Search

As the reform of electricity marketization advances, the virtual power plant (VPP), as an emerging market participant, is being progressively incorporated into the trading scope of both the electricity spot market (ESM) and the reserve market (RM). Due to its flexible regulation capabilities, the bidding strategies of VPPs in the electricity market are complex and variable, presenting new challenges to the joint operation of power markets. In this context, first, based on the interactive game relationships among multiple market entities under the operating mechanisms of the spot and reserve markets, this paper designs a principal-follower game framework for transactions in these markets with the VPP as the main entity. Second, it constructs a two-level joint optimized bidding strategy model for VPP participation in the spot and reserve markets, where the inner layer is the VPP’s optimized bidding strategy model aiming to maximize the total revenue from the spot and reserve markets; the outer layer is the spot and reserve market clearing model targeting the minimization of the total electricity purchasing cost for society. Finally, the simulation case analysis shows that the proposed method can achieve joint optimal operation of the electricity spot and reserve markets by considering the VPP bidding strategy, resulting in a 1.42% reduction in the total electricity purchase cost of the power market.

Under the condition of a small time scale (e.g. second), distributed photovoltaic (PV) power generation output has the problems of strongly fluctuating and difficult to accurately simulate. It affects the control strategy and operation mode of hybrid energy systems. To address this problem, a data-driven small-scale distributed PV plant power output model on a 1-second time scale is proposed for the generation of second-by-second PV power output scenarios in hybrid energy systems. Firstly, this work analyzes the characteristics of PV power output at the 1-second time scale based on the probability distribution of power output fluctuations. Secondly, an index system that characterizes the PV power output fluctuation characteristics at the 1-second time scale is constructed. Then, using the data-driven method, a BP neural network model is constructed to simulate the PV power output at the 1-second time scale. Finally, a simulation is performed using the measured data from the PV plant. The findings demonstrate that compared to PV power output models in seconds based on Pearson systematic random numbers: (1) The correlation coefficient (r) of the proposed model is more than 0.8, in a higher degree of fit; (2) The root mean square error (RMSE) of the proposed model achieves 0.005, generally representing a 37.12% reduction. Overall, both the time scale and model accuracy of this model have deep potential value in PV power output modeling and system regulation.

Designing power electronics converters for harsh environments is challenging due to the absence of components' performance under harsh conditions, the frequent transition and data-passing among various software, and the time-consuming and computationally heavy work flow. This paper promotes using design automation to address the aforementioned design challenges. The implementations include public-accessible component databases, automated co-action among circuit simulators and finite element simulations to perform electrical, electromagnetic and thermal co-design, and finally an exclusion-based work flow with hierarchical computation to reduce computational load. The theorized framework is exemplified on designing a real world 175 °C 1.5 kW Three-level Neutral-point-clamped dc-dc converter. A database containing the high-temperature characteristics of SiC MOSFETs and ferrites is established and shared through a web application with graphical user interface. In 310 min, the program, which includes computationally heavy 3D finite element simulation, delivers design output after evaluating the converter's electrical, electromagnetic and thermal performance under 10 million parameter sets. Finally, a 1.5 kW dc-dc converter prototype is built and tested in 175 °C ambient temperature to verify the quality of the design output.

Rolling bearings are essential in the industrial field as a critical component of mechanical systems. Therefore, accurately predicting the remaining useful life of rolling bearings is vital to the safety and reliability of mechanical operation. However, traditional life prediction methods often have problems such as insufficient feature extraction and poor model generalization capabilities, which lead to more significant errors. To solve the above problems, this paper proposes a novel remaining useful life (RUL) prediction method of rolling bearings based on integrated multi-head attention (MHA), improved temporal convolutional network (TCN), and bidirectional long short-term memory (BiLSTM). This method utilizes an improved TCN-BiLSTM network to capture dependencies in sequences and extract global features from signals. In the meantime, MHA is introduced to fully capture the degradation information of the bearing and ultimately predict the life of the bearing. Finally, the bearing life prediction process is fully demonstrated through novel three-dimensional feature visualization. To verify the effectiveness of this method, this paper conducted RUL prediction experiments using the IEEE PHM 2012 dataset and the XJTU-SY dataset, respectively. Many experiments are organized to test the performance, and the experimental results show that this method has higher prediction accuracy and robustness than other methods.

In recent years, “Artificial Intelligence (AI)” has become a focal point of discussion. AI music composition, an interdisciplinary field blending computer science and musicology, has emerged as a prominent area of research. Despite rapid advancements in AI music creation technology, there remains a dearth of comprehensive surveys addressing the core technologies within this domain. To address this gap, this study conducted a comprehensive search across multiple databases spanning a 23-year period (2000–2023) on the topic of “AI music composition.” Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) standard for literature screening, the study systematically organized the development status, frontier hotspots, and technical evaluations of the field. Drawing from literature data, the study verified Price’s Law, Lotka’s Law, and Bradford’s Law—three scientific productivity laws—while summarizing the current landscape from four perspectives: authors, organizations, countries, and journals. Subsequently, utilizing VOSviewer and CiteSpace, two technical software tools, the study conducted an in-depth analysis consisting of four steps: clustering, time zone, burst words, and high-frequency referenced literature. The study presented the evolution trajectory of frontiers and hotspots through visualization maps. Finally, building upon quantitative statistical insights, the study qualitatively expanded research efforts by organizing and evaluating the latest AI music generation algorithm technologies. The systematic literature analysis, both quantitative and qualitative, aims to furnish researchers and practitioners in related fields with systematic references.

This study presents insights into high-speed optical wireless communication (OWC) within plastic pipes, introducing a Gbps-capable alternative for challenging environments. Utilizing a 1W LED with five wavelengths, the experiment explores signal power, attenuation, and bandwidth characteristics. Notably, the blue LED achieves an unprecedented 58.64 MHz bandwidth, red and purple LEDs demonstrate novel bandwidths of approximately 25.23 MHz, and green and yellow LEDs exhibit unique bandwidths of 23.75 MHz and 9.62 MHz, respectively. The attenuation parameters for different wavelengths provide numerous insights, showcasing the novelty of this research and its potential applications in robot communication within plastic pipes. Concurrently, the paper introduces an approach to address the temperature impact on five distinct wavelength LEDs in OWC. By focusing on variations in LED bandwidth and optical power, an optimal heat sink design is proposed. This design achieves a remarkable minimum temperature of 27.06°C and reduces the chip LED device’s response time from 15 to 9 seconds. The significance lies in the novelty of the proposed heat sink, which incorporates variables such as fin thickness, height, air gap width, number of fins, and airflow rate, marking a substantial advancement in thermal management for OWC systems.

