Applied Mechanics and Materials Vols. 543-547

Abstract: A quantum genetic algorithm (QGA) with good global optimization ability and fast convergence speed is proposed to solve the parameter selection problems of least squares support vector machine (LSSVM) on wind turbine gearbox fault diagnosis model. The method can convert the LSSVM model parameter selection into optimization. It overcomes the problem that GA is easy to fall into local optimum in the optimization process and it also improves the optimization performance. A series experiments are carried out on the data sets of UCI database. Compared with GA—LSSVM and CV—LSSVM, the classification accuracy is improved. Finally QGA—LSSVM model is applied to the wind turbine gearbox diagnosis and a good result is achieved.

Abstract: This paper aims to solve the problem of traffic flow detection through video analysis. To tackle with this problem, we propose a novel traffic flow detection algorithm based on Gaussian Mixture Background Modeling which mainly utilizes the advantages of two existing traffic flow detection methods combined with some other technologies, such as improved canny edge detection algorithm, Gaussian Mixture Background Modeling method and background difference techniques. Extensive experiments and comparative studies show the superiority of our traffic flow detection algorithm in real time performance and accuracy.

Abstract: Test coverage information is usually used to compute the suspiciousness to locate the software errors in the current fault localization techniques, but this technique usually do not consider the reliance information within the target program, and the precision is also very low. A novel fault localization technique based on fine grained slicing spectrum is proposed in this paper, which can increase the efficiency of fault localization. This technique analyzes the reliance information under fine grained level, and selects the check points which are prone to be faulty, and the faulty statements is located according to the suspicious result. Experimental results show that this technique has better efficiency than the current techniques.

Abstract: The fluorescence spectrum of pesticides whose structures are very similar overlap in a certain wavelength range. To the classification and recognition of overlapping fluorescence spectrum, BP network has the shortcomings of slow training speed and high error rate in this paper. The network topology is given, wavelet basis is selected and its network algorithm is designed to carry out the design of experimental system. By using the RBF network and BP network separately, the simulation research of fluorescence spectrum classification of carbofuran and carbaryl has been done. The results show that RBF network a higher resolution. To minor structural differences of spectrum, it has a better recognition capability and higher measuring accuracy.

Abstract: Various environmental factors will adversely influence accelerometers working in the field. Mixed inputs, including external causes and the intended acceleration, are important contributors to the problem. As a consequence, calibrating accelerometers with multiple accelerations is required by manufacturers and customers. We introduce a multi-component mechanical system called vibrafuge to provide multiple accelerations. In this paper, we deduce a general model of the accelerations reproduced by an ideal vibrafuge with homogeneous transformation of matrix. Based on the model, we have an elementary analysis of the components of the output acceleration and point out two byproducts. Investigation in this paper could lay the groundwork for vibrafuge testing of accelerometers.

Abstract: According to the surface quality inspection requirements of honeycomb ceramics, an intelligent image inspection and sorting system based on machine vision was designed. The working flow of the designed system includes four steps: top inspection, bottom inspection, side inspection and sorting. Top images and bottom images are captured with area scan CCD cameras. The workpiece rotation (360°) was driven by the turntable unit driven by a servo motor and an encoder while workpiece side images were captured with line scan CCD cameras. The captured images were sent to IPC inspection software for processing and judgment. Then the judgment results were sent to the PLC electrical control system. Unqualified products were picked out by the PLC-controlled sorting unit. The designed system realized the rapid surface inspection and accurate sorting of honeycomb ceramics and improved the production efficiency.

Abstract: Rapid, accurate identification and location of abnormal situation are significant for campus security. The traditional abnormal detections on campus are mainly artificial, which is difficult to conduct accurate and timely action. This paper proposes an Android monitoring platform for campus security based on visual multi-state monitoring states. It uses multiple sensors to collect system information thus realizing advantages complementary, uses wavelet transform in data processing, based on campus feature extraction, to realize image fusion in the pixel domain, greatly improving the output SNR, which improve the results of identification and location of the abnormal situation on campus. Finally, we verify the algorithm by the disaster source in actual disaster environment. Results show that the fusion algorithm in this paper greatly improves performance of campus security monitoring system to identify and locate the source disaster.

Abstract: The linear motor pump bears the pumping work and other important work, once the valve failure, it threats to the safety of staff seriously. In order to ensure the safety of underground working environment, the abnormal current signal, vibration signals, differential pressure signal and absolute pressure signal are taken as the recognition base, a portable linear motor pump fault detection system is designed. The modular design process of hardware system and software system in detail are given. The data which collected from Victory Oil Field are taken for system test and detection instrument assembly. Test results show that the system can effectively collect failure signal and make an accurate diagnosis. It has the good performance in man-machine interaction, and it has very nice engineering application prospect.

Abstract: The node self-localization is the basis of target localization for wireless sensor network (WSN), the WSN nodes localization algorithms have two types based on distance and non distance. The node localization based on RSSI is simple and widely used in application. According to the traditional WSN nodes localization algorithm, the RSSI signal intensity changes greatly and with nonlinearity. And it is converted into distance feature with a large deviation, which leads to inaccurate positioning and localization. In order to solve this problem, a sensor node localization algorithm is proposed based on fuzzy RSSI distance. The nodes information is collected based on RSSI ranging method. And the location information is processed with fuzzy operation. The disturbance from the environmental factors for the positioning is solved. The accuracy of the node localization is improved. Simulation result shows that this algorithm can locate the sensor nodes accurately. The localization accuracy is high, and the performance of nodes localization is better than the traditional algorithm. It has good application value in the WSN nodes distribution and localization design.

Abstract: This paper presents a design schema about remote electrocardiograph (ECG) monitoring system. There are three parts in this system which include hardware detection system based on MSP430, Zigbee wireless sensor network and data management software system on PC. This paper mainly provides information for study and design of the data management software system using visual C++ program to realize immediate communication of medical monitoring. The data management software system will achieve many functions such as serial communication, real-time dynamic display of 12-lead ECG waveform and physiological parameter which include Heart Rate (HR).