Further results of recursive evaluation for compound distribution with the severity distribution of mixed type

Yang et al. [J.P. Yang, S.H. Cheng, Q. Wu, Recursive equations for compound distribution with the severity distribution of the mixed type, Science in China Series A 48 (2005) 594-609] investigated a recursive procedure for a kind of compound distributions with the number of claims belonging to (a,b)-family and the severity distribution of the mixed type. In this paper, we extend their results by assuming that the claim number belongs to a larger class. As applications, the excess-of-loss reinsurance treaty is discussed and concrete examples are considered in some detail.     

A framework for reviewing domain specific conceptual models

Conceptual models are used in understanding and communicating the domain of interest during analysis phase of system development. As they are used in early phases, errors and omissions may propagate to later phases and may be very costly to correct. This paper proposes a framework for evaluating conceptual models when represented in a domain specific language based on UML constructs. The framework describes the main aspects to be considered when conceptual models are represented in a domain specific language, presents a classification of semantic issues and some evaluation indicators. The indicators can, in principle, identify situations in the models where inconsistencies or incompleteness might occur. Whether these are real concerns might depend on domain semantics, hence these are semantic, not syntactic checks. The use of the proposed review framework is illustrated in the context of two conceptual models in a domain specific notation, KAMA. With reviews based on the framework, it is possible to spot semantic issues which are not noticed by case tools and help the analyst to identify more information about the domain.     

PIRF-Nav 2.0: Fast and online incremental appearance-based loop-closure detection in an indoor environment

This paper presents a fast and online incremental solution for an appearance-based loop-closure detection problem in a dynamic indoor environment. Closing the loop in a dynamic environment has been an important topic in robotics for decades. Recently, PIRF-Nav has been reported as being successful in achieving high recall rate at precision 1. However, PIRF-Nav has three main disadvantages: (i) the computational expense of PIRF-Nav is beyond real-time, (ii) it utilizes a large amount of memory in the redundant process of keeping signatures of places, and (iii) it is ill-suited to an indoor environment. These factors hinder the use of PIRF-Nav in a general environment for long-term, high-speed mobile robotic applications. Therefore, this paper proposes two techniques: (i) new modified PIRF extraction that makes the system more suitable for an indoor environment and (ii) new dictionary management that can eliminate redundant searching and conserve memory consumption. The results show that our proposed method can complete tasks up to 12 times faster than PIRF-Nav with only a slight percentage decline in recall. In addition, we collected additional data from a university canteen crowded during lunch time. Even in this crowded indoor environment, our proposed method has better real-time processing performance compared with other methods.     

Visual quality recognition of nonwovens using generalized Gaussian density model and robust Bayesian neural network

This work is dedicated to develop an algorithm for the visual quality recognition of nonwoven materials, in which image analysis and neural network are involved in feature extraction and pattern recognition stage, respectively. During the feature extraction stage, each image is decomposed into four levels using the 9-7 bi-orthogonal wavelet base. Then the wavelet coefficients in each subband are independently modeled by the generalized Gaussian density (GGD) model to calculate the scale and shape parameters with maximum likelihood (ML) estimator as texture features. While for the recognition stage, the robust Bayesian neural network is employed to classify the 625 nonwoven samples into five visual quality grades, i.e., 125 samples for each grade. Finally, we carry out the outlier detection of the training set using the outlier probability and select the most suitable model structure and parameters from 40 Bayesian neural networks using the Occam's razor. When 18 relevant textural features are extracted for each sample based on the GGD model, the average recognition accuracy of the test set arranges from 88% to 98.4% according to the different number of the hidden neurons in the Bayesian neural network.     

Modeling the high pH sensitivity of Suspended Gate Field Effect Transistor (SGFET)

High sensitivity to chemical species of sub-micron gap Suspended-Gate FETs (more than 200 mV/pH for example) is explained from the charge distribution induced by the high field in the sub-micronic gap under the gate-bridge. Modeling of Metal-Electrolyte-Insulator-Silicon (MEIS) capacitor, which is the basic vertical structure of the transistor, is performed to highlight this effect through the response to the pH change of the solution filling the gap. The analytical model is based on the 2D-numerical resolution of Poisson's equation. The response of quasi-static C(V) plots versus pH is simulated using both electrolyte charge distribution and site-binding theory considering the influence of sites densities on silicon nitride. Device modeling and simulated/experimental electrical characteristics are presented. Effect of the gap thickness on the pH sensitivity is also discussed in this study.     