Accurately mapping crop cultivation types is essential for the sustainable development of precision agriculture. Environmental restrictions on crop growth, such as soil salinization in arid zones, generally lead to spatial crop growth heterogeneity within cropland fields, which in turn generates differences in the spectral responses reflected in optical remote sensing images of the same croplands and leads to pixel-scale crop-mapping misclassifications. Thus, through this article, we proposed a method to solve this problem at the geoparcel scale by integrating geometric features from a Gaofen-2 high-resolution remote sensing image and the spectral-temporal features derived from Sentinel-2 time series. The results showed that cropland parcels could be accurately extracted from Gaofen-2 images by employing the U-Net semantic segmentation model with an overall accuracy (OA) reaching 97% and a kappa coefficient of 0.95. Then, geoparcel-scale crop types were mapped based on prior crop phenology knowledge and the corresponding Sentinel-2 time series using the time-weighted dynamic time warping (TWDTW) classification algorithm. The parcel-based TWDTW algorithm had an OA of 99.64%, a kappa coefficient of 0.99, and optimal spatial homogeneity in the results, thus outperforming the pixel-based TWDTW method. These results provide a potential solution for mapping crops under spatially heterogeneous cropland conditions affected by various environmental constraints.

Existing research on road safety mainly focuses on normal driving conditions, and seldom considers the impact of poor environment on driving safety. This study focuses on the braking safety of vehicles under sandy road conditions, combines the theories of traffic engineering and cardiophysiology, Through the real vehicle test, the cardiophysiological data of drivers braking under different conditions (including different road alignment, different braking speeds, and different road sand thicknesses) are collected, and the average of the heart rate growth rate and EEG data are analyzed in the process of braking, so as to determine the critical hazardous state of the braking under the condition of sandy roads. The results of the study show that: under different road linear conditions, braking in the combined section of curved slopes is the most dangerous, and the average value of heart rate growth rate of the driver is 26.3%, and the average value of PSD integral value of EEG $\alpha /\beta $ is 0.68, which is in the state of nervousness; under the conditions of different braking speeds, the heart rate growth rate of the driver is more than 27% of the nervousness threshold, and the ratio value of EEG $\alpha /\beta $ wave is less than 1, which is the same as that of EEG $\alpha /\beta $ wave. The ratio of EEG $\alpha $ -wave to $\beta $ -wave is lower than 1, and the degree of alertness is higher; under the condition of different road surface sand thickness, when the thickness of road surface sand is 0-2mm, the average value of the driver’s heart rate growth rate is the maximum of 26.2%, the braking distance is increased obviously, and the $\alpha /\beta $ is reduced significantly, and the driver has a potential safety problem. The results of the study are of great significance to improve the safety of driving on desert highways and reduce the risk of accidents.

Cryogenic propulsion with hydrogen fuel cells replacing fossil fuels is a promising solution to cut carbon emissions in the aviation sector. Hydrogen will also be used for cooling the superconducting machines and power converter circuits. This article aims to test devices suitable for power electronic converters supplying a 1.6 MW superconducting machine. SiC MOSFET and Si IGBT modules with ratings of 1200 V and more than 450 A are selected to assess their performance at different temperatures. Four tests are conducted to determine: 1) the forward voltage drop, 2) the breakdown voltage, 3) the switching behavior, and 4) their operation with two modules in parallel. A bespoke current sensing rig has been developed that avoids the need to extend conductors outside the cryogenic zone in the switching losses measurement test. This configuration introduces minimal stray inductance into the circuit, which minimizes errors in the measurement. One of the aims of this article is to assess the suitability of different module technologies for SiC MOSFET and Si IGBTs in cryogenic applications. Six power modules (SiC MOSFETs and Si IGBTs) were evaluated at both room and cryogenic temperatures. Three of the modules employed conventional bond wire technology, while the other three utilized solid cover (SLC) technology that has no internal bond wires. It was found that the modules which employed SLC technology were the only ones able to survive the extreme temperatures. Following this, a comparison was made between the energy losses of the three SLC modules (two Si IGBTs and one SiC MOSFET) that were able to withstand low temperatures. The results indicated that the performance of the SiC MOSFET module worsens at cryogenic temperatures, whereas the performance of the Si IGBT modules improves with decreasing temperatures. Finally, an inverter simulation was conducted with each of the IGBT modules to estimate the efficiency.

While neural networks have been achieving increasingly significant excitement in solving classification tasks such as natural language processing, their lack of interpretability becomes a great challenge for neural networks to be deployed in certain high-stakes human-centered applications. To address this issue, we propose a new approach for generating interpretable predictions by inferring a simple three-layer neural network with threshold activations, so that it can benefit from effective neural network training algorithms and at the same time, produce human-understandable explanations for the results. In particular, the hidden layer neurons in the proposed model are trained with floating point weights and binary output activations. The output neuron is also trainable as a threshold logic function that implements a disjunctive operation, forming the logical-OR of the first-level threshold logic functions. This neural network can be trained using state-of-the-art training methods to achieve high prediction accuracy. An important feature of the proposed architecture is that only a simple greedy algorithm is required to provide an explanation with the prediction that is human-understandable. In comparison with other explainable decision models, our proposed approach achieves more accurate predictions on a broad set of tabular data classification datasets.

Alzheimer’s disease (AD) is one of the most common neurodegenerative diseases in the world. To reduce the incidence of AD, it’s essential to quantify the AD conversion risk of mild cognitive impaired (MCI) individuals. Here, we propose an AD conversion risk estimation system (CRES), which contains an automated MRI feature extractor, brain age estimation (BAE) module, and AD conversion risk estimation module. The CRES is trained on 634 normal controls (NC) from the public IXI and OASIS cohorts, then it is evaluated on 462 subjects (106 NC, 102 stable MCI (sMCI), 124 progressive MCI (pMCI) and 130 AD) from the ADNI dataset. Experimental results show that the MRI derived age gap (AG, chronological age subtracted from the estimated brain age) significantly distinguish NC, sMCI, pMCI and AD groups with ${p}$ -value $=0.000017$ . Considering AG as the primary factor, incorporating gender and Minimum Mental State Examination (MMSE) for more robust Cox multi-variate hazard analysis, we concluded that each additional year in AG is associated with 4.57% greater AD conversion risk for the MCI group. Furthermore, a nomogram was drawn to describe MCI conversion risk at the individual level in the next 1 year, 3 years, 5 years and even 8 years from baseline. This work demonstrates that CRES can estimate AG based on MRI data, evaluate AD conversion risk of the MCI subjects, and identify the individuals with high AD conversion risk, which is valuable for effective intervention and diagnosis within an early period.

