QUANTIFICATION OF METHANE (MARSH) GAS USING RESILIENT BACKPROPAGATION NEURAL NETWORK
Authors/Creators
- 1. Department of Electronics and Communication Engineering, Rajiv Gandhi University Arunachal Pradesh 791112, India
Description
This study unequivocally demonstrates that robust propagation neural networks can be used to quantify marsh or methane gases. Methane has lower explosive limit (LEL) is 5.0% (5000 ppm) and its upper explosive limit is 15.0% (150000 ppm) by volume with air. The main purposed of this study is to classify the concentration level below the LEL of the methane gas present in the air by using the resilient back-propagation neural network to prevent the various hazardous caused by the methane. After the suggested network was trained with defaults free parameters, it provided a very high quantification rate with two hidden layers, and tan-sigmoid activation functions. The neural network with twelve neurons in the first hidden layer and fifteen hidden neurons in the second hidden layers gives first converging rate with more accuracy. From the regression graph shows the training, validation, testing, and overall data set display high regression values that are very close to 1. The regression values quantify how closely the network outputs resemble the targets. The training, validation, testing, and overall regression values are 0.97, 1%, 0.97, and 0.98 correspondingly. This indicates that the suggested network is 98% accurate overall.
Files
April 2023-151-154.pdf
Additional details
References
- 1. Anthony, Katey Walter, MacIntyre, Sally (2016). Nocturnal escape route for marsh gas. Nature. 535 (7612): 363–365. 2. B. Mondal, M.S. Meetei, J. Das, C. Roy Chaudhuri, H. Saha (2015). Quantitative recognition of flammable and toxic gases with artificial neural network using metal oxide gas sensors in embedded platform. Engineering Science and Technology, an International Journal. 18(2):229-234. 3. Simon Haykin, (2011). Neural Networks and Learning Machines. PHI Learning Private Limited. Third Edition. 4. T. Jayalakshmi, A. Santhakumaran (2011). Statistical Normalization and Back Propagation for Classification. International Journal of Computer Theory and Engineering. 3(1) ):89-93. 5. Guang-Bin Huang (2003). Learning Capability and Storage Capacity of Two-Hidden-Layer Feed forward Networks. IEEE Transactions On Neural Networks. 14(2):274 - 281. 6. D. Stathakis (2009). How many hidden layers and nodes?. International Journal of Remote 7. Sensing. 30(8): 2133–2147. 8. Ritaban Dutta, K.R. Kashwan, M. Bhuyan, E.L. Hines, J.W. Gardner (2003). Electronic nose based tea quality standardization. Neural Networks. 16(5-6): 847-853. 9. Shubhneesh Kumar, Maibam Sanju Meetei(2014). Forecast and Analysis of Short Term Electric Load of New South Wales Region using ANN. International journal of engineering research & technology. 3(6): 1669-1671. 10. Kurt Hornik, Maxwell Stinchcombe, Halbert White(1989). Multilayer feedforward networks are universal approximators. Neural Networks. 2(5):359-366. 11. Howard Demuth, Mark Beale. Neural Network Toolbox. The MathWorks. Version 4.