Stochastic Global Optimization Method for Solving Constrained Engineering Design Optimization Problems

This work presents a stochastic global optimization (SGO) approach, which integrates an artificial immune algorithm and a particle swarm optimization (AIA-PSO) approach. to solve constrained engineering design optimization problems (e.g., tension/compression string design and pressure vessel design problems), the proposed AIA-PSO algorithm uses a penalty function method to transform a constrained engineering design optimization problem to an unconstrained optimization problem. Based on an external AIA approach, the constriction coefficient, cognitive parameter, social parameter, penalty parameter and mutation probability of an internal PSO algorithm are optimized. Constrained engineering design optimization problems are then solved using the internal PSO algorithm. Moreover, numerical results obtained using the proposed AIA-PSO algorithm is compared with those of published individual genetic algorithm (GA) with AIA methods and hybrid algorithms. Experimental results indicate that the optimum parameter settings of the internal PSO algorithm can be obtained using the external AIA approach. Also, the proposed AIA-PSO algorithm performs significantly better than those of some published individual GA with AIA approaches and hybrid algorithms for solving the pressure vessel design problem. Therefore, the proposed AIA-PSO algorithm can be considered as a promising SGO approach for solving constrained engineering design optimization problems.