基于遗传神经网络的失能老人健康评估  

The Health Assessment of the Disabled Elderly Based on Genetic Neural Network

作　　者：王帅[1] 陶晋宜[1] 杨刚[2] 张晓雪 WANG Shuai;TAO Jin-yi;YANG Gang;ZHANG Xiao-xue(Institu te of Elec trical and Power Engineering, Taiyuan University of Technology, Taiyuan 030024, China;Institute of Electronic Engineering, Xidian University, Xian 710071, China;Neurology, Affiliated Hospital of Shanxi University of Chinese Medicine, Taiyuan 030024, China)

机构地区：[1]太原理工大学电气与动力工程学院,山西太原030024 [2]西安电子科技大学电子工程学院,陕西西安710071 [3]山西中医药大学附属医院神经内科,山西太原030024

出　　处：《数学的实践与认识》2019年第10期219-225,共7页Mathematics in Practice and Theory

基　　金：工业和信息化部2018年通信软科学研究项目基金(2018-R-21);超高速电路设计与电磁兼容教育部重点实验室开放式基金(2017KFKT/B11,2017KFKT/B04)

摘　　要：针对人体健康评估所需参数多,专业性较强,家人对家庭卧床失能老人的健康状况难以准确判断等问题,与医院、养老机构和老年社区合作,提出一种基于遗传神经网络的家庭简单健康评估方法.首先挑选出家庭易测且能反映健康状况的5项生理参数作为特征向量,并对应设置3个健康等级,然后利用遗传神经网络建立生理参数与健康等级的映射关系,最后与一般神经网络进行对比.仿真结果表明,遗传神经网络模型预测误差更小,准确率更高,验证了方法在老人健康评估中的可行性.Since there are many parameters needed and the professionalism is relatively strong for the assessment of human health, the family members are unable to accurately determine the health status of the disabled elderly, a simple health assessment method is proposed based on genetic neural network by cooperating with hospitals, pension institutions and elderly communities. Firstly, five physiological parameters which can be easily measured and reflect the health status are selected as feature vectors, and three health grades are set up. Then, the mapping relationship between physiological parameters and health grades is established by genetic neural network. Finally, the model is compared with the general neural network model. The simulation results show that, the genetic neural network model has smaller prediction error and higher accuracy, which verifies the feasibility of this method in assessing the health of the elderly.

关 键 词：神经网络 遗传算法 失能老人 健康评估 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] R161.7[自动化与计算机技术—控制科学与工程]