基于人工神经网络描述的编程

Neural Network Programming Based on Neural Network Description

目的在神经网络结构和算法描述语言的基础上，提出神经网络编程的概念，建立神经网络仿真系统．方法采用具有强类型、并发性和面向对象等特点的Ａｄａ语言作为系统的宿主语言，建立基于结构、算法描述的神经网络仿真系统．结果为神经网络的工程应用研究提供手段．结论采用神经网络编程，可以将神经网络与传统的编程语言在通用计算机的基础上统一起来，达到更新更高的处理能力．

Aim To Introduce the concept of ANN programming based on the description of neural network′s structure,neural network′s algorithm,and to build up the neural network simulation system.Methods Using Ada as the basic programming language and making full use of its ability in programming large systems in conjunction with properties such as strict type limitation and object oriented programming.Results Systems are built up based on description of neural network′s structure and neural network′s algorithm.Conclusion The ability of programming language and the power of neural network are combined to solve more difficult problems.