A Study Employing the Taguchi Approach, Regression Analysis, and Artificial Neural Network to Analyze the Performance of a Single-Cylinder Piston-Type Compressed Air Engine

Researchers are looking for new energy sources or ways to optimize energy use because of the overuse of fossil fuels. In order to lessen the amount of carbon footprint in the atmosphere, a lot of research is being done on creating more environmentally friendly and efficient methods of producing the energy needed for propulsion. In order to create a cleaner, more sustainable, and emission-free future, this work presents a physical model of a Compressed Air Engine (CAE), which is established from an existing 4-stroke, 100-CC petrol engine. To achieve performance on CAE, a traditional four-stroke internal combustion engine that had been scrapped was converted to a two-stroke compressed air engine. Through experimental research, the output torque and brake power are determined. Through the use of experiment design, on a test bench of the piston type compressed air engine, the brake power has been examined under various loading scenarios. The L9 Taguchi method is utilized for parametric analysis and experimentation in experiment design. Regression analysis was used in the development of the mathematical model. Confirmation runs show that the results derived from Artificial Neural Networks (ANNs) and regression analysis match with the optimized values. The findings demonstrate that CAE operates economically at 5 kg/cm2 of pressure and can produce up to 0.93 kW of power at varying loads of 1.5 kg to 2.5 kg. Major Findings: The findings demonstrate that the compressed air engine which was a 4-stroke single-cylinder internal combustion engine was converted into a 2-stroke air engine by modifying the cam operates economically at 5 kg/cm2 of pressure and can produce up to 0.93 kW of power at varying loads of 1.5 kg to 2.5 kg.