An intelligent gas concentration estimation system using neural network implemented microcontroller

  • Authors:

  • Ali Gulbag;Fevzullah Temurtas

  • Affiliations:

  • Institute of Science & Technology, Sakarya University, Adapazari, Turkey;Department of Computer Engineering, Sakarya University, Adapazari, Turkey

  • Venue:
  • AI'04 Proceedings of the 17th Australian joint conference on Advances in Artificial Intelligence
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

The use of microcontroller in neural network realizations is cheaper than those specific neural chips In this study, an intelligent gas concentration estimation system is described A neural network (NN) structure with tapped time delays was used for the concentration estimation of CCI4 gas from the trend of transient sensor responses After training of the NN, the updated weights and biases were applied to the embedded neural network implemented on the 8051 microcontroller The microcontroller based gas concentration estimation system performs NN based concentration estimation, the data acquisition and user interface tasks This system can estimate the gas concentrations of CCI4 with an average error of 1.5 % before the sensor response time The results show that the appropriateness of the system is observed.