Search

The multispectral image captured by the Sentinel-2 satellite contains 13 spectral bands with different resolutions, which may hider some of the subsequent applications. In this article, we design a novel method to super-resolve 20- and 60-m coarse bands of the S2 images to 10 m, achieving a complete dataset at the 10-m resolution. To tackle this inverse problem, we leverage the deep image prior expressed by the convolution neural network (CNN). Specifically, a plain ResNet architecture is adopted, and the 3-D separable convolution is utilized to better capture the spatial–spectral features. The loss function is tailored based on the degradation model, enforcing the network output obeying the degradation process. Meanwhile, a network parameter initialization strategy is designed to further mine the abundant fine information provided by existing 10-m bands. The network parameters are inferred solely from the observed S2 image in a self-supervised manner without involving any extra training data. Finally, the network outputs the super-resolution result. On the one hand, our method could utilize the high model capacity of CNNs and work without large amounts of training data required by many deep learning techniques. On the other hand, the degradation process is fully considered, and each module in our work is interpretable. Numerical results on synthetic and real data illustrate that our method could outperform compared state-of-the-art methods.

This article proposes a decoupled control scheme for the modular multilevel converter (MMC) to reduce the total harmonic distortion (THD) and eliminate one specific harmonic. First, to realize the decoupled control between the ac and dc paths of the MMC, an improved nearest level control (INLC) method is proposed. It applies the round function to generate the ac output voltage levels, instead of the arm output voltage levels. Thus, a staircase wave is generated by the INLC method. Then, by adding an additional pulse to each quarter of the staircase wave, one specific harmonic can be eliminated without optimizing each conduction angle. Furthermore, a decoupled circulating current fuzzy control method is proposed to suppress the second-order harmonic and balance the arm energy. By this decoupled control structure, the output voltage and the circulating current can be controlled independently. Taking the fifth-order harmonic as an example, simulation and experimental results validate that the proposed scheme can simultaneously realize THD reduction and one specific harmonic elimination in the MMC output voltage. [ABSTRACT FROM AUTHOR]

The open-circuit fault detection and localization (FDL) technique can improve the reliability of the modular multilevel converter (MMC). However, the conventional software-based FDL methods usually have a heavy computation burden or a limited localization speed. This article proposes a simplified and fast software-based FDL approach for the grid-connected MMC. First, the errors between the measured state variables (the output current and the circulating current) and their estimated values are calculated. By comparing these errors with their threshold values, the switch fault can not only be detected but also be localized to the specific arm. Then, the capacitor voltages in this faulty arm are collected, and the submodule (SM) with the highest capacitor voltage is selected. To confirm the switch fault in this SM, a modified Pauta criterion is presented to check the abnormal voltage data. As a result, the computation burden of the proposed software-based FDL approach is significantly reduced, and the faulty SM can be localized in a short period. Simulation and experimental results verify that the proposed approach can effectively detect and localize different open-circuit faults, and it is immune to the step of power references. [ABSTRACT FROM AUTHOR]

A power electronic transformer (PET) can flexibly regulate the power flow of a network and facilitate the integration of distributed energy resources via AC/DC networks. With the continuous expansion of the scale of multiregional interconnection networks, the traditional centralized optimization method faces great challenges. This paper proposes a distributed convex optimal power flow (OPF) model by leveraging the hierarchical structural characteristics of PET enabled AC/DC distribution networks. An analytical target cascading method is introduced to divide the centralized OPF problem into several subproblems that can make the subnetworks operate individually and coordinate with each other. The PET flexibility is also harnessed to reduce network power loss and voltage deviation. To improve the computational efficiency and algorithm convergence, the convex relaxation and linear approximation methods are proposed. This allows us to transform the original nonconvex nonlinear problem into a successive convex programming problem, which can be easily solved by an iterative solution algorithm. Simulation results verify the feasibility and efficiency of the proposed approach. [ABSTRACT FROM AUTHOR]

The environment sound classification(ESC) is of great significance to the monitoring and control of urban noise. Aiming at the curse of dimensionality phenomenon in ESC, a feature dimensionality reduction architecture combining attention and mutual information is proposed. In order to match the two-dimensional MFCC (Mel Frequency Cepstral Coefficients) feature matrix, the proposed method separates and reconstructs the feature frames of different samples, and achieves the effect of dimensionality reduction by making decisions on the information entropy between the feature frames and labels. In addition, the method combines LSTM (Long Short-Term Memory) model with attention mechanism to ensure the recognition accuracy of the model after dimensionality reduction. Ten urban acoustic events from UrbanSound8k (US8K) dataset are selected to verify the performances of the proposed method by simulation experiments, which are also compared with the existing classification methods. The simulation results show that by combining the attention mechanism and mutual information, the recognition accuracy of the proposed method on the UrbanSound8k dataset is 95.16%, and the parameter scale is the smallest, only 0.92M. Moreover, the model parameter scale is adjustable by dynamic frame retention mechanism to balance the recognition accuracy and speed. This method not only ensures a high classification accuracy, but also can reduce computing power consumption and storage space of monitoring equipment, which shows a better practical performance for urban acoustic events recognition. [ABSTRACT FROM AUTHOR]

