In order to improve the accuracy of target parameter estimation,the quaternion theory was applied to the twodimensional DOA (Direction of Arrival) estimation of the uniform circular acoustic vector array. Receiving signal model based on the quaternion model was established, quaternion vector was deduced, and a quaternion field spatial spectrum estimation algorithm was presented. Considering software and hardware of algorithm realization, theoretical analyses show that the required storage space and computing operations of the quaternion model significantly decrease, and the simulation results show that the proposed MUSIC (Multiple Signal Classification) algorithm has higher resolution, stronger antijamming capability and the improved accuracy of the signal parameter estimation.

In order to capture and track the target accurately in complex scene, combined with the target feature, video tracking algorithm which based on gradient projection is put forward. The algorithm based on prior knowledge of object, we open window for the video and conduct gradient projection in the window to obtain the position information of object areas. By feature extraction and morphological analysis of target extracting feature parameters, to determine goals and capture. Using the object centroid coordinates to update the information of tracking window and achieve the target tracking.The results fully prove that the method reduces the amount of computation, gets strong real-time and tracks target accurately and stably, for the scene of experiment illumination change and suspected object interference has a strong robustness.

The drilling equipment simulation system based on 3DS Max was designed. Models combined accurateness and simplicity were created based on a variey of drilling equipment. Animations were added with 3DS animation technology to the model, the drilling equipment were rendered segmented and sub-structured, the simulation results output in AVI (Audio Video Interleaved) format, by the late integration, output the video tutorial with the animationg, sound, words together. Technical training tools of the drilling company was promoted to a certain extent through the system, reducing the accident rateof the drilling company.

In order to promote the digital and intelligent development of the frequency stabilization of diode laser, a digital assistant frequency stabilization method based on C8051F020 is proposed. The variation of frequency is obtained by detecting saturated absorption signal, and the system feeds back the error amount of frequency change to the current control and temperature control to keep the frequency stabilization of the diode laser. The hardware and the software of the frequency stabilization system are outlined. Experiment shows that it can automatically achieve frequency stabilization after power on, and can work continuously and stably in the long term, and losing lock condition can be quickly locked, the stability of the laser is 8.39×10^-11(1 s), 7.99×10^-12(1 024 s).

To solve the problem of the difficulty in estimating the parameters of Weibull distribution model, a new algorithm based on FPSO(Fuzzy Particle Swarm Optimization) with SA(Simulated Annealing) for estimating the parameters of Weibull distribution model was presented. In this intelligent algorithm, the inertia factor of PSO was turned by adopting fuzzy logic reasoning to improve the convergence rate of the PSO(Particle Swarm Optimization) according to the characteristic of particle's motion trajectory in longitudinal direction and lateral direction. Then combine the FPSO and SA, use the update formulas of FPSO as the state generating function of SA, and use Metropolis algorithm to accept the new state with probability to get the global optimal parameters' estimations. The parameter estimation method based on the intelligent algorithm above mentioned was used to estimate the parameters of Weibull distribution model, and the parameter estimation accuracy was improved. Practical application in the complex distribution model parameter estimation showed that the method was feasible and effective.

For a better pressure tracking performance of DCT (Dual Clutch Transmission) clutch electro\|hydraulic control valves, a control-oriented nonlinear DCT electro-hydraulic control valve model is presented and the backstepping technique is used to derive the nonlinear controller. To verify the effectiveness of the proposed controller, a simulation model of DCT clutch electro\|hydraulic control valve and clutch is established in AMESim environment and simulation combining Simulink and AMESim are conducted tracking different references, with the influence of the bulk modulus changing and temperature changing of oil considered. As comparison, simulation results with the control of a well\|toned PI controller under the same circumstance as that of the proposed controller are given. Finally, simulation results are provided to demonstrate the effectiveness of the proposed controller.

The current Bluetooth technology is high-power, not suitable for multiuser laboratory research and development applications. The system uses the STM32 processors Speex speech coding and decoding algorithm technology to achieve infrared wireless communication in a short distance via voice transmission module, relaying node module and receiver module enhances the function of the voice device. Test results show that the output noise voltage of the receiving device is smaller than 0.1 V, the relaying station can change the transmission direction in any degrees, the speech signal safe, stable transmission range up to 8 m.

To resolve the problem of access point overload in WLAN (Wireless Local Area Network), a loadbalancing algorithm based on the access point selection and adaptive handoff is presented. In the algorithm, RSSI (Received Signal Strength Indication), MAC (Media Access Control) frame transmission delay and other f
actors are taken into consideration to select the access point. Meanwhile, it makes a positive response to the station of gusty traffic with minimal access point handoff to balance the entire network load. The algorithm can explicitly select the access point of the idle channel, and balance the load of access points. Then it effectively improves system throughput and reduces transmission delay of the packet. Compared with the traditional algorithms, it improves system throughput by 13.6%.

