Search

This paper is devoted to real-time optimal attitude reorientation control of rigid spacecraft control. Particularly, two typical practical problems - actuator misalignment and forbidden pointing constraints are considered. Within the framework of adaptive dynamic programming (ADP), a novel constrained optimal attitude control scheme is proposed. In this design, a special reward function is developed to characterize the environment feedback and deal with the pointing constraints. Notably, a novel argument term is introduced to the reward function for overcoming the inevitable difficulty in actuator misalignment. By virtue of the Lyapunov stability theory, the ultimate boundedness of state error and the optimality of the proposed method can be guaranteed. Finally, the effectiveness and performance of the developed ADP-based controller are evaluated by not only numerical simulations but also experimental tests with a hardware-in-loop platform.

Most of the classification methods of urban functional areas nowadays are only based on single source data analysis and modeling, which can not make full use of the multi-scale and multi-source data that is easy to obtain. Therefore, this paper proposed a classification model of urban functional areas based on multi-modal machine learning, by analyzing regional remote sensing images and behavior data of visitors in the area, using the combination of supervised methods extracted the deep-seated features and relationships of kinds of data, filtered and merged the overall and local features of the data. The model used dual branch neural network combining SE-ResNeXt and Dual Path Network (DPN) to automatically mined and fused the overall characteristics of multi-source data, and used the designed feature engineering to deeply mine the behavior data of users to obtain more association information, then combined the algorithm based on Gradient Boosting Decision Tree to learn the characteristics of different levels and obtained the classification probability for different levels of features. Finally, we continued to use the algorithm based on the Gradient Boosting Decision Tree to learn the probability distribution of different levels of features to obtain the final prediction results of urban functional area classification. Through the analysis and experimental verification of real data sets, the results showed that MM-UrbanFAC model can effectively integrate the features of multi-modal data. Compared with a single classifier, the integration framework based on gradient lifting tree improved the prediction performance, this method can effectively integrate the results of multiple models and accurately classify urban functional areas, and the model can provide reference for tourism recommendation, urban land planning and urban construction.

Neighborhood reconstruction is proved effective for dimensionality reduction because of the preservation of manifold structure. Conventional neighborhood preserving embedding (NPE) method first learns the affinity relationship or reconstruction relationship in the original space, and then learns the projection matrix to preserve the learned local information in low-dimensional space. However, the pre-learned manifold information may be inaccurate due to the noises and irrelevant features in real-world data. The performance of dimensionality reduction would be influenced as well. Besides, NPE and its variants only aim to preserve the local reconstruction relationship but ignore the fuzzy membership relationship between samples and cluster prototypes. To address these issues, we propose an adaptive neighborhood preserving discriminant projection model, where sparse reconstruction coefficients are updated in the process of dimensionality reduction to eliminate the influence of noises and irrelevant features. Meanwhile, we also learn the fuzzy membership relationships between data points and cluster prototypes to gather the samples belonging to the same class together in low-dimensional space. Neighborhood reconstruction learning and clustering are seamlessly connected in the learned subspace. To solve this model, an iterative algorithm is developed. The experimental results of recognition accuracy show the superiorities of the proposed methods over the state-of-the-arts.

Fuzzy K-Means (FKM) clustering is of great importance for analyzing unlabeled data. FKM algorithms assign each data point to multiple clusters with some degree of certainty measured by the membership function. In these methods, the fuzzy membership degree matrix is obtained based on the calculation of the distance between data points in the original space. However, this operation may lead to suboptimal results because of the influence of noises and redundant features. Besides, some FKM clustering methods ignore the importance of the weighting exponent. In this paper, we propose a novel FKM method called Fuzzy K-Means Clustering With Discriminative Embedding. Within this method, we simultaneously conduct dimensionality reduction along with fuzzy membership degree learning. To retain most information in the embedding subspace and improve the robustness of this method, principal component analysis is incorporated into our framework. An iterative optimization algorithm is proposed to solve the model. To validate the efficacy of the proposed method, we perform comprehensive analyses, including convergence behavior, parameter determination and computational complexity. Moreover, we also match a appropriate weighting exponent for each data set. Experimental results on benchmark data sets show that the proposed method is more discriminative and effective for clustering tasks.

Previous deep learning methods usually required large‐scale annotated data, which is computationally exhaustive and unrealistic in certain scenarios. Therefore, few‐shot learning, where only a few annotated training images are available for training, has attracted increasing attention these days, showing huge potential in practical applications, such as portable equipment or security inspection, and so on. However, current few‐shot learning methods usually neglect the valuable semantic correlations between samples, thereby failing in extracting discriminating relations to achieve accurate predictive results. In this work, extending on a recent state‐of‐the‐art few‐shot learning method, transductive relation‐propagation network (TRPN), which considers the correlations between training samples, a constrained relation‐propagation network is proposed to further regularise the distilled correlations and thus achieve favourable few‐shot classification performance. The proposed framework contains three main components, namely preprocess module, relational propagation module, and relation constraint module. First, sample features are extracted and a relation graph node is constructed by treating the relation of each support–query pair as a graph node in the preprocess module. After that, in the relation propagation module (RPM), the valuable information of support–query pairs is modelled and propagated to directly generate the relational representations for further prediction. Then, a relation constraint module is introduced to regularise the relational representations and make it consistent with the ground‐truth relations as much as possible. With the guidance of the effective RPM and relation constraint module, the relational representations of the support–query pairs are distinguishable and thus can achieve accurate predictive results. Comprehensive experiments conducted on widely used benchmarks validate the effectiveness of our method compared to state‐of‐the‐art few‐shot classification approaches.

