一种群体趋化性算法及其应用的研究

Study of a New Group Chemotaxis Algorithm and Its ApplIcation

摘要: 提出了一种新型神经网络群体趋化性算法，将其用于多层前向神经网络的学习过程，表明其具有学习速度快、精度高的特点。进行了基于这一算法的尿素生产过程神经网络模型化研究。仿真结果表明了这一算法的有效性。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应用。

Abstract: This paper puts forward a new group chemotaxis algorithm used as the learning algorithm for BP neural networks.Both training speed and accuracy of neural networks are increased.This new algorithm is verified through the modelling application in the complex urea production process.