Conventional federated learning (FL) approaches are ineffective in scenarios where clients have significant differences in the distributions of their local data. The Non-IID data distribution in the client data causes a drift in the local model updates from the global optima, which significantly impacts the performance of the trained models. In this article, we present a new algorithm called FLIS that aims to address this problem by grouping clients into clusters that have jointly trainable data distributions. This is achieved by comparing the inference similarity of client models. Our proposed framework captures settings where different groups of users may have their own objectives (learning tasks), but by aggregating their data with others in the same cluster (same learning task), superior models can be derived via more efficient and personalized federated learning. We present experimental results to demonstrate the benefits of FLIS over the state-of-the-art approaches on the CIFAR-100/10, SVHN, and FMNIST datasets. Our code is available at https://github.com/MMorafah/FLIS.

We demonstrate a polarization-insensitive 2 × 2 thermo-optic switch on a 220-nm silicon-on-insulator platform. This device is based on a balanced Mach-Zehnder interferometer with arms consisting of square-cross-section waveguides. In simulations, the Mach-Zehnder switch (MZS) exhibits extinction ratios (ERs) > 9.5 dB for the transverse electric (TE) and transverse magnetic (TM) polarized modes from 1500 nm to 1600 nm. Experimental results verify the broadband operation of the MZS: switching the two polarizations with ERs > 10 dB over an 85-nm wavelength range. Across the C-band (1530 $\sim$ 1565 nm), the fabricated MZS has insertion losses of 0.2 $\sim$ 4.3 dB and ERs > 16 dB (14 dB) at the cross (bar) port for both the TE and TM polarizations. The polarization-dependent losses are around 2 dB within a 100-nm wavelength range, showing the low polarization diversity of the MZS. The proposed MZS is a promising building block for implementing on-chip polarization-division-multiplexed optical interconnects.

Given the detrimental effect of spectral variations in a hyperspectral image (HSI), this article investigates to recover its discriminative representation to improve the classification performance. We propose a new method, namely local low-rank approximation with superpixel-guided locality preserving graph (LLRA-SLPG), which can reduce the spectral variations and preserve the local manifold structure of an HSI. Specifically, the LLRA-SLPG method first clusters pixels of an HSI into several groups (i.e., superpixels). By taking advantage of the local manifold structure, a Laplacian graph is constructed from the superpixels to ensure that a typical pixel should be similar to its neighbors within the same superpixel. The LLRA-SLPG model can increase the compactness of pixels belonging to the same class by reducing spectral variations while promoting local consistency via the Laplacian graph. The objective function of the LLRA-SLPG model can be solved efficiently in an iterative manner. Experimental results on four benchmark datasets validate the superiority of the LLRA-SLPG model over state-of-the-art methods, particularly in cases where only extremely few training pixels are available.

In the direction of arrival (DOA) estimation of sparse array received signals, the estimation accuracy of the grid search method in compressed sensing is improved with the increase of over-complete redundant dictionary elements. However, the increase of over-complete redundant dictionary elements will lead to a significant increase in the computational complexity of this method. In order to reduce the computational complexity caused by over-complete redundant dictionary division, based on the equivalent received signal of large aperture continuous difference co-array generated by sparse array, a DOA estimation method using discrete Fourier transform (DFT) spectrum of signal for initial estimation is proposed in this paper. After obtaining the DFT spectrum of the equivalent signal, based on the correspondence between the DFT spectrum and the actual angle value, this paper proposes a new strategy for dividing the over-complete redundant dictionary. In the process of fine angle search, this paper applies Taylor expansion to orthogonal matching pursuit (OMP) algorithm to obtain higher estimation accuracy. Numerical simulation results demonstrate the advantages of the proposed estimation method over the other methods.

Symmetric Positive Definite (SPD) data are increasingly prevalent in dictionary learning recently. SPD data are the typical non-Euclidean data and cannot constitute a Euclidean space. Therefore, many dictionary learning algorithms cannot be directly adopted on SPD data. Reproducing Kernel Hilbert Spaces (RKHS) is now commonly used to deal with this problem. However, RKHS is an infinite-dimensional Hilbert space, rather than a Euclidean space, resulting in the inability of the dictionary learning to be directly used on SPD data. In this paper, we propose a novel dictionary learning algorithm for SPD data, which is based on the Riemannian Manifold Tangent Space (RMTS). Since RMTS is based on a finite-dimensional Hilbert space, i.e., Euclidean space, most machine learning algorithms developed on Euclidean space can be directly applied to RMTS. The introduction of RMTS provides a better mathematical platform for machine learning of non-Euclidean data. The proposed RMTS method first selects a point on a Riemannian manifold as an anchor point. It transforms the other points on the Riemannian manifold into tangent vectors of the geodesics between these points and the anchor point, with the tangent vectors at the anchor point. We set the length of the tangent vector equal to the length of the geodesic. As a result, such tangent vectors have an explicit geometric meaning, such as direction information, while the RKHS method may cause some geometric meaning to be lost in the original data during the mapping process. In addition, the proposed algorithm adds a regular term of local homomorphism between SPD data and its RMTS dictionary coding coefficients, so that the similarity of SPD data and its RMTS dictionary coding coefficients are as close as possible. Experimental results on four public datasets demonstrate that the proposed algorithm significantly outperforms five state-of-the-art algorithms. This work opens a new pathway towards SPD data dictionary learning methods.

The rapid development of remote sensing technology has led to a sharp increase in the amount of synthetic aperture radar (SAR) measurements, which put forward higher requirements for remote sensing image processing. As an important application of SAR, fast and accurate ship detection has always been a research hotspot. In this article, an improved lightweight RetinaNet for ship detection in SAR images is proposed. Compared with the standard RetinaNet, the shallow convolutional layers of the backbone are replaced by ghost modules and the number of the deep convolutional layers is reduced. The spatial and channel attention modules are embedded into the model to enhance detectability. K-means clustering algorithm is applied to adjust the initial aspect ratios of the model. The effectiveness and robustness of the proposed method is demonstrated by numerical experiments with SSDD dataset, Gaofen-3 mini dataset, and a large-scale SAR image of Hisea-1 satellite, it is shown that the proposed method can significantly reduce the floating-point operations and the number of parameters without decreasing the detection accuracy and recall ratio. Moreover, the experimental results also show the proposed model's robustness and the ability to detect ship targets in small datasets.