Background Several computational tools for predicting protein Ubiquitylation and SUMOylation sites have been proposed to study their regulatory roles in gene location, gene expression, and genome replication. However, existing methods generally rely on feature engineering, and ignore the natural similarity between the two types of protein translational modification. This study is the first all-in-one deep network to predict protein Ubiquitylation and SUMOylation sites from protein sequences as well as their crosstalk sites simultaneously. Our deep learning architecture integrates several meta classifiers that apply deep neural networks to protein sequence information and physico-chemical properties, which were trained on multi-label classification mode for simultaneously identifying protein Ubiquitylation and SUMOylation as well as their crosstalk sites. Results The promising AUCs of our method on Ubiquitylation, SUMOylation and crosstalk sites achieved 0.838, 0.888, and 0.862 respectively on tenfold cross-validation. The corresponding APs reached 0.683, 0.804 and 0.552, which also validated our effectiveness. Conclusions The proposed architecture managed to classify ubiquitylated and SUMOylated lysine residues along with their crosstalk sites, and outperformed other well-known Ubiquitylation and SUMOylation site prediction tools.

In this article, a deep reinforcement learning (RL)-based control approach with enhanced learning efficiency and effectiveness is proposed to address the wind farm control problem. Specifically, a novel composite experience replay (CER) strategy is designed and embedded in the deep deterministic policy gradient (DDPG) algorithm. CER provides a new sampling scheme that can mine the information of stored transitions in-depth by making a tradeoff between rewards and temporal difference (TD) errors. Modified importance-sampling weights are introduced to the training process of neural networks (NNs) to deal with the distribution mismatching problem induced by CER. Then, our CER-DDPG approach is applied to optimizing the total power production of wind farms. The main challenge of this control problem comes from the strong wake effects among wind turbines and the stochastic features of environments, rendering it intractable for conventional control approaches. A reward regularization process is designed along with the CER-DDPG, which employs an additional NN to handle the bias of rewards caused by the stochastic wind speeds. Tests with a dynamic wind farm simulator (WFSim) show that our method achieves higher rewards with less training costs than conventional deep RL-based control approaches, and it has the ability to increase the total power generation of wind farms with different specifications.\ud

Lane marking detection is a fundamental yet challenging task for traffic scene understanding. Previous works generally predict traffic line segmentation and obtain lane marking coordinates with a post-processing step, which are vulnerable to various challenging environments including occlusion, illumination variation, shadow, background clutter. In this paper, we propose a novel method, named as Lane marking Regression Network (LRN), which can simultaneously consider confidence area detection and field inference for producing more precise lane markings in an unified encoder-decoder framework. For obtaining confidence detection of lane area, we introduce the graph diffusion mechanism to aggregate the contextual information for better detecting hard areas. The field inference considers the geometric relationship between lane area and lane markings by adopting a novel constraint regression function. Comprehensive evaluations on CULane and GD datasets well demonstrate the significant superiority of our proposed LRN over other state-of-the-arts for lane marking detection.

Neighborhood reconstruction is a good recipe to learn the local manifold structure. Representation-based discriminant analysis methods normally learn the reconstruction relationship between each sample and all the other samples. However, reconstruction graphs constructed in these methods have three limitations: 1) they cannot guarantee the local sparsity of reconstruction coefficients; 2) heterogeneous samples may own nonzero coefficients; and 3) they learn the manifold information prior to the process of dimensionality reduction. Due to the existence of noise and redundant features in the original space, the prelearned manifold structure may be inaccurate. Accordingly, the performance of dimensionality reduction would be affected. In this article, we propose a joint model to simultaneously learn the affinity relationship, reconstruction relationship, and projection matrix. In this model, we actively assign neighbors for each sample and learn the inter-reconstruction coefficients between each sample and their neighbors with the same label information in the process of dimensionality reduction. Specifically, a sparse constraint is employed to ensure the sparsity of neighbors and reconstruction coefficients. The whitening constraint is imposed on the projection matrix to remove the relevance between features. An iterative algorithm is proposed to solve this method. Extensive experiments on toy data and public datasets show the superiority of the proposed method.

Owing to the increasing penetration of renewable power generation, the modern power system faces great challenges in frequency regulations and reduced system inertia. Hence, renewable energy is expected to take over part of the frequency regulation responsibilities from the gas or hydro plants and contribute to the system inertia. In this article, we investigate the feasibility of frequency regulation by the offshore hydrostatic wind turbine (HWT). The simulation model is transformed from NREL (National Renewable Energy Laboratory) 5-MW gearbox-equipped wind turbine model within FAST (fatigue, aerodynamics, structures, and turbulence) code. With proposed coordinated control scheme and the hydrostatic transmission configuration of the HWT, the `continuously variable gearbox ratio' in turbulent wind conditions can be realised to maintain the constant generator speed, so that the HWT can be connected to the grid without power converters in-between. To test the performances of the control scheme, the HWT is connected to a 5-bus grid model and operates with different frequency events. The simulation results indicate that the proposed control scheme is a promising solution for offshore HWT to participated in frequency response in the modern power system.\ud

Pegylated recombinant human granulocyte colony-stimulating factor (PEG-rhG-CSF, pegfilgrastim) is a long-acting derivative of recombinant human granulocyte colony-stimulating factor with limited renal clearance and a longer half-life. It is used for the prevention of febrile neutropenia, owing to its capacity to promote neutrophil recovery. In this study, the pharmacokinetics, pharmacodynamics, safety, and immunogenicity of 2 formulations of PEG-rhG-CSF were evaluated in healthy Chinese subjects. Twenty-four male subjects who received a single dose of subcutaneous PEG-rhG-CSF 100 µg/kg were randomized to either treatment A (3 mg/mL) or treatment B (1 mg/mL). Noncompartmental pharmacokinetic parameters of PEG-rhG-CSF were derived from serum concentration-time data. In addition, absolute neutrophil count (ANC) as a pharmacodynamic maker, immunogenicity through antidrug antibody testing, and safety were evaluated. The mean area under the concentration-time curve from time zero to the last quantifiable concentration (AUC0-t ) and the mean maximum concentration (Cmax ) of PEG-rhG-CSF after treatment A were 5070 ng·h/mL and 125 ng/mL, respectively; these values were comparable to those measured after treatment B (5340 ng·h/mL and 123 ng/mL, respectively). The mean value of area under the △ANC (baseline-adjusted ANC)-time curve and the maximum △ANC values were 4380 × 109 h/L and 33.1 × 109 /L, respectively, in the treatment A group, and 5170 × 109 h/L and 38.6 × 109 /L, respectively, in the treatment B group. The pharmacokinetic and pharmacodynamic profiles were similar for the 2 PEG-rhG-CSF formulations following a single dose of 100 µg/kg. The safety and immunogenicity profiles were also similar, with no significant differences. The dose adjustment of PEG-rhG-CSF was not considered necessary for formulation transformation.

