论文部分内容阅读
基于过程先验知识与神经网络结构选择相结合的思想，提出了一类新型神经网络结构，并具体研究了其学习算法．同时给出了一种新的样本抽取方法及其相应的网络训练步骤．这些思想和方法均在复杂尿素生产过程的建模和优化中得到了应用．基于所建神经网络模型知识的专家系统已在某化肥厂得到成功的应用．
Based on the idea of process prior knowledge and neural network structure selection, a new type of neural network structure is proposed and its learning algorithm is studied in detail. At the same time, a new sample extraction method and its corresponding network training steps are given. These ideas and methods have been applied in the modeling and optimization of complex urea production processes. An expert system based on knowledge of the built neural network model has been successfully applied in a fertilizer plant.