This paper proposes a UAV cooperative path planning method based on improved two-way rapid expansion random tree (RRT-Connect) algorithm.Firstlyin view of the low efficiency of the basic RRT-Connect algorithm and the limitation of random samplingan improved RRT-Connect algorithm is proposed.By introducing the target gravity and controlling the sampling rangethe invalid samplings are reduced and the convergence speed is improved.The using of chaotic sequences ensures the sampling nodes have a certain randomness.Based on the improved algorithmthe constraints of the path planning model are considered and the artificial potential field is introduced to propose a collaborative path planning method.Simulation experiments verify that the improved algorithm is superior to the original algorithmand the effectiveness of the collaborative path planning method is proved.

Aiming at the requirement of both high detection accuracy and high detection speed in the object detection fieldBalanced YOLO (B-YOLO)a one-stage object detection algorithm is proposed.Firsta spatial attention mechanism is introduced in the algorithmand the multi-scale maximum pooling layer is used to increase the range of the receptive field.Thenthe cross-stage local connection structure and the pass-through layer are adopted to optimize the backbone network structure and improve the computational efficiency.Finallya bottom-to-top path is added to the multi-scale detection structureand splicing is used for transverse connection to effectively fuse the deep semantic information with the shallow location information. Experimental results show that the algorithm has achieved a good balance between accuracy and speed.

The multi-objective optimization method is adopted to study the modeling of Weapon-Target Assignment (WTA) in ground target attacking.Firstlybased on the analysis to the characteristics of the ground target attackinga multi-objective optimization mathematical model is established.The optimization objectives are defined as the minimum risk of aircraft in penetration and the maximum target damage efficiencyand a quantitative method for evaluating the threat of air defense fire power to the aircraft is proposed based on grid map.Thenthe Non-dominated Sorted Genetic Algorithm-Ⅱ(NSGA-Ⅱ) is adopted to solve the problemand the components of the algorithm are designed.Finallya simulation example is built for experimental analysis.The simulation results show that the model and method in this paper can effectively solve WTA problem of ground target attackingwhich provides a reference for engineering implementation.

In order to solve the problem of mobile robot path planninga global dynamic path planning method is designed combining improved A* algorithm with Dynamic Window Approach(DWA).Firstbased on the traditional A* algorithm and the JPS algorithmthe child nodes are extended and jumped to improve the efficiency of path planning.Secondlythe planned path is smoothed and optimized by using Floyd algorithm.FinallyDWA is integrated so that the A* algorithm can perform global dynamic path planning.Based on the Matlab platformsimulation experiments of the global dynamic path planning algorithm are conducted for 8 kinds of grid maps.The analysis result shows that the fusion algorithm is greatly improved in efficiency and smoothnessand can perform dynamic obstacle avoidance.The global dynamic path planning algorithm after fusion and improvement has excellent path planning capabilities.

To the consistency problem of the heterogeneous multi-agent system with first-order linear and quadratic non-linear terms under the condition of undirected topologythe sufficient condition for the heterogeneous system to achieve consistency was obtained by modifying an existing distributed control protocoland by using graph theoryLyapunov stability theory and the theory of Lasalle invariant set.Simulink was used to build a heterogeneous multi-agent platform with nonlinear linkswhich was composed of quad-rotor UAVs and small intelligent vehicles.The theoretical results of this paper were verified on this platform.The final numerical simulation results showed that the heterogeneous multi-agent system could achieve consistency under this protocol if the conditions in the paper were met.

Aiming at the radars classification and recognition of the interference signal under complicated electromagnetic environmentwe studied the Choi-Williams Distribution (CWD) time-frequency images of the RF noise interferencenoise-amplitude-modulation interferencenoise-frequency-modulation interferencerange gate pull off interference at a constant speedand speed gate pull off interference.The AlexNet convolution of deep learning neural network model was adopted to automatically extract the image features of detailsso as to realize the classification and recognition of radar intereference signal.The simulation results showed that:1) The recognition rate of the network increases rapidly with the increase of the interference-to-noise ratio in the range of -10 dB to 0 dBand the recognition rate is basically close to 100% when the interference-to-noise ratio is above 0 dB;and 2) In the full range of interference-to-noise ratiothe networks recognition accuracy rate is 99.25%and the recognition effect is good.