This paper investigates the effects power oscillation damping (POD) controller could have on a wind turbine structural system. Most of the published work in this area has been done using relatively simple aerodynamic and structural models of a wind turbine which cannot be used to investigate the detailed interactions between electrical and mechanical components of the wind turbine. Therefore, a detailed model that combines electrical, structural and aerodynamic characteristics of a grid-connected Doubly Fed Induction Generator (DFIG) based wind turbine has been developed by adapting the NREL (National Renewable Energy Laboratory) 5 MW wind turbine model within FAST (Fatigue, Aerodynamics, Structures, and Turbulence) code. This detailed model is used to evaluate the effects of POD controller on the wind turbine system. The results appear to indicate that the effects of POD control on the WT structural system are comparable or less significant as those caused by wind speed variations. Furthermore, the results also reveal that the effects of a transient three-phase short circuit fault on the WT structural system are much larger than those caused by the POD controller.

BACKGROUND Patellar tendon rupture is a rare disease, and reports regarding patellar tendon reconstruction with ligament augmentation reconstruction system (LARS) ligaments are limited, with only three reports available in the literature. LARS ligaments are made of polyethylene terephthalate and have been certified as a more favorable option than other tendon transplants. To our knowledge, this is the first report of patellar tendon reconstruction with LARS for suture fixation due to poor quality of the tendon after multiple operations to enable early mobilization and quick rehabilitation. CASE SUMMARY A 65-year-old woman had limited ability in extending her leg and an inability to perform a straight leg raise after multiple operations due to patella fracture. The patient underwent patellar tendon reconstruction with LARS artificial ligaments. After 12 mo of follow-up, the patient was able to perform a straight leg raise, and the incision healed well without complications. The Lysholm score was 95 and the range of motion of the knee was 0-130°. CONCLUSION This study revealed that patellar tendon reconstruction with LARS artificial ligaments is possible in a patient with a patellar tendon rupture who required rapid postoperative recovery.

Linear discriminant analysis (LDA) is the most widely used supervised dimensionality reduction approach. After removing the null space of the total scatter matrix St via principal component analysis (PCA), the LDA algorithm can avoid the small sample size problem. Most existing supervised dimensionality reduction methods extract the principal component of data first, and then conduct LDA on it. However, "most variance" is very often the most important, but not always in PCA. Thus, this two-step strategy may not be able to obtain the most discriminant information for classification tasks. Different from traditional approaches which conduct PCA and LDA in sequence, we propose a novel method referred to as joint principal component and discriminant analysis (JPCDA) for dimensionality reduction. Using this method, we are able to not only avoid the small sample size problem but also extract discriminant information for classification tasks. An iterative optimization algorithm is proposed to solve the method. To validate the efficacy of the proposed method, we perform extensive experiments on several benchmark data sets in comparison with some state-of-the-art dimensionality reduction methods. A large number of experimental results illustrate that the proposed method has quite promising classification performance.

This paper proposes a reinforcement learning (RL)-based six-degree-of-freedom (6-DOF) control scheme for the final phase proximity operations of spacecraft. The main novelty of the proposed method are from two aspects: 1) the closed-loop performance can be improved in real-time through the RL technique, achieving an online approximate optimal control subject to the full 6-DOF nonlinear dynamics of spacecraft; 2) Nontrivial motion constraints of proximity operations are considered and strictly obeyed during the whole control process. As a stepping stone, the dual-quaternion formalism is employed to characterize the 6-DOF dynamics model and motion constraints. Then, an RL-based control scheme is developed under the dual-quaternion algebraic framework to approximate the optimal control solution subject to a cost function and a Hamilton-Jacobi-Bellman equation. In addition, a specially designed barrier function is embedded in the reward function to avoid motion constraint violations. The Lyapunov-based stability analysis guarantees the ultimate boundedness of state errors and the weight of NN estimation errors. Besides, we also show that a PD-like controller under dual-quaternion formulation can be employed as the initial control policy to trigger the online learning process. The boundedness of it is proved by a special Lyapunov strictification method. Simulation results of prototypical spacecraft missions with proximity operations are provided to illustrate the effectiveness of the proposed method.

This paper investigates the optimal power generation control for a two-stage horizontal-axis tidal turbine system based on backstepping disturbance rejection control (BDRC), which is a new control framework for high-order nonlinear systems. The tidal turbine system is designed with the main structure being described. The dynamics of the tidal turbine system is then formulated based on the integration of the dynamics of its constitute components. The tidal turbine experiences large uncertainties and unknown dynamics from non-uniform operating thrust and fatigue forces, variations and turbulence in tidal flow velocities induced by waves and wind. The proposed BDRC is a unique control concept which has superior performance in dealing with these large uncertainties without requiring much information about the turbine dynamics. Consequently, the BDRC system is synthesized in two control loops for the control of the turbine dynamics (outer loop) and the q-axis current dynamics (inner loop) respectively to track the optimal tidal turbine speed and hence maintain the optimal power generations. Both the stability and convergence of the two closed control loops are subsequently analyzed and proved. The simulations are conducted in MATLAB/Simulink to compare the performance of the developed BDRC system with a sliding mode control method used in the literature. In addition, the proposed BDRC algorithm is extended to a general nonlinear strict-feedback system with high uncertainties and external/internal disturbances. The proposed BDRC has significantly extended the traditional ADRC and does not need any differential operations, thereby totally eliminating the inherent problems of “explosion of complexity” and repeated differentiations of virtual control variables in traditional backstepping control.

