A novel hybrid algorithm of GSA with Kepler algorithm for numerical optimization

Abstract

It is now well recognized that pure algorithms can be promisingly improved by hybridization with other techniques. One of the relatively new metaheuristic algorithms is Gravitational Search Algorithm (GSA) which is based on the Newton laws. In this paper, to enhance the performance of GSA, a novel algorithm called “Kepler”, inspired by the astrophysics, is introduced. The Kepler algorithm is based on the principle of the first Kepler law. The hybridization of GSA and Kepler algorithm is an efficient approach to provide much stronger specialization in intensification and/or diversification. The performance of GSA–Kepler is evaluated by applying it to 14 benchmark functions with 20–1000 dimensions and the optimal approximation of linear system as a practical optimization problem. The results obtained reveal that the proposed hybrid algorithm is robust enough to optimize the benchmark functions and practical optimization problems.

Keywords

• Numerical function optimization
• Hybrid algorithm
• Swarm intelligence
• Astrophysical concepts
• Gravitational Search Algorithm
• Kepler algorithm