Microwave-Sensor Array for Decoupling Detection of Distance, Shape, Dielectric, and Morphology

Usually, four sensors are required to detect an object’s distance, shape, morphology, and permittivity. This work proposes a two-port microwave sensor array to comprehensively detect these parameters. The array consists of transmission line and nine sensitive independent units designed as spiral resonators. The distance (or shape) detection range is 0–0.5 mm, with the sensitivities of 161 MHz/mm by frequency shift or 1.29 dB/mm by amplitude change and the limit of detection (LOD) of 0.07 mm. The array can also scan the morphology by detecting the thickness with a maximum sensitivity of 214 MHz/mm (or 0.67 dB/mm) and an LOD of 0.039 mm, respectively. The sensor can also examine material-under-test’s (MUT’s) overall uniformity by array and measure its permittivity, showing a sensitivity of 94 MHz/
$\Delta \varepsilon $
. A neural network algorithm is proposed to decouple the distance, thickness, and permittivity. Moreover, the detection performance, such as sensitivity, resolution, and precision, is determined by the microwave sensor array, for which this work can be regarded as a polit study and can be applied in a wide application after certain optimization.