In future penetration combat operations under the complex electromagnetic environment, the cooperative manned/unmanned aerial vehicles will face the problem of jamming resource allocation.To solve the problem, a method for the optimization and effectiveness evaluation of the dual-aircraft blinking jamming strategy is proposed.Under the condition of blinking jamming of cooperative manned/unmanned aerial vehicles and according to the battlefield situation, the proposed method can quickly calculate the optimal blinking period, the lateral and vertical spaces of the formation, the penetration probability and its improvement in real time, so as to evaluate the jamming performance.The simulation results show that the proposed blinking jamming strategy can greatly improve the penetration probability when the formation of manned/unmanned aerial vehicles is performing a penetration operation mission.

To solve the problems of frequent collision of elephant flows, severe link congestion and unbalanced load in data center networks, a multi-path traffic scheduling algorithm based on Ant Colony Optimization (ACO) is proposed.The algorithm combines the multi-path characteristics of network topology with the advantages of Software Defined Network (SDN) global view.First, the large and small flows are detected by using the sFlow controller.Then, the K short path algorithm is used to add the non-shortest paths on the basis of the shortest path, and calculate the criticality of each path in the path library.The less critical path is selected as the candidate path to reduce the cost of the scheduling algorithm.Finally, according to the characteristics of the large and small flows, the link transmission cost and the pheromone's update method in the ACO are set respectively, and the multi-path traffic scheduling is realized by calculating the number of optimal paths and assigning weights.Simulation results show that under different-type traffic modes, the proposed algorithm effectively improves the average network bisection bandwidth, lowers the maximum link utilization, and reduces the average transmission delay.

For exo-atmospheric interception, an adaptive sliding-mode guidance law is presented for accelerating the missile without available aerodynamic force.The missile intercepts the target through fixed axial acceleration and corresponding attitude adjustment.Zero-effort-miss distance is chosen as sliding surface for guidance law derivation.The guidance law comprises the equivalent controller and the uncertainty controller.The equivalent controller is used for eliminating the influence of line-of-sight rotation to reduce miss distance, while the uncertainty controller adapts to the effects of external disturbances such as target maneuver to ensure the rate and stability when reaching the sliding surface.The designed guidance law can be transformed to an attitude control problem and be applied to the exo-atmospheric interceptor with attitude control device.Numerical simulation and analysis are conducted on the derived sliding mode guidance law.The results show that the guidance law has strong robustness and good trajectory characteristics, and can realize target interception effectively.

To solve the problem of poor performance of the traditional Kernel Correlation Filter (KCF) under complex situations such as target scale variation, occlusion and deformation, a KCF tracking algorithm with adaptive model update based on Radon transform is proposed.By using the motion information, the optical flow method and the inter-frame difference method are adopted to predict the possible region of the target, which significantly reduces the search range and improves the speed of the algorithm.The Radon transform is insensitive to noise and invariant to moment translation and scale change.The optimal scale is determined by using the peak value of the matching degree of the moment features, which improves the accuracy of the algorithm while reducing the calculation amount.According to the nonlinear relationship between the learning rate and the peak value of the response graph in the model updating strategy, a parabolic learning rate curve is constructed to adaptively update the model.The tracking accuracy of the algorithm is guaranteed even when the target is temporarily lost or is false.The experimental results show that the proposed algorithm has good real-time performance, high success rate and high tracking accuracy.

Most of the existing fusion methods cannot well preserve all the significant features of the source images.To solve the problem, a novel multi-sensor image fusion method driven by multiple gradient features is proposed.Two-scale decomposition is performed on the source images by using the Gaussian filter for obtaining the base layers and the detail layers of the source image.A morphological gradient operator is used to extract the gradient feature of the base layers and the detail layers to construct the saliency maps, which are optimized by Gradient Domain Guided Filtering(GDGF).The fusion of the base layers and the detail layers is guided by the optimized saliency maps for obtaining the final fused image.The experimental results of test image sets demonstrate the superiority of the proposed method in the subjective and objective evaluation.

