Applied Research Of Intelligent Optimization Algorithm

With the development of science and technology,optimization problem plays an increasingly important role in all walks of life,such as engineering architecture design,image processing,economic load distribution,classification and clustering.However,the optimization problems in these fields are usually multi-modal,multi-dimensional,non differentiable and constrained.Traditional mathematical optimization methods can not solve these problems.In recent years,meta heuristic intelligent optimization algorithm based on bionics has been paid attention and research by more and more researchers.Because of its advantages of simple operation and high efficiency,it has been widely used in many fields.Among the many meta heuristic algorithms,firefly algorithm and artificial bee colony algorithm show better performance in many optimization problems.This paper focuses on the firefly algorithm and artificial bee colony algorithm,analyzes the advantages and disadvantages of the two algorithms,and improves the algorithm from the perspective of dimension update strategy,update formula and balance global and local search performance,and applies the improved algorithm to practical application.The main work of this paper is as follows:(1)Although the traditional firefly algorithm has good performance in some problems,it has some problems such as slow convergence speed and easy to fall into local optimum in complex problems.To solve these problems,this paper improves the update strategy of traditional firefly algorithm.First of all,the concept of elite group is introduced.The optimal individual updating adopts elite group leading and finite dimension updating strategy to maximize the quality of the optimal individual.The finite dimension updating strategy improves the accuracy of searching the optimal individual.The firefly updating adopts the optimal individual leading and finite dimensional updating strategy,and the firefly individual update is led by the optimal individual,which speeds up the optimization speed of the algorithm.The finite dimension updating strategy makes fireflies search more finely,balancing the search speed and precision of the algorithm.Then,the standard test function results of the improved algorithm are compared with that of several classical meta heuristic algorithms,and the experimental results prove the effectiveness of the improved algorithm.Finally,several algorithms are applied to image segmentation,and the experimental results verify the effectiveness of the improved algorithm in image segmentation.(2)Traditional artificial bee colony algorithm has strong global search ability,but it has some problems such as slow convergence speed and poor precision.In order to solve these problems,this paper improves the update strategy of the traditional artificial bee colony algorithm in the stage of employed bee and onlooker bee.Firstly,the concept of elite group is introduced.In the stage of employed bee,the elite learning strategy is adopted,and the employed bee randomly selects an individual in the elite group to learn,which improves the optimization speed of the algorithm.The chaotic operator is introduced to improve the distribution of random numbers,which improves the local search ability of the algorithm,and balances the global and local search ability of the algorithm.In the onlooker bee stage,the selection and difference dimension updating strategies are adopted.When the onlooker bee is the optimal individual,it learns from the suboptimal individual,which makes the onlooker bee search around the high-quality honey source as much as possible,which improves the optimization speed of the algorithm,and the remaining onlooker bees learn from the optimal individual.The difference dimension updating strategy is used to find all dimensions with different values to update at the same time,which improves the optimization speed of the algorithm.Then,the the standard test function results of the improved algorithm are compared with that of several classical meta heuristic algorithms,and the experimental results prove the effectiveness of the improved algorithm.Finally,several algorithms are applied to the image segmentation of robot vision system,and the experimental results verify the effectiveness of the improved algorithm in image segmentation of robot vision system.