In SINS, the inertial components are directly mounted on the carrier.The error can be divided into deterministic error and random drift error (dynamic error), in which, the formercan be compensated.In this case, the initial alignment of the pure static base can achieve a high accuracy.In the practical application, the dynamic error is directly reflected in the inertial device because of influence of the external environment (wind, vibration and disturbance, etc.).At this time, the error model is not linear.Under this situation, unscented Kalman filter (UKF) and adaptive unscented Kalman filter (AUKF) were designed respectively.We introduced the principle of adaptive estimation into the original UKF algorithm, adjusted the contribution of the kinetic model to navigation solution.AUKF algorithm can automatically balance the weight ratio of state and observation information in filtering to adjust the covariance of state vector and observation vector in real-time, thereby to improve the system performance.Experimental results showed that the use of adaptive UKF algorithm in comparion with the normal UKF algorithm can obtain better accuracy and reliability of self alignment.

To the nonlinear model errors caused by the swaying base, unscented Kalman filters and Adaptive Unscented Kalman Filter (AUKF) are designed (UKF) for fine alignment respectively. In this paper, the adaptive estimation principle is introduced into the UKF algorithm. AUKF algorithm can balance automatically the right of the state information and observation information in the filtering result, so that to real-time adjust the covariance of the state vector and observation vector. The experimental results show: Compared with normal UKF algorithm, the adaptive UKF algorithm can eliminate the appearance of abnormal error and improve the accuracy and reliability of self alignment in the swaying base of SINS. In the following three direction east, north and day, the angle accuracy of misalignment is increased by 0.2', 0.2' and 5.0'; convergence time is shortened by 10 s, 28s and 27s respectively.

To real-time problem of nonlinear filtering in the initial alignment of large azimuth misalignment angle, unscented Kalman filter (UKF) and adaptive unscented Kalman filter (AUKF) are designed for fine alignment respectively. Based on the analysis of Sigma point sampling strategy in KFU filter, we introduced the principles of adaptive estimation into nonlinear initial alignment process of large azimuth misalignment. In the case of large initial value error and dynamic model disturbance, the contribution of the contribution of the kinetic model to navigation solution can be adjusted. Simulation results showed that AUKF algorithm compared with UKF's can obtain better alignment accuracy and rapidity.

H∞ filtering is a representative method of robust control. In the SINS/GPS integrated navigation system, to solve the limitation of Kalman in the system model and noise, this paper puts forward an application of H∞ filtering algorithm, which has strong robust performance in integrated navigation system. The filter equation of H∞ and Kalman algorithm is given. The integrated navigation (SINS/GPS) uses the output difference between SINS and GPS as input value of the filter, and then the error of integrated navigation system is estimated and corrected by one filtering method in real time. The accuracy and robustness are analyzed and compared between the two kinds of filtering algorithm. The simulation result shows that the H∞ filtering has better stability and robustness in colored noise. Through this research, H∞ filtering algorithm can well solve the uncertainty of the noise model and statistical characteristics. H∞ filtering algorithm is more suitable for the application of SINS/GPS to integrated navigation system.

Motion estimation plays a very important role in video coding because it can eliminate the temporal redundancy effectively to achieve good coding efficiency. However, motion estimation suffers from high computational complexity, which makes it difficult to meet the demands of real-time encoding. A fast matching algorithm, which uses histogram ordering model to predict matching error, is presented to reduce the complexity of motion estimation. In this algorithm, the correlation of motion vector is firstly utilized to predict initial search center. Then, block is divided into 1×4 sub-blocks, the edge strength of sub-blocks, which is used to determine the calculation order of sub-blocks, is analyzed by pixel histogram of current block. Finally, the accumulated partial distortion of sub-blocks is used to adaptively predict the total matching error of current candidate block according to the edge strength of sub-blocks, which can make an early decision for an impossible candidate block before complete distortion computation and effectively reduce the calculation complexity of matching criterion. Experimental results indicate that, taking no account of overhead computation, the computational costs of proposed algorithm are averagely reduced by 92.62%than that of PDS algorithm while the image quality degradation is quite small, which could be ignored.

