The real-world optimization problems which we encounter nowadays are becoming more and more complicated and difficult to solve by traditional heuristic methods. For those problems, metaheuristic algorithms have proven to be efficient and reliable techniques for finding near optimal solutions with a reasonable computational cost. Most of the today’s successful metaheuristic methods were inspired by swarm behavior of animals, biological systems and natural phenomena.
Recently, gravitational search algorithm (GSA) has been newly introduced, imitating the law of gravity in physics. The GSA is distinguished from other metaheuristic algorithms for its unique concepts and operators and it is regarded as one of the powerful metaheuristic algorithms. Actually, it has been widely utilized to solve various kinds of optimization problems arising in many fields, for example, image processing, effective solving of aircraft landing planning (ALP) problems in the air traffic control, predicting the structure of protein and estimating the minimum ignition energy.
Hybrid algorithms of GSA with other optimization methods have been presented in several works. Still, a large number of open problems exist for GSA despite these exertions.
Choe Thae Ryong, a researcher at the Faculty of Applied Mathematics, has investigated a method for improvement of GSA by mixing it with the invasive weed optimization, thus proposing a new hybrid algorithm called IWO-GSA. In IWO-GSA, the agents generate new seeds and scatter them at each iteration of GSA. Then, elite agents are selected according to their fitness.
He applied IWO-GSA to 23 benchmark functions and compared it with GSA and IWO. From the results of numerical experiments, he found that the proposed algorithm is superior to the competitors for most benchmark functions.
© 2021 KumChaek University of Technology