Piezoelectric vibration energy harvesters as an autonomous power source for wireless sensor networks have attracted increasing attention in recent years. Conventional vibration harvesters are linear mass-spring devices with single beam. In order to improve the output power of vibration energy harvesters, a piezomagnetoelastic beam array configuration for vibration energy harvester is proposed. It consists of piezoelectric beam array harvester coupled to permanent magnets to create a bistable system. Electromechanical coupling model describing the proposed nonlinear vibration energy harvester is built along with computing simulations. The simulation analysis shows that the proposed nonlinear vibration energy harvester generates larger output power within several excitation frequency ranges compared to conventional vibration harvesters.

An adaptive slicing algorithm based on stereo lithography (STL) model was proposed. The main principle is, firstly, reestablishing the topological information of STL model. Then the intersection points were calculated and sequenced by using intersection points' tracking method and marking method. The slicing thickness was automatically controlled according to the area deviation ratio between two adjacent slices. Then the adaptive slicing process based on STL model was implemented. Based on the algorithms, a rapid prototyping slicing software STLSlicer based on STL model was developed. With real cases, it is demonstrated that the software runs stably with high efficiency and good reliability.

This paper discusses how to optimize the digital graphic program with cache system used in GPU/CPU architecture to gain more FPS. Firstly, we introduce the basic principle of cache system summarily; secondly, we discuss the three main organization and mapping technologies of cache system in detail, and then compare these three cache mapping solutions by giving examples; thirdly, illustrate the cache-friendly programming style, how to avoid cache race and cache split in programming.

Aimed to lack of incarnating the contents of driving knowledge in the distributed vehicle driving simulator (VDS), such as road sign, Highway Code and driving skill, this paper brings forward the framework of driving knowledge expert system based on WEB, assorts the driving knowledge, and constructs driving knowledge database which may interface timely. Thus, the driving knowledge can be melted in the process of virtual driving. Therefore we can not only learn driving knowledge through expert system of driving knowledge, but also experience the feeling of driving on the real road during the process of virtual driving.

The paper presents a distributed interactive vehicle simulation system for driving training based on the distributed interactive simulation (DIS) technology. It describes the system requirements, presents the system design particularly the vision subsystem design, and discusses the key techniques for the implementation of the proposed system including real-time communication, model partition, and dynamic camera control. A prototype system has been deployed to validate the proposed concepts and techniques and proved that the Distributed Interactive Simulation techniques can meet the needs of "Immersion", "Real time" and "Interaction " of the proposed vehicle simulation system.

With the rapid development of Internet, the fields related with distributed virtual reality are greatly enlarged. As one of the excellent virtual reality software development tools, Virtools is remarkable for its strong capability of interaction. Through an in-depth study on Virtools, the authors have developed a distributed driving simulator system based on the 3D engine, and studied the distributed network system with a real case. At the same time, some researches on the fundamental techniques of distributed virtual reality, such as real-time prediction, dead reckoning, and so on, have been conducted in the paper

Based on distributed virtual reality technology, the vehicle simulator is studied in this paper. A model and construction of vehicle simulator is put forward. The vehicle simulator system is composed of six parts including input and output system, controlling system, sense, feedback system, etc. 3D modeling on virtual sense, dynamic model of vehicle control and the description of knowledge of controlling vehicle using ANN (artificial neural network) and fuzzy theory, etc., are discussed. The integration of the key technology above discussed is useful for improving the performance of immersion and interaction of the vehicle simulator. The vehicle simulator will be applied to training driver and studying performance of driver.

This paper is concerned with the design and implementation of a control architecture for a mobile robot that is to navigate in dynamic unknown indoor environments. It is based on the framework of Open Robot Control Software @ KTH, which is discussed and evaluated in this paper. As a hybrid architecture, it is decomposed into several basic components which can be classified as either deliberative or reactive. Each component can concurrently execute and communicate with another using unified communication interfaces. Scalability and portability and reusability are the goals of the design.

Three-dimensional terrain generation is one of important factors of the natural scenes in virtual environment. It is also a significant composition of view system in virtual reality field. This paper introduces the necessary of terrain simplification in virtual browse, and concludes the existing algorithm. Then reduces the realtime algorithm of dynamic level of detail terrain rendering and describes its realizing process in detail. At last, this paper also presents the method of removing cracks on the meshes boundary.