In this paper different neural networks (NN) are compared for modelling a variable valve timing spark-ignition (VVT SI) engine. The overall system is divided for each output into five neural multi-input single output (MISO) subsystems. Three kinds of NN, multilayer Perceptron (MLP), pseudo-linear radial basis function (PLRBF), and local linear model tree (LOLIMOT) networks, are used to model each subsystem. Real data were collected when the engine was under different operating conditions and these data are used in training and validation of the developed neural models. The obtained models are finally tested in a real-time online model configuration on the test bench. The neural models run independently of the engine in parallel mode. The model outputs are compared with process output and compared among different models. These models performed well and can be used in the model-based engine control and optimization, and for hardware in the loop systems.
HolzmannH.HalfmannCh.IsermannR.Neuro-fuzzy modelling of automotive SI-engine characteristics. In Proceedings of IFAC Symposium onAdvances in Automotive Control 98, Mohican State Park, Loudonville, Ohio, USA, 26 February-1 March 1998, 156–160.
YuD. L.YuD. U.GommJ. B.WilliamsD.Model predictive control of a chemical process based on an adaptive neural network. Proceedings of IFAC 15th World Congress, Barcelona, Spain, 21–26 July 2002.
6.
NellesO.Non-linear System Identification with Local Linear Neuro-Fuzzy Models, Shaker Verlag, Aachen, Germany, 1999.
7.
NellesO.Non-linear System Identification, (Springer, Berlin), 2001.
8.
NellesO.On the identification with neural networks as series-parallel and parallel models. International Conference on Artificial Neural Networks, Paris, October 1995.
9.
GommJ. B.WilliamsD.EvansJ. T.DohertyS. K.LisboaP. J. G.Enhancing the non-linear modelling capabilities of MLP neural networks using spread encoding, Fuzzy Sets and Systems, 1996, 79 (1), 113–126.
10.
YuD. L.WilliamsD.GommJ. B.On-line implementation of a neural network model predictive controller, Control Systems Research Group. Proceedings of the IEE Workshop onPractical experience with predictive control, Middlesbrough, UK, 16 February 2000.
11.
LightbodyG.IrwinG. W.Direct neural model reference adaptive controlIEE Proceedings Control Theory Appl., 1995, 142 (1), 31–43.
12.
MüllerR.HembergerH.Neural adaptive ignition control, SAE Special Publication, 1998, 1356, 97–102.
13.
HafnerM.SchülerM.NellesO.Neural net models for diesel engines—simulation and exhaust optimization. In European Congress onIntelligent Techniques and Soft Computing, EUFIT 98, 7–10 September, Aachen, Germany, 1998.
14.
HolzmannH.HalfmannCh.IsermannR.Neuro-fuzzy modelling of automotive SI-engine characteristics. IFAC Workshop onAdvances in Automotive Control, Mohican State Park, Ohio, USA, 27 February-1 March 1998.
KlütingM.FlierlR.UngerH.PoggelJ.Throttle-free load control through fully variable valve train systems—thermodynamics and technology, 7, Aachener Kolloquium Fahrzeug- und Motorentechnik, 1998.
17.
EschT.PinschingerM.GöbelT.Luft- und Kraftstoffzumessung bei Ottomotoren mit variabler Ventilsteuerung, MTZ Motortechnische Zeitschrift 57, 1996, No. 2.
18.
ChenS.BillingsS. A.Practical identification of NARMAX models using radial basis functions, Int. J. Control, 1990, 52 (6), 1327–1350.
19.
ChenS.BillingsS. A.Representations of non-linear systems: The NARMAX model, Int. J. Control, 1989, 49 (3), 1013–1032.
20.
NellesO.ErnstS.IsermannR.Neuronale Netze zur Identifikation nichtlinearer, dynamischer Systeme: Ein Überblick, Automatisierungstechnik 45, 1997, 6, 251–262.
