Xu Weisheng

A Fast Harmonic Detection Method Based on Recursive DFT

This paper proposed a fast harmonic detection method based on recursive discrete Fourier transform, which is suitable for real-time calculation. The main features of the proposed method are that (1) it can be used to detect only specific harmonic components, (2) the recursive method needs few calculation, so harmonic detection and compensation can be implemented on-line, (3) spectrum update is implemented in a step larger than one sampling period to avoid over-sampling, (4) time-domain windowing has been taken into account to reduce spectral leakage caused by asynchronous sampling. These characteristics make the proposed method especially compatible with hybrid active filter, in which structure only specific harmonic components concerned are needed to be analyzed and harmonic compensation is implemented almost immediately.

Area Harmony Dominating Rectification Method for SIFT Image Matching

This paper presents a method of verifying matching key-point couples extracted by SIFT algorithm from two relative images. This method defines dependability of matching couples with local area features which express the harmony of the tested couple with nearest matching couples. In this paper, pictures that have similar contents are first analyzed by SIFT algorithm. After this operation, original matching key-point couples are found out. Then the result of SIFT are examined by our method. In the pilot study, experiments have proved the method behaves fast, efficiently and stably. The data from experiment shows that the method is capable of discriminating correct matching couples from incorrect couples, so it can eliminate mismatching key-point couples automatically in the same time protect the matching couples. In most tests, it increases the matching rate from about 70% to more than 95%.

Tensorial Kernel Principal Component Analysis for Action Recognition

We propose the Tensorial Kernel Principal Component Analysis (TKPCA) for dimensionality reduction and feature extraction from tensor objects, which extends the conventional Principal Component Analysis (PCA) in two perspectives: working directly with multidimensional data (tensors) in their native state and generalizing an existing linear technique to its nonlinear version by applying the kernel trick. Our method aims to remedy the shortcomings of multilinear subspace learning (tensorial PCA) developed recently in modelling the nonlinear manifold of tensor objects and brings together the desirable properties of kernel methods and tensor decompositions for significant performance gain when the data are multidimensional and nonlinear dependencies do exist. Our approach begins by formulating TKPCA as an optimization problem. Then, we develop a kernel function based on Grassmann Manifold that can directly take tensorial representation as parameters instead of traditional vectorized representation. Furthermore, a TKPCA-based tensor object recognition is also proposed for application of the action recognition. Experiments with real action datasets show that the proposed method is insensitive to both noise and occlusion and performs well compared with state-of-the-art algorithms.

A high-speed and high-resolution CMOS comparator with three-stage preamplifier

The accuracy of A/D and D/A converters depend largely upon their inner comparators. To guarantee 12-bit high resolution for an A/D converter, a precise CMOS comparator consisting of a three-stage differential preamplifier together with a positive feedback latch is proposed. Circuit structure, gain, the principle of input offset voltage storage and latching time constant for the comparator will be analyzed and optimized in this article. With 0.5 μm HYNIX mixed signal technology, the simulation result shows that the circuit has a precision of 400 μV at 20 MHz. The test result shows that the circuit has a precision of 600 μV at 16 MHz, and dissipates only 78 μW of power dissipation at 5 V. The size of the chip is 210 × 180 μm2. The comparator has been successfully used in a 10 MSPS 12-bit A/D converter. The circuit can be also used in a less than 13-bit A/D converter.

Design of Neural Network Output Layer in Fault Diagnosis of Analog Circuit

In this paper, we have developed a new way to design the output layer of neural network in fault diagnosis of analog circuit. The primary focus of our method is to simplify the network architecture while offering the value information of faulty component. A comparison of our work with reference indicates that our work ensures a robust and fast fault diagnosis in analog circuit including effects due to tolerances. The concepts of our work can be extended to detect multiple faulty components.

Design of a Fluorescent Lamp Control Platform Based on Power-line Carrier

This paper introduced a design of a fluorescent lamp control platform based on the technology of power-line carrier, which was consistent to the trend of green lighting and intelligent building. And the key point of this platform was to make the communication among those transceivers, which are connected just by power-line, dependable and stable. So this paper focused on the circuitry designing around the PLC chip SC1128.

On-line current sensing circuit for the over-current protection of the electronic ballast

A new on-chip current sensing circuit for the over-current protection of the electronic ballast is presented in this paper. The proposed scheme increases the sense accuracy and noise immunity over the full temperature range of the devices and can avoid fault over-current protection. Simulation results verify the accuracy of the on-chip current sensing circuit.

A Petri net based model for business reengineering

Business reengineering (BR) is a new concept of organization and management, and its model is of great importance. First, through enhancing the SLANG language and a model for software process development, a Petri net based business process model is described for BR. A hybrid modeling method that combines top-down and bottom-up principles is then explored. Finally, the model and its modeling method are illustrated by an example.

Improving the 3PLM-Based Computerized Adaptive Testing System with Multi-Agent Item Bank

A multi-agent-based computerized adaptive testing (CAT) system on IRT (item response theory) is introduced, which uses the three-parameter logistic model (3PLM) of IRT as the theoretical model. The paper explores and researches into the steps of constructing item bank , determining the beginning item, response, estimating ability, item selection and terminating condition, and a IRT-based item bank on multi-agent system is put forward here. Aiming at the limits of the most commonly used Newton-Raphson iterative algorithm in the ability estimating, median value algorithm is applied to calculate maximum likelihood estimation (MLE) instead of that, and utilizes a series of adjusting strategy to solve some other shortages in the logistic median value model. This CAT system has more superior function and lower calculation quantity, with fast convergence and the accuracy of ability estimating, which will be adapted further in modern test system.

A Method of Modeling and Solution Based on Non-zero-sum Cooperative and Dynamic Games for Nodal Enterprises in Supply Chain

Based on some features of supply chain, game theory is utilized to analyze the behaviors of nodal enterprises in supply chain, the results of analysis show that the process of games among nodal enterprises belongs to non-zero-sum cooperative games, dynamic games and incomplete information games. Then, for the nodal enterprises in supply chain, the model of non-zero-sum cooperative and dynamic games is established. Furthermore, the core method is improved by introducing a subsidy variable. Solution of model by applying the improved core method and the Shapley value method combined with cooperative game theory is used as the criterion of payoff allocation for nodal enterprises. Finally, researching potentials and suggestions about supply chain coordination and incentive are proposed. The work of this paper provides the theory basis and the direction of further research for the coordination mechanism and the incentive mechanism among nodal enterprises.