This paper reports a pioneering demonstration platform of a cryogenic propulsion unit at liquid nitrogen temperature. A high temperature superconducting (HTS) machine is connected with a cryogenic power rectifier in a generator mode to prove the feasibility of a cryogenic propulsion unit for electric aircraft propulsion applications. Machine operation was carried out with a special focus on the total heat dissipation inside the HTS AC windings. Different types of 2G HTS tapes were tested to provided data for stator coils’ design. The transient operation was carried out to represent a short-circuit failure in one of the power electronic devices. The test shows AC loss performance of the HTS windings using calorimetric method during the short circuit event, indicating the importance of developing protection schemes for the cryogenic propulsion units to prevent damage to the HTS components. The paper provides initial insights into the interaction between superconducting machines and cryogenic power electronics within a cryogenic propulsion unit. It is an important first step to understand and further develop cryogenic propulsion technology for future electric aircraft.

Previous studies on magnetization AC loss of high temperature superconducting (HTS) coated conductors (CCs) taken into account the ideally sinusoidal external magnetic field, or non-sinusoidal by considering only the harmonics in phase with the fundamental component. However, realistically magnetic field often contains harmonics with different phase angles and amplitudes. In this paper, magnetization AC loss in HTS CC was studied, considering the phase angle $\varphi $ of each harmonic; assuming superconductor is subjected to external magnetic field which is distorted by the $3^{\mathrm {rd}}$ and the $5^{\mathrm {th}}$ harmonics. Phase angle of each harmonic was considered in the range of 0 to $\pi $ and amplitude of fundamental magnetic field, was assumed as 10, 20, 50, and 100 mT. Different distortion level of magnetic field was considered too. It is concluded that phase angle of field harmonic has a significant effect on magnetization loss of HTS CCs.

Recently, the design of sparse linear array for direction of arrival (DOA) estimation of non-circular (NC) signals has attracted great attention because the difference and sum co-array offered by non-circularity increases the aperture of virtual linear array. In this paper, we present a coprime array with shifted and flipped sub-array for the DOA of non-circular signals. By shifting one sub-array, the proposed array structure achieves a higher number of consecutive lags than the prototype coprime array with the same number of sensors. Then, through flipping the shifted sub-array with the zero point as the symmetry point, the number of sensor pairs with small separation is significantly reduced, making the resulting structure much sparser. For the proposed array structure, we derive the closed-form expression for the number of consecutive lags, the optimal distribution of two sub-arrays with a given number of sensors and the weight function. Numerical simulations are conducted to verify the superiority of the proposed array over the existing sparse arrays.

Spatial sampling, finite bandwidth, and overlying-strata shielding are three key issues to affect the spatial resolution of seismic imaging for deep targets. Some factors have a great impact on the horizontal resolution, whereas others influence the vertical resolution. How to quantify these effects remains controversial for complex media. Most previous studies on seismic acquisition geometries focus on the horizontal resolution for layered media but neglecting to measure the vertical resolution especially in complex media. Conventional criteria for vertical resolution are based on the theory of geometric seismology with the assumption of a simple medium. As a practical alternative for resolution estimation in complex media, numerical methods with wavefield extrapolation for focal-beam analysis can provide comprehensive insight into the combined effect of acquisition geometries, bandlimited frequencies, and complex media on the horizontal and vertical spatial resolutions of acquisition geometries. We incorporate some classic criteria into the focal-beam numerical analysis to measure the spatial resolutions. Four parameters are used to quantify the performance of acquisition geometries. The horizontal (vertical) resolution is defined as the main-lobe width of a focal beam along the horizontal (vertical) direction, whereas the square root of the peak-to-total ratio of energies is referred to as the horizontal (vertical) sharpness. These parameters describe the horizontal and vertical spatial resolution and sharpness to image the target. Numerical examples with typical acquisition geometries demonstrate the performance of numerical resolution analyses in complex media.

Prediction of Chlorophyll-a (Chl-a) concentration is significant for marine ecology and environmental protection. This paper presents an integrated approach to forecast seasonal Chl-a concentration in coastal waters. Before modeling, feature construction procedures, such as simplification, combination, and normalization, are conducted to identify the potentially significant features. The feature extraction method based on Random Forest (RF) and eXtreme Gradient BOOSTing (XGBoost) is applied to select relevant variables. Then, we propose a Cluster-stacking-based approach which includes a station-oriented clustering model and a stacking-based regression model. The former model is used to divide the observation stations into several groups, thus partitions the study region into several sub-regions and the study dataset into several subsets according to the corresponding stations. In each subset, single regression models including K-Nearest Neighbor (KNN), Support Vector Regression (SVR), Multi-Layer Perceptron regression (MLP) and XGBoost are established in level 0 space and integrated by RF in level 1 space via stacked generalization. We compare the performance of the Cluster-stacking model with that of Cluster-KNN, Cluster-SVR, Cluster-MLP, Cluster-XGBoost and the regression stacking model without cluster. The model evaluation shows that the Cluster-stacking-based approach outperforms others in forecasting Chl-a concentration with a coefficient of determination (R2) of 0.848 and a mean absolute error (MAE) of 0.665 ug/l.

Numerous research studies have been performed to help develop advanced control algorithms for semi-active seat suspension. This paper experimentally investigates a state observer-based Takagi-Sugeno (T-S) fuzzy controller for a semi-active seat suspension by equipping an electrorheological (ER) damper. A new ER damper prototype is designed, assembled, and tested. Then, a T-S fuzzy model is established to describe the ER seat suspension, which can facilitate the H∞ controller design considering the multiobjective optimization. A state observer is established and integrated into the controller to estimate the state information for the T-S fuzzy model in real-time. Additionally, the experimental validation of the control algorithm is critical in the practical application. A seat suspension test rig is built to validate the effectiveness of the proposed controller. The presented control algorithm is evaluated by comparing the corresponding test results to those with a skyhook controller. The experimental results demonstrate that the proposed T-S fuzzy control method, compared to the traditional control method, can further improve the performance of an ER seat suspension system.

As the concept of variable stiffness and variable damping (VSVD) has increasingly drawn attention, suspensions applied with magnetorheological (MR) dampers to achieve varying stiffness and damping have been an attractive method to improve vehicle performance and driver comfort further. As highly nonlinearity of MR damper dynamics and coupled interconnections in the case of multi-output control, to build a direct control system for VSVD suspension based on multiple MR dampers is difficult. Applying Takagi-Sugeno (T-S) fuzzy model on the VSVD system enables the linear control theory to be directly utilized to build the multi-output controller for multi-MR dampers. In this paper, a T-S fuzzy model is established to describe an MR VSVD suspension model, and then an H∞ controller that considers the multi-input/multi-output (MIMO) coupled interconnections characteristic and multi-object optimization is designed. To estimate state information for the T-S fuzzy model in real-time, a state observer is designed and integrated in the controller. Then, the performance of the VSVD control algorithm was evaluated by numerical simulation. The results demonstrate that the T-S fuzzy model-based H∞ controller outperforms the independent control method for a VSVD suspension system with multi-MR dampers.