As the one of the key technologies and an important part in the design and operation of the new generation aircraftPrognostics and Health Management (PHM) technology has been widely concerned by numerous experts and scholars at home and abroad.In view of the unmanned and collaborative features of the future aviation equipmentan analysis is made to the operation scene of the PHM system for manned and unmanned aerial vehicle cooperative fleet by using the system engineering method.The system requirements are studied by modeling the system case diagram and system activity diagram.Thenthe top-level architecture of PHM for manned and unmanned aerial vehicle cooperative fleet is proposedwhich contains functional architecture designlogic architecture design and data transmission service.Some key technologies involved in the system design are also discussed.The research has certain significance for the design of the PHM system for manned and unmanned aerial vehicle cooperative fleet.

To solve the problem that the forward direction of radar does not have enough Doppler bandwidth thus it is difficult to use the Synthetic Aperture Radar (SAR) imaging technologydeep researches have been carried out on the radar forward imaging technology from multiple aspectsand have been applied in missile precision striketerrain recognitiontarget positioning and aircraft autonomous landing.This paper comprehensively introduces the history of the researches on forward-looking imaging technologyand make classification to the technology systematically according to imaging principleradar model and detection signals.The basic principles and characteristics of the imaging technology based on convolution inversion single pulse imaging technologymicrowave correlation technologyarray radar imaging and bistatic forward-looking imaging are mainly described.The research status and development trend of various forward-looking imaging technologies are summarized.

The design of the air conditioning system of the three-axis air bearing test bed directly affects the accuracy of the control system and the air-conditioning consumption.In this paper the thrust allocation problem in the air conditioning control of the three-axis air bearing test bed is studied.Aiming at the redundant thrust allocation problem two methods are designed to optimize the thrust allocation of the three-axis air bearing test bed.The corresponding theoretical analysis and experimental research are carried out from the perspectives of expense and accuracy.The rationality and effectiveness of the proposed method are verified by mathematical simulation and experiment.The research results meet the requirements of finding a good cold air thrust allocation scheme quickly after the expected torque is given the steps of thrust allocation are simplified and the problem of waste of cold air caused by the redundant allocation of the thruster is effectively solved and the control accuracy of air conditioning is within the design range.The experimental results show that the accuracy of air-conditioning control can reach 0.4° and the task of attitude control with only four thrusters at a time is realized.

Avionics system is an important part of modern fighterand it is an important support means to achieve mission integrationcapability integration and resource integration.The performance of combat aircraft is closely related to the effectiveness of avionics system.Aiming at the problems of complex evaluation index system of avionics system and great divergence of expert evaluationsthis paper designs an effectiveness evaluation method for avionics system based on wavelet neural network methodand presents an avionics effectiveness evaluation system.Based on the Analytic Hierarchy Process (AHP)the index system is constructed.Finallythe weights are obtained through judging and scoring of the expertsand the judgment value is modified through wavelet neural network.Compared with avionics effectiveness evaluation under specific capabilitythe evaluation system has more abundant indexes and good correction effect on the weights of the indexeswhich provides a tool prototype support for avionics system effectiveness evaluation.

The detection of Foreign Object Debris (FOD) is of great significance to airports safety operations.To this enda set of UAV-based FOD detecting and locating system is designed.The system adopts the roundabout route planning strategy of multi-UAV collaboration to collect images across the airport runway.For FOD location marking probleman automatic aligning and stitching algorithm based on RootSIFT feature matching is proposed and used to restore the panoramic image of the airport runway.Finallythe Faster R-CNN detector is designed to realize the detecting and locating of FOD.The test results based on the measured data show thatthe system can improve the FOD detection efficiency and obtain a satisfactory detecting and positioning effect.

A combined low-pass filtering technology for space Fiber Optic Gyroscope (FOG) is proposedwhich is easy to implement in FPGA.When the momentum wheeli.e.the actuator of satelliterotates at high speedit will cause low-amplitudehigh-frequency vibration of the satellitewhich is harmful to the satellite.If the bandwidth of FOG is too highthe high-frequency vibration will lead to greater noiseand even mixed output of FOGwhich will result in degraded control quality of the satellite.In consideration of both cost and effect of the filterwe propose a design scheme of FIR + IIR combined filter.The scheme uses five hundred and sixteen logic unitswhich only accounts for 13.8% of the chip resources of A54SX72A.It can be added to the FPGA program of FOG closed digital loop.The advantage is that it is easy to realize high-speed filtering and improve the effect of filtering out the high-frequency interference signals.In this waythe users dont have to design filters separately in other terminals.The configuration cost of the whole satellite control system is reducedand the application configuration of space FOG is simplified.

