Probabilistic graphical models for semi-supervised traffic classification

Traffic classification using machine learning continues to be an active research area. The majority of work in this area uses off-the-shelf machine learning tools and treats them as black-box classifiers. This approach turns all the modelling complexity into a feature selection problem. In this paper, we build a problem-specific solution to the traffic classification problem by designing a custom probabilistic graphical model. Graphical models are a modular framework to design classifiers which incorporate domain-specific knowledge. More specifically, our solution introduces semi-supervised learning which means we learn from both labelled and unlabelled traffic flows. We show that our solution performs competitively compared to previous approaches while using less data and simpler features. Copyright © 2010 ACM.

Experimental investigation of a control method for SI-HCCI-SI transition in a multi-cylinder gasoline engine

In HCCI engines, the Air/Fuel Ratio (AFR) and Residual Gas Fraction (RGF) are difficult to control during the SI-HCCI-SI transition, and this may result in incomplete combustion and/or high pressure raise rates. As a result, there may be undesirably high engine load fluctuations. The objectives of this work are to further understand this process and develop control methods to minimize these load fluctuations. This paper presents data on instantaneous AFR and RGF measurements, both taken by novel experimental techniques. The data provides an insight into the cyclic AFR and RGF fluctuations during the switch. These results suggest that the relatively slow change in the intake Manifold Air Pressure (MAP) and actuation time of the Variable Valve Timing (VVT) are the main causes of undesired AFR and RGF fluctuations, and hence an unacceptable Net IMEP (NIMEP) fluctuation. We also found large cylinder-to-cylinder AFR variations during the transition. Therefore, besides throttle opening control and VVT shifting, cyclic and individual cylinder fuel injection control is necessary to achieve a smooth transition. The control method was developed and implemented in a test engine, and the result was a considerably reduced NIMEP fluctuation during the mode switch. The instantaneous AFR and RGF measurements could furthermore be adopted to develop more sophisticated control methods for SI-HCCI-SI transitions. © 2010 SAE International.

Robust approach to dynamic statistical process control

Statistical Process Control (SPC) technique are well established across a wide range of industries. In particular, the plotting of key steady state variables with their statistical limit against time (Shewart charting) is a common approach for monitoring the normality of production. This paper aims with extending Shewart charting techniques to the quality monitoring of variables driven by uncertain dynamic processes, which has particular application in the process industries where it is desirable to monitor process variables on-line as well as final product. The robust approach to dynamic SPC is based on previous work on guaranteed cost filtering for linear systems and is intended to provide a basis for both a wide application of SPC monitoring and also motivate unstructured fault detection.