Search

Multivariate time-series (MTS) clustering is a fundamental technique in data mining with a wide range of real-world applications. To date, though some approaches have been developed, they suffer from various drawbacks, such as high computational cost or loss of information. Most existing approaches are single-view methods without considering the benefits of mutual-support multiple views. Moreover, due to its data structure, MTS data cannot be handled well by most multiview clustering methods. Toward this end, we propose a consistent and specific non-negative matrix factorization-based multiview clustering (CSMVC) method for MTS clustering. The proposed method constructs a multilayer graph to represent the original MTS data and generates multiple views with a subspace technique. The obtained multiview data are processed through a novel non-negative matrix factorization (NMF) method, which can explore the view-consistent and view-specific information simultaneously. Furthermore, an alternating optimization scheme is proposed to solve the corresponding optimization problem. We conduct extensive experiments on 13 benchmark datasets and the results demonstrate the superiority of our proposed method against other state-of-the-art algorithms under a wide range of evaluation metrics.

The existing small-signal analysis models do not adequately describe the impact of the types of disturbance on the torque-frequency dynamics of the multi-VSC parallel system. In this paper, based on the principle of electromechanical analogy, a unified equivalent admittance modeling method is proposed by analogizing the rotor dynamic model of a synchronous generator (SG) to the circuit network, which has the advantages of simple modeling, clear physical meaning and easy expansion. Firstly, the equivalent admittance model of a single SG/VSC is established, and then easily extended to the multi-VSC system. The proposed modelling method not only applicable to a variety of VSCs, but also can handle different types of disturbance. More importantly, the circuit theory can be applied to analyze the torque-frequency dynamic characteristics. Subsequently, the equivalence between the proposed model and the state-space model under different damping terms is compared. In addition, the response characteristics of two different damping terms are compared. Next, the dynamic response characteristics of single VSC and multi-VSC parallel systems under different disturbances are analyzed. Finally, related simulations and experiments are executed to verify the accuracy and effectiveness of the proposed method.

Federated learning (FL), as a type of collaborative machine learning framework, is capable of preserving private data from mobile terminals (MTs) while training the data into useful models. Nevertheless, from a viewpoint of information theory, it is still possible for a curious server to infer private information from the shared models uploaded by MTs. To address this problem, we first make use of the concept of local differential privacy (LDP), and propose a user-level differential privacy (UDP) algorithm by adding artificial noise to the shared models before uploading them to servers. According to our analysis, the UDP framework can realize $(\epsilon_{i}, \delta_{i})$-LDP for the $i$-th MT with adjustable privacy protection levels by varying the variances of the artificial noise processes. We then derive a theoretical convergence upper-bound for the UDP algorithm. It reveals that there exists an optimal number of communication rounds to achieve the best learning performance. More importantly, we propose a communication rounds discounting (CRD) method. Compared with the heuristic search method, the proposed CRD method can achieve a much better trade-off between the computational complexity of searching and the convergence performance. Extensive experiments indicate that our UDP algorithm using the proposed CRD method can effectively improve both the training efficiency and model quality for the given privacy protection levels.

Lightning strikes in substations may result in extremely uneven ground potential distribution, which will threaten the safety and stability of equipment in substation. However, the main parameters for evaluating the performance of grounding grids are grounding resistance and impulse impedance, and little attention is paid to the distribution of the transient ground potential differences. In addition, when measuring the impulse impedance of a grounding grid, the length of the current leads should be very large, which requires a high-capacity impulse generator to suppress the influence of lead inductance on current waveform. In this paper, a test method is proposed for the transient characteristics of grounding grid with the current-return point located in the same grid. The feasibility of the method is illustrated by analyzing the ground impedance between the current injection point and the return point, longitudinal current of the grounding conductor, transient ground potential rise and transient ground potential difference. The effects of parameters such as soil resistivity, lightning waveform, size of the grid, spacing and materials of the grid conductor are discussed to demonstrate that the method can be applied to most substations. Finally, the practicality of the method is verified by experiments in a 220kV substation.

This paper presents a general method to obtain the improved common ground quasi-Z-source series dc-dc converters. Utilizing the negative output characteristic of the quasi-Z-source converter (QZSC), the output voltage of the improved converter is the sum of the input voltage and the output voltage of original QZSC, which means that part of the energy is directly supplied by the input voltage source without passing through QZSC. Therefore, compared with the original QZSC, the proposed improved converter has advantages such as lower voltage stress, lower current stress, higher voltage gain and higher efficiency. More importantly, all these improvements are obtained with no obvious drawback. No additional component is required, the input and output remain common ground, and if the input current of original converter and the load current are continuous, the input current of improved converter is also continuous. In order to prove the feasibility of proposed method, firstly, the improved converters of three existing QZSCs are obtained. Then, the operation principle, parameter design, and comparison are introduced in sequence. Finally, the above analyses are verified by simulation and experiment results.

