Aiming at the disadvantages of poor development ability and imbalance between exploration and development of Artificial Bee Colony Algorithm,multi-strategy dynamic clustering artificial bee colony algorithm was proposed.Firstly,the population is divided into multiple subgroups by using fitness ranking and random grouping strategies,so that it can search different regions at the same time.Secondly,in the search process,the dynamic subgroup strategy is combined to update the individuals in the excellent subgroup according to the fitness level.Different ordinary subgroups compete to produce offspring according to the success rate of their search strategy,and the population number of each ordinary subgroup is dynamically adjusted.Finally,the multi-strategy selection mechanism is used to design different search strategies for each subgroup.By strengthening the guiding role of excellent subgroups,the diversity of common subgroups in exploration and development is increased,and the balance between exploration and development of the algorithm is achieved.The simulation results of 9 benchmark test functions show that compared with other improved algorithms,the proposed algorithm has higher convergence accuracy and stronger searching ability.
维普
万方数据