Because traditional Linux distribution has the problem of complicated drivers, graphical desktop of  extreme resource consuming and excess software, we build a system which is optimized for a particular processor. Using GNU(GNU's Not UNIX) tool chain, Linux 2.6.25 kernel and related components we propose a micro system solution in USB storage devices. In the process of building the tool chain, confusing compile order has been marked and explained. We have a detailed discussion of establishing cross-compiler environment and the use of the toolchain. The solutions of root permissions, the kernel USB related module selection, and grub boot failure are elaborated. This solution, based on the host system of opensuse 11.4 and the open source, has achieved the purpose of completing the Linux system development in x86 PC and USB flash disk. The result shows that in the low-cost test bench the system takes about 600 MByte of storage space, provides a complete teaching programming requirement, and ensures faster start-up and operation rate.

In order to improve the traditional proton magnetometer comprehensive performance, and provide reference basis for developing Overhauser magneto
meter, JPM-1 proton magnetometer is introduced, which is based on DSP(Digital Signal Processor) TMS320F2812 and is different from previous domestic
 magnetometer. DSP is used to acquire and process the precession signal, and the frequency counting method based on zerocrossing counting algorithm is employed. Initial signal strength and the noise characteristics are discussed, and the measured results fit well with the calculated results. Final resu
lts show that the initial signal strength is about 0.4 μV with SNR(Signal to Noise Ratio) 32 ∶1, and 0.27 nT sensitivity is obtained in field measurement.

On the market, the function of traditional water dispenser is relatively simple, low intelliget. In order to meet the technical requirements on the market and update the traditional “sensor-relay” mechanism, we use MCU(Micro Controller Unit) as the control core, adding temperature detection, the water level
detection, temperature setting and mode setting, liquid crystal display circuit and so on, to design standardization and intellectuality dispenser electronic co
ntrol system structure, which significantly improves its overall performance and used functions. Experimental test show that it can achieve automatic cont
rol of the dispenser and realize energy-saving security and intelligence integration of the dispenser.

In order to solve machine vision reduce the effects of background and noise on the video feature extraction, and to improve recognition results. We propose
 a human action analysis method by combining the local space-time interest points and global accumulated edge image feature based on the study of human action representation and the full consideration of the advantages and disadvantage of the local and global features. First, local space-time points of interest and global accumulated edge image feature are extracted from the video sequences. Then we use the weighted dictionary to combine them together organically, finally we use the minimums distance method for human action analysis and identification. This method can effectively obtain the spatial characteristics of human, human edge contour and the human movement trends and intensity, the experimental show that our method is faster and gains a higher recognition accuracy generally inceased by 2% to 5%.

The problem of faulty localization during program debugging is studied, and execution trace and dynamic slicing are applied to faulty localization. Irrelevant statements of fault localization are included in execution trace, accuracy of fault localization is affected by these statements. The influence of irrelevant tatements is reduced by dynamic slicing technology based on execution trace. The model of suspicious degrees computing based on execution trace and dynamic slicing had been established, and suspicious degree of each statements could be computed by this model. After sorting these statements by their suspicious degrees, recommendation of fault localization would be reached. The experimental results are compared with other algorithms, and the method of faulty localization based on execution trace and dynamic slicing is proved effective.

A method which aims at attribute reduction based on assignment set in numerical data sets is proposed. Firstly, the concept and some important properties of assignment set based on numerical data sets was presented. Then, authors defined dicernibility matrix based on assignment set, and proposed a new
 method for attribute reduction based on discernibility matrix in numerical data sets. Finally, an algorithm for computing discernibility matrix was presented.
Example analysis show that this method can effectively handle attribute reduction in numerical decision systems.

Because of the influence of complex image background, illumination changes, facial rotation and some other factors, face detection in complex background is much more difficult, lower accuracy and slower speed. Adaboost algorithm was used for face detection, and the test process in OpenCV was implemented. Face detection experiments were performed on images with facial rotation and complex background. The detection accuracy rate was 85% and 99% respectively, the average detection time of each picture was 16.67 ms and 76 ms. Experimental results show that the face detection algorithm can accurately and quickly realize face detectionin in complex background, and can satisfy the requirements of real-time face recognition system.

In order to solve the problem of data organization and storage as electronic computer, a DNA computing model of stack based on hairpin structure is proposed. The method for data storage is described. And data operation method is demonstrated. The idea of this paper is feasible to solve data organization problem for DNA computer so that it can help DNA computer development for practical application.

To solve the microarray data problem that the data has small sample size and each dimension of the sample is high, therefor there is a lot of interfering information with redundant information in the data. The multi-stage algorithm of informative gene set selection is discussed in this paper. First, the difference of single gene in different condition is considered, genes with more differences in special condition are selected as candidate gene set. Then, the genes with less correlation are rejected. Finally, the better gene sets based on the global classification performance of any set are selected. The experiment result shows that the algorithm is feasible and effective for informative gene set selection. The feature set of genes in both fitness of classification (classification ca
pability metrics) and classification accuracy is more accurate, and more efficient than the raw data.

This paper studies the contagion risk in the payment system based on network model. Using current research result for reference, we describe the financial contagion through delay-differential equation and get the analytical solution with the help of renewal process. Then we analyze the contagion risk in different time-delay parameters based on the analytical solution. The results show that contagion risk significantly increases as time passed. Time-delay parameter also has a very strong effect on the contagion risk. Then we present related suggestions and policy implications which may improve the construction of China national automatic payment system and prudent macro-economic management both in theory and practice.