In battlefield environment,it is difficult to determine the level of threat of multiple incoming targets.To solve the problem,a multi-target threat assessment method based on double-layer variable weight and TOPSIS-gray correlation is proposed.Firstly,Analytic Hierarchy Process (AHP) is used to calculate the subjective weights of target indexes,the improved entropy weight method is used to calculate the objective weights of target indexes,and the subjective and objective weights are then combined to obtain the first-layer weight.Then,a comprehensive equilibrium function is constructed,which integrates the variable weight theory with the first-layer weight to obtain the final weight.Finally,according to the decision-makers preference to the TOPSIS method or the gray correlation method,the multi-target threat assessment results are obtained.The simulation results show that: 1) The proposed method is more in line with the actual situation by combining the subjective and objective factors of the indexes with the impact of state value on the weights;2) The ranking results comprehensively consider the changes in the distance between the indexes and the changes of the index shape,and a comparison with other methods also proves the effectiveness of the proposed method.

In the UAV flight control system,when the UAV adopts a single height sensor,the measurement accuracy is low,and the traditional Kalman filter is prone to be divergent.To solve the problem,a method of fusing UAV altitude information of different sensors is proposed based on the improved Kalman filter.Firstly,the noise reduction algorithm based on ARIMA model is used to reduce the noise of the original measurement data of the three kinds of sensors.After the noise reduction,the height information of the sensors is fused for the first time by using the Kalman filter algorithm.Then,the fusion result is fused for the second time with the noise-reduced differential GPS data by using the method of recursively weighted least squares.Computational analysis shows that,compared with the traditional Kalman filter algorithm,the Root Mean Square Error (RMSE) of the height estimation is reduced by 39.6% and the maximum deviation is reduced by 31.7%.The simulation results show that the positioning accuracy of the obtained results in the vertical direction is effectively improved,and the preliminary ability to deal with abnormal conditions is guaranteed,which ensures the accuracy and reliability of the UAV flight control system.

In the re-entry phase of Reusable Launch Vehicles (RLVs),there are unknown disturbances, uncertainty and Partial Effectiveness Loss Fault (PELF) of control surfaces.To solve the problems,a robust fault-tolerant control system is designed based on Incremental Backstepping (IBS) and Tracking Differentiator Disturbance Compensator (TDDC).Firstly,IBS control laws are designed for the loops of attitude angles and angular rate respectively,and an error integral term is introduced to enhance the robustness of the system.Secondly,a TDDC based on Sigmoid function applicable to IBS control laws is constructed to estimate and compensate for the slow variable terms and high-order terms after Taylor expansion,which are omitted by the traditional IBS.Finally,the simulation results show that the control laws presented here have higher command tracking accuracy than the traditional IBS.

The ground-attacking UAV has become an indispensable part of military operations,whose autonomy level will determine the upper limit of UAV operations.In view of the small data size of some of the evaluation indexes,on the basis of the established evaluation index system,the Gray Relational Analysis (GRA) method is used to calculate the gray relational degree between the comparison sequence and the reference sequence,whose weighted relational degree is then obtained by using the judgment matrix of Analytic Hierarchy Process (AHP), so that the autonomous capability of the ground-attacking UAV is evaluated.Three types of ground-attacking UAVs, namely,MQ-1 Predator,MQ-1C Gray Eagle and MQ-9B Guardian,are used to verify the autonomous capability evaluation system through simulation.The results show that the method can distinguish the level of autonomous capability of different types of ground-attacking UAVs,which has a guiding significance for the application of the quantitative model of autonomous capability evaluation to the large/medium-sized UAVs on active service.

Based on the status of rotorcraft research at home and abroad,a kind of trirotor aircraft with a tiltable tail rotor is studied.Firstly,the mechanical structure of the trirotor aircraft is introduced,the forces and overall torque of the aircraft are analyzed,and the mathematical model of the aircraft is established by using Newton-Euler equation.Then,in view of the fact that the system is highly nonlinear with strong coupling,and is easily be affected by disturbances,the Active Disturbance Rejection Control(ADRC) is used to control the system.Finally,a flight control simulation and a disturbance rejection test are conducted based on Matlab/Simulink,and then the controller is compared with PID controller for further analysis.Simulation results show that,compared with PID controller,ADRC can realize a more rapid and stable adjustment to the attitude and height with stronger anti-disturbance ability.

