Applied Mechanics and Materials Vols. 143-144

Abstract: Aimed at the nonlinear properties of motor rotor vibration signal,a fault diagnosis method based on kernel principal component analysis (KPCA) and support vector machines (SVM) was proposed. Initial feature vectors of motor vibration signal were mapped into higher-dimensional space with kernel function. Then the PCA method was used to analyze the data in the high dimensional space to extract the nonlinear features which is used as training sample of SVM fault classifier. Then the rotor fault is identified using the trained classifier. The classification effect of KPCA-SVM is compared with PCA-SVM and SVM. The result shows that the method based on KPCA-SVM can identify motor rotor fault efficiently and fulfill fault classification accurately.

Abstract: Satellite-based is a very complex system, including of many subsystems involving space, channel, ground, command and so on. The simulation system must rely on the joint co-simulation. Based on HLA (High-Level Architecture) system standard, from the perspective of distributed simulation, a simple satellite design schema was introduced in this article, which can achieve a joint multi-system co-simulation.

Abstract: Vibration signals of rotating machinery on the base are very weak and always buried in noisy noise; the common denoising methods have become powerless. It presents an ensemble empirical mode decomposition method (EEMD) that is used to denoise for the base vibration signal, which not only to overcome the problem of mode mixing, but also to avoid the selection of wavelet basis function and decomposition level of the problem. Experimental results of simulation and measured data show that EEMD method can effectively reduce the base vibration signal noise, which is better than the wavelet and EMD denoising method.

Abstract: Vibration is one of the most important factors affecting the accuracy and precision of machine bed. Under the definite structure, different materials and fillings will affect the natural frequency of the machine bed. 3 kinds of candidate materials (HT300.20#.20# & concrete) are selected to build different machine bed. Based on FEM, modal analyses are performed on the different machine beds, 4 sets of natural frequency are sampled on each one. Aiming to better performance in vibration, the machine bed is optimized with higher natural frequency. Through experimental study on the machine bed, it is clear that the optimized machine bed can meet the needs to perform better in machining.

Abstract: Elastic tube bundles are universally used in heat transfer enhancement by flow-induced vibration in heat exchangers, and the study of the heat transfer performance is of importance. The structure of conical spiral tube bundle heat exchanger was introduced first, and the structure of pulsation flow generator was also introduced. The frequency of pulsation flow was discussed. Finally, in condition of same shell side diameter, the heat transfer and natural frequency of the conical spiral tube bundle were compared with the planar elastic tube bundle. The results show that the natural frequency of conical spiral tube bundle was smaller, the heat transfer performance of conical spiral tube bundle was better than the planar elastic tube bundle.

Abstract: This paper describes a strategy for fault diagnosis of mechatronic equipments. The basic idea of information fusion is not only to take into account on-site sensor measurements, but also other factors that could potentially illustrate the essential course of an equipment's evolving faults. For example, these distributed factors might include components information, overhaul test results, usage data, expert experience, and others. Integrating the above produces a more refined diagnosis analysis compared to a local decision based on on-site information alone. However, measurements and other evidential contents were originally meant to serve different goals, i.e. the multi sources might be highly heterogeneous. To allow a unified representation we propose in this paper a formalization modeling based on ontology. Finally, we demonstrate the diagnostic synthesis process in terms of a certain information fusion method, i.e. Bayesian Network.

Abstract: To lower the cost of passive gravity navigation, a simple and easy passive gravity navigation, which is composed of a rate azimuth inertial platform with a gravity sensor on it, and a digital gravity map, is proposed. Based on the local observability theory, the relationship between the observation of the integrated system and the gravity field’s roughness is deduced by the singular value of the observation matrix. The simulation of the integrated navigation system is carried out with Matlab/Simulink and gravity anomaly map whose resolution is five over one thousand. Simulation results show that, in the area with evident gravity anomaly, the positioning error of the RAPINS/Gravity integrated navigation system, combined of moderate precision accelerometers and gyroscopes, is small, which can meet up with the demand of underwater passive navigation.

Abstract: Fractal geometry is an useful approach in pattern recognition. Many fractal recognition methods use global analysis of the shape. In this paper we present a new fractal recognition method based on a dependence graph obtained from the PIFS. Moreover, this method uses local analysis of the shape which improves the recognition rate. The recognition algorithms have been tested to provide a feasible classification of the possible errors present in our similar object images datebase.

Abstract: In the presented paper, we proposed a common color model and designed the color judgment method, which is based on the HSV model. This method will translate the RGB values of the points in video images to HSV values, and use HSV values to recognize the color. After that, software of real-time video object recognition was developed based on color features, which is also based on their search of target color identification. Besides, the system is developed by VC based on OpenCV, which has achieved the goal of real-time video motion detection and object color recognition. Finally, the experimental results indicate that the algorithm is accurate and similar to human recognition of the moving objects in videos view, which demonstrates the good performance of the target identification and color judgment.

Abstract: A measurement based on image process is proposed for detecting the hole on a barrel. The barrel image is separated from background with the image edge detection technology. The different detection operators are compared and the sobel operator is used for edge detection. The actual space position of the hole on barrel is computed by circle detecting method. Gradient hough transform is used for detecting circle position information, including circle radius and two dimension coordinates. The stepping motor drive system is consists of a control card, stepping motor drive units and two two-phase hybrid stepping motors. Automatic positioning is achieved through stepping motor drive system.