This paper presents two mechanisms to improve Ant Colony Optimization (ACO) for design of truss structures. The Sub-Optimization Mechanism causes the reduction in the search space, the size of the trail matrix and the decision vector; consequently, the optimization process is performed in smaller time. The Scatter Memory Mechanism introduces a matrix for saving the best scatter order and enables the ACO-based algorithms to reduce the number of ants and intensity of the analyses. The proposed mechanisms applied to ACS and ASrank are tested on two benchmark trusses from literature. The results for different algorithms demonstrate the efficiency of the proposed mechanisms.

No abstract