In the cooperative penetration application scenario, there is an immediate need to optimize the communication topology and wing aircraft formation structure. A formation configuration optimization design method based on penetration thoroughfares is proposed to address the issues of reference benchmark selection and modeling of aircraft/interception force/battlefield relationships in formation configuration optimization. The leader aircraft’s role is obtained by the optimization of the communication topology. The formation configuration is referenced by the geometric centers of the leader aircraft and each team leader in turn. Explicit expressions for the formation configuration are designed to eliminate the dependence on obtaining prior information about the leader aircraft. A penetration thoroughfares model is established for winged aircraft to ensure their advantages in detection, anti-detection, and maneuver avoidance capabilities at various battlefield grids. To address the issue of balancing information sharing and network load on communication topology optimization, a communication topology is constructed under the constraints of formation configuration. In order to optimize both formation configuration and communication topology, a topology switching strategy depending on battlefield conditions is suggested after a communication topology optimization approach based on minimal spanning tree and optimal rigid graph is presented. Finally, the effectiveness of the designed optimization method is verified by taking the cooperative penetration against military threats for winged aircraft as an example.

For the problem of multi-robot collaborative search and source localization in unknown environments, a distributed collaborative search and coverage method based on Voronoi diagrams is proposed. In order to assure safety, this method initially addresses collision problems caused by the physical dimensions and positioning faults of the robots by building Voronoi buffer zones based on each robot’s placement uncertainty radius. Utilizing sparse Gaussian process regression and the centroidal Voronoi tessellation (CVT) algorithm with an uncertainty regularization term, the distribution of the unknown concentration field is reconstructed to achieve collaborative coverage. An adaptive environmental exploration strategy is proposed to enable environment exploration without prior information. Simulation experiments demonstrate that this method can rapidly complete exploration of unknown environments and accurately locate the position of the pollution source.

Full trailers are at a significant risk of collision since the local planning algorithm for regular passenger cars does not completely account for the entire trailer system. To address this issue, a refined dynamic window approach (DWA) is proposed specifically for off-axis full trailer systems to enable proactive obstacle avoidance for unmanned full trailer systems on unstructured roads. Initially, the sampling of the towing vehicle’s speed constructs a velocity vector space. The motion paths of both vehicles are then forecasted using the kinematic model of the system and the data that were sampled. Subsequently, introducing sub-cost functions related to the target point’s position, an evaluation function tailored to the trailer system is proposed. Finally, the optimal velocity is selected based on the evaluation function to ensure the system safely reaches the target point. Experimental results demonstrate the method’s reliable safety in obstacle avoidance tasks, with a minimum distance of 0.83 meters between the towing vehicle and obstacle boundaries in real vehicle experiments, and 0.89 meters for the full trailer from obstacle boundaries.

A joint apron-taxiway assignment model is designed with an emphasis on improving system resilience in order to alleviate the impact of severe weather on aircraft taxiing and enhance the resilience of airport surface operations. Firstly, the performance of the surface operation system is characterized by the taxiing time of aircraft, and the resilience is quantified by the loss and recovery of system performance. Then, based on the topological network structure among the runway, taxiway, and apron, the joint apron-taxiway assignment model is established considering the taxiing time, passenger boarding time and system operation resilience. A three-step algorithm based on linear iteration is developed, taking into account the complex nonlinear model. Finally, a case study based on the Tianjin Binhai International Airport is conducted. The surface operating system’s recovery speed is enhanced by 16.67% and its average performance is raised by 20.68% following optimization. In addition, the average resilience and recovery speed are increased by 20.33% and 27.15%, separately. It indicates that the optimized scheme can facilitate the system’s adaptation to severe weather, reduce the performance loss in the adaptive period, propel the recovery speed of system performance, and ensure its relative stability.

Using factor graph (FG), a master-slave cooperative localization technique is suggested to meet the high-precision positioning needs of autonomous underwater vehicle (AUV) clusters. First, the state equation and measurement equation for a master-slave AUV cooperative localization system are formulated, and a corresponding FG model is constructed. Second, message passing between nodes within the FG model is derived using the sum-product algorithm (SPA), leading to the acquisition of the probability density function (PDF) for the slave AUV’s position. In order to carry out useful experimental verification, a one-master-one-slave cooperative localization test platform is subsequently set up utilizing ground vehicles, GPS, inertial equipment, and data link equipment. The experimental results demonstrate that the proposed cooperative localization algorithm can enhance positioning accuracy by 18.60% compared to the conventional extended Kalman filter (EKF)-based cooperative localization algorithm. Additionally, the results indicate that ranging errors significantly impact the accuracy of cooperative localization

