An adaptive evolution strategy for constrained optimisation problems in engineering design

Nature inspired algorithms (NIAs) are extensively employed to solve non-linear non-convex constrained optimisation problems (COPs) since the traditional methods show poor global convergence performance. Engineering design domain involves highly complex COPs studied extensively by various NIAs. Since the global optimum for almost all benchmark problems are already identified, improving the solution is, in general, not possible. However, an improvement in terms of number of objective function evaluations (FES) and reliability is still likely. Inspired by the work by Kusakci and Can (2013), this paper proposes an evolution strategy (ES) with a CMA-like mutation operator and a ranking-based constraint-handling method. During the design stage, a set of preliminary experiments conducted on the benchmark set, and some modifications are made to improve the performance of the algorithm. The two competing mutation strategies, adaptive initialisation of the population size, and ranking-based constrained-handling strategy contribute effectively to the aim of performance improvement. The results indicate that the modified algorithm is able to find the global optimum in less FES and with higher reliability when compared with the benchmarked methods.