Solution to Combinatorial Problems by Using Chaotic Global Optimization Method on A Simplex

抄録

The discretized maps of gradient models with Euler’s method generate chaos, which can be used for global optimization by applying the chaotic annealing. In this paper, we apply this approach to solve the traveling salesman problem. We formulate it as continuous optimization problems with particular constraints, and newly derive gradient projection dynamics whose discretized maps generate chaos on a simplex.