Real-time traffic forecasting acts as a critical enabling service for IoT-driven Intelligent Transportation Systems (ITS). While existing Spatiotemporal Graph Neural Networks (STGNNs) achieve superior forecasting accuracy, their intensive computational complexity and high latency create a deployment bottleneck for resource-constrained IoT edge devices. To address this resource-accuracy mismatch, we propose a novel framework termed Dynamic Hub-Aware Knowledge Distillation (DHKD). Unlike traditional uniform distillation paradigms, DHKD introduces a topology-aware strategy to transfer knowledge from a complex teacher to a lightweight Spatiotemporal Multi-Layer Perceptron (STMLP) student model. Specifically, we design a dynamic hub-aware gating (DHAG) mechanism that adaptively identifies time-varying pivotal sensing nodes (hubs), ensuring that the student model prioritizes the most information-dense spatiotemporal regions. Furthermore, we develop a multi-level distillation strategy that aligns both intermediate features and final predictions through contrastive learning and attention mechanisms. Extensive experiments conducted on four real-world datasets demonstrate that DHKD significantly reduces inference latency while maintaining state-of-the-art performance, validating its viability for deployment on resource-constrained Internet of Vehicles (IoV) edge devices. The implementation code is available at https://anonymous.4open.science/r/DHKD.

The assignment and scheduling of AAV power inspection tasks constitute a representative Mixed-Integer Nonlinear Programming (MINLP) problem. However, due to the complexity of the workflow, manual methods and existing heuristics struggle to balance time costs and solution quality for large-scale tasks. To address this challenge, this paper introduces a rolling window approach for batch processing and proposes a clustering strategy before scheduling. Building on this strategy, we develop a rolling dispatch algorithm utilizing task-adaptive clustering (RDA-TAC), which alleviates problem complexity and enhances search efficiency. Specifically, we present a novel MINLP model designed to minimize total equivalent endurance losses, incorporating multi-level and sequential inspection constraints derived from actual inspection processes. We then batch and segment large-scale tasks, treating task fragments as the smallest units for assignment, and adaptively adjust them to convert the problem into a single-drone path planning task. Through nine scenarios based on real power grid data, RDA-TAC attains optimal solutions for small-scale tasks and outperforms manual and heuristic methods in more complex cases, demonstrating a tenfold increase in solving speed. Note to Practitioners—Autonomous aerial vehicles (AAVs) are the leading method for power line inspection, offering rapid deployment, cost-effectiveness, and improved safety. However, traditional manual assignment and path planning algorithms often struggle with large-scale and complex tasks. To address this, we propose a strategy that clusters tasks before scheduling, reducing complexity and improving efficiency. Empirical validation demonstrates that our method balances solution time and quality effectively. Comparative results are included in the attached digital format.

To tackle challenges of low profitability and high response delays in multitask competitive production environments with fluctuating capacity, this article studies a collaborative optimization approach of task admission and service scheduling with dynamic pricing aid. First, in the context of platform-aggregated manufacturing service collaboration, we introduce service queues to account for response delays and develop a novel nonlinear profit optimization model. This model optimizes task admission, service scheduling, and pricing decisions simultaneously, to adapt the admitted task load to the fluctuating capacity and enhance the throughput utilities of heterogeneous services. To solve this large-scale, nonlinear optimization problem, we then propose a novel distributed online task admission and service scheduling optimization strategy by constructing a Lyapunov quadratic function. It coordinates the optimal decisions for each service in a computationally efficient manner without requiring prior knowledge of task statistics or data training. Moreover, we analytically illustrate that our approach can achieve the optimal time average profit while bounding time average queue length over temporal fluctuations. Numerical results from real workload traces demonstrate the effectiveness of our approach compared to three existing strategies, offering valuable insights for platform operations.

This article focuses on the distributed adaptive consensus control problem for nonlinear systems with unknown parameters and denial-of-service (DoS) attacks. The communication channels among subsystems are directed and DoS attacks are executed on subsystems to jam the communication transmission. Besides, only part of these subsystems can access states of the desired reference system. To actively alleviate attack effects on the consensus performance, a distributed adaptive consensus control scheme with an active-defense mechanism is proposed. The active-defense mechanism consists of an attack-detection algorithm and a switching strategy. The distributed adaptive consensus controller is designed with normalized damping terms in control inputs and parameter update laws. In the presence of DoS attacks, a stability condition is derived based on the designed control scheme, which guarantees that the consensus errors are globally uniformly bounded under arbitrary switching dwell-time. Experimental results are provided to validate the effectiveness of the proposed control scheme with the active-defense mechanism.

In this article, the distributed adaptive secure formation control problem for mobile robots is considered. The communication channels among robots are undirected and suffer from denial-of-service (DoS) attacks. Only part of the robots can access the reference trajectories. To mitigate attack effects on the formation performance, a switching strategy for communication channels is designed. Besides, an adaptive secure control scheme is proposed with a distributed adaptive trajectory estimator and an adaptive tracking controller for each robot. In the estimator, damping and normalizing terms are introduced to alleviate attack effects on the system performance. Moreover, these terms can also remove the constraints on the lower bounded dwell time of switching topologies. By applying the backstepping technique, an adaptive tracking control scheme is proposed for robots with uncertainties to track estimator states. According to the proposed secure control scheme, a stability condition is provided, such that the boundedness of formation errors can be guaranteed under DoS attacks and arbitrary switching dwell time. Experimental results are given to illustrate the effectiveness of the proposed secure control scheme.

In this article, the synchronization issue of certain edge-colored networks is investigated. We start with a linear subspace determined by the colored edges and show that the invariant kernel of this linear subspace is precisely consists of all the synchronous states. Moreover, this invariant kernel is characterized by a cluster of algebraic equations. Especially, for network described by a polynomial vector field, due to the property of Noetherian rings, this invariant kernel can be determined by a finite number of algebraic equations. Furthermore, the equivalence between network synchronization and the asymptotic behavior of the aforementioned invariant kernel are proved. Based on the colored edges and the invariant kennel, we decompose the original network twice, arriving at a three-layer network: the first two layers, referred to as the external part, can only synchronize to an equilibrium point, while the third layer, known as the internal part, possesses the same colored edges. For this three-layer network, we construct two Lyapunov-type functions for the external part and the internal part, respectively, to establish the synchronization criteria. In particular, our criteria involve linear matrix inequalities and polynomial inequalities of smaller scale, which can be solved by the existing semi-definite programming tools. Finally, this article provides three examples to illustrate the effectiveness and advantages of the theoretical results presented.

In this article, we search for polynomial Lyapunov functions beyond the quadratic form to investigate the synchronization problems of nonlinearly coupled complex networks. First, with a relaxed assumption than the quadratic condition, a synchronization criterion is established for nonlinearly coupled networks with asymmetric coupling matrices. Compared with the existing synchronization criteria, our results are less conservative and have a wider application. Second, the synchronization problem for polynomial networks is characterized as the sum-of-squares (SOS) optimization one. In this way, polynomial Lyapunov functions can be obtained efficiently with SOS programming tools. Furthermore, it is shown that the local synchronization of certain nonpolynomial networks can also be analyzed by using the SOS optimization method through the Taylor series expansion. Finally, three numerical examples are presented to verify the effectiveness and less conservatism of our analytical results.

This article presents a rolling optimization algorithm (ROA) for minimizing total vehicle delay for real-time traffic signal control in urban road networks, where the vehicle delay minimization problem is formulated as a quadratically constrained quadratic programming problem, which is NP-hard. The programming problem is relaxed and resolved using the ROA in a distributed manner. In particular, we introduce network partition to a large-scale urban road network and then present a regional ROA to eliminate the low efficiency caused by the large number of decision variables. Numerical experiments are performed on one of Beijing’s district road networks to validate the efficiency of the proposed method in benchmark against several typical techniques.

This article investigates the consensus tracking problem of the heterogeneous multivehicle systems (MVSs) under a repeatable control environment. First, a unified iterative learning control (ILC) algorithm is presented for all autonomous vehicles, each of which is governed by both discrete- and continuous-time nonlinear dynamics. Then, several consensus criteria for MVSs with switching topology and external disturbances are established based on our proposed distributed ILC protocols. For discrete-time systems, all vehicles can perfectly track to the common reference trajectory over a specified finite time interval, and the corresponding digraphs may not have spanning trees. Existing approaches dealing with the continuous-time systems generally require that all vehicles have strictly identical initial conditions, being too ideal in practice. We relax this unpractical assumption and propose an extra distributed initial state learning protocol such that vehicles can take different initial states, leading to the fact that the finite time tracking is achieved ultimately regardless of the initial errors. Finally, a numerical example demonstrates the effectiveness of our theoretical results.

In this paper, the synchronization problem of complex networks with linearly diffusively coupled nonidentical nodes is investigated. Starting with the boundedness condition of network trajectories, we introduce an invariant set such that it contains all limit points of ultimately synchronous trajectories. Then, we develop a decomposition technique for the heterogeneous network. With this decomposition, the synchronization of the network can be investigated by the convergence of one decomposed network and the synchronization of the other decomposed homogeneous-like network. Moreover, for a particular case that the invariant set is a linear subspace, conditional synchronization analysis is provided to reduce the coupling complexity between the two decomposed networks. It is noted that our decomposition technique is quite simple yet general: by this technique, the synchronization of various heterogeneous complex networks can be transformed into the stability of nonlinear systems and synchronization of homogeneous-like complex networks. Finally, we present several numerical examples to demonstrate the effectiveness of the theoretical results. In particular, we use an example to show that our theoretical procedure is also feasible for some heterogeneous networks with a general invariant submanifold instead of linear subspace.

Nowadays, production-oriented manufacturing is transforming to service-oriented one, which leads to an increasing demand for high reliability of manufacturing systems. However, a popular approach to measuring terminal reliability of complex networked systems is based on graph theory, which has been shown to be an NP-hard problem. Though the NP-hard problem can be avoided by employing a statistical measure method for terminal reliability of random networks based on percolation theory, there is still lack of a general assessment approach to calculating terminal reliability of heterogeneous complex networks in intelligent manufacturing. In this paper, we propose a novel fractal-based approach to measuring the terminal reliability of heterogeneous networks. With help of the renormalization procedure that coarse grains a network into boxes containing nodes within given lateral size and inverse renormalization, which gives a fractal network growth model, a fractal network approximation of an arbitrary complex network is obtained. This fractal network topology can be described by a superposition of fractal elements based on fractal theory. Following this description, terminal reliability is the function of reliability of fractal elements. Then, a reliability assessment algorithm with computational complexity O(N2) based on fractal elements is developed. Numerical simulation is performed on a real network and a fractal network to validate the effectiveness of our method.

An increasing number of cities are implementing bicycle sharing systems to reduce traffic congestion. Determining the locations of bicycle stations is one of the fundamental challenges in planning of such systems. This paper provides a novel solution on it. The min-plus algebra is introduced to model transport systems for accessibility analysis. A unified model in the sense of the min-plus algebra for an integrated system with both buses and bicycles is presented to dynamically describe the state transitions of passengers in the system. A new accessibility is proposed with regards to a general index ω such as the geographical distance and the travel time. A necessary and sufficient condition on the accessibility is then provided. The minimization of bicycle stations under the accessibility is formulated to be a 0-1 integer programming problem. The case studies on two cities, Ningbo and Hangzhou, were performed, which show that compared with the current layouts of urban transportation networks, the proposed public bicycle sharing systems have remarkable advantages in topological characteristics and robustness against failures.

This paper investigates the local consensus issue for multiagent systems with nonlinear dynamics and time-varying delays. By defining a weighted average state of the agents and applying local linearization, we show that the local consensus of the agents in a directed communication network can be guaranteed by the asymptotic stability of several decoupled delayed systems. Then, by employing a novel Lyapunov-Krasovskii functional, proposing a new extended reciprocally convex approach and using some matrix analysis, we derive a sufficient condition for the local consensus in terms of linear matrix inequalities associated with the dynamics of the agents, the eigenvalues of Laplacian matrix, and the time-varying delay. Finally, two numerical examples are provided to show the effectiveness of the analytical results.

The outlier detection problem for dynamic systems is formulated as a matrix decomposition problem with low rank and sparse matrices, and further recast as a semidefinite programming problem. A fast algorithm is presented to solve the resulting problem while keeping the solution matrix structure and it can greatly reduce the computational cost over the standard interior-point method. The computational burden is further reduced by proper construction of subsets of the raw data without violating low-rank property of the involved matrix. The proposed method can make exact detection of outliers in case of no or little noise in output observations. In case of significant noise, a novel approach based on under-sampling with averaging is developed to denoise while retaining the saliency of outliers, and so-filtered data enables successful outlier detection with the proposed method while the existing filtering methods fail. Use of recovered "clean" data from the proposed method can give much better parameter estimation compared with that based on the raw data.

Number estimation of controllers is a fundamental question in pinning synchronization of complex networks. This paper studies the problem of controller number in synchronizing a complex network of coupled dynamical systems by means of pinning. For a complex network with a symmetric coupling matrix and full coupling between the nodes, we formulate network synchronization via pinning as a linear matrix inequality criterion, and provide a lower bound and an upper bound of the controller number for a given complex network with fixed architecture. Several numerical examples with Barabási-Albert network topologies are provided to verify our theoretical results.