The literature shows that the Gravitational Search Algorithm (GSA) is really competitive compared to Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) on benchmark functions. However, local optima entrapment and slow convergence are hindrances when solving real engineering problems. Such issues originate from slow movement of masses due to nearly equal weight proportional to the number of iterations. In this study, 10 chaotic maps tune the gravitational constant (. G) to overcome these problems. The gravitational constant balances exploration and exploitation, so chaotic maps are allowed to perform this duty in this study. Ten unconstrained benchmark functions examine the proposed Chaotic GSA (CGSA) algorithms. This work also considers finding the optimal design for welded beam and pressure vessel designs to prove the applicability of the proposed method. The results prove that chaotic maps improve the performance of GSA.
|Title of host publication||Handbook of Neural Computation|
|Number of pages||16|
|Publication status||Published - 1 Jan 2017|
- Constrained optimization
- Gravitational Search Algorithm
- Population-based algorithm
- Stochastic optimization