论文部分内容阅读
通过建立神经元输出和与其相邻神经元阈值之间的相互耦合,提出了一种输出-阈值耦合神经网络,实现了对在通常脉冲耦合神经网络中所存在的自动波现象的模拟,并基于它的自动波现象,提出了一种用输出-阈值耦合神经网络求解最短路问题的方法,该方法具有所需神经元数目少、神经元和网络的结构简单、大规模并行计算等特点,可用于求解非对称赋权图单起点多终点的最短路问题,其所需的计算量(迭代次数)仅正比于最短路的长度,而与图的复杂程度、所存在的通路总数等无关。最后给出了最短路求解的例子。
By establishing the output of neuron and the coupling of its adjacent neuron threshold, an output-threshold coupling neural network is proposed to simulate the auto-wave phenomena existing in the usual pulse-coupled neural networks Its auto-wave phenomenon, this paper presents a method of solving the shortest path problem using output-threshold coupled neural network. This method has the characteristics of fewer required neurons, simple structure of neurons and networks, large-scale parallel computation and so on. In solving the shortest path problem with as many endpoints as the starting point of the asymmetric weighted graph, the amount of computation required (the number of iterations) is only proportional to the length of the shortest path, not to the complexity of the graph, the total number of paths existing, and so on. Finally, an example of solving the shortest path is given.