A Version of NL-SHADE-RSP Algorithm with Midpoint for CEC 2022 Single Objective Bound Constrained Problems

This paper presents an enhanced version of NL-SHADE-RSP, which won CEC'2021 competition on single objective bound-constrained numerical optimization for shifted and rotated shifted functions. The proposed version uses the midpoint of the population to estimate the optimum. The midpoint fitness is also used to introduce a restart trigger. For large populations, the midpoint is calculated after splitting the population into two parts by the k-means algorithm. Other introduced modifications include changing the bound constrain handling method and reducing population size. The performance of the proposed approach is evaluated on the CEC 2022 benchmark for single objective bound-constrained numerical optimization. The results confirm that each proposed modification gradually improves the algorithm's ranking on the benchmark.