运动体脂监测系统中基于PSO-ELM算法的生物电阻抗-体脂率预测模型研究  

Research on Bioelectrical Impedance and Body Fat Rate Prediction Model Based on PSO-ELM Algorithm in Sports Body Fat Monitoring System

作　　者：涂锐 TU Rui(Xi’an academy of fine arts,Xi’an 710065,China)

机构地区：[1]西安美术学院,西安710065

出　　处：《自动化与仪器仪表》2024年第5期29-32,37,共5页Automation & Instrumentation

基　　金：《新时代背景下陕西体育博物馆的文化宣传路径研究》(2023535)。

摘　　要：为实时监控运动员的体脂率,同时做出预防与合理的运动训练安排,研究设计了基于卡尔曼滤波算法的生物阻抗检测方法,并提出一种基于粒子群算法优化的极限学习机(PSO-ELM)的生物电阻抗-体脂率预测模型和运动体质监测系统。研究结果显示,卡尔曼滤波算法处理后的人体电阻抗信号最大误差不超过±0.01。在皮肤干湿度、饮水与时间条件下,数据前后的阻抗值差异非常小。与主流预测模型的对比中,基于PSO-ELM的生物电阻抗-体脂率预测模型的相关系数最大为0.985 4。综上所述,研究提出的系统能满足运动员的日常体脂率检测的功能需求,并实现实时检测。In order to monitor the body fat rate of athletes in real-time and make preventive and reasonable exercise training arrangements,a bioimpedance detection method based on Kalman filter algorithm was studied and designed.A bioimpedance body fat rate prediction model and exercise fitness monitoring system were proposed based on Particle Swarm Optimization Extreme Learning Machine(PSO-ELM).The research results show that the maximum error of the human body electrical impedance signal processed by the Kalman filtering algorithm does not exceed±0.01.The difference in impedance values before and after the data is very small under conditions of skin dryness,water intake,and time.In comparison with mainstream prediction models,the maximum correlation coefficient of the PSO-ELM based bioelectrical impedance body fat rate prediction model is 0.9854.In summary,the system proposed in the study can meet the functional requirements of daily body fat rate detection for athletes and achieve real-time detection.

关 键 词：体脂 监测系统 PSO-ELM算法 网络化 生物电阻抗 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]