Couplet is a unique art form in Chinese traditional culture. The development of deep neural network (DNN) technology makes it possible for computers to automatically generate couplets. Especially, Transformer is a DNN-based ``Encoder-Decoder'' framework, and widely used in natural language processing (NLP). However, the existed Transformer mechanism cannot fully exploit the essential linguistic knowledge in Chinese, including the special format and requirements of Chinese couplets. Therefore, this article adapts the Transformer mechanism to generate meaningful Chinese couplets. Specifically, the contributions of our work are threefold. First, considering the fact that the words in the corresponding positions of the antecedent clause and the subsequent clause in a Chinese couplet always have same part-of-speech (pos, i.e., word class), pos information is intentionally added into the Transformer to improve the accuracy of the conceived couplet. Second, to deal with the large number of unregistered and low-frequency words in Chinese couplet, a specific unregistered/low-frequency word processing mechanism (UWP) is designed and combined with the Transformer model. Third, to further improve the coherence of couplets, we incorporate the polish mechanisms (PMs) into Transformer model. In terms of three evaluation criteria including bilingual evaluation understudy (BLEU), perplexity, and human evaluation, the experimental results demonstrate the effectiveness of our designed Chinese couplet generation system.

High-Voltage power Integrated Circuits (HVICs) are widely used to realize high-efficiency power conversions (e.g., AC/DC conversion), gate drivers for power devices and LED lighting, and so on. The Bipolar-CMOS-DMOS (BCD) process is proposed to fabricate devices with bipolar, CMOS, and DMOS modes, and thereby realize the single-chip integration of HVICs. The basic integrated technologies of HVICs include High-Voltage (HV) integrated device technology, HV interconnection technology, and isolation technology. The HV integrated device is the core of HVICs. The basic requirements of the HV integrated device are high breakdown voltage, low specific on-resistance, and process compatibility with low-voltage circuits. The REduced SURFace field (RESURF) technology and junction termination technology are developed to optimize the surface field of integration power devices and breakdown voltage. Furthermore, the ENhanced DIelectric layer Field (ENDIF) and REduced BULk Field (REBULF) technologies are proposed to optimize bulk fields. The double/triple RESURF technologies are further developed, and the superjunction concept is introduced to integrated power devices and to reduce the specific on-resistance. This work presents a comprehensive review of these technologies, including the innovation technologies of the authors' group, such as ENDIF and REBULF, substrate termination technology prospective integrated technologies and HVICs in wide band gap semiconductor materials are also discussed.

Multi-terminal modular multilevel converter (MMC) power transmission is the current development trend in DC transmission. However, owing to its fast control and dynamic characteristics, it may interact with the other inductive devices or the grid and cause stability issues. This paper describes the use of the impedance analysis method in the frequency domain to analyze system stability. First, the multi-harmonic linearization method is used to model the impedance of a three-terminal MMC system. To make the model more accurate, the modeling considers the effect of frequency coupling caused by an asymmetric controller and the AC sampling site position. This paper then proposes an equivalent method of the grid impedance for the three-terminal system that can accurately judge the stability of the grid-connected system when the AC side is coupled. This work provides a theoretical basis for the prediction and suppression of the oscillation of MMC grid-connected system with coupling on the AC side. Finally, a three-terminal MMC model is built in Matlab/Simulink and the Nyquist stability criterion is used to verify the correctness of the analysis.

In this paper, a new vision- and grating-sensor-based intelligent unmanned settlement (IUS) system is proposed for convenience stores to automatically recognize the shopping behavior of customers, record their identities, and generate invoices. First, we design a new IUS architecture, which includes a shelf module and exit module. To achieve automatic settlement for each customer, a shopping event detection method is proposed. In this method, a vision-based human pose estimation algorithm is used to detect a human form standing in front of a shelf. The hand actions of each customer are detected by a grating sensor, and an image recognition method based on a convolutional neural network (CNN) is applied to recognize the items in the hands of customers. To reduce the image annotation workload, we propose a semisupervised training method for the recognition network. Based on hand action detection and item recognition, a shopping event recognition method is designed for the system, and a facial image of the customer corresponding to each shopping behavior is captured. Finally, each detected shopping event is added to the invoice of the corresponding customer via a facial recognition method. To verify the effectiveness of the proposed IUS system, we have built a handheld item image dataset and a shopping event dataset for an unmanned convenience store. The experimental results show that the proposed system can accurately recognize shopping behaviors and generate invoices.

Based on the analysis of the operation characteristics of photovoltaic cells, a novel virtual synchronous generator adaptive power regulation strategy for two-stage three-phase photovoltaic system with high permeability of clean energy is proposed. By improving the maximum power point tracking process of photovoltaic cells, the variable power point tracking control of photovoltaic cells is realized, and the control effect of DC/DC converter on the output power of photovoltaic cells is fully exerted. Without additional energy storage equipment, the energy supply required for inertia support and primary regulation of virtual synchronous generator is accurately realized. Combined with inverter VSG control strategy, the control of photovoltaic virtual synchronous generator is realized.

The current small-signal models of inverters that mimic synchronous generators usually use the steady-state equation of feeder line and load, and ignore the dynamic characteristics of voltage and current controller of inner loop. Thus, some important dynamics of the system will be omitted. To capture the dynamics more accurately, firstly, a small-signal model of a three-phase isolated inverter that mimics a synchronous generator was developed in this paper, and a comparison between the corresponding simulink model and the small-signal model was presented to validate the model correctness. Then, participation factor was calculated, which showed the relation between the state variables and eigenvalues. Based on the model, several root locus plots as a function of the system parameters was obtained. An individual eigenvalue sensitivity analysis of a certain eigenvalue close to the real axis disclosed different parameters’ contribution to the oscillation mode. Finally, a disturbance test with simulink model and the small-signal model under multiple sets of parameters was carried out and the results can accurately reflect the oscillation mode, which verified the correctness of the parameters effect on the conjugate eigenvalues.

