Analysis of the impact of genetic algorithm parameters on optimization efficiency in MATLAB

This paper presents an analysis of the impact of genetic algorithm (GA) parameters on the efficiency of optimizing complex nonlinear functions using the MATLAB Optimization Toolbox. The study focuses on the Rastrigin function, known for its complex structure and multiple local minima. Key GA parameters, including population size, mutation and crossover probabilities, and stopping conditions, are considered. Experimental results demonstrate that proper parameter tuning significantly enhances the algorithm's ability to find the global minimum while reducing the likelihood of premature convergence. The findings highlight the importance of adapting GA parameters to specific optimization tasks and demonstrate the potential of GA applications in engineering and scientific domains. Limitations of the method are discussed, and future research directions, including the development of hybrid approaches, are proposed.

1. Haupt R.L., Haupt S.E. Practical Genetic Algorithms. – Wiley, 2004. - p. 253.

2. Simon D. Evolutionary Optimization Algorithms. – Wiley, 2017. - p. 776.

3. Dai Q., Liu N. Alleviating the problem of local minima in Backpropagation through competitive learning. Neurocomputing. - 2012. - 94(1). - P. 152–158. DOI: 10.1016/j.neucom.2012.03.011

4. Shad R., Doris L. Optimization techniques in machine learning: develop and analyze optimization algorithms for machine learning, such as stochastic gradient descent, convex optimization, and non-convex optimization. – Mathematics, 2024. - p. 19.

5. Kramer O. Genetic Algorithm Essentials. – Springer, 2017. - p. 94.

6. Deb K. Optimization for Engineering Design: Algorithms and Examples. PHI Learning. - 2012. - p. 440.

7. MathWorks. MATLAB Optimization Toolbox Documentation. 2024. URL: https://www.mathworks.com/help/optim/index.html