Embedded system to recognize the heat power of a fuel gas and to classificate the quality of alcohol fuel

This work presents the result obtained to develop an electronic nose to recognize the fuel gas heat power. As a first approach, synthetic data was generated for each sensor. It was considered the use of raw data and the use of a principal component analysis (PCA) to reduce the number of sensors. Two topologies of neural networks have been used, the backpropagation and learning vector quantization (LVQ). A fuzzy inference system (FIS) also has been used as a solution to this problem.