查表计算的人工神经网络方法研究

Artificial Neural Network Method for Table-Lookup Calculation

提出了一种构建神经网络模型的归纳综合法,并以零件工时定额计算为例介绍了该方法的应用。深入研究了神经网络的计算精度,阐明了该方法具有网络模型数量少、计算准确、使用简便的优点,是解决查表计算化的有效方法。

The synthetic inductive method for constructing a neural network model is put forward. The application of this method is introduced by taking the manhour rating calculation of the parts as an example. The calculating accuracy of the neural network is studied. Some advantages of this method, such as a small amount of network model, accurate calculation and simple to use, which is an effective method for table-lookup calculation are described.