Analysis of Parameters Influence in a MOX Gas Sensor Model

Metal oxide (MOX) gas sensors present great potential for early detection of dangerous substances or undesirable environmental conditions in various industries for their excellent cost-benefit. Even though some works show promising results regarding MOX sensors’ sensitivity and suitability, this type of device is heavily influenced by environmental conditions, leading to low selectivity and preventing application on non-controlled atmospheres. In this paper, we aim to address that issue by analyzing how the uncertainties of a MOX gas sensor’s parameters impact the model output. This study is done through uncertainty propagation analyses, using both the general uncertainty propagation equation and the Monte Carlo method, which we applied to an already validated MOX model. Thus, we make possible an analytical understanding of the sensor sensitivity to each one of the model parameters. Results show that sensor measurements are more sensitive to model parameters than to gases concentrations.