Applied Mechanics and Materials Vols. 201-202

Abstract: In a high-power fiber laser welding process, the thermal radiation of a weld pool contains plenty of information for welding quality, in which the pool width can reflect the welding stability. Thus, extracting the welding pool width of high-power fiber laser welding based on infrared thermal imaging is an important method for monitoring the weld seam quality. In this paper, we studied the 304 stainless steel welded by a 10kW high-power fiber laser continuously. A near-infrared high-speed sensing camera was used to capture the weld pool images. Image algorithms such as median filtering, gray scale stretching, cutting, dynamic threshold mathematical morphology were applied to extract the weld pool image edge, analyze and detect the weld pool width. Welding experimental results showed that the proposed methods could extract the weld pool width, which could reflect the stability status of high-power fiber laser welding process accurately.

Abstract: We proposed the structure and technics process in corrugated single plane pressure control system through the combination of weighing module SIWAREX and PLC S7-300. We also introduced the system work principle, the hardware configuration and PLC software design. As the special weighing module is applied here, the stability of the pressure load signal has been significantly improved to satisfy the pressure requirements of the board production process. The pressure can be set from the central control room in the production line or the workshop office through network to realize intelligentl control in production management so as to save the production data and avoid the paper break or bad conglutination problem. This design will have a very good application in the area of corrugated packaging in the future.

Abstract: In this paper, a method for estimating induction motor’s rotor speed is proposed. The proposed rotor speed estimation method is based on model reference adaptive identification theory. By applying the proposed method, the induction motor control system can estimate the rotor speed of the induction motor precisely. To improve the rotor speed estimation performance of the system, one input filter and one output filter are introduced into the speed sensorless control system. The introduced input filter and output filter enhance the estimation accuracy and improve the reliability and robustness of the system. The speed sensorless control system based on proposed method was built with Simulink blocks in Matlab platform. The simulation results indicate that the proposed method can operate stably in whole range of speed with preferable identification precision of the rotor speed.

Abstract: The explosive growth of world-wide-web and the emergence of e-commerce make the Collaborative Filtering (CF) algorithms widely used in recommender systems to identify a set of items that will be of interest for a certain user. However, with the rapid growth in the number of users and items, the quality of recommendation decreases. Improved CF algorithms are needed to provide a large number of users with the ability to efficiently obtain useful information. Different user-based approaches for computing similarities (e.g., correlation-based vs. adjusted-cosine) and different techniques for obtaining recommendations (e.g., weighted sum vs. regression model) are presented in the paper. The main steps of the algorithms are divided into two steps. The first is pre-computation of the similarities and pre-selecting of the N-nearest neighbors in the training phase, and the second is online prediction computation. All the algorithms require time linear in the number of items for online predicting. The experiments suggest that the user-based regression algorithms provide better qualities than those not using regression, and the average MAE drop is by 11.16%. Furthermore, the correlation-based regression gets the best MAE drop by 19.41%, and it provides better quality than the best available classic Pearson algorithms.

Abstract: This paper first analyzed the development of Kinetic Sculpture, then proposed the new concept of the Matrix Kinetic Sculpture. By introducing two typical works of the Matrix Kinetic Sculpture, the paper summarized the unique characteristic forms. The paper researched the technology realization of the Matrix Kinetic Sculpture according to one typical case, including the distributed control system and the lifting mechanism. Basing on the Fluid Dynamics theory, the Shallow Water Equations has been applied to the feature design of the Matrix Kinetic Sculpture to enhance its beauty of moving shape. Finally the paper came up with several suggestions for the innovation of Matrix Kinetic Sculpture.

Abstract: This paper introduced a new modular multi-level converter (MMC), which could enhance the voltage and power level by sub-converter modules in series and was easy to extend to any level of output. Its structure and working mechanism were described.By analyzing the new modular multi-level converter (MMC) working mode, its sub-module four power devices’ current distribution is detailed analysis. It can further guide the choice of the device. The simulation results show that the analysis is correct and effective. It can provide a valid theoretical basis for the choice of the MMC sub-module power devices.

Abstract: Image scrambling is an approach to secure the image data by scrambling the image into an unintelligible format. Based on M sequence which can be produced by a series shift registers, this paper proposes a method of scrambling 2-bit watermark image, in which the effectiveness is controlled by given private keys. The user can change the security keys to generate many different M sequences. In watermark embedding algorithm, block DCT with size of 8×8 is performed on the given image, the relationship between the selected DCT middle frequency coefficients is used to embed scrambled watermark. When extracting watermark from the watermarked image no original image is needed. The simulation results show the proposed scheme can guarantee high visual quality of watermarked image, and the same time attain a good robustness to common image processing and attacks.

Abstract: Aviation ad hoc network has dynamic topology and uneven distribution of nodes, so it is difficult to establish node-disjoint multipath routes. Aviation ad hoc network multipath routing of path segmentation based build multipath routes between source node and the destination node in path segmentation. Each segmentation has one to three concurrent paths established by the topology of network, through which it leads to loading equilibrium. To ensure the accuracy of the data transmission of links with hop-by-hop verifying transmission mechanism and to reduce the cost of frequently routing update and end-to-end delay with routing fragment strategy. Finally, the routing protocol is simulated with the simulation software NS2 and the result shows that it has a good performance.

Abstract: Driver model is the most essential part of intelligent driving system. In Driver-Vehicle- Environment system, a driver is a pre-condition, to correct pedal input and the steering wheel angle position in right time, which makes the vehicle drive in expected path. A new driver model is introduced. Firstly, a real-time algorithm based on dynamic parameters of vehicle body is proposed to estimate body posture in the next step moment. Secondly, vehicle’s posture is compared with the target trajectory to get a deviation quantity. And a correction value for angle at the steering wheel is obtained according to a driver model based on fuzzy gain modulated PID control, which is completed in Simulink workspace. Finally, driver model is verified by two typical driving maneuvers in simulation. And results show that the driver model achieves a good performance and suitability in different applications of Driver-Vehicle-Environment.

Abstract: Collaborative Filtering (CF) algorithms are widely used in recommender systems to deal with information overload. However, with the rapid growth in the amount of information and the number of visitors to web sites in recent years, CF researchers are facing challenges with improving the quality of recommendations for users with sparse data and improving the scalability of the CF algorithms. To address these issues, an incremental user-based algorithm combined with item-based approach is proposed in this paper. By using N-nearest users and N-nearest items in the prediction generation, the algorithm requires an O(N) space for storing necessary similarities for the online prediction computation and at the same time gets improvement of scalability. The experiments suggest that the incremental user-based algorithm provides better quality than the best available classic Pearson correlation-based CF algorithms when the data set is sparse.