Paper Info
Title
Optimal Control of a Variable Geometry Turbocharged Diesel Engine Using Neural Networks: Applications on the ETC Test Cycle
Abstract
Modern diesel engines are typically equipped with variable geometry turbo-compressor, exhaust gas recirculation (EGR) system, common rail injection system, and post-treatment devices in order to increase their power while respecting the emis- sions standards. Consequently, the control of diesel engines has become a difficult task involving five to ten control variables that interact with each other and that are highly nonlinear. Actually, the control schemes of the engines are all based on static lookup tables identified on test-benches; the values of the control variables are interpolated using these tables and then, they are corrected, online, by using the control techniques in order to obtain better engine's response under dynamic conditions. In this paper, we are interested in developing a mathematical optimization process that search for the optimal control schemes of the diesel engines under static and dynamic conditions. First, we suggest modeling a turbocharged diesel engine and its opacity using the mean value model which requires limited experiments; the model's simula- tions are in excellent agreement with the experimental data. Then the created model is integrated in a dynamic optimization process based on the Broyden-Fletcher-Goldfarb-Shanno (BFGS) algo- rithm. The optimization results show the reduction of the opacity while enhancing the engine's effective power. Finally, we proposed a practical control technique based on the neural networks in order to apply these control schemes online to the engine. The neural controller is integrated into the engine's simulations and is used to control the engine in real time on the European tran- sient cycle (ETC). The results confirm the validity of the neural controller. making the search for the optimal control schemes highly com- plex and time consuming. Actually, the control of the engine is based on two dimensional static lookup tables with inputs the crankshaft angular speed and the fuel flow rate, these tables, called basic lookup tables, cover the whole functioning area of the engine and are identified from steady state experiments using experimental optimization process. The primary values of the control variables are computed using these maps. Then these values are adjusted online using corrected lookup tables that take into consideration the changes in the engine's environ- ment and the evolution of the engine's states variables detected in real time using different sensors. Finally, in order to meet the emissions standards and to increase the efficiency of the engine and enhance its response under dynamic charge, the corrected values are modified using the control techniques (6)-(8). These methods are characterized by control parameters that are tuned experimentally on a dynamic test-bench. Therefore, the control of the engine is clearly very difficult and time consuming; it requires steady state and dynamic experiments and highly depends upon the judgment of the experimental staff where we cannot eliminate the possibility for human errors especially with the rising number of the control variables. Consequently, the need of a reliable optimization tool has become a necessity that occupied the engines' producers for the last two decades. Such a process requires the buildup of consistent engine and emissions models to replace the expensive experiments and to predict the engine's response when varying the control variables. In this paper, we presented the principal control algorithms of the engine and we discussed the different strategies com- monly used to control the air management system. Then, we proposed a new methodology to control the engine. First, we used the mean value model to describe the engine's states vari- ables and exhaust emissions (9)-(12); the model is based on the ideal gas state equation, the mass and energy conservation prin- ciples, the fundamental principle of the dynamic and semi-em- pirical equations that describe the relations between different states variables of the engine. We adopted this modeling tech- nique because it is precise and simple enough to be used in a mathematical optimization process, and it requires limited ex- perimental data to identify the model's parameters. Then the engine's model is integrated in a dynamic optimization process based on the BFGS algorithm (13). Our objective is to find the optimal values of the engine's control variables over dynamic courses in order to create dynamic lookup tables that can be di- rectly applied to the corresponding actuators. This optimization
Year
DOI
Venue
2009
10.1109/TCST.2008.2001049
IEEE Trans. Contr. Sys. Techn.
Keywords
Field
DocType
Optimal control,Geometry,Diesel engines,Neural networks,Table lookup,Humans,Fuels,Steady-state,System testing,Rails
Intelligent control,Internal combustion engine,Optimal control,Fuel injection,Control theory,Turbocharger,Control engineering,Exhaust gas recirculation,Control variable,Diesel engine,Geometry,Mathematics
Journal
Volume
Issue
ISSN
17
2
1063-6536
Citations 
PageRank 
References 
8
1.02
5
Authors
3
Name
Order
Citations
PageRank
Rabih Omran181.02
Rafic Younes2265.62
Jean-Claude Champoussin381.02