In view of low conduction loss and low cost, thyristors are good candidates for configuring solid state circuit breakers (SSCBs). However, due to the half-controlled characteristic of thyristors, it is difficult for the thyristor-based SSCBs (TCBs) to achieve a fully controllable current breaking capability. In this paper, a novel bidirectional TCB is proposed, which can fully control the current breaking process with simple structure and control strategy. Besides, the proposed TCB can break current without creating current surges to the source and load, thereby enhancing the reliability of direct current (dc) systems. Typical advantages of high power efficiency and low cost can also be obtained from the novel topology. Operation process and circuit design of the proposed TCB are presented. The effectiveness of the proposed solution is verified by experimental results.

In this article, to reduce the deviation of frequency/voltage and power sharing in islanded microgrids, a distributed coordinated control based on multiagent system (MAS) is proposed. A hierarchical control structure is designed in this strategy. For the poor dynamic performance in the primary control, coordinated compensation terms are added to the traditional droop control-loop to reduce the transient fluctuation in frequency/voltage under load change. In the secondary control, an event-triggered consensus protocol is applied to restore frequency/voltage, obtain equal per-unit power sharing, and relieve communication pressure. Also, the communication network is viewed as an unreliable switching graph under communication channel interruption. To this end, a data prediction compensation mechanism is proposed to ensure the secondary control on the right track. Furthermore, to predict the lost data caused by communication channel interruption, an improved exponential smoothing model based on time series is proposed. Finally, several simulation scenarios have been carried out in MATLAB/Simulink to illustrate the validity of the theoretical results.

For continuous monitoring of cardiovascular diseases, this paper presents a novel framework for heart sound acquisition. The proposed approach uses compressed sensing for signal sampling, and a two-stage reconstruction is developed for reconstruction. The first stage aims to give a tentative recovered signal, on which a peak detection technique is developed to identify whether there is a peak in current segment and, if so, its location. With such information, an adaptive dictionary is selected for the second round reconstruction. Because the selected dictionary is adaptive to the morphology of current frame, the signal reconstruction performance is consequently promoted. Experiment results indicate that a satisfactory performance can be obtained when the frame length is 256 and the signal morphology is divided into 16 categories. Furthermore, the proposed algorithm is compared with a series of counterparts and the results well demonstrate the advantages of our proposal, especially at high compression ratios.

Critical conduction mode (CRM) or discontinuous conduction mode (DCM) boost power factor correction (PFC) converter is widely used in low power applications for its advantages of zero-current turn-on of the switch, no reverse recovery in diode and simple control. The input current of fixed switching frequency control CRM boost PFC converter and the input current of constant duty-cycle control DCM boost PFC converter mainly contain third harmonic. Moreover, the phase of the third harmonics is opposite. This paper proposes a harmonic elimination parallel control to connect these two converters in parallel. The input current third harmonic of the parallel converter can be eliminated completely so as to improve the PF to nearly unity in the whole input voltage range. This parallel method is an interesting way to extend power, and is different from the existing method of parallel DCM boost PFC converter or parallel CRM boost PFC converter. The experimental results verify the accuracy of the theoretical analysis.

Nowadays, much effort focuses on research for the multi-receiver WPT system based on magnetic resonance. However, its output characteristics are significantly deteriorated by the variation of transfer spacing, loads, and cross-couplings, which exerts harmful effects on the performances of cascading bucks before loads if needed. Inspired by the work of a single-receiver WPT system based on the parity-time-symmetric model, in this paper, the self-oscillating source is first applied to the multi-receiver WPT system to tackle these drawbacks. Based on the coupled-mode theory, the proposed system's modeling is established and deduced to analyze its performance. The model analysis shows how the system can achieve zero-voltage switching simply via choosing the minimum dead time and leading time whose calculation method is provided. Meanwhile, a transmitting coil for a uniform magnetic field is designed to improve the free-positioning performance. The analysis is validated on a 36 W prototype, including the effects of the variation of transfer spacing, cross-couplings, and loads on the output characteristics, efficiency, and operating frequency. The proposed system is demonstrated to provide a much more robust power transfer than the conventional magnetic-resonant multi-receiver WPT system, which helps extend the transfer distance and improve the system efficiency.

A novel Cuk-based bridgeless rectifier is proposed for receiving side of wireless power transfer (WPT) system. The proposed converter can not only help improve overall efficiency, but also is with features of wide power modulation range and very low output current ripple. In terms of the output current source characteristics of series-series-compensated WPT system, a bridgeless rectifier and a Cuk converter are combined and an input inductor of Cuk converter is cut down to form the proposed single-stage converter. Besides, a synchronous operation method is derived for the proposed converter to realize the optimal working point of transmission power and efficiency. Operation principles of the converter and its impact on performance of WPT system are analyzed. Also, design methods for system parameters are given. A prototype is built up to verify the feasibility of the proposed topology and the operation method. It is noted that the proposed converter can well adjust the output power of WPT system to realize the desired working point via adjusting the duty cycle at different transfer distances and loads.

