THE CREATION OF A DIESEL ENGINE USING GENETIC ALGORITHM PREDICTION MODELS AND MACHINE LEARNING TO VERIFY FUEL CONSUMPTION, EMISSIONS, AND HEAT TRANSFER

Abstract

The advancement of diesel engine design has become increasingly reliant on computational techniques to optimize performance while minimizing fuel consumption, emissions, and heat transfer losses. This research explores the integration of genetic algorithm (GA) prediction models and machine learning (ML) techniques to enhance diesel engine efficiency. The study employs GA to generate optimal engine configurations by simulating various design parameters, selecting the most efficient solutions based on performance criteria. Machine learning models are then trained on experimental and simulated data to validate the accuracy of fuel consumption, emission levels, and thermal characteristics. The combined approach enables real-time optimization by iteratively refining engine parameters through adaptive learning. By leveraging data-driven methodologies, this study provides a novel framework for designing high-efficiency diesel engines with lower environmental impact. The results demonstrate that ML-assisted GA optimization significantly improves fuel economy and emission compliance while enhancing thermal performance. This research contributes to the on-going efforts in sustainable engine development, offering a predictive and adaptable design methodology for future diesel engines. Traditional methods rely on empirical correlations or computational fluid dynamics (CFD) simulations, which are computationally expensive and time-consuming. This thesis explores the application of machine learning (ML) techniques to predict and analyze heat transfer in IC engines. By leveraging historical engine performance data and modern ML algorithms, we aim to provide an efficient, accurate, and scalable solution for IC engine heat transfer analysis. The findings demonstrate the potential of ML approaches to enhance the efficiency of engine development cycles.