An Enhanced Multi-Sensor Detection Scheme With Drift Calibration and Array Optimization

In this paper, a multi-metal oxide sensor (MOS) detection system aims at identifying herbal medicines is proposed. To improve the long-term stability of the system and reduce the number of sensors, we derive a combined scheme with drift calibration and array optimization design. Specifically, a subspace projection matrix considering multiple constraints is first optimized to compensate the sensor drift. Subsequence, a sensor array optimization via an embedded joint mutual information framework is developed to remove redundant information in the target domain. Finally, the experimental results show that compared with the existing approaches, the proposed calibration framework can obtain the best system performance with the minimum number of sensors.