Advanced Materials Research Vols. 791-793

Abstract: Through computer-related technologies, virtual reality allows the user to appreciate the induction of vision, auditory, tactile sensor and immersive feel in an interactive virtual environment. Because of the above-mentioned features, it has become one of the hot topics of the direction quickly. This paper discusses virtual reality simulation system, takes sports training simulation as an example, analyzes its key technologies and makes more people to understand the virtual reality simulation system. It provides some scientific ideas for the combination of virtual reality simulation with many other areas of development. The discussions about the key technology of virtual reality simulation system in this paper provide some reference for the development of many other cross-cutting areas combined with the virtual reality to a certain extent.

Abstract: At present, there are many commonly used maximum power point tracking (MPPT) control methods for solar photovoltaic system which have different degrees of output fluctuations. In order to solve the problem and improve the response speed further, a novel MPPT algorithm based on the extreme learning machine was put forward in this paper. Using the extreme learning machine which has small error, fast speed and simple structure train a mathematical model with the environmental temperature and light intensity as input and the maximum power point as output, the model can be used as the controller of the photovoltaic cells. The simulation results show that the error of the model meets the requirements, when the ambient temperature and light intensity change, it can respond quickly so as to photovoltaic cells work at the maximum power point and operate stably.

Abstract: For analyzing the accuracy of wind power prediction, an analyzing model combined with multi-leaner and dynamic weight distribution is proposed. With this method, Numerical Weather Prediction (NWP), Wind power data (historical) and weather data (historical) are structured into several sample sets, each set has a different weight value, which determined by the training errors, these sample set is trained by different learner algorithm with a weight too. Finally, using these models to predict the outputs. The experiments indicate the effectiveness of the method this paper proposed. Compared with Single model of Support Vector Machine and Artificial Neural Network, the combination method has better performance in both calculation accuracy and generalization.

Abstract: In order to resolve the vehicle routing problem with soft time window, a kind of Partheno-genetic Algorithms combined with Simulated Annealing was proposed in the paper, inverse operator and 2-change operator were presented.Centre point was replaced by the dummy natural number, then it is easy to made use of the available methods using by TSP. A selection method with tournament of three copies can keep the diversity of population. The simulation results show that new algorithm can effectively resole VRPTW, and get better results than common GA, new algorithms searching efficiency and convergence probability are effectively enhanced.

Abstract: In this paper, we present a building earthquake disaster simulation system. The system can estimate the destruction of buildings by different earthquakes, and analyze the quake-proof ability of different buildings. Moreover, based on the multi-input method of certain dynamical analysis, the system uses 3D, GIS, and VS.NET, and implements a building earthquake disaster simulation platform, which can present the destroyed results visually.

Abstract: The paper allocated a method to optimize the management pattern under cloud computing environment according to the current situation of Enterprise Information Management. The method used Ant Colony Optimization (ACO) as basic foundation to satisfy the property of Cloud Computing. CloudSim was also set up as the simulation to imitate the cloud environment and the test of computing. The algorithm prognosticated the capability of the potential available resource node when being allocated and analyzed the usage of bandwidth, the quality of networks and the response time. This algorithm met the needs of limitation of resource with better performance and has shorter response time.

Abstract: LTE system uses OFDM, MIMO and other key technologies, then signal detection is of great importance in LTE system. Several classical signal detection algorithms are analyzed and discussed, and an improved V-BLAST algorithm is presented in this paper. Meantime, the simulations are conducted with Matlab to analyze their performances. Compared with the traditional algorithm, the improved one has much better performance, which could be applied to TD-LTE RF conformance test system.

Abstract: The dynamic bit-locking backoff (DBLBO) anti-collision algorithm was proposed on the basis of bit-locking backoff (BLBO) anti-collision algorithm . If there is only one collision bit When the reader is searching the locked collision bit , the reader can identify directly without queries.The proposed algorithm gives full consideration to the number of queries and throughput of the system.The analysis on simulation result indicates that DBLBO performs significantly better than the existing BLBO anti-collision algorithms. It is suitable for the RFID anti-collision protocol in a greater deal.

Abstract: Operation and trouble shooting of ballast water system becomes a legal training item for marine engineers in STCW convention of IMO. Multi-mode simulating training system designed combines the actual devices with the simulating system. Ballast console mode improves the third dimension of training. Simulating software interfaces can meet the requirement of more trainees. The new configuration of the simulating system is benefit to the learning of actual devices. Convenient adjustment of the velocity of system process improves the efficiency of training. Several years of operation proved that the system is economical, reliable and the training effect is perfect.

Abstract: The high-order schemes have attracted more and more attention in computational fluid dynamics (CFD) simulations. As a kind of high-order schemes, weighted compact nonlinear schemes (WCNSs) have been widely applied. In recent years, the highly parallel graphics processing unit (GPU) is rapidly gaining maturity as a powerful engine for high performance computer. This paper studies the heterogeneous parallel computation and implementation of a high-order CFD program on Tianhe-1A supercomputer system. The CFD program is intended for the solution of the Navier-Stokes equations on multi-block Cartesian meshes for aerodynamics research. The solver utilizes the high-order WCNS scheme for space discretization and Jacobi iteration method for time discretization. The performance analyses show that the single-GPU solver achieves about 8× speed-ups relative to a serial computation on a CPU core.