脐动脉血流时间序列PS-WOA-BP神经网络诊断模型研究  

Study on PS-WOA-BP Neural Network Diagnosis Model of Umbilical Artery Blood Flow Time Series

作　　者：俞凯君 Yu Kaijun(Graduate School of Sciences and Technology for Innovation,Yamaguchi University,Yamaguchi 755-8611,Japan)

机构地区：[1]日本山口大学创成科学研究科,日本山口755-8611

出　　处：《黑龙江科学》2023年第2期123-127,共5页Heilongjiang Science

摘　　要：为解决人为经验导致的早期脐血流病理诊断准确率不高问题,提出人工智能算法与混沌理论相结合的智能诊断模型。根据脐血流时间信号具有混沌特性,采用WOA算法与BP神经网络,构建PS-WOA-BP智能诊断模型,与传统BP算法和WOA-BP神经网络构建的诊断模型进行对比。结果表明,脐血流二维相空间混沌序列包含更多的代表性特征,以此构建的PS-WOA-BP智能诊断模型具有较高的准确度和良好的泛化能力。In order to solve the problems of low accuracy of early umbilical blood flow pathological diagnosis,the study proposes the intelligent diagnosis model under the combination of artificial intelligent algorithm and chaos theory.According to the chaos characteristic of umbilical artery flow time signal,the study adopts WOA algorithm and BP neural network,constructs PS-WOA-BP intelligent diagnosis model,and compares with traditional BP algorithm and WOA-BP neural network.The results show that two-dimension phase space chaos sequence of umbilical artery blood flow contains more representative characteristics.Based on this,PS-WOA-BP intelligent diagnosis model is qualified with better accurate and generalization ability.

关 键 词：脐血流 时间序列 鲸鱼算法 BP神经网络 

分 类 号：R714.2[医药卫生—妇产科学]