基于可视化和数据融合的医疗设备异常运行状态检测方法  

Abnormal Operation State Detection Method of Medical Equipment Based on Visualization and Data Fusion

作　　者：吴昊[1] 林建阳[1] Wu Hao;Lin Jianyang(Asset Management Department,the First Affiliated Hospital of China Medical University,Shenyang 110001,China)

机构地区：[1]中国医科大学附属第一医院资产管理部,沈阳110001

出　　处：《科技通报》2022年第11期37-40,45,共5页Bulletin of Science and Technology

基　　金：国家自然科学基金青年科学基金资助项目(81302841)。

摘　　要：医疗设备运行状态难以检测,一旦出现异常情况,只能强制停机检查,降低了工作效率,增加了维护成本,为此,提出了基于可视化和数据融合的医疗设备异常运行状态检测方法。预处理运行数据,通过融合、分析和可视化等操作剔除冗余参数值,以正常运行状态下参数值为标准,构建健康状态评估模型,运用滑动窗口残差统计法和加权平均方差消除运行状态中干扰因素和不确定因素,设置合理的预警偏离阈值和检测阈值,确保医疗设备异常运行状态检测精度。实例测试结果表明,在人为添加不同程度的异常信号环境中,所提方法均能有效检测医疗设备异常运行状态,帮助工作人员根据检测情况采取相应的处理措施。It is difficult to detect the operation state of medical equipment. Once there is an abnormal situation, it can only be forced to stop for inspection, which reduces the work efficiency and increases the maintenance cost. Therefore, a detection method of abnormal operation state of medical equipment based on visualization and data fusion is proposed. Preprocess the operation data, eliminate redundant parameter values through fusion, analysis and visualization, build a health state evaluation model based on the parameter values in normal operation state, eliminate interference factors and uncertain factors in operation state by using sliding window residual statistics method and weighted average variance, and set reasonable early warning deviation threshold and detection threshold, ensure the detection accuracy of abnormal operation state of medical equipment. The example test results show that in the environment of artificially adding different degrees of abnormal signals, the proposed methods can effectively detect the abnormal operation state of medical equipment and help the staff to take corresponding treatment measures according to the detection situation.

关 键 词：数据融合 医疗设备 滑动窗口残差统计法 加权平均方差 

分 类 号：TH323.3[机械工程—机械制造及自动化]