Adaptive & parallel simulated annealing genetic algorithm based on cloud model

Due to the “premature” phenomenon and poor local search ability of genetic algorithm, an improved genetic algorithm, adaptive and parallel simulated annealing genetic algorithm based on cloud model (PCASAGA), is proposed in this paper. This algorithm integrates cloud model, multi-populations optimization mechanism, parallel techniques, simulated annealing algorithm and adaptive mechanism. It applies qualitative reasoning technology - cloud model to the regulation of crossover probability and mutation probability to improve the adaptive ability. The use of new multi-threading building blocks TBB parallel technology has greatly enhanced the operational efficiency of the algorithm. simulation results illustrate that PCASAGA has better convergence speed and optimal results than original genetic algorithm, and takes full advantage of the current multi-core resources of computers.