Model based learning of sigma points in unscented Kalman filtering

The unscented Kalman filter (UKF) is a widely used method in control and time series applications. The UKF suffers from arbitrary parameters necessary for sigma point placement, potentially causing it to perform poorly in nonlinear problems. We show how to treat sigma point placement in a UKF as a learning problem in a model based view. We demonstrate that learning to place the sigma points correctly from data can make sigma point collapse much less likely. Learning can result in a significant increase in predictive performance over default settings of the parameters in the UKF and other filters designed to avoid the problems of the UKF, such as the GP-ADF. At the same time, we maintain a lower computational complexity than the other methods. We call our method UKF-L. © 2011 Elsevier B.V.

Cold-gas chemical vapor deposition to identify the key precursor for rapidly growing vertically-aligned single-wall and few-wall carbon nanotubes from pyrolyzed ethanol

Vertically-aligned carbon nanotubes (VA-CNTs) were rapidly grown from ethanol and their chemistry has been studied using a "cold-gas" chemical vapor deposition (CVD) method. Ethanol vapor was preheated in a furnace, cooled down and then flowed over cobalt catalysts upon ribbon-shaped substrates at 800 °C, while keeping the gas unheated. CNTs were obtained from ethanol on a sub-micrometer scale without preheating, but on a millimeter scale with preheating at 1000 °C. Acetylene was predicted to be the direct precursor by gas chromatography and gas-phase kinetic simulation, and actually led to millimeter-tall VA-CNTs without preheating when fed with hydrogen and water. There was, however a difference in CNT structure, i.e. mainly few-wall tubes from pyrolyzed ethanol and mainly single-wall tubes for unheated acetylene, and the by-products from ethanol pyrolysis possibly caused this difference. The "cold-gas" CVD, in which the gas-phase and catalytic reactions are separately controlled, allowed us to further understand CNT growth. © 2012 Elsevier Ltd. All rights reserved.

Robust filtering and smoothing with gaussian processes

We propose a principled algorithm for robust Bayesian filtering and smoothing in nonlinear stochastic dynamic systems when both the transition function and the measurement function are described by non-parametric Gaussian process (GP) models. GPs are gaining increasing importance in signal processing, machine learning, robotics, and control for representing unknown system functions by posterior probability distributions. This modern way of system identification is more robust than finding point estimates of a parametric function representation. Our principled filtering/smoothing approach for GP dynamic systems is based on analytic moment matching in the context of the forward-backward algorithm. Our numerical evaluations demonstrate the robustness of the proposed approach in situations where other state-of-the-art Gaussian filters and smoothers can fail. © 2011 IEEE.

Effect of laser heating temperature on coating characteristics of Stellite 6 deposited by cold spray

Laser-assisted cold spray (LCS) is a new coating and fabrication process which combines some advantages of CS: solid-state deposition, retain their initial composition and high build rate with the ability to deposit materials which are either difficult or impossible to deposit using cold spray alone. Stellite 6 powder is deposited on medium carbon steels by LCS using N 2 as carrier gas pressure. The topography, cross section thickness, structure of the coatings is examined by SEM, optical microscopy, EDX. The results show that thickness and fluctuation of coating are improved with increased deposition site temperature. Porosity of coating is affected by N 2 and deposition site temperature. In this paper, it presents optimal coating using N 2 at a pressure of 3 MPa and temperature of 450°C and deposition site temperature of 1100°C.

Market-pull and technology-push in manufacturing start-ups in emerging industries

Purpose - As traditional manufacturing, previously vital to the UK economy, is increasingly outsourced to lower-cost locations, policy makers seek leadership in emerging industries by encouraging innovative start-up firms to pursue competitive opportunities. Emerging industries can either be those where a technology exists but the corresponding downstream value chain is unclear, or a new technology may subvert the existing value chain to satisfy existing customer needs. Hence, this area shows evidence of both technology-push and market-pull forces. The purpose of this paper is to focus on market-pull and technology-push orientations in manufacturing ventures, specifically examining how and why this orientation shifts during the firm's formative years. Design/methodology/approach - A multiple case study approach of 25 UK start-ups in emerging industries is used to examine this seldom explored area. The authors offer two models of dynamic business-orientation in start-ups and explain the common reasons for shifts in orientation and why these two orientations do not generally co-exist during early firm development. Findings - Separate evolution paths were found for strategic orientation in manufacturing start-ups and separate reasons for them to shift in their early development. Technology-push start-ups often changed to a market-pull orientation because of new partners, new market information or shift in management priorities. In contrast, many of the start-ups beginning with a market-pull orientation shifted to a technology-push orientation because early market experiences necessitated a focus on improving processes in order to increase productivity or meet partner specifications, or meet a demand for complementary products. Originality/value - While a significant body of work exists regarding manufacturing strategy in established firms, little work has been found that investigates how manufacturing strategy emerges in start-up companies, particularly those in emerging industries. © Emerald Group Publishing Limited.

