基于改进TS云推理网络的板形模式识别方法  

Method of flatness pattern recognition based on improved T S cloud inference network

作　　者：张秀玲[1,2] 赵文保[1,2] 张少宇[1,2] 徐腾[1,2] 

机构地区：[1]燕山大学河北省工业计算机控制工程重点实验室,河北秦皇岛066004 [2]国家冷轧板带装备及工艺工程技术研究中心,河北秦皇岛066004

出　　处：《中南大学学报（自然科学版）》2013年第2期580-586,共7页Journal of Central South University:Science and Technology

基　　金：国家自然科学基金资助项目(50675186)

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。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.

关 键 词：云模型 TS模糊神经网络 最速下降法 板形 模式识别 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]