When the system model and noise statistical characteristics are known, the conventional Kalman filtering algorithm is suitable. In most cases, the noise statistics are unknown. To improve the alignment precision and convergence speed of strap-down inertial navigation system, an initial alignment method based on Sage-Husa adaptive filter is proposed. Automatic on-line estimation and correction for the noise parameters, the state of the system and the state estimate covariance by the observed data. Using forgetting factor can limit memory length of the filter, which could enhance the effect the newly observed data acts on the present estimation. Thus, enable the system to achieve the best filtering effect. Through simulation verifiable, the adaptive Kalman filter algorithm, improve the convergence speed and alignment accuracy effectively.

Optic-electro carrier landing measuring equipment is applied on the accurate sliding-route and attitude observation for the landing of carrier aircraft,and the Line of Sight(LOS)stabilization system is helpful to decrease the error of tracking and measuring when the equipment works in the waved situation.This paper presents a design of LOS stabilization system based on the deck-strapped Inertia Navigation System(INS).The structure of the system is firstly introduced.And then modeling analysis of the control system is realized.Finally,the simulation and experimental result is given that the design has good effect on resisting the disturbance in inertia system.

To meet the increasing requirement for small angle measurement in the field of military and industry, the authors had designed the 2-D optoelectronic autocollimator based on area CMOS image sensor to pursuit higher precision and stability. For hardware circuit, the authors chose the DSP+FPGA structure to build a reliable and high-speed platform. For software algorithms, frame subtraction, Gaussian and Median filtering were effective to reduce the noise of original images. After threshold segmentation, weighted centroid locating algorithm was selected to calculate the sub-pixel location for laser spot. Finally, angle conversion was performed to get the 2-D misalignment angles. Experimental results show that, with the help of the methods of 2-D image processing, the autocollimator can obtain high precision within 0.2" and the capability of anti-interference for misalignment angle measuring.

利用MT9M413C36STM、TMS320VC33PGE研制了一种高帧频CMOS图像传感器成像、显示及数据处理系统.根据所用器件的特点,文章对图像传感器成像、显示及数据处理原理和时序进行了分析.给出了系统相关的硬件电路,介绍了设计重点.在QuartusⅡ8.0及CC4.1开发环境下,使用VHDL、AHDL、C语言进行了驱动程序编写和调试.结果表明,该系统在1 280×1 024@60 Hz逐行扫描模式下可以稳定地工作,已在重点课题中得到了批量应用.

With TMS320VC33 , EPF10K50 and CY7C1061BV33 , the hardware based on SRAM“Ping ,pong” frame-storage structure is designed .The working sequence and design key-point are analyzed .Under Quartus 8 and CC4 .1 ,VHDL ,AHDL ,C language driver and image process is programmed and debugged .The results show that the system can work stably at 1 280 × 1 024@60 Hz mode .T he system has been applied into industry .

In order to achieve real-time imaging and simultaneous display,a lary array CMOS camera system with simple structure and portable size is designed.Making use of EPF10K50,image timing of MT9M413C36STM,display timing of SXGA and multiplex of output data are designed.Image sensor data is divided into two way signals.One way data outputs for processor electric circuit to deal with,the other way data through ADV7127 video AD transformation outputs to display simultaneity in SXGA.The related hardware electric circuit is designed,and the device driver is written using the VHDL language.Under the development environment of the Quartus 8.0,it is debugged.Experimental results show that this system can stably work in 1 280×1 024@60 Hz line-by-line scan.

When the self-collimation angle measuring equipment is working in outside field, it will be affected by a variety of factors such as airflow, the stray light, the temperature and so on. The display value will be unstable and the system will get low measurement accuracy. This problem is on one hand due to the increasing noise which participates in imaging of linear CCD. On the other hand, the sampling images become jitter both in location and amplitude. According to these problems, we can reduce these interferences by using hardware filtering and software method combined with adaptive integration time, image superposition and weighted barycenter algorithm in grayscale. The software for the system is realized based on the high-performance DSP platform. Experiment results show that, after the realization of these methods, the angle measuring equipment has the advantages of high-speed, stability, getting better repeatability and the ability of anti-interference.

In order to solve the problem of high non-uniformity of LED display images which is caused by the edges of LED display panel during module splicing, transitional compensation algorithm is proposed by improving the existed correction technique named the correction technique based on CCD. First, introduce the three development stages of the LED display panel. Then, the realization progress of the transitional compensation algorithm is described in detail after elaborating the principle of transitional compensation. Finally, the algorithm is emplaned in a LED video control system to control the LED display panel whose display area is 1280×960, and which is spliced by 30×20 LED modules whose size are 64×32. Experimental results show that this algorithm is able to reduce non-uniformity of LED display images from 29.3% before correcting to 0.95%.

We compare the fourth-order Runge-Kutta algorithm with the three-sample algorithm for improving the SINS accuracy under noise disturbances. With classical coning motion, the gyro output carrying Gaussian white noise is analyzed by means of simulation and the results show that anti-jamming effects of the fourth-order Runge-Kutta algorithm is more effective .

In order to improve the initial alignment accuracy and convergence rate of the SINS system, proposed the improved UKF algorithm (AUKF) based on the Unscented Kalman Filter (UKF). Noise statistical characteristics are mostly unknown in real systems, when it was effected by the initial value errors and dynamic model errors, AUKF algorithm can real-time adjust the covariance of the state vector and observation vector, and balance the right ratio of the state information and observation information in the filter results, thereby improving the system performance. The experimental results show: The Improved UKF Algorithm enhances the convergence speed and alignment accuracy effectively.

The sum of squared differences(SSD) is used as the distortion metric for inter mode decision in traditional audio video coding standard(AVS) rate distortion optimization(RDO),but it is not consistent with human vision system(HVS) quite well after a lot of investigations.Compared with the other methods,the structural similarity(SSIM) proposed recently for assessing image quality accords with HVS much more well.Therefore,in order to improve coding efficiency,the SSIM is introduced into Lagrangian cost function to correct the expression of distortion metric.Based on experiments,the empirical formula of Lagrangian parameter is established in this paper.Experimental results indicate that the bit rate using the proposed algorithm is averagely reduced by 13.22% than that using the mode selection algorithm of AVS.Especially,when the sequences have massive stationary blocks,the bite rate is saved more than 30% with QP=10.At the same time,the quality of the reconstructed image is only decreased by 0.14% which could be ignored.

In order to satisfy the requirement of size,power and cost of the micro inertial navigation system(MINS),an integrated navigation system based on ADIS16364 micro inertial measurement unit(MIMU) of AD Corporation and dual-core processor OMAP3530 of TI Corporation was designed.Initial alignment could not be done by MIMU itself because of its low precision,so a method of using electrical compass in initial alignment is presented.The traditional linear filtering method of Kalman filter could not be used in the system because of the large initial alignment error,thus the nonlinear filtering method of unscented Kalman filter(UKF) was applied in data fusion of the integrated navigation system.Experiment results show that the scheme is feasible and meets the system requirement.

Focusing on the contradiction of the real-time image processing and real-time tracking in the real-time video signal processing system,a new real-time image processing system based on DSP and FPGA which can improve the system performance was introduced.DSP was used as the main processor to fast and effectively handle the complex recognition tracking algorithm because of its high-speed operation ability.FPGA was used as coprocessor to complete video image receiving,storing,preprocessing,etc.to make the design more flexibly.The centroid tracking and relevant track algorithm were adopted in the system.Experimental results showed that the system can track the moving targets in real-time quickly and stably.

In order to overcome the shortcomings of standard unscented Kalman filter (UKF), which are obviously influenced by the error of initial value and the model error of system, adaptive UKF which is based on the adaptive principle is applied in initial alignment of the MINS/GPS integrated navigation system. Coarse alignment cannot be done by micro inertial measurement unit (MIMU) itself because of its low precision, the method of using magnetometer to assist it with coarse alignment is presented. Simulation results show that the adaptive UKF can overcome the influences of initial values error and inaccurate system model, and improve the convergence speed and alignment accuracy effectively.

In order to improve the transmission velocity in mulitipath fading wireless channel, the high speed OFDM technology receives increasing attentions in mobile communication. Modulation programs are designed with VHDL based on the principle of OFDM in the paper. First, after OFDM fundamental is introduced; two main advantages are obtained via discussing spectrum utilization ratio of OFDM. Then, from the results of VHDL simulation, the realization of baseband operations, such as interleaver, subcarrier modulation, IFFT and adding CP, are presented. Finally, implemented programs are validated on the actual implement system. Experimental results indicate that setup time corresponding to transmission velocity is only 71.05µs and steady time is approximately 6 times as setup time, that is, not only achieving the high speed transmission, but also supplying adequate modulation time.