Advanced Materials Research Vols. 403-408

Abstract: Traditional clustering ensemble combines all of the available clustering partitions to get the final clustering result. But in supervised classification area,it has been known that selective classifier ensembles can always achieve better solutions.Following the selective classifier ensembles,the question of clustering ensemble is defined as clustering ensemble selection.The paper introduces the concept of clustering ensemble selection and gives the survey of clustering ensemble selection algorithms.

Abstract: A novel spatial filtering collimator is developed for the laser indication application. The mass produced, inexpensive, red and blue semiconductor laser modulars can easily be optimized as fine line laser indicator source. We have theoretically investigated the results, simulated the designs, and measure the laser beam spots and the laser fine lines. The spatial filtering collimator is designed and optimized for practical used and the results are very satisfactory.

Abstract: The understanding and description of behaviors for road vehicles is a hot topic of intelligent visual surveillance system. Trajectory analysis is one of the basic problems in behavior understanding, from which anomalies can be detected and also accidents can be predicted. In this paper, we proposed a hierarchical self-organizing neural network model to learn trajectory distribution pattern and a probability model for accident recognition. Sample data including motion trajectories are first get by real-time vehicle tracking. The self-organizing neural network algorithm is then applied to learn activity patterns from the sample trajectories. Using the learned patterns, we consider anomaly detection as well as object behavior prediction. Experiments in actual road scene show the effectiveness of the proposed algorithm.

Abstract: This paper researching on vehicle collaboration distribution of TPL alliance, discussing the solution to the arrangement of vehicle routing problem which is base on “multi-deport” vehicle distribution of the triangle network. As for the key situation of vehicle distribution, clusters the supply and demand nodes into sliced area in order to constrain triangle distribution model and realize the triangle closed-loop route of vehicle distribution in both pickup and delivering. By comparing relative advantages of the traditional distribution model and the triangle distribution model on many code aspects; then stimulate the problem with improved ant-algorithm; finally analyzes the results of stimulation. The results imply that the triangle network of vehicle distribution can reduce unload ratio and the cost of transportation, can improve full-load ratio and can achieve the optimal allocation of transport resources.

Abstract: A system of wireless sensor network(WSN) based on laser rangefinder is studied, which adopts low power consumption ZigBee chip and SimpliciTI network protocol. 8 laser rangefinders are used to collecy the distance of the test points. The collection data is sent to upper computer by wireless data transmission module. It is rapid and real time to monitor the test point. The system can be installed in complex test environment to perform the monitor of all kinds of architecture structure. It dispenses with long time power supply and the user interface is friendly and has abundance function.

Abstract: In order to identify projection light color information of the color structured light rapidly and accurately, this paper proposes a projection light color correction algorithm based on “Dichromatic Reflection Model”(DRM). This algorithm revises the effect on projection light color which caused by object surface’s color and highlight, through estimating three-channel spectral reflectance of the object, separating specular and diffuse components. The experiment shows that the proposed method has more advantages than the color recognition of color structured light which based on DRM, such as high robustness, fast calculation and high accuracy.

Abstract: As the need for broadband access in high speed railway, it becomes necessary to understand the impact of mobility on traditional parameters of WLAN standards always working under stationary environment, such as throughput. In this paper, we provide the analysis of the impact of high speed in IEEE 802.11b PHY layer, and then propose the derivation of saturation throughput by applying 802.11b MAC schemes in TGC system. Our results show that WLAN standards have promising potential to be used in high speed environment.

Abstract: Severe energy constraints and limited computing abilities of the nodes in a network present a major challenge in the design and deployment of a wireless sensor network. This thesis aims to present energy efficient algorithm for data delivery in a sensor network. The methodology of data delivery presented in this thesis is based on affine transform and performs coordination compression. Experimental results show that the algorithm can reduce about half of the data amount delivered by the sensor.

Abstract: For a mini remotely operated vehicle (mini-ROV), the performance of the propulsion system has a direct impact on its underwater running and operation. A proper designed propulsion module will improve the maneuverability and controllablity of the mini ROV. In this paper, we propose a kind of propulsion modular series taking lower drag, higher efficiency and lighter weight as design objectives. Simulation and experiment results are also presented to illustrate performance of a specific designed thruster.

Abstract: Time-varying parameters identification in linear system is considered, which can be changed into time-invariant coefficient polynomials after Taylor expansion. Using response data to establish the time-varying autoregressive moving average (TV-ARMA) model, then utilizing least-square algorithm to obtain time-invariant coefficients of time-varying parameters. According to error analysis, to reduce errors and improve accuracy, the estimation time is divided into small internals and the above method is used in each interval. Simulation shows that, under certain error condition, the time-varying parameters obtained by the method have good agreement with the theoretical values; the measures taken have strong anti-interference and high efficiency.