Research On Light Ray Optimization Algorithm Based On Parallel Computing

With the development and progress of society,there are constantly complex and changeable practical engineering problems in the production and life.In the face of these problems,mathematicians and engineering scientists usually establish mathematical models,which are transformed into optimization problems to solve.However,these problems often have the characteristics of high dimension,large amount of data and long calculation time.By using traditional method and numerical algorithm often fail to meet the increasing demands of People.Therefore,scientists have proposed some intelligent optimization algorithms based on the optimization phenomenon in nature.Such algorithm has strong stability,easy to understand and suitable for parallel computing etc.Therefore,the research of intelligent optimization algorithm is more and more meaningful.Light Ray Optimization Algorithm is an intelligent optimization algorithm based on Fermat's principle and light propagation characteristics.This algorithm completely simulates the light's propagation in different media.This paper introduces the idea of parallel computing to solve the problem that the algorithm is sensitive to the initial direction.The algorithm is improved and optimized,and the convergence speed of the algorithm is improved.This paper mainly completes the following three tasks:Firstly,we use the statistical method to analyze the initial direction and the number of ite rations(ie,the optimization time)of the Light Ray Optimization Algorithm.It is proved that the number of iterations of the Light Ray Optimization Algorithm in the four quadrants corresponds to a normal distribution and the difference between the corresponding iterations is significant,It provides theoretical support for the parallel optimization algorithm to improve search efficiency..Secondly,the propose Light Ray Optimization Algorithm based on parallel computing,presents a design scheme and specific iterative steps of Parallel Light Ray Optimization Algorithm.According to the spherical function,the algorithm generates four directions to search the optimal point at the same time and analyzes the performance of the obtained data.Thirdly,the numerical experiments of five typical intelligent algorithm test functions are carried out by using the Parallel Light Ray Optimization Algorithm.The experiments show that the Parallel Light Ray Optimization Algorithm is more efficiently than the Serial Light Ray Optimization Algorithm in dealing with the optimization problem with many minimal points.