Novel Adaptive Genetic Algorithm for Reactive Power Optimization of Power System

摘要

In order to avoid the premature convergence and improve convergence rate, a novel adaptive genetic algorithm for reactive power optimization is discussed in detail. In reproduction operator, the method of retaining optimal individual is used to ensure the convergence and at the same time, the competition method is also adopted to keep the better dispersal of all individuals. In Mutation operator, the mutation probability P_m is improved based on adaptive genetic algorithm. When fitness of individuals in the population tends to be identical, P_m can be adjusted to make bigger, and the local convergence can be avoided greatly. The algorithm has been applied to IEEE 30-bus testing system. The test shows that this algorithm is feasible and practical.