Applied Mechanics and Materials Vols. 571-572

Abstract: Activity recognition is a challenging problem for context-aware systems and applications. Many studies in this field has mainly adopted techniques based on supervised or semi-supervised learning algorithms to recognize activities by movement patterns gathered through sensors, but these existing systems suffer from complex issues for feature representations of sensor data and multi-sensors integration. In this paper, we propose a novel feature learning method for activity recognition based on entropy and construct an activity recognition model with multi-class AdaBoost algorithm. Experiments on sensor data from a real dataset demonstrate the significant potential of our method to extract features for activity recognition. The experimental results also show recognition model based on multi-class AdaBoost is effective. The average precision and recall for six activities are 95.9% and 95.9%, respectively, which are higher than results obtained by using other methods such as Support Vector Machine (SVM) or K-Nearest Neighbor (KNN).

Abstract: In deep research and analysis on WSN security routing protocols and according to the rational development trend of the sensor nodes, this paper proposes a new WSN security routing protocol based on the theory of the static game—GBR. GBR describes the interactions in a static game model using rational characteristics of nodes. When selecting the forwarding node, neighboring nodes decide whether to participate in forwarding data through game, and prevent the destruction of malicious nodes by calculating the utility function. They control and adjust the game process and results through incentive mechanism. GBR (Game theory Based Routing algorithm) reduces packet loss rate by the above mechanisms and strategies and effectively controls the self-ish behavior of nodes. In addition, GBR pursues security routing and makes design and improvement in the energy conservation, which considers node security and regards the energy of the nodes as an importing reference index, in order to balance the network energy consumption.

Abstract: The traditional smart security systems have poor extensity, delayed alarm systems, and cannot be controlled through the remote network. To solve the shortcomings, this article proposed a smart security system based on Sensor Web. The system used Sensor Observation Service to achieve the data transmission and storage in real time. Users can send requests to Sensor Observation Service to get home environment information in real time via Internet. The system also implemented Sensor Event Service, so that users can get home security alarm information in real time by subscribing the events they are interested in.

Abstract: An improved ANN controller is presented inspired from hormone modulation function. This ANN controller consists of the main ANN controller and the conventional controller. To increase the learning efficiency, the slop of the excitation function is changed by the correcting parameters according to the hormone modulation law. To improve the control accuracy, we chose the accumulation of control error during the regulating process. And to avoid the integrated saturation, we judge the input of BP based on the absolute value of error. The main ANN controller adjusts the control input of the secondary conventional controller. To testify the effectiveness of the improved ANN controller, we apply it on the experiment device of offshore oil platform. The results show that the improved ANN controller has better control performance than the conventional controller and the normal ANN controller.

Abstract: Improving the traditional remote monitoring system requires the implementation of a wireless sensor front-end and to develop communication server and client applications. All the relevant control devices must be developed parallel to one another, which likely entails involved developments and difficulty in controlling the development cycle as well as cost. Based on the Ayla IOT (Internet Of Things) cloud platform, this paper proposes a design method of the remote monitoring system for boilers. The Ayla cloud platform provides the standard IOT control services that are idea for this task, for it only allows for the development of crucial elements such as the front controller, client application, Internet-based communication, state synchronization, as well as message driven, SNS release, database storage, and short message gateways. What’s more, these functions do not require separate developments, which greatly reduce the burden of parallel development of within an IOT remote control system. This in turn reduces costs and maintenance. Furthermore, the solution system can also be applied to the wireless control systems of electronic control devices, which should further anticipate its far reaching applications.

Abstract: A new Cuk converter with high step-down ratio was proposed. The operating principle and operating mode of the new Cuk converter were analyzed in detail and the EL model of the new Cuk converter was established. Based on passivity-based control theory, a passivity-based controller of the new Cuk converter was designed by the method called damping injection. The controller can enable the new Cuk converter achieve good dynamic and static performance and robustness to the load. Compared with Buck converters, the new Cuk converter has a higher step-down capability and a smaller ripple of the input current. Simulation results show that the new Cuk converter is feasible.

Abstract: Target localization was generally achieved by range and azimuth measurement for USBL (Ultra-Short Base Line) positioning system, of which the positioning accuracy was restrained by the array aperture. It usually adopted redundant array elements to solve the inconsistence between high precision and phase ambiguity. However, this routine would introduce the additional phase shift calibration problem for the complex and asymmetric array elements, which was not easy to get good results. To solve this problem, an anti-ambiguity method called dual-pulse signal is proposed in this paper to ensure the correct positioning for the improved array formation. Numerical simulations indicate that the proposed method can effectively improve the positioning precision of the USBL positioning system.

Abstract: In this paper, combined with task-driven way, the PBL teaching method was introduced into PLC control technology course. Four stages were adopted to implement the teaching process. Firstly, teachers instruct students to find problems. Secondly, students search materials and discuss the assigned tasks. Thirdly, teachers introduce some basic knowledge and instruct students to discuss and solve problems. Finally, teachers summarize the rules. It is beneficial to improve students’ learning initiative and the ability of discovering, analyzing and solving problems.

Abstract: On the basis of Dick Stella algorithm and Bessel Curve, this paper proposes an improved algorithm for real-time obstacle avoidance control of mobile robots. With strong robustness, this algorithm can effectively solve the problem of robot path planning and tracking while meeting the dynamic constraint conditions. The simulation results show that the improved algorithm shortens the robot moving period and remarkably improves the motion planning in robot obstacle avoidance.

Abstract: Sheet incremental forming is a new sheet metal dieless forming technology. This paper introduced the fundamentals of the sheet incremental forming process. Based on the principle of “layered manufacture” in rapid prototype technology, this process resolves the intricate three-dimensional geometry information of the workpiece into a series of two-dimensional data, which can be used by an NC system to control a forming tool to make a curvilinear movement over the raw sheet metal layer by layer until the component wanted is formed. This paper introduced the sheet incremental forming system and metal digital forming technology. An FEM model of the incremental forming process is established, and a typical process is analyzed to instruct the parameters selection and the optimization of the forming tracks.