In this work, a physics-informed deep learning model is developed to achieve the reconstruction of the three-dimensional (3-D) spatiotemporal wind field in front of a wind turbine, by combining the 3-D Navier–Stokes equations and the scanning LIDAR measurements. To the best of the authors’ knowledge, this is for the first time that the full 3-D spatiotemporal wind field reconstruction is achieved based on real-time measurements and flow physics. The proposed method is evaluated using high-fidelity large eddy simulations. The results show that the wind vector field in the whole 3-D domain is predicted very accurately based on only scalar line-of-sight LIDAR measurements at sparse locations. Specifically, at the baseline case, the prediction errors for the streamwise, spanwise and vertical velocity fields are 0 . 263 m/s, 0 . 397 m/s and 0 . 361 m/s, respectively. The prediction errors for the horizontal and vertical direction fields are 2 . 84 ° and 2 . 58 ° which are important in tackling yaw misalignment and turbine tilt control, respectively. Further analysis shows that the 3-D wind features are captured clearly, including the evolutions of flow structures, the wind shear in vertical direction, the blade-level speed variations due to turbine rotation, and the speed variations modulated by the turbulent wind. Also, the developed model achieves short-term wind forecasting without the commonly-used Taylor’s frozen turbulence hypothesis. Furthermore it is very useful in advancing other wind energy research fields e.g. wind turbine control & monitoring, power forecasting, and resource assessments because the 3-D spatiotemporal information is important for them but not available with current sensor and prediction technologies.

Motivation: Discovering long noncoding RNA (lncRNA)–disease associations is a fundamental and critical part in understanding disease etiology and pathogenesis. However, only a few lncRNA–disease associations have been identified because of the time-consuming and expensive biological experiments. As a result, an efficient computational method is of great importance and urgently needed for identifying potential lncRNA–disease associations. With the ability of exploiting node features and relationships in network, graph-based learning models have been commonly utilized by these biomolecular association predictions. However, the capability of these methods in comprehensively fusing node features, heterogeneous topological structures and semantic information is distant from optimal or even satisfactory. Moreover, there are still limitations in modeling complex associations between lncRNAs and diseases. Results: In this paper, we develop a novel heterogeneous graph attention network framework based on meta-paths for predicting lncRNA–disease associations, denoted as HGATLDA. At first, we conduct a heterogeneous network by incorporating lncRNA and disease feature structural graphs, and lncRNA–disease topological structural graph. Then, for the heterogeneous graph, we conduct multiple metapath-based subgraphs and then utilize graph attention network to learn node embeddings from neighbors of these homogeneous and heterogeneous subgraphs. Next, we implement attention mechanism to adaptively assign weights to multiple metapath-based subgraphs and get more semantic information. In addition, we combine neural inductive matrix completion to reconstruct lncRNA–disease associations, which is applied for capturing complicated associations between lncRNAs and diseases. Moreover, we incorporate cost-sensitive neural network into the loss function to tackle the commonly imbalance problem in lncRNA–disease association prediction. Finally, extensive experimental results demonstrate the effectiveness of our proposed framework.

Glutarylation is a type of post-translational modification that occurs on lysine residues. It plays an irreplaceable role in various cellular functions. Therefore, identification of glutarylation sites is significant for understanding the molecular mechanism of glutarylation. In this study, we proposed a method named DEXGB_Glu to identify lysine glutarylation sites using XGBoost as classifier which was optimized by differential evolution algorithm. Aiming at the imbalance between positive samples and negative samples, Borderline-SMOTE method was employed to synthesize positive samples, increasing their amount equal to negative samples. Then, Tomek links technique was applied to filter out noise data. Analysis of this method and its results showed that differential evolution algorithm obviously improved the performance and the combination of Borderline-SMOTE and Tomek links effectively solved the imbalance between positive samples and negative samples. Finally, the performance of this method was much better than other methods in prediction of glutarylation sites. The data and code are available on https://github.com/ningq669/DEXGB_Glu.

The brain–computer interface (BCI) control technology that utilizes motor imagery to perform the desired action instead of manual operation will be widely used in smart environments. However, most of the research lacks robust feature representation of multi-channel EEG series, resulting in low intention recognition accuracy. This article proposes an EEG2Image based Denoised-ConvNets (called EID) to enhance feature representation of the intention recognition task. Specifically, we perform signal decomposition, slicing, and image mapping to decrease the noise from the irrelevant frequency bands. After that, we construct the Denoised-ConvNets structure to learn the colorspace and spatial variations of image objects without cropping new training images precisely. Toward further utilizing the color and spatial transformation layers, the colorspace and colored area of image objects have been enhanced and enlarged, respectively. In the multi-classification scenario, extensive experiments on publicly available EEG datasets confirm that the proposed method has better performance than state-of-the-art methods.

Lysine carboxylation is one of the most crucial type of post-translation modification, which plays a significant role in catalytic mechanisms. Therefore, it is essential to study lysine carboxylation and explore its biological mechanism. Compared with traditional experimental methods that are labor-intensive and time-consuming, computational methods are much more convenience and faster. Therefore, it is urgent to establish an accurate carboxylation identification model. Herein we proposed a method, named pQLyCar for identification of lysine carboxylation using SVM as classifier. In pQLyCar, a peptide-based dynamic query-driven sample rescaling strategy (pDQD-SR) is proposed to address the class imbalance of training data, which builds a specific prediction model for each query sample. KNN algorithm calculates distance between samples according to original sequences instead of feature vectors. Information entropy is applied to select optimal size of sliding window and various types of sequence- and position-based features are incorporated for construction of feature space, including residues composition (RC), K-space and position-special amino acid propensity (PSAAP). Finally, the performance of pQLyCar is measured with a specificity of 96.49% and a sensibility of 99.59% using jackknife test method, which indicated that pQLyCar method can be a useful tool for prediction of lysine carboxylation sites.

