摘要

In this paper a novel evolutionary global search strategy called Colonial Competitive Algorithm (CCA) is utilized to determine an optimal decoupling matrix and also to tune a decentralized PID controller for the multi-input multi-output evaporator plant. Recently, introduced CCA has proven its excellent capabilities, such as faster convergence and better global minimum achievement. First, parameters of the evaporator plant are identified based on prediction error model algorithm. The decoupling matrix for the identified plant transfer matrix is obtained through shrinking the Gershgorin bands by minimization of an appropriate cost function in frequency domain, to make the compensated plant diagonally dominant. Then a decentralized PID controller for the MIMO evaporator is designed based on CCA. The simulation results verify the superiority of the proposed method to the conventional approaches and also to the centralized MIMO PID controller designed with CCA, along With its less complexity.

  • 出版日期2009-10