This work proposes a multi-objective collision avoidance optimization technique based on model predictive control (MPC) to reduce the probability of rear-end collisions during aircraft taxiing and for passenger comfort. Firstly, the longitudinal kinematic model of the airplane is established. Considering the safety of aircraft taxiing and passenger comfort design objective function and constraints. Secondly, the design of variable weight functions using relative velocity and spacing as parameters. Introducing it into the MPC to optimize security weights. The desired acceleration is obtained by solving the variable weight MPC control strategy using the sequential quadratic programming (SQP) algorithm, and analyzing the stability of variable weight MPCs. Lastly, simulation tests are used to confirm that the proposed algorithm can prevent collisions under two common operating situations The experimental results show that the proposed algorithm is useful in achieving the deceleration collision avoidance, and optimized acceleration change amplitude improves passenger comfort.

This paper addresses the problem of task execution for unmanned aerial vehicles (UAV) swarms, considering the coupling characteristics of UAV task allocation and trajectory planning as well as the no-fly zone constraints. A task planning algorithm is proposed that can make the UAV swarm reach the target positions in the shortest time simultaneously. A "hovering waiting and dynamic speed adjustment" method is used to synchronize the arrival time of each UAV, Dubins curves are used to design the pathways, and an upgraded particle swarm optimization (PSO) algorithm with particle swarm mutation is used to optimize the task allocation scheme. Finally, the effectiveness of the algorithm is evaluated and verified in a simulation environment based on the six-degree-of-freedom dynamics model and the dynamic inverse control model. In contrast to the conventional PSO algorithm approach, the simulation results demonstrate that this enhanced PSO algorithm is capable of successfully escaping the local optimum and achieving a better allocation scheme. Under the control of the proposed algorithm, the maximum deviation of flight time among multiple UAVs is only 0.5%, meeting the requirements of a saturation attack.

This paper proposeds a self-supervised signal separation method based on a linear convolutive mixing process. The method comprises three components: a linear convolutive mixer, a semantic token extractor, and a query-based signal separator. During the training phase, source signals undergo convolutional mixing within the mixer, which is a better mimic of the realistic optical fiber sensing process when compared with the linear simultaneous mixing process, resulting in a mixed signal that could be used for the self-supervised learning of the separator. The source signals' embeddings are then produced by the semantic token extractor and used as query tokens in the separator. Finally, mixed signal and source embeddings are combined and fed into the separator to produce the target source signal. The proposed method could be even used in a zero-shot setting. And the number of training samples could be expanded with this random combination of mixed signal and source embedding. In an interior setting, experimental optical fiber sensor data are gathered, including cyclical vibrations and human motions like jogging. The results of the signal separation experiments demonstrate the effectiveness of the proposed method.

Resources are typically optimized using the radar cross section (RCS) statistical model in the detection process of conventional collocated multiple-input multiple-output (MIMO) radar networks. However, the RCS of stealth targets changes dynamically, which can lead to the degradation of target tracking accuracy or even target loss. To address this problem, a collocated MIMO radar networked system resource optimization allocation algorithm for stealth target tracking is proposed. Firstly, the target state is estimated using the covariance intersection (CI) fusion filtering algorithm, and the predicted Bayesian Cramér-Rao lower bound (BCRLB) under the CI fusion criterion is derived. After that, the target RCS is predicted based on the property that the target RCS is related to the radar predicted observation angle, and the objective function is consisted of the weighted sum of individual target BCRLB. Consequently, a beam and power optimization algorithm under the RCS predicted model is established. Subsequently, a contribution-based fast solution algorithm is proposed to solve the model. In comparison to the RCS statistical model strategy, simulation results demonstrate that the proposed algorithm can efficiently use the target RCS information to achieve a better resource allocation, which can increase the accuracy of stealth target tracking, under the stealth target RCS dynamically changing scenario.

For high-maneuverability aircraft, maneuver loads constitute the primary design constraint for airframe structural strength, significantly impacting structural mass and fatigue damage accumulation. To address the requirements for lighter airframes and extended service life, a wing maneuver load control methodology was developed utilizing normal acceleration load factor as the feedback parameter and implementing active control surface deflection. Focusing on typical extreme maneuvers of conventional-configuration high-maneuverability aircraft, an optimal load control strategy was derived through systematic evaluation of wing control surface deflection effects, thereby establishing deflection parameters for subsequent simulations. Comparative analyses of wing maneuver load control effectiveness were conducted for multiple threshold-initiated strategies. Results demonstrate that initiating control at 75% of the maximum normal load factor and applying a 5° deflection command reduces peak wing bending moment by 10%. This approach shows significant potential in reducing structural load-bearing requirements and mitigating fatigue damage in high-maneuverability aircraft, supporting structural integrity enhancement.

