Zhang, S, Lin, M, Zou, X, Su, S, Zhang, W, Zhang, X & Guo, Z 2020, 'LSTM-based air quality predicted model for large cities in China', Nature Environment and Pollution Technology, vol. 19, no. 1, pp. 229-236.
© 2020 Technoscience Publications. All rights reserved. In this paper, the LSTM model is used to predict the PM2.5 concentrations in five representative Chinese cities with the GDP exceeding 1 trillion Yuan, including Beijing, Chengdu, Shanghai, Shenzhen and Wuhan. The PM2.5 concentration data in 2015-2017 are selected for training, and the results are optimized to achieve an efficient solution by adjusting the parameters. Based on the optimized solution, a test is carried out to predict the PM2.5 concentration in 2018, and the results are compared with the real value obtained from the monitoring centre. According to the comparison results, the correlation coefficient of Wuhan and Chengdu is 0.86724 and 0.80070, which are the highest in these five cities. While the correlation coefficient of Shenzhen and Shanghai, are 0.78225, 0.72147, Beijing, as the capital city of China achieved the lowest correlation coefficient which is 0.64118. The LSTM-based predictive model has relatively good reliability and transferability. More effective predictive results can be achieved by implementing deep learning to analyse PM2.5 concentration.
Liu, T, Zhang, W, Yuwono, M, Zhang, M, Ueland, M, Forbes, SL & Su, SW 2020, 'A data-driven meat freshness monitoring and evaluation method using rapid centroid estimation and hidden Markov models', SENSORS AND ACTUATORS B-CHEMICAL, vol. 311.View/Download from: Publisher's site
Liu, T, Zhang, W, Ye, L, Ueland, M, Forbes, SL & Su, SW 2019, 'A novel multi-odour identification by electronic nose using non-parametric modelling-based feature extraction and time-series classification', Sensors and Actuators, B: Chemical, vol. 298.View/Download from: Publisher's site
© 2019 Elsevier B.V. The electronic nose (e-nose) is an olfaction system that consists of an array of chemical sensors and effective machine learning algorithms for the detection of various target odours. Feature extraction and classification methods are of great importance in improving the performance of the e-nose system. In this paper, a novel odour identification method is presented. Firstly, we use the kernel-based system modelling approach to extract odour features. Its solution is a series of finite impulse responses which containing discriminant information of different odours. In addition, a parameter optimisation method based on normalised mean square error and information entropy is proposed to optimise the kernel function. The entropy is effective in preventing the finite impulse responses from overfitting. Multi-odour classification is achieved based on Gaussian mixture density hidden Markov model (GMM-HMM) considering the characteristic of the extracted features. Also, parameter selection for GMM-HMM is realised according to BIC index and cross-validation. Then, we validate the performance of the proposed feature extraction method in resistance to noise and compare it with other existed features. The modelling-based feature reached the highest performance even without applying any filtering or smoothing techniques. Finally, we compare the proposed combination of feature extraction and classification algorithms with other approaches. The proposed method outperformed other approaches reaching 93.56% in sensitivity and 98.71% in specificity. The results demonstrate that the proposed method is applicable in e-nose-based odour identification.
Zhang, W, Liu, T, Ye, L, Ueland, M, Forbes, SL & Su, SW 2019, 'A novel data pre-processing method for odour detection and identification system', Sensors and Actuators, A: Physical, vol. 287, pp. 113-120.View/Download from: Publisher's site
© 2018 Elsevier B.V. This paper presents a novel electronic nose (E-nose) data pre-processing method, based on a recently developed non-parametric kernel-based modelling (KBM) approach. The proposed method is tested by an automated odour detection and classification system, named “NOS.E” developed by the NOS.E team in University of Technology Sydney. Experimental results show that when extracting the derivative-related features from signals collected by the NOS.E, the proposed non-parametric KBM odour data pre-processing method achieves more reliable and stable pre-processing results comparing with other pre-processing methods such as wavelet package correlation filter (WPCF), mean filter (MF), polynomial curve fitting (PCF) and locally weighted regression (LWR). Based on these derivative-related features, the NOS.E can achieve a 96.23% accuracy of classification with the popular Support Vector Machine (SVM) classifier.
