Improved Gravitational Search Algorithm (GSA) Using Fuzzy Logic

Subject Areas : Control systems

Omid Mokhlesi ¹ (MSC/khorasan Institute for Higher Education, Mashhad)
Seyed Hamid Zahiri ² (Associate Professor/ University of Birjand)
Naser Mehrshad ³ (Assistant Professor / University of Birjand)
Seyed Mohammad Razavi ⁴ (Assistant Professor/ University of Birjand)

Received: 2012-07-04 Accepted : 2013-05-04 Published : 2013-06-01

Keywords: scattering coefficient, Parametric fuzzy controller, Gravitational search algorithm (GSA), Fuzzy Population model,

Abstract :

Researchers tendency to use different collective intelligence as the search methods to optimize complex engineering problems has increased because of the high performance of this algorithms. Gravitational search algorithm (GSA) is among these algorithms. This algorithm is inspired by Newton's laws of physics and gravitational attraction. Random masses are agents who have searched for the space. This paper presents a new Fuzzy Population GSA model called FPGSA. The proposed method is a combination of parametric fuzzy controller and gravitational search algorithm. The space being searched using this combined reasonable and accurate method. In the collective intelligence algorithms, population size influences the final answer so that for a large population, a better response is obtained but the algorithm execution time is longer. To overcome this problem, a new parameter called the dispersion coefficient is added to the algorithm. Implementation results show that by controlling this factor, system performance can be improved.