Research On Fibonacci Tree Optimization Algorithm For Multimode Optimization And Its Application

In the field of optimization,the optimization problem is often a multi-modal optimization problem,which has many global optimal solutions and local optimal solutions.In the optimization process,the traditional intelligent optimization algorithm is often difficult to obtain multiple optimal solutions and valuable local solutions,therefore,the multimodal optimization algorithm has become an important research direction in the field of optimization.In recent years,although the multimodal optimization algorithm has achieved good results,there are some deficiencies.Therefore,based on the Fibonacci tree optimization algorithm(Fibonacci Tree Optimization,FTO),by introducing the distance parameter in the algorithm,this thesis puts forward a method suitable for multimodal optimization FTO algorithm.The algorithm retains the basic Fibonacci tree optimization algorithm global optimization characteristics,in the optimization process,process the wholly exploration and local optimization alternately,store the optimization process information through the Fibonacci tree structure,to realize the sharing and memory of information search,and use distance parameter to achieve the multi-mode optimization ability of the algorithm,finally gives the algorithm strong wholly optimality and multimode optimization ability.This thesis first introduces the basic idea and algorithm structure of Fibonacci tree optimization algorithm,and gives the realization process of the basic FTO algorithm.Secondly,the thesis analyzes the multi model of the basic FTO algorithm.Through the analysis,it is found that the optimization principle of the basic FTO algorithm determines that it has a strong global optimization ability while lacks multimode capability.For this reason,this thesis discusses in detail and finds out that adding the distance parameter to the algorithm can solve the problem of weak multi model optimization ability,and can give the implementation of the FTO algorithm after adding the distance parameter.Then,verifies the characteristics and effectiveness of the proposed algorithm by the experiments.In the experimental part of this thesis,firstly,the characteristics of the algorithm are analyzed and discussed by using 10 kinds of test functions with different characteristics.Experimental results show that the improved FTO algorithm has strong optimization ability of multimode;discusses the effects of size of the distance parameters and algorithm structure to the optimize performance;proves that the improved FTO algorithm is higher in the universality and stability in multi modal optimization problem.Then compares the improved FTO algorithm with different multimodal optimization algorithms,the experimental results show that the improved FTO algorithm is a kind of feasible and effective multimode optimization algorithm,and is not affected by the testing problems,has strong stability of optimization,high success rate and better convergence precision.Finally,applies the improved FTO algorithm to optimize the design of a practical optimization problem in power transformer,the optimization results show that the algorithm can obtain a plurality of groups which can meet the design requirements of the program,and each result is better than the original artificial scheme,the time required is far lower than that of the original scheme.This verifies that the algorithm has obvious advantages and application prospect in practical engineering optimization problems.