Intelligent transportation systems (ITS) have been developed rapidly over the last few decades because of global urbanization and industrialization. ITS involve a wide range of different technologies and applications such as automatic road enforcement, dynamic traffic light sequence, and as a result, a significant number of scientific papers have been published in the field of ITS. In this paper, we present a useful insight into the development of ITS area by systematically analyzing the publications over the period of 20 years. First, we identify the most cited papers and most impactful authors in the field. Second, in the aspect of topic analysis, we identify some active keywords. To do so, we develop a keyword co-occurrence network to find topics in the ITS field. Finally, for the collaboration pattern analysis, we construct two networks to interpret collaboration patterns, including a co-authorship network, and an author co-keyword network to show the development and research tendency of ITS. Some most interesting findings from our investigation include the following: 1) Besides the USA, China and Europe have begun to play an increasingly significant role in this field and 2) GPS , traffic control , and road safety show an upward trend from the analysis of the evolution of ITS research topics, given the rise of new research areas such as autonomous vehicles. Our first-hand investigation and analysis of the literature provides a valuable reference to research activities in the development of ITS field and presents worthy insights on the current status and future technical trends.

Although the conventional active appearance model (AAM) has achieved some success for face alignment, it still suffers from the generalization problem when be applied to unseen subjects and images. To deal with the generalization problem of AAM, we first reformulate the original AAM as sparsity-regularized AAM, which can achieve more compact/better shape and appearance priors by selecting nearest neighbors as the bases of the shape and appearance model. To speed up the fitting procedure, the sparsity in sparsity-regularized AAM is approximated by using the locality (i.e., K-nearest neighbor), and thus inducing the locality-constrained active appearance-model (LC-AAM). The LC-AAM solves a constrained AAM-like fitting problem with the K-nearest neighbors as the bases of shape and appearance model. To alleviate the adverse influence of inaccurate K-nearest neighbor results, the locality constraint is further embedded in the discriminative fitting method denoted as LC-DFM, which can find better K-nearest neighbor results by employing shape-indexed feature, and can also tolerate some inaccurate neighbors benefited from the regression model rather than the generative model in AAM. Extensive experiments on several datasets demonstrate that our methods outperform the state-of-the-arts in both detection accuracy and generalization ability.

Optical coherence tomography (OCT) employs near-infrared light to image the microstructure of different tissues. Clinically, it has been used to image the walls of coronary arteries. In research settings, one of the applications for OCT is visualizing endocardial and subendocardial structures. The present experiment sought to determine whether OCT can identify native conduction tissues in adult porcine hearts. During the study, the right atrial endocardial surfaces of excised adult porcine hearts were exposed. The triangle of Koch was imaged with the OCT system and the conduction tissue was identified. The area was then prepared for histologic examination with Masson's trichrome stain. The results of histologic preparations and OCT images were then compared. Ultimately, nine porcine hearts were examined using this methodology. OCT imaging successfully identified subendocardial structures presumed to be the compact atrioventricular node. Histologic images of the preparations delineated the different tissue types and conduction tissue was easily identified. The location of distinctive hyporeflective areas in the OCT images correlated with the location of conduction tissue in the histology images. In light of the findings of this study, it is suggested that atrioventricular nodal tissue can be identified by OCT in freshly dissected unfixed porcine hearts. OCT images distinguished the differentiated conduction tissue in close proximity with the endocardium, myofibers, and fibrous tissue, and the success of this was verified with histology. This technology may be useful for the direct visualization of the native conduction system during procedures in the operating room and electrophysiology laboratory. Further studies with perfused tissue samples and live animal experiments are needed to better assess the efficacy of this novel application.

Background Acetylation on lysine is a widespread post-translational modification which is reversible and plays a crucial role in some biological activities. To better understand the mechanism, it is necessary to identify acetylation sites in proteins accurately. Computational methods are popular because they are more convenient and faster than experimental methods. In this study, we proposed a new computational method to predict acetylation sites in human by combining sequence features and structural features including physicochemical property (PCP), position specific score matrix (PSSM), auto covariation (AC), residue composition (RC), secondary structure (SS) and accessible surface area (ASA), which can well characterize the information of acetylated lysine sites. Besides, a two-step feature selection was applied, which combined mRMR and IFS. It finally trained a cascade classifier based on SVM, which successfully solved the imbalance between positive samples and negative samples and covered all negative sample information. Results The performance of this method is measured with a specificity of 72.19% and a sensibility of 76.71% on independent dataset which shows that a cascade SVM classifier outperforms single SVM classifier. Conclusions In addition to the analysis of experimental results, we also made a systematic and comprehensive analysis of the acetylation data.

The language history questionnaire (LHQ) is an important tool for assessing the linguistic background and language proficiency of multilinguals or second language learners. Previously we developed a generic LHQ based on the most commonly asked questions in published studies (Li, Sepanski & Zhao, 2006) and provided a web-based interface (LHQ 2.0) that has flexibility in functionality, accuracy in data recording, and privacy for users and data (Li, Zhang, Tsai & Puls, 2014). LHQ3 (version 3) introduces new functions, developed in response to many comments/requests from users. One important improvement is the addition of an automatic scoring system, in that the new interface automatically calculates aggregated scores for language proficiency, language dominance, and language immersion levels. Finally, LHQ3 allows researchers to assign different weights to the modules when calculating the aggregated scores, addressing the issue of different focuses that different researchers put on multilingual speakers’ language usage and background.

N6-methyladenosine (m6A) is the one of the most important RNA modifications, playing the role of splicing events, mRNA exporting and stability to cell differentiation. Because of wide distribution of m6A in genes, identification of m6A sites in RNA sequences has significant importance for basic biomedical research and drug development. High-throughput laboratory methods are time consuming and costly. Nowadays, effective computational methods are much desirable because of its convenience and fast speed. Thus, in this article, we proposed a new method to improve the performance of the m6A prediction by using the combined features of deep features and original features with extreme gradient boosting optimized by particle swarm optimization (PXGB). The proposed PXGB algorithm uses three kinds of features, i.e., position-specific nucleotide propensity (PSNP), position-specific dinucleotide propensity (PSDP), and the traditional nucleotide composition (NC). By 10-fold cross validation, the performance of PXGB was measured with an AUC of 0.8390 and an MCC of 0.5234. Additionally, PXGB was compared with the existing methods, and the higher MCC and AUC of PXGB demonstrated that PXGB was effective to predict m6A sites. The predictor proposed in this study might help to predict more m6A sites and guide related experimental validation.