In the field of airborne multi-sensor data fusion,the sensors errors in spatial measurement will have adverse effects on state estimation accuracy.To solve the problem,a novel error estimation and compensation algorithm is proposed.The time-invariant errors,i.e.the fixed part of measurement bias,can be quantitatively acquired by using the Minimum Mean Square Error (MMSE) based parameter estimation method.Furthermore,error registration is combined with target track estimation,which will effectively improve state estimation accuracy.Simulation results show that the improved algorithm can effectively suppress the effects of system errors.

Synthetic Aperture Radar (SAR) target classification is generally performed by feature extracting and classification decision-making.The Block-Matching and 3D filtering (BM3D) denoising algorithm is applied to SAR images to relieve noise corruption.Afterwards,the Extreme Learning Machine (ELM) is used to classify the denoised SAR images.ELM has high classification efficiency and precision.In addition, its sensitivity to noise corruption can be effectively relieved by Bi-dimensional Empirical Mode Decomposition (BEMD) denoising algorithm.Therefore,the overall classification performance can be enhanced by combining the strengths of BM3D with that of ELM.The proposed method is tested on the MSTAR dataset and the results have proved its validity and robustness.

The application of fixed-time control in nonlinear pure-feedback systems is studied.To solve the problem that the traditional backstepping method cannot deal with the pure-feedback system,an unconventional coordinate transformation is introduced.Based on Lyapunov stability theorem,it is proved that the proposed control algorithm ensures that the system tracks the given signal within fixed time,and the convergence time is independent with the initial state of the system.Finally,the effectiveness of the proposed algorithm is verified by two simulation examples.

In order to improve the matching accuracy of the original RANSAC algorithm,an improved RANSAC algorithm based on the strategy of twice matching is proposed.Of all the sample points,the feature points with excessive matching errors are eliminated at first,so as to reduce iteration times and accelerate the matching process.Then,the reverse matching method is used for second-time matching,so as to ensure the matching accuracy and reliability of the algorithm.The experimental results show that:1) The matching accuracy and speed of the original algorithm are improved;and 2) The improved algorithm has strong robustness and high reliability under the interference of illumination changes and motion blur.

In the background of sea clutter,it is difficult to detect small targets on sea surface,and the traditional Fractional Fourier Transform (FRFT) algorithm has unstable detection performance,large computation amount and low detection efficiency.To solve the problems,a new fast algorithm for detecting small targets in sea clutter is proposed.CEMD-RobustICA denoising algorithm is used to suppress sea clutter,so as to weaken its influence on the performance of Bi-orthogonal Fourier Transform (BFT) and Concise Fractional Fourier Transform (CFRFT).BFT-CFRFT algorithm is used to detect small targets, which transforms the two-dimensional search of the traditional algorithm into two one-dimensional searches,so that the amount of computation is greatly reduced.IPIX measurement data is used to verify the new algorithm,and the experimental results show that:1) The new algorithm can effectively suppress sea clutter,greatly reduce the amount of computation,and improve the accuracy and timeliness of target detection with the best performance in HH polarization conditions;2) The new algorithm has high detection efficiency and stable performance,which can be widely used in practical projects.

The follow-up system of a maritime weapon is nonlinear under the influence of unknown disturbances.To solve the problem,a nonlinear predictive control strategy based on Sliding Mode Disturbance Observer (SMDO) is proposed.The compound control algorithm uses SMDO to approximate the fluctuation of sea waves,load variation and system uncertainty,and then conducts feedforward compensation and model correction of predictive control based on the output of the observer,which keeps the rolling optimization of predictive control close to the actual system,so as to improve the performance of the whole control strategy.When designing the SMDO-based nonlinear predictive controller,the predictive control method for the nonlinear system based on a continuous time period is adopted and then combined with SMDO,so that the design of the composite control law for stability control of the firing line of maritime weapons is finally completed.The simulation results show that the control strategy can effectively improve the anti-jamming ability,hit accuracy and dynamically-stable aiming performance of maritime weapons.

The original YOLO V3 algorithm has low accuracy in small target detection and is prone to missed detection and false detection.To solve the problem,a ship target detection algorithm based on the improved YOLO V3 is proposed.Firstly,based on the structure of the original YOLO V3 network,an additional output scale is derived from the backbone network,whose feature information is spliced with that of the prior output scale,so as to obtain a feature vector with richer semantic information.Secondly,based on the data set,the clustering is improved,in which process the distance measurement formula is improved, and the number of anchor boxes and the corresponding parameters are reset.Finally,the loss function of the improved YOLO V3 is optimized,so as to improve the overall performance of the model.Analysis and experimental results on the test data set show that the average detection accuracy of the improved algorithm is 83.98%,which is 6.72% higher than that of the original YOLO V3.