Thin-film electrode based droplet detection for microfluidic systems

We report on a droplet-producing microfluidic system with electrical impedance-based detection. The microfluidic devices are made of polydimethylsiloxane (PDMS) and glass with thin film electrodes connected to an impedance-monitoring circuit. Immiscible fluids containing the hydrophobic and hydrophilic phases are injected with syringe pumps and spontaneously break into water-in-oil droplet trains. When a droplet passes between a pair of electrodes in a medium having different electrical conductivity, the resulting impedance change signals the presence of the particle for closed-loop feedback during processing. The circuit produces a digital pulse for input into a computer control system. The droplet detector allows estimation of a droplet's arrival time at the microfluidic chip outlet for dispensing applications. Droplet detection is required in applications that count, sort, and direct microfluidic droplets. Because of their low cost and simplicity, microelectrode-based droplet detection techniques should find applications in digital microfluidics and in three-dimensional printing technology for rapid prototyping and biotechnology.     

Liquid-state motion sensing

This paper demonstrates a liquid droplet-based motion sensing system which has the advantages of simple fabrication, low power consumption and digital signal processing. The sensor consists of a dielectric substrate patterned with an array of microelectrodes, and a saline droplet as the proof mass. Once an external linear acceleration is applied, the inertial force moves the droplet on the micropatterned substrate. The acceleration is determined from the movement profile detected by the microelectrodes. In order to enhance the threshold and the sensitivity of motion sensing, two surface treatment approaches are utilized to create superhydrophobic surfaces. The result shows that the minimal sliding angle that can move a 20 μl droplet on the superhydrophobic surface is lower than 1°, corresponding to a threshold of lower than 0.017 g. A lumped-parameter model is developed to estimate the dynamic behavior of the proposed system. The result shows that the frequency response of the droplet-based sensor is more significant at low frequencies than at high frequencies, which is distinct from solid-state accelerometers. Measurement under a constant acceleration shows that the predicted value derived from the measurement has a good match with the actual applied acceleration, validating the proposed system as a viable alternative for motion sensing.     

A multichannel surface plasmon resonance sensor using a new spectral readout system without moving optics

Surface plasmon resonance (SPR) sensors with spectral interrogation provide a high refractive index resolution, a large dynamic range and a fixed optical detection module. In this work, we propose a new multichannel spectral detection unit that uses only one spectrometer to measure the reflection spectrum from multiple sensing spots serially without any mechanical movement. This spectral detection unit is designed based on a spatial light modulator (SLM) configured as a programmable optical aperture for the spectrometer. To demonstrate this concept, a five-channel laboratory SPR prototype was built based on the proposed multichannel detection unit, and we evaluated the device's sensitivity and resolution using a refractive index test. Refractive index resolution of 1.4 × 10⁻⁶ refractive index units (RIU) can be reached using the five-channel prototype. This sensor is suitable for low-cost multichannel biosensing applications that do not contain fast kinetics.     

Towards a GC-based microsystem for benzene and 1,3 butadiene detection: Pre-concentrator characterization

The performance of a gas pre-concentrator/gas chromatographic column/gas sensor system was studied step by step towards benzene vapor in presence of 1,3 butadiene vapor. The pre-concentrator, based on an activated carbon of high adsorption capacity, was first characterized using a GC/FID system. By this way, the adsorption capacity of each pollutant was evaluated, first separately, then mixed with different proportions in air. It was shown that the affinity of the pre-concentrator for benzene was higher than for butadiene. Once the optimal pre-concentration parameters were adjusted, we replaced the FID detector by a metal oxide gas sensor. The use of such gas pre-concentrator allowed decreasing the detection limit of the system towards benzene in presence of butadiene by at least ten times.