文献类型：期刊文献

中文题名：改进量子蝙蝠算法的研究及应用

英文题名：Research and application of improved quantum-behaved bat algorithm

作者：周志垚[1];孙自强[1]

机构：[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

年份：2019

卷号：40

期号：1

起止页码：84

中文期刊名：计算机工程与设计

外文期刊名：Computer Engineering and Design

收录：CSTPCD;;北大核心:【北大核心2017】；

基金：中央高校基本科研业务费重点科研基地创新基金项目(222201717006)

语种：中文

中文关键词：量子蝙蝠;自然选择;优化因子;求解精度;管式反应器

外文关键词：quantum-behaved bat;natural selection;optimization factor;accuracy;tubular reactor

摘要：针对量子蝙蝠算法求解精度低、易陷入局部最优等缺点,提出一种改进型量子蝙蝠算法。引入自然选择,在每次迭代过程中对整个种群适应度值进行排序,用部分较好个体的位置替换部分较差个体的位置,保存个体历史最优适应度函数值。针对蝙蝠算法的频率引入优化因子,使蝙蝠在迭代初期发出较高频率进行全局搜索猎物,在迭代后期降低频率,提高局部搜索能力。对4个标准测试函数进行测试,测试结果表明,改进型量子蝙蝠算法有更好的收敛速度和求解精度。将改进的算法应用于典型化工过程的动态优化问题中,优化结果接近于最优值,性能良好,结果验证了该算法的有效性。
To solve the problem that quantum-behaved bat algorithm has low convergence accuracy and it easily falls into local optimization,an improved quantum-behaved bat algorithm was proposed.Natural selection was introduced.In each iteration,the fitness values of the whole population were sorted,and the positions of some worse individuals were replaced by the better individuals to preserve the individual historical optimal fitness values.Optimization factor was introduced in the frequency of bat algorithm to make bats have higher frequency in the initial iteration for fine global search capability.The local search capability was improved by reducing frequency in the end of iteration.The results of simulating 4 standard benchmark functions show that the improved quantum-behaved bat algorithm has better convergence speed and accuracy.The proposed algorithm was applied to the dynamic optimization of a typical chemical process.The result is close to the optimal value and the performance is good,which verifies the validity of the proposed algorithm.