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 CCl4 gas from the trend of the 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 CCl4 with an average error of 1.5 %, before the sensor response time. The results show that the appropriateness of the system is observed.