Wu, Y, Liu, T, Ling, SH, Szymanski, J, Zhang, W & Su, SW 2019, 'Air Quality Monitoring for Vulnerable Groups in Residential Environments Using a Multiple Hazard Gas Detector.', Sensors, vol. 19, no. 2.View/Download from: Publisher's site
This paper presents a smart "e-nose" device to monitor indoor hazardous air. Indoor hazardous odor is a threat for seniors, infants, children, pregnant women, disabled residents, and patients. To overcome the limitations of using existing non-intelligent, slow-responding, deficient gas sensors, we propose a novel artificial-intelligent-based multiple hazard gas detector (MHGD) system that is mounted on a motor vehicle-based robot which can be remotely controlled. First, we optimized the sensor array for the classification of three hazardous gases, including cigarette smoke, inflammable ethanol, and off-flavor from spoiled food, using an e-nose with a mixing chamber. The mixing chamber can prevent the impact of environmental changes. We compared the classification results of all combinations of sensors, and selected the one with the highest accuracy (98.88%) as the optimal sensor array for the MHGD. The optimal sensor array was then mounted on the MHGD to detect and classify the target gases without a mixing chamber but in a controlled environment. Finally, we tested the MHGD under these conditions, and achieved an acceptable accuracy (70.00%).
Liu, T, Zhang, W, McLean, P, Ueland, M, Forbes, SL & Su, SW 2018, 'Electronic Nose-Based Odor Classification using Genetic Algorithms and Fuzzy Support Vector Machines', International Journal of Fuzzy Systems, vol. 20, no. 4, pp. 1309-1320.View/Download from: Publisher's site
© 2018, Taiwan Fuzzy Systems Association and Springer-Verlag GmbH Germany, part of Springer Nature. Electronic nose devices consisting of a matrix of sensors to sense the smell of various target gases have received considerable attention during the past two decades. This paper presents an efficient classification algorithm for a self-designed electronic nose, which integrates both genetic algorithms (GAs) and fuzzy support vector machines (FSVMs) to detect the target odor. GAs are applied to select the informative features and the optimal model parameters of FSVMs. FSVMs are adopted as fitness evaluation criterion and the sequent odor classifier, which can reduce the outlier effects and provide a robust and accurate classification. This proposed algorithm has been compared with some commonly used learning algorithms, such as support vector machine, the k-nearest neighbors and other combination algorithms. This study is based on experimental data collected from the response of the UTS NOS.E, which is the electronic nose system developed by the University of Technology Sydney NOS.E team. In comparison with other approaches, the experiment results show that the proposed odor classification algorithm can significantly improve the classification accuracy by selecting high-quality features and reach to 92.05% classification accuracy.
Zhang, W, Chiu, C, Szymanski, J, Chaczko, Z, Su, S & Zhou, J 2020, 'Internet of Things is Changing Teaching of Technical Subjects at UTS' in Klempous, R & Nikodem, J (eds), Smart Innovations in Engineering and Technology: Intelligent Engineering and Informatics, Springer Nature, Cham, Switzerland, pp. 131-143.View/Download from: Publisher's site
Zhang, W, Liu, T, Zhang, M, Zhang, Y, Li, H, Ueland, M, Forbes, SL, Rosalind Wang, X & Su, SW 2018, 'NOS.E: A New Fast Response Electronic Nose Health Monitoring System.', 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), International Conference of the IEEE Engineering in Medicine and Biology Society, IEEE, USA, pp. 4977-4980.View/Download from: Publisher's site
We present a practical electronic nose (e-nose) sys-tem, NOS.E, for the rapid detection and identification of human health conditions. By detecting the changes in the composition of an individual's respiratory gases, which have been shown to be linked to changes in metabolism, e-nose systems can be used to characterize the physical health condition. We demonstrated our system's viability with a simple data set consists of breath collected under three different scenarios from one volunteer. Our preliminary results show the popular classifier SVM can discriminate NOS.E's responses under the three scenarios with high performance. In future work, we will aim to gather a more varied data set to test NOS.E's abilities.
Zhang, W, Szymanski, J, Chiu, C, Chaczko, Z, Su, S & Zhou, J 2017, 'How the Internet of Things is Changing Teaching of Technical Subjects at UTS', 5th Asia-Pacific Conference on Computer-Aided System Engineering, Asia-Pacific Conference on Computer Aided System Engineering, APCASE Foundation, Guilin, Guangxi, China, pp. 71-77.