An Improved Particle Swarm Optimization Algorithm For Optimal Operation Of Cascade Reservoirs

Yong Peng
2012-01-01
Abstract:For solving the optimal operation problems of cascade reservoirs, an improved particle swarm optimization (IPSO) algorithm is proposed. To increase the searching efficiency, the IPSO introduces crossover operator and mutation operator, which are similar to the operators in genetic algorithm. Crossover is that the positions of the particles in solution space are making arithmetical crossover at a certain probability, and mutation is that the particles randomly make some one-dimensional component of velocity vector change into zero at a certain probability. To increase the convergence speed, the initial particle swarm generates randomly under certain conditions, and penalty function method is employed to deal with the boundary conditions and the inequality constraints. Taking cascade reservoirs for example to compare the IPSO with dynamic programming algorithm and conventional PSO, the results show that the IPSO has faster calculation speed and satisfying optimal operation results.
What problem does this paper attempt to address?