This brief introduces a fully distributed fault-tolerant control scheme based on a dual-channel dynamic event-triggered mechanism (DETM) for multi-agent systems (MASs) affected by both external disturbances and actuator faults, aiming to achieve leader-following bipartite consensus. To reconstruct unknown system states, external disturbances, and actuator faults, a series of observers is developed utilizing only local output measurements. Based on these estimates, a fully distributed fault-tolerant control law is formulated to ensure leader-following bipartite consensus, while simultaneously enhancing robustness against disturbances and resilience to faults. To alleviate communication burdens, dual-channel DETMs are implemented on both the sensor-to-observer (S-O) and observer-to-controller (O-C) channels, where internal dynamic variables are employed to adaptively adjust the triggering thresholds. The efficacy of the proposed method is validated via simulations studies on robotic manipulator systems.

For interconnected nonlinear systems, the global continuous-time/event-triggered decentralized output feedback prescribed-time (PT) stabilization problem is discussed in this paper. A strategy employing dual time-varying gains with distinct properties is introduced to facilitate the design of the observer and controller. Using these gains within a nonrecursive design approach, a continuous-time decentralized output feedback PT controller is developed to achieve PT stabilization. Further, an event-triggered decentralized output feedback PT controller is constructed to reduce communication resources while fulfilling the control objective. A numerical example and a hardware-in-the- loop simulation of a 2-degree-of-freedom (DOF) helicopter system are employed to illustrate the availability of the decentralized output feedback PT controller.

MOTIVATION: Accurate prediction of protein-ligand binding (PLB) relationships plays a crucial role in drug discovery, which helps identify drugs that modulate the activity of specific targets. Traditional biological assays for measuring PLB relationships are time consuming and costly. In addition, models for predicting PLB relationships have been developed and widely used in drug discovery tasks. However, learning more accurate PLB representations is essential to meet the stringent standards required for drug discovery. RESULTS: We propose an image-based PLB representation learning framework, called ImagePLB, which equips ligand representation learner (LRL) and protein representation learner (PRL) to accept 3D multi-view ligand images and protein graphs as input, respectively, and learns rich interaction information between ligand and protein through a binding representation learner (BRL). Considering the scarcity of protein-ligand pairs, we further propose a multi-level next trajectory prediction (MLNTP) task to pre-train ImagePLB on the 4D flexible dynamics trajectory of 16 972 complexes, including ligand level, protein level, and complex level, to learn information related to trajectories. Besides, by introducing trajectory regularization (TR), we effectively alleviate the problem of high (even almost identical) feature similarity caused by adjacent trajectories. Compared with the current state-of-the-art methods, ImagePLB has achieved competitive improvements on PLB-related prediction tasks, including protein-ligand affinity and efficacy prediction tasks. This study opens the door to the image-based PLB learning paradigm. AVAILABILITY AND IMPLEMENTATION: All data and implementation details of code can be obtained from https://github.com/HongxinXiang/ImagePLB.

The epidemic of infectious diseases has a significant impact on society, the economy, and people’s lives. Social media, with its high user participation and rapid information dissemination, plays a crucial role in shaping public opinion. Fine-grained sentiment analysis of public opinion on infectious diseases can provide valuable insights for improving the quality of public services. However, there are few relevant studies on Chinese data due to language complexity and low resources. Moreover, most of the existing approaches utilize the Graph Neural Network (GCN) method by syntactic dependency trees to construct graphs of text, which ignore the potential link relationships between aspects and words. Therefore, to address this limitation, in this article, we propose a new method based on GCN using aspect-specific heterogeneous graphs, named ASHGCN, which combines BiLSTM, heterogeneous graphs, GCN, the mask and the attention mechanism. We mine social media posts related to COVID-19 for aspect-based sentiment analysis task (ABSA) for ten aspect entity types in both Chinese and English data. The heterogeneous graph is designed with two node types (aspect nodes and non-aspect nodes) and four edge connection types, including various relationships between aspect entities, and between aspect entities and non-aspect entities. In addition, we release a Chinese dataset and an English dataset that include medical and named entities, along with corresponding sentiment labels. Experiments on our datasets, as well as two public datasets, demonstrate that our method greatly improves performance in the ABSA task. Ablation experiments and case studies further support the effectiveness of the proposed approach.

The issue of adaptive dynamic event-triggered asymptotic tracking control for strict-feedback nonlinear systems with unknown functions and full-state prescribed-performance constraints is discussed in this paper. A novel self-adjusting performance function (SAPF) is constructed by merging a continuous function with a finite-time performance function. By associating a transformation function and SAPF, the state constraints problem is recast into analyzing the boundedness of the new variables. Moreover, a new dual dynamic variable-dependent event-triggered mechanism is provided to reduce redundant data transmission. By using a command-filter technique and neural network method, a control scheme is proposed to guarantee the system output asymptotically tracks the reference signal and all system states satisfy specified constraints. Lastly, an applied example is introduced to illustrate the effectiveness of the proposed scheme. Note to Practitioners—In reality, many practical systems, such as flexible manipulators and unmanned underwater vehicles, need to operate in a constrained region. Motivated by this, we design controller for nonlinear systems to guarantee steady-state and transient performance. To reduce the conservatism, a novel SAPF is constructed and then a prescribed performance control strategy is proposed, which can not only achieve the asymptotical tracking of the nonlinear system, but also ensure the constraint performance. In addition, to decrease the transmitted data via communication network, a new dual dynamic variable-dependent event-triggered mechanism is proposed. This method has been illustrated to be feasible via a simulation example.

This study investigates bipartite containment control algorithm for multi-agent systems which are vulnerable to DoS attacks and external disturbances. Disturbances are generated via a series of exogenous nonlinear systems with a one-sided Lipschitz condition. To estimate unknown system states and external disturbances, a set of decentralized state observers and disturbance observers is proposed. Then a fully distributed control protocol is adopted to avoid the utilization of global information. An attack compensator is introduced to mitigate the adverse impacts of DoS attacks. Furthermore, Zeno-free dynamic event-triggered mechanisms that do not require continuous communication between neighboring agents are presented to conserve limited communication resources. In the end, satellite flight systems are introduced to illustrate the feasibility of the designed control scheme.

In this study, the problem of mean square exponential $l_{2}-l_{\infty }$ control is investigated for switched-Markovian jump systems (SMJSs). SMJSs are subject to deterministic switching obeying mode-dependent average dwell time (MDADT) and stochastic switching complying to Markov chain. mode-dependent transition probability (MDTP) and edge-dependent transition probability (EDTP) are proposed. MDTP describes the transition probability (TP) of Markovian jump systems (MJSs) under deterministic switching, and EDTP portrays the TP among MJSs affected by deterministic switching, which is associated with two different deterministic switching modes. Using multiple discontinuous Lyapunov function technology, the mean square exponential stability with $l_{2}-l_{\infty }$ performance is guaranteed by the MDADT method, MDTP and EDTP. Certain solvable sufficient conditions are obtained for the controller. Finally, two numerical examples and a practical example are provided to illustrate the validity of the obtained results.

This study discusses the problem of event-triggered (ET) asymptotic tracking control for parametric strict feedback nonlinear systems (SFNSs) with time-varying disturbances and unknown control directions. A unified dynamic threshold method is proposed by combining a unified function with a self-adjustable performance function. In contrast to previous research, the results of this study provide a unified framework in which global or semi-global performance can be achieved through simple parameter selection while excluding the conservativeness of the constraint thresholds owing to the artificial selection of a uniform performance function. The basic lemma based on the Nussbaum function frequently is extended to adapt to the case in which the coefficients are multiple bounded functions. By fusing a first-order differentiator, the reduplicative derivation of the virtual controller in the backstepping process is obviated. Moreover, an ET mechanism with two dynamic variables is constructed to reduce the burden of data transmission. The developed controller can guarantee the boundedness of all signals in the closed-loop system, full-state constraints performance, and asymptotic tracking control performance. Finally, the feasibility of the proposed scheme is attested by two examples.

The molecular property prediction (MPP) plays a crucial role in the drug discovery process, providing valuable insights for molecule evaluation and screening. Although deep learning has achieved numerous advances in this area, its success often depends on the availability of substantial labeled data. The few-shot MPP is a more challenging scenario, which aims to identify unseen property with only few available molecules. In this paper, we propose an attribute-guided prototype network (APN) to address the challenge. APN first introduces an molecular attribute extractor, which can not only extract three different types of fingerprint attributes (single fingerprint attributes, dual fingerprint attributes, triplet fingerprint attributes) by considering seven circular-based, five path-based, and two substructure-based fingerprints, but also automatically extract deep attributes from self-supervised learning methods. Furthermore, APN designs the Attribute-Guided Dual-channel Attention module to learn the relationship between the molecular graphs and attributes and refine the local and global representation of the molecules. Compared with existing works, APN leverages high-level human-defined attributes and helps the model to explicitly generalize knowledge in molecular graphs. Experiments on benchmark datasets show that APN can achieve state-of-the-art performance in most cases and demonstrate that the attributes are effective for improving few-shot MPP performance. In addition, the strong generalization ability of APN is verified by conducting experiments on data from different domains.

This paper focuses on addressing the problems of finite-time boundedness and guaranteed cost control in switched systems under asynchronous switching. To reduce redundant information transmission and alleviate data congestion of sensor nodes, two schemes are proposed: the event-triggered scheme (ETS) and the round-robin protocol (RRP). These schemes are designed to ensure that the system exhibits good dynamic characteristics while reducing communication resources. In the field of finite-time control, a switching signal is designed using the admissible edge-dependent average dwell time (AED-ADT) method. This method involves a slow AED-ADT switching and a fast AED-ADT switching, which are respectively suitable for finite-time stable and finite-time unstable situations of the controlled system within the asynchronous switching interval. By constructing a double-mode dependent Lyapunov function, the finite-time bounded criterion and the controller gain of the switched systems are obtained. Finally, the validity of the proposed results is showcased by implementing a buck-boost voltage circuit model.

A decentralized dynamic event-triggered output feedback adaptive fixed-time (DDETOFAFxT) funnel controller is described for a class of interconnected nonlinear systems (INSs). A novel dynamic event-triggered mechanism is designed, which includes a triggering control input, fixed threshold, decreasing function of tracking error, and a dynamic variable. To obtain the unknown states, a decentralized linear filter is designed. By introducing a prescribed funnel and using an adding a power integrator technique and a neural network method, a DDETOFAFxT funnel controller is designed to obtain better tracking performance and effectively alleviate the computational burden. Furthermore, it is ensured that the tracking error falls into a preset performance funnel. A simulation example is presented to demonstrate the availability of the designed control scheme.

In this article, a dynamic event-triggered adaptive neural network decentralized nonrecursive output-feedback control scheme for nonlinear interconnected systems under hybrid cyber attacks is first proposed, where the hybrid cyber attacks, containing deception attacks and denial-of-service (DoS) attacks, obey Bernoulli distribution. Taking into account the restrictions imposed by network transmission, a dynamic event-triggered mechanism is introduced to fulfill the analysis and design task. A decentralized linear observer is established to estimate the unknown states. Meanwhile, a decentralized adaptive output-feedback controller is developed in a nonrecursive method with the help of neural network technology. The proposed control scheme can ensure that all signals in the closed-loop system are bounded. Furthermore, Zeno phenomenon can be effectively avoided. Finally, simulation results validate the feasibility of the proposed control scheme.

Graph Convolutional Network (GCN) is an important method for learning graph representations of nodes. For large-scale graphs, the GCN could meet with the neighborhood expansion phenomenon, which makes the model complexity high and the training time long. An efficient solution is to adopt graph sampling techniques, such as node sampling and random walk sampling. However, the existing sampling methods still suffer from aggregating too many neighbor nodes and ignoring node feature information. Therefore, in this paper, we propose a new subgraph sampling method, namely, Similarity-Aware Random Walk (SARW), for GCN with large-scale graphs. A novel similarity index between two adjacent nodes is proposed, describing the relationship of nodes with their neighbors. Then, we design a sampling probability expression between adjacent nodes using node feature information, degree information, neighbor set information, etc. Moreover, we prove the unbiasedness of the SARW-based GCN model for node representations. The simplified version of SARW (SSARW) has a much smaller variance, which indicates the effectiveness of our subgraph sampling method in large-scale graphs for GCN learning. Experiments on six datasets show our method achieves superior performance over the state-of-the-art graph sampling approaches for the large-scale graph node classification task.

The problem of stabilizing switched linear systems under asynchronous switching is addressed. The admissible edge-dependent average dwell time method is applied to design a switching signal that comprises slow admissible edge-dependent average dwell time and fast admissible edge-dependent average dwell time. Under this switching signal, the restriction that the maximum delay of asynchronous switching is known in advance is removed. The constructed Lyapunov function is associated with both the system mode and controller mode. The stabilization criteria and the corresponding algorithm are presented to obtain the controller gains and to design the switching signal. Finally, two examples are given to demonstrate the effectiveness of the proposed results.

This paper combines the traditional output regulation (OR) theory with fuzzy identification method, and designs a distributed output regulator for the multi-agent system (MAS). In view of the situation that the follower model in the MAS cannot be obtained accurately, the fuzzy model is established to make it approximate to the considered nonlinear discrete follower agent system first. By employing fuzzy identification method and OR theory, a distributed controller is proposed, make follower agents track the reference signal given by the dynamic leader. Two numerical examples demonstrate the obtained results.

This study investigates the problem of adaptive decentralized tracking control for a class of interconnected nonlinear systems with input quantization, unknown function, and time-delay, where the time-delay and interconnection terms are supposed to be bounded by some completely unknown functions. An adaptive decentralized tracking controller is constructed via the backstepping method and neural network technique, where a sliding-mode differentiator is presented to estimate the derivative of the virtual control law and reduce the complexity of the control scheme. On the basis of Lyapunov analysis scheme and graph theory, all the signals of the closed-loop system are uniformly ultimately bounded. Finally, an application example of an inverted pendulum system is given to demonstrate the effectiveness of the developed methods.

Literature analysis facilitates researchers better understanding the development of science and technology. The conventional literature analysis focuses on the topics, authors, abstracts, keywords, references, etc., and rarely pays attention to the content of papers. In the field of machine learning, the involved methods (M) and datasets (D) are key information in papers. The extraction and mining of M and D are useful for discipline analysis and algorithm recommendation. In this paper, we propose a novel entity recognition model, called MDER, and constructe datasets from the papers of the PAKDD conferences (2009-2019). Some preliminary experiments are conducted to assess the extraction performance and the mining results are visualized.

In this paper, a dynamic output feedback tracking controller is studied for a class of discrete-time system with multiple disturbances. A composite anti-disturbance tracking controller is proposed based on disturbance observer and H∞control schemes, which can guarantee system have a good anti-disturbance performance. A solvable sufficient condition is given via linear matrix inequality. Finally, a numerical example is employed to demonstrate the effectiveness of the proposed control approach.