Real-Time Modeling and Correction of Sensor Inputs and Outputs for Increased Accuracy and Reliability

Abstract

Mathematical modeling was employed for automatic, real-time interrogation and correction of amperometric electrochemical gas sensors (AES) deployed in instruments for industrial hygiene applications. The interrogation involved two steps which were analyzed using various mathematical modeling techniques to determine the “goodness” of the sensor and to correct its output for small changes due to aging and environmental conditions. The first interrogation was a physical interrogation of the flow path into the sensor by supplying a driving force at the face of the sensor, either by exhaled human breath, or by acoustic sound pressure applied to diffusion barriers protecting the sensor, ensuring that the flow path was sufficiently open for detection of hazardous gases. The second interrogation was an electronic interrogation of the sensor itself, achieved by applying a small potential or current pulse to the working electrode of the sensor and deconvoluting its response to make small corrections to the sensor signal. These two steps provide periodic, real-time, and automatic surveillance and correction of AES, increasing robustness, reliability and accuracy, and user confidence, while decreasing calibration frequency, maintenance, and cost-of-ownership. Validation testing over sensors of various designs indicated a greater than 98% reliability of the sensor/instrument interrogation technique.