| Summary: | In an ion-sensitive field-effect transistor (ISFET) sensor, the ions within the sample media undergo multiple environments influenced reactions occurring molecules from these reactions to accumulate upon the gate oxide layer. The change in charge affects the conductance in the ISFET channels; consequently, the changes of conductance within the source and the drain will produce an electrical signal. The most common problem is drift happens when the electrical signal output gradually changes independent of the measured sample. The primary goal of this study is to investigate a reliable artificial neural network model to classify and predict the error of low-pressure chemical vapour deposition SixNy ISFET pH sensor and implement the drift compensation. Such models could be later used to encounter the drift problems that usually exist in chemical sensors. Three units of ISFET sensors were used to calibrate with three types of pH buffer solutions viz. pH 4, pH 7 and pH 10. Artificial neural networks were applied to construct black-box multiple-input multiple-output models of the ISFET data where the percentage accuracy value was used to assess the model’s performances in classifying while the mean squared error (MSE) and the coefficient of determination (R2) parameter used in determining the best models in predicting the error in the ISFET sensors. Concerning the model structure in classification, Pattern Recognition Neural Network (PATTERNNET) proved to perform better than Function Fitting (FITNET) networks with 100% accuracy. The network configuration in PATTERNNET, a dual-layered network with 30 nodes on the first hidden layer and 3 nodes on the second hidden layer achieved the best results. As for the prediction, the NARX-BR model with 75 delays produce an efficient model in predicting the error of ISFET data set. The value of MSE = 4.8814e-5 and R2 = 0.99930 for the NARX-BR model revealed that the model capable in predicting the error. The drift compensation applied and the drift issues in the ISFET sensors has successfully solved. As a result, this study demonstrates great potential in developing artificial neural networks to stave off the drift issues in ISFET low-pressure chemical vapour deposition SixNy ISFET pH sensor.
|
|---|