The proposed improvement to the Rapidly-exploring Random Tree (RRT) path planning algorithm is aimed at addressing the issue of slow convergence speed caused by boundary information in the original algorithm, by introducing a phase search approach. The initial approach involves employing a three-stage search strategy to generate sampling points that are specifically oriented toward real-time sampling failure rate, thereby significantly reducing the number of redundant nodes. Simultaneously, a balanced exploration strategy is introduced, enhancing the algorithmos convergence speed by constructing two randomly growing trees for searching. Secondly, a path-pruning strategy is implemented, effectively reducing the path length. Finally, the bidirectional exploration technique from the improved algorithm is applied to the traditional RRT algorithm based on boundary information, and comparative experiments are conducted. The experimental results demonstrate that, compared to the traditional boundary-based RRT method, the proposed improved algorithm reduces the running time by 13.4% and decreases the path length by 9.51%.

This paper deals with formation control of multiagent systems (MASs) subject to disturbances and input constraints, and an active disturbance rejection formation control (ADRFC) approach based on distributed observation is proposed. Firstly, the states and disturbances of tracking error systems are estimated by a distributed intermediate estimator (DIE). Secondly, an improved distributed model predictive control (IDMPC) is proposed to achieve formation control of MASs under disturbances and input constraints. Then, the robust ADRFC method is designed by combining the DIE and IDMPC, and the stability of the whole system is analyzed. Finally, experiments are conducted on a MAS consisting of four mobile robots to illustrate the effectiveness and superiority of the proposed method.

Networked microgrids (NMGs) have emerged to tackle the intricate challenge posed by today's increasing power demands. However, the configuration of NMGs is subject to change due to revamping, faults, or privacy issues within each microgrid (MG), and existing control mechanisms cannot realize configurable control with a privacy guarantee. To address this control problem, we offer a fresh perspective on the control framework by proposing a configurable decentralized-distributed control strategy using directed graphs with minimal communication and less privacy disclosure risk. Specifically, one distributed generator (DG) in each MG is designated as the leader to ensure complete privacy protection in a decentralized manner, eliminating the need to exchange information among other NMGs. Follower DGs communicate exclusively with their leader DGs within clusters, enhancing the control performance. Furthermore, by introducing a novel leaky integral control law, the proposed configurable control framework can adapt to different NMG configurations without redesigning controllers. The stability of the entire NMG system is analyzed rigorously. Finally, the effectiveness of the proposed control framework is demonstrated in solving secondary frequency restoration and active power-sharing problems of NMGs. The experimental test system is built on the OPAL-RT system, and the results show the superiority of our proposed control method under decentralized, distributed, and hybrid NMG configurations.

Graph neural networks (GNNs) have achieved remarkable success in various real-world applications. However, recent studies highlight the vulnerability of GNNs to malicious perturbations. Previous adversaries primarily focus on graph modifications or node injections to existing graphs, yielding promising results but with notable limitations. Graph modification attack (GMA) requires manipulation of the original graph, which is often impractical, while graph injection attack (GIA) necessitates training a surrogate model in the black-box setting, leading to significant performance degradation due to divergence between the surrogate architecture and the actual victim model. Furthermore, most methods concentrate on a single attack goal and lack a generalizable adversary to develop distinct attack strategies for diverse goals, thus limiting precise control over victim model behavior in real-world scenarios. To address these issues, we present a gradient-free generalizable adversary that injects a single malicious node to manipulate the classification result of a target node in the black-box evasion setting. Specifically, we model the single-node injection label specificity attack as a Markov decision process (MDP) and propose gradient-free generalizable single node injection attack, namely G2-SNIA, a reinforcement learning framework employing proximal policy optimization (PPO). By directly querying the victim model, G2-SNIA learns patterns from exploration to achieve diverse attack goals with extremely limited attack budgets. Through comprehensive experiments over three acknowledged benchmark datasets and four prominent GNNs in the most challenging and realistic scenario, we demonstrate the superior performance of our proposed G2-SNIA over the existing state-of-the-art baselines. Moreover, by comparing G2-SNIA with multiple white-box evasion baselines, we confirm its capacity to generate solutions comparable to those of the best adversaries.