Sensitivity study of a semi-automatic training set generator

This paper addresses the problem of assessing the robustness with respect to change in parameters of an integrated training and classification routine. Extensions to traditional multivariate statistical methods are applied to perform the classification. Training sets are grown from one or a few seed points by a method that ensures spatial and spectral closeness of observations. Spatial closeness is obtained by requiring connectivity. Spectral closeness is obtained by excluding observations that have high Mahalanobis distances to the training class mean. The marginal effects of changes in the parameters that are input to the seed growing algorithm are evaluated. Initially, the seed is expanded to a small area in order to allow for the estimation of a dispersion matrix. This expansion is controlled by upper limits for the spatial and Euclidean spectral distances from the seed point. Second, after this initial expansion, the growing of the training set is controlled by an upper limit for the Mahalanobis distance to the current estimate of the class centre. Also, the estimates of class centres and dispersion matrices may be continuously updated, or the initial estimates may be used. Finally, the effect of the operator's choice of seed among a number of potential seed points is evaluated. An evaluation of the sensitivity of the seed algorithm with respect to parameter settings is carried out by applying it to the classification of minerals commonly encountered in siliciclastic or carbonate rocks from twelve chemical elements mapped from thin sections by energy-dispersive spectroscopy (EDS) in a scanning electron microscope (SEM), using a standard quadratic classifier. The performance for each parameter setting is measured by the overall misclassification rate on an independently generated validation set. The integrated training and classification method is presently used as a routine petrographical analysis method at Norsk Hydro Research Centre.     

A full-scale experimental set-up for assessing the energy performance of radiant wall and active chilled beam for cooling buildings

Full-scale experiments under both steady-state and dynamic conditions have been performed to compare the energy performance of a radiant wall and an active chilled beam. From these experiments, it has been observed that the radiant wall is a more secure and efficient way of removing heat from the test room than the active chilled beam. The energy saving, which can be estimated to around 10%, is due to increased ventilation losses. The asymmetry between air and radiant temperature, the air temperature gradient and the possible short-circuit between inlet and outlet play an equally important role in decreasing the cooling need of the radiant wall compared to the active chilled beam. It has also been observed that the type and repartition of heat load have an influence on the cooling demand. Regarding the comfort level, both terminals met the general requirements, except at high solar heat gains: overheating has been observed due to the absence of solar shading and the limited cooling capacity of the terminals. No local discomfort has been observed although some segments of the thermal manikin were slightly colder.     

Asynchronous variational contact mechanics

An asynchronous, variational method for simulating elastica in complex contact and impact scenarios is developed. Asynchronous variational integrators [1] (AVIs) are extended to handle contact forces by associating different time steps to forces instead of to spatial elements. By discretizing a barrier potential by an infinite sum of nested quadratic potentials, these extended AVIs are used to resolve contact while obeying momentum- and energy-conservation laws. A series of two- and three-dimensional examples illustrate the robustness and good energy behavior of the method.     

Mobility Impact on Cluster Based MAC Layer Protocols in Wireless Sensor Networks

Wireless sensor networks (WSNs) enable a wide variety of applications resulting in still increasing requirements for the protocols supporting the operations. The medium access control (MAC) layer protocols are essential for improving the performance of an application and its quality of service because MAC protocols influence channel capacity utilization, network delay, energy consumption, and scalability. The contribution of this paper is two novel cluster-based time division multiple access (TDMA) scheduling MACs for WSNs and an analysis of the mobility impact on both. The proposed MAC layer protocols support real time applications where the cluster-based scheduling improves the scalability and also improves the performance in varying network conditions. The paper presents the design, implementation and performance evaluation of the proposed cluster based TDMA scheduling algorithms green conflict free (GCF) and multicolor-GCF (M-GCF) for high complexity and high requirement applications of WSNs under both low and high mobility conditions. The comparative evaluation shows that the M-GCF algorithm has better slot sharing and less conflicts with reduced communication energy consumption, delay, and good throughput under static and low mobility conditions while the GCF algorithm has better performance in high mobility scenarios. The paper also defines the mobility threshold that decides the use of the GCF- and M-GCF algorithms according to the mobility requirement of application.     

城市森林

重庆是中国最年轻的直辖市,城市规划建设大有可为,完全有条件建成一个大家公认的宜居城市,一个环境优美的城市,一个交通顺畅的城市,甚至变成一个森林城市。一个有希望、有活力的地方,要敢于设想和设计自己的美好未来。——薄熙来(重庆市长)     

DCB-specimen loaded with uneven bending moments

A double cantilever beam specimen loaded with uneven bending moments (DCB-UBM) is proposed for mixed mode fracture mechanics characterisation of adhesive joints, laminates and multilayers. A linear elastic fracture mechanics analysis gives the energy release rate and mode mixity analytically for both isotropic and orthotropic materials. By varying the ratio between the two applied moments, the crack tip stress state can be varied from pure mode I to pure mode II for the same specimen geometry. The specimen allows stable crack growth. A special test fixture is developed to create uneven bending moments. As a preliminary example, the DCB-UBM specimen was used for characterising fracture of adhesive joints between two laminates of thermoset glass fibre reinforced plastic.     

Fe-BEA Zeolite Catalysts for NH3-SCR of NOx

Iron-containing zeolites are known to be promising catalysts for the NH₃-SCR reaction. Here, we will investigate the catalytic activity of iron-based BEA catalysts, which was found to exhibit improved activities compared to previously described iron-containing zeolite catalysts, such as ZSM-5 and ZSM-12. Series of Fe-BEA zeolite catalysts were prepared using a range of different preparation methods. Furthermore, we found that an iron concentration around 3 wt% on BEA showed a small optimum in SCR activity compared to the other iron loadings studied.     

The effect of gelation on statically and dynamically freeze-cast structures

Freeze-casting is a technique used to produce structures with anisotropic porosity in the form of well-defined microchannels throughout a sample. Here, this technique is used on the magnetocaloric ceramic La_0.66Ca_0.26Sr_0.07 Mn_1.05O₃. We show that a dynamic freezing profile, where the temperature is decreased continuously at −10 K/min, results in homogeneous, lamellar channels with widths of ∼15 µm, while static freezing, where the temperature is kept constant at 177 K, results in channels of increasing size away from the initial ice crystal nucleation site. The effect of gelation before freeze-casting is also investigated. Gelation inhibits ice crystal growth, which significantly changes the morphology by making channel cross sections less elongated, while additionally introducing more dendrites and ceramic bridges in the structure. The latter significantly dominates the flow path through the gelated structures, affecting the calculated tortuosity, which increases to τ ≈ 4 when compared to non-gelated samples where calculated tortuosities are in the range of ∼1.3 to ∼3. Finally, we present a systematic and automatic approach for evaluating channel and wall sizes and calculating tortuosities. This is based on analysis of images obtained by scanning electron microscopy using a continuous particle size distribution method and the TauFactor application in MATLAB^®.     

Gender gaps in scientific performance: a longitudinal matching study of health sciences researchers

Scientometrics - The existence of gender disparities in academia is well documented. Many explanations have been proposed and productivity is one of the most used variables to explain a possible...