Modeling with Multilayer Perceptron for Detection of Fuel Adulteration using Python Programming

Abstract

Vimal Babu U, Ramakrishan M, Nagamani M, Anjaneya Prasad P, Tagore Finny P, Kumar Goud E, Sai srujana T and Saidi Reddy M

Adulteration of fuel is introduction of an unknown substance into motor spirit unlawfully or not permitted resulting the product does not conform to the needs and specifications. Normally cheaper boiling point range hydrocarbons having more or less similar composition are added as additives leading to alter and degrade the quality of the base fuels. This method is adopted by the trading community for their quick illegal profits. This is coming as tail pipe exhaust in automobile lead to environmental pollution as well as human hazard. Ethanol and methanol added illegally to increase octane levels caused fuel pipes to leak exhaust. In order to detect the pollutants there shall be proper way both at laboratory level as well as statute. Artificial Neural Networks technique to analyze the fuel adulteration is a precise technique than any other existing methods. The gasoline, hydrocarbon fractions are detected at the in-situ with the help of Internet of Things and can be controlling through the remote and that data can be collected through the smattering. This data will help in the finding the impurities in the gasoline, diesel pollutants released into the air from tailpipe exhaust. So in this paper we are using some advance computational technique called Multilayer perceptron (MPL) to identify the impurities in the fuels. This will reduce the global warming and toxic diseases. Multilayer perceptron (MLP) is one of the most efficient techniques for Detection of fuel adulterants; MLP is class of feed forward in the artificial neural network. It consists of Three layers i.e., input layer, hidden layer and output layer. For the recognition and 3D objects estimation from a 2D single perspective view Multilayer perceptron is used.

PDF