|
自动化学报 2007
Multi-objective Control Problem of Bilinear Systems
|
Abstract:
The control problem of bilinear systems with multiple objectives is studied.First,the multi-objective control problem is converted into a single objective optimal control problem using the utility function technology.The utility function is a nonlinear function of multiple quadratic performance indices,and therefore it is non-separable in the sense of dynamic programming.Then,to overcome this difficulty,a two-level optimal control algorithm is proposed.At the lower level,the formulated auxiliary Lagrangian problem is of a parametric bilinear-quadratic structure and it is solved by dynamic programming.Finally,the weighting vector in the auxiliary Lagrangian problem is adjusted by the upper level iteratively.This two-level process repeats until an optimal condition is satisfied.