MODEL PREDICTIVE CONTROL STRATEGY FOR ENERGY OPTIMIZATION IN CNG-CONVERTED VEHICLES IN NIGERIA

Abstract

The increasing adoption of Compressed Natural Gas (CNG) as an alternative transportation fuel in Nigeria, catalyzed by the removal of petroleum subsidies and the implementation of the Presidential Compressed Natural Gas Initiative (PCNGI), has accelerated the conversion of petrol-powered vehicles using aftermarket conversion kits. However, many converted vehicles experience poor fuel economy, unstable engine performance, and inconsistent Air-Fuel Ratio (AFR) regulation due to inadequate calibration methods. Most existing calibration practices rely on manual tuning and static mapping techniques that cannot effectively adapt to varying driving conditions. This study presents a Model Predictive Control (MPC)-based strategy for optimizing fuel consumption and AFR regulation in CNG-converted vehicles. A dynamic model of the CNG engine system was developed based on relationships among air intake, fuel injection, and combustion. The model was integrated into an MPC framework capable of predicting future engine behavior and optimizing injection timing to satisfy AFR and operational constraints. The study adopts a Python-based simulation environment utilizing Jupyter Notebook, Streamlit, SciPy, and optimization libraries. The proposed controller is evaluated against a conventional Proportional-Integral-Derivative (PID) controller using fuel consumption, AFR tracking accuracy, response time, overshoot, and control effort as performance indices. The study demonstrates that MPC provides an 8.6% improvement in fuel efficiency, a 61% faster response time (12.4 s vs. 31.8 s), and tighter AFR regulation ( vs. AFR units) across varying operating conditions, thereby supporting Nigeria's transition toward cleaner, more sustainable transport energy systems. The paper recommends the adoption of optimization in the calibration process to ensure  more efficient CNG converted vehicles