Advanced Materials Research Vols. 760-762

Abstract: To master the variation regularity of finance, obtain greater benefits in stock investment. study of the support vector machine and application in prediction of stock market. The simulated annealing algorithm to optimize the least squares support vector machine prediction model, and the least square support vector machine and simulated annealing algorithm is described, given the optimal prediction model. Through the research on the simulation of the Hang Seng Index, shows that this method is simple, fast convergence, the algorithm with high accuracy. Has the actual guiding sense for investors, the stock market of the financial firm to operate.

Abstract: This paper proposes that elements of supplier risk assessment can be divided into two parts, indicators of core competencies and executive capacity. It uses simple statistical method to establish indicators of supplier core competencies, and Delphi technique and Network AHP to establish indicators of supplier executive capacity. We can find out the relationship among the key indicators of supplier risk by the survey of Delphi technique, the weighting and ranking by Network AHP and set up the system by Super Decision.

Abstract: Target tracking algorithm based on Mean-Shift and Kalman filter does well in linear tracking. However, the algorithm might lose the target when the trace of mobile target is curve or the acceleration is not constant. To cope with these drawbacks, this paper proposes Target Tracking Analysis Based on Corner Registration. The algorithm modifies the initial iteration center of Mean-Shift by using the corner features combined with affine transformation theory and then the Mean-Shift can track the target. The theoretical analysis and the experimental results demonstrate that this method can overcome the drawbacks we talk above and make achievements in target tracking.

Abstract: According to the requirements of modern education theory and technology, combined with the actual teaching conditions of Computer Operating System, an item bank system is designed and realized by adopting the VB 6.0 and SQL Server 2000. The system benefits for separating testing from teaching and settles the problem of randomness and unrepeatability caused by manual paper generation. In this paper, some key techniques, such as system function design, the database table design, making-up strategy, Word-paper making etc. are introduced, which have very strong universal property and operability.

Abstract: Firstly this article briefly describes the technology of topology optimization. Then the design space of frame of aircraft deicing vehicle is established and the model of finite element is gained by meshing. Loads and DOF constraints are applied based on the analysis of working conditions. The results of topology optimization are attained by calculation. On the basis of aforementioned results, the analysis is made. Finally, this paper makes a summary and outlook.

Abstract: A technique of high-speed network intrusion detection system based on packet sampling theory is proposed. Starting with basic principles of packet sampling, this paper first analyses the significant mathematical conclusion of sampling strategies, then after discussing current strategies, mechanism and performance of different packet sampling methods, we specify an efficient strategy of packet sampling. Results show that this method can attain above 55% accurate rate with below 1% false rate in 94 specified attacking cases from DARPA 2000 IDS evaluation dataset.

Abstract: TU-T has recommended a standardized scheme for achieving sub-50ms service protection in ring topologies through Ethernet Ring Protection Switching (ERPS) specified in G.8032 recommendation^[. ERPS provides a means to reliably achieve carrier-class network requirements for Ethernet topologies forming a closed loop. This article outlines an Ethernet ring protection switching mechanism and explains how its automatic protection switching (APS) protocol works. In addition, this article illustrates a sample scenario with failure and recovery.

Abstract: Nowadays, it becomes a hot research topic for autonomous flight of Quadrotor in the complex environment and the realization of fully autonomous flight is still a big challenge. The path planning of unmanned aerial vehicle is a key problem for its autonomous flight. For the path planning of Quadrotor, using the quantum particle swarm optimization algorithm, and made a lot of simulation and actual flight experiments. The results of simulation and actual flight experiment show that the using of QPSO for the path planning of Quadrotor is able to obtain a satisfactory result.

Abstract: To describe the performance of a new kind of approximation algorithm using B-Spline weight functions on feedforward neural networks, the analysis of generalization is proposed in this paper. The neural networks architecture is very simple and the number of weight functions is independent of the number of patterns. Three important theorems are proved, which mean that, by increasing the density of the knots, the upper bound of the networks error can be made to approach zero. The results show that the new algorithm has good property of generalization and high learning speed.

Abstract: An optimization-based algorithm is presented for the optimal operation of hydropower systems with cascaded hydro-plants. Continuous reservoir dynamics and constraints, discrete operating states, and hydraulic coupling of cascaded hydro-plants are considered in an integrated way. The main idea is to decompose the cascaded hydro-plants into individual plants, and then use Fuzzy Dynamic Programming (FDP), rather traditional Dynamic Programming (DP), to solve the subproblem of each plant while considering the hydraulic coupling of the plants. Numerical test shows that this method converges very fast, and is efficient and effective to deal with hydropower system with cascaded hydro-plants.