摘要
本文以神经网络理论为基础，将并行处理方法用于求解固定需求交通平衡分配问题之中，从而减少了利用平衡交通分配模型进行交通分配的时间，降低了分配过程的复杂性，为该模型在实际工程中得到应用提供可能。
This paper presents a parallel processing method for the transportation equilibrium assignment problem. The method based on the neural network theory would substantially reduce the computation time if the neural computer will be available in near future.The neural network consists of mutually connected 'neurons', which change the states inparallel to the descent direction of the Lyapunov function. In the method proposed in thispaper, the link flows are represented by the neural states, and we can obtain the equilibrium flows by the parallel change of the neural states according to the equations of neuraldynamics.
出处
《中国公路学报》
EI
CAS
CSCD
北大核心
1997年第1期84-89,共6页
China Journal of Highway and Transport
基金
国家自然科学基金
