on a more complicated structure. The parame trization problem involves choosing P, the nature of control in each time interval, and the length of each time interval. Here we consider piecewise constant control and piecewise linear control parametrizations with constant stage lengths in the first two examples, and variable stage lengths in the third example. In each case we examine how the choice of P affects the resulting optimal control policy, and provide better re sults for these systems than have been reported in the literature.

Important Links:

Go Back