Abstract:Pattern recognition is the core component of the electronic nose (E-nose). Traditional machine learning algorithms highly rely on the feature data selected manually for model training and testing. A complete experiment must be performed before the data can be further processed. To realize the automatic extraction of response features and simplify the model's training and application process, a multi-task convolutional neural network (MTL-CNN) with a dual-block knowledge-sharing structure is designed to train a model for the E-nose system. This model can simultaneously perform three different classification tasks, for the purposes of target discrimination, concentration prediction, and state judgment. Only a few consecutive seconds of response data are needed to be input into the trained model to obtain various information about the E-nose. With the utilization of an unmanned gas-sensing test system, large-scale measurements of the E-nose can be carried out automatically. A baseline tracking algorithm (BTA) is designed based on the relative changes of short-term data, reducing the impact of long-term shifts. Over thousands of gas response processes and more than 10 million sensing data have participated in the training of the deep learning model. The 5-fold cross-validation method shows that the fully trained model has an outstanding generalization performance. After the baseline is tracked automatically, the accuracy of three tasks towards 12 kinds of volatile organic compounds (VOCs) is about 95% (type recognition: 95.2%, concentration prediction: 92.1%, status judgment: 97.3%) using only 4 seconds of sensing data during the response status of the E-nose. Our work shows the distinct advantages of combining "big data" and "deep learning" in the gas-sensing field and further proves that the application of MTL-CNN can significantly improve the training and application efficiency of the E-nose.

Active Instance Selection for Drift Calibration of an Electronic Nose

Boost AI Power: Data Augmentation Strategies with Unlabeled Data and Conformal Prediction, a Case in Alternative Herbal Medicine Discrimination with Electronic Nose

Conformal Prediction Based On K-Nearest Neighbors For Discrimination Of Ginsengs By A Home-Made Electronic Nose

Advances in drift compensation algorithms for electronic nose technology

A pattern recognition method for electronic noses based on an olfactory neural network

Gas identification with drift counteraction for electronic noses using augmented convolutional neural network

An Effective Approach to Handling Noise and Drift in Electronic Noses.

Online Sensor Drift Compensation for E-Nose Systems Using Domain Adaptation and Extreme Learning Machine

Open Set Domain Adaptation for Electronic Nose Drift Compensation on Uncertain Category Data

Covariance Matrix Adaptation Evolutionary Strategy for Drift Correction of Electronic Nose Data

Wasserstein Distance Learns Domain Invariant Feature Representations for Drift Compensation of E-Nose

Conformal predictions for electronic nose system: An application to tea classification

An Empirical Evaluation of Meta Residual Network for Classifying Sensor Drift Samples

Unsupervised Attention-Based Multi-Source Domain Adaptation Framework for Drift Compensation in Electronic Nose Systems

Online Drift Compensation Framework Based on Active Learning for Gas Classification and Concentration Prediction

Domain adaptive subspace transfer model for sensor drift compensation in biologically inspired electronic nose

A Novel Label Disentangling Subspace Learning Based on Domain Adaptation for Drift E-nose Data Classification

Anti-Drift in Electronic Nose via Dimensionality Reduction: A Discriminative Subspace Projection Approach

Drift Compensation for Electronic Nose by Semi-Supervised Domain Adaption

Target discrimination, concentration prediction, and status judgment of electronic nose system based on large-scale measurement and multi-task deep learning

A Novel Detection Framework Via Drift Compensation for Inter Board Differences