Gravitation Algorithm Design For Solving Complex Fucntion Optimization Problems

Gravitation algorithm is based on Newton’s law of universal gravitation physics and simulation of the universe, all matter has mass attract each other and to a large gathering of model substance, which is a global optimization algorithm, gravitation algorithm is a complex function optimization algorithm for solving the problem of providing a kind of innovative solutions, showing broad prospects of development of universal gravitation algorithm is a swarm intelligence optimization algorithm, the basic idea: mutual attraction by population among substance produced swarm intelligence optimization search, the mechanism and principle is very simple, even if the algorithm has a good performance optimization,while maintaining the swarm intelligence background.Therefore,GSA algorithm have been widely applied to function optimization, multiobjective programming, fuzzy systems control and other fields.The traditional gravitation algorithm has two shortcomings: a local search capability is relatively weak, and the other is prone to stagnation.On one hand is dynamically adjusted weighting factor, the introduction of a mutation in the search process operators in order to improve the local search ability of the algorithm, the other is to increase the diversity of the population, while enhancing the ability of local optima jumping, avoid blind calculation and accelerate the convergence speed to avoid these two problems, for solving complex function optimization problem, we do the research work the following two areas:1.According to the gravitation algorithm’s weak local convergence, the paper proposes an adaptive chaotic mutation gravitational search algorithm,which introduces the average particle distance and chaotic search for variation, improves the local search ability of gravitation algorithm, increases the diversity of physical population, and treats the infeasible material after variation by constraint boundary mutation. The experimental results show that the new algorithm has a higher convergence precision, a faster convergence speed and can effectively avoid premature convergence problem,compared with the basic gravitation algorithm (SGSA) and chaos gravitation algorithm (CGSA).2.1n order to balance this contradiction between the gravitation algorithm search speed and function optimization solutions accuracy,this paper combines the optimal foraging theory and gravitation algorithm,and introduces the quality difference between groups of substances and distance ratio as a measure of individual energy efficiency in gravitation algorithm so as to improve the computational algorithm efficiency.and does a new substance of a new strategy to cross-border mutation to increase the diversity of physical substances.Numerical experiments show that the algorithm is better than SGSA, CGSA in solving accuracy.