To address the challenge of the real-time weapon target intelligent assignment in modern high dynamic air defense operation, a neural network based air defense weapon target intelligent assignment method is proposed in this paper. An optimization model for weapon target assignment is first established considering the number of weapons and interception capability in order to optimize the damage efficiency. A neural network based weapon target intelligent assignment framework is then developed, where a neural network is constructed based on the existed training sample set and employed to assign the weapons to the targets effectively and efficiently. The comparison results illustrate that the neural network based air defense weapon target assignment method can obtain better weapon target assignment solution compared with the conventional discrete differential evolution based assignment method. Moreover, the computational cost of the proposed method can be reduced by 99. 9%, which demonstrates the effectiveness and practicability of the research work.

With the rapid development of artificial intelligence and computer vision, using intelligent object detection technology to identify the aircraft and UAV has become a common military defense means. Aiming at the feasibility of confusing and deceiving the object detection model based on deep learning network from the digital image level, and materializing the digital attack results in the future, this paper proposes a limited area adversarial attack method for vision-based aircraft intelligent object detection. The digital image of aircraft is used to simulate the adversarial attack of YOLO v3 object detection network. The effectiveness of the proposed method is verified by the experiments on the aircraft categories in PASCAL_VOC_2007 dataset.  The work of this paper provides the basis for further research on whether the digital analog attack can be transferred to the aircraft entity.

With the continuous innovation of information technology, the UAV cluster system will occupy a dominant position in the future battlefield. Analyzing and studying its defense strategy and interception means is particularly important for China’s unmanned technology innovation and system construction. Combining with the requirements of anti-group combat missions and based on the OODA cycle theory, this paper establishes a combat network model. It constructs an index system of anti-UAV cluster combat capability. Using the improved information entropy method, it proposes an efficient and accurate combat effectiveness evaluation method. Through the effectiveness evaluation of an anti-UAV cluster combat process, this paper verifies the feasibility of this method.

Combining with the characteristics of the integrated air and space distributed combat, the development status and drawbacks of the existing space-air integration defense system are described systematacially in the paper. Moreover, the development of the space-air integration distributed defense system (SAIDDS) is analyzed in details. Finally, the key technologies of the SAIDDS are summarized, and the corresponding development proposals are also proposed in the paper.

Aiming at the threat of intelligent swarming in the future air defense, this article focuses on the impact of intelligent swarming on air defense operations, including detection and identification, interception, command and control. It analyzes the development trend of foreign anti-swarming equipment technology from aspects of detection and identification, soft and hard damage. Finally, the relevant strategic thinking of countering intelligent swarming is proposed, which can provide an important reference for the design of the capability architecture of the anti-intelligent swarming operation system and equipment technology development.

To conduct the anti-cluster UAV (unmanned aerial vehicle) air defense system demonstration, this paper constructs a system capability evaluation index system and a system effectiveness evaluation index system from the per-spective of system capability and system effectiveness. The TOPSIS (technique for order preference by similarity to an ideal solution）method and simulation method are used to evaluate the system capability and system effectiveness of anti-cluster UAV equipment, and then complete the evaluation of contribution rate of anti-cluster UAV equipment system. An example is used to verify the feasibility of the method, which can support the demonstration and research of the equipment to a certain extent.

To solve the problem of low accuracy for multiple targets tracking in case of low SNR(signal noise ratio), a MIMO radar DOA(direction of arrival) tracking method based on SDE model is proposed in this paper to correct error of tracking model. The stochastic differential equation and the particle filter based on Markov Chain Monte Carlo (MCMC) are used for modeling and continuous reasoning for the multiple targets. The effectiveness of this method is verified by the simulation results, the tracking accuracy of motion multiple targets can be improved by more than 20% in case of low SNR.

The low resolution and poor contrast of infrared image make the line detection of infrared image more difficult. Hough transform has strong anti-interference and robustness, and it is a very effective line detection method in the field of computer vision. In this paper, a method of infrared image line detection based on improved Hough transform is proposed. Experimental results show that under the conditions of low resolution and fuzzy edge, the method can detect the edge line well, and determine the endpoint and length of the edge line.

针对多无人机协同飞行情况下的路径规划问题，为了解决传统路径规划算法搜索效率低、难以处理复杂约束的缺点，提出了一种基于Radau伪谱法的无人机集群路径规划方法。首先，建立了无人机的动力学模型，考虑了集群飞行过程中的约束条件，将其转化为非线性约束；然后，确定了路径规划的性能指标，建立了无人机集群路径规划的最优控制模型；最后，使用Radau伪谱法求解最优控制问题，并分别进行了无人机集群集结任务仿真和编队飞行任务仿真。仿真结果表明，基于Radau伪谱法的无人机集群路径规划算法可以得到满足状态约束、路径约束和控制约束的飞行轨迹，符合多架无人机协同飞行的需求，具有较强的实用价值。

In order to realize the UAV target recognition and detection in complex background, an improved Faster R-CNN network model is proposed. Aiming at the difference between the target size in the field of view and the labeled training dataset, three different scales of receptive field parallel detection and weight sharing feature extraction structure Tri-VGG network are designed, and a training strategy is designed to avoid network over fitting. The optimized network detection average precision (average precision, AP) reaches 90. 9%, and the real-time performance reaches 24 frames per second, which meets actual needs.

