Industrial & Engineering Chemistry Research, Vol.36, No.6, 2279-2286, 1997
Extension of Iterative Dynamic-Programming to Multiobjective Optimal-Control Problems
Multiobjective optimal control problems are converted into a single-objective optimal control problem including constraints. As a result, while the threshold level of constraints is specified, such single-objective optimal control problems can be solved by dynamic programming. Interactive programming procedures are usually required to find a satisfactory solution to multiobjective optimal control problems. In order to obtain the trade-off information in the interactive programming, the multiplier updating method incorporated in iterative dynamic programming is introduced to estimate the Lagrange multipliers of the optimal control problems. Two chemical processes are used to investigate the effect of using Lagrange multiplier updating in iterative dynamic programming.
Keywords:BATCH POLYMERIZATION REACTOR;TIME-OPTIMAL-CONTROL;OPTIMAL TEMPERATURE;OPTIMIZATION;COMPUTATION;POLICY