Search citation statements
Paper Sections
Citation Types
Year Published
Publication Types
Relationship
Authors
Journals
The integration of analytical functions into machine learning-based engine models represents a significant advancement in predictive performance and operational efficiency. This paper focuses on the development of hybrid approaches to model engine combustion and temperature indices and on the synergistic effects of combining traditional analytical methods with modern machine learning techniques (such as artificial neural networks) to enhance the accuracy and robustness of such models. The main innovative contribution of this paper is the integration of analytical functions to improve the extrapolation capabilities of the data-driven models. In this work, it is demonstrated that the integrated models achieve superior predictive accuracy and generalization performance across dynamic engine operating conditions, with respect to purely neural network-based models. Furthermore, the analytical corrective functions force the output of the complete model to follow a physical trend and to assume consistent values also outside the domain of values assumed by the input features during the training procedure of the neural networks. This study highlights the potential of this integrative approach based on the implementation of the effects superposition principle. Such an approach also allows us to solve one of the intrinsic issues of data-driven modeling, without increasing the complexity of the training data’s collection and pre-processing.
The integration of analytical functions into machine learning-based engine models represents a significant advancement in predictive performance and operational efficiency. This paper focuses on the development of hybrid approaches to model engine combustion and temperature indices and on the synergistic effects of combining traditional analytical methods with modern machine learning techniques (such as artificial neural networks) to enhance the accuracy and robustness of such models. The main innovative contribution of this paper is the integration of analytical functions to improve the extrapolation capabilities of the data-driven models. In this work, it is demonstrated that the integrated models achieve superior predictive accuracy and generalization performance across dynamic engine operating conditions, with respect to purely neural network-based models. Furthermore, the analytical corrective functions force the output of the complete model to follow a physical trend and to assume consistent values also outside the domain of values assumed by the input features during the training procedure of the neural networks. This study highlights the potential of this integrative approach based on the implementation of the effects superposition principle. Such an approach also allows us to solve one of the intrinsic issues of data-driven modeling, without increasing the complexity of the training data’s collection and pre-processing.
<div>A previously developed piston damage and exhaust gas temperature models are coupled to manage the combustion process and thereby increasing the overall energy conversion efficiency. The proposed model-based control algorithm is developed and validated in a software-in-the-loop simulation environment, and then the controller is deployed in a rapid control prototyping device and tested online at the test bench. In the first part of the article, the exhaust gas temperature model is reversed and converted into a control function, which is then implemented in a piston damage-based spark advance controller. In this way, more aggressive calibrations are actuated to target a certain piston damage speed and exhaust gas temperature at the turbine inlet. A more anticipated spark advance results in a lower exhaust gas temperature, and such decrease is converted into lowering the fuel enrichment with respect to the production calibrations. Moreover, the pollutant emissions associated with production calibrations and the implementation of the developed controller are compared through a GT-Power combustion model.</div> <div>Finally, the complete controller is validated for both the transient and steady-state conditions, reproducing a real vehicle maneuver at the engine test bench. The results demonstrate that the combination of an accurate estimation of the damage induced by knock and the value of the exhaust gas temperature allows to reduce the brake specific fuel consumption by up to 20%. Moreover, the stoichiometric area of the engine operating field is extended by 20%, and the GT-Power simulations show a maximum CO reduction of about 50%.</div>
<div class="section abstract"><div class="htmlview paragraph">It is well known that ethanol can be used in spark-ignition (SI) engines as a pure fuel or blended with gasoline. High enthalpy of vaporization of alcohols can affect air-fuel mixture formation prior to ignition and may form thicker liquid films around the intake valves, on the cylinder wall and piston crown. These liquid films can result in mixture non-homogeneities inside the combustion chamber and hence strongly influence the cyclic variability of early combustion stages. Starting from these considerations, the paper reports an experimental study of the initial phases of the combustion process in a single cylinder SI engine fueled with commercial gasoline and anhydrous ethanol, as well as their blend (50%<sub>vol</sub> alcohol). The engine was optically accessible and equipped with the cylinder head of a commercial power unit for two-wheel applications, with the same geometrical specifications (bore, stroke, compression ratio). Ultra-violet (UV) natural emission spectroscopy measurements ranging from 250nm to 470nm wavelength and simultaneous thermodynamic analysis were used to better understand the effect of ethanol content on flame kernel inception and development. All experiments were conducted at wide open throttle (WOT), with stoichiometric air-fuel mixtures, fixing the engine speed at 2000rpm. Optical investigations allowed to follow the evolution of chemical species that marked the spark discharge (cyano CN and hydroxyl OH radicals) as well as flame front initial growth (OH and carbyne CH radicals). Vibrational and rotational temperatures were calculated during the arc and glow phase by the ratio between the emission intensity of CN and OH radicals. Results were compared with adiabatic flame temperature traces obtained by applying a two zone model.</div></div>
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.