The finite-time containment control for multi-agent systems only enables the followers be in the leaders convex hull, which does not realize the consensus control of the followers.To solve the problem, a containment consensus control algorithm in the form of piecewise function with finite-time convergence is proposed.In addition, in view of the unknown velocity information, a velocity observer is designed to estimate the velocity information.Through Lyapunovs second theorem and homogeneous finite-time stability theorem, it is proved that the leaders can realize finite-time containment control of the followers and the followers can achieve finite-time consensus.Finally, the effectiveness of the control algorithm is proved by a large number of numerical simulations.

In this paper, the mean-square exponential stability is studied for neutral neural networks with time-varying delays, uncertain parameters and Lévy noise.Based on the sliding mode control, the generalized It? formula and the Lyapunov stability theory, several sufficient conditions are obtained for the reachability of the specified integral Sliding Mode Surface (SMS) and the mean-square exponential stability of neutral neural networks on SMS.Finally, the effectiveness of the obtained result is verified by a numerical simulation.

The problem of finite-time consensus tracking control is investigated for the second-order leader-follower Multi-Agent Systems (MAS) with disturbance and unmeasured velocity.Firstly, to solve the problem of unmeasured velocity and disturbances, a finite-time super-twisting observer and a sliding-mode disturbance observer are designed respectively to quickly estimate and compensate for the disturbance.A fast, finite-time arctangent tracking differentiator is introduced by using the observed velocity to quickly estimate the leaders input information.Secondly, based on Integral Sliding Mode (ISM), a finite-time tracking controller is designed, which can effectively improve the response speed and tracking accuracy of the system, and enhance the anti-disturbance performance and robustness of the system.Finally, several numerical simulations are provided to demonstrate the effectiveness of the proposed methodology.

In view of small radar scattering area and weak echo energy of weak targets, which lead to a large number of false-alarm clutters in the tracking process and the problem of false tracking and missed tracking, this paper proposes a GM-PHD algorithm assisted by amplitude information.By establishing amplitude likelihood functions of the weak target and the false-alarm clutter, the algorithm uses the amplitude information to improve the recognition rate of the target and the clutter, and improves the state estimation and potential estimation of the multiple targets.In addition, the algorithm is extended by Unscented Kalman Filtering (UKF) to adapt to nonlinear tracking.The simulation results show that the proposed algorithm has better performance on the multi-target's state estimation and potential estimation than the traditional algorithm of Gaussian mixture probability hypothesis filtering.

To solve the problem that the determination of the order of an effective rank in Singular Value Decomposition (SVD) algorithm is not convincing enough and it is difficult to be applied in engineering, a new method utilizing noise characteristics is proposed.First, the singular value sequence is obtained by phase space reconstruction.Secondly, the fractal dimension of the filtered signal corresponding to each point is obtained.Finally, the optimal order of an effective rank is determined by the actual fractal dimension of noise.Simulation examples show the effectiveness of the proposed method, and it is compared with two typical algorithms to illustrate its advantages resulted from making full use of noise characteristics, and the variation of the algorithm performance with SNR is given.

Aiming at the position estimation error caused by far-point extrapolation in the existing ballistic extrapolation algorithm of the gun reconnaissance radar, a ballistic extrapolation algorithm based on the inverse extended Kalman filter is proposed.Based on the traditional ballistic extrapolation algorithm, a filtering model of seven-dimensional state vectors including the trajectory coefficients is established in the algorithm, and the end point of forward filtering is regarded as the start point of inverse filtering to implement the inverse extended Kalman filtering.Starting from the end point of the inverse filtering, the fourth-order Runge-Kutta equation is used to extrapolate towards the ballistic starting point.Simulation experiments and results show that the extrapolation accuracy of the proposed ballistic extrapolation algorithm based on the inverse extended Kalman filter is improved by about 50% compared with that of the original algorithm.