When a UAV formation is performing transportation tasks,the external interference may make the movement of the UAVs be out of sync,destroy the formation and damage the transported goods.To solve this problem,a UAV formation control method is proposed,in which keeping the formation stable is its priority.By using the virtual-structure formation flight strategy,Proportional-Differential (PD) controller of the individual quadrotor is designed,to which the loop coupling synchronous controller of the whole formation is added.The validity of this method is verified by simulation.Compared with the method of estimating the virtual leaders state,this method ensures that the UAVs can return to the formation in shorter time with smaller damage on the formation after being disturbed.Moreover,this method can make the yaw angle respond synchronously.

To realize the attitude control of a quadrotor aircraft with uncertain model in the presence of unknown external interferences,a robust adaptive controller based on the super twisting algorithm is proposed.This method combines the super twisting algorithm with robust adaptive control.The former is used to suppress the chattering phenomenon of the system,and the latter can effectively compensate for the uncertainty of the model,so that the anti-interference performance of the system can be enhanced.A Lyapunov function is constructed,and the stability of the closed-loop system is proved.In order to verify the designed controller,a simulation is conducted and a quadrotor platform is built for flight experiments.The simulation results show that the designed controller has faster convergence rate and stronger robustness,and the flight experiment verifies the feasibility of the proposed controller,which can realize stable control of the attitude of a quadrotor aircraft.

For the DOA estimation of uniform linear arrays under narrowband signal model,an efficient and fast direction-finding method based on spatial spectrum estimation is proposed.The principle of DOA estimation is presented,the relationship between spatial sampling rate and the unambiguous DOA estimation range is studied,and the factors affecting DOA estimation accuracy are analyzed.Based on this method,a simulation of DOA estimation under a specific uniform array model is given.The results show that high DOA estimation accuracy can be achieved by using this method.

The traditional polarization imaging technology has large mechanical vibration and the system is unstable.To solve the problems,a dual-camera polarization imaging technology based on Stokes vector is proposed.Under the action of a liquid-crystal drive controller,two Liquid Crystal Variable Retarders (LCVR) are driven to perform flexible modulation,the cameras are used to collect the polarized image of the target,thus the component image of the Stokes vector is obtained,and the images of polarization degree and polarization angle are calculated.The method uses dual-camera simultaneous imaging technology and avoids the problem that the information collected by a single camera is incomplete.The method can realize full Stokes measurement and obtain linear and circular polarization images of the target at the same time.Experimental results show that the polarization degree (Dop) image contains more detailed information than the linear polarization degree (Dolp) image,and has an improvement of more than 5% in the image evaluation indexes,indicating that the Dop image can better restore the measured target.Using this method can improve the ability to identify static targets.

To remove the influence of disturbances on the tracking accuracy of the stable platform,a new composite Sliding Mode Control (SMC) algorithm is proposed.Firstly,robust H∞ mixed sensitivity control is used to design a Robust Disturbance Observer (RDOB) and effectively optimize its filter,so as to estimate and compensate for equivalent disturbances rapidly and accurately and improve the disturbance rejection ability of the system.Secondly,in order to improve the tracking accuracy of the system,the finite time theory is used to design a New Sliding Mode Controller (NSMC),and the convergence of the system in a finite time is proved by using the Lyapunov function.The simulation results have proved the effectiveness of the control strategy.

In order to satisfy the demand of NETD test for thermal imaging systems,four time components and three space components of noise of thermal imaging systems are discussed based on the 3D noise model, and the conventional noise computation method is analyzed.On which basis,a simple algorithm based on standard deviation is put forward,which can reduce the amount of computation and enhance test efficiency.The engineering test system is designed,and the calculation method and test process of NETD index are presented.Through experiment,two test algorithms for 3D noise are verified and analyzed.A comparison test with the imported equipment is conducted,and it is proved that the NETD test result of the proposed system is accurate and repeatable.It has been applied to the performance test of several kinds of thermal imaging equipment,which meets the requirement of engineering support.