The bifactor model is a promising alternative to traditional modeling techniques for studying the predictive validity of hierarchical constructs. However, no study to date has systematically examined the influence of cross-loadings on the estimation of regression coefficients in bifactor predictive models. Therefore, we present a systematic examination of the statistical performance of six modeling strategies to handle cross-loadings in bifactor predictive models: structural equation modeling (SEM), exploratory structural equation modeling (ESEM) with target rotation, Bayesian structural equation modeling (BSEM), and each of the three with augmentation. Results revealed four clear patterns: 1) forcing even small cross-loadings to zero was detrimental to empirical identification, estimation bias, power and Type I error rates; 2) the performance of ESEM with target rotation was unexpectedly weak; 3) augmented BSEM had satisfactory performance in an absolute sense and outperformed the other five strategies across most conditions; 4) augmentation improved the performance of ESEM and SEM, although the degree of improvement was not as substantial as that of BSEM. In addition, we also presented an empirical example to show the feasibility of the proposed approach. Overall, these findings can help users of bifactor predictive models design better studies, choose more appropriate analytical strategies, and obtain more reliable results. Implications, limitations, and future directions are discussed.

Nowadays, with the development of communication technology and its application in islanded microgrid, the pure power grid has gradually become a kind of cyber-physical system. In this system, the communication data and consensus algorithm are widely used in the secondary control of energy resources. However, in the communication process, there are also risks of network attacks, such as denial of service attack. To deal with this kind of attack existing in the data uploading channel during secondary control, the evolutionary game-based defense mechanism is designed, and the main works are as follows: firstly, the caused influence by attack on the control effect is analyzed through constructing a small-signal model. Secondly, two defense strategies including ``Adjacent prediction'' and ``Path reconstruction'' are designed. Facing different attack situations, the most suitable strategy can be selected via the evolutionary game. Thirdly, a game-based active defense strategy is designed, which can simulate the attack probability in the near future so that the defense can be prepared in advance. Finally, based on Laplace transformation and H $_{∞} $ robust theories, the parameter design in defense strategies is completed. The effectiveness of the above strategies is shown in multiple case studies.

To improve the voltage and power sharing of distributed generations, the hierarchical controls are widely used in microgrids while the dependent communication network makes it difficult to ensure the performance under limited bandwidth. In this paper, a predictive voltage hierarchical controller is designed. With the delayed secondary PI compensation signals, an inner-loop robust control is designed in the primary controller, ensuring stable voltage tracking. Then, a predictive controller is designed to provide neighbor and local predictions, such that the data integrity is improved by an accurate predictive compensation under bidirectional data-loss. Final case study results verify the effectiveness of the proposed method.

Emerging power sources and load demands penetrating the power system on demand side lead to diversified dynamic processes of distribution networks and thus create challenges for transient simulation of large‐scale active distribution networks. This paper proposes a network‐decomposition‐based multi‐rate parallel transient simulation technique for active distribution networks. Prominent advantages of the proposed method are accentuated by balancing computational complexities of CPU cores and devised core‐oriented adaptive step sizes in terms of the simulated elements with various dynamic characteristics. First, an optimised decomposition strategy of active distribution networks is proposed with the goals of balancing the computational burden of each core and minimising interactive data between cores. On that basis, coordinating models of multi‐port networks and equivalent models of interfaces are established. Then, a parallel simulation method of coordinated adaptive step sizes is proposed to accelerate the transient simulation. Case studies on eight feeders connecting to a common 10‐kV bus are carried out to verify the simulation speed and precision of the proposed parallel simulation method.

Inversion of large-scale time-domain transient electromagnetic (TEM) surveys is computationally expensive and time-consuming. The calculation of partial derivatives for the Jacobian matrix is by far the most computationally intensive task, as this requires calculation of a significant number of forward responses. We propose to accelerate the inversion process by predicting partial derivatives using an artificial neural network. Network training data for resistivity models for a broad range of geological settings are generated by computing partial derivatives as symmetric differences between two forward responses. Given that certain applications have larger tolerances for modeling inaccuracy and varying degrees of flexibility throughout the different phases of interpretation, we present four inversion schemes that provide a tunable balance between computational time and inversion accuracy when modeling TEM datasets. We improve speed and maintain accuracy with a hybrid framework, where the neural network derivatives are used initially and switched to full numerical derivatives in the final iterations. We also present a full neural network solution where neural network forward and derivatives are used throughout the inversion. In a least-squares inversion framework, a speedup factor exceeding 70 is obtained on the calculation of derivatives, and the inversion process is expedited ~36 times when the full neural network solution is used. Field examples show that the full nonlinear inversion and the hybrid approach gives identical results, whereas the full neural network inversion results in higher deviation but provides a reasonable indication about the overall subsurface geology.

