In order to achieve the diagnosis of fault categories of permanent magnet synchronous motor, fault feature is extracted by using wavelet function according to the different frequency band, then normalized date sample as the SOM(Self Organizing Map) network input, the SOM field function is constructed with the wavelet function and the second excited neurons are formed to update weights, so the local optimization of SOM is avoided. The fault data extracted is regarded as the input samples of SOM neural networks in order to train the network, and appropriate neuron index stimulated when a specific fault occurs is obtained. the feasibility has been demonstrated by simulation results.

In order to solve the problems of delay time existed in tactical data link communication and nonlinearity of measurement information, a delay time compensation algorithm based on unscented kalman filtering is proposed combined with a current statistical model. Nonlinear transformation is applied to state prediction, avoiding accuracy decline brought by model linearization. By analyzing delay time and its impact on tactical data link, the current statistical model of the target is established, unscented kalman filtering is adopted to track the target, and delay time compensation algorithm based on the UKF(unscented kalman filtering) algorithm is added to compensate displacement error and velocity error. The simulation results have verified indicate the effectiveness of the proposed method.

The seismic exploration technology needs vibroseis to provide seismic wave with highly autocorrelation function. To solve this problem, we put forward a method of using m sequence coding pulse signal to design vibroseis signal. Firstly, we use Matlab to generate m sequence and pulse signal, then we use m sequence modulating pulse signal to generate scanning signal, finally we realize scanning signal through hardware circuit. We use the GEIST seismograph to acquire the scan signal, and use Matlab to analyze the data. The results show that the seismic wave signal has high signal to noise ratio.

In order to fill the vacancy of accessing campus information on mobile devices, a campus information inquiry system is designed and implemented. The mobile client applies popular J2ME Java (Plalform Micro Edition) MIDP and J2ME Polish framework to adapt a variety of mobile operating systems. GSM (Global System for Mobile Communication) Modem hardware, modular architecture and half-synchronous/half-asynchronous framework are applied to the server implementation to provide smooth update, ensuring the fairness and reliability of services on the server side, to give a solution on the interaction between the server and clients and a way of implementing queries through GSM messages. The benchmark results verify the availability, reliability and security of the system.

In order to improve the informationalized level of distributing containers and reduce the problems caused by manual operations.We using embedded processor as core, design a data terminal for distributing containers, which combines RFID, GPS and WIFI. We first introduce the proposal of the whole system, and discuss the hardware interface and the flow of the software. At last the experiments of managing containers distribution are carried out. The results indicate that the terminal is able to improve the informationalized level of distributing containers and has a good performance. Compared to manual operations, this terminal has such advantages as high efficiency, easy operations and low error rate.

In order to deal with the reality of large scale microcosmic management process simulation, a novel application scheme is proposed as. Based on war game process simulations, an artificial emotion application model is presented for management process simulation. Its emotion information inputs and data process flows are given. The model utilized time slice methods and event methods to establish microcosmic management process driver mechanism to synchronize management processes and virtual emotion states. Further more, an artificial emotion memory model is presented to deal with the consistency and cooperation indetermination among multi simulation objects. Two implement examples are given to describe the model flows. Experiment results show that artificial emotion can make microcosmic management process simulations factual, and given high efficiency to simulation results.

In order to assist an effective decision for the online lending borrower, we apply Logistic regression in to calculate the probability for the securing a loan. Based on the Logistic regression of the data from PPdai.com, we build new kinds of tools for decision makers to help the borrower to quantify her strategic options to get loans. In the train
ing set of the prediction model, the comprehensive prediction rate reached 89.6% and the comprehensive prediction rate in validation set is 85.8%. Such result indicates that our model is valid.

It is necessary to give more efficient method for forecasting the estimate telecommunication. A model based on Wavelet Neural Network is introduced in telecommunications to forecast the income and the uesers. To build wavelet basis NN (Neural Network), the sigmoid function is replaced with the wavelet in NN, and adaptive-learning rate and initialization of parameters are also used to obtain high probability and high convergence speed. When a model of wavelet neural network is established to forecast the gross service in telecommunications, subjective guided data is introduced in consideration of the impact of industry convergence on future telecom industry. 

The traditional algorithm of 3D face reconstruction is inefficient and difficult to meet the requirements of practical application. To address this problem, a 3D face fast reconstruction algorithm is proposed. The algorithm based on improved ASM (Active Shape Model) is used to automatically and accurately positioning the feature points, and the improved ASMbased algorithm is used to normalized 3D face database. The sparse 3D morphable model is established to fast reconstruction 3D face from the 2D frontal face; Finally, shading information of 2D image is fully used as a prior knowledge to constrain each vertex's normal located in feature region of 2D face. Experimental results certify that the proposed method obtains faster speed to reconstruction 3D face from the 2D frontal face. keeping a quite high accuracy results and having a high recognition rate.

