Advanced Materials Research Vols. 403-408

Abstract: The paper completes image registration of medical CT and DR through improved SUSAN algorithm. First, the paper divides the image into a certain number of anatomical point regions with some significance, and sets different grayscale D-value for different regions to make corners distribute equably. Second, the paper uses the size of the USAN from limited corner regions to reduce the number of corners and the computation. Then the paper adopts confidence algorithm, find the correct match points by setting threshold, and thus completes image registration.

Abstract: This paper presents a wideband low noise amplifier (LNA) for multi-standard radio applications. The low noise characteristic and input matching are simultaneously achieved by active-feedback technique. Bond-wire inductors and electrostatic devices (ESDs) are co-designed to improve the chip performance. Implemented in 0.18-μm CMOS process, the core size of the fully integrated LNA circuits is 535 μm×425 μm without any passive on-chip inductor. The simulated gain and the minimal noise figure of the CMOS LNA are 17.5 dB and 2.0 dB, respectively. The LNA achieves a -3dB bandwidth of 3.1 GHz. And the simulated IIP3 is -4.4 dBm at 2.5 GHz. Operating at 1.8V, the LNA draws a current of 7.7 mA.

Abstract: This paper studied the key technoligies and scientific issues of the smart grid dispatching, combining the status and trends of the development of smart grid, and has done research and analysis from the four aspects respectively as follow, the security and defense of a large network and panoramic information and integrated technology platform of smart grid dispatching, the multi-level integration of security check rechnology, energy saving intelligent optimization technology and multi-dimensional intelligent visualization technology of large grid. Summarized the existing equipment and industrial in the power network dispatching and gave some suggestions for developing the technologies of smart grid dispathching.

Abstract: With the development of wireless network technology, heterogeneous wireless networkis ubiquitous. Thus, selection of an optimal access network canmeet the demands of customersto the highest degree. A non-optimal network selection can result in bad effects such as poor service or higher cost.The traditionalapproach for network selection is based on Received Signal Strength (RSS). However, this paper aims to propose a Multi AttributesDecisionMaking (MADM) algorithm,which combined the parameters of network quality, system state, terminal state and user preferences, to select the optimal network for Accessing Multiple Services Simultaneously (AMSS).This paper analyzesservices classification,handoff related network attributes and the strong association attributes of services, and concerns both of the objective weights and subjective weights when calculating the attributes weights under AMSS. This proposal uses Entropy Method to assign the objective weights and Analytic Hierarchy Processto assign the subjective weights. In the end of this article, a Matlab simulation is performed to evaluate theproposed network selection mechanism based on Technique for Order Preference by Similarity to Ideal Solution(TOPSIS).

Abstract: In this paper a multi-function GPS data collection system based on GSM Short Message Service is designed. By using a self-defining communication protocol, data transmission between a mobile terminal and a monitor center is achieved. The information collected by the mobile terminal is displayed upon the monitor center and a path of the mobile terminal can be shown on Google Earth. Experimental results show that the system not only reaches a satisfying positioning accuracy, but also runs robustly at a low-power state.

Abstract: Concurrent engineering has been used in many OEMs in order to face up to the fierce competition situation. Modular development teams develop common modules which are applied in the same platform-based variants and its product family in order to save design costs and reduce development time. The R&D department should construct a concurrent engineering team to quicken the pace of new product development and to ensure product quality by means of open-loop and close-loop cycle, and frame a platform to frequent communication and quick decision for the concurrent engineering team.

Abstract: This paper builds the Artificial Neural Network model on the amelioration of the traditional BP arithmetic, which is applied to the study of magnitype punching mould quotation system. And the quotation scheme based on the ANN is proposed. By adjusting the error, the article resolves the multi-factor, non-linearity problem in the quotation system.

Abstract: With the great demand for broadband communication in high speed railway, new access technologies should be adapted to high mobility which brings Doppler frequency shift and fast handoff. The definition of cooperative range is presented and its influence on network connectivity is studied. The result shows that cooperation can greatly extend a wireless user’s communication range. A novel scheme of cooperative access architecture is proposed to support high-speed users with broadband communication.

Abstract: This paper presents a control method for prosthetic hand's grip force. Firstly, the correlated characters of the SEMG signal are calculated and input into the neural network in order to extract the estimation of muscle strength. Secondly, the estimation of muscle strength and the real-time prosthetic hand’s grip force are input into the designed fuzzy logic controller in order to adjust the motor speed, and indirectly control the grip force. This method could greatly reduce the individual difference of SEMG signal, make it can be used by Non-specific patient, and the experiment demonstrates that the control of prosthetic hand’s grip force is effective.

Abstract: Engine torque estimation function is the base of engine torque control. This paper establishes the model for engine torque estimation respectively under steady condition and unsteady condition based on BP Neural network, and develops a new engine torque real-time estimation method. The experiment results under steady condition and unsteady condition show that the engine torque estimation model can estimate the engine output torque and the precision is remarkable.