Research On MIEC Sensor Application And Its Signal Processing Method

Based on the advantages of non-invasive measurement,the laboratory designed and developed the Motion Induced Eddy Current Sensor for vibration signal measurement.As a non-intrusive sensor,MIEC sensor has the advantages of simple structure,low price and ensuring the integrity of the structure of the measured object.It can be used not only for the status monitoring of electromechanical equipment,but also for human health monitoring in the field of smart home.Aiming at indoor human health monitoring and vibration monitoring of rotating equipment,this paper takes MIEC sensor as the signal acquisition device,and mainly studies the application of MIEC sensor in vibration signal acquisition and its signal processing methods.The main work includes three parts:(1)Research on the application of MIEC sensors.The structure and measuring principle of MIEC sensor are studied.And for the vibration signal measurement of human body and electromechanical equipment,the conditioning circuit of MIEC sensor is designed and the non-intrusive deployment of MIEC sensor is carried out.For the detection of human heart function,the MIEC sensor is installed on the office seat,and the measurement of the BCG signal is realized by the MIEC sensor collecting the human body vibration signal on the seat.For the collection of bearing vibration signals,the MIEC sensor is arranged on the base of the rotating vibration table to realize the body surface measurement of the bearing vibration signals.(2)Research on event signal extraction algorithm.For event signals with impulsive properties,such as BCG signals measured by MIEC sensors and ECG signals and bearing fault signals commonly used in the field of human and equipment health monitoring,an event signal extraction algorithm based on phase space reconstruction is studied.The algorithm includes three parts: peak characteristic curve calculation,time series segmentation and similarity segment search.Through the phase space reconstruction and grid calculation,the peak characteristic curve of the original signal is obtained,which can realize the local enhancement of the event characteristic signal.Signal segmentation is based on the peak characteristic curve,which ensures that the split sub-signals have complete event characteristics.The segmented sub-signals are embedded in the two-dimensional phase space for similarity calculation,which solves the problem of similarity measurement of unequal-length signals,and the clustering algorithm is used to find similar features of sub-signals with event features.The algorithm is used to process the measured BCG signals and the data in public databases.The results obtained verify the effectiveness of the event signal extraction algorithm.(3)Research on fault information extraction algorithm.The signal measured by the MIEC sensor on the body surface has a low signal-to-noise ratio.Therefore,for the problem of extracting fault information from vibration signals with low signal-to-noise ratio,a time-frequency joint analysis algorithm based on Teager energy operator is studied.The Teager energy operator introduces the concept of energy when analyzing the signal,so that the impulse response component of the fault in the signal is enhanced,and it is easier to observe in the result.In the calculation process,only three sampling points are needed to obtain an output point,the calculation amount is small,and it has good time resolution and instantaneous response characteristics.Combined with the empirical mode decomposition algorithm,in the obtained time-frequency analysis results,the bearing fault characteristics are more obvious than the results of the short-time Fourier transform.