Considering that the ground test equipment for space cameras has the characteristics of large amount of data and rapid storing velocity, the design of a high-speed large capacity storage system is implemented by using FPGA as the control core and the solid NAND Flash with excellent performance as a storage medium. To improve the storing velocity of the system, the Flash array is controlled by using two-plane operation, parallel technology and pipeline technology. The system uses bad block management after analyzing bad block feature of Flash. Experimental results show that the system can accomplish the record of a large number of high-speed original data from space camera, which ensures the real-time performance and reliability of data recording.

Considering the bounded uncertainties without the prior knowledge of the bounds in heavyweight airdrop operations, an adaptive sliding mode control law with finite-time convergent performance is proposed.This control law is composed of two parts of nominal control law and compensated control law, the former is used to achieve finite-time stabilization performance, and the latter is used to reject the uncertainties and improve the robustness.The switching gains of the compensated control law are compensated using adaptive algorithms, and the knowledge of the bounds of the uncertainties is not required to be known in advance.Meanwhile, the severe chattering of the sliding mode control that caused by high switching gains is effectively reduced.Theoretical analysis indicates that the proposed method can ensure finite-time convergent performance of the tracking error of air speed and pitch angle.Simulation results prove the effectiveness and superiority of the control method.

The Simultaneous Localization and Mapping (SLAM) method based on monocular vision is one of the most significant research directions for the robot autonomous walking, and the image feature matching technology is the key technology of this method.The SLAM matching method based on Scale Invariant Feature Transform (SIFT) algorithm has the advantage of abundant and stable feature points, but it also has some disadvantages in speed and accuracy.Therefore, considering that the SIFT descriptor has high dimensions and long matching time, an improved SIFT algorithm is proposed.The original 128-dimensional feature descriptor is reduced to a 24-dimensional feature descriptor with a rectangular inner side and round outer side.In the matching process, trilinear interpolation and RANSAC algorithm, etc.are applied to eliminate the mismatching of matching results.The experimental results finally show that the improved SIFT algorithm not only has good robustness to the changes of angle and illumination conditions and improves the matching speed and accuracy, but also meets the needs of SLAM synchronous map construction.

Since the single-sensor based head pose measuring system of the Helmet-Mounted Display (HMD) suffers from low accuracy and poor precision, a novel algorithm based on the information fusion of MARG sensor is proposed.First, the complementary filtering is combined with PI algorithm to estimate the gyroscope drift error, and then the pose data fusion is implemented by using the Extended Kalman Filter (EKF).We made experiments to compare the performance of the conventional single EKF and the proposed algorithm in filtering by using the measuring system composed of a CC1310, a gyroscope, an accelerometer and a magnetometer.The results demonstrate that the proposed algorithm can restrain the drift of the gyroscope and the noise of the accelerometer, and consequently improve the stability and precision of the head pose algorithm.

In the AC servo system of a rocket launcher, it is difficult for the traditional PID control method to get good performance due to the complex nonlinearity of the moment of inertia, friction, variable loads and internal and external disturbances under different working conditions.To solve the problem, we established the mathematical model of the AC servo system of rocket launcher according to its composition, and designed a Fractional Order PID (FOPID) controller based on neural network Active Disturbance Rejection Control (ADRC).In order to reduce the parameters'calculation cost of the fractional order controller and improve its dynamic performance, a control algorithm of Radical Basis Function (RBF) neural network is adopted for on-line self-tuning of the two parameters of FOPID, integral order and differential order.The simulation test results show that the control strategy can effectively suppress the position disturbance with fast response speed and no overshoot.

In order to realize the anti-sea combat capability requirement generation of the new types of naval guns, the framework model is established, the anti-sea combat mission is decomposed and the capability index system of the naval gun is constructed.The mapping models are established, including the detection probability model, the maximum detection distance and tracking distance model, the single-shot hitting probability model, the reaction time model and so on.Taking the system accuracy when shooting a target at 25 km away with the given single-shot hitting probability as the standard, the simulation experiment analyzes the capability requirements to the new naval gun at different firing distances under different initial velocities. The model method can quantitatively generate the capability indicator requirements that satisfy the operational mission, and has theoretical and practical significance for scientific development of naval gun weapon equipment.

Considering the problems of “difficult acquisition of test data and diverse failure mechanism” of the radar seeker of a missile, we proposed a competition failure state prediction model related to degraded failure and sudden failure based on the analysis to the characteristics of the degraded data and failure data.Aiming at the characteristics of small sample, nonlinearity, and uncertainty of the degraded data, we used Relevance Vector Machine (RVM) to regress and predict the distributed parameter sequence.Then the kernel parameter and embedded dimensions were both optimized via Quantum-Behaved Particle Swarm Optimization(QPSO) and Hannan-Quinn(H-Q) criterion.Taking into account the correlation between the sudden failure and the degraded failure, a position-scale model was introduced to describe the correlation between the sudden failure distribution parameters and the degraded data, thus to predict the failure state of the next stage.Prediction result of an example verified the feasibility and rationality of the model.

