HYBRID GENETIC ALGORITHM / PARTICLE SWARM OPTIMIZATION-FUZZY BASED AUTOMATIC GENERATION CONTROL

摘要

This paper presents a comparative optimization performance and transient performance studies among Hybrid of Genetic Algorithm-Sugeno Fuzzy Logic (GA-SFL), Hybrid of Evolutionary Particle Swarm-Sugeno Fuzzy Logic (EPSO-SFL) and Hybrid of Hybrid Particle Swarm with Constriction factor Approach-Sugeno Fuzzy Logic (HPSOCFA-SFL) in automatic generation control. In a large scale power system, automatic generation control in response to area load changes and abnormal imprecise system operating parameters essentially means very fast minimization of area frequency deviations and mutual tie line power flow deviations of the areas for satisfactory and stable operation of the system. To achieve this, the abovementioned techniques may be used to determine nominal /off-nominal optimal Proportional-Integral-Derivative gains of PID controller employed in each area. HPSOCFA / HPSOCFA-SFL and GA/ GA-SFL based optimal gains result in similar optimal transient responses of frequency and tie line changes. EPSO / EPSO-SFL yields suboptimal responses. The action of Sugeno Fuzzy logic is to adaptively manipulate the GA / EPSO /HPSOCFA based off-line nominal optimal gains, resulting in off-nominal optimal gains and off-nominal optimal transient responses under on-line varying system parameters and load conditions.