Prediction of Autoignition and Flame Properties for Multicomponent Fuels Using Machine Learning Techniques

Shah, Neel; Zhao, Peng; DelVescovo, Dan; Ge, Haiwen

doi:10.4271/2019-01-1049

Cited by 10 publications

(4 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To exemplify, Shah N. et al developed deep neural networks and random forest algorithms to predict ignition delay times, flame speeds, octane ratings, and crank angle with 50% of total heat release in HCCI engine conditions (CA50) values. 22 In the experimental part of this study, the EGR heat exchanger was removed from the main engine cooling system and replaced by a novel proposed EGR cooling system consisting of a controllable 12 V dc pump and fan to set the determined exhaust gas temperatures entering the intake manifold. The effects of EGR gas temperature transferred to the intake manifold under different engine speed and load conditions on NOx and BSFC were investigated by the proposed EGR cooling system.…”

Section: Introductionmentioning

confidence: 99%

The design and development of a diesel engine electromechanical EGR cooling system based on machine learning-genetic algorithm prediction models to reduce emission and fuel consumption

Kaleli

Akolaş

2021

Proceedings of the Institution of Mechanical Engineers, Part C:

View full text Add to dashboard Cite

Data-driven modelling techniques have recently been used in the development of engine design and control systems for defining the engine in-cylinder complex combustion process. The aim of this investigation is to improve exhaust emissions and fuel consumption by designing an electromechanical exhaust gas recirculation (EGR) cooling system consisting of an electric water pump and fan unlike conventional systems. To determine the effects of the EGR ratio and the temperature of the exhaust gas entering the intake manifold on the diesel engine parameters of nitrogen oxides (NOx) emission and brake specific fuel consumption (BSFC), four learning (ML) algorithms were adapted according to statistical evaluation criteria such as root mean squared error (RMSE), coefficient of determination (R2), mean squared error (MSE) and mean absolute error (MAE). The hyper parameters of the selected best model among four learning algorithms were determined by using grid search method. The results showed that the Gaussian process regression model (GPR) outperformed other ML models according to success error prediction of NOx and BSFC. Then, performance of the designed electromechanical EGR cooling system was analyzed under global driving conditions, the New European driving cycle (NEDC), and the world-wide harmonized light duty test procedure (WLTP). In these test cycles, global optimization was utilized with the GPR model as the objective function based on minimizing NOx and BSFC. Consequently, this study demonstrates the potential of the proposed system based on ML-GA to reduce NOx and BSFC by achieving reductions of 13.7%(NEDC)–9.98%(WLTP) and 2.61%(NEDC)–2.07%(WLTP) in NEDC and WLTP conditions, respectively compared to the conventional EGR cooling approach.

show abstract

Section: Introductionmentioning

confidence: 99%

The design and development of a diesel engine electromechanical EGR cooling system based on machine learning-genetic algorithm prediction models to reduce emission and fuel consumption

Kaleli

Akolaş

2021

Proceedings of the Institution of Mechanical Engineers, Part C:

View full text Add to dashboard Cite

show abstract

“…Badra et al (2020) optimized the combustion system of a gasoline compression ignition engine using CFD and machine learning-grid gradient algorithm. Shah et al (2019) used machine learning techniques to predict the ignition delay, flame speed, octane rating, and combustion phasing of multicomponent gasoline surrogates in homogenous charge compression ignition engines.…”

Section: Introductionmentioning

confidence: 99%

CFD Optimization of the Pre-Chamber Geometry for a Gasoline Spark Ignition Engine

Bakir

Yadav

et al. 2021

Front. Mech. Eng.

Self Cite

View full text Add to dashboard Cite

In the present paper, an efficient optimization method based on Bayesian updating strategy is developed for the design of a spark-ignition engine equipped with pre-chamber. 3D computational fluid dynamics (CFD) simulation coupled with strategies including design of experiment, genetic algorithm, and machine learning methods is used to optimize the pre-chamber with desired combustion phasing. The optimization process starts from a design of experiment matrix of 11 design parameters, which are used to analytically characterize the pre-chamber geometry and set up the 3D combustion CFD. Taking CA50 as the single objective, the CFD results are then used to train the machine learning models. Different machine learning models are evaluated based on their Root Mean Square Error. Five machine learning models from five different categories are selected for second round evaluation. The trained machine learning model is used in the genetic algorithm optimization, which yields the optimized configuration and is again justified by CFD. The new CFD results based on the optimized design are added into the database to further refine the machine learning model. After 24 iterations for each selected machine learning models, the medium Gaussian support vector machine model is identified as the best method for the present application. Iterations using the medium Gaussian support vector machine model continue until a satisfactory result is achieved. Detailed combustion analysis is conducted to investigate the physical mechanism about how the design of pre-chamber influences the engine's performance. It is found that larger volume of the upper part of the pre-chamber results in stronger jet flow and turbulent intensity which further accelerates the flame propagation inside the pre-chamber, dominating the contrary effects from reduced pressure and temperature. Regression analysis shows that the radius of the pre-chamber is the most influential design parameter. The current work not only sheds light on the optimization of engine design, but also has demonstrated a general strategy applicable to the purpose of arbitrary engine optimization and mechanical system design.

show abstract

“…Apart from these studies, other studies include modeling and simulation of wind turbine on rotor performance by predicting power and torque characteristics, which focus on the use of heuristically methods, including artificial neural network (ANN) (Kalogirou, 2000;Sargolzaei & Kianifar, 2009). The data-driven modelling and prediction methods based on machine learning (ML) have emerged as a tool for catching the complex dynamic systems (Hafner & Isermann, 2003;Shah, Zhao, Delvescovo, & Ge, 2019). The designed ML model for the system can be considered as an input-output function learning directly from attributes of experimental data (Russell & Norvig, 2016) The purpose of this study is implementing different ML algorithms and comparing their rotor torque prediction performance of wind turbines such as Savonius and vertical turbines.…”

Section: Introductionmentioning

confidence: 99%

Torque Prediction Based Performance Analysis of Small-Scale Wind Turbines Using Data Driven Modelling Methods

Kaleli

Yalçıner

2021

Erzincan Üniversitesi Fen Bilimleri Enstitüsü Dergisi

View full text Add to dashboard Cite

In this study, the rotor torque of wind turbines has been predicted using machine learning approach based on real time data which have been collected for the designed small scale Savonius and four leaves rotors. The tip speed ratio (TSR) has been selected as the main input parameter in machine learning modelling technique which are linear regression (LR), support vector machine (SVM) regression and Gaussian process (GP) regression. The hyperparameter of these models have been defined by grid search method. RMSE, determination coefficient, MSE and MAE have been used to evaluate the predictive performance of the models to experimental data. The rotor torque modelling results show the efficiency of wind turbines can be maximized with high estimation accuracy of models. On the other hand, it has been also observed that torque of the Savonius type wind turbine is higher than the four leaves turbine

show abstract

Prediction of Autoignition and Flame Properties for Multicomponent Fuels Using Machine Learning Techniques

Cited by 10 publications

References 19 publications

The design and development of a diesel engine electromechanical EGR cooling system based on machine learning-genetic algorithm prediction models to reduce emission and fuel consumption

The design and development of a diesel engine electromechanical EGR cooling system based on machine learning-genetic algorithm prediction models to reduce emission and fuel consumption

CFD Optimization of the Pre-Chamber Geometry for a Gasoline Spark Ignition Engine

Torque Prediction Based Performance Analysis of Small-Scale Wind Turbines Using Data Driven Modelling Methods

Contact Info

Product

Resources

About