Chaotic Status-Based Optimization: Deterministic Chaos Integration for Congress on Evolutionary Computation-2022 Benchmarks and Engineering Design Problems

Metaheuristic optimizers often suffer from premature convergence and instability in multimodal and constrained search landscapes. Although the recently proposed Status-Based Optimization demonstrates competitive performance, its reliance on stochastic randomization may limit the reliability of convergence. This study introduces a deterministic chaos-driven extension, i.e., Chaotic Status-Based Optimization, in which all pseudo-random components are replaced by ergodic one-dimensional chaotic sequences while preserving the original algorithmic structure and computational complexity. To comprehensively assess the suggested framework, ten one-dimensional chaotic maps are combined with initialization, coefficient generation, selection, and boundary handling. The framework is evaluated on the Congress on Evolutionary Computation 2022 benchmark suite at dimensions 10 and 20, together with three constrained engineering design problems. The Tent-map variant achieved the best overall Friedman mean rank (4.4167 at dimension 20) and demonstrated statistically significant improvements over the baseline method (Wilcoxon rank-sum test (Mann-Whitney U), p < 0.05) with medium-to-large effect sizes. The computational cost remained comparable to that of the original algorithm. All variants attained the best-known solutions for the welded beam (1.6952), pressure vessel (5885.10), and spring design (0.012663) problems. These findings establish deterministic chaos as a reproducible and scalable mechanism for controlling diversity. Future work will extend the framework to multi-objective and large-scale optimization with adaptive chaotic-map selection strategies.