The controller design problem of a class of UAV (unmanned aerial vehicle) formation systems with bounded disturbance is studied in this paper. Based on adaptive integral sliding mode control frame, a novel adaptive controller is proposed. The adaptive law is introduced to estimate the upper bound of disturbance. The effect of disturbance can be cancelled in the designed protocol. The stability of the closed UAV formation system is analyzed by Lyapnov method. Simulation results validate that the controller proposed in this paper can guarantee the robustness of the closed-loop system, the continuity of the control signal and the signal is relatively small. Hence, the method is prone to apply in practical engineering.

To solve the problem of time cooperation reentry guidance for hypersonic gliding vehicles, a reentry guidance method based on dynamic angle of attack profile is proposed, which can realize the terminal time constraint. Firstly, the angle of attack profile and the height-velocity profile are determined by two parameters. The two profiles can ensure that the aircraft meets terminal state constraints and process constraints. The remaining range and time are predicted online by longitudinal guidance via iteratively correcting two parameters. Combined with the lateral heading angle corridor, the time-constrained entry guidance is implemented. The adaptability of the proposed method in different tasks is verified. The simulation results show that the method is effective to satisfy the time constraint and other terminal constraints.

采用风洞试验和数值方法对影响折叠翼外翼气动特性的因素进行研究，包括外翼折叠角、攻角、马赫数、气流滚转角、弹身截短长度等。研究表明：“×”型状态下折叠角为30°时外翼载荷最大；外翼载荷对马赫数变化不敏感，马赫数为1.2时气动力系数最大；弹身截短的长度取决于头部压力区是否会影响到尾翼区；不同气流滚转角下的外翼载荷大小与外翼当地气流偏角正相关。研究结论可作为折叠翼简化设计的依据。

In view of the possibility that the rapid development of the small commercial unmanned aerial vehicle(UAV) market will cause serious harm to public social security, the threat scenarios of small commercial UAVs are introduced. According to DJI Phantom 4 UAV, the target characteristic data such as its shape, flight parameters, communication frequency, radar cross section, infrared radiation intensity and acoustics are analyzed. The early warning technology and development status suitable for detecting small commercial UAV are analyzed from the aspects of radar, photoelectric, radio and acoustic detection principles, technical difficulties and foreign advanced equipment, and the application suggestions for different threat scenarios are put forward. The research has reference value to small commercial UAV counter technology.

Aiming at the engagement of large maneuver target by air defense missiles, a sliding mode guidance law based on immersion and invariance methodology is designed. First, the relative motion of the missile and the target is established. Secondly, in order to accelerate the convergence rate and hit precision of the guidance law, the finite-time convergent guidance law is studied and the backstepping method is applied. For the large maneuver target, the immer-sion and invariance estimator is applied to estimate and compensate the target acceleration online, and the nonlinear function in the estimator during guidance is studied. Finally, the stability of the proposed guidance law is analyzed by the Lyapunov function and the effectiveness of the proposed guidance law is verified by the numerical simulation.

Aiming at the problem of track generation of unknown maneuvering target in clutter environment, an algorithm combining multi hypothesis track with interactive multi model (MHT-IMM) filtering is proposed. By making full use of all kinds of characteristic information of plot and historical accumulated information of fusion track, the acceleration of maneuvering target can be predicted, and appropriate filter output track can be selected adaptively. The maneuvering coverage is wide, and the correctness and real-time performance of track generation are improved. The digital simulation results show that the method can meet the requirements of track generation for large and small targets in complex environment, and the track is complete with high tracking accuracy.

Based on the ant colony algorithm, this paper studies the cluster cooperation problem of multi-UAV system in cooperative reconnaissance of complex target group, and optimizes the ant colony algorithm through cross strategy to improve the accuracy of the algorithm and form a better track scheme and reasonable scheduling strategy. The solution of this problem can not only deepen the research of multi-task allocation and multi-objective cooperative field, but also provide assistant decision-making for troops to carry out combat tasks.

The traditional task assignment algorithm mainly solves the single target optimization under the specific problem, which is difficult to adapt to the complex situation under the real battlefield cooperative strike task. In view of the above problem, this paper constructs a multi-objective optimization model of UAV task assignment by considering the factors such as flight path length, target strike effect, target value and weapon cost in the cooperative strike mission of UAV to determine the task assignment scheme based on NSGA-Ⅲ algorithm, and the Pareto optimal solution set is obtained. The simulation results show that the multi-objective task assignment model and its solution proposed in this paper can provide an effective and reliable diversified assignment for cooperative strike.