Study Of The Improved Ant Colony Algorithm And Its Application On Structural Optimization

Ant colony algorithm is a new and potential intelligent algorithm in optimizational field,with excellent search ability,strong robustness and easy integration of other algorithms.its disadvantage is long searching time and easy to fall into the local optimal iteration.The principle of Ant colony algorithm and its application in structural optimization are reviewed.According to the characteristics of the structural optimization and the shortcomings of the algorithm,the Ant colony algorithm is improved on the routing scheme,the pheromone updating rule and the pheromone controlling strategy.The search process is divided into two stages by combining the pheromone dynamic block-updating strategy and the pheromone overall reduction strategy.A structural optimization model of truss structure is built based on the improved algorithm.Taking ten bar plane truss as an example,the superiority of the improved algorithm and the applicability and effectiveness in structural optimization are tested and verified.The results show that the improved Ant colony algorithm can gain a better optimization result.Compared with the Genetic algorithm,the improvedalgorithm is improved by 2.17%.And compared with the MMAS with the local search,the improved algorithm is improved by nearly 1%,and the number of iterations is only about 9.61%.A structural optimization model of the performance-based seismic optimal design is built taking the structural displacement as the performance index.A Pushover analysis method to analyze the seismic performance of the structure is introduced and verified by comparison of the analysis results and the experimental data of two frame models.Taking a three-story four-bay frame and a nine-story five-bay frame as examples,the application of the improved Ant colony algorithm in performance-based seismic design is realized.The research shows that the Pushover analysis results and test data have good fit,the maximum relative error is 13.94% and 12.17% respectively.the Pushover analysis method and the program is reliable,which can get the structural deformation curve and the elastic-plastic state of each node of the member.In the example of three-story four-bay frame,compared to the previous study of the ACO and GA algorithm,the improved Ant colony algorithm in solving quality are improved by 1.71% and 10.03%,and the computing times is only 22.69% and 13.01% of two algorithms;In the example of nine-story five-bay frame,the improved algorithm in solution quality are improved by 4.25% and 10.96%,and the computing times is only 55.16% and 33.57%.Two applications not only confirm the applicability and validity of the improved Ant colony algorithm in structural optimization,but also provide a feasible design method to optimize the structure.It also shows that the improved algorithm can gain better solution and possess stronger stability.The search efficiency and the solving speed are also improved significantly.The improvement is significant in improving the search efficiency and solving speed of the algorithm,and the two stage search strategy can greatly improve the efficiency of search and optimization by reducing the search scope.