Continuously variable speed wind turbines (CVSWTs) without using fixed speed gearboxes continue being as viable options for modern wind power industry. CVSWTs have an infinite number of transmission ratios in contrast to specific gear ratios of a standard wind turbine. This transmission flexibility allows the generator to maintain a constant speed regardless of highly turbulent wind speeds. The development of CVSWTs has made great progress in recent years due to the high demand for renewable energy sources. Therefore, this paper serves to provide a systematic review of the CVSWTs including mechanical, hydraulic and hydro-mechanical type CVSWTs. The operating principles, design topologies and basic characteristics are described and analyzed. The comparisons and new trends of the CVSWTs are also discussed, which indicate the feasibility of applying CVSWTs in wind power industry and the high potential of using hydro-viscous transmission in megawatt CVSWTs.

An in-line fiber Mach–Zehnder interferometer (MZI) for refractive index (RI) sensing was proposed, and the interferometer was fabricated by cascading two peanut-shape structures in the single mode fiber using the commercial fusion splicer. A model of the peanut-shape profile was used to calculate the power distribution of the modes in the device. The RI sensing performances of the peanut-shape structure in-line fiber MZIs with two different arm length (30 and 100 mm) were investigated. As the sensor with a short arm length of 30 mm, the RI sensitivity of − 67 . 355 nm/RIU was achieved for the RI range of 1.33–1.37. As the sensor with a long arm length of 100 mm, the simultaneous RI and temperature measurement was realized to eliminate the temperature and RI cross sensitivity, and the highest RI and temperature sensitivities of − 67 . 953 nm/RIU and 0.073 nm/°C were obtained, respectively.

Geothermal energy is a renewable resource that is constantly available. The low geothermal well operating lifetime is the main challenge in using this type of renewable energy. This problem can be covered by the aid of solar system (hybrid system). For complicated renewable energy systems, finding the optimum design parameters and operating conditions require to develop experimental apparatus or sophisticated thermodynamic models. Hence, in this study, artificial intelligence (AI) approach is proposed for modeling the geothermal organic Rankin cycle (GORC) equipped with solar thermal unit. Indeed, the current study depicts how AI methods can meticulously simulate the operation of a complicated renewable energy system. The developed intelligent methods are adaptive neuro-fuzzy inference system (ANFIS) optimized with particle swarm optimization (PSO) algorithm (ANFIS-PSO) and multilayer perceptron (MLP) neural network optimized with PSO algorithm (MLP-PSO). The models are composed based on the main design parameters of the geothermal system that are solar radiation, well temperature, working fluid mass flow rate, turbine output pressure, surface area of the solar collector and preheater inlet pressure. The intelligent models use the mentioned input variables to predict the net power output, energy efficiency, exergy efficiency and levelized cost of energy (LCOE) of the GORC. Energy, exergy and economic analyses are carried out for the low global warming potential (GWP) refrigerants. It was found out that although the intelligent models can meticulously predict the targets, ANFIS-PSO performs better than MLP-PSO for modeling the GORC with solar system. Root mean square error of this model for prediction of power generation, energy efficiency, exergy efficiency and LCOE was 12.023 (kW), 3.587 × 10 - 4 , 3.278 × 10 - 4 and 1.332 × 10 - 4 , respectively.

Person Re-Identification (person re-ID) is an image retrieval task which identifies the same person in different camera views. Generally, a good person re-ID model requires a large dataset containing over 100000 images to reduce the risk of over-fitting. Most current handcrafted person re-ID datasets, however, are insufficient for training a learning model with high generalization ability. In addition, the lacking of images with various levels of occlusion is still remaining in most existing datasets. Motivated by these two problems, this paper proposes a new data augmentation method called Random Linear Interpolation that can enlarge the sizes of person re-ID datasets and improve the generalization ability of the learning model. The key enabler of our approach is generating fused images by interpolating pairs of original images. In other words, the innovation of the proposed approach is considering data augmentation between two random samples. Plenty of experimental results demonstrates that the proposed method is effective to improve baseline models. On Market1501 and DukeMTMC-reID datasets, our approach can achieve 92.71% and 82.19% rank-1 accuracy, respectively.

We transform the National Renewable Energy Laboratory (NREL) 5-MW geared equipped monopile wind turbine model into a hydrostatic wind turbine (HWT) by replacing its drivetrain with a hydrostatic transmission (HST) drivetrain. Then, we design an $\mathcal {H}_{\infty }$ loop-shaping torque controller (to regulate the motor displacement) and a linear parameter varying (LPV) blade pitch controller for the HWT. To enhance the performances of the pitch control system during the transition region around the rated wind speed, we add an antiwindup (AW) compensator to the LPV controller, which would otherwise have had undesirable system responses due to pitch saturation. The LPV AW pitch controller uses the steady rotor effective wind speed as the scheduling parameter, which is estimated by the light detection and ranging preview. The simulations based on the transformed NREL 5-MW HWT model show that our torque controller achieves a very good tracking behavior, while our pitch controller (no matter with or without AW) gets much improved overall performances over a gain-scheduled PI pitch controller.

This article explores the deep neural learning (DNL) based predictive control approach for offshore wind farm using high-fidelity large eddy simulations (LES) data. The DNL architecture is defined by combining the long short-term memory (LSTM) units with convolutional neural networks (CNN) for feature extraction and prediction of the offshore wind farm. This hybrid CNN-LSTM model is developed based on the dynamic models of the wind farm and wind turbines as well as higher fidelity LES data. Then, distributed and decentralized model predictive control (MPC) methods are developed based on the hybrid model for maximizing the wind farm power generation and minimizing the usage of the control commands. Extensive simulations based on a two-turbine and a nine-turbine wind farm cases demonstrate the high prediction accuracy (97% or more) of the trained CNN-LSTM models. They also show that the distributed MPC can achieve up to 38% increase in power generation at farm scale than the decentralized MPC. The computational time of the distributed MPC is around 0.7 s at each time step, which is sufficiently fast as a real-time control solution to wind farm operations.

