Novel numerical and computational strategies based on differential evolutionary strategy

Evolutionary algorithm for complex process optimization based on differential evolutionary strategy (DEACOP) that has a similar framework structure of scatter search is proposed. This algorithm not only retained the original algorithm's advantages, but also made improvements in three areas: above all, in order to maintain the diversity of the population, the set RefSet2 is selected from those individuals generated by Latin hypercube uniform sampling, according to minimum Euclidean distance to set RefSet1 is the highest. Furthermore, differential mutation with scaling factor and differential crossover strategy is introduced to replace linear combination method of evolutionary algorithm for complex-process optimization (EACOP). Finally, local search method is adopted to improve the trial solution generated at “go-beyond strategy” stages. The results show that the algorithm is able to use fewer adjustable parameters to complete to search and get feasible mathematical solution.