Two Improved Artificial Bee Colony Algorithms

Optimization technology plays a vital role in the engineering design,opera-tions research,production process,information science and so on,it is an effec-tive method for solving complex problems.However,many practical problems are not continuous or have many local minimas,such that it is difficult to solve the complex problems by using the traditional method.As an important random searching method,evolutionary computation has been successfully applied to solve many non-convex,discontinuous and multimodal optimization problems and received extensive attention.Artificial bee colony algorithm is an important kind of swarm intelligence optimization algorithm,which has the advantages of less control parameters,easy implementation,self-organization and so on.However,due to the random-ness for choosing parent individual,artificial bee colony algorithm can't well bal-ance the exploration and exploitation of search,such that it is easily to fall into local optimum and occurs premature convergence phenomenon.Io improve the performance of artificial bee colony algorithm,thesis preserents two improved artificial bee colony algorithm.1.To overcome the weak search ability and imbalance in computing resource allocation,thesis presents an improved artificial bee colony algorithm.To en-hance the neighborhood region search of the elites and global best solution,two new strategies are designed for the employed phase and onlooker phase,respec-tively.Then a local learning strategy is used to the onlooker bees chosen probably to speed up the convergence speed and enhance the global optimization ability.Finally,to balance the exploration and exploitation of search effectively,criss-cross search method(CS)is used to enhance performance of the onlooker bees and global best solution.The proposed algorithm is compared with six excellent meta-heuristic algorithms on 10 classical benchmark functions and 30 CEC2014 benchmark functions with different dimensions.Experimental results show that the proposed algorithm is very competitive.2.To take full use of the optimal individual characteristics and control the search range,thesis presents an artificial bee colony algorithm based on cloud model and neighborhood search.The superior subpopulation of current individ-ual is firstly found and two new strategies are designed for the employed phase and onlooker phase,respectively.Then,the cloud model is employed to real-ize the transform of qualitative concepts and quantitative values and the search range is dynamically adjusted by setting the optimal solution as the expected value and controlling the certainty degree.The proposed algorithm not only can effectively use the individual characteristics of the superior subpopulation to en-hance the search ability of the algorithm in different stages,but also enhances the robustness of the algorithm by introducing the cloud model.The proposed algorithms are compared with several meta-heuristic algo-rithms on 10 classical benchmark functions and 30 CEC2014 benchmark functions with different dimensions.The experimental results show that the proposed al-gorithms improves the global search ability and improve the convergence speed and accuracy.