A hybrid global optimization algorithm based on particle swarm optimization and hill-climbing search and its engineering application

作者：Chen GC^*; Yu JS

来源：Dynamics of Continuous Discrete and Impulsive Systems: Series B; Applications and Algorithms , 2005, 1: 145-150.

A hybrid global optimization algorithm based on particle swarm optimization (PSO) and hill-climbing search (HC) is proposed. The new method assumes that some stochastic particles are placed in an initial neighborhood of the best particle P, of PSO since the first iteration of PSO. Then the best individual of the stochastic particles is found and noted as P-gn. if P-gn is better than P-g P-g is replaced with P-gn and the next iteration of PSO goes on. If P-g, is not better than P-g, the neighborhood width of the best particle P-g is broadened, the stochastic particles are placed renewedly and the next iteration of PSO does not go on until the best individual P-g, of stochastic particle is better than Pg or the neighborhood width exceeds the scheduled width. This new algorithm is called hill-climbing particle swarm optimization algorithm with variable width's neighborhood (vwnHCPSO). Then, vwnHCPSO, hill-climbing particle swarm optimization algorithm with invariable width's neighborhood (HCPSO) and PSO are employed to resolve five widely used test function's optimization problems. Results show that vwnHCPSO has greater efficiency, better optimization performance and more advantages in many aspects than HCPSO and PSO. Next, vwnHCPSO is employed to train artificial neural network (NN) to construct a practical soft-sensor of light diesel oil flash point of main fractionator of fluid catalytic cracking unit (FCCU). The obtained results indicate and the new method proposed by this paper is feasible and effective in soft-sensor of light diesel oil flash point.

• 出版日期2005
• 单位东华大学