论文部分内容阅读
分析了Hopfield－Tank模型在收敛性、稳健性、优化率以及计算速度方面存在的问题，根据外部惩罚函数法的基本思想提出了一种新的基于Hopfield－Tank模型的快速神经网络方法。对TSP的能量函数进行了改进，并对我国31个城市的TSP进行了软件模拟，得出了15640km的最短路径，在收敛性、稳健性、优化率以及计算速度方面的结果都十分满意。
The problems of Hopfield-Tank model in convergence, robustness, optimization rate and calculation speed are analyzed. A new fast neural network method based on Hopfield-Tank model is proposed according to the basic idea of external penalty function method. The energy function of TSP is improved. The software TSP of 31 cities in China is simulated and the shortest path of 15640km is obtained. The convergence, robustness, optimization rate and calculation speed are all satisfied.