Chun Shun Cheung

Effect of Waste Cooking Oil Biodiesel on the Properties of Particulate from a DI Diesel Engine

The present work focuses on the effect of waste cooking oil biodiesel on the particulate mass, number concentration, nanostructure, and oxidative reactivity under different engine speeds and engine loads. Particulate samples were collected from the diluted exhaust of a medium-duty direct injection diesel engine and were used to analyze the physico-chemical properties via the transmission electron microscope (TEM) and the thermogravimetric analyzer/differential scanning calorimeter (TGA/DSC). The TEM images reveal that smaller primary particles are formed at higher engine speed, lower engine load, or using biodiesel. Quantitative analysis of the nanostructures indicates more soot with more disordered configuration, in which shorter and more curved graphene layer is prevailing at lower engine load or when using biodiesel. Furthermore, the TGA results infer that the soot oxidative reactivity is closely related to the nanostructure properties and the effect of engine load is more pronounced than the effect of engine speed. Also biodiesel soot has faster oxidative reactivity than diesel soot. Moreover, results obtained for B30 (30% biodiesel and 70% diesel fuel) lie in between those for biodiesel and diesel fuel. Copyright 2015 American Association for Aerosol Research

Modelling of diesel engine performance using advanced machine learning methods under scarce and exponential data set

Abstract Traditional methods on creating diesel engine models include the analytical methods like multi-zone models and the intelligent based models like artificial neural network (ANN) based models. However, those analytical models require excessive assumptions while those ANN models have many drawbacks such as the tendency to overfitting and the difficulties to determine the optimal network structure. In this paper, several emerging advanced machine learning techniques, including least squares support vector machine (LS-SVM), relevance vector machine (RVM), basic extreme learning machine (ELM) and kernel based ELM, are newly applied to the modelling of diesel engine performance. Experiments were carried out to collect sample data for model training and verification. Limited by the experiment conditions, only 24 sample data sets were acquired, resulting in data scarcity. Six-fold cross-validation is therefore adopted to address this issue. Some of the sample data are also found to suffer from the problem of data exponentiality, where the engine performance output grows up exponentially along the engine speed and engine torque. This seriously deteriorates the prediction accuracy. Thus, logarithmic transformation of dependent variables is utilized to pre-process the data. Besides, a hybrid of leave-one-out cross-validation and Bayesian inference is, for the first time, proposed for the selection of hyperparameters of kernel based ELM. A comparison among the advanced machine learning techniques, along with two traditional types of ANN models, namely back propagation neural network (BPNN) and radial basis function neural network (RBFNN), is conducted. The model evaluation is made based on the time complexity, space complexity, and prediction accuracy. The evaluation results show that kernel based ELM with the logarithmic transformation and hybrid inference is far better than basic ELM, LS-SVM, RVM, BPNN and RBFNN, in terms of prediction accuracy and training time.

DIESEL ENGINE MODELLING USING EXTREME LEARNING MACHINE UNDER SCARCE AND EXPONENTIAL DATA SETS

To predict the performance of a diesel engine, current practice relies on the use of black-box identification where numerous experiments must be carried out in order to obtain numerical values for model training. Although many diesel engine models based on artificial neural networks (ANNs) have already been developed, they have many drawbacks such as local minima, user burden on selection of optimal network structure, large training data size and poor generalization performance, making themselves difficult to be put into practice. This paper proposes to use extreme learning machine (ELM), which can overcome most of the aforementioned drawbacks, to model the emission characteristics and the brake-specific fuel consumption of the diesel engine under scarce and exponential sample data sets. The resulting ELM model is compared with those developed using popular ANNs such as radial basis function neural network (RBFNN) and advanced techniques such as support vector machine (SVM) and its variants, namely least squares support vector machine (LS-SVM) and relevance vector machine (RVM). Furthermore, some emission outputs of diesel engines suffer from the problem of exponentiality (i.e., the output y grows up exponentially along input x) that will deteriorate the prediction accuracy. A logarithmic transformation is therefore applied to preprocess and post-process the sample data sets in order to improve the prediction accuracy of the model. Evaluation results show that ELM with the logarithmic transformation is better than SVM, LS-SVM, RVM and RBFNN with/without the logarithmic transformation, regardless the model accuracy and training time.

