Improvement of Neural Network Accuracy for Engine Simulations

Brahma, Indranil; Rutland, Christopher J.

doi:10.4271/2003-01-3227

Cited by 34 publications

(15 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Applying the principle of energy conservation to the crankshaft gives (15) is the moment of inertia of the engine, it is a periodic function of the crankshaft angle due to the repeated motion of the pistons and connecting rods, but for simplicity, in this paper, Fig. 5.…”

Section: B Exhaust Manifoldmentioning

confidence: 99%

Optimal Control of a Variable Geometry Turbocharged Diesel Engine Using Neural Networks: Applications on the ETC Test Cycle

Omran

Younès

Champoussin

2009

IEEE Trans. Contr. Syst. Technol.

View full text Add to dashboard Cite

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 emissions 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 simulations 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) algorithm. 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 transient cycle (ETC). The results confirm the validity of the neural controller.Index Terms-Diesel engines, dynamic optimization process, mean value model, optimal control and neural networks.

show abstract

Section: B Exhaust Manifoldmentioning

confidence: 99%

Optimal Control of a Variable Geometry Turbocharged Diesel Engine Using Neural Networks: Applications on the ETC Test Cycle

Omran

Younès

Champoussin

2009

IEEE Trans. Contr. Syst. Technol.

View full text Add to dashboard Cite

show abstract

“…Accurate predictions are usually achieved by utilising back propagation neural networks with Levenberg-Marquardt (LM) and Scaled Conjugate Gradient (SCG) as the learning algorithm (Lucas et al 2001;Brahma, He, and Rutland 2003;Cortes, Urquiza and Hernández 2009;Ghobadian et al 2009;Obodeh and Ajuwa 2009;Kiani et al 2010;Togun and Baysec 2010;Moradi et al 2013). Typically, the activation functions for hidden layer was set to be logsigmoid (logsig) and output layer was set as linear (purelin) (Ghobadian et al 2009;Kiani et al 2010).…”

Section: Introductionmentioning

confidence: 99%

A neural network model for the prediction of compression ignition engine performance at different injection timings

Kullolli

Sakthivel

Ilangkumaran³

2014

International Journal of Ambient Energy

View full text Add to dashboard Cite

Rapid depletion of fossil fuel and continuous increase in gasoline prices have stimulated the search of alternative fuels. This paper deals with the prediction of engine performance, emission and combustion characteristics of compression ignition engine fuelled with fish oil biodiesel using artificial neural network (ANN). Experimental investigations are carried out in a single cylinder constant speed direct injection diesel engine under variable load conditions at different injection timings −21 0 , 24 0 and 27 0 bTDC. The performance, combustion and emission characteristics are measured using an exhaust gas analyser, smoke meter, piezoelectric pressure transducer and crank angle encoder for different fuel blends and engine load conditions. For training the neural network, feed-forward back propagation algorithm is used. The developed ANN model predicts the performance, combustions and exhaust emissions with a correlation coefficients (R) of 0.97-0.99 and a mean relative error of 0.62-4.826%. The root mean square errors are found to be low. The developed model has found to predict accurately the engine performance, combustion and emission parameters at different injection timings.

show abstract

“…A structured test plan or DOEs are not usually required and any amount of nonlinearity can be handled. A lot of work [40][41][42][43] has gone into utilizing the power of neural networks for engine modeling and control. Qiang et al 44 46 Brahma et al, 47 and Bayesian techniques 48 as well as Gaussian processes 49 50 has dealt with modeling for steady-state calibration in which the authors compared empirical methods for steady-state engine modeling: global regression, localized regression, and neural networks (found to work best) and presented a new variant of localized regression.…”

Section: Introductionmentioning

confidence: 99%

Analysis and prediction of transient opacity spikes using dimensional modeling

Brahma

2013

International Journal of Engine Research

Self Cite

View full text Add to dashboard Cite

Dimensional modeling, GT-Power in particular, has been used for two related purposes-to quantify and understand the inaccuracies of transient engine flow estimates that cause transient smoke spikes and to improve empirical models of opacity or particulate matter used for engine calibration. It has been proposed by dimensional modeling that exhaust gas recirculation flow rate was significantly underestimated and volumetric efficiency was overestimated by the electronic control module during the turbocharger lag period of an electronically controlled heavy duty diesel engine. Factoring in cylinder-to-cylinder variation, it has been shown that the electronic control module estimated fuel-Oxygen ratio was lower than actual by up to 35% during the turbocharger lag period but within 2% of actual elsewhere, thus hindering fuel-Oxygen ratio limit-based smoke control. The dimensional modeling of transient flow was enabled with a new method of simulating transient data in which the manifold pressures and exhaust gas recirculation system flow resistance, characterized as a function of exhaust gas recirculation valve position at each measured transient data point, were replicated by quasi-static or transient simulation to predict engine flows. Dimensional modeling was also used to transform the engine operating parameter model input space to a more fundamental lower dimensional space so that a nearest neighbor approach could be used to predict smoke emissions. This new approach, intended for engine calibration and control modeling, was termed the ''nonparametric reduced dimensionality'' approach. It was used to predict federal test procedure cumulative particulate matter within 7% of measured value, based solely on steady-state training data. Very little correlation between the model inputs in the transformed space was observed as compared to the engine operating parameter space. This more uniform, smaller, shrunken model input space might explain how the nonparametric reduced dimensionality approach model could successfully predict federal test procedure emissions when roughly 40% of all transient points were classified as outliers as per the steady-state training data.with steady-state data has been used for this purpose as an alternative to direct measurements of dynamic flow. Further GT-Power predictions have been used to propose a new empirical modeling method that could be useful for engine calibration and engine controls architecture design. This new approach, termed the

show abstract

Improvement of Neural Network Accuracy for Engine Simulations

Cited by 34 publications

References 11 publications

Optimal Control of a Variable Geometry Turbocharged Diesel Engine Using Neural Networks: Applications on the ETC Test Cycle

Optimal Control of a Variable Geometry Turbocharged Diesel Engine Using Neural Networks: Applications on the ETC Test Cycle

A neural network model for the prediction of compression ignition engine performance at different injection timings

Analysis and prediction of transient opacity spikes using dimensional modeling

Contact Info

Product

Resources

About