用随机神经网络优化求解改进算法的研究

Improved algorithm research of optimization solution with random neural network

随机神经网络是一种仿照实际的生物神经网络的生理机制而定义的网络,其网络结构及应用具有自身的特点。在详细讨论了动态随机神经网络求解典型NP优化问题TSP的算法的同时,特别提出了一种有效改进算法,使得参数在简单选取的情况下保证能量函数的下降,在组合优化问题上具有普遍意义,并且在10城市TSP对改进算法进行验证,指出RNN是解决TSP问题的有效途径。

Random neural network is defined according to the actually biologic neural networks. It has its own peculiarities on the structure and the applications. The algorithm on the typical optimal problems - TSP with dynamical random neural network is elaborated. Especially,an effective improved algorithm is put foreword. The decline of the energy function is ensured by the simply selected parameter that has universal significance on combinatorial optimization. The improved algorithm is tested in solving 10-city TSP, and random neural network is verified to be an effective way to solve TSP.