Prediction of Co Concentration in Different Conditions Based on Gaussian-Tcn

Abstract

Accurate prediction of gas concentrations can help people to know if there are hazardous gases in the environment, thus protecting people's health. The performance of the electronic nose (E-nose) in predicting gas concentrations depends on its algorithm. In recent years, recurrent networks have demonstrated their suitability for processing time series data. Typical models are  Short Term Memory (LSTM) neural networks and Gated Recursive Units (GRU). But traditional recurrent neural networks use a gate structure to drop some information randomly, which can cause important features to be discarded. The temporal convolutional neural Network (TCN) based on convolutional structures is more suitable for time series prediction. To enhance the accurate model forecasts,  the key parameters in the model are first optimized in this paper. We then adapted the residual structure of TCN and replaced the activation function of TCN with Gaussian error linear units (GELUs), And the modified model is named Gaussian-TCN. Experimental results show that the proposed Gaussian-TCN outperforms traditional recurrent network algorithms in terms of prediction accuracy.

Keywords: gas concentration prediction, Electronic nose, TCN, GELU