Unsupervised domain adaptive object detection aims to adapt a well-trained detector from its original source domain with rich labeled data to a new target domain with unlabeled data. Recently, mainstream approaches perform this task through adversarial learning, yet still suffer from two limitations. First, they mainly align marginal distribution by unsupervised cross-domain feature matching, and ignore each feature's categorical and positional information that can be exploited for conditional alignment; Second, they treat all classes as equally important for transferring cross-domain knowledge and ignore that different classes usually have different transferability. In this paper, we propose a joint adaptive detection framework (JADF) to address the above challenges. First, an end-to-end joint adversarial adaptation framework for object detection is proposed, which aligns both marginal and conditional distributions between domains without introducing any extra hyperparameter. Next, to consider the transferability of each object class, a metric for class-wise transferability assessment is proposed, which is incorporated into the JADF objective for domain adaptation. Further, an extended study from unsupervised domain adaptation (UDA) to unsupervised few-shot domain adaptation (UFDA) is conducted, where only a few unlabeled training images are available in unlabeled target domain. Extensive experiments validate that JADF is effective in both the UDA and UFDA settings, achieving significant performance gains over existing state-of-the-art cross-domain detection methods., Accepted by IEEE T-MM, 2021, the code is available at https://github.com/BOBrown/JADF-caffe

The conventional, L2-norm-based, regularization term in electromagnetic (EM) inversions implements smooth constraints on model complexity in the space domain, which can smoothen the boundaries of complex underground structures. To improve the resolution of 3-D frequency-domain airborne EM (AEM) inversions, we propose a new algorithm for sparse-regularized inversion based on the shearlet transform. Unlike traditional methods that invert the model parameters in the space domain, we first transform the 3-D resistivity model into the frequency domain and then invert the sparse coefficients using an L1-norm measure to ensure the sparseness of the solution. Finally, we transform the shearlet coefficients back to the space domain to update the model. The shearlet transform has inherent multiscale and multidirectional properties, making it capable of effectively extracting complex geometries such as curved boundaries. We adopt the finite-difference method and the iteratively reweighted least-squares scheme for our 3-D AEM modeling and inversions and apply the ``moving footprint'' technique to speed up the inversion. Tests using synthetic data show that sparse-regularized inversion based on the shearlet transform can obtain more-focused inversion results than conventional smoothness-constrained inversions based on the L2-norm. Tests using field survey data also reveal that the new method can achieve more realistic underground structures.

Image generation by the conditional generative adversarial network (CGAN) can provide rich samples for supervised classification tasks. However, the unstable gradient updating problem makes it difficult to be used for image generation, especially for the high-resolution SAR image. In this letter, an improved residual condition generation network is proposed to enhance the quality of generated images and classification accuracy for hard negative examples. First, a residual convolutional-based block is built to clear the detail texture of different types of targets. Then, the gradient penalty and Wasserstein loss are used as discriminators to improve the similarity between real samples and the intra diversity of a generated images. We test the proposed method on generation and classification of three kinds of commercial ships using C-band 3-m Gaofen-3 (GF-3) SAR images. Experimental results show that our method can generate high-quality ship samples and achieve good classification accuracy.

In this paper, a new Multi‐Focus Image Fusion (MFIF) method based on multi‐source joint layering and Convolutional Sparse Representation (CSR) is proposed. Based on the characteristics of multi‐focus source images, a multi‐source joint layering regularization model was designed to divide the sources into a common base‐layer and respective focus detail‐layers. This strategy can overcome the defects caused by source layering separately effectively. In detail‐layer fusion, CSR was employed to extract and global features. It can avoid detail blur and high computational cost caused by image blocking in the conventional sparse representation model. The proposed detail‐layer fusion rule combined the CSR coefficient maps pairwise with the window based select‐max rule. In the experiments, the optimal layering parameter was selected by experiments at first, and then five recently proposed specific MFIF or general image fusion algorithms were contrasted with the proposed method by plenty of subjective and objective experimental comparisons. The experimental results demonstrated the superiority of the authors’ method.

The great success of deep learning poses urgent challenges for understanding its working mechanism and rationality. The depth, structure, and massive size of the data are recognized to be three key ingredients for deep learning. Most of the recent theoretical studies for deep learning focus on the necessity and advantages of depth and structures of neural networks. In this paper, we aim at rigorous verification of the importance of massive data in embodying the out-performance of deep learning. To approximate and learn spatially sparse and smooth functions, we establish a novel sampling theorem in learning theory to show the necessity of massive data. We then prove that implementing the classical empirical risk minimization on some deep nets facilitates in realization of the optimal learning rates derived in the sampling theorem. This perhaps explains why deep learning performs so well in the era of big data., Comment: 15pages, 4 figures

For a multi-load wireless power transfer (WPT) system, it has always been a challenge to maintain constant output power with constant efficiency for each load when the coupling conditions change. Besides, the unbalanced power distribution and cross-coupling among loads are also the key issues. In order to solve these problems, a novel multi-load WPT system combining the parity-time (PT) symmetry principle and time-sharing control strategy is proposed. Firstly, the multi-load PT-symmetric WPT (PT-WPT) system with single natural resonant frequency is modeled and analyzed based on the circuit model. Then, a multi-load PT-WPT system with time-sharing control strategy is constructed by adding a controllable resonant capacitor array at the transmitter. The time-sharing control strategy is realized by periodically turning on/off a controllable resonant capacitor selectively, which enables a single transmitter to supply power to multiple loads with different natural resonant frequencies, and each load realizes PT transmission characteristics. Further, using the time-sharing control strategy, the power distribution among multiple loads is realized by only controlling the duty cycle of each load. Finally, two dual-load PT-WPT experiments are designed and carried out. Experimental results show that each load can maintain constant output power with constant transfer efficiency about 90% in a wide coupling range, and realize power distribution at desired values simultaneously.

