Experimental analysis and parameter optimization on the reduction of NOx from diesel engine using RSM and ANN Model

Area of Science:

• Environmental Engineering
• Chemical Engineering
• Plasma Science

Background:

• Nitrogen oxides (NOx) are major air pollutants from combustion sources like vehicles and power plants.
• Nonthermal plasma (NTP) technology, particularly Dielectric Barrier Discharge (DBD) reactors, offers a promising method for NOx conversion.
• Optimizing NTP parameters is crucial for efficient NOx removal from exhaust gases.

Purpose of the Study:

• To investigate the feasibility of a DBD reactor-based NTP process for removing NOx from diesel engine exhaust.
• To optimize NTP operating parameters including NOx concentration, gas flow rate, applied plasma voltage, and electrode gap.
• To compare the predictive accuracy of Response Surface Methodology (RSM) and Artificial Neural Network (ANN) models for NOx removal and energy efficiency.

Main Methods:

• Utilized a Dielectric Barrier Discharge (DBD) reactor for nonthermal plasma (NTP) generation.
• Investigated four operating parameters: NOx concentration (300-400 ppm), gas flow rate (2-6 lpm), applied plasma voltage (20-30 kVpp), and electrode gap (3-5 mm).
• Employed Box-Behnken design (BBD) for RSM and Artificial Neural Network (ANN) for process optimization and performance comparison using R², MSE, RMSE, and MAPE metrics.

Main Results:

• Both RSM and ANN models showed strong agreement with experimental results.
• The ANN model demonstrated superior performance over RSM in predicting NOx removal efficiency and energy efficiency, indicated by lower error metrics (MSE, RMSE, MAPE) and higher R² values.
• Experimental validation at optimal RSM conditions achieved a maximum NOx reduction of 60.5% and an energy efficiency of 66.24 g/J.

Conclusions:

• The Artificial Neural Network (ANN) model is more accurate than Response Surface Methodology (RSM) for predicting NOx removal and energy efficiency in NTP processes.
• The DBD-based NTP technology is feasible for effective NOx reduction from diesel engine exhaust.
• Further research can leverage ANN models for enhanced optimization of plasma-based pollution control technologies.