摘要
医疗器械的振动信号通常包含大量的噪声和干扰,导致振动信号难以从背景噪声中分离,无法有效提取振动信号中对故障检测有用潜变量,且振动信号具有非线性和非平稳特性,使得振动故障检测更加困难。因此,提出基于相对变换PLS的医疗器械振动故障检测方法。基于相对变换理论将医疗器械振动信号从原始空间变换到相对空间,增强振动信号的可区分性,使其从背景噪声中分离;利用PLS算法提取振动信号的潜变量;为处理非线性和非平稳性问题,将提取的振动信号潜变量作为输入,通过支持向量数据描述算法(Support Vector Data Description,SVDD),实现医疗器械的振动故障检测。实验结果表明,上述方法振动故障检测的IG值为0.97,且具有较高的检测精度。
The vibration signal of medical instruments usually contains a lot of noise and interference,which makes it difficult to separate the vibration signal from the background noise,and it is impossible to effectively extract the useful latent variables in the vibration signal for fault detection.Moreover,the vibration signal has nonlinear and non-stationary characteristics,which makes the vibration fault detection more difficult.Therefore,a vibration fault detection method of medical devices based on relative transformation PLS is proposed.Based on the relative transformation theory,the vibration signal of medical instruments is transformed from the original space to the relative space to enhance the distinguishability of the vibration signal and separate it from the background noise.Using PLS algorithm to extract the latent variable of vibration signal;In order to deal with the nonlinear and nonstationary problems,the vibration fault detection of medical devices is realized by using the support vector data description(SVDD)algorithm with the extracted latent variables of vibration signals as input.The experimental results show that the I c value of vibration fault detection by the above method is 0.97,and it has high detection accuracy.
作者
王梦媛
魏代清
卢贞燕
魏柏文
WANG Meng-yuan;WEI Dai-qing;LU Zhen-yan;WEI Bo-wen(Department of Medical Equipment,Affiliated Hospital of Southwest Medical University,Luzhou Sichuan 646000,China;Department of Medical Affairs,Affiliated Hospital of Southwest Medical University,Luzhou Sichuan 646000,China;Infrastructure Department,Affiliated Hospital of Southwest Medical University,Luzhou Sichuan 646000,China)
出处
《计算机仿真》
2025年第4期446-450,共5页
Computer Simulation
基金
四川省科技厅自然科学基金(2023NSFSC1872)。
关键词
相对变换
医疗器械
振动故障检测
Relative transformation
Medical devices
Vibration fault detection