Modelling and Prediction of Diesel Engine Performance using Relevance Vector Machine

Diesel engines are being increasingly adopted by many car manufacturers today, yet no exact mathematical diesel engine model exists due to its highly nonlinear nature. In the current literature, black-box identification has been widely used for diesel engine modelling and many artificial neural network (ANN) based models have been developed. However, ANN has many drawbacks such as multiple local minima, user burden on selection of optimal network structure, large training data size, and over-fitting risk. To overcome these drawbacks, this article proposes to apply an emerging machine learning technique, relevance vector machine (RVM), to model and predict the diesel engine performance. The property of global optimal solution of RVM allows the model to be trained using only a few experimental data sets. In this study, the inputs of the model are engine speed, load, and cooling water temperature, while the output parameters are the brake-specific fuel consumption and the amount of exhaust emissions like nitrogen oxides and carbon dioxide. Experimental results show that the model accuracy is satisfactory even the training data is scarce. Moreover, the model accuracy is compared with that using typical ANN. Evaluation results also show that RVM is superior to typical ANN approach.

Influence of waste cooking oil biodiesel on the nanostructure and volatility of particles emitted by a direct-injection diesel engine

ABSTRACT To reduce air pollution and the reliance on fossil fuel, biodiesel has been widely investigated as an alternative fuel for diesel engines. The purpose of this study is to investigate the influence of waste cooking oil (WCO) biodiesel on the physical properties and the oxidation reactivity of the particles emitted by a diesel engine operating on WCO biodiesel as the main fuel. Experiments were conducted on a direct-injection diesel engine fueled with biodiesel, B75 (75% biodiesel and 25% diesel on volume basis, v/v), B50, B20, and diesel fuel, at five engine loads and at an engine speed of 1920 rev/min. Particulate samples were collected to analyze the particulate nanostructure, volatility, and oxidation characteristics. Biodiesel or low-load operation leads to smaller primary particles and more disordered nanostructures having shorter and more curved graphene layers. It can be found that particles from biodiesel, blended fuels, or low-load operation have higher volatile mass fractions and faster oxidation reaction rates than particles from diesel or heavy-load operation. The higher oxidation reaction rates are due mainly to the smaller particle size, the more disordered nanostructure, and the higher volatile mass fraction. It is also found that changes in primary particle size and particulate nanostructure are not directly proportional to the biodiesel content, while changes in particulate volatility and particulate oxidation reactivity are proportional to the biodiesel content. The use of biodiesel can enhance particulate oxidation reactivity and the regeneration of soot particles in an after-treatment device. Copyright

Modelling and prediction of particulate matter, NO x , and performance of a diesel vehicle engine under rare data using relevance vector machine

Traditionally, the performance maps and emissions of a diesel engine are obtained empirically through many testes on the dynamometers because no exact mathematical engine model exists. In the current literature, many artificial-neural-network- (ANN-) based approaches have been developed for diesel engine modelling. However, the drawbacks of ANN would make itself difficult to be put into some practices including multiple local minima, user burden on selection of optimal network structure, large training data size, and overfitting risk. To overcome the drawbacks, this paper proposes to apply one emerging technique, relevance vector machine (RVM), to model the diesel engine, and to predict the emissions and engine performance. With RVM, only a few experimental data sets can train the model due to the property of global optimal solution. In this study, the engine speed, load, and coolant temperature are used as the input parameters, while the brake thermal efficiency, brake-specific fuel consumption, concentrations of nitrogen oxides, and particulate matter are used as the output parameters. Experimental results show the model accuracy is fairly good even the training data is scarce. Moreover, the model accuracy is compared with that using typical ANN. Evaluation results also show that RVM is superior to typical ANN approach.

Effect of biodiesel on PAH, OPAH, and NPAH emissions from a direct injection diesel engine

Polycyclic aromatic hydrocarbon (PAH), oxy- and nitro-derivate PAH (OPAH and NPAH) emissions from a direct injection diesel engine fueled with conventional fossil diesel (D), waste cooking oil biodiesel (B100), and their two blends (B20 and B50) were compared. The results show that B100 can reduce low molecular weight PAHs such as naphthalene, acenaphthylene, and fluorene as much as 90% compared with diesel. However, the emissions of high molecular weight PAHs including benzo[b]fluoranthene, benzo[k]fluoranthene, and benzo[a]pyrene decrease slightly when using B100. The emission levels for PAHs and OPAHs present comparable, while NPAH emission levels are five to ten times lower than those of PAHs and OPAHs. Compared with diesel, PAH and NPAH emissions significantly decrease. On the contrary, an increase trend of OPAH emission has been observed with adding biodiesel. For the specific parent PAHs and its oxygenated and nitrated derivatives, the fractions of parent PAHs gradually decrease with increasing biodiesel content in the blends, while the corresponding oxygenated and nitrated derivative fractions observably increase, especially for the high molecular weight compounds. Considering the increase of OPAH and NPAH fractions in total particle-phase PAHs when using biodiesel, in-depth biodiesel cytotoxicity assessment should be conducted.