Research and application of Distributed Parallel Genetic Algorithm Based on PC Cluster

A distributed parallel genetic algorithm based on PC cluster was proposed, aiming at the disadvantage of traditional genetic algorithm, such as the bad searching quality and long computation time. It adopts the improved genetic simulated annealing algorithm and distributed parallel technique MPI, to implement the distributed computing on PC cluster. The algorithm uses the individual migration strategy to collaboratively optimize every process. The dynamic demes are adopted to balance the CPU load. Computing efficiency is introduced to judge the computing load state. A singular function and an IEEE 14 test system in power system are tested. The results reveal that the algorithm has a good stable searching capacity and good parallel efficiency. Hence the proposed method should have a bright future in the application.