Study On Improved ABC Algorithm And Its Application

Artificial bee colony algorithm,a kind of swarm intelligent optimization algorithm,which has the advantage of less control parameters,easy to implement,simple calculation,etc.,is widely used in many fields.However,the classical artificial bee colony algorithm also has some defects similarly to other intelligent algorithms,such as premature convergence,easy falling into local optimum,and slow convergence in late evolutionary stages.In order to overcome the shortcoming of falling into local optimum,a new algorithm based on Tabu Search was put up.Introduce the Tabu Search algorithm in basic artificial bee colony algorithm to jump out the local search.Through adding a Tabu list-T in the process of bee searching for nectar to store the nectar information which already searched by bees,at the same time,a new searching mechanism obtained.By which avoiding loops in the search path,so as to strengthen the capability of local search.Ten standard test functions,including both unimodal function and multi-peak function,were used to verify the performance of the improved artificial bee colony algorithm based on Tabu list.Set up the numerical simulation experiment,and used the programming language of MATLABR2014a to get the experimental results.In the results,very few of function may not reach to the theoretical optimal value in certain iterations,but it good performed in the whole.The case of bad performance was analyzed in detail.Additionally,the improved artificial bee colony algorithm was compared to the standard artificial bee colony algorithm in the question of 30-dimensions.Analysis shows that the improved artificial bee colony algorithm has better search performance.Furthermore,the improved artificial bee colony algorithm is applied to predict the amount of ground subsidence due to its good performance.The classical integral probability method was introduced in,which has a solid theoretical basis in the prediction of surface subsidence.And based on the present research about integral probability method and surface subsidence,the modified predict model was proposed.After optimizing the parameters of predict model based on the improved artificial bee colony algorithm,a new model of subsidence prediction was optimized obtained.Then,the optimized model was applied to engineering examples.The final results shows that the error range of the predicted subsidence and the actual subsidence amount is(0,0.0026),which fully demonstrates the feasibility and effectiveness of the improved artificial bee colony algorithm in the prediction of surface subsidence,which is of great theoretical significance for guiding on-site construction.Finally,summarized the research content,pointed out the need for further research on improved algorithm,and put forward the development directions of the artificial bee colony algorithm.