We designed a distributed fiber optic sensor based on a hybrid Michelson-Sagnac interferometer to overcome the defect that the conventional hybrid structure cannot locate when the disturbance position is close to the midpoint of the Sagnac ring. The sensor has excellent and flat frequency response for no-blind zone location. With the demodulation algorithm based on 3 × 3 optical coupler and simple mathematical operations, we can get two phase signals with time difference for location and pattern recognition. The phase signals were converted to two-dimensional images by Gramian Angular Field. Tagging images to create a dataset and feeding dataset into a convolutional neural network to achieve pattern recognition. After experiments, the frequency response capability of the structure was verified, and it was able to achieve location within 0-20 km with location errors of $ \pm $ 42 m. In a six-category pattern recognition, the testing set accuracy was 97.83%. [ABSTRACT FROM AUTHOR]

This article investigates the operation of cascaded H-bridge (CHB) based static synchronous compensator (STATCOM) after failure in its submodules (SMs). Once a switch fails, the corresponding faulty SM is bypassed. In the post-fault operation, a zero-sequence voltage (ZSV) is injected to the converter phases to generate the maximum possible line-to-line voltage. Using a clamping technique, the fundamental component of injected ZSV (FZSV) is minimized according to the operating point of CHB STATCOM. But, FZSV leads to unequal flow of active power into the converter phases. In the proposed method, negative-sequence current is controlled to manage the active power flow within CHB phases. By doing this, the effect of FZSV on active power of CHB phases is compensated and the voltages of capacitors are kept balanced. This method ensures the continuous operation of CHB STATCOM with the minimum over-design and guarantees the voltage balance of capacitors after a failure. The presented simulation and experimental results validate the feasibility and effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

With the high penetration of the power electronic loads in the grid, the stability of static synchronous compensator (STATCOM) devices is greatly challenged. However, the conventional control methods for the modular multilevel converter (MMC) based STATCOM only consider the stability with small signal disturbances. This article proposes a novel dual-layer back-stepping control (BSC) for the MMC-based STATCOM. In the first layer, the BSC aims to regulate the sum of the capacitor energy and the reactive output current. In the second layer, the BSC aims to control the circulating current. Therefore, the proposed method possesses a fast dynamic response and accurate tracking with the Lyapunov stability of the MMC-based STATCOM. Compared with the arm-control-based BSC for MMC-based inverters, the proposed method has a simplified structure and a reduced computation burden. Moreover, the proposed method realizes the independent control between the output current and the circulating current. The simulation and experimental results verify the effectiveness of the proposed method. In addition, its robustness toward different circuit parameters and the operation ability under unbalanced grid fault is also verified. [ABSTRACT FROM AUTHOR]

The model predictive control (MPC) methods have been widely applied due to the fast dynamic performance and multiple control targets. However, they show certain limits when applied to the multilevel converters, such as unfixed switching frequency, large computation burden, and complex weighting factors selection. This article proposes a novel modulated MPC (M2PC) method for cascaded H-bridge (CHB) converter-based static synchronous compensator (STATCOM). Instead of multitarget integration and weighting factor design, the proposed method disintegrates the voltage balancing control process from the cost function. The evaluated control sets for output current control is decided by the search step and the search range in each period. By proper design and selection of the adaptive search step for the output voltage references under α–β coordinate, the search range can be reduced, which greatly reduces the number of evaluated control sets in each control period. Therefore, the dynamic of output characteristics and the low computation burden of the MPC method are guaranteed simultaneously. In addition, with the simple and independent integration of the capacitor voltage balancing control, the proposed method can be easily applied in the CHB converter-based STATCOM application. Furthermore, this article provides a delay compensation for the prediction process, and it provides a fixed switching frequency for each H-bridge cell. Simulation and experimental results verify the effectiveness of the proposed M2PC method. [ABSTRACT FROM AUTHOR]

For hyperspectral imagery (HSI) classification, most of the studies focus on how to improve the classification accuracy, while the influence of sampling strategy for classification performance attracts little attention. For now, random sampling (RS) is the most adopted strategy. That is, for a hyperspectral image, a certain number of labeled samples are randomly selected as the training set, and the remaining labeled samples are taken as the test set. However, the RS strategy will produce over optimistic results when used for performance evaluation because of the overlap between training set and test set. Though spectral-spatial classification methods benefit most from the RS strategy, the pixel-wise classification methods can also benefit from it because of the high spectral correlation between training and test samples. However, in practical applications, the RS strategy is not feasible. Because the training and test samples are often collected from different locations. In this situation, the correlation between training and test samples will decrease dramatically and the performance of HSI classification methods will be affected. In this article, a nonoverlapped sampling method is adopted to reduce the correlation between training and test samples and different classic classification methods are evaluated. Experimental results show that the classification performance of all methods drops a lot when nonoverlapped sampling strategy is adopted. After the analysis of some important factors for HSI classification, we also propose a cotraining-based classification method to relief the influence of sampling strategy and obtains much better performance compared with those classic spectral-spatial classification methods. [ABSTRACT FROM AUTHOR]