Study Of Function Optimization Method Based On Artificial Memory Principles

With the development and widely use of swarm intelligence optimization algorithm,more and more experts and scholars to join group of intelligence research field, especially the use of the swarm intelligence algorithm in complex function optimization. Swarm intelligent optimization and artificial life has been became more attentions and hot research in recent years. Under this based,through analyses and researches the special mechanism of article memory, introduced the unique memory mechanism of human brain into application of function optimization algorithm and according to the definite problem to building the definite theory, mathematical model and algorithm. This offers effective solutions to solve the practical application problems of complex function optimization, overcome the existing defect of intelligent algorithms in complex function problem.First, this paper form complete sets of article the memory theory though analyzes and improving the existing memory theory. It illustrate the flaws of existing swarm intelligence algorithm in complex function optimization, and established a function optimization algorithm based on the basic model of memory; in the algorithm, the test solution is automatic classification base of the quality by association test solution and memory, it fast convergence, high accuracy and strong effectiveness base on the characteristics and basic principle of artificial memory. Secondly, in this paper, combines the theory of memory principle with traditional ant colony algorithm, substituting forgetting and update of article memory for hormone accumulation and evaporation; and building the ant own memory banks and ant swarm memory banks to the optimal path search. Instantaneous, short-term and long-term memory can adjust ant swarm synthesis memory and storage, update validpath in different range of time, as this avoids the traditional ant colony algorithm order to improve the algorithm efficiency of search and the accuracy to preset number of ant colony and the number of iterations. Finally, the paper apply artificial memory into particle swarm algorithm, and recording the memory storage, update with searching behavior of particle swarm. Particle swarm can according to the change of the memory value in the memory banks to adjust the searching behavior and updates the corresponding memory,improve the particle swarm optimization speed.In this paper, three new build optimization models was analyzed and validated by actual examples, and give the corresponding implementation step of algorithms. The experiments show the design of the three algorithms are the rationality and effective of these algorithms in solving the problems of complex function optimization. In conclusion,this paper gives a further study of the theory and technology application model of article memory principle on application of function optimization problem, and provides theoretical fulcrum and solutions thinking for the thorough studies of memory principle on applications of the algorithm optimization.