A small object detection method based on the combination of global and local attention mechanism is proposed in this paper to detect small objects distributed in images. object detection model based on local attention...
详细信息
ISBN:
(纸本)9798350321050
A small object detection method based on the combination of global and local attention mechanism is proposed in this paper to detect small objects distributed in images. object detection model based on local attention mechanism has good detection accuracy and speed. However, its performance will be reduced due to the smaller size of the object, especially in the case of missed detection and false detection, and the proposed Global Local Detection Model (GLD) can solve this problem. Specifically, a model solution of the Global and Local Combined Attention Block (GL-CAB) combining deep global features and shallow local features of the network is proposed to solve the problem of small object missed detection. On the one hand, the model focuses on small object s in the local and global ranges, and on the other hand, it supplements the small object information lost during the down-sampling process. Aiming at the situation of pseudo-information generated by small object feature fusion, a multi-branch feature pyramid network (MB-FPN) is proposed. Multi-input is used to form multi-scale feature maps for multi-feature fusion on different branches, which reduces the formation of pseudo-information and enhances the extraction of detailed features of small object by the network. Then, the AU-AIR and VOC2007 datasets are selected for experimental training, and the object detection evaluation indicators (AP, AR, F1, mAP, and FPS) are introduced for comparative analysis. Finally, the simulation results show that the proposed method has better performance to solve the problem of missed detection and false detection of small object.
With the help of neural network-based representation learning, significant progress has been recently made in data-driven online dynamic stability assessment (DSA) of complex electric power systems. However, without s...
详细信息
With the help of neural network-based representation learning, significant progress has been recently made in data-driven online dynamic stability assessment (DSA) of complex electric power systems. However, without sufficient attention to diverse data loss conditions in practice, the existing data-driven DSA solutions' performance could be largely degraded due to practical defective input data. To address this problem, this work develops a robust representation learning approach to enhance DSA performance against multiple input data loss conditions in practice. Specifically, focusing on the short-term voltage stability (SVS) issue, an ensemble representation learning scheme (ERLS) is carefully designed to achieve data loss-tolerant online SVS assessment: 1) based on an efficient data masking technique, various missing data conditions are handled and augmented in a unified manner for lossy learningdataset preparation;2) the emerging spatial-temporal graph convolutional network (STGCN) is leveraged to derive multiple diversified base learners with strong capability in SVS feature learning and representation;and 3) with massive SVS scenarios deeply grouped into a number of clusters, these STGCN-enabled base learners are distinctly assembled for each cluster via multilinear regression (MLR) to realize ensemble SVS assessment. Such a divide-and-conquer ensemble strategy results in highly robust SVS assessment performance when faced with various severe data loss conditions. Numerical tests on the benchmark Nordic test system illustrate the efficacy of the proposed approach.
Stacked autoencoder (SAE) has been applied to the prediction of manufacturing process data quality owing to the excellent feature selection capabilities. Nevertheless, a challenge arises in the accumulation of losses ...
详细信息
This literature review extends and contributes to research on the development of data-driven optimal control. Previous reviews have documented the development of model-based and data-drivencontrol in isolation and ha...
详细信息
This literature review extends and contributes to research on the development of data-driven optimal control. Previous reviews have documented the development of model-based and data-drivencontrol in isolation and have not critically reviewed reinforcement learning approaches for adaptive data-driven optimal control frameworks. The presented review discusses the development of model-based to model-free adaptive controllers, highlighting the use of data in control frameworks. In data-drivencontrol frameworks, reinforcement learning methods may be used to derive the optimal policy for dynamical systems. Attractive characteristics of these methods include not requiring a mathematical model of complex systems, their inherent adaptive control capabilities, being an unsupervised learning technique and their decision-making abilities, which are both an advantage and motivation behind this approach. This review considers previous reviews on these topics, including recent work on data-drivencontrol methods. In addition, this review shows the use of data to derive system dynamics, determine the control policy using feedback information, and tune fixed controllers. Furthermore, the review summarises various data-driven methods and their corresponding characteristics. Finally, the review provides a taxonomy, a timeline and a concise narrative of the development of model-based to model-free data-driven adaptive control and underlines the limitations of these techniques due to the lack of theoretical analysis. Areas of further work include theoretical analysis on stability and robustness for data-drivencontrolsystems, explainability of black-box policy learning techniques and an evaluation of the impact of the extension of system simulators to include digital twins.
learning Model Predictive control (LMPC) is a data-driven approach to MPC that enhances closed-loop performance by leveraging data from successive task iterations to approximate the solution of optimal control problem...