Aiming at the assignment problem of multimissile and multitarget engagement (MME) scenario, this paper proposes a multi-to-multi energy optimal task allocation method based on interception capture region from the perspective of guidance. The characteristics of the capture region and optimal energy cost in the scenario of three-dimensional realistic true proportional navigation (3D-RTPN) intercepting arbitrary maneuvering targets are analyzed. The weight matrix is then built with the intention of achieving both the lowest total energy consumption and successful interception. In order to achieve multimissile and multitarget assignment (MMA), the adaptable Hungarian algorithm (AHA) is used. Numerical simulation is used to confirm the MMA strategy's efficacy. The effectiveness of the proposed method is verified by numerical simulation.

In order to effectively solve the problem of low-cost navigation and localization of UAV swarm under the satellite partial denial environment, a cooperative navigation method for UAV swarm based on the attitude heading reference system (AHRS) is proposed. Firstly, the design of the 3D position estimation model is completed using the AHRS as the basis. Secondly, the cooperative dilution of precision (CDOP) is used to finish the optimal node selection in a distributed cooperative navigation filter based on inter-aircraft range, which lessens the navigation system's computational load. Algorithms for fault identification and isolation are then used to diagnose the disrupted cooperative measurement data and reconfigure the system. Finally, the solution of the absolute position is accomplished using the distributed cooperative navigation algorithm. Simulation and experiments demonstrate that this algorithm effectively resolves problems such as excessive reliance on satellite navigation and slow processing of large-scale navigation data. Compared with traditional multi-source fusion algorithms, this approach significantly reduces hardware costs while meeting high-precision positioning requirements for a large-scale UAV swarm at a lower cost.

Carrier-based aircraft are an important part of the aircraft carrier battle group. There are many technical challenges in practical application, including the landing technology. This work proposes a sliding mode observer and nonlinear dynamic inversion technology-based direct lift control system for carrier-based aircraft, aiming to address the issues of multivariable coupling and difficult landing environments. In order to account for the impact of airwake on control accuracy, this research designs an adaptive sliding mode observer that can accurately assess the impact of external disturbances on the carrier aircraft's motion. To realize the decoupling of control inputs, a direct lift control system for carrier-based aircraft is established by using nonlinear dynamic inverse control technology, and a self-adjusting pigeon-inspired optimization (SAPIO) algorithm is proposed for parameter tuning of the system. The simulation results show that the proposed direct lift control system has higher control accuracy than the traditional proportional-integral-differential control system.

The error accumulation problem in srereo visual-inertial odometry (VIO) systems based on nonlinear optimization is serious when operating for extended periods in low-texture environments. Therefore, we propose an online temporal calibration method for the stereo VIO system based on nonlinear optimization. This approach makes full use of the benefits of stereo cameras by constructing error factors using epipolar constraints, which enhances system robustness and state estimation accuracy while lessening the detrimental effect of feature point mismatches on time offset calibration. It is suitable for low-cost, self-assembled systems. Experiments on public datasets show that the proposed calibration method has higher accuracy and faster convergence speed than current advanced calibration methods, thereby improving the accuracy and robustness of system state estimation. Experiments in real-world scenarios also validate the effectiveness of the proposed method.

Addressing the issues of low accuracy in terrain elevation matching and poor real-time performance in iterative search methods, we propose a neural network-based method for terrain contour-aided navigation. This study focuses on two-dimensional contour feature matching to enhance the robustness of matching algorithms under elevation noise. Considering the rotational and translational invariance characteristics of wavelet transforms, we extract contour edge features using wavelet transform sub-bands. Furthermore, we present a contour edge feature matching algorithm based on neural networks that replaces the conventional iterative search matching process by using multiple sub-networks for classification recognition, greatly enhancing the algorithm's matching accuracy and real-time performance. In comparison to terrain elevation matching, simulation results show that the suggested approach improves the matching success rate by more than 30% and reduces the matching time by more than 97% when compared to iterative search-based terrain contour matching techniques.

