Multi-objective Structural Optimization Of Truss Structures Based On Group Search Optimizer

Structural optimization is a research focus in recent years. As traditional optimization methods can not solve it well, more and more intelligent optimization algorithms were introduced into the field, mainly a genetic algorithm and particle swarm optimization. In this paper a new intelligent algorithm-Group Search Optimizer (GSO) was introduced for structural multi-objective optimization. GSO has simple program, fast convergence speed, good computational efficiency. For high-dimensional variable problem group search optimizer has obvious advantages.Based on GSO and Multi-objective Group Search Optimizer (MGSO), this paper analyzed the performance when GSO applied on multi-objective optimization. In terms of some feasible solutions from MGSO, the multi-objective application of GSO is practicable. In order to enhance the efficiency of MGSO, focused on the disadvantages of MGSO, a GSO-based optimization method for constrained multi-objective optimization problems (CMOPGSO) is proposed, together with its flow chart. At last, the algorithm is applied on several structural multi-objective optimal of the truss.CMOPGSO comparing with MGSO propose three mainly improved aspects. Firstly, the concept of transition-feasible region is introduced to handle the constrained conditions, and a feasible solution "filter" is set to make all the final Pareto solutions are feasible solution. Secondly, Dealer’s principle was employed to build the non-dominated set, which can reduce the number of contrast effectively. At last, in order to avoid getting into the local optimum untimely and improve the solution precision, the selection of the producer was achieved based on the combination of Tabu Search algorithm and Crowded Mechanism. After verification, this method can effectively improve the distribution of the solutions.CMOPGSO will be applied to the practical optimization of engineering structures. In this paper,3static multi-objective optimization examples of the truss and4dynamic multi-objective optimization examples of the truss are carried out, and the results of each example are analyzed in detail. The result shows that CMOPGSO can converge to many well non-dominated solutions and the distribution of the solutions have a well guarantee. The superiority of CMOPGSO is much more obvious, especially in the situation of complicated calculation and much more constraints.