Particulate matter emission during start-up and transient operation of a spark-ignition engine

In order to understand why emissions of Particulate Matter (PM) from Spark-Ignition (SI) automobiles peak during periods of transient operation such as rapid accelerations, a study of controlled, repeatable transients was performed. Time-resolved engine-out PM emissions from a modern four-cylinder engine during transient load and air/fuel ratio operation were examined, and the results could be fit in most cases to a first order time response. The time constants for the transient response are similar to those measured for changes in intake valve temperature, reflecting the strong dependence of PM emissions on the amount of liquid fuel in the combustion chamber. In only one unrepeatable case did the time response differ from a first order function: showing an overshoot in PM emissions during transition from the initial to the final steady state PM emission level. PM emissions during controlled, motored start-up experiments show a peak at start-up followed by a period during which emissions are either relatively constant or drift somewhat. When the fuel injection and ignition are shut off, PM emissions also peak briefly, but rapidly decay to low levels. Qualitative implications on the study and modeling of PM emissions during transient engine operation are discussed. Copyright © 1999 Society of Automotive Engineers, Inc.

Deterministic cold cathode electron emission from carbon nanofibre arrays.

The ability to accurately design carbon nanofibre (CN) field emitters with predictable electron emission characteristics will enable their use as electron sources in various applications such as microwave amplifiers, electron microscopy, parallel beam electron lithography and advanced Xray sources. Here, highly uniform CN arrays of controlled diameter, pitch and length were fabricated using plasma enhanced chemical vapour deposition and their individual emission characteristics and field enhancement factors were probed using scanning anode field emission mapping. For a pitch of 10 µm and a CN length of 5 µm, the directly measured enhancement factors of individual CNs was 242, which was in excellent agreement with conventional geometry estimates (240). We show here direct empirical evidence that in regular arrays of vertically aligned CNs the overall enhancement factor is reduced when the pitch between emitters is less than half the emitter height, in accordance to our electrostatic simulations. Individual emitters showed narrow Gaussian-like field enhancement distributions, in excellent agreement with electric field simulations.

Single camera pose estimation using Bayesian filtering and Kinect motion priors

Traditional approaches to upper body pose estimation using monocular vision rely on complex body models and a large variety of geometric constraints. We argue that this is not ideal and somewhat inelegant as it results in large processing burdens, and instead attempt to incorporate these constraints through priors obtained directly from training data. A prior distribution covering the probability of a human pose occurring is used to incorporate likely human poses. This distribution is obtained offline, by fitting a Gaussian mixture model to a large dataset of recorded human body poses, tracked using a Kinect sensor. We combine this prior information with a random walk transition model to obtain an upper body model, suitable for use within a recursive Bayesian filtering framework. Our model can be viewed as a mixture of discrete Ornstein-Uhlenbeck processes, in that states behave as random walks, but drift towards a set of typically observed poses. This model is combined with measurements of the human head and hand positions, using recursive Bayesian estimation to incorporate temporal information. Measurements are obtained using face detection and a simple skin colour hand detector, trained using the detected face. The suggested model is designed with analytical tractability in mind and we show that the pose tracking can be Rao-Blackwellised using the mixture Kalman filter, allowing for computational efficiency while still incorporating bio-mechanical properties of the upper body. In addition, the use of the proposed upper body model allows reliable three-dimensional pose estimates to be obtained indirectly for a number of joints that are often difficult to detect using traditional object recognition strategies. Comparisons with Kinect sensor results and the state of the art in 2D pose estimation highlight the efficacy of the proposed approach.