摘要
针对Hopfield神经网络解旅行商问题(TSP)经常出现无效解和局部优化解。将模拟退火智能算法与Hopfield神经网络相结合,提出了一种混合优化算法(SA-HNN),同时合理地修改了Hopfield神经网络的能量函数,确立网络参数。这种方法在很大程度上避免了Hopfield神经网络优化陷入局部极小的缺陷,大量实验证明了该算法具有收敛速度快,可避免无效解,易获得全局最优解等优点。
For the Hopfield neural network in solving the traveling salesman problem（TSP） often getting unvalid and not optimal solution, the simulating anneal algorithm with the Hopfield neural network is combines, and bring forward a sort of the mixed optimization algorithm （SA-HNN）. At the same time, the energy function of the Hopfield neural network is modified reasonably. The network parameter is also established rationally. This algorithm can avoid the local optimal solution effectively. A number of simulations show that the algorithm has many advantages such as faster convergence rate, avoiding most local energy minima, easily obtaining effectual and global optimal solution.
出处
《科学技术与工程》
2008年第14期3937-3939,共3页
Science Technology and Engineering
