A hybrid pattern search method for solving unconstrained optimization problems

In solving engineering optimization problems, the current Evolutionary Programming (EP) has slow convergence rates on most problems, and if there is more than one local optimum in the problem, the obtained optimal solution may not necessarily be the global optimum. This paper describes a new approach for solving unconstrained optimization problems with either discrete or continuous design variables. The proposed approach is a pattern search method that is based on univariate search hybridized with the Shaking Optimization Algorithm “SOA”. The computational analysis shows that, for the selected benchmark problems, the proposed approach is a powerful search and optimization technique that may yield better solutions to engineering problems than those obtained using current algorithms for both the solution efficiency and the number of iterations.