Improved Firefly Algorithm And Application

The swarm intelligence optimization algorithm is one of the areas in recent years,experts have drawn widespread attention, it is a process of interaction and informationforaging insects novel biomimetic simulation algorithms. Firefly algorithm as aswarm intelligence algorithm, also favored by experts and scholars, but also one ofthe areas has been widely studied. Firefly algorithm is based on the process of transferof information between peers in nature or attract fireflies simulated food comes weakluminous fireflies will always move to the luminous intensity of fireflies, on this basis,scholars have developed a Firefly algorithm, using its solve the optimization problem,the basic algorithm for each fireflies firefly is a separate entity, they have their ownperception and search radius, so it is looking for local optima and global optimumsolution have a very good ability, and algorithm can process concurrently, saving timeoptimization. However, there are also a little less than the algorithm itself. Forexample, looking at local optimum and global optimum balance between weakposition in the iterative process of convergence is slow and other issues. To solvethese problems we firefly algorithm for standard two improvements:(1)change the value of ξ. size ξ values in the balance between global and local searchplay a crucial role, so a reasonable value ξ, ξ values means that the size of thealgorithm is to find whether the global optimal solution, we propose a ξ improvedmethods to improve the algorithm balance between local and global search andimprove the accuracy of the algorithm.(2)change the value of γ0. In the standard Firefly algorithm, γ0is a constant changedoes not occur in the optimization process. The valueγ0control algorithm searchmethods, in order to accelerate the convergence rate, we introduce the parameter pvalues γ0be controlled in the early iteration of the algorithm major global search, andin later iterations performed in all local optima Looking for a local optimal solution, so to speed up the convergence rate.The improved algorithm both in terms of the balance between global and localsearch, or on the convergence rate is better than the standard Firefly algorithm. Thispaper then discusses two applications improved Firefly algorithm.(1)Firefly improved algorithm application node localization algorithm. One of thehottest areas of wireless sensor network node localization has been also been studied,and its many applications have penetrated people’s lives. For example, monitoring ofthe harsh environment, target tracking, and application in the medical area. In manyapplications, the most important issue is the sensor node localization problem, thecorrect node positions to make people get the right information, in order to correctlydetermine the next step, we have improved the fireflies in the node localizationalgorithm has a good effect, compared with the traditional location algorithm, theerror is greatly reduced, so that the node positioning accuracy is improved greatly.(2) Job shop scheduling problem has been in the field is also of concern, the use oflimited resources to create a higher interest, but also a matter of concern, however, ishow reasonable scheduling in this respect is the main question. How critical rationaluse of limited machine, complete the processing of all processes as soon as possible,so to minimize the completion time is such a problem in the job shop scheduling.Job-shop scheduling problem is a discrete function of the problem, so we use aparticular encoding algorithm post discrete, play a good role in the scheduling ofmachines and processes. Compared to the standard used in the firefly job shopscheduling, and improved algorithm better, so that the maximum completion time issmaller.