Cam-Phasing Optimization Using Artificial Neural Networks as Surrogate Models-Maximizing Torque Output

Wu, Bin; Prucka, Robert; Filipi, Zoran; Kramer, Denise; Ohl, Gregory L.

doi:10.4271/2005-01-3757

Cited by 25 publications

(25 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We use it to model gas dynamics in the intake and exhaust systems from the air filter to the tailpipe. SIS is a research code written in FORTRAN language that has been refined over time and used routinely at the University of Michigan for a variety of simulation studies (Filipi and Assanis, 2000;Wu et al, 2005Wu et al, , 2006. The combustion sub-model in the code is based on the turbulent flame entrainment model proposed by Tabaczynski et al (1977Tabaczynski et al ( , 1980 and further refined by Poulos and Heywood (1983).…”

Section: Introductionmentioning

confidence: 99%

Improving the predictiveness of the quasi-d combustion model for spark ignition engines with flexible intake systems

Lee

Filipi

2011

Int.J Automot. Technol.

View full text Add to dashboard Cite

Fast and predictive simulation tools are prerequisites for pursuing simulation based engine control development. A particularly attractive tradeoff between speed and fidelity is achieved with a co-simulation approach that marries a commercial gas dynamic code WAVE TM with an in-house quasi-dimensional combustion model. Gas dynamics are critical for predicting the effect of wave action in intake and exhaust systems, while the quasi-D turbulent flame entrainment model provides sensitivity to variations of composition and turbulence in the cylinder. This paper proposes a calibration procedure for such a tool that maximizes its range of validity and therefore achieves a fully predictive combustion model for the analysis of a high degree of freedom (HDOF) engines. Inclusion of a charge motion control device in the intake runner presented a particular challenge, since anything altering the flow upstream of the intake valve remains "invisible" to the zero-D turbulence model applied to the cylinder control volume. The solution is based on the use of turbulence multiplier and scheduling of its value. Consequently, proposed calibration procedure considers two scalar variables (dissipation constant C β and turbulence multiplier C M ), and the refinements of flame front area maps to capture details of the spark-plug design, i.e. the actual distance between the spark and the surface of the cylinder head. The procedure is demonstrated using an SI engine system with dual-independent cam phasing and charge motion control valves (CMCV) in the intake runner. A limited number of iterations led to convergence, thanks to a small number of adjustable constants. After calibrating constants at the reference operating point, the predictions are validated for a range of engine speeds, loads and residual fractions.

show abstract

Section: Introductionmentioning

confidence: 99%

Improving the predictiveness of the quasi-d combustion model for spark ignition engines with flexible intake systems

Lee

Filipi

2011

Int.J Automot. Technol.

View full text Add to dashboard Cite

show abstract

“…Others have carried out some form of search but this is unspecified [17], whilst some use trial and error methods [18], [19]. Of those which actually perform a search, some are quadratic [9], [10], whilst other use a linear search with equal number of nodes in the first and second hidden layers [20], [21]. Nonetheless it may be that the authors are basing their searches from past experience.…”

Section: Anecdotalmentioning

confidence: 99%

On the Optimal Node Ratio between Hidden Layers: A Probabilistic Study

Thomas¹,

Walters²,

Gheytassi³

et al. 2016

IJMLC

View full text Add to dashboard Cite

Abstract-Two-hidden layer feedforward neural networks (TLFNs) have been shown to outperform single-hidden-layer neural networks (SLFNs) for function approximation in many cases. However, their added complexity makes them more difficult to find. Given a constant number of hidden nodes n h , this paper investigates how their allocation between the first and second hidden layers (n h = n 1 + n 2 ) affects the likelihood of finding the best generaliser. The experiments were carried out over a total of ten public domain datasets with n h = 8 and 16. The findings were that the heuristic n 1 = 0.5n h + 1 has an average probability of at least 0.85 of finding a network with a generalisation error within 0.18% of the best generaliser. Furthermore, the worst case over all data sets was within 0.23% for n h = 8, and within 0.15% for n h = 16. These findings could be used to reduce the complexity of the search for TLFNs from quadratic to linear, or alternatively for 'topology mapping' between TLFNs and SLFNs, given the same number of hidden nodes, to compare their performance.Index Terms-ANN, optimal node ratio, topology mapping, two-hidden-layer feedforward, function approximation.

show abstract

“…The constraint on exhaust temperature is introduced to protect the catalyst from overheating. Thus, the optimization problem is formulated as follows [18]:…”

Section: Maximizing Engine Performancementioning

confidence: 99%

“…After calibrating the model constants with a limited number of experimental data, the simulation tool is able to predict engine behavior with sufficient fidelity for generating a robust calibration [18]. The predictiveness of physics-based models used in a high-fidelity simulation allows investigations of new system configurations, even those that have not been built yet.…”

Section: Introductionmentioning

confidence: 99%

A Simulation-Based Approach for Developing Optimal Calibrations for Engines with Variable Valve Actuation

Filipi

Prucka

et al. 2007

Oil & Gas Science and Technology - Rev. IFP

View full text Add to dashboard Cite

Résumé -Une approche basée sur la simulation pour le développement de réglages optimaux des moteurs avec système d'actionnement variable de soupape -La technologie d'actionnement variable de soupape (Variable Valve Actuation, VVA) présente un fort potentiel pour augmenter les performances et réduire consommation et pollution. Les avantages du VVA proviennent d'une meilleure aération et de la possibilité de contrôler les résidus internes. Par contre, l'addition de variables de contrôle supplémentaires dans un moteur VVA augmente la complexité du système, ce qui nécessite la mise au point d'une stratégie de contrôle optimale afin d'en tirer tous les bénéfices. L'approche traditionnelle reposant sur l'expérimentation s'avère trop onéreuse dans ce cas du fait de l'augmentation exponentielle du nombre de tests. Ce projet propose comme alternative de définir le réglage des soupapes comme un problème d'optimisation. Il identifie les simulations nécessaires au développement d'une méthode générique capable d'intégrer un nombre croissant de degrés de liberté afin de développer un outil de grande fidélité pour la prédiction des effets de différents dispositifs. Puisque la résolution d'un problème d'optimisation requiert l'évaluation de centaines de fonctions, l'utilisation directe d'une simulation hautefidélité demande des temps de calculs beaucoup trop longs. En remplacement, un réseau artificiel de neurones est entraîné avec des résultats de simulation haute-fidélité et utilisé pour représenter la réponse du moteur à différentes combinaisons de variables de contrôle, et ce avec des temps de calcul beaucoup plus courts. Cet article décrit une méthode basée sur la simulation, apporte des détails sur les techniques de modélisation haute-fidélité et démontre une application à un prototype de moteur à essence à quatre cylindres et double arbre à cames indépendant. Abstract

show abstract

Cam-Phasing Optimization Using Artificial Neural Networks as Surrogate Models-Maximizing Torque Output

Cited by 25 publications

References 58 publications

Improving the predictiveness of the quasi-d combustion model for spark ignition engines with flexible intake systems

Improving the predictiveness of the quasi-d combustion model for spark ignition engines with flexible intake systems

On the Optimal Node Ratio between Hidden Layers: A Probabilistic Study

A Simulation-Based Approach for Developing Optimal Calibrations for Engines with Variable Valve Actuation

Contact Info

Product

Resources

About