Prediction of Efficient Operating Conditions Inside a Heavy-Duty Natural Gas Spark Ignition Engine Using Artificial Neural Networks

2021 ◽  
Author(s):  
Jinlong Liu ◽  
Cosmin Dumitrescu ◽  
Christopher Ulishney
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Natural Gas ◽  
Spark Ignition ◽  
Operating Conditions ◽  
Spark Ignition Engine ◽  
Heavy Duty ◽  
Artificial Neural

On the use of artificial neural networks to model the performance and emissions of a heavy-duty natural gas spark ignition engine

2021 ◽  
pp. 146808742110344
Author(s):  
Qiao Huang ◽  
Jinlong Liu ◽  
Christopher Ulishney ◽  
Cosmin E Dumitrescu
Keyword(s):  
Neural Network ◽  
Natural Gas ◽  
Combustion Process ◽  
Spark Ignition ◽  
Spark Ignition Engine ◽  
Heavy Duty ◽  
Network Algorithm ◽  
Performance And Emissions ◽  
Neural Network Algorithm ◽  
Artificial Neural

The use of computational models for internal combustion engine development is ubiquitous. Numerical simulations using simpler to complex physical models can predict engine’s performance and emissions, but they require large computational capabilities. By comparison, statistical methodologies are more economical tools in terms of time and resources. This paper investigated the use of an artificial neural network algorithm to simulate the nonlinear combustion process inside the cylinder. Three engine control variables (i.e. spark timing, mixture equivalence ratio, and engine speed) were set as the model inputs. Outputs included peak cylinder pressure and its location, maximum pressure rise rate, indicated mean effective pressure, ignition lag, combustion phasing, burn duration, exhaust temperature, and engine-out emissions (i.e. nitrogen oxides, carbon monoxide, and unburned hydrocarbons). Eighty percent of the experimental data from a heavy-duty natural gas spark ignition engine were utilized to train the model. The perceptions accurately learned the combustion characteristics and predicted engine responses with acceptable errors, evidenced by close-to-unity coefficient of determination and close-to-zero root-mean-square error. Moreover, the regressors captured the effect of key operating variables on the engine response, suggesting the well-trained models successfully identified the complex relationships and can help assist engine analysis. Overall, the neural network algorithm was appropriate for the application investigated in this study.

Experimental Investigation of a Heavy-Duty Compression-Ignition Engine Retrofitted to Natural Gas Spark-Ignition Operation

2019 ◽  
Vol 141 (11) ◽  
Author(s):  
Jinlong Liu ◽  
Hemanth Kumar Bommisetty ◽  
Cosmin Emil Dumitrescu
Keyword(s):  
Natural Gas ◽  
Spark Ignition ◽  
Operating Conditions ◽  
Compression Ignition ◽  
Heavy Duty ◽  
Compression Ignition Engine ◽  
Engine Operation ◽  
Cycle Variation ◽  
Spark Timing ◽  
Ci Engines

Heavy-duty compression-ignition (CI) engines converted to natural gas (NG) operation can reduce the dependence on petroleum-based fuels and curtail greenhouse gas emissions. Such an engine was converted to premixed NG spark-ignition (SI) operation through the addition of a gas injector in the intake manifold and of a spark plug in place of the diesel injector. Engine performance and combustion characteristics were investigated at several lean-burn operating conditions that changed fuel composition, spark timing, equivalence ratio, and engine speed. While the engine operation was stable, the reentrant bowl-in-piston (a characteristic of a CI engine) influenced the combustion event such as producing a significant late combustion, particularly for advanced spark timing. This was due to an important fraction of the fuel burning late in the squish region, which affected the end of combustion, the combustion duration, and the cycle-to-cycle variation. However, the lower cycle-to-cycle variation, stable combustion event, and the lack of knocking suggest a successful conversion of conventional diesel engines to NG SI operation using the approach described here.

Performance Estimation of Direct Methanol Fuel Cell Using Artificial Neural Network

2015 ◽  
Author(s):  
M. A. Rafe Biswas ◽  
Melvin D. Robinson
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Networks ◽  
Fuel Cell ◽  
Direct Methanol Fuel Cell ◽  
Cell Voltage ◽  
Operating Conditions ◽  
Direct Methanol ◽  
Artificial Neural ◽  
Methanol Fuel Cell

A direct methanol fuel cell can convert chemical energy in the form of a liquid fuel into electrical energy to power devices, while simultaneously operating at low temperatures and producing virtually no greenhouse gases. Since the direct methanol fuel cell performance characteristics are inherently nonlinear and complex, it can be postulated that artificial neural networks represent a marked improvement in performance prediction capabilities. Artificial neural networks have long been used as a tool in predictive modeling. In this work, an artificial neural network is employed to predict the performance of a direct methanol fuel cell under various operating conditions. This work on the experimental analysis of a uniquely designed fuel cell and the computational modeling of a unique algorithm has not been found in prior literature outside of the authors and their affiliations. The fuel cell input variables for the performance analysis consist not only of the methanol concentration, fuel cell temperature, and current density, but also the number of cells and anode flow rate. The addition of the two typically unconventional variables allows for a more distinctive model when compared to prior neural network models. The key performance indicator of our neural network model is the cell voltage, which is an average voltage across the stack and ranges from 0 to 0:8V. Experimental studies were carried out using DMFC stacks custom-fabricated, with a membrane electrode assembly consisting of an additional unique liquid barrier layer to minimize water loss through the cathode side to the atmosphere. To determine the best fit of the model to the experimental cell voltage data, the model is trained using two different second order training algorithms: OWO-Newton and Levenberg-Marquardt (LM). The OWO-Newton algorithm has a topology that is slightly different from the topology of the LM algorithm by the employment of bypass weights. It can be concluded that the application of artificial neural networks can rapidly construct a predictive model of the cell voltage for a wide range of operating conditions with an accuracy of 10−3 to 10−4. The results were comparable with existing literature. The added dimensionality of the number of cells provided insight into scalability where the coefficient of the determination of the results for the two multi-cell stacks using LM algorithm were up to 0:9998. The model was also evaluated with empirical data of a single-cell stack.

