A new method of improving Wang and Smith's chaotic neural network
Proceedings of the 4th World Congress on Intelligent Control and Automation (Cat. No.02EX527) - Tập 3 - Trang 1994-1997 vol.3
Tóm tắt
In this paper, a novel method is proposed, with the aid of which the change of time step, /spl Delta/t, depends on two parameters according to the value of the cost function of combinatorial optimization problems. We have applied Wang and Smith's (1998) method and the proposed method to 10 city traveling salesman problem (TSP), respectively. The simulation results show that the performance of the proposed method is better than that of Wang and Smith's chaotic neural networks.
Từ khóa
#Chaos #Neural networks #Cellular neural networks #Bifurcation #Cost function #Traveling salesman problems #Optimization methods #Cities and towns #Simulated annealing #Hopfield neural networksTài liệu tham khảo
10.1103/PhysRevE.51.R2693
10.1109/72.701185
teren, 1999, A unified framework for chaotic neural network approaches to combinatorial optimization, IEEE Trans on Neural Networks, 10, 978, 10.1109/72.774279
10.1073/pnas.81.10.3088
hopfield, 1985, Neural computation of decisions in optimization problems, Biological Cyborenetics, 52, 141
jiao, 1990, Theory of Neural Network System, 52
10.1016/0893-6080(95)00033-V
he, 2000, Chaotic neural networks and their applications, Proceedings of the 3rd World Congress on Intelligent Control and Automation, 826