APPLICATION OF ITERATIVE DYNAMIC-PROGRAMMING TO OPTIMAL-CONTROL OF NONSEPARABLE PROBLEMS

Convergence properties of iterative dynamic programming are examined w ith respect to solving non-separable optimal control problems. As sugg ested by LUUS and TASSONE [1], the best values available from the prev ious iteration are used for those variables which are required from up stream. As iterations continue, the values tend to converge to the opt imal values. Although the convergence is not monotonic, it is neverthe less fast. Two non-separable optimisation problems are used to test th e viability of this approach. It is found that iterative dynamic progr amming (IDP) is considerably more efficient than direct search optimis ation when the number of stages is large. To solve a 100-stage non-sep arable optimisation problem with 3 state variables and 3 control varia bles requires less than one minute of computation time with IDP.