Task allocation model based on single objective function can not provide more useful information for the fire control decision makers. In order to make up the deficiency, the wastage cost of UCAV (Unmanned Combat Aerial Vehicle) and damage value of target are treated as two optimization objective functions of the task allocation for MultiUCAV cooperatively attacking multiple targets, and a new task allocation model is established. Based on the optimization model, an improved NSGA-II(Nondominated Sorting Genetic Algorithm II) with elitist strategy is adopted for searching the Pareto optimal solutions of the task allocation of cooperative attacking multiple targets for MultiUCAV. The decision makers select the best task allocation scheme according to their preferences. Simulation results demonstrate that the algorithm of task allocation is convergent and effective.

Aimed at the large number of regression test cases caused by the features of graphical input/output, event driven, random event trigger in GUI (Graphical User Interface) software, and on the basis of GUI event model, a mathematical model of regression test suite optimization for GUI software is constructed. The objective function and constraints in the model are given. and an ant algorithm is presented to solve the problem. Ant pheromone update rules and ant path selection rules in the algorithm are set. Simulation results show that under the premise that coverage is guaranteed, this method can reduce both the number and length of test case effectively.

In order to limit the speed of vehicle on highway, an speed measurement system based on CMOS（Complementary Metal Oxide Semiconductor）digital camera with rolling shutter is established and its calibrating technology and speed measurement algorithms are investigated. Rolling shutter digital camera's speed measurement is that using exposure of rolling shutter with a certain delay, producing a certain distortion when take photo of the moving object, and exploiting the image processing techniques to process and calculate the part of the image distortion, then get the speed parameters of the moving object. The camera calibration is the foundation and prerequisite for
 CMOS digital camera measuring speed of the moving object. Based on the principle of the camera imaging, the text analyzes the camera calibration theory. According to the characteristics of the camera's measuring speed, the calibration method fit for speed measurement. Following introduces the methods and procedures of camera calibration experiment, then finishes the experiments. Experimental results indicate that the error of the speed measurement system using this calculation method is within 2.5%. It can satisfy the speed measurement system.

In order to reduce the PSO (Particle Swarm Optimization) in time and space on the complexity of the problem, the GPU(Graphic Processing Unit)for parallel computing method is analized. Using GPU parallel characteristic, Particle PSO path search process in GPU platform is achied in paralled, Function test results show that the GPU mode is greater than the CPU mode calculation improve search speed.

The traditional experiment course is usually carried out in a specialized laboratory, which could produce undesirable results due to the factors such as time limit and equipment constraints. Therefore, it is very necessary to develop a set of virtual experimental system of computer organization principles. The system was designed for the laboratory equipment in our school. In order to realize this virtual experimental system, the specification of objectoriented technology was followed, UML（Unified Modeling Language） method was adopted for modeling, and the language VC++ was used. The innovation of the system is that the UML based design for functional unit was adpoted, and the complete simulation of the device was realized. After a practical use of 327 students of two grades, the satisfaction rate reached 81%, which laid the foundation of the virtual laboratory in our school.

Based on the idea of virtual manufacturing and computer simulation, using Microsoft Visual Studio 2008 development platform, C#, XNA(Windows Xbox Next Generation Architecture) and 3ds MAX development tools, a virtual NC(Numerical Control) milling machine simulation system is realized. This system is able to accurately simulate the
real various operations of NC milling machine at different speeds and NC code. Using the system users can reach the initial experience of operation and NC milling programming and so on. It helps save the time of using real NC milling machine, which will reduce studying and training costs.

In order to solve the regulators can be real-time monitoring of radiation enviroment woorkers dosimeter, we introduce a design proposal of personal radiation dosimeter based on the GPRS(General Packet Radio Service). Through analyzing the functional module parameters in detail, then system circuit design is completed. The designed system has realized the remote monitoring of personal radiation dosimeters, and the circuit structure is simple, it has low cost, low power consumption advantages and soon, which is widely used in radiation monitoring.

In order to provide a userfriendly interface to control smart appliances for home users, we use virtual reality technology to create 3D models of the house, and allow the users to interact with the appliance control system immersingly. The virtual tour system based on Microsofts DXUT(the DirectX Utility Library) framework is designed using Visual C++ 2010 as the development platform, in the roaming process, user can select smart appliances and interact with them. In actual use, the user get a very good experience and the controling of intelligent home appliances are also very convenient.

Based on target detection range and the distance, small test target, target and background low contrast problems, puts forward the characteristics based on linear template matching algorithm of target detection range in positioning method, in order to cross a features instead of the two dimensional linear rectangular template matching calculations, reduce the computational complexityin matching algorithm and raised the speed of the algorithm and ensure the accuracy. using the CCD（Charge Coupled Device） camera shooting range of target image sequences the simulation experiment is conducted. the experimental results show that the characteristics in straight line instead of a rectangular template matching algorithm related calculation of template, and improve the system of the target tracking guarantee the accuracy rate algorithm, and has very strong robustness.