Abstract: Ant colony algorithm,in recent years,emerges as a novel approach of bionic meta-heuristic algorithm in the field of optimization.Though it is widely applied in the discrete space area,it is relatively less researched in solving continuous function optimization.Aiming at overcoming the shortage of long time in searching for continuous function with ant colony algorithm,the paper proposes an improved ant colony algorithm for solving continuous function optimization,which is based on the original methods of continuous function optimization.The improvement is directed against the total amount of pheromone and size of ant colony within all the subintervals.It leads-in a function that varieswith increase of the iterations,in the hope of increasing the convergence speed of ant colony algorithm after its improvement.And numerical simulation results indicate that,comparing with the algorithm proposed by References,thisimproved algorithm offers better solution for continuous space optimization problems,hence it is an effective new way to solve problems alike.

Key words: ant colony algorithm, continuous function optimization, pheromone