engleski

Application of machine learning and Box-Behnken design in optimization of engine characteristics operated with a dual-fuel mode of algal biodiesel and waste-derived biogas

Waste-derived biogas and third-generation algal biodiesel are attractive alternative fuels to substitute fossil diesel in a diesel engine. However, the use of biodiesel as pilot liquid fuel and biogas as the main fuel in a diesel engine is a complex and extremely non-linear process. The current research aims to forecast and optimize the combustion and exhaust emission parameters of a variable compression dual-fuel combustion engine. Experiment data were collected at multiple engine loads, compression ratio, pilot fuel injection pressure, and timings. Using the experimental data, a multi-layer perceptron network was used to create an Artificial Neural Network (ANN) based prognostic model. Brake thermal efficiency, flow rates of biogas, peak in-cylinder pressure, carbon dioxide, unburnt hydrocarbons, oxides of nitrogen, and carbon monoxidewere all predicted using the established prognostic model. The predictive model's robustness is demonstrated by statistical metrics such as R (0.9723 – 0.988) and R2 (0.9453 – 0.9761), Nash-Sutcliffe model efficiency (94% – 97%), and mean absolute percentage error (0.013 – 0.128), Kling-Gupta efficiency (0.9548 – 0.9836), Theil's U2 model uncertainty (0.162 – 0.368) were used to validate the generated model, demonstrating good model quality. The Multi-Output Response Surface Methodology (RSM) was used to optimize the parameters of dual-fuel combustion. Using the desirability technique, the trade-off analysis between emission and efficiency showed that 84% engine load, 244 bar of fuel injection pressure, 28 oCA bTDC of injection timing, and 17.5 of compression ratio are the best operation conditions. The RSM research results were confirmed using an experimental investigation, and errors were less than 9%. It was revealed that ANN-linked RSM is a good hybrid method for modeling, prediction, and optimization of a dual- fuel engine's performance.

Biofuel ; biogas, prediction model ; optimization ; neural networks ; engine behaviors

nije evidentirano