In the case of blind signal sparsity, in order to solve the problem of over-fitting which is caused by the termination conditions of OMP(Orthogonal Matching Pursuit) algorithm, an AOMP(Adptive Orthogonal Matching Pursuit) algorithm is proposed. It estimates reconstruction error of the original signal through additional observations in the iteration process. It adaptively determines the iteration corresponding to the best spectrum reconstruction by finding the minimum error point. Simulation results show that the proposed adaptive OMP algorithm can effectively improve signal reconstruction performance under low SNR and the reconstruction effect is as good as the case of known signal sparsity.

Because the existing wireless image transmission system has poor anti-interference ability and low confidentiality issues, an image transmission system based on VLC(Visible Light Communication) is designed.51 processor is used as the main controller, using visible LED(Light Emitting Diode) high-velocity modulation as information carrier, visible light of high-speed modulation emitted from the LED is utilized as information carrier,photodiode is utilized to receive optical carrier signals, the image information is obtained through the PC,wireless image transmission based on visible light communication is realized. The results show that the maximum bit transfer rate of the device is 0. 1 Mbit/ s when the communication distance is 0. 7 m, the BER(Bit Error
Rate) is less than 10-5. The system effectively overcomes the problems inherent in the conventional image transmission system of the wireless, it is a good supplement for short-range wireless communications technology,having a good prospects.

Traditional application MIL-STD-1553B ( Military Standard 1553 Bus) data bus requires many peripheral chips. Based on the study of MIL-STD-1553 bus protocol, 1553B interface module and FPGA (Field Programmable Gate Array) are designed. I use transformer coupling, through HI-1567 converter, send the signal to the FPGA. And I use abundant logic resources in FPGA to complete the bus data gathering, encoding, decoding and communicate with the PC. And I comprehensively analyze the principles and processes of FPGA control 1553B bus. Then, I give specific method of FPGA control USB interface. Finally, FPGA received data is displayed on the PC. The experiment results show that the plan using FPGA to realize 1553B bus interface, reduces a lot of peripheral circuits, improves the reliability and real-time performance of the system, completes the bus data sending and receiving, achieves a lossless data transfer.

BaLuF5 :Yb3+, Er3+ nanocrystals are synthesized with the method of high-temperature thermal decomposition. The average particle diameter of nanocrystals is 13 nm. The nanocrystals exhibit strong emission at 1530 nm. The fluorescence emission spectrum showed that the FWHM(Full Width at Half Maximum) is about 50 nm. The synthesized nanocrystals were doped into the SU-8 polymer as the core material of the optical waveguide amplifiers. The waveguide amplifier based on SU-8 doped by BaLuF5 -Yb3+,Er3+ nanocrystals was fabricated with the SiO2 upper cladding using photolithography and development process. For an 1 530 nm wavelength input signal power of 0. 1 mW and a 980 nm wavelength pump power of 280 mW, a relative optical
gain of 3. 95 dB for waveguide length of 1. 1 cm was achieved.

To solve the weak robustness of UKF(Unscented Kalman Filtering) of the DD-PMSG(Direct-Driven Permanent Synchronous Magnetic Generator) without position sensor, a rotor speed estimation method based on RUKF(Robust Unscented Kalman Filtering) algorithm is proposed. The DD-PMSG model is transformed into the augmented equations with electromechanical state and stator parameters. By applying the RUKF to the augmented model, the PMSG states will be estimated, and the uncertain parameters dynamic, in which the latter participates in the iterative process of state estimation and improves the robustness of state estimation against parameter uncertainties. The stronger robustness can be shown by comparing with the UKF method in simulation results.

Aiming at the vehicle roll rate sensor fault, a new model based method was proposed, focusing on fault diagnosis and fault reconfiguration. Comparing with the traditional method, an exact proof is given, which gets the strong robust residual, and the high high-precision signal of the faulty sensor. The multi-input-multi-output nonlinear system, which fulfills the Lipchitz Condition, could be took place by a equivalent model which is a one-dimension linear state equation with a model uncertainty, in the sensor fault diagnosis problem. The residual generator is designed by the classical observer theory, and through the eigenstructure assignment theory,a proof is addressed that the model uncertainty, actuator fault and disturbance could be perfectly decoupled from residual. The effectiveness of the result in testified by the 8-dof model in Simulink.

A control scheme is proposed based on improved exponent trending law to deal with chattering problem in UAS-SMC(Unidirectional Auxiliary Surfaces-Sliding Mode Control). First, to ensure the smoothness of system in the control range, T-S model is built to deal with aircrafts model change with control variable. Then an advanced sliding mode control with unidirectional auxiliary surfaces is analyzed, and an improved exponent trending law is proposed. The control scheme is designed based on fuzzy model and the new trending law. The simulation results show that the scheme can guarantee the rapidity of approaching process and suppress the chattering effectively.

Considering the condition of filter divergence problem in a relatively short time caused by small misalignment angle, based on the quadrotor intergrated navigation system, the central difference of kalman filtering algorithm is applied to the low cost integrated navigation system. First this paper realizes the hardware circuit of quadrotor integrated navigation system, the MPU-6050 chip is used for the attitude sensor, choosing NEO-6M module from u-Blox company to get a GPS(College of Science) signal. Then the central difference of kalman filtering algorithm is used to realize the multi-sensor data fusion, and to get the navigation information.Finally the system can be applied to actual quadrotor platform for experiment. The results of the experiment show that the design of quadrotor integrated navigation system based on CDKF(Central Difference Kalman Filter) algorithm can effectively restrain the sensor noise, overcome the gyroscope drift phenomenon, avoid the shortcomings of traditional EKF(Extended Calman Filter) algorithm, and provide accurate navigation information for aircraft.