详细信息
learning Model Predictive control (LMPC) is a data-driven approach to MPC that enhances closed-loop performance by leveraging data from successive task iterations to approximate the solution of optimal control problems. The value function in LMPC is pivotal for performance enhancement, but its discrete nature-where each point corresponds to a data point-renders the LMPC problem computationally intensive due to its mixed-integer nature. This letter introduces a novel method to construct the LMPC value function. The proposed value function is a piecewise affine approximation that interpolates the discrete data points of the original value function, resulting in a nonlinear relaxation of the mixed-integer LMPC problem. By connecting the discrete data points with piecewise affine segments, the essential characteristics of the original value function are preserved. The proposed algorithm's effectiveness is demonstrated through numerical simulations in autonomous racing.
In the field of visual Simultaneous Localization and Mapping (SLAM), dynamic environment poses a significant challenge to the accuracy and robustness of systems, especially for monocular visual systems. Existing monoc...
详细信息
Aiming at the localization problem of robots in indoor environment with less characteristic information (such as long corridors and foyers), a localization system based on natural vision labels is designed to achieve ...
详细信息
This article studies the robust bipartite consensus problems for heterogeneous nonlinear nonaffine discrete-time multiagent systems (MASs) with fixed and switching topologies against data dropout and unknown disturban...
详细信息
This article studies the robust bipartite consensus problems for heterogeneous nonlinear nonaffine discrete-time multiagent systems (MASs) with fixed and switching topologies against data dropout and unknown disturbances. The controlled system's virtual linear data model is first developed by employing the pseudopartial derivative technique, and a distributed combined measurement error function is established utilizing a signed graph theory. Then, an input gain compensation scheme is formulated to mitigate the effects of data dropout in both feedback and forward channels. Moreover, a data-drivenlearning-based robust bipartite consensus control (LRBCC) scheme based on a radial basis function neural network observer is proposed to estimate the unknown disturbance, using the online input/output data without requiring any information on the mathematical dynamics. The stability analysis of the proposed LRBCC approach is given. Finally, simulation and hardware testing results further demonstrate the designed method's correctness and effectiveness.
We consider a microgrid system with sparse structure, where the security is susceptible to the impact of potential faults. To enhance the reliability and scalability of the microgrid system, this paper investigates th...
详细信息
data-driven approaches are promising to address the modeling issues of modern power electronics-based power systems, due to the black-box feature. Frequency-domain analysis has been applied to address the emerging sma...
详细信息
data-driven approaches are promising to address the modeling issues of modern power electronics-based power systems, due to the black-box feature. Frequency-domain analysis has been applied to address the emerging small-signal oscillation issues caused by converter control interactions. However, the frequency-domain model of a power electronic system is linearized around a specific operating condition. It thus requires measurement or identification of frequency-domain models repeatedly at many operating points (OPs) due to the wide operation range of the power systems, which brings significant computation and data burden. This article addresses this challenge by developing a deep learning approach using multilayer feedforward neural networks (FNNs) to train the frequency-domain impedance model of power electronic systems that is continuous of OP. Distinguished from the prior neural network designs relying on trial-and-error and sufficient data size, this article proposes to design the FNN based on latent features of power electronic systems, i.e., the number of system poles and zeros. To further investigate the impacts of data quantity and quality, learning procedures from a small dataset are developed, and K-medoids clustering based on dynamic time wrapping is used to reveal insights into multivariable sensitivity, which helps improve the data quality. The proposed approaches for the FNN design and learning have been proven simple, effective, and optimal based on case studies on a power electronic converter, and future prospects in its industrial applications are also discussed.
暂无评论