Research On The Fruit Fly Algorithm Based On Traction Mechanism And Its Application

The swarm intelligence algorithm is an important method to solve the optimization problem.This method mainly simulates some biological behavior in nature,and obtains the result of optimization problem in the feasible solution space through progressive iteration.Compared with the traditional optimization algorithms,such as gradient descent,the swarm intelligence optimization algorithm has strong robustness and the implementation is simpler.Swarm intelligence algorithm provides a new method to solve the complex parameter optimize problems;many scholars have been concerned about this method and has used it to solve many optimization problems.As a new type of swarm intelligence algorithm,the fruit fly optimization algorithm(FOA)has many advantages,such as low computation,fast convergence,easily to implement and so on.At present,the FOA has been widely used in the field of scientific research and production.However,in practical application,the FOA also has the defect that it is easy to fall into the local optimal and cannot traverse the entire solution space.These limitations restrict the application range of FOA and has become an urgent problem in the research of FOA.In this dissertation,after researching on the optimization process of FOA,proposing an improved scheme for the defects of the FOA named TFOA(Traction Fruit Fly Optimization Algorithm)and testing the performance of the improved algorithm in both continuous and discrete environment.In addition,by using the characteristics of basic FOA,the improved algorithm is applied to the localization problem of wireless sensor networks;the application scope of FOA in the field of wireless sensor networks is expanded.Experiments shows that compared with the traditional location algorithm,the location algorithm combined with TFOA has higher positioning accuracy and stronger ability to resist ranging errors.Compared with particle swarm optimization algorithm and genetic algorithm,the localization algorithm in this paper has faster convergence speed and better location effect.The main content of this paper is as follows.(1)The defect of FOA is proved theoretically and the convergence of the algorithm is analyzed in detail.Aiming at the defect of the FOA,the traction mechanism and the new search radius calculation method are introduced based on the original algorithm,which solves the problem that the original algorithm cannot traverse the real number domain and is easy to fall into the local optimum.(2)In order to verify the performance of TFOA,The algorithm and other four improved algorithms are used to solve the extreme value of 12 benchmark functions to verify the algorithm's optimization ability in continuous environment.Then,by optimizing the web service composition problem,we further verify the validity of TFOA in discrete environment.(3)TFOA is used to solve the problem of node location in wireless sensor network.After transforming the node localization problem into a constrained optimization problem,using TFOA combined with centroid positioning algorithm and optimal concussion strategy to locate wireless sensor network nodes.In the simulation environment,we compared the TFOA with traditional location algorithm based on ranging,particle swarm optimization algorithm(PSO)and genetic algorithm(GA),to verify the localization performance of the TFOA.