多策略果蝇优化算法及其应用  

Multi-strategy fruit fly optimization algorithm and its application

作　　者：钟伟民[1] 牛进伟 梁毅[1] 孔祥东[1] 钱锋[1] 

机构地区：[1]华东理工大学化工过程先进控制和优化技术教育部重点实验室,上海200237

出　　处：《化工学报》2015年第12期4888-4894,共7页CIESC Journal

基　　金：国家科技支撑计划课题项目(2015BAF22B02);国家自然科学基金项目(21376077;61422303);上海市人才发展基金项目;上海市重点学科建设项目(B504)~~

摘　　要：针对果蝇算法容易陷入局部极值、收敛速度慢和收敛精度低的问题,基于粒子群优化算法中社会认知因子和差分演化算法的变异算子,提出了一种多策略果蝇优化算法(SFOA)。对于味道浓度值劣于平均味道浓度的个体,采用社会认知变异因子产生下一代个体,加快收敛速度。对于味道浓度值优于平均味道浓度的个体,引入差分向量,提高算法跳出局部极值的能力。经过8个测试函数的仿真实验对比,SFOA具有更好的全局搜索能力、更快的收敛速度和更高的收敛精度。最后,将改进后的果蝇算法运用到GE气化炉操作优化中,以有效合成气产率最大化为优化目标,氧煤比和水煤浆浓度为决策变量,结果表明,SFOA能够快速找到最优值,证明了多策略果蝇优化算法的有效性。For the demerits of fruit fly optimization algorithm（FOA）, such as easily falling into local optimum, slow convergence rate and low convergence precision, an improved FOA is proposed called multi-strategy fruit fly optimization algorithm. It is based on social cognitive part of particle swarm optimization algorithm（PSO） and the differential vector of differential evolution（DE）. For those individuals whose smell concentration values are worse than the average concentration, using social cognition operator generates next generation to accelerate the convergence rate. For others, introducing the differential vector improves the ability to jump out the local optimum. Through the simulation on eight benchmarks and comparison with other algorithms, the experimental results show that SFOA has better global search capability, fast convergence and higher convergence precision. Finally the SFOA is also applied to optimize the operation of a GE gasification process, which is to maximize the syngas yield with two decision variables, i.e., oxygen-coal ratio and coal concentration. The results show that SFOA can quickly find the optimal value, which demonstrates the effectiveness of SFOA.

关 键 词：果蝇算法 味道浓度判定值 粒子群算法 差分进化 煤气化 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]