This paper deals with the problem of dynamic trajectory planning for a group of unmanned aerial vehicles (UAVs) in unknown environments. The existing methods often suffer from excessive computational burden, which creates the gap between theoretical approaches and practical swarm deployment. To overcome these limitations, this paper proposes a distributed cooperative planning system (DCPS). The system consists of two levels: a trajectory planning level and a cooperative trajectory planning level. On the trajectory planning level, a kinodynamic Gaussian potential B-spline (KGPB) approach is designed by combining the local kinodynamic-A-star method and the Gaussian potential field B-spline method. Specifically, in the front-end, a trajectory is first generated by using the local kinodynamic-A-star method based on kinematics-dynamics constraints. And then, in the back-end, the trajectory is further optimized by using the Gaussian potential B-spline (GPB) method. On the cooperative trajectory planning level, a back field neighbor replanning (BFNP) approach is proposed, where each UAV only needs to communicate with the nearest neighbors. According to the potential collision region of the front UAV, the trajectory of the current UAV is dynamically improved to ensure safe and stable flight such that communication costs are significantly improved. Finally, the simulation results demonstrate that the proposed DCPS achieves at least a 27% increase in average velocity and a 24% reduction in traversal time for the UAV swarm compared to prior methods. The experimental outcomes provide further validation that the proposed DCPS can generate efficient and safe trajectories. For particular cases, the UAV operates at 97% of its maximum possible velocity. The proposed DCPS provides a reliable solution for a group of unmanned aerial vehicles in unknown environments.

In this paper, we propose a multi-agent reinforcement learning approach to address the design problem of control strategy for multi-agent consensus. Distributed control policy is learned automatically through reinforcement learning algorithm, thereby circumventing the intricate process of controller design. To be specific, the proposed approach exhibits notable consistency performance in scenarios wherein the agents’ initial positions differs from that in the training environment. We confirm the effectiveness of our approach by using several initial positions of agents and different dimensionality of positions in the simulations.

In this paper, two technologies are proposed to deal with the problem of flight safty of multiple unmanned aerial vehicles (UAVs) in unknown environments. One technology is to optimize the front-end path generated by traditional path planning methods in order to better match the dynamics of UAVs to obtain the back-end movement trajectories of UAVs. The other technology is to introduce the collision detection adjustment region such that collision avoidance can be realized for multiple UAVs by dynamic replanning of UAV’s trajectory under local neighborhood communication. Finally, according to simulation and real-world experimental results, the effectiveness of the proposed technologies is verified for the flight safty of multiple UAVs in unknown environments.

This article deals with the problem of target tracking and detecting in unknown environments by designing two new algorithms for an autonomous aerial vehicle (AAV). First, an auto-Gaussian-GRU-predictive (AGUP) algorithm is designed to solve the tracking problem of a dynamic target in unknown environments. By integrating Gaussian process regression and gated recurrent unit neural networks, the AGUP algorithm can predict the motion trajectory of a dynamic target. Second, a Tabu search interpolated B-spline (TBL) algorithm is also proposed to solve the problem of optimal path planning for multiple stationary targets. The TBL algorithm can efficiently plan the visiting paths and also can enable the path smooth. Third, both AGUP and TBL algorithms are combined with the model predictive control (MPC) approach in order to guide AAVs to track and detect the targets. Finally, simulation and experimental results show that the AGUP-MPC algorithm exhibits excellent tracking capability. In addition, the TBL-MPC algorithm effectively plans the optimal and smooth detection path and controls AAVs to orderly visit multiple stationary targets.

In this paper, the real-time motion planning framework for unmanned aerial vehicle (UAV) is proposed to solve the dynamic target tracking problem in complex environments. The framework is applicable to 3D path planning and trajectory optimization of UAV, which can effectively reduce the unsmoothness during UAV flight and achieve real-time tracking of dynamic targets in 3D space. The framework is divided into two parts: the front-end uses a graph search-based method to find the shortest path for the UAV to approach the dynamic target, while the back-end uses the GP (Gaussian Potential)-B-Spline soft-constrained trajectory optimization method to optimize the shortest path of the front-end, and design the trajectory that conforms to the motion for the UAV. The results show that the method exhibits excellent tracking performance in complex environments and has a wide potential for practical applications.

This paper mainly considers the navigation problem of unmanned aerial vehicle (UAV) in an unknown environment. Traditional path planning method relies on accurate model parameters and environment maps, which has poor adaptability. Therefore, this paper adopts the deep reinforcement learning algorithm to accomplish the navigation task. The classical proximal policy optimization (PPO) algorithm lacks the perception of the correlation between UAV action and state makes difficult for the UAV to choose the optimal path, thus affecting the success rate and speed of navigation. To solve this problem, this paper adds a long short-term memory (LSTM) network to the policy and evaluation network of the PPO algorithm so that the UAV can refer to the preceding status and action information during path planning. The method is extended to three-dimensional motion space. Simulation results demonstrate that the LSTM-PPO algorithm designed in this paper can complete navigation tasks in unknown environments, and show stability in continuous state space and continuous action space. Meanwhile, compared with the PPO algorithm, the success rate of navigation and average arrival time is significantly improved.

Betweenness centrality (BC) is an important metrics in graph analysis which indicates critical vertices in large-scale networks based on shortest path enumeration. Typically, a BC algorithm constructs a shortest-path DAG for each vertex to calculate its BC score. However, for emerging real-world graphs, even the state-of-the-art BC algorithm will introduce a number of redundancies, as suggested by the existence of articulation points. Articulation points imply some common sub-DAGs in the DAGs for different vertices, but existing algorithms do not leverage such information and miss the optimization opportunity.

This paper presents a novel auction-based structure to multi-robot coalition formation problem. The structure, which is called multi-robot Coalition Structure Generation based on Credit Mechanism (CoSGCrM), contains a sub-optimal coalition member selection algorithm with an analysis of its soundness and completeness. A credit mechanism is introduced to reduce the complexity for the coalition leader in making a decision as well as to restrict the profit-oriented robot member in bidding for coalitions. Simulations are given to compare with first-price auction algorithm and the results show the viability of the proposed structure in both simple and complex tasks environments.

The number of malicious applications (apps) targeting the Android system has exploded in recent years. The evolution of malware makes it difficult to detect for static analysis tools. Various behavior-based malware detection techniques to mitigate this problem have been proposed. The drawbacks of the existing approaches are: the behavior features extracted from a single source lead to the low detection accuracy and the detection process is too complex. Especially it is unsuitable for smart phones with limited computing power. In this paper, we extract sensitive behavior features from three sources: API calls, native code dynamic execution, and system calls. We propose a sensitive behavior feature vector for representation multi-source behavior features uniformly. Our sensitive behavior representation is able to automatically describe the low-level OS-specific behaviors and high-level application-specific behaviors of an Android malware. Based on the unified behavior feature representation, w e provide a light weight decision function to differentiate a given application benign or malicious. We tested the effectiveness of our approach against real malware and the results of our experiments show that its detection accuracy up to 96% with acceptable performance overhead. For a given threshold t (t=9), we can detect the advanced malware family effectively.

This paper addresses a box transport problem that requires the cooperation of multiple mobile robots. A geometric-based distributed formation control strategy is proposed for robots to push the box to the target, which might be static or dynamic. Velocity and hardware constraints are considered in the advanced planning of the trajectory. Information sharing is included because the robots used as box pushers cannot acquire the required environmental information from their local sensors. Simulation results show the effectiveness of the proposed distributed cooperation strategy.