Nonlinearity compensation of DIC-based multi-sensor measurement

The development of an efficient framework with an advanced algorithm for single or multi-sensors measurement using non-ADC (nADC) based direct interfacing circuits (DICs) is essential for better understanding of underlying phenomena. The major issue in the DIC system is the significant level of nonlinearity due to external DIC components and the counter of the microcontroller. In this paper, a set of error compensating techniques (ECT) were investigated and the best model was implemented for single sensor measurement which reduced the nonlinearity to 0.02%. Under the improved ECT framework, the algorithm was extended for multi-sensor measurement in an E-Nose setup. A neural network model was integrated with the system which was implemented using a microcontroller for effective discrimination of various gases. The efficacy of the proposed model of ECT was enlightened by comparing with ADC based approach.