In a data access control system oriented toward the cloud storage environment, a data owner defines attribute-based access control policies for data files to realize fine-grained data sharing. However, the existing schemes have defects in user execution efficiency and user privacy protection, and they do not consider the problems of user revocation and attribute updates. To this end, we propose a ciphertext policy attribute-based encryption method with verifiable outsourced decryption; this requires a user to complete decryption with the help of a server, but the results of the outsourced decryption can be verified independently. With this new encryption scheme and the technique of k-times anonymous authentication, a new fine-grained data access control system was constructed; this system allows a server to provide users with outsourced decryption services, and users’ computation cost is independent of the size of the underlying access control policy. Moreover, the number of outsourced decryption requests is limited. In addition, the new system supports user revocation and attribute updates and it is provably secure under formal proofs. An efficiency analysis shows that it can be compared with other similar systems in terms of performance, despite the addition of several practical properties.

This paper proposes a receiver position estimation method suitable for multi-transmittermulti-TXwireless power transfer systems from a new perspective of machine learning. The novel method uses the voltage and current of the transmitter side, and is realized through self-learning, which can solve the problems of complex electromagnetic environment and slow response in the traditional method. First, the switching criterion of the multi-TX WPT system based on receiver position is established. In addition, the receiver position estimation method is determined by comparing the estimation accuracy of the BP and SVR algorithms. On this basis, a swithcing control strategy of multi-TX WPT system based on receiver position estimation is further proposed. Finally, the experimental prototype with FPGA controller is designed to verify the correctness of the theoretical analysis. The average relative error of the position estimation method is about 5%, and the system efficiency is stable near 90% with the misalignment conditions.

In recent years, the parity-time (PT) symmetry principle has been introduced into wireless power transfer systems, which can improve the stability of wireless power transfer within a certain range. In this paper, we analyze the generalized PT symmetry conditions of four basic wireless power transfer topologies based on circuit theory and coupled mode theory. The results show that the series-parallel topology meets the PT-symmetry condition based on the coupled mode theory, although it does not meet the condition obtained by the circuit theory. The symmetry between the energy of the two-resonance system is established in the series-parallel topology. According to this unique characteristic, a constant-voltage wireless power transfer system for spinal cord stimulation is designed, which is independent of the transfer distance and load in a wide range.

Cordless kitchen appliances powered by wireless power transfer (WPT) technology can operate without the clutter of cords, making them safer and easier to clean. In order to allow cordless kitchen appliances to be placed more freely, this paper proposed a parity-time (PT)-symmetric WPT system with stable output power in the whole coupling range. Here, a new implementation of a PT-symmetric circuit using a combination of self-oscillating methods and pulse width modulation (SO-PWM) is presented. The self-oscillating mode is activated in the strong coupling region, which guarantees a constant output power and constant transfer efficiency against the coupling coefficient variation. In the weak coupling region, the PWM mode with a fixed frequency and a variable duty cycle is adopted, and a control strategy based on primary-side-only parameters is presented to obtain stable power transmission. Besides, a corresponding coupling-region detection method based on the reflected resistance is proposed to ensure smooth switching between SO mode and PWM mode. The advantage is that the stable output power can be maintained in the whole coupling region without any extra DC-DC converter and dual-side communication requirements, while maintaining the same transfer efficiency as the standard PT-symmetric WPT system. A 500-W experimental prototype is built to verify the theoretical analysis. The results show that with a transfer efficiency of over 91.6%, the misalignment tolerance can be increased by 172%.

Multiproduct pipelines play an important role in ensuring the downstream energy supply as they transport about 70% of refined products. To guarantee the safe and stable operation of pipelines during the scheduling horizon, it is necessary to make a reasonable schedule. During the past two decades, the multiproduct pipeline scheduling problem has been widely studied in the world. In this paper, the bibliometric analysis method is adopted to conduct quantitative analysis and visual research on related literature. Then, through the review of the existing articles, the research framework of multiproduct pipeline scheduling problem is constructed. Finally, the research defects and trends are analyzed. The aim of the paper is to provide a comprehensive understanding of multiproduct pipeline scheduling as well as some directions and inspiration for the future research.

In this article, a novel isolated circuit breaker (ISCB) that realizes robust interruption capability is proposed for direct current microgird protection. First, the proposed ISCB isolates the main current path and the commutating circuit during the fault-clearing state, thereby not causing fault current surges on both the input and output sides, and effectively protecting the source and load. Second, the ISCB can automatically charge the commutating capacitor to offer interruption capability before the reclosing of the breaker, which ensures a safe restart of microgrids after faults (especially temporary faults). Third, due to the introduced voltage clamping circuits, the ISCB can successfully interrupt faults without being effected by parasitic parameters on microgrids. Finally, the use of the insulated gate triggered thyristor with strong tolerance to current and voltage surges can also enhance the robustness of the ISCB. Consequently, the ISCB features higher fault interruption robustness than the existing breakers. Moreover, design guidelines are provided through mathematical modeling, and then based on the design guidelines, a laboratory prototype is developed to validate the feasibility of the ISCB.

