Application of evolutionary computation for automatic generation control in restructured power system

摘要

In this paper, the four-area multi-units automatic generation control is studied in restructured power system. There are various types of ancillary services present in power system. One of these ancillary services is load following with frequency control which comes broadly under Automatic Generation Control in restructured power system. The prime objective of the paper is to introduce some novel evolutionary computation based techniques applied independently to obtain optimal gain parameters for optimal transient performances under various system operating conditions. Computational results and transient performances are compared to finally determine the best optimization technique for this problem. A novel particle swarm based algorithm namely, modified chaotic ant swarm optimization (MCASO) and real coded GA (RGA) prove to be competitively the best. A conventional PSO and binary coded GA are the other two techniques, yielding sub-optimal performances. A DISCO can contract individually and multilaterally with a GENCO for power and these transactions are done under the supervision of the ISO. In this paper, the concept of DISCO participation matrix is used to simulate the bilateral contracts in the four-area diagram. The computed values of generators' participations and tie-line power exchanges match with the corresponding actual values obtained by MATLAB-SIMULINK. Optimal transient responses are determined by substituting the optimal gains in the MATLAB-SIMULINK based four-area multi-units diagram.