Practical genetic algorithms
Multi-Objective Optimization Using Evolutionary Algorithms
Multi-Objective Optimization Using Evolutionary Algorithms
| Hi-index | 0.00 |
This paper presents the comparative performances between a computationally enhanced steady state genetic algorithm (CSGA), standard steady state genetic algorithms (SSGA) and elitist genetic algorithms (EGA) in improving the voltage profile and conducting reactive power planning (RPP) in respect to loss minimisation scheme. Although the SSGA is time efficient but tend to be biased towards exploitation whereby more highly fit chromosomes are accumulated to venture into the next generation of genetic algorithms (GA). Consequently, the SSGA is susceptible to converge prematurely onto a local optimum. The selection elitism incorporated with the variant population spin technique and steady state mechanisms are integrated into the development of the SSGA. The EGA with an appropriate population composition may strike a fair balance between exploitation and exploration in achieving an acceptable optimum solution. Thus, the paper adopts the reading produced by the EGA as the benchmark for drawing any judgment. Any inferior result shall be deemed as a premature convergent. However, the EGA has the flaw of a moderate convergent rate despite of a good on line performance. CSGA is a modified technique of applying exploration factors on the SSGA. In each generation, the CSGA randomly resettles the chromosomes in the neighbourhood of the potential optimum solution while employing extra protective measure to inhibit from trapping onto a local minimum. Identical initial population is supplied to the individual mechanism of GAs for obtaining The proposed CSGA technique has been validated on the IEEE Reliability Test System (IEEE-RTS) and demonstrated profound performances over the SSGA and EGA.