Image object detection technology has greatly improved the work efficiency of the security inspection and further guaranteed public security. However, the differences in imaging standards among different types of security inspection machines, the complexity of X-ray images, and the expensive cost of data annotation have constrained further research of object detection technology based on X-ray security inspection images. To improve the universality of our item detection system, we extend the dataset using a style transfer approach to account for variations in X-ray imaging hues of the same substance across various security equipment manufacturers. A refined feature pyramid network structure is proposed to extract richer semantic information from different levels in response to the significant differences in the size of similar objects to be recognized in X-ray images. A fine-grained classification module, which is simple to plug into the general object detectors, is what we suggest in order to increase detection accuracy even more. Meanwhile, this dataset contains 56659 X-ray images, featuring 37 types of contraband, with each image being high-quality annotated. This is a larger publicly available X-ray image dataset in terms of both the variety of contraband types and the number of images. Based on comparative experiments conducted on this X-ray contraband dataset, the model structure proposed in this article achieved an approximate 0.056 improvement in mean average precision (mAP) compared to the baseline model.

This study aims to tackle the collaborative path planning issue in multi-unmanned vehicle systems using deep reinforcement learning. We’ve devised an efficient path planning framework by first establishing kinematic and mathematical models for differential-drive unmanned vehicles and collaborative obstacle avoidance scenarios. Then, we addressed the challenges of slow training, low sampling efficiency, and poor adaptability of reinforcement learning in complex dynamic scenarios. For cooperative obstacle avoidance and pursuit, we suggested an improved twin delayed deep deterministic policy gradient (AE-TD3) algorithm. By introducing random noise to pursuing unmanned vehicle actions, exploration in unknown environments is improved, leading to efficient and stable collaborative obstacle avoidance and pursuit. Our method is validated by simulation results, which show faster convergence and a 16.7% reduction in pursuit time when compared to the twin delayed deep deterministic policy gradient (TD3) algorithm.

To improve the anti-interference performance of the controller during the test mass release phase of the space-borne gravitational wave detection mission, disturbance-observer (DOB) based trumpet tube model predictive control (MPC) is proposed for the steady-state establishment of the test mass. On the one hand, the controller’s performance in terms of disturbance immunity is enhanced by the DOB, and high precision estimation is achieved by reducing the DOB design problem to the standard $ {H}_{\mathrm{\infty }} $ mixed sensitivity optimization problem using virtual loop technology. On the other hand, the trumpet tube MPC is designed, and the active set method is used to solve the optimization problem, and the high-precision test mass anti-disturbance control is realized under strong interference and strong execution constraints. Finally, the proposed method is verified by simulation on the full degree of freedom simulation platform of spacecraft-double test masses. Step matching interference and sine matching interference at 0.1 Hz are proposed based on the fundamentals of noise and interference. The results show that the DOB can accurately estimate the disturbance, and the method can realize high-precision control of the test mass under interference. The measurement noise is also inhibited.

Based on the theory of prescribed-time convergence consensus, a cooperative interception guidance law for multiple flight vehicles under regional confinement constraints is designed. In the line-of-sight (LOS) direction, a prescribed-time cooperative guidance law is designed to ensure that the impact time errors and the consensus errors of impact time converge to zero, thereby causing the impact times to tend toward consistency and satisfying the requirement of simultaneous interception at a designated time. In the vertical direction of the LOS, by integrating sliding mode control, a prescribed-time convergent sliding mode surface and a guidance law with LOS angle constraints are designed to drive the LOS angle errors and LOS angular rates of each flight vehicle to zero. This enables multiple flight vehicles to intercept the target at their respective specified LOS angles, thereby meeting the desired LOS angle requirements. The designs along the LOS direction and vertical to the LOS direction enable the prescribed-time convergence cooperative guidance law to simultaneously satisfy the dual constraints of impact time and LOS angle. Theoretical analysis demonstrates that the proposed guidance method ensures multiple flight vehicles intercept the target simultaneously at the desired LOS angles. Simulation results verify the correctness and effectiveness of the proposed method.

This paper proposes a calculation method for the all-domain fire field for the threat assessment of beyond visual range (BVR) air combat situations. To overcome the drawbacks of conventional situation threat assessment techniques, such as high subjectivity and an inability to meet real-time computing requirements, the all-domain fire field calculation is split into offline single aircraft fire field calculation and online aggregation calculation, taking into account the limitations of missile-based computing resources. Firstly, a BVR air combat simulation environment was established, taking into account the detection error of missile seekers and the delay of servo response. Secondly, based on the Monte Carlo method, considering the deviation of pilot behavior, key decision points for maneuver are divided and control variables with normal distribution are introduced to calculate the success rate. Furthermore, based on the independent probability event formula, the single aircraft fire field is aggregated. Finally, calculate the gradient feature representation model of the entire fire field, and design a decision aiding system for one-on-one beyond visual range air combat scenarios. This work can confirm the all-domain firing field’s conceptual design and provide further evidence for the study of decision-assistance system design and threat assessment techniques for BVR air combat situations.