Wake effects impose significant aerodynamic interactions among wind turbines. To improve the wind farm operating performance, practical wind farm online control considering wake effects becomes very important. To achieve online optimal wind farm control while responding to grid demands, this paper proposes a novel optimal wind farm supervisory control (SC) model and its explicit solutions. From the controller modelling perspective, the two major wind farm operating modes, the maximum power point tracking mode and the set‐point tracking mode, are first analysed and unified in one optimisation model while considering wake effects. In this way, wind farm power production and rotor kinetic energy reserve can be simultaneously considered to conveniently modify the operation mode in response to different grid demands. Aside from controller modelling, the collocation method is first introduced to address the online application problem of such wake‐effect aware optimal WF control. Although a few optimisation algorithms have been proposed to find the optimum offline, online optimal control is still challenging because of the computational complexity brought by wake model non‐linearity and non‐convexity. The proposed collocation method explicitly approximates the optimal solutions to the proposed supervisory control model, through which only a direct algebraic operation is required for online optimal control instead of repeated optimisations. Case studies are carried out on different wind farms under various wind conditions, showing that the wind farm power production potential and releasable power reserve are improved compared to traditional greedy control in both modes. The accuracy of the collocation method is verified. A detailed analysis of the wind farm production capacity under different wind speeds and directions is also provided.

Pulmonary vein isolation with radiofrequency ablation (RFA) has become the most common procedure to treat Atrial Fibrillation (AF). However, current RFA lesion formation is guided only with indirect information (e.g. temperature, impedance, contact force), which does not guarantee transmural lesions. Non-transmural lesions are understood to contribute to AF recurrence. We have previously demonstrated that polarization-sensitive optical coherence tomography (PSOCT) can monitor RFA lesion transmurality in the left atrium (LA) of living swine. However, it requires expert interpretation. Here, we demonstrate quantitative image quality and lesion transmurality evaluation metrics applied to the in vivo LA RFA PSOCT lesion monitoring data.

Transseptal puncture (TSP) is commonly conducted under the guidance of fluoroscopy and/or intracardiac echocardiography (ICE) at the fossa ovalis (FO) to gain percutaneous access to the left atrium for intracardiac procedures. Issues with traditional TSP include: additional vascular access through a sheath, and fluoroscopy exposes patients to ionizing radiation. TSP, if not done appropriately can result in serious complications. We studied the feasibility of optical coherence tomography (OCT) guidance of TSP with ex vivo and in vivo experiments. Results show that OCT can provide detailed structure information to identify FO allowing for safe TSP.

Periprosthetic joint infections (PJIs) have become the most catastrophic complication for patients after arthroplasty. Although previous studies have found that many biomarkers have good performance for diagnosing PJI, early diagnosis remains challenging and a gold standard is lacking. This study aimed to investigate the diagnostic accuracy of plasma fibrinogen (FIB) in detecting PJI compared to other traditional biomarks (CRP, WBC and ESR). A total of 156 patients (including 57 PJI and 99 non-PJI patients) who underwent revision arthroplasty were retrospectively reviewed from 01/2014 to 01/2020. The diagnostic criteria of PJI were mainly based on the definition from the evidence-based definition for periprosthetic joint infection in 2018. The optimal plasma FIB predictive cutoff was 4.20 g/L, the sensitivity of the plasma fibrinogen was 0.860, the specificity was 0.900, the positive predictive value (PPV) was 0.831, and the negative predictive value (NPV) was 0.908. The area under the curve (AUC) value of plasma fibrinogen was 0.916 (95% CI 0.869–0.964), and the CRP, ESR and WBC levels had AUCs of 0.901, 0.822 and 0.647, respectively. Plasma FIB demonstrated better diagnostic strength compared with that of other serum biomarkers before revision arthroplasty. It represents a new horizon for the diagnosis of PJI due to the diagnosis values and cost-effective features.

The motor imagery brain-computer interface uses the human brain intention to achieve better control. The main technical problems are feature representation and classification of signal features for specific thinking activities. Inspired by the structure and function of the human brain, we construct a neural computing model to explore the critical issues in the representation and real-time recognition of the state of specific thinking activities. In consideration of the physiological structure and the information processing process of the brain, we construct a multi-scale cascaded Conv-GRU model and extract high-resolution feature information from the dual spatio-temporal dimension, effectively removing signal noise, improving the signal-to-noise ratio, and reducing information loss. Extensive experiments demonstrate that our model has a low dependence on training data size and outperforms state-of-the-art multi-intention recognition methods.

The vortex force map (VFM) method is a promising low-cost technique for predicting unsteady aerodynamic load from the vorticity field. Developed under potential flow assumption and applied to high-resolution computational fluid dynamics (CFD) data, the original VFM method needs to be further explored for viscous flows, especially at low Reynolds numbers and with incomplete knowledge of the flow data. In this paper, the VFM method is extended to general viscous flows, with both forces contributed from vorticity in the flowfield and on the body surface considered. We find that viscosity contributes to not only the skin-friction force but also a pressure force term. We also show that the extended vortex pressure force maps for general flow problems without the potential flow assumption lead to vanishing contributions from the far-field. The extended VFM method is applied to CFD flowfields for the starting flow of a cylinder and a NACA0012 airfoil, and it is used to analyze the force contribution effect of a given vortex. Moreover, acceptable accuracy of the method on truncated domains and under coarse sampling of typical particle image velocimetry measurement size is demonstrated.

Feature selection aims to select a subset of features to optimize the performances of downstream predictive tasks. Recently, multi-agent reinforced feature selection (MARFS) has been introduced to automate feature selection, by creating agents for each feature to select or deselect corresponding features. Although MARFS enjoys the automation of the selection process, MARFS suffers from not just the data complexity in terms of contents and dimensionality, but also the exponentially-increasing computational costs with regard to the number of agents. The raised concern leads to a new research question: Can we simplify the selection process of agents under reinforcement learning context so as to improve the efficiency and costs of feature selection? To address the question, we develop a single-agent reinforced feature selection approach integrated with restructured choice strategy. Specifically, the restructured choice strategy includes: 1) we exploit only one single agent to handle the selection task of multiple features, instead of using multiple agents. 2) we develop a scanning method to empower the single agent to make multiple selection/deselection decisions in each round of scanning. 3) we exploit the relevance to predictive labels of features to prioritize the scanning orders of the agent for multiple features. 4) we propose a convolutional auto-encoder algorithm, integrated with the encoded index information of features, to improve state representation. 5) we design a reward scheme that take into account both prediction accuracy and feature redundancy to facilitate the exploration process. Finally, we present extensive experimental results to demonstrate the efficiency and effectiveness of the proposed method.

