基于改进蚁狮优化算法的黄酒发酵过程模型的参数辨识

Model parameter identification of rice wine fermentation process based on an improved ant lion algorithm

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.