Laminar flame speed correlations for methane, ethane, propane and their mixtures, and natural gas and gasoline for spark-ignition engine simulations

2017 ◽  
Vol 18 (9) ◽  
pp. 951-970 ◽  
Author(s):  
Riccardo Amirante ◽  
Elia Distaso ◽  
Paolo Tamburrano ◽  
Rolf D Reitz
Keyword(s):  
Natural Gas ◽  
Laminar Flame ◽  
Spark Ignition ◽  
Flame Speed ◽  
Operating Conditions ◽  
Spark Ignition Engine ◽  
Computational Time ◽  
Laminar Flame Speed ◽  
Spark Ignition Engines ◽  
Empirical Correlations

The laminar flame speed plays an important role in spark-ignition engines, as well as in many other combustion applications, such as in designing burners and predicting explosions. For this reason, it has been object of extensive research. Analytical correlations that allow it to be calculated have been developed and are used in engine simulations. They are usually preferred to detailed chemical kinetic models for saving computational time. Therefore, an accurate as possible formulation for such expressions is needed for successful simulations. However, many previous empirical correlations have been based on a limited set of experimental measurements, which have been often carried out over a limited range of operating conditions. Thus, it can result in low accuracy and usability. In this study, measurements of laminar flame speeds obtained by several workers are collected, compared and critically analyzed with the aim to develop more accurate empirical correlations for laminar flame speeds as a function of equivalence ratio and unburned mixture temperature and pressure over a wide range of operating conditions, namely [Formula: see text], [Formula: see text] and [Formula: see text]. The purpose is to provide simple and workable expressions for modeling the laminar flame speed of practical fuels used in spark-ignition engines. Pure compounds, such as methane and propane and binary mixtures of methane/ethane and methane/propane, as well as more complex fuels including natural gas and gasoline, are considered. A comparison with available empirical correlations in the literature is also provided.

Lean-Burn Characteristics of a Heavy-Duty Diesel Engine Retrofitted to Natural Gas Spark Ignition

2018 ◽  
Author(s):  
Jinlong Liu ◽  
Cosmin E. Dumitrescu
Keyword(s):  
Natural Gas ◽  
Combustion Chamber ◽  
Spark Ignition ◽  
Operating Conditions ◽  
Combustion Stability ◽  
Fuel Injector ◽  
Heavy Duty ◽  
Lean Burn ◽  
Spark Timing ◽  
Heavy Duty Diesel

Increased utilization of natural-gas (NG) in the transportation sector can decrease the use of petroleum-based fuels and reduce greenhouse-gas emissions. Heavy-duty diesel engines retrofitted to NG spark ignition (SI) can achieve higher efficiencies and low NOx, CO, and HC emissions when operated under lean-burn conditions. To investigate the SI lean-burn combustion phenomena in a bowl-in-piston combustion chamber, a conventional heavy-duty direct-injection CI engine was converted to SI operation by replacing the fuel injector with a spark plug and by fumigating NG in the intake manifold. Steady-state engine experiments and numerical simulations were performed at several operating conditions that changed spark timing, engine speed, and mixture equivalence ratio. Results suggested a two-zone NG combustion inside the diesel-like combustion chamber. More frequent and significant late burn (including double-peak heat release rate) was observed for advanced spark timing. This was due to the chamber geometry affecting the local flame speed, which resulted in a faster and thicker flame in the bowl but a slower and thinner flame in the squish volume. Good combustion stability (COVIMEP < 3 %), moderate rate of pressure rise, and lack of knocking showed promise for heavy-duty CI engines converted to NG SI operation.

scholarly journals Modelling Acetification with Artificial Neural Networks and Comparison with Alternative Procedures

Processes ◽  
2020 ◽  
Vol 8 (7) ◽  
pp. 749 ◽  
Author(s):  
Jorge E. Jiménez-Hornero ◽  
Inés María Santos-Dueñas ◽  
Isidoro García-García
Keyword(s):  
Neural Networks ◽  
Acetic Acid ◽  
Artificial Neural Networks ◽  
First Principles ◽  
Goodness Of Fit ◽  
Black Box ◽  
Operating Conditions ◽  
Box Models ◽  
Artificial Neural ◽  
Black Box Models

Modelling techniques allow certain processes to be characterized and optimized without the need for experimentation. One of the crucial steps in vinegar production is the biotransformation of ethanol into acetic acid by acetic bacteria. This step has been extensively studied by using two predictive models: first-principles models and black-box models. The fact that first-principles models are less accurate than black-box models under extreme bacterial growth conditions suggests that the kinetic equations used by the former, and hence their goodness of fit, can be further improved. By contrast, black-box models predict acetic acid production accurately enough under virtually any operating conditions. In this work, we trained black-box models based on Artificial Neural Networks (ANNs) of the multilayer perceptron (MLP) type and containing a single hidden layer to model acetification. The small number of data typically available for a bioprocess makes it rather difficult to identify the most suitable type of ANN architecture in terms of indices such as the mean square error (MSE). This places ANN methodology at a disadvantage against alternative techniques and, especially, polynomial modelling.

Sign in / Sign up

Export Citation Format

Share Document