Electrical impedance sensing system design for abnormal object detection
Journal
IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM
Journal Volume
2021-July
Pages
1313-1318
Date Issued
2021
Author(s)
Abstract
This paper proposed a design for an electrical impedance (EI) sensing system. For the part of physical modeling, the harmonic electric fields of the EI sensing system are formulated by the distributed parameter element (DPE) method to calculate the electrode potentials for several injection patterns of different abnormal object distributions, and the computed electrode potentials are feed into a deep neural network (DNN) to estimate the location and size of the abnormal object. For the part of system development, an electric circuit that integrates the multiplexer and Howland pump is utilized to switch the current injection electrodes and control the injection currents. The harmonic electric fields computed by the DPE method are verified by the FEA software, and the effects of utilizing the DNN for abnormal object detection are numerically validated. The proposed design, along with a prototype of the EI sensing system, which is conducted on two kinds of materials, phantom and biological objects, have been experimentally compared. ? 2021 IEEE.
Subjects
Deep neural networks
Distributed parameter control systems
Electric fields
Electric impedance
Electric impedance measurement
Electric switches
Electrodes
Intelligent mechatronics
Object recognition
Abnormal object detections
Biological objects
Distributed parameter elements
Electrical impedance
Electrode potentials
Harmonic electric field
Object distribution
System development
Object detection
SDGs
Type
conference paper