In this paper, a reliability aware multi-objective predictive control strategy for wind farm based on machine learning and heuristic optimizations is proposed. A wind farm model with wake interactions and the actuator health informed wind farm reliability model are constructed. The wind farm model is then represented by training a relevance vector machine (RVM), with lower computational cost and higher efficiency. Then, based on the RVM model, a reliability aware multi-objective predictive control approach for the wind farm is readily designed and implemented by using five typical state of the art meta-heuristic evolutionary algorithms including the third evolution step of generalized differential evolution (GDE3), the multi-objective evolutionary algorithm based on decomposition (MOEA/D), the multi-objective particle swarm optimization (MOPSO), the multi-objective grasshopper optimization algorithm (MOGOA), and the non-dominated sorting genetic algorithm III (NSGA-III). The computational experimental results using the FLOw Redirection and Induction in Steady-state (FLORIS) and under different inflow wind speeds and directions demonstrate that the relative accuracy of the RVM model is more than 97%, and that the proposed control algorithm can largely reduce thrust loads (by around 20% on average) and improve the wind farm reliability while maintaining similar level of power production in comparison with a conventional predictive control approach. In addition, the proposed control method allows a trade-off between these objectives and its computational load can be properly reduced.

Reliable prediction of wind turbine wakes is essential for the optimal design and operation of wind farms. In order to achieve this, the parameter uncertainty of analytical wake model is investigated for the first time. Specifically, large eddy simulations (LES) of wind farms are carried out with different turbine yaw angles, based on SOWFA (Simulator fOr Wind Farm Applications) platform. The generated high-fidelity flow field data is used to infer the low-fidelity model’s parameters within the Bayesian uncertainty quantification framework. After model calibration, the posterior model check shows that the predicted mean velocity profile with the quantified uncertainty matches well with the high-fidelity CFD data. The prediction of other quantities, such as wind farm flow field and turbine power generation, is also carried out. The results show that the wake model with the model parameters specified by their posterior distributions can be seen as the stochastic extension of the original wake model. As most of the existing wake models are static, the resulting stochastic model shows a great advantage over the original model, as it can give not only the static wind farm properties but also their statistical distributions.\ud \ud

This paper studies the data-driven structural control of monopile wind turbine towers based on machine learning approach, by using an active tuned mass damper (TMD) located in the nacelle. The adaptive dynamic programming (ADP) approach is employed to obtain the optimal controller which is derived on the modern large-scale machine learning platform Tensorflow. The proposed network structure includes three simple three-layer neural networks (NNs), i.e. a plant network, a critic network, and an action network. The plant network is used to capture the fully nonlinear dynamics of the structural system while the action network is used to approximate the optimal controller. Their training requires the gradient information flowing through the whole network. The automatic differentiation is used in this paper for all the gradient derivations, which greatly improves the employed ADP algorithm’s ability in solving complex practical problems. The simulation results of structural control of monopile turbine towers show that on average the active TMD achieves 15% performance improvement on tower fatigue load reduction over a passive TMD, with small active power consumption (less than 0.24% of the turbine’s nominal power production). Besides, the controller design considers the trade-off between control performance and power consumption.

The structural control of floating wind turbines using active tuned mass damper is investigated in this article. To our knowledge, this is for the first time that reinforcement learning-based control approach is employed to this type of application. Specifically, an adaptive dynamic programming (ADP) algorithm is used to derive the optimal control law based on the nonlinear structural dynamics, and the large-scale machine learning platform Tensorflow is employed for the design and implementation of the neural network (NN) structure. Three fully connected NNs, i.e., a plant network, a critic network, and an action network, are included in the proposed NN structure. Their training requires the gradient information flowing through the whole network, which is tackled by automatic differentiation, a popular technique for deriving the gradients of complex networks automatically. While to our knowledge, the network structures in the existing literature are rather simple and the training of the hidden layer is usually ignored. This allows their gradients to be derived analytically, which is infeasible with complex network structures. Thus, automatic differentiation greatly improves the employed ADP algorithm's ability in solving complex problems. The simulation results of structural control of floating wind turbines show that ADP controller performs very well in both normal and extreme conditions, with the standard deviation of the platform pitch displacement being reduced by around 40%. A clear advantage of ADP controllers over the H∞ controller is observed, especially in extreme conditions. Moreover, our design considers the tradeoff between the control performance and power consumption.\ud

Atrial fibrillation (AF) is the most common sustained arrhythmia in the western world. Pulmonary vein isolation (PVI) using radiofrequency ablation (RFA) is frequently conducted to treat AF. However, current PVI procedures for lesion formation are guided only with indirect information, which may lead to non-transmural lesions, and contribute to the high recurrence of AF. Therefore, direct lesion quality feedback may potentially improve PVI efficacy. To study the real-time direct guidance capability of polarization sensitive optical coherence tomography (PSOCT), a custom-designed integrated PSOCT-RFA catheter was prototyped and tested in RFA procedures in the left atria of living swine.

Ubiquitylation, a typical post-translational modification (PTM), plays an important role in signal transduction, apoptosis and cell proliferation. A ubiquitylation like PTM, sumoylation also may affect gene mapping, expression and genomic replication. Over the past two decades, machine learning has been widely employed in protein ubiquitylation and sumoylation site prediction tools. These existing tools require feature engineering, but failed to provide general interpretable features and probably underutilized the growing amount of data. This prompted us to propose a deep learning-based model that integrates multiple convolution and fully-connected layers of seven supervised learning sub-models to extract deep representations from protein sequences and physico-chemical properties (PCPs). Especially, we divided PCPs into 6 clusters and customized deep networks accordingly for handling the high correlations among one cluster. A stacking ensemble strategy was applied to combