Using artificial physics to solve global optimization problems

Heuristics are quite an effective kind of methods to solve global optimization problems, which utilizes sample solution(s) searching the feasible regions of the problems in various intelligent ways. Inspired by physical rule, this paper proposes a stochastic global optimization algorithm based on physicomimetics framework. In the algorithm, a population of sample individuals search a global optimum in the problem space driven by virtual forces, which simulate the process of the system continually evolving from initial higher potential energy to lower one until a minimum is reached. Each individual has a mass, position and velocity. The mass of each individual corresponds to a user defined function of the value of an objective function to be optimized. An attraction-repulsion rule is constructed and used to move individuals towards the optimality. Experimental simulations show that the algorithm is effective.