基于改进T-S云推理网络的板形模式识别方法

来源期刊:中南大学学报(自然科学版)2013年第2期

论文作者:张秀玲 赵文保 张少宇 徐腾

文章页码:580 - 586

关键词:云模型;T-S模糊神经网络;最速下降法;板形;模式识别

Key words:cloud model; T-S fuzzy neural network; steepest descent method; flatness; pattern recognition

摘    要:将云模型与T-S模糊神经网络相结合,利用正态云代替模糊神经网络的隶属度函数,构成T-S云推理网络。该网络综合考虑了模糊逻辑的快速性和云模型处理数据的不确定性,增强了网络处理数据的能力,同时分析了T-S云推理网络的结构和稳定性。在超熵确定的情况下,使用最速下降法辨识了T-S云推理网络的参数,将该网络应用于板形模式识别,并与T-S模糊神经网络作了对比。仿真结果表明:T-S云推理网络抗干扰能力较强,能够识别出常见的板形缺陷,并且识别精度较高。

Abstract: T-S cloud inference network was built based on the cloud model and T-S fuzzy neural network, and the normal cloud is used to take the place of membership function. The network takes into account synthetically the rapidity of fuzzy logic and the uncertainty of cloud model for processing data, and enhances the network ability of processing data. At the same time, the structure and stability of T-S cloud inference network were analyzed. When ultra-entropy was determined, the steepest descent method was used to identify the parameters of T-S inference network cloud. The network was applied to the recognition of flatness pattern and comparison with T-S fuzzy neural network. The simulation results demonstrate that T-S cloud inference network has a stronger ability of anti-interference, and it can recognize common defects in flatness with higher identify precision.

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号