ITSA: An Adaptive Tree Seed Algorithm Based on Differential Evolution with Tree Migration and Seed Intelligent Generation
Jiansheng Huang, Ruixiang Song, Lingna Li, Yvzhu Ji · 2024
An Adaptive Tree Seed Algorithm based on Differential Evolution with Tree Migration and Seed Intelligent Generation (Improvements-Tree Seed Optimization, ITSA) is proposed in this paper to address the issues with the Tree Seed Algorithm (TSA). The Improved Tree Seed Algorithm (ITSA) introduces significant enhancements to overcome limitations in the original Tree-Seed Algorithm (TSA). By incorporating a Sine map initialization, ITSA ensures a more balanced and effective population distribution compared to TSA's random mapping. Moreover, the introduction of velocity vectors in seed updates enhances population diversity and convergence, particularly in later optimization stages. The integration with Differential Evolution (DE) further bolsters global exploration capabilities. These modifications collectively address challenges such as premature convergence and susceptibility to local optima in TSA. Comparative experiments demonstrate that ITSA outperforms TSA, EST-TSA, and other benchmark algorithms, showcasing its superior optimization prowess, robust convergence speed, and enhanced population diversity.