A reaction model of metal oxide gas sensors and a recognition method by pattern matching

The metal oxide (MOX) gas sensors have distinct response patterns with different gas sorts. In this paper, a reaction model of MOX gas sensors is represented to simulate the sensor response patterns. Six model parameters of each sensor responding to each gas sort could be calculated out in the sensor response pattern simulations. The sensor response curves also could be reconstructed by the model with the six parameters. In this way, a performance database of sensors could be built with these model parameters. A pattern matching method based on the sensor performance database was also proposed for gas sort classification without any usual pattern recognition methods. Eleven volatile organic compounds (VOCs) were measured by a gas sensor array of 6 TGS gas sensors for model validation. The mean simulation error of all sensors was 3.98%. The performance database of these sensors was well built by model calculation. The correct recognition ratio of the pattern matching method was 100%. The results indicated that the proposed reaction model might have many advantages in real applications. The work in this paper is a step towards a general applicable model for MOX gas sensors.