2020-08-012024-05-17https://scholars.lib.ntu.edu.tw/handle/123456789/679230This research proposes the development of point-surface-volume electrical conductivity (PSV-EC) intelligent sensing systems for non-ferrous metallic and biological objects via magnetic/eddy-current (M/EC) and electrical impedance (EI) sensing, including the forward/inverse modeling method and development of sensing systems. In the forward method, the distributed parameter method formulates the two and three dimensional magnetic/eddy-current fields of vector potential fields and electrical and thermal fields of scalar potential fields into state-space representation for frequency and time domain analysis. The modeling methods will be verified by commercial finite element analysis software and experimental data. The inverse methods are obtained by integrating the closed-form solutions from the distributed parameter element method and some techniques, such as the frequency response mapping, Tikhonov regularization, and deep neural network. The developing systems include point-surface-volume metallic-magnetic/eddy-current (PSV-M-M/EC), point-surface-volume biomagnetic/eddy-current (PSV-Bio-M/EC), and point-surface-volume bio-electrical impedance (PSV-Bio-EI) sensing systems. The point sensor, the basic sensing unit, is utilized to measure the electrical conductivity of a point in objects. In the surface and volume sensing systems, the sensing units are distributed on the surface and volume of the objects for estimations of the electrical conductivity distributions. The metallic-magnetic/eddy-current sensing systems measure magnetic fields generated from induced eddy-currents by anisotropic magnetoresistance (AMR) sensors. The bio-magnetic/eddy-current sensing systems measure the magnetic fields generated from the induced eddy-currents by sensing coils. The bio-electrical-impedance sensing systems measure the electric fields of biological objects by electrodes. The magnetic/eddy-current systems are non-contact methods and do not deteriorate the objects. The electrical impedance systems have stronger signals for analysis. After calibration, the PSV-EC sensing systems integrated with the forward/inverse modeling will be utilized to estimate the electrical conductivity distributions of the metallic to biological objects for some applications, such as non-destructive detection for metallic and abnormal biological objects.magnetic/eddy-current sensing, electrical impedance sensing, electrical conductivity estimation, forward and inverse model