改进的多目标启发式粒子群算法及其在桁架结构设计中的应用

Improved Multi-objective Heuristic Particle Swarm Optimizer and Its Application in Truss Structural Design

针对工程结构多目标优化设计中出现的约束条件处理能力差、编程复杂,计算效率低且收敛精度差等问题,对启发式粒子群算法(HPSO)进行改进,提出了多目标启发式粒子群算法(MOHPSO),并与多目标粒子群算法(MOPSO)和改进的多目标群搜索算法(IMGSO)进行比较。通过对15杆平面桁架、40杆平面桁架和72杆空间桁架3个经典算例的计算,证明了所提出的MOHPSO算法的有效性。结果表明:MOHPSO算法具有收敛精度高、约束处理能力强、全局最优解选取更合理、非劣解集维护效率高等特点。

According to the common problems in the multi-objective optimization of engineering structures, such as difficulties in dealing with the constraints, the complexity of programming, low calculating efficiency and bad convergence precision, a multi-objective heuristic particle swarm optimizer （MOHPSO） was proposed by improving the heuristic particle swarm optimizer （HPSO）. Then the MOHPSO was compared with multi-objective particle swarm optimizer （MOPSO） and improved multi-objective group search optimizer （IMGSO）. Through three classic examples of 15-bar plane truss, 40-bar plane truss and 72-bar spatial truss structure, the validity of MOHPSO was proved. The results show that the MOHPSO has better convergence accuracy, constraint handling is powerful, the global optimal solution selection is more reasonable and the maintenance efficiency of the non-inferior-solution set is much higher.