To solve the problem that the accuracy of the acquired SAR image information is reduced due to the interference of external environment, a low-rank approximation method for high-order SAR images based on Tensorial Singular Value Decomposition (TSVD) is proposed.Firstly, on the basis of classical Singular Value Decomposition(SVD), the classical two-dimensional matrix is extended to the tensorial high-order matrix by using the neighborhood selection method.Secondly, the classical matrix algorithm is extended to the algorithm related to the tensorial matrix by using “t-product” model, and the specific implementation process of TSVD is obtained.Finally, the performance of low-rank approximation under the conditions of TSVD is compared with that of the classical SVD by using structural similarity and PSNR.The simulation results show that, compared with the classical SVD, the TSVD fully considers the interaction and spatial structure between image pixels, and with the expansion of the order of the tensorial matrix, the higher the similarity of image structure and the higher the PSNR.The method can be applied to the low-rank approximation and reconstruction of high-order images.

VIO-SLAM (Visual Inertial Odometer Simultaneous Localization and Mapping)means that mobile robots use cameras and IMUs as external sensors to build external maps while positioning themselves.This paper reviews the VIO-SLAM positioning system from such modules as the front-end visual inertial navigation odometer, back-end optimization, loop detection and the mapping.As to the visual inertial navigation odometer, the similarities and differences between the feature point method and the optical flow method, and the process of IMU pre-integration are stated.Back-end optimization mainly points out how to deal with the noise in VIO-SLAM process and the bias of the gyroscope and acceleration in IMU.Loop detection mainly expounds how to solve the problem of position estimation drifting with time.In the mapping part, several main mapping methods and strategies are given according to the purpose of mobile robots.Finally, the development trend and prospect of VIO-SLAM are given.

Initial alignment is one of the key technologies in the inertial navigation system, and the aligned time and alignment accuracy will directly affect the performance of the inertial navigation system.This paper analyzes current research difficulties in initial alignment of the inertial navigation system from the following three aspects:Modeling errors, random errors of the inertial sensor, environmental disturbances and other unpredictable factors.Then, this paper focuses on the current research progress of initial alignment in terms of three types of methods:Traditional two-stage alignment, nonlinear alignment and optimization-based alignment, as well as motion alignment.In future research, the speed, accuracy and applicability of initial alignment of the inertial navigation system will be further improved by taking manipulation factors as disturbances, adding them to the error model, and using a new type of strong tracking filter to deal with underwater nonlinear non-Gaussian noise.

The existing artificial potential field method is prone to fall into local minimum and unable to escape from it when the UAV confronts some typical continuous obstacles (U-shaped, L-shaped, line-shaped, etc.).To solve the problem, the mathematical model of the electric potential field is used, and the concept of simulated equipotential line is added to the existing artificial potential field method.The repulsive force function of typical continuous obstacles is redefined, and an improved artificial potential field method based on the simulated equipotential line is presented.The method is able to escape from local minimum, and is then applied to the UAV's path planning.The rationality of the improved algorithm is verified by simulation experiments.While maintaining the advantages of high speed and simple implementation of the artificial potential field method, the problem of the UAV's path planning with the artificial potential field method when the UAV is faced with continuous obstacles is solved.

In view of the problem of initial alignment of the moving base caused by the disturbance when an individual shooter carries the small ammunition on the shoulder, an initial alignment method that can reduce the disturbance caused by the irregular movement of the human body is studied.In view of the characteristics of the moving base alignment of the inertial guidance system, an error model is established.The accelerometer is used to acquire the data in real time for coarse alignment.In the process of refined initial alignment, the extended Kalman filter is used and the errors of position and velocity are taken as an observation for estimation.The attitude matrix is updated in real time, and finally the accuracy of the proposed method is analyzed by using the covariance of each axial attitude angle.The results show that using this method can effectively improve the initial alignment accuracy of the small ammunition carried by an individual shooter.

The quadrotor UAV has complex aerodynamic characteristics and is prone to be interfered, which will greatly affect the stability of the UAV.In order to enhance the anti-interference ability of the UAV's attitude control system, an attitude control system is designed based on the improved Active Disturbance Rejection Control (ADRC).The proposed method combines the traditional ADRC technology with the global fast terminal sliding mode control technology, which is used to optimize the feedback control law of the nonlinear state error in the ADRC system.In addition, the ADRC system is redesigned, and its stability is proved by using Lyapunov theory.Experimental results show that the improved ADRC system has faster response speed and stronger anti-interference ability in attitude control.

