模拟退火粒子群算法在化工过程综合中的应用  

Application of Simulated Annealing Particle Swarm Algorithm in Chemical Process Synthesis

作　　者：李学文[1] 金思毅[1] 陶少辉[1] 

机构地区：[1]青岛科技大学化工学院,山东青岛266042

出　　处：《青岛科技大学学报（自然科学版）》2011年第1期80-84,共5页Journal of Qingdao University of Science and Technology:Natural Science Edition

摘　　要：将一种模拟退火粒子群算法应用于化工过程综合。由于粒子群算法后期进化速度变慢,并易陷入局部极值点,本研究将模拟退火思想应用到粒子群算法中,通过Me-tropolis准则提高粒子跳出局部极值的能力,并在降温过程中加快了算法后期的进化速度。模拟退火粒子群算法在换热器设计实例和反应器网络综合问题中的应用结果表明,该算法的性能较粒子群算法有了较大改善,同时也表明该算法用于求解化工过程综合问题是可行和有效的。A hybrid particle swarm optimizaiton algorithm（PSO） based on Simulated Annealing（SA） is introduced for solving chemical process synthesis problems.Because the PSO algorithm has a slow convergence speed and is easy to drop into a local minimum,the SA is integrated into the PSO to effectively avoid the local minimum by Metropolis rule and improve evolutionary rate during temperature decreasing procedures.The simulated annealing particle swarm algorithm is applied to a heat exchanger design example and a reactor network synthesis problem.The calculation results have shown that the performance of the hybrid algorithm is better than that of the particle swarm optimization,and the hybrid algorithm is effective and feasible for solving chemical process synthesis problems.

关 键 词：模拟退火粒子群算法 粒子群算法 化工过程综合 反应器网络综合 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]