In distributed multi-sensor fusion system, the uncertainty of target tracking quality due to different detection performance of the sensors has not been fully considered by the traditional track fusion algorithms, which may result in degraded track quality after fusion.A method of track fusion algorithm based on track quality entropy is proposed here.Firstly, the quality entropy models of all the local tracks of the fusion center are constructed, sorted according to value of the entropy, and divided into different uncertainty levels.Secondly, a sort queue of track quality entropy is choosed as the reference queue, and the gray theory is used to cluster the tracks of different local nodes from the same target source.Lastly, the clustered tracks are grouped according to quality uncertainty level, and the track state with low uncertainty level will be fed back to local nodes for local fusion.This method can improve the quality of local node tracking, and enhance the correctness of track fusion.Simulation verifies the feasibility and effectiveness of the method.

Cognitive electronic warfare requires smarter algorithms to adapt to the increasingly complex electromagnetic environment.As to the frequency prediction for frequency-agile radar with variational parameters, a frequency prediction method based on differential evolution algorithm is proposed according to the real data and the data analysis after eliminating the outliers.The proposed method can not only improve the prediction accuracy and reduce the demand for data consistency, but also be adaptable to the variation of frequency of the tuning parameters through online analysis, thus it has higher application value and robustness.

The satellite navigation signals are susceptible to interferences under a complex environment, which may result in the failure of UAV satellite positioning.An autonomous positioning technology for micro-UAVs based on multi-point constraint is proposed.Through the on-board photoelectric measurement platform, multiple ground targets are observed.By using the geometric relationship that the ground target point, image point and the camera optical center are collinear, the autonomous positioning mathematical model of the UAV is established.The UAVs own spatial position is obtained by using the least square estimation.The feasibility and robustness of the method are verified by experiments.The experimental results show that the method can obtain the position information of the UAV within the allowable error range.

Aimed at the contradiction between the accuracy and speed of the traditional air-to-air missile attack area in calculation process, a boundary search algorithm based on adaptive step-size golden section method is proposed.The three degree-of-freedom ballistic models are established for the missile and the target.Then, the miss distance is used as the adaptive variable to improve the size of the traditional golden section method.Simulation results show that the algorithm can improve the attack area calculation accuracy to a large extent under the premise of increasing calculation speed.The method can meet the requirements of airborne fire control system on real-time performance and accuracy for the attack area calculation.

The traditional correlation filter tracking algorithms have low precision when the target is in fast motion, occluded, or under complex background. To solve the problem, a correlation filter target tracking algorithm with adaptive context sensing was proposed. Based on the framework of the correlation filter algorithm, the boundary effects brought by the cyclic shifts and the fixed learning rate were mainly improved. Firstly, an adaptive sampling strategy based on response map was used to sample the context information in the classifier training stage. Then, a segmented learning rate adjustment strategy was used to make the algorithm more adaptive to the changes of target. Finally, the performance of the proposed algorithm was verified on the standard data set. The experimental results showed that the proposed algorithm can improve the tracking precision of DCF and SAMF algorithms. It not only has better robustness in the case of fast motion, occlusion, and complex background, but also can be used as a framework integrated into most of the correlation filter trackers.

A method for multi-sensor track association based on interval number and DS evidence theory is presented.Through analyzing the influence of error on parameter measurement, the interval numbers of target parameter true value detected by the sensors are established.The relative support level between the interval numbers is calculated, based on which the support level of track association about target parameter is obtained.Then, the support levels about track association of different parameters are combined based on DS evidence theory, and thus the problem of track association is solved.Examples show that this method is effective for multi-sensor track association.

A state amplification based method is proposed for real-time filtering of MEMS random error, which is used to overcome the disadvantage of civil MEMS gyroscopes of low precision, and to solve the problem of traditional MEMS that random error time series modeling needs zero-mean processing and cannot be modeled online.In the method, the mean of time series is considered as an unknown parameter, and an intercept term is added to the traditional ARMA model.A classical Kalman filter and an adaptive Kalman filter are designed by using the state amplification method based on this model, in which the filter has no zero-mean requirement for the measured data.Static and rocking experiments show that this method can significantly improve the accuracy of MEMS gyroscopes.

Aiming at the typical avionics Fibre Channel (FC) switched network, considering the difference between the impact of the overall failure and the independent port failure on the network reliability, and also considering the demand of analysis to different redundancies, we built failure models of the node, the interchanger, the link, the port to port network, and of the node to node network, thus to provide an intuitionistic modeling idea based on matrix.A failure sample generating method and a network reliability calculating method based on simulation were provided, and a quantitative analysis was made to the avionics FC switched network reliability by use of a number of cases.The models and methods above can provide a theoretical support for reliability design of avionics FC switched network.

In the ARINC661 specification,there is a large amount of data interaction between UA and CDS with short transmission period, and it is difficult to comprehensively and meticulously analyze the instructions in real-time monitoring process.To solve the problems, we studied the offline detection technology of the cockpit display system, and designed such methods as ring data batch storage,simulation UA, and anomaly positioning, which can reproduce the instruction interaction process in an offline environment.The test results showed that the design is reasonable,reliable, and can detect interactive commands in an offline environment.

Aiming at the problem of how to eliminate the interference of the data collected by the six-component balance in the wind tunnel test, a processing method combining EMD algorithm with wavelet transform is proposed.Considering the interference components that may be contained in the data from the perspective of the design of the balance, the time-frequency relationship between the interference components and the data is given, and the rationality of the method is theoretically explained.The data is restored by simulation test, and the error is compared with that of the traditional interference elimination algorithm.It is verified that the method has higher accuracy.Finally, the real six-component balance data is processed to further prove the practical feasibility of the method.

Considering the fault characteristics of UAV sensors of multiple and complex fault types, we modeled and predicted the time delay by using grey model and Elman neural network, and used the principle of minimum variance to determine the best combination of model weighting coefficients.The method was applied to the fault diagnosis of UAV sensors.The simulation result showed that the combined prediction model has a high precision in predicting the fault diagnosis delay, and proved the validity of the proposed diagnostic method.

The large-size and high-precision target tracking turntable is mainly used for simulating the target motion characteristics of seeker, radar and other detection equipment, and testing the following performance of the detection equipment.It is an important facility for semi-physical simulation in the field of aeronautics and astronautics.In this paper, the working principle of the turntable and the hardware composition of the control system are introduced.Then, the mathematical model of the turntable is established by combining the mathematical model of the permanent magnet DC torque motor with the dynamic equation of the turntable.On this basis, a three-loop control system based on the PID controller is proposed.Simulation experiments show that the system meets the specifications within the technical requirement range.

Comprehensive analysis of the various attributes of the targets and the comprehensive sorting are the core issues in the composite team's decision-making. Aiming at the shortcomings of traditional target ranking methods, this paper proposes a fuzzy comprehensive sorting method that combines the Improved Analytic Hierarchy Process (IAHP) with the Entropy Weight Method (EWM).On the basis of making full use of the previous evaluation indicators, the psychological value of the target is included in the evaluation index system, which improves the comprehensiveness of the evaluation.In the process of calculating the weight of the evaluation index, the IAHP-EWM is used, and the subjective and objective characteristics of expert scoring are taken into consideration, thus the rationality of weight calculation is improved.Based on the evaluation index and the weights, a multi-level fuzzy evaluation algorithm is proposed for target sorting. Eight important goals in the scenario are sorted by using the method.The results show that the ranking result is in line with the commander's cognition, which can improve the rationality and effectiveness of decision-making.

Gear pair gap is an important parameter in gear mechanical design and gear meshing.In order to improve the measurement precision of the gear pair gap and solve the problem that it is impossible to measure the gap of the gear pair which is equipped with a circular grating photoelectric code disk on both gears as an angle sensor due to the limitation of structure, an automatic measuring technique based on the photoelectric angle measurement module is proposed for measuring the gap of the gear pair.To collect the angles of the two gears, a micro-photoelectric angle measurement module is installed on the driving gear and a photoelectric encoder is mounted on the driven gear, and then data processing and software control are performed to realize the automatic measurement of the gear pair gap.The main error sources affecting the automatic detection system are given, and an error analysis is carried out.Finally, the automatic measurement results of the gap of bevel gear pair are given through experiments.

In the conventional Synthetic Aperture Radar and Ground Moving Target Indication (SAR/GMTI) system, the excessively high azimuth velocity may cause velocity ambiguity.To solve the problem of radial velocity ambiguity of moving target, a kind of multi-channel interferometric of SAR/GMTI scheme is presented.By setting up four receiving channels with non-uniform intervals and designing each channel spacing reasonably, the method carries out interferometric cancellation respectively on the sub-images of two echo signals.Then, two velocity sets are obtained, and the intersection of the two sets is the correct estimation of the target velocity.Therefore, the problem of velocity ambiguity is solved and the radial velocity of the target is estimated correctly.The validity of the method has been demonstrated by analysis and computer simulation.