The parameter uncertainty and environmental interferences may result in unstable attitude of quadrotor aircraftsand the traditional PID control cant meet the control requirements of the quadrotors attitude stability and maneuverability.Aiming at the probleman Extended State Observer (ESO) neural network RBF PID controller is proposed.Firstlythe extension characteristics of ESO and nonlinear functions are used to estimate and compensate for disturbances to reduce system errors.Secondlythe ESOs estimated value of the system output is used as the input of the RBF neural networkto make the gradient information more accurate and better optimize the parameters of the incremental PID.Finallya Gaussian function is adopted as the excitation function of the neural networkand the model control parameters are adjusted by using the self-adaptability and self-learning ability of the RBF neural network.The Matlab simulation experiment shows that:in the unknown interference environmentthe ESOs RBF neural network PID controller can significantly improve the anti-interference ability of the system,and has a smaller overshoot and better robustness.

Test optimization selection is an important step in testability design.This paper mainly studies the test optimization selection problem under the condition of unreliable test.Firstlythe problem is reduced to a multi-objective problem for analysis.On this basisthe mathematical model of the problem is established with the test quantitytest cost and false alarm rate as the objectivesand the fault detection rate and isolation rate as the constraints.Thenbased on the Bayesian network testability modelthe MOPSO-NSGA2 algorithm is used to solve the problem.Finallythe method is used to make test optimization selection for the radio altimeter of a certain missileand a comparison is made with MOPSO algorithm and NSGA-2 algorithm.The results verify the effectiveness and feasibility of the method.

The support system of wind tunnel model has large motor power and complex operating environmentthus the balance signal is influenced seriously by motor electromagnetic interferencewhich may affect the accuracy of experimental data.To solve the problema technology for interference signal processing based on wavelet reconstruction is proposed.Firstlythe frequency threshold of wavelet reconstruction is determined based on the balance step test principle.Thenwavelet basis of db10 is usedwhich divides the signal wavelet into 10 layers.Analysis is made to each layer by Fast Fourier Transform (FFT)the signals under 1 Hz are reservedand thus signal reconstruction is realized.Through the use of wavelet reconstruction technologyElectromagnetic Interference (EMI) is significantly suppressed and the accuracy of the processed data meets the requirement on experimental accuracy.

In view of network management of miniaturizedlow-cost airborne stores mounted in quantitythe application characteristics of airborne store bus network and three typical store bus networks are studied at first.Thenthe characteristicstopology structurebasic protocolextension protocol of the EBR1553 bus network and its application in airborne store network are analyzed.Finallythe feasibilities of store EBR1553 bus network technology and its application are verified through the developed simulation environment.

Aiming at the complicated carrier-landing environment and the heavy operating burden of the pilot in manual approach and precision landing control of the carrier aircrafta two-layer Nonlinear Dynamic Inversion (NDI) control method is usedand Direct Lift Control (DLC) is introduced to achieve trajectory control of the precision carrier landing.Control allocation is applied to deal with multiple control surfaces and decoupling of control surface effects of DLC and conventional surfaces.From the aspects of trajectory and attitude decouplingsimulation analysis is made to the designed control frameworkand it is demonstrated that it is effective in achieving the precision carrier landing similar to Magic Carpet.Finallythe E-2C aircraft model is built and simulated for verification of the designed method.The results show that the control method can perfectly accomplish trajectory control in carrier landing.

In the presence of range-false-target deceptive jammingthe radar receives false signal with high fidelitywhich is difficult to be filtered out effectively from the signal level.The false signal may enter into radar data processor and form false plotsand then form false tracks after filtering and tracking.To solve this problema target tracking technique is proposed based on radar network data fusion.Firstlythe position measurement data of each radar in the radar network are preprocessed respectively.Thencorrelation processing is carried out in the fusion centerand the equivalent position measurements are obtained to realize accurate positioning for targets.Finallytarget tracking is realized.The technique can overcome range-false-target deceptive jamming effectivelyand is also appropriate for non-jamming situation.Simulation results verify the application value of the technique.