神经网络理论在图书馆运行监控系统中的应用

设计了一个基于BP人工神经网络理论的图书馆运行监控系统；该系统根据图书馆里丢失物的特征，通过图像识别进行判断，来保障图书馆的正常运行。研究了图像识别的学习速度及识别准确率与隐含层神经元个数及步幅(α)的关系。计算结果显示，该系统设计原理可靠，运行有效。

Application Running Monitor System for Libraries Based on Neural Network Theory

This paper designs a running monitor system for libraries based on BP artificial neural network. According to the characteristics of things lost in libraries, the system makes libraries run normally by means of picture identification. Also the paper studies the relationship between the learning speed, and identification accuracy of picture identification and the number of neuron in hidden layer,pace(a). Computation results prove the system reliable in design and effective in practice.