Answers are grounded in this paper and cite their sources.
The Fourier Transform Optimizer (FTO) provides a frequency-domain search framework for metaheuristic optimization, but its original spectral operations do not explicitly consider spectral-energy distribution, amplitude–phase-aware information exchange, or multi-resolution spectral manipulation. To address these limitations, this study proposes a Spectral-Domain Multi-Strategy Fourier Transform Optimizer, termed MIFTO. Three complementary mechanisms are introduced. First, Spectral Energy Redistribution (SER) adaptively regulates the relative contribution of spectral components through normalized spectral energy and bounded energy redistribution. Second, Cross-Spectral Alignment (CSA) performs phase-reference-preserving cross-individual spectral recombination while maintaining the conjugate symmetry required for real-valued reconstruction. Third, Multi-Resolution Spectral Search (MRSS) partitions the spectrum into low-, middle-, and high-order subspaces and applies differentiated update mechanisms at different spectral resolutions. The proposed method is evaluated on the CEC2017 benchmark suite at 30 and 50 dimensions and the CEC2022 suite at 10 and 20 dimensions against eleven representative optimization algorithms. Convergence analysis, statistical distributions, population diversity, Wilcoxon signed-rank tests, Friedman mean-rank analysis, ablation experiments, and computational-cost analysis are conducted. The results show that MIFTO achieves favorable optimization accuracy and robustness on the majority of the investigated benchmark functions. MIFTO is further applied to robot path planning in 20 × 20, 40 × 40, and 60 × 60 static grid environments. The corresponding mean path lengths are 28.4494, 60.2801, and 107.5165, respectively, and MIFTO obtains the first Friedman rank in all three scenarios. These results demonstrate the effectiveness of restructuring the internal spectral search process of FTO and indicate the potential of MIFTO for numerical and constrained path-planning problems.
No takes yet. Share an insight, caveat, or question.
Lan et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: