In this paper, we investigate the Source-Free Domain Adaptation (SFDA), where a well-trained model adapts to an unlabeled target domain without access to source data. Previous SFDA methods mainly relied on convolutional neural networks, which struggle with domain shifts due to their local focus. To address this, we propose the Object-centric Perception Source-Free Transformer (OP-SFT), which leverages the self-attention mechanism of Transformers to focus on relevant target regions, improving adaptability to domain shifts. We also introduce self-supervised knowledge distillation to enhance semantic perception and a confidence-based k-means clustering method for more accurate pseudo-label generation. Extensive experiments demonstrate that our OP-SFT achieves significant adaptation performance across four widely-used domain adaptation benchmark datasets compared to other state-of-the-art baselines. The code is available at https://github.com/Weilong-Gao/OP-SFT.

Multi-source Domain Adaptation (MSDA) aims to adapt models trained on multiple labeled source domains to an unlabeled target domain. Recent MSDA methods based on Generative Adversarial Networks (GANs) implicitly capture the image distribution, which can lead to limited sample fidelity and result in misalignment of pixel-level information between the sources and the target domain. Moreover, when samples from different sources interact during training, significant misalignment across various source domains can occur. In this study, we introduce a novel MSDA framework called Adaptive Causal Diffusion Networks (ACDN) to address these challenges. ACDN integrates a diffusive domain adaptation model for effective, high-fidelity adaptation between the source and target domains, incorporating Granger-causal inference to ensure that the assigned weights for each source domain are closely related to their respective contributions to the decision-making process. Experimental results show that ACDN outperforms existing methods significantly across real-world domain adaptation benchmarks.

Remaining useful life (RUL) prediction plays a pivotal role in prognostics and health management (PHM) systems, which enhances the reliability of operating equipment and reduces maintenance costs. With the advent of Industrial Internet of Things (IIoT) technology, it becomes feasible to obtain preformance-degradation data that precisely mirrors the health status of equipment. This facilitates real-time process monitoring of device status and promotes intelligent predictive maintenance methods, thereby achieving more accurate RUL estimations. Nonetheless, IIoT systems often suffer from sensing or communication failures in practical industrial scenarios, leading to fragmented multivariate time-series data with missing observations from partial edge devices, which severely restricts the performance of RUL prediction. To address this issue, a graph-based representation learning method is skillfully proposed for RUL prediction with missing values. Specifically, the observations and features in multivariate time-series are regarded as two distinct nodes types in a semi-bipartite graph. Furthermore, the observed values are viewed as real edges whereas the dependency between timestamps and the correlation between features as virtual edges. In this scheme, the representation learning for multivariate time-series is expressed as the graph-level tasks. The acquired representations in the observation missing mode possess the capability to extract meaningful information and unveil underlying patterns from fragmented data, which bolsters the performance of RUL estimations through direct imputation or end-to-end architecture. The effectiveness and robustness of the proposed method are validated through the results of comparative experiments under different missing patterns using the C-MAPSS dataset, demonstrating its capability to enhance predictive maintenance in IIoT-enabled systems while reducing unexpected failures and maintenance costs.

Information granularity and information granules are fundamental concepts that permeate the entire area of granular computing. With this regard, the principle of justifiable granularity was proposed by Pedrycz, and subsequently a general two-phase framework of designing information granules based on Fuzzy C-means clustering was successfully developed. This design process leads to information granules that are likely to intersect each other in substantially overlapping clusters, which inevitably leads to some ambiguity and misperception as well as loss of semantic clarity of information granules. This limitation is largely due to imprecise description of boundary-overlapping data in the existing algorithms. To address this issue, the rough k -means clustering is introduced in an innovative way into Pedrycz's two-phase information granulation framework, together with the proposed local boundary fuzzy metric. To further strengthen the characteristics of support and inhibition of boundary-overlapping data, an augmented parametric version of the principle is refined. On this basis, a local boundary fuzzified rough k -means-based information granulation algorithm is developed. In this manner, the generated granules are unique and representative whilst ensuring clearer boundaries. The validity and performance of this algorithm are demonstrated through the results of comparative experiments.

In this paper, we propose a novel framework called Du-al Re-weighting Multi-source Network (DRMN) to address the task of Multi-source Domain Adaptation (MSDA). Two challenges exist in MSDA: i) the domain discrepancies a-mong the multiple source domains, and ii) the domain mis-match between target and source domains. We propose du-al re- weighting mechanisms including source distribution re-weighting and sample selected re-weighting. Source distribution re-weighting mechanism can match the estimated source label distribution and the unknown target label distribution to adapt the classifier. Sample selected re-weighting mechanism can select highly confident target data as pseudo-labeled sam-ples to integrate the information from different sources, and further improve the classification performance. We find that DRMN can show competitive performance with respect to the state-of-the-art on different real-world datasets.

For a class of nontriangular nonlinear systems in presence of unknown disturbances, we propose a predictor-based neural dynamic surface control (PNDSC) strategy in this paper. This nontriangular system is transformed via the mean value theorem, and a predictor is then constructed. To avoid an algebraic loop problem, partial state vectors are employed as input signals of neural networks (NNs) for approximating unknown dynamics, and compensation items are designed to compensate for approximation errors from NNs. Different from the traditional NDSC, the PNDSC in this paper utilizes prediction errors to update learning parameters for improving NNs’ learning behaviors with overlarge adaptive gains. On the basis of improved NNs’ approximation behaviors, a predictor-based NNs disturbance observer (PNNDO) is constructed for compensation for external disturbances and approximation errors from NNs. Furthermore, with predictors, a normalization method of weights is developed to reduce the number of online learning parameters. On the basis of the aforementioned result, measurement noises are taken into account in our predictor-based neural control strategy. We employ predictor states, rather than measurement information paralyzed by noises, in design of our control strategy. This reduces high-frequency oscillations in control input. A Lyapunov-based stability analysis shows that all signals are ultimately bounded in the closed-loop system. Finally, the effectiveness of the proposed control strategy is verified by a numerical example and a permanent magnet brushless DC motor system.

We study Universal Domain Adaptation (UniDA) problem, which is recently proposed. Different from existing domain adaptation (DA) methods, e.g., Closed set, Open set and Partial DA, UniDA does not need any prior knowledge about the overlap across the source and target label sets. We have two challenges in UniDA problem: i) Domain shift. ii) Category shift. Towards tackling above challenges, we formulate a universal adaptation network called Dual Contrastive Network (DCN), where a contrastive module and a transferability rule are included. The experimental results reveal that DCN can work stably on different UniDA settings and exceeds the state-of-the-art performance across five benchmarks against existing DA methods.

In this article, a cyber-physical cooperative control strategy is proposed for islanded microgrid (MG), which divides the MG into cyber and physical layers. And the main designs in these two layers are two event-triggered mechanisms, where one mechanism is used to improve the voltage and frequency stability of MG considering the packet loss problem, the other is used to reduce the communication burden in the control process. More specifically, the control process of the first mechanism can be understood as we use these event-triggered mechanisms to complete the secondary control in the physical layer based on the information in the cyber layer. In this mechanism, the packet loss situation in one communication channel is divided into three categories: 1) to handle the case where the loss rate is small, an adaptive virtual leader-following consensus controller (AVLFCC) is proposed in the cyber layer; 2) to handle the case where the loss rate is large and the forecasted data can be used, a hybrid forecast supplement method (HFSM) is proposed in the physical layer; and 3) to handle the case where the loss rate is large and the forecasted data cannot be used, a path reconstruction method combined with a novel sliding-mode control (SMC) is proposed in the cyber layer. In the second mechanism, an event-triggered protocol is designed for the consensus controller to reduce the communication burden based on the designs in 1)-3). Finally, based on these designs in the two mechanisms, a novel secondary controller is designed. And the experimental results have confirmed the validity of the contributed strategy.

This article proposes a fuzzy k-means cluster based generalized predictive control (GPC) method for a 1000 MW ultra supercritical power plant to improve of the boiler combustion efficiency. First, to fully use the statistic characteristic of the historical data, a fuzzy k-mean cluster network (FKN) is well constructed to derive the local linear models, and the nonlinear dynamic process of studied system is elaborately approximated by the fuzzy combination of the local linear models. Then, a global GPC method is proposed to improve the control performance by using the membership of the current FKN. Different from the traditional GPC, the advantage of proposed GPC is that local GPC is fuzzily combined together to achieve the purpose of global GPC by a scheduling algorithm. Finally, an example illustrates that the proposed control strategy can achieve the satisfactory performance.

This article investigates resilient event-triggered load frequency control (LFC) of multiarea power systems under nonideal network environments. Under the sample-data framework, a novel QoS-dependent event-triggered communication (QEC) scheme is presented to deal with nonideal network environments while preserving the desired control performance and improving the communication efficiency. In comparison with some traditional state-dependent event-triggered communication schemes, since the proposed QEC depends not only on the state of controlled plant but also on the QoS of communication network, higher communication efficiency can be achieved. Then, a resilient LFC is well developed based on the proposed QEC, where “resilient” implies that: 1) for a normal QoS case, less packets are transmitted to save the communication resources and 2) for an abnormal QoS case, more packets are transmitted to mitigate the influence of nonideal QoS. Moreover, the proposed method provides a better way to balance the control performance and communication resource by reasonably choosing the parameters of QEC. Finally, the simulation results show the effectiveness of the proposed method.

Rough K-Means (RKM) is an efficient clustering algorithm for overlapping datasets, and has captured increasing attention in recent years. RKM algorithms are the main focus on the further description of uncertain objects located in boundary regions in order to improve the performance. However, most available RKM algorithms fail to pay attention to the influence of imbalanced clusters, together with imbalanced spatial distributions (i.e., the cluster density) and differing cluster sizes (i.e., the number of object ratios). This paper seeks to address this deficiency and examines in detail some adverse effects caused by imbalanced clusters. To mitigate adverse effects of imbalanced clusters and decrease the computational cost, an interval type-2 fuzzy local measure for the RKM clustering is proposed, on the basis of which, a novel RKM clustering algorithm has been developed that specifically gives due consideration to imbalanced clusters. The effectiveness and superiority of this algorithm are demonstrated through simulation and experimental analysis.

Due to the strong intermittency and volatility and the increasing proportion of photovoltaic (PV) power in the power grid, the PV power prediction becomes more and more important for the reliability of the power grid. Neural network is a popular model that used for PV power prediction. However, traditional neural networks prediction model that relies solely on the offline training cannot adapt well to the dynamic changes of PV power station. To cope with this problem, the very short-term online forecasting model for PV power based on two-stage resource allocation network (RAN) is presented. Firstly, the RAN model is offline trained to determine the initial structure. Thereafter, the initial RAN model is used for online forecasting, in this stage, the forecasting model is further updated. The simulation results show that the two-stage RAN model can effectively improve the forecasting accuracy of the PV power output.

With the rapid development of microgrids, generator excitation control for multi-machine systems to improve the stability of power systems has become a key technical problem. This paper presents an excitation controller design for a typical two-machine system. According to the characteristics of strong nonlinearity, load disturbance and time-varying uncertainty, conventional PID control schemes cannot meet the high quality requirement of excitation control for two- machine systems. A Resource Allocation Network (RAN) based neural PID adaptive generator excitation control is proposed for two-machine systems. The parameters of the PID controller can be adjusted dynamically according to the RAN-enabled online model. The validity of the proposed control strategy is demonstrated by the simulation results.