In-Car Modelling of Emissions with Dynamic Artificial Neural Networks

Hentschel, Robert; Cernat, RA; Varchmin, J.-U.

doi:10.4271/2001-01-3383

Cited by 9 publications

(2 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Neural networks [30][31][32] have been employed successfully for a variety of engine modelling problems [33][34][35][36][37][38][39][40][41][42][43][44][45]. Neural networks are powerful function approximators and are prime candidates for steady state engine calibration models.…”

Section: Neural Network Modelsmentioning

confidence: 99%

Development of the nearest neighbour multivariate localized regression modelling technique for steady state engine calibration and comparison with neural networks and global regression

Brahma

Sharp

Richter

et al. 2008

International Journal of Engine Research

View full text Add to dashboard Cite

A new modelling technique has been developed to aid steady state diesel engine calibration by accurately predicting engine system response and emissions at steady state operating conditions. This new modelling technique, referred to as the nearest neighbour multivariate localized regression (NNMLR) in this work, is built upon the particular localized regression technique for multiple independent variables developed by M. C. Sharp et al. at Cummins Inc., Columbus and referred to as the multivariate localized regression (MLR) technique in this work. Among other advantages, MLR has been demonstrated to generalize better than other similar data-driven empirical modelling techniques such as global regression. Although MLR has been proven and tested, it is computationally expensive which makes it unwieldy with current optimization schemes, particularly random search methods; NNMLR is significantly faster than MLR. Additionally, the primary localization parameter which remains fixed in MLR over the entire dataset is allowed to vary over the dataset in NNMLR. Therefore, in NNMLR it is the dataset that decides the degree of localization. This variable degree of localization is unique to each response being modelled and can change with data density and with the complexity of the relationship between the dependent and independent variables. This greater degree of localization makes NNMLR slightly more accurate than the current MLR scheme. The motivation for developing NNMLR was to reduce optimization routine runtimes while maintaining accuracy at MLR levels or even improving on these levels. The other modelling approach that promises similar or better accuracy levels and model runtimes is neural networks. Neural networks are increasingly being applied over a wide range of applications, hence they were examined for their suitability for modelling engine responses. Their robustness to noise, training data requirements, and ability to extrapolate were compared with localized and global regression. In the following sections, after background that describes the MLR technique, the algorithm for NNMLR is described. Some heuristics are unavoidable and these have been discussed and justified. Results have been presented for many datasets of different sizes over different sets of dependent variables, comparing MLR, NNMLR, global regression, and neural networks. In addition to performance over standard engine responses, test functions have been used to create pseudo-responses, so that noise can be added in a controlled manner. Accuracy and robustness of MLR, NNMLR, global regression, and neural networks over different levels of noisy data have been studied. The study shows how neural networks work better than other approaches for steady state calibration in terms of mean accuracy, maximum error, robustness to noise, training data size, ability to handle non-linearities as well as ability to extrapolate. However, localized regression (MLR and NNMLR) produces comparable performance and the reasons why a regression-based approac...

show abstract

Section: Neural Network Modelsmentioning

confidence: 99%

Development of the nearest neighbour multivariate localized regression modelling technique for steady state engine calibration and comparison with neural networks and global regression

Brahma

Sharp

Richter

et al. 2008

International Journal of Engine Research

View full text Add to dashboard Cite

show abstract

“…In-car digital measurement data processing system along with a dynamic ANN was made by Hentschel et al[33] as a vehicle emission modeling, employing the advantage of vehicle-specific parameters. They implanted a constant volume sampling by using the FSN (Filter Smoke Number) method that represented an integral method of measurement.Their claim was to consider the vehicle-specific parameters in a diesel engine car by using the dynamic ANN.…”

mentioning

confidence: 99%

Effects of artificial neural networks characterization on prediction of diesel engine emissions

Tehranian¹

View full text Add to dashboard Cite

Effects of Artificial Neural Networks Characterization on Prediction of Diesel Engine Emissions Azadeh Tehranian More than a century after its invention, diesel remains the fuel of choice for buses and freight trucks. Diesel exhaust contains three gases that are regulated by the United States Environmental Protection Agency (EPA), as well as particulate matter (PM). There is a societal need both to lower emissions and to predict or model emissions more accurately for inventory purposes. Engine modeling, and real time control are the most indispensable steps towards lowering engine emissions, and it is argued that this modeling can be achieved by implementation of Artificial Neural Networks (ANN). Effects of ANN design, architecture, and learning parameters on the accuracy of emissions predictions were studied along with the variation of embedded activation functions. An optimization strategy was followed to attain the most suitable network in the defined framework for five emissions of NO x , PM, HC, CO, and CO 2. The emissions data were obtained from five engine transient test schedules, namely the E-CSHVR, ETC, FTP, E-Highway and E-WVU-5 Peak schedules. These were performed on a 550 hp General Electric DC engine dynamometer-testing unit at the West Virginia University Alternative Fuels, Engine and Emissions Research Center. The 3-Layer and Jump Connection networks were the most promising architectures and it was found that the radial basis functions such as the Gaussian and Gaussian Complement functions outperform the sigmoidal functions in all of the examined architectures. The accuracy of an excellent typical instance of CO 2 prediction was as good as 0.009% error of accumulated value over the course of a FTP cycle. DEDICATION To Afsaneh, for her profound empathy, To Nick and Roya, for their remarkable nobility, To Nasser, for his exclusive sincerity. iii WBP Weight associated with back propagation algorithm ŷ ANN prediction of variable y xxii 157 .

show abstract

Improvement of Neural Network Accuracy for Engine Simulations

Brahma¹,

Rutland²

2003

SAE Technical Paper Series

View full text Add to dashboard Cite

In-Car Modelling of Emissions with Dynamic Artificial Neural Networks

Cited by 9 publications

References 1 publication

Development of the nearest neighbour multivariate localized regression modelling technique for steady state engine calibration and comparison with neural networks and global regression

Development of the nearest neighbour multivariate localized regression modelling technique for steady state engine calibration and comparison with neural networks and global regression

Effects of artificial neural networks characterization on prediction of diesel engine emissions

Improvement of Neural Network Accuracy for Engine Simulations

Contact Info

Product

Resources

About