摘要
目前对麻醉呼吸机活瓣关闭不全故障检测结果存在较大误差率,因此,提出基于进化算法的麻醉呼吸机活瓣关闭不全故障自动检测。从麻醉呼吸机运行过程中随机提取变异故障参数,生成故障数值初始种群,得到其参数值。按照时间周期延拓的离散变化提取故障特征,随后利用进化算法对特征数据合集扩散概率计算,使扩散故障因子聚集,从而完成故障因子进化自动检测。设计仿真实验,通过对比基于进化算法的麻醉呼吸机活瓣关闭不全故障自动检测与基于网络神经的麻醉呼吸机活瓣关闭不全故障自动检测,分析两种检测方法辨识误差,证明研究有效性。
At present,there is a large error rate in the detection results of anaesthesia ventilator valve closure incomplete fault,so the automatic detection of anesthesia ventilator valve closure incomplete fault based on evolutionary algorithm is proposed.The variable fault parameters were randomly extracted from the operation of the anesthetic ventilator,and the initial population of the fault value was generated.the the fault feature is extracted according to the discrete change of the time period extension,and then the evolutionary algorithm is used to calculate the diffusion probability of the feature data combination set to make the diffusion fault factor aggregate,thus the automatic detection of the fault factor evolution is completed.A simulation experiment was designed to compare the anesthesia ventilator valve closure failure automatically based on evolutionary algorithm.The detection is compared with the automatic detection of closed valve failure of anesthesia ventilator based on network nerve,and the identification error of the two detection methods is analyzed to prove the effectiveness of the research.
作者
徐咏梅
XU Yongmei(Second Affiliated Hospital of PLA Army Military Medical University,Chongqing 400037,China)
出处
《自动化与仪器仪表》
2020年第10期161-164,共4页
Automation & Instrumentation
关键词
进化算法
麻醉
呼吸机
自动检测
evolutionary algorithm
anesthesia
ventilator
automatic detection