结构平面布置的神经网络实现

The Use of Neural Networks in Plane Layout

改进ＢＰ神经网络算法，有效地克服了神经网络技术应用中普遍存在的学习收敛速度慢及陷入局部极小的问题，并成功应用于结构平面布置系统的设计．

Slow specd of learning convergence and local minimun error function are common problems in the use of neural network. A modifited algorithm can overcome these deficiencies,and be used in design of plane layou.