简单介绍了神经网络的理论、方法、计算、应用与实现，讨论了神经网络较之于冯·诺依曼式数字计算机在计算方法上的区别及其在处理组合优化问题上的优势。探索了神经网络在邮政网络规划中的应用，给出了神经网络用于邮政通信网结构研究的方法和过程，论述该方法在最小成本优化方面的应用，建立相应的神经网络模型，开发相应的软件，并给出了计算实例。 The theory, method, calculation, application and realization of neural network are briefly introduced. The difference between neural network and Von Neumann digital computer and the advantage in dealing with combinatorial optimization are discussed. This paper explores the application of neural network in postal network planning and gives the method and process of neural network for postal communication network structure research. It discusses the application of this method in the optimization of minimum cost, establishes the corresponding neural network model and develops the corresponding Software, and gives a calculation example.