Implementation of Chaotic Reverse Slime Mould Algorithm Based on the Dandelion Optimizer

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Machine Learning

Background:

• Slime Mould Algorithm (SMA) and Dandelion Optimizer are metaheuristic algorithms.
• Existing algorithms may suffer from slow convergence and local optima entrapment.
• Parameter optimization for machine learning models like Extreme Learning Machine (ELM) is crucial for performance.

Purpose of the Study:

• To develop a novel hybrid optimization algorithm with improved convergence speed and global search capabilities.
• To enhance the performance of the Slime Mould Algorithm by incorporating chaotic mapping, Brownian motion, Lévy flight, and specular reflection learning.
• To apply the improved algorithm for optimizing Extreme Learning Machine (ELM) parameters in power load forecasting.

Main Methods:

• Hybridization of Slime Mould Algorithm (SMA) with a mixed Dandelion Optimizer.
• Integration of Bernoulli chaotic mapping for population diversity.
• Incorporation of Brownian motion and Lévy flight for enhanced global and local search.
• Application of specular reflection learning to avoid local optima in later iterations.

Main Results:

• The proposed hybrid algorithm demonstrated superior convergence speed and precision on standard test functions compared to existing methods.
• Experimental validation showed significant improvements in optimization performance.
• The optimized Extreme Learning Machine (ELM) model achieved effective power load forecasting.

Conclusions:

• The developed hybrid algorithm effectively addresses the limitations of traditional optimization techniques.
• The enhanced SMA variant offers robust performance for complex optimization tasks.
• The method proves effective for practical engineering applications, specifically in power load forecasting.