Support vector machine for a diesel engine performance and NOx emission control-oriented model. Issue 2 (2020)

• Record Type: Journal Article
• Authors: Aliramezani, Masoud; Norouzi, Armin; Koch, Charles Robert
• Is Part Of: IFAC-PapersOnLine. Volume 53:Issue 2(2020)
• Page Start: 13976
• Page End: 13981
• Publication Date: 2020
• Subjects: Control Oriented Mode -- Machine Learning -- Support Vector Machine (SVM) -- Diesel engine -- Emissions; Automatic control -- Periodicals; 629.805
• DOI: 10.1016/j.ifacol.2020.12.916
• Languages: English
• ISSNs: 2405-8963

Abstract

Abstract: A control oriented diesel engine NOx emission and Break Mean Effective Pressure (BMEP) model is developed using Support Vector Machine (SVM). Steady state experimental data from a medium duty diesel engine is used to develop BMEP and NOx emission model using Support Vector Machine (SVM). The engine speed, the amount of injected fuel and the injection rail pressure are used as input variables to predict the steady state engine NOx emission and BMEP. The steady state model results were then implemented in the control oriented model. A fast response electrochemical NOx sensor is used to experimentally study the engine transient NOx emission and to verify the transient response of the control oriented model. The results show that the SVM algorithm is capable of accurately learning the engine BMEP and NOx which improves the accuracy of the control oriented model compared to a conventional regression algorithm (trust-region) used in the literature. The control oriented model results closely match the experiments in both transient and steady state conditions with a root mean square error of 0.26 (bar) and 10 (ppm) for BMEP and NOx respectively.