Research On Improved Artificial Bee Colony Algorithm And Its Application

Optimization problems widely exist in daily life,for example,we want to travel to multiple cities,and what kind of route can make the shortest distance and the lowest cost.Traditional optimization methods can be used to solve simple optimization problems effectively.However,as the problems gets more complicated,the traditional optimization methods are often limited in solving optimization problems.Intelligent optimization algorithm has the advantages of parallel computation and independent of gradient information,so it has gradually entered the view of researchers.Artificial bee colony algorithm is an intelligent optimization algorithm based on the foraging behavior of bee colonies.It has the advantages of strong exploration ability,fast convergence speed and so on,providing new ideas for solving practical problems in science and life.For the lacking of exploitation ability of artificial bee colony algorithm,this thesis makes some improvements,which can solve nonlinear and multi-modal optimization problems effectively.And the improved algorithm is applied for solving portfolio optimization problems.The research contents of this thesis are as follows:1.An improved artificial bee colony algorithm is proposed.First,the combination of Pareto dominance criterion and feasibility rule are used to deal with the constraints,so as to select the better individual more reasonably.On the evolutionary stage,the evolution strategy is adaptively selected according to the current environment,and the search space is precisely explored by using the real-time dynamic characteristics.The algorithm is tested on two benchmark function sets,and the experimental results fully reflect the superiority of the algorithm.Finally,the improved algorithm is used to solve the portfolio optimization problems,which meets the needs of investors to invest rationally.2.An improved multi-objective artificial bee colony algorithm is proposed.First of all,random individuals and optimal individuals are used to guide the search of the population in both the employed bee and onlooker bee stages,so as to improve the algorithm ability to explore the better region.Then,a new archive set updating strategy is proposed to preserve the non-dominant solutions with better convergence and distribution,and which allots more computing resources to the better individuals.Finally,the improved algorithm is tested on two benchmark function sets.And the test results show that the proposed algorithm can solve the multi-objective optimization problem effectively.