Application of computational intelligence in classifying chemical gases using electronic nose technology

• Main Author: Illahi, Ana Antoniette C.
• Format: text
• Language: English
• Published: Animo Repository 2023
• Subjects: Computational intelligence; Olfactory sensors; Electrical and Electronics; Systems and Communications
• Online Access: https://animorepository.dlsu.edu.ph/etdd_ece/6
• Institution: De La Salle University

The classification of chemical gases is a critical issue in various industries, such as food, healthcare, and environmental monitoring. The development of electronic nose (e-nose) technology has provided a cost-effective and non-invasive solution for gas detection. In this research need to know the effectiveness of using computational intelligence (CI) techniques for classifying chemical gases using e-nose technology. Developing an e-nose system consisting of a sensor array and an algorithm based on CI techniques. The sensor array was designed to detect volatile organic compounds (VOCs) in the gas samples, and the CI algorithm was used to classify the samples based on their gas concentration. SVM, FFNN, and GPR is use for regression analysis while LSTM, BILSTM, and GRU is used for classification task. The performance of the developed e-nose system was evaluated using a real-world dataset of six different chemical gases namely Carbon Monoxide Toluene, Methane, Ammonia, Ethanol, and Isobutylene. The results showed that the developed system achieved high classification accuracy, with LSTM and BILSTM, and GRU achieving accuracy rates of 97.14%, 98.29%, and 98.8%, respectively. E-nose technology has potential applications in various industries that require gas detection and classification. The results of this research demonstrate the effectiveness of using CI techniques in developing e-nose systems for gas classification. Future work can focus on improving the performance of the system by optimizing the sensor array design and exploring other machine learning algorithm.