The use of regional covariance descriptors to generate feature data represented by Symmetric Positive Definite (SPD) matrices from images or videos has become increasingly common in machine learning. However, SPD data itself does not constitute a vector space, and dictionary learning involves a large number of linear operations, so dictionary learning cannot be performed directly on SPD data. For this reason, a more common method is to map the SPD data to the Reproducing Kernel Hilbert Space (RKHS). The so-called kernel learning is to find the most suitable RKHS for specific tasks. RKHS can be uniquely generated by a kernel function. Therefore, RKHS learning can also be considered as kernel learning. In this article, there are two main contributions. The first contribution is to propose a framework which based on Kernel Learning and Riemannian Metric (KLRM). Usually the learnable kernel function framework is to learn some parameters in the kernel function. The second contribution is dictionary learning by applying KLRM to SPD data. The SPD data is transformed into the RKHS generated by KLRM, and RKHS after training provides the most suitable working space for dictionary learning. Under the proposed framework, we design a positive definite kernel function, which is defined by the Log-Euclidean metric. This function can be transformed into a corresponding Riemannian kernel. The experimental results provided in this paper is compared with other state-of-the-art algorithms for SPD data dictionary learning and show that the proposed algorithm achieves better results.

Aiming the problems that the clinical data of different patients is difficult for reasonable representation and the time interval between medical events is different, which lead to the difficulty of clinical prediction, a clinical prediction model based on the long short-term memory (LSTM) network optimized by fruit fly optimization algorithm in health time series data is proposed. First, FastText method is used to represent the interpretable vector of medical events, which can extract the concept relationship rich in medical information more effectively. Then, considering the strong dependence of clinical data on time stamp, LSTM network is used to model clinical events for better extraction of long-term and short-term information, so as to improve the prediction performance of the model. Finally, the fruit fly optimization algorithm is used to find the optimal super parameters of LSTM network, which can improve the training efficiency and prediction precision of the network. Experimental results on MIMIC datasets show that the prediction precision, Recall@k and MAP@k of the proposed model are better than those of other models. The validity of the model is proved.

Users in heterogeneous wireless networks may generate massive amounts of data that are delay-sensitive or require computation-intensive processing. Owing to computation ability and battery capacity limitations, wireless users (WUs) cannot easily process such data in a timely manner, and mobile edge computing (MEC) is increasingly being used to resolve this issue. Specifically, data generated by WUs can be offloaded to the MEC server for processing, which has greater computing power than WUs. However, as the location of MEC servers is fixed, unmanned aerial vehicles (UAVs) have been considered a promising solution in heterogeneous wireless networks. In this study, we design an UAV-assisted computation offloading scheme in an MEC framework with renewable power supply. The proposed model considers the instability of energy arrival, stochastic computation tasks generated by WUs, and a time-varying channel state. Owing to the complexity of the state, it is difficult to use traditional Markov decision process (MDP) with complete prior knowledge for offloading optimization. Accordingly, we propose UAV-assisted computation offloading for MEC based on deep reinforcement learning (UACODRL) to minimize the total cost, which is the weighted sum of the delay, energy consumption, and bandwidth cost. We first use the K-Means algorithm for classification to reduce the dimension of the action space. Subsequently, we use UACODRL to find the near-optimal offloading scheme to minimize the total cost. Simulations demonstrate that UACODRL converges satisfactorily and performs better than four baseline schemes with different parameter configurations.

With the development of 5G technology, the requirements for data communication and computation in emerging 5G-enabled vehicular networks are becoming increasingly stringent. Computation-intensive or delay-sensitive tasks generated by vehicles need to be processed in real time. Mobile edge computing (MEC) is an appropriate solution. Wireless users or vehicles can offload computation tasks to the MEC server due to it has strong computation ability and is closer to the wireless users or vehicles. However, the communication and computation resources of the single MEC are not sufficient for executing the continuously generated computation-intensive or delay-sensitive tasks. We consider migrating computation tasks to other MEC servers to reduce the computation and communication pressure on current MEC server. In this article, we construct an MEC-based computation offloading framework for vehicular networks, which considers time-varying channel states and stochastically arriving computation tasks. To minimize the total cost of the proposed MEC framework, which consists of the delay cost, energy computation cost, and bandwidth cost, we propose a deep reinforcement learning-based computation migration and resource allocation (RLCMRA) scheme that requires no prior knowledge. The RLCMRA algorithm can obtain the optimal offloading and migration policy by adaptive learning to maximize the average cumulative reward (minimize the total cost). Extensive numerical results show that the proposed RLCMRA algorithm can adaptively learn the optimal policy and outperform four other baseline algorithms.

Current is no longer sinusoidal in modern electric networks because of widespread use of power electronic-based equipments and nonlinear loads. Usually AC loss is calculated for pure sinusoidal current, while it is no longer accurate when current is nonsinusoidal. On the other hand, efficiency of cooling system in large scale power devices is dependent on accurate estimation and prediction of the heat load caused by AC loss in design stage. Therefore, estimation of nonsinusoidal AC loss of high temperature superconducting (HTS) material would be of great interest for designers of large-scale superconducting devices. In this paper, at first nonsinusoidal AC loss of a typical HTS tape was calculated under distorted currents using H-formulation finite element method. Then, a range of artificial intelligence (AI) models were implemented to predict AC loss of a typical HTS tape. In order to find the best and more adaptive AI model for nonsinusoidal AC loss prediction, different regression models are evaluated using Support Vector Machine regression model, Generalized Linear regression model, Decision Tree regression model, Feed Forward Neural Network based model, Adaptive Neuro Fuzzy Inference System based model, and Radial Basis Function Neural Network (RBFNN) based model. In order to evaluate robustness of developed models cross-validation technique is implemented on experimental data. To compare the performance of different AI models, four prediction measures were used: Theil's U coefficients (U_Accuracy and U_Quality), Root Mean Square Error (RMSE) and Regression value (R-value). Obtained results show that best performance belongs to RBFNN based model and then ANFIS based model. U coefficients and RMSE values are obtained less than 0.005 and R-Value is become close to one by using RBFNN based model for testing data, which indicates high accuracy prediction performance.

Due to the target motion, range cell migration (RCM) and Doppler frequency migration (DFM) always occur. That is harmful to the signal enhancement and detection. In order to solve the problem, a novel three-dimensional (3-D) coherent integration (TDCI) based algorithm is proposed in this article which consists of three stages. Firstly, a 3-D space is generated by the autocorrelation function. After that, TDCI algorithm is realized and TDCI domain is obtained in which the motion parameters can be accurately estimated. Finally, compensating off the RCM and DFM by the estimates, the target signal can be accumulated and detected in range-Doppler frequency domain. Theoretical analyses and simulation experiments are given to demonstrate that the proposed algorithm is able to deal with the problems of velocity ambiguity, shadow effect, and cross-term with superior resolution. Comparisons with several representative algorithms lead us to the conclusion that the proposed algorithm can strike a good balance between computation cost and anti-noise performance. In the end, real measured data processing and result analysis are carried out, which further verify the effectiveness of the proposed algorithm.

