Improved Particle Swarm Algorithm And Its Application In Structural Optimization

Nowadays,it is an urgent problem in the field of structural optimization to quickly and accurately meet some engineering requirements such as the best performance and economic engineering demands when solving the problems under the numerous complex constraints.It is a common way to choose an algorithm to do static optimization in structural engineering,but the proportion of dynamic problems is much higher than static problems in practical projects and the relatively few studies existing on the optimal design of structural dynamic problems.The main contents of this paper are the improvement of particle swarm algorithm,the application of the improved particle swarm algorithm on the optimization design of truss structure and on the seismic optimal design of frame structure respectively.1?The particle swarm algorithm is briefly introduced from the basic idea of the algorithm,mathematical description and model analysis.Meanwhile,the general parameter settings and algorithm complexity analysis of particle swarm algorithm are given.2 ? The standard particle swarm optimization is improved from the inertia weight value,the speed update mode and the location updatemethod.Additionaly,the convergence of improved particle swarm algorithm is proved and the general steps of the structural optimal design to the improved particle swarm algorithm is proposed.3?The four classic examples which is the 52 bar plane truss,the 200 bar plane truss,the 25 bar space truss and the 160 bar space truss are used to discuss the performances of the improved particle swarm algorithm in the problems of the discrete optimization of plane truss,the continuous optimization of plane truss,the discrete optimization of space truss and the continuous optimization of space truss respectively.Also,the effectiveness and the applicability of the improved particle swarm algorithm in the optimization design of truss structures are evaluated from the quality,the speed and stability of solution.The results show that compared with the standard particle swarm algorithm,the improved particle swarm algorithm in solving quality is improved by 17.19%,296.42% and 3.35% in the 52 bar plane truss problem,the 200 bar plane truss problem and the 25 bar space truss problem respectively,and the number of iterations is accounted about 36.26% and 78.49% in the 200 bar plane truss problem and the 25 bar space truss problem respectively,and the computing times is accounted about 50% in the 25 bar space truss problem.In addition,the improved particle swarm algorithm in solving quality is improved by 3.09% and the number of iterations is accounted about 1.6% compared with the eACO.4?A nonlinear elastic-plastic Push-over analysis method is introduced.Based on this,a performance-based seismic optimal design model is built and the two optimization examples of the four-bay three-story steel frame and the five-bay nine-story steel frame are solved,which realizing the application of improved particle swarm optimization in seismic dynamic optimization.The results show that compared with the specific best results,the improved particle swarm algorithm in solving quality is improved by 5.28% and 0.44%,and the computing times is accounted about 60.48% and 51.18% in the four-bay three-story steel frame and the five-bay nine-story steel frame respectively.