Multi-objective optimization for decentralized PI control of the ALSTOM benchmark problem

In this paper, based on the analysis of multi-objective optimization for decentralized PI control of the ALSTOM benchmark problem, deficiencies in optimization results are pointed out. Specially, two deficiencies in Xue´s study are proposed, i.e. performance tests are done when the gasifier is unstable, and the sub-objective functions of multi-objective optimization cannot reflect the change of outputs´ relative integral of absolute error reasonably at each load. Multi-objective optimization algorithm with modifying sub-objective functions are used to eliminate the two deficiencies and applied to choose parameters of decentralized PI control, better dynamic responses performances are achieved, including the decrease of the RIAE sum in pressure disturbance tests, bigger scale of rejecting coal quality change.