Compared with the past questionnaire survey, this paper applies the intelligent algorithm developed rapidly in recent years to identify the tendency of customers to buy financial products in the market. In addition, for the single state customer classification indicators based on the previous demographic information and action information, it is proposed to combine the action of market activities with demographic information; that is, the static integrated customer classification index is further combined with the improved neural network model to study the classification and preference of enterprise financial customers. Firstly, the enterprise financial customer classification model based on neural network algorithm is studied. Aiming at the shortcomings of easy falling into the local optimal solution of neural network algorithm, slow convergence speed of algorithm, and difficult setting of network structure, combined with the characteristics of genetic algorithm, the concept of adaptive genetic neural network algorithm is proposed. Then, the design of adaptive genetic neural network model is studied. Secondly, combined with the customer data of a financial enterprise and the characteristics of enterprise finance, this paper analyzes the risk influencing factors of enterprise financial customers, analyzes the customer data, evaluates the enterprise financial customers through the adaptive genetic neural network model, and realizes the classification of enterprise financial customers. Through an example, it is proved that the enterprise financial customer classification and preference model based on the adaptive genetic neural network algorithm discussed in this paper has better customer classification accuracy and can provide better method support for enterprise financial customer management.

Understanding mechanisms of coexistence is a central topic in ecology. Mathematical analysis of models of competition between two identical species moving at different rates of symmetric diffusion in heterogeneous environments show that the slower mover excludes the faster one. The models have not been tested empirically and lack inclusions of a component of directed movement toward favourable areas. To address these gaps, we extended previous theory by explicitly including exploitable resource dynamics and directed movement. We tested the mathematical results experimentally using laboratory populations of the nematode worm, Caenorhabditis elegans. Our results not only support the previous theory that the species diffusing at a slower rate prevails in heterogeneous environments but also reveal that moderate levels of a directed movement component on top of the diffusive movement allow species to coexist. Our results broaden the theory of species coexistence in heterogeneous space and provide empirical confirmation of the mathematical predictions.

Pine wilt disease (PWD) is currently one of the main causes of large-scale forest destruction. To control the spread of PWD, it is essential to detect affected pine trees quickly. This study investigated the feasibility of using the object-oriented multi-scale segmentation algorithm to identify trees discolored by PWD. We used an unmanned aerial vehicle (UAV) platform equipped with an RGB digital camera to obtain high spatial resolution images, and multi-scale segmentation was applied to delineate the tree crown, coupling the use of object-oriented classification to classify trees discolored by PWD. Then, the optimal segmentation scale was implemented using the estimation of scale parameter (ESP2) plug-in. The feature space of the segmentation results was optimized, and appropriate features were selected for classification. The results showed that the optimal scale, shape, and compactness values of the tree crown segmentation algorithm were 56, 0.5, and 0.8, respectively. The producer’s accuracy (PA), user’s accuracy (UA), and F1 score were 0.722, 0.605, and 0.658, respectively. There were no significant classification errors in the final classification results, and the low accuracy was attributed to the low number of objects count caused by incorrect segmentation. The multi-scale segmentation and object-oriented classification method could accurately identify trees discolored by PWD with a straightforward and rapid processing. This study provides a technical method for monitoring the occurrence of PWD and identifying the discolored trees of disease using UAV-based high-resolution images.

This paper presents an improved reconstruction approach for compressed sensing based ECG acquisition in Internet of Medical Things. The proposed method exploits the concepts of adaptive overcomplete dictionary and QRS detection in CS domain. Based on whether there is a QRS complex, the ECG frames to be reconstructed are divided into several categories and corresponding overcomplete dictionaries are trained to fit these different kinds of ECG frames. Specifically, QRS detection is first performed directly on the compressed measurements to determine the QRS morphology without actually reconstructing the signal in advance, and then a suitable overcomplete dictionary is chosen for the signal reconstruction. Because the selected dictionaries well fit the ECG frames, the reconstruction quality is consequently improved. Comparative experiments have been conducted, and the results well demonstrate the performance of the proposed approach.

In order to improve the target recognition effect of small UAV (unmanned aerial vehicle) remote sensing image, this paper proposes a new super-resolution reconstruction method based on the recurrent convolutional network, which can achieve different degrees of super-resolution effect by controlling the number of cycles. Moreover, it can control the number of iterations of small UAVs with different degrees of blur and can be better adapted to the recognition scenarios of UAVs. In addition, this paper studies the target recognition method of small UAV remote sensing image, combines fuzzy clustering method to construct the intelligent remote sensing image target recognition model, combines it with the UAV structure, realizes remote sensing recognition by UAV, and designs experiments to analyze the effect of remote sensing recognition. Further, this paper improves the recognition algorithm and positioning algorithm of remote sensing image, so that recognition and positioning of UAV video remote sensing image can get better results. Finally, this paper verifies the performance of the system through simulation experiments. The research results show that the method proposed in this paper has certain reliability.

