Review of chaotic mapping enabled nature-inspired algorithms.

Area of Science:

• Computer Engineering
• Optimization Algorithms
• Computational Intelligence

Background:

• Pseudo-random numbers are crucial in computer engineering for simulations and optimization.
• Chaotic maps offer an alternative to traditional pseudo-random number generators.
• Chaotic improved optimization algorithms have been reported to yield better performance in literature.

Purpose of the Study:

• To investigate the effectiveness of chaotic maps in enhancing optimization algorithms.
• To analyze the impact of different chaotic maps and improvement strategies on algorithm performance.
• To provide insights into the selection of appropriate chaotic maps for optimization tasks.

Main Methods:

• Collected and analyzed 19 classical chaotic maps generating pseudo-random numbers between 0 and 1.
• Summarized four types of chaotic improvements applied to optimization algorithms.
• Conducted simulation experiments using the Grey Wolf Optimization (GWO) and Sine Cosine (SC) algorithms.

Main Results:

• The performance enhancement from chaotic improvements was found to be uncertain across different algorithms, improvement types, and benchmark functions.
• The Bernoulli map demonstrated potential as a generally effective choice for most scenarios.
• Code for the study is publicly available for reproducibility.

Conclusions:

• Chaotic map integration into optimization algorithms does not guarantee universal performance improvement.
• The choice of chaotic map and its application strategy significantly influences optimization outcomes.
• The Bernoulli map emerges as a potentially robust option for enhancing optimization algorithms in various contexts.