The RGBD semantic segmentation model based on the standard 2D convolution kernel mostly takes the depth map as a single channel.Due to the limitation of convolution kernel characteristics, the geometric structure information brought by the depth information cannot be fully exploited.To overcome this defect, this paper constructs depth-sensitive convolution kernels and a pooling layer to make rich mining of depth information,and uses depth-sensitive spatial pyramid pooling to extract multi-scale information, so as to realize the segmentation of objects of different scales.Results of experiment on NYU v2 and SUN RGB-D datasets show that this method effectively improves the overall semantic segmentation accuracy.

In order to improve the torque tracking accuracy of a load simulator for a servo system, a method of composite terminal sliding mode control based on the fuzzy RBF neural network is proposed.Firstly, by analyzing the system's composition and working principle of the servo load simulator, the torque motor model is simplified.According to the models of the torque sensor and the inertia disk, the simplified equivalent model of the servo load simulator is established.Then, a fast terminal sliding mode controller is designed.In order to improve the dynamic quality of the sliding mode, the parameters of the sliding mode surface are dynamically adjusted by using the method of the fuzzy neural network.At the same time, in order to improve the learning and training speed of the fuzzy RBF neural network, the nearest-neighbor hierarchical clustering and the conjugate gradient algorithm are used to adjust the parameters of the network, and local optimization is carried out to improve the performance of the algorithm.Finally, simulations are conducted, and the results show that the method can improve the system's control accuracy and has good dynamic characteristics.

The tethered Unmanned Aerial Vehicle (UAV) is a new type of aircraft developed for long-duration aerial tasks, which combines a rotorcraft with a ground tethered unit.In actual operation, the launch and recovery of a tethered cable will lead to changes of dynamic performance of the tethered UAV, which will affect its trajectory tracking control, and so does the lumped disturbances which contains unmodeled dynamics and gust disturbances.To solve the above problems, a robust adaptive Sliding Mode Controller (SMC) based on the disturbance observer is proposed.Firstly, the catenary theory is introduced to describe the tension of the tethered cable.The dynamic model of the whole system is established by using Newton-Euler equation.Secondly, under the framework of Lyapunov theory, the disturbance observer is used to estimate and compensate for lumped disturbances and an adaptive SMC is designed to ensure the convergence of trajectory tracking errors.Finally, the effectiveness of the proposed controller is verified through the analysis of simulation cases and experiment.The results show that the proposed controller has a higher trajectory tracking precision compared with the terminal SMC and it can help the tethered UAV to track complex trajectories.

The hole-set method is one of the most widely-used methods for measuring the divergence angle of airborne laser detection system in the process of alignment and inspection.In view of such factors as the quality of incident beam, sensitivity of power meter and positioning accuracy of the stop, errors will occur between the measured value and the real value of the divergence angle.In this paper, the theory of laser transmission and transformation is adopted to analyze the influence of the above-mentioned factors on the measurement results.Our findings can provide instructions for the selection of such parameters as the diameter of the stop, the focal length of the lenses, sensitivity of power meter and positioning accuracy of the stop in the process of divergence angle measurement, so as to reduce the measurement error and improve the measurement accuracy.

The flocking control problem of a group of quadrotor aerial vehicles is studied under limited distance for real-time information exchanges and bounded constraints on the input controller.The method of combining the bounded function with the modified saturation function is presented to design the bounded total-thrust controller in the position subsystem, which will adjust the positions and velocities of the group of quadrotor aerial vehicles autonomously, so as to realize ideal collision avoidance and cohesion maintenance, as well as velocity consensus.The generalized bounded torque controller is designed by combining the bounded transformation term with the modified nested saturation function to adjust the attitude of the aerial vehicles autonomously, so as to achieve the yaw angle consensus in the rotational subsystem.Finally, simulation results have proved the validness of the designed input controller.