Because the existing personnel positioning system is difficult to locate the personel position in mine accidents, a system based on double Mach-Zehnder interferometer structure is designed. The structure and positioning principles of the system is described, the positioning method of the system is easy to implement, with the advantages of a passive, flame-proof, anti-electromagnetic interference and remote control and so all, having great practical value. To improve the positioning accuracy of the system, combined with wavelet threshold noising features, a wavelet threshold noising cross-correlation delay estimation algorithm is preposed. The simulation results show that the algorithm can effectively improve the positioning accuracy of the system.

Aiming at the disadvantages of contact measurement density instrument, such as detection range limitation, we put forward a kind of non-contact volume measurement method. This method can be used to calculate the density by detecting the inner variation of container air pressure to indirectly measure the non-hollow solid density. A wide range of the measurement, a notable advantage of this device, can nondestructively measure the non-hollow solid matter which does not have chemical reaction with air. This method can also avoid the damage and the error brought by single drainage destruction. Within the range of pressure endurance, the container can increase the pressure of the compressed air chamber by controlling the air compressors properly, improving the measurement accuracy and making the error less than 3% .

A task assignment method was proposed to address the task assignment issue of camera networks in the intelligent traffic systems. The workload of camera node and the tasks quality of service are both taken into consideration with the introduction of fields of task. A task transfer approach has been put forward on the basis of the relative value between the adjacent camera nodes to ensure tasks quality of service. The proposed method has been implemented in the embedded smart camera networks system. The experiment result demonstrate that the method could improve both the system average resource utilization of about 50% and the success ratio of the task assignment is about 40% when compared to conventional method.

In order to satisfy the users needs of being in real-time control of the content on the large screen,a novel smart mobile control system is presented. It can control the mosaic screen by way of graphic operation on the platform of Android. An intelligent layout algorithm based on the world coordinate is also proposed. The images intelligent layout has been realized with the layout algorithm under the client edit mode. The communication between client and server is established by socket program with TCP/ IP protocol. And it has been widely applied to many cities with good social and economic benefits.

In order to solve the singular describing of candidate target in object tracking, an algorithm based on adaptive feature fusion of visual significant feature is proposed. The quaternion model of target composed of intensity, motion, color and the change of color feature is built, and the PQFT (Phase spectrum of Quaternion Fourier Transform) is used to extract saliency map. According to similarity coefficient, the fusion weights are adaptively adjusted and the color feature is combined to ensure accurate tracking. The experimental results demonstrate that the approach can effectively overcome the part occlusion and the interference of background, realizing the accurate tracking under the case of complex background.

To solve the problem that multiple instance learing can cause error accumulation in object tracking algorithm, a new object tracking algorithm was proposed based on co-training and online multiple instance learning. This algorithm uses co-training to overcome errors accumulation caused by self-training and improve the robustness of tracking performance based on online multiple instance learning. The center error between tracking results and ideal object location was used as evaluation criteria. The semi-supervised learning tracking algorithm and traditional multiple instance learning tracking algorithm were simulated on the video frames which come from standard video library. The tracking results show that the algorithm performance is more superior, and the center error curves further demonstrate the experimental results. Experimental results show that the proposed approach can track the target well in the complex conditions, and it has a better robustness.

According to the image fusion of optical and SAR(Synthetic Aperture Radar), the multimodal and multiscale features including pixel features, texture features and edge features were analyzed in order to improve the traditional homologous image registration and fusion algorithm. Then the improved SURF(Speeded Up Robust Features) operator, texture analysis and contour extraction algorithm were adopted to obtain the multimodal and multiscale features of the heterologous images. By standardization algorithm of the fuzzy scale and dimension, the differences between the feature pairs of the heterologous images were overcome, which made the matching of the feature pairs available. The accuracy of registration and fusion were ensured through the method of fuzzy
correlation coefficient, and the registration and fusion of optical-SAR images were completed. Finally, the modified algorithm was verified and compared with the traditional fusion methods. Experimental results show that the multimodal registration and fusion algorithm can improve the precision and adaptability of optical-SAR registration. The average accuracy rate of registration and fusion can reach to 87. 7%, which can satisfy the requirement of high precision registration and fusion.

In order to achieve skeleton identification function of body feeling controller, engineers must implement human body detection. We present a real-time detection method based on rgb-d data and the effect of its implementation on body detection. First, simplify the original 3D point cloud. Second, remove the ground plane. Third, for the rest of the point cloud data, the preliminary classification is performed to get human point cloud data cluster. Again after the first classification do the secondary fine classification, and then implement multiple human body detection. This method can detect multiple human body in static, and also can detect multiple human body in walking. Experimental results show that this method has good real-time performance,and CPU real-time processing speed can reach 25 frames per second. No matter to the static human body or the walking human body, the method is more accurate and rapid in detection, and the human body detection accuracy can reach 86%.

In order to improve the accuracy of network performance pre judgment, proposes a prediction model based on time series connected to two yuan rate, congestion rate sequence into the prediction process of docking pass rate. The calculation method of correlation using multivariate time series, fuzzy time series method and construction method of fuzzy rule matrix. The method overcomes the disadvantages of traditional network performance prediction method, considered only a single time series, without considering the drawbacks of other variables and forecast variables correlation of. Simulation results show that, the model can efficiently predict the network connection rate, provides the theory basis and the new scheme for comprehensive prediction of network performance.