Evolutionary algorithm integrating stress heuristics for truss optimization (Q1826354)

Optimal design of truss structures subjected to stress and displacement constraints is considered. The optimal design problem is reduced to a problem of combinatorial optimization since the optimization variables should be selected from catalogues of the available cross-sections. An evolutionary algorithm is proposed through special genetic operators, which use problem specific knowledge. For example, an intelligent mutation operator randomly decreases the cross-section of under-stressed truss, and randomly increases cross-section of over-stressed truss. To demonstrate the performance of the proposed algorithm, two versions of a benchmark problem of minimum weight 10 bar truss are solved. The first design was subjected to stress constraints, and the second design was subjected to stress and displacement constraints.