Microgrids equipped with hybrid energy storage systems (ESSs) are increasingly critical for balancing the intermittency of renewable energy sources and the fluctuations in demand. This article introduces a novel multiagent imitation learning (MAIL) framework for real-time energy management in microgrids, particularly under real-time pricing conditions. The approach leverages a problem decomposition strategy, which separates the energy management task into two phases: first, optimal actions for each ESS are estimated using individual agents, each employing a deep neural network trained to emulate an ideal mixed-integer linear programming solver; then, a one-step online optimization reacts to these decisions to optimize the remaining system components holistically. Rigorous proofs establish the equivalence of the decomposed subproblems to the original optimization problem, ensuring integrity and effectiveness. Comparative simulation studies with real-world data reveal that our MAIL framework offers significant cost advantages and superior training efficiency relative to both traditional single and emerging multiagent reinforcement learning methods. Notably, the operational costs achieved through our approach are markedly lower than those of its counterparts, closely approximating the theoretical minimum. This underscores the method’s proficiency and its potential as an effective solution for microgrid energy management. Moreover, the MAIL strategy’s multiagent design allows for straightforward scalability to more extensive systems involving multiple ESSs and shows promising potential for cooperative management across interconnected microgrids.

With limited annotated data, the existing deep learning-based ground-penetrating radar (GPR) data inversion methods have difficulty effectively reconstructing the relative permittivity (RP) of subsurface defects. To address this issue, a GPR data inversion method based on semisupervised deep learning is proposed. A synthetic aperture radar (SAR) feature fusion branch is designed in the encoding stage of the RP generator to enrich the feature representation of the defect shape profiles. A multidimensional parallel attention module is designed to aggregate defect-related invariant features, which thereby enhances the capability for inversing new defect types with new shapes. In addition, foreground defect loss is introduced to improve the performance of the defect boundary reconstruction. The validation of the proposed method is performed using both synthetic and real data. Comparative experiments with the existing methods indicate that in scenarios with limited annotated data, the proposed method exhibits superior performance for imaging new subsurface defect types that are absent in the labeled target domain.

Periodically monitoring the pavement cracks is of great importance to many transportation infrastructures. This paper proposed an unsupervised deep-learning-based method to match the cracks in multi-temporal unmanned aerial vehicle (UAV) images and identify the changes of pavement cracks over time. A regional focus module was specially designed to enforce the network to focus on regions where cracks were located and enhance its capacity for small-crack identification. Moreover, a data augmentation method which combined Poisson blending and random projective transformations was introduced for generating images with crack variations for model training. The superiority of the method was validated using actual image collected from real pavements. The experimental results showed that the proposed method outperformed the feature-based method and existing unsupervised deep learning-based UAV image registration method.

This study investigates economical scheduling of charging for an electric vehicle (EV) in a typical household with an intelligent charging management system. The problem formulation considers rooftop solar power generation, time-varying domestic energy consumption, real-time pricing of electricity, and user preferences. This task traditionally takes the form of a mixed-integer linear programming (MILP) problem, but we demonstrate its equivalence to linear programming (LP) to reduce computational complexity. The LP problem can be solved to global optimality if all future information is known, which is unrealistic in practice and replaced with forecasting. Learning-based methods such as deep reinforcement learning (DRL) eliminate the need for a forecaster and make online decisions rapidly using a learned policy. We propose an approach based on imitation learning that leverages the knowledge of an LP expert by learning from its optimal demonstrations instead of learning from scratch in DRL. Our approach trains a deep neural network (DNN) based policy efficiently in a supervised manner and incorporates a safety post-processing mechanism that enforces strict constraint satisfaction. Numerical studies on real-world data show that the proposed approach achieves$23~ \sim ~220$times speedup compared to DRL for DNN training, and the total electricity cost is far lower than DRL as well, which is strikingly close to the lower bound in theory. Our implementation code can be found athttps://github.com/ZhenhaoH/IL_EVCS.

Ground-penetrating radar (GPR) with different center frequencies can detect defects at different depths with a range of resolutions enabling it to be used for subsurface defect inspection. However, the existing deep learning methods cannot accurately invert the permittivity from GPR data of different frequencies, due to the limited number of labeled GPR images for every center frequency. To tackle this challenge, a semi-supervised deep neural network-based cross-frequency GPR data inversion method was proposed, which enables the generalized model to be trained on the GPR data with one frequency (source domain) for migration to other frequencies (target domain). The method was trained in a semi-supervised manner using a small number of paired GPR data with permittivity labels and a large amount of unlabeled GPR data without corresponding permittivity maps. An adversarial learning mechanism together with a novel random perturbation strategy was designed to improve the global inversion performance for a large-scale structure and avoid a discontinuity in the reconstructed shapes. Furthermore, a mean teacher architecture is introduced to improve the inversion accuracy of detailed information from the unlabeled GPR data under different perturbation conditions. The ablation and comparative experiments results indicated that the proposed method outperforms other methods and can be effectively generalized to GPR B-Scan data with different frequencies and signal-to-noise ratios. In addition, sandbox model testing was conducted and the results indicate that this method can transfer the knowledge from the synthetic data domain to the real data domain with satisfactory results.

Ground penetrating radar (GPR) enables infrastructure inspection using consecutive and long survey lines. However, the existing GPR data processing methods may lead to distortions or dislocations in the reconstructed shapes of the detected objects or even inconsistencies of the inverted dielectric values when performing inversion or object identification directly using GPR data obtained from the consecutive and long survey lines. To overcome these issues, this study proposed a novel deep neural network (DNN) architecture named GPRI2Net to simultaneously reconstruct the permittivity maps and categorize the object class labels from the GPR data of consecutive and long survey lines. GPRI2Net combined a convolutional neural network (CNN) based on DenseUnet and a recurrent neural network (RNN) based on bidirectional convolutional long short-term memory (Bi-ConvLSTM) to exploit the contextual information in and between the B-Scan segments extracted from the GPR data of a consecutive and long survey line. In addition, GPRI2Net performed inversion and object identification simultaneously using one network, which highly shared features between the two tasks and greatly reduced the computational complexity. Validation experiments were performed at two levels: first, using synthetic data based on the tunnel liner defects model and then using a sandbox model test in a realistic scenario. The results demonstrated that GPRI2Net can reconstruct consecutive permittivity maps and categorize the object class labels from GPR data with different dominant frequencies and survey line lengths and achieved a superior performance using the synthetic data compared to several other methods. Moreover, GPRI2Net also achieved satisfactory results using real-world GPR data.