In order to achieve the stable operation of different-capacity parallel distributed generators, this paper proposes an improved droop control strategy according to the control strategy of parallel synchronous generators. By the improved control strategy, the active power and reactive power of parallel distributed generators can be automatically allocated by ratio of distributed generators capacity, and the function of plug and play of the distributed generators in microgrids can be realized. It will improve the flexibility of distributed generator’s interconnection and solve the technological problem of different-capacity distributed generator interconnection. In addition, the proposed control strategy can keep unchanged while switching microgrid operation mode, realizing the smooth transition between grid mode and island mode, and improving the stability and reliability of microgrid. Simulation results verify that the improved droop control strategy can make the different-capacity parallel distributed generators work stably under static and transient operation.

The vehicle information acquired by current video-based traffic surveillance methods is usually onefold. In this paper, a traffic surveillance system for obtaining comprehensive vehicle information, including type, number of axles, 3D bounding box, speed, length, current driving lane, and traffic volume, is proposed based on instance segmentation which is realized by Mask R-CNN. The work mainly contains: 1) An annotated image dataset which contains car, coach, truck, and wheel is established for training Mask R-CNN; 2) The method for identifying the number of axles based on MaskIoU is proposed; 3) The 3D bounding box is generated based on the high-quality vehicle segmentation, which is more accurate than that generated by background subtraction methods; 4) For calculating the vehicle speed and length, the reference points on the road plane are determined by the lane with dashed line and one vanishing point to calculate the homography matrix; 5) In order to keep the obtained vehicle information reliability, the results identified from multi-frames are obtained by means of the tracking method SORT for processing. The proposed system is verified in different scenes. The average recognition accuracies of the vehicle types and number of axles are more than 97% and 88% in three testing videos. The speed errors of more than 90% vehicles are less than 4%. And the ratios of both missing and repetition in vehicle counting are low in general.

The aim is to improve the efficiency of medical data processing and establish a sound medical data management system. To apply distributed parallel classification algorithms in the field of hospital intelligent guidance, a Parallel Random Forest (PRF) classification algorithm is proposed based on the Apache Spark cloud computing platform. Given sparse cluster loss in variable density distribution data sets, an Adaptive Domain Density Peak Clustering (ADDPC) method is proposed. Here, a Bilayer Parallel Training-Convolutional Neural Network (BPT-CNN) model based on distributed computing is proposed to detect and classify colon cancer nuclei more accurately through the large-scale parallel deep learning (DL) algorithm. Then, the performance of the proposed model is evaluated through case analysis. The results show that the PRF algorithm based on distributed cloud computing platform can independently design data-parallel tasks, thereby optimizing the data communication cost and efficiency. ADDPC algorithm can adaptively measure domain density and merge sparse clusters to prevent data loss and fragmentation. The BPT-CNN model improves the performance of the algorithm and balances the workload of each task in the algorithm. The results have a significant reference value for solving problems in medical data processing.

The wireless power transfer (WPT) system based on the parity-time (PT) symmetry has been proved to be a promising WPT system due to its characteristics of constant efficiency and output power. However, the PT-symmetric range is limited, and it is necessary to widen the PT-symmetric region to extend the distance of constant and efficient transmission. The novel PT-symmetry-based WPT system with an inductor added on the receiving circuit is proposed in this article. First, the model of the proposed system is established by utilizing the coupled-mode theory, and the performance differences between the proposed system and the original PT-symmetric system are studied, and it is found that the critical coupling coefficient of the proposed system is smaller than that of the original PT-symmetric system. Second, to improve the efficiency of the proposed structure, the method of increasing the inductance of the transmitting coil is also put forward. Finally, the experimental results are basically consistent with the theoretical results, and the validity of the proposed approach is verified.

Mining the relationship structures among the investors plays a vital role in promoting economic development as well as preventing financial risks, especially in the context of big data. This article proposes fast networking approaches from investment big data to explore three underlying structures, namely, investment pedigrees, investment groups, and structural holes. Inspired by disjoint sets and path compression, we first present a pedigree classification algorithm to identify investment pedigrees. Second, through introducing a pruning strategy and a data structure termed as “2-tuple list,” we develop a novel linear-time structure mining algorithm in network (SMAN) for investigating investment groups and structural holes from the investment pedigree. Finally, we show that our SMAN has higher clustering accuracy and efficiency than other existing algorithms on a variety of real-world tasks in terms of normalized mutual information (NMI) values. Our method is particularly well suited for mining the underlying structures from investment big data.

Autonomous underwater vehicle (AUV) has been employed in oceanography applications based on a reliable navigation. The complex underwater environment leads to more velocity measurement errors of AUV, so it is difficult to determine the accurate navigation and positioning information. To solve the problem, a novel variational Bayesian-based filter for inaccurate input (VBFII) is proposed to determine the state information under the complex marine condition of inaccurate input. Firstly, the velocities are assumed to follow the Gaussian distribution, which can better describe the model of input information. Secondly, the augmentation method is used to augment the state vector and error covariance matrix to simplify calculation. The augmented state vector, the augmented predicted error covariance and measurement error noise matrices are derived more accurately based on the variational Bayesian (VB) approach. The experiment results show that the proposed VBFII has better estimation accuracy and robustness than other comparison algorithms.