This article proposes a safe landing control strategy for unmanned aerial vehicle (UAVs) by integrating control barrier functions with neural network controllers. Initially, control barrier functions and UAV’s dynamical models are introduced, providing a theoretical foundation for subsequent algorithm design. Then, a control approach is proposed that uses the level set method to design control barrier functions and combine them with neural network controllers to successfully ensure UAV safety during obstacle avoidance and safe landing. Simulation experiments then validate the effectiveness of the proposed control strategy in obstacle avoidance and safe landing, demonstrating the UAV’s safe obstacle avoidance capabilities under limited maneuverability and attitude constraints. The success of the suggested algorithm is finally summed up, and potential research avenues are examined.

For the quadrotor unmanned aerial vehicle (QUAV) attitude tracking problem under external disturbance and model uncertainty, a fixed-time command-filtered control approach is developed based on the composite adaptive radial basis function (RBF) neural network. Firstly, a fixed-time command filter based on the hyperbolic tangent function is proposed, which avoids the differential explosion problem during the derivation of virtual control and eliminates the singularity phenomena of traditional command filters with fractional order effectively. Secondly, the online approximation impact is enhanced by using a RBF neural network to approximate the model uncertainty and designing the adaptive adjustment law of neural network weights based on the tracking deviation. Additionally, combined with the backstepping method and disturbance observer, a fixed-time control strategy for the QUAV system is established, and the external disturbance is estimated and compensated by the disturbance observer, enabling rapid and accurate tracking of desired attitudes. The stability of the proposed control strategy is rigorously proved via Lyapunov theory. Finally, the effectiveness of the control strategy is verified by numerical simulation.

The Hp adaptive pseudo-spectral method is used to optimize and solve the velocity vector control problem. The track coordinate system is used to create a nonlinear aircraft dynamics model. During the modeling, the dynamic responses of the aircraft’s overload, thrust and roll are described in the form of dynamic links, and the angle of attack limitation function in the actual flight control law is realized through path constraints. Different tactical requirements are achieved by setting the control quantity, state quantity, and objective function, and then the pseudo-spectral method is used for optimization and solution. Based on the Hp adaptive pseudo-spectral method, the simulation results show that the velocity vector control optimization method is effective and that it is feasible to handle different restrictions in certain situations.

A lightweight method for detecting and tracking small unmanned aerial vehicle (UAV) targets in complex environments, such as urban and industrial areas, is proposed. Leveraging the CenterNet target detection algorithm as its foundation, this method integrates multi-level feature fusion and a rapid spatial pyramid pooling (SPPF) structure while employing the MobileNet lightweight backbone network to ensure precise detection of small UAV targets. An enhanced DeepSORT-based multi-target tracking technique is presented to overcome the inherent instability in monitoring UAV targets with telescopic cameras. This method utilizes an adaptive noise Kalman filter (NSA Kalman Filter) for target trajectory prediction and incorporates a camera motion compensation module and BYTE target association algorithm to achieve accurate tracking of multiple UAV targets. Furthermore, a dataset for detecting and tracking small UAV targets is constructed, and the proposed algorithm is trained, tested, and validated on the embedded Jetson NX device. Experimental results demonstrate a reduction of 56.9% in average model parameter count, a 1.18% increase in mAP, and a 66.5% reduction in average computational load. With an average model size of 14.5 MB and an average processing time per frame of 36.4 ms on the Jetson NX platform, the algorithm's efficacy in accomplishing accurate identification, real-time operation, and appropriateness for deployment on edge devices with constrained computational resources is confirmed.

In this paper, the cooperative path planning problem in games for the unmanned aerial vehicles system is addressed under conditions of unknown dynamics and input constraints. By planning their routes and avoiding collisions and prohibited areas, friendly and enemy unmanned aerial vehicles must catch up to each other in the game. The trajectory of the opposing unmanned aerial vehicles is predicted to assist path planning by a long short term memory (LSTM) model with an attention mechanism. By creating the value function, the cooperative path planning issue is transformed into an optimum control problem with input restrictions. A method based on integral reinforcement learning is designed to achieve optimal control using the historical data, without the knowledge of inertial parameters. The results of the simulation confirm the efficacy of the proposed method.