There are input constraints and uncertainties in the control of coaxial multi-rotor UAVs.To solve the problems,a robust UAV control method based on the adaptive observer is proposed.Firstly,the mathematical models of position and attitude of a coaxial 12-rotor UAV with uncertainties are established.In the design of robust control law of the position loop,an auxiliary observer is introduced to compensate for the effect of input constraints.Then,in the design of robust control law of the attitude loop,an adaptive observer is introduced to suppress the effect of uncertainties.These methods help realize global asymptotic stability of the coaxial multi-rotor UAV.Simulation results show that the proposed robust control law has better stability, accuracy and rapidity,and can realize the robust control of coaxial multi-rotor UAVs within a shorter time with the maximum error in position and attitude tracking of only 0.05 m and 0.2° respectively.

As for the optimization design of phase-coded signal sets of MIMO radar,an improved Sine- Cosine Algorithm (SCA) is proposed.Through the introduction of Lévy flight,the algorithm expands the search range for the position of target points to prevent partial optimization.The algorithm sets the control parameters,improves the convergence rate,and balances the global search and local search ability.Based on the Particle Swarm Optimization (PSO) weight theory,the adaptive updating equation is changed to improve the convergence precision,so as to obtain the best result.Based on the improved SCA,a phase-coded signal set of MIMO radar is designed with good correlation performance,and the convergence rate and convergence precision are obviously improved.

When the signal of Global Positioning System (GPS) is weak or rejected,the UAV cant implement autonomous navigation.To solve the problem,an improved ORB-LK optical flow algorithm is proposed to estimate the velocity of UAVs.Firstly,the feature points on the object are extracted by using the ORB algorithm,and the optical flow velocity of the image is calculated by using the LK optical flow method.Then,the forward/backward matching algorithm is used to determine matched pairs of the points,and the PROSAC algorithm is used to eliminate the mismatched pairs,so as to realize fine matching.Finally,the UAV rotation velocity measured by Inertial Measurement Unit (IMU) and the height measured by the sonar are used to calculate the translation velocity,so as to provide velocity information for the autonomous navigation of UAVs.Simulation experiments with Matlab show that the average operation time of the improved algorithm in the public image data set is 0.622 s,which meets the requirements of real-time application of the UAV.Compared with the original ORB-LK algorithm,the improved algorithm greatly improves the matching accuracy,and can estimate the velocity accurately.

The gyro-stabilized platform of the optoelectronic pod can only compensate for the attitude change of an aircraft in inertial stabilization state,but cannot compensate for the linear motion of the aircraft.In the course of flight,the Line of Sight (LOS) of the optoelectronic pod moves fast along with the linear motion of the aircraft.This brings difficulty for the operator to capture and aim the target promptly.If the operator wants to keep an immovable region under surveillance for a long time,he has to operate the pod constantly to keep the LOS pointing at the target.This will increase the burden and operation complexity of the operator.In this paper,a dynamic tracking algorithm based on INS information is introduced.The velocity of linear motion of the aircraft is projected onto the coordinate system of the platform.The dynamically-compensated angular velocity of the platform is calculated and is then compensated for in real-time,which drives the pod to move and keeps the LOS pointing at the immovable ground scene.The experimental results indicate that the dynamic compensation technology based on INS information of the aircraft can compensate for the linear motion of the aircraft in real time and keep the imaging scene of the optoelectronic pod steady,which is convenient for the operators manipulation and can alleviate the burden of the operator.

Flatness is an important inspection index in quality control of the production of industrial work pieces.The most widely-used Coordinate Measuring Machine (CMM) technology has a small number of sampling points and low acquisition efficiency.To solve the problem,a non-contact and simple flatness measurement system based on fringe projection technology is proposed.The device projects fringe patterns onto the surface of the work piece under test,and by calculating the phase of modulated fringe images and calibrating the phase/height mapping relationship,the sample points can be efficiently acquired on the full field of the measured plane.The least squares method is used to calculate the ideal plane,and the deviation from this plane is used to calculate the flatness error.Several experiments are conducted,and the results demonstrate that the proposed system can be applied to the flatness measurement of general industrial work pieces with few measurement errors and high repeatability.

As to the qualitative requirement assessment of maintainability of photoelectric reconnaissance equipment,a method for quantitative evaluation of qualitative requirements of maintainability is proposed based on the integration of maintainability design with maintenance events according to the idea of designing an integrated test.Analytic Hierarchy Process (AHP) is used to progressively quantify each qualitative index,and a method is presented to objectively test and assess the qualitative requirements of maintainability, which provides a reference for the testing and assessing of qualitative requirements of maintainability of photoelectric reconnaissance equipment under the new equipment identification system.