摘要
针对基于Levenberg-Marquardt方法辨识黄酒发酵过程模型参数时易陷入局部最优,收敛速度慢,很难准确获取具有强泛化能力的模型参数的问题,提出了一种具有莱维飞行机制和柯西变异的蚁狮优化算法(ant lion optimization with Levy flight and Cauchy mutation,LCALO),该算法采用基于莱维飞行和柯西变异来解决这类问题。莱维飞行可以提高算法的全局搜索能力,而柯西变异有助于避免陷入局部最优。结果表明,相比于遗传算法、粒子群算法和蚁狮算法,LCALO的收敛速度快,具有全局搜索能力和局部开发能力好的优点。最后将改进算法应用于黄酒发酵模型的参数辨识,仿真结果证明该算法具有较好的参数辨识能力。
For identifying the model parameters of rice wine fermentation process based on the Levenberg-Marquardt method,it is easy to fall into local optimum and slow to converge.This paper proposed an enhanced ant lion optimization algorithm called LCALO(ant lion optimization with Levy flight and Cauchy mutation,LCALO),which employed Levy flight and Cauchy mutation to overcome this problem.Levy flight could improve the global search ability of the algorithm,and the Cauchy mutation with a long tail helped trapped ant lions escape from local optima.The results showed that compared with the genetic algorithm,the particle swarm algorithm and the ant lion algorithm,the LCALO had the advantages of faster convergence speed,better global search ability,and local development ability.Finally,the improved algorithm was applied to the parameter identification of a rice wine fermentation model.Simulation results proved that the algorithm had good identification ability.
作者
宗原
刘登峰
刘以安
ZONG Yuan;LIU Dengfeng;LIU Yian(School of Internet of Things,Jiangnan University,Wuxi 214122,China;Key Laboratory of Light Industry Process Control Ministry of Education(Jiangnan University),Wuxi 214122,China)
出处
《食品与发酵工业》
CAS
CSCD
北大核心
2021年第2期153-159,共7页
Food and Fermentation Industries
基金
国家自然科学基金青年项目(21706096)
江苏省自然科学基金青年项目(BK20160162)
江苏省博士后科研项目(1601009A)
第62批中国博士后科学基金面上资助(2017M621627)。
关键词
蚁狮优化算法
莱维飞行机制
收敛速度
黄酒发酵
参数辨识
ant lion optimization algorithm
Levi′s flight mechanism
convergence speed
rice wine fermentation
parameter identification
作者简介
第一作者:宗原,硕士研究生;通讯作者:刘登峰,副教授,E-mail:liudf@jiangnan.edu.cn。
