张开生,黄谦

• 1:陕西科技大学电气与信息工程学院

摘要(Abstract)：

针对传统脉诊存在易受主观因素影响、诊断结果可靠性不高等问题,提出基于粒子群优化BP神经网络的脉象识别方法。粒子群算法中评判粒子好坏的适应度函数采用神经网络的输出误差,以此获得最优粒子的位置向量,并把其值作为BP神经网络的初始权值和阈值。在Matlab中建立基于BP算法、PSO-BP算法和GA-BP算法的三种ANN模型用于脉象信号的识别。实验结果表明,在识别脉象时,优化后的算法降低了传统BP神经网络的输出误差,提高了识别精度,PSO-BP算法明显改善了传统BP神经网络的泛化能力。

关键词(KeyWords)： 脉象识别;粒子群算法;输出误差;误差反向传播算法;神经网络;泛化能力

Abstract:

Keywords:

基金项目(Foundation): 陕西省科技计划项目(2017GY-063)~~

作者(Author): 张开生,黄谦

DOI: 10.16652/j.issn.1004-373x.2018.03.023

参考文献(References)：

• [1]魏红,徐刚.从中医整体、动态、平衡观论脉诊客观化研究[J].中医杂志,2014,55(1):25-27.WEI Hong,XU Gang.Objective study of pulse diagnosis from the integral,dynamic and balanced view of traditional Chinese medicine[J].Journal of traditional Chinese medicine,2014,55(1):25-27.
• [2]焦琪玉,庞春颖.基于DSP的脉象仪的设计与实现[J].中国医疗器械杂志,2013,37(2):112-115.JIAO Qiyu,PANG Chunying.Design and realization of pulse instrument based on DSP[J].Chinese journal of medical instrumentation,2013,37(2):112-115.
• [3]宫树杰,陈明锐.基于脉象信号处理的肝硬化诊断方法研究[J].计算机应用与软件,2014,31(6):69-71.GONG Shujie,CHEN Mingrui.Diagnosis of liver cirrhosis based on pulse signal processing[J].Computer applications and software,2014,31(6):69-71.
• [4]燕海霞,王忆勤,郭睿.中医常见脉象脉图特征分析及识别研究[J].中华中医药杂志,2010,25(12):2295-2298.YAN Haixia,WANG Yiqin,GUO Rui.Analysis and recognition of common pulse patterns of traditional Chinese medicine[J].Journal of traditional Chinese medicine and pharmacy,2010,25(12):2295-2298.
• [5]郭红霞,师义民.中医脉象的BP神经网络分类方法研究[J].计算机工程与应用,2005(32):187-189.GUO Hongxia,SHI Yimin.Study on classification method of Chinese pulse by BP neural network[J].Computer engineering and applications,2005(32):187-189.
• [6]张德慧,张德育,刘清云,等.基于粒子群算法的BP神经网络优化技术[J].计算机工程与设计,2015,36(5):1321-1326.ZHANG Dehui,ZHANG Deyu,LIU Qingyun,et al.Optimization of BP neural network based on particle swarm optimization[J].Computer engineering and design,2015,36(5):1321-1336.
• [7]张璐,雷雪梅.基于粒子群优化BP神经网络的养肠胃菜谱判定[J].计算机科学,2016,43(z2):63-66.ZHANG Lu,LEI Xuemei.Decision of menu good for stomach based on particle swarm optimization and BP neural network[J].Computer science,2016,43(S2):63-66.
• [8]岳沛平,李训铭.基于小波变换的中医脉象信号特征提取与分析[J].医疗卫生装备,2006,27(1):23-25.YUE Peiping,LI Xunming.Feature extraction and analysis of pulse signal based on wavelet transform[J].Medical and health equipment,2006,27(1):23-25.
• [9]张坤艳.脉象特征信息的提取与识别方法研究[D].青岛:山东科技大学,2007.ZHANG Kunyan.Study on extraction and recognition of pulse feature information[D].Qingdao:Shandong University of Science and Technology,2007.
• [10]ERIC WONG W,YU Qi.BP neural network based effective fault localization[J].International journal of software engineering and knowledge engineering,2009,19(4):573-597.
• [11]钟珞,饶文碧,邹承明.人工神经网络及其融合应用技术[M].北京:科学出版社,2007.ZHONG Luo,RAO Wenbi,ZOU Chengming.Artificial neural network and its application technology[M].Beijing:Science Press,2007.
• [12]张少迪.基于PSO-BP神经网络的短期负荷预测算法[J].现代电子技术,2013,36(12):155-158.ZHANG Shaodi.Short term load forecasting algorithm based on PSO-BP neural network[J].Modern electronics technique,2013,36(12):155-158.
• [13]WANG Hongbo,ZHAO Xiaoqi,TU Xuyan,et al.Cooperative velocity updating model based particle swarm optimization[J].Applied intelligence,2014,40(2):322-342.
• [14]谷宇,吕晓琪,赵瑛,等.基于PSO-SVM的乳腺肿瘤辅助诊断研究[J].计算机仿真,2015,32(5):344-349.GU Yu,LüXiaoqi,ZHAO Ying,et al.Study on auxiliary diagnosis of breast cancer based on PSO-SVM[J].Computer simulation,2015,32(5):344-349.
• [15]王小川,史峰,郁磊,等.Matlab神经网络43个案例分析[M].北京:北京航空航天大学出版社,2013.WANG Xiaochuan,SHI Feng,YU Lei,et al.43 cases analysis of Matlab neural network[M].Beijing:Beijing University of Aeronautics and Astronautics Press,2013.