In this paper, we apply deep reinforcement learning (DRL) to solve the flocking control problem of multi-robot systems in complex environments with dynamic obstacles. Starting from the traditional flocking model, we propose a DRL framework for implementing multi-robot flocking control, eliminating the tedious work of modeling and control designing. We adopt the multi-agent deep deterministic policy gradient (MADDPG) algorithm, which additionally uses the information of multiple robots in the learning process to better predict the actions that robots will take. To address the problems such as low learning efficiency and slow convergence speed of the MADDPG algorithm, this paper studies a prioritized experience replay (PER) mechanism and proposes the Prioritized Experience Replay-MADDPG (PER-MADDPG) algorithm. Based on the temporal difference (TD) error, a priority evaluation function is designed to determine which experiences are sampled preferentially from the replay buffer. In the end, the simulation results verify the effectiveness of the proposed algorithm. It has a faster convergence speed and enables the robot group to complete the flocking task in the environment with obstacles.

Smart home scheduling, facilitated by advanced metering, monitoring, and manipulation technology, plays an important role in the energy transition in terms of accommodating intermittent renewable energy and improving energy consumption efficiency. The key functionalities of home energy scheduling are usually implemented by leveraging the flexibility of household appliances, such as thermostatically controlled loads (TCLs) and energy storage units, to improve the peak-to-average ratio for utilities and reduce energy bills for customers. However, the consumption patterns of appliances are greatly influenced by a variety of factors, including real-time tariffs, ambient temperature profiles, indoor activities, and solar irradiance. Hence, smart home energy scheduling is a challenging task because most of these impacting factors are stochastic and difficult to predict. To properly model and manage the uncertainty factors associated with smart home appliance scheduling, an economic model predictive control (MPC)-based bilevel smart scheduling scheme is proposed in this paper. The comprehensive modeling of distributed generation and household appliances is performed at the single-household level. The home energy scheduling problem is formulated on two levels, with the upper level emphasizing the economic impact and the lower level focusing on capturing TCL responses. The correlations among different TCLs and their performance under the influence of various uncertainty factors, such as environmental impacts and user behaviors, are considered. The efficiency of the proposed MPC-based bilevel optimization model and the effectiveness of the home energy scheduling strategy in managing uncertainties are validated and illustrated in numerical studies.

The epicardial fat plays a key role in the development of many cardiovascular diseases. It is necessary and useful to precisely segment this fat from CT scans in clinical studies. However, it is not feasible to manually segment this fat in clinical practice, as the workload and cost for technicians or physicians is quite high. In this work, we propose a novel method for automatic segmentation and quantification of epicardial fat from CT scans accurately. In detail, dual U-Nets with the morphological processing layer is used for this goal. The first network is based on the U-Net framework to detect the pericardium, before segmenting its inside region. A morphological layer is concatenated as the following layer of the first network, to refine and obtain the ideal inside region of the pericardium. While the second network is also applied using U-Net as its backbone to find and segment the epicardial fat of the processed inside region from the pericardium using the first network. Our proposed method obtains the highest mean Dice similarity (91.19%), correlation coefficient (0.9304) compared to other state-of-art methods on a cardiac CT dataset with 20 patients. The results indicate our proposed method is effective for quantifying epicardial fat automatically.

The increasing penetration of highly intermittent wind generation could seriously jeopardize the operation reliability of power systems and increase the risk of electricity outages. To this end, this paper proposes a novel data-driven method for operation risk assessment of wind-integrated power systems. Firstly, a new approach is presented to model the uncertainty of wind power in lead time. The proposed approach employs k-means clustering and mixture models (MMs) to construct time-dependent probability distributions of wind power. The proposed approach can also capture the complicated statistical features of wind power such as multimodality. Then, a non-sequential Monte Carlo simulation (NSMCS) technique is adopted to evaluate the operation risk indices. To improve the computation performance of NSMCS, a cross-entropy based importance sampling (CE-IS) technique is applied. The CE-IS technique is modified to include the proposed model of wind power. The method is validated on a modified IEEE 24-bus reliability test system (RTS) and a modified IEEE 3-area RTS while employing the historical data of wind generation. The simulation results verify the importance of accurate modeling of short-term uncertainty of wind power for operation risk assessment. Further case studies have been performed to analyze the impact of transmission systems on operation risk indices. The computational performance of the framework is also examined.

Mining a document structure from multiple data sources in terms of their underlying topics has become an important task of document clustering. The traditional document clustering approach cannot be applied directly to the multi-source document clustering problem. There are three typical difficulties: 1) The topics of different data sources are related but not the same. 2) Usually, each data source has its own focus on topics. 3) The number of clusters of the data sources are not necessarily the same and are not known beforehand. In this paper, based on our previous research, we design a novel multi-source document clustering model, namely, the hierarchical Dirichlet multinomial allocation (HDMA) model, to solve all the above problems. The HDMA model is investigated with a two-step hierarchical topic generation process. Topics are learnt to share their general characteristics across data source, while at the same time preserve the local characteristics of the data source. Each data source is applied with an exclusive topic partition to learn the source-level topic emphasis. A Gibbs sampling algorithm is then used to learn the number of clusters for each data source as well as the parameters of the HDMA model at the same time. Experimental results demonstrate that the HDMA model is effective.

A trench-gate field stop insulated gate bipolar transistor (TFS-IGBT) with a novel hybrid p+/n collector structure is proposed to enhance the trade-off relationship between the on-state voltage drop (Von) and the turn-off energy loss (Eoff). The proposed hybrid collector structure consists of a p+/n layer between the p± collector and the field stop layer. During turn-on, the p+ regions in the hybrid p+/n layer provide high carrier injection efficiency. During turn-off transient, the n-regions in the hybrid p+/n collector provide fast carrier extraction paths for holes. 2-D numerical simulations comparing a conventional TFS-IGBT, an injection efficiency controlled IGBT with the proposed hybrid p+/n collector TFS-IGBT (HC-IGBT) having similar device structures show that for a 1.2 kV rating and with Von at 1.5 V, the HC-IGBT has an advantage in Eoff reduction by 69% and 22%, respectively. Finally, backside mask alignment is not needed for fabricating the proposed hybrid p+/n collector structure.

Direction-of-arrival (DOA) estimation in coprime array involves the problem that there is the coexistence of both uncorrelated signals and coherent signals in the multipath environment. An approach to estimate DOA of uncorrelated and coherent signals separately, based on the coprime linear array and coprime planar array, is proposed in this paper. The uncorrelated signals are estimated first, where the root multiple signal classification (root-MUSIC) is applied in the coprime linear array and unitary estimating signal parameters via rotational invariance techniques (Unitary-ESPRIT) is utilized in the coprime planar array. Those subspace algorithms are of low complexity compared with the spectral peak search method. We then eliminate the components of noises and uncorrelated signals and reconstruct a covariance matrix of coherent signals. Finally, we use the root-MUSIC or Unitary-ESPRIT to resolve the one-dimensional or two-dimensional DOAs of coherent signals, respectively. The simulation results demonstrate the computational efficiency and high accuracy of the proposed algorithm. The results also prove that this algorithm can estimate the number of signals more than that of subarray sensors and separate two signals from the same angle.

In this paper, a novel compact wide-frequency tunable filter with switchable bandpass and bandstop frequency response is presented. By employing two PIN diodes between the source and load port, the bandpass and bandstop filtering response can be switched, respectively. Three varactor diodes have been used to tuning the operating frequency of the tunable filter at bandpass and bandstop frequency response. The even- and odd-mode method is applied to analyze the presented tunable filter. When PIN diode is forward-biased, the filter has the bandstop response, and the filter has bandpass response while it is reversed. To verify the proposed tunable filter circuits, the tunable bandpass filter and bandstop filter are designed, fabricated, and measured. It is demonstrated that the center frequency of the fabricated tunable filter is tuned from 0.95 to 1.35 GHz. The measured insertion loss in the band is greater than 13 and 11 dB for bandstop mode and bandpass mode, respectively. Moreover, the fabricated tunable filter has a compact size of 0.24λg × 0.24λg (λg is the wavelength of the center frequency during the tunable frequency range).

A nuclear spin co-magnetometer operated in the spin-exchange relaxation-free (SERF) regime is a promising tool for long-term navigation application for its abilities to sense inertial rotation and to suppress environmental magnetic field disturbance. The magnetic field response model of a K-Rb-21Ne nuclear spin co-magnetometer is derived based on the state space method and the model is experimentally validated on a tabletop SERF co-magnetometer. The theoretical and experimental results indicate that the relaxation rate of the nuclear spins limits the field-suppression ability. The results here not only give insight into the nature of self-compensate characteristic but also provide a precise model for the estimation of the magnetic noise-induced rotation measurement error in a SERF co-magnetometer.

Currently, blockchain technology has been widely researched and is being applied to many other fields, such as financial transactions, E-government, logistics, and supply-chain system. It can be used to store and maintain transaction data through the de-centralized model. In this paper, we propose an autonomous resource request transaction framework based on blockchain in a social network, in which all kinds of resources in the social community can be traded through blockchain technology. When a user needs to acquire some resources from a community, the user may make a transaction with the members from the community through blockchain technology while the members autonomously negotiate each other to reach an agreement. The proposed framework provides an incentive mechanism to encourage community members to disseminate the resources through a smart contract.

The sensitivity of a dual-axis K-Rb-21Ne co-magnetometer is subject to the accuracy of the magnetic field compensation and the calibration coefficient, which are determined by the orthogonality of the coil system. In this paper, an ingenious method of measuring the non-orthogonal angles between the x- and z-axes coils and that between the y- and z-axes coils is demonstrated. This in situ measurement method is based on the magnetic field responses of the dual-axis co-magnetometer. The non-orthogonal angle can be obtained by solving the relevant trigonometric function of the geometric relationship. Finally, the nonorthogonal angles of about 7.93° between the z- and y-axes coils and 7.39° between the z- and x-axes coils are obtained by fitting the response formulas. Besides, the influence of the applied transverse magnetic field on the accuracy of the non-orthogonal angle calibration is analyzed. This method is also verified to be valid at different working temperatures.

Deep Neural Networks (DNNs) provide state-of-the-art results for most machine learning and computer vision tasks. However, they have been found susceptible to adversarial examples. In the recent literature, many ways of generating adversarial examples have been discovered. In this work, we propose a novel method to generate adversarial examples with generative adversarial networks (GANs). Compared to traditional optimisation-based methods, our method provides a fast yet powerful alternative for adversary generation. Unlike other GAN-based approaches in the literature which learn to generate an intermediate perturbation vector, our method generates adversarial examples from benign input images in a straightforward manner. By directly generating adversarial examples from given input images, our method produces perturbations that better align with the underlying edge and shape contained in the inputs, hence more natural-looking and imperceptible to human eyes. We evaluate our method on the MNIST and the CIFAR-10 dataset and demonstrate that it outperforms the state-of-the-art GAN-based attack AdvGAN with similar attack capability in terms of distortion. We show that our method produces competitive results to notable optimisation-based attacks in the literature including the strongest Carlini & Wagner (CW) attack.

In this paper, the angle-of-arrival (AOA) measurements are adapted to locate a target using the UAV swarms equipped with passive receivers. The measurement noise is considered to be target-to-receiver distance dependent. The Cramer-Rao low bound (CRLB) of the AOA localization is calculated, and the optimal deployments are explored through changing angular separations and distances. Then, a distributed collaborative autonomous generation (DCAG) method is proposed based on the deep neural network (NN). The off-line training and on-line application rules are applied to generate the optimal heading angles for the UAV swarms in the AOA localization. The simulation results show that through the DCAG method, the generated heading angles for UAV swarms enhance the localization accuracy and stability.

This paper studies the power control in a two-hop decode-and-forward relay system over fading channel. The relay node is an energy harvesting node and is equipped with a data buffer and a rechargeable battery with finite capacity. Under the condition that any prior and statistical knowledge about the harvested energy and channel fading is unknown, based on the Lyapunov optimization framework, an online power control strategy for the source node and the relay node is proposed to maximize the average transmission rate and minimize the average energy consumption of the source node. First, the optimization problem is decomposed into two sub-problems: minimizing the average energy consumption of the source node and maximizing the average transmission rate of the relay node. Then, we use the Lyapunov optimization framework to transform the minimization and maximization problems to the minimization of the upper bounds of two “drift-plus-penalty”s and solve them. The solution of the optimization problem only relies on the current channel coefficient, the state of the data buffer, the energy level of the battery in the current time slot, and the average transmission rate of the past time slots. Only a few calculations are needed to get the optimal power of the source node and the relay node. Finally, simulation is done to verify the performance of the proposed algorithm. The simulation results show that the transmission rate difference between the proposed algorithm and the ideal offline water filling algorithm is small, which has full knowledge of energy and channels in advance and is without the constraints of the causality of energy and data, and the overflow of the battery and data buffer. The proposed algorithm is greatly superior to the online greedy algorithm and the semi-greedy algorithm.

Different from other human behaviors, sign language has the characteristics of limited local motion of upper limb and meticulous hand action. Some sign language gestures are ambiguous in RGB video due to the influence of lighting and background color, which affects the recognition accuracy. We propose a multimodal deep learning architecture for sign language recognition which effectively combines RGB-D input and two-stream spatiotemporal networks. Depth videos, as an effective compensation of RGB input, can supply additional distance information about the signer's hands. A novel sampling method called ARSS (Aligned Random Sampling in Segments) is put forward to select and align optimal RGB-D video frames, which improves the capacity utilization of multimodal data and reduces the redundancy. We get the hand ROI by joints information of RGB data for local focus in spatial stream. D-shift Net is proposed as depth motion feature extraction in temporal stream, which fully utilizes three dimensional motion information of the sign language. Both streams are fused by convolutional fusion layer to get complementary features. Our approach explored the multimodal information and enhanced the recognition precision. It obtains the state-the-of-art performance on the datasets of CSL (96.7%) and IsoGD (63.78%).

The transient process of pumped storage system (PSS) experienced pump mode, turbine mode and switching mode in pumped outage condition, spanning hump zone and S-shaped region. The conflicting dynamic indicators make it difficult to meet the requirements of the transient process. How to specify optimal operating strategies and improve the reliability of PSS in the pumped outage condition has become a key issue. This problem is addressed in this work based on coordinated multi-objective optimization, and three stages are included: nonlinear modeling, strategy optimization and decision making. A trade-off is embodied in each stage. Firstly, the real-time accurate equivalent circuit model is applied for PSS. The spatiotemporal discrete analysis is presented to enable the trade-off between computing efficiency and model accuracy in stage 1. Then, the neighborhood-search chaotic mutation multi-objective gravitational search algorithm, integrating multiple constraints, is performed to solve the Pareto front. It achieves the trade-off between algorithm convergence and searching diversity in stage 2. Further, fuzzy analytic hierarchy process is innovatively introduced to select the most compatible solution, which comprehensively considers the quantitative and qualitative factors. It accomplishes the trade-off in conflicting objectives and indicators in stage 3. Compared with the on-site operation in pumped outage condition, the maximum reversing speed decreased from 545 rpm to 161.45 rpm, the maximum water pressure of volute can be reduced by 4.55%, and the minimum draft tube pressure can be improved by 6.94%. The coordinated multi-objective optimization provides a novel technical guidance for the safe and stable operation of PSS.

Direction-of-arrival (DOA) estimation using nested linear array ignores the information from repeated sensors, thus involves the problem of losing accuracy. Moreover, non-circular feature of signals is rarely considered and it results in discontinuity of the virtual array, namely, holes appear. This paper first provide an improved data averaging method to increase the accuracy of DOA estimation by fusing data from covariance and elliptic covariance. And then an algorithm based on matrix completion theory in nested array that greatly extends the degrees of freedom (DOF) and is able to find more sources than number of physical sensors is presented. The algorithm reconstructs the covariance matrix of the virtual linear array, which exactly has the same shift invariance as the uniform linear array but a higher aperture. Considering about the missing elements of the virtual array covariance matrix, we apply matrix completion theory to solve the problem. Finally, true DOAs of multiple signals can be obtained through subspace algorithms. Numerical results demonstrate that the proposed algorithms can obtain high accuracy while underdetermined DOA estimation is realized.

A compensation method for a magnetoelectric velocity sensor (MVS) is always necessary, which can lower the resonance frequency of the measuring system and subsequently extend the measuring bandwidth. In this paper, a novel compensation method is proposed based on the BP neural network under the TensorFlow architecture. Comparing with the existing methods, the new method does not depend upon an accurate model of the MVS any more, whose parameters are badly influenced by the temperature. The dynamic compensator is connected with the sensor. The BP neural network algorithm is used to identify compensation parameters. The dynamic compensator works at state of the optimum parameter all the time to compensate the dynamic performance of MVS by training the weights and thresholds of the neural network. The experiment results show that velocity measurement deviation is within ±5% error band by the dynamic compensator, which can reduce the measurement deviation caused by the variation of temperature and improve the measurement accuracy. The bandwidth can be as low as 0.28Hz. The dynamic compensator is superior to Random Forests and RBF Neutral Network in implement in FPGA/CPLD. Its' accuracy is superior to zero-pole compensation method. This method leads a new way to weaken the temperature variation characteristics of the velocity sensor and improve the measurement performance.

Source localization based on the hybrid time-difference-of-arrival (TDOA) and frequency-difference-of-arrival (FDOA) measurements from distributed sensors is an essential problem in wireless sensor networks (WSNs). In this paper, we mainly study the optimal sensor deployment and velocity configuration of UAV swarms mounted with TDOA and FDOA based sensors. Explicit solutions of optimal sensor deployment and velocity configuration are acquired in both static and movable source scenarios based on the Fisher information matrix (FIM). Both centralized and decentralized localization are explored to meet different types of localization methods. Path planning problem of UAV swarms in TDOA/FDOA localization is also studied with constraints. Simulations verify its efficiency with path planning in TDOA and FDOA localization.

Reconnaissance mission has a wide application in both civil and military fields, which provides intelligence and basis for the following decision-making to accomplish certain goals. Due to numerous advantages of UAV swarms such as strong flexibility, high efficiency, and low cost, conducting reconnaissance missions by UAV swarms has become a trend of future. However, the path planning problem of UAV swarms is a key challenge in the aspect of model construction, algorithm, selection and high computational complexity, especially when the mission is complicated. In this paper, various distributed particle swarm optimization (DPSO)-based path planning algorithms are proposed for UAV swarms conducting a reconnaissance mission, in which targets are gathered in the form of clusters and different tactic needs are taken into consideration. Three algorithms named the maximum density convergence DPSO algorithm (MDC-DPSO), the fast cross-over DPSO algorithm (FCO-DPSO), and the accurate coverage exploration DPSO algorithm (ACE-DPSO) are proposed correspond to the needs of fast convergence, random cross-over, and accurate search, respectively. Different fitness functions and search strategies are specifically designed considering the mobility and communication constraints of the UAV swarms. Besides, the jump-out mechanism and revisit mechanism are designed to save invalid search efforts and avoid falling into local optimum. The simulation results demonstrate that the proposed algorithms are effective in generating paths for UAV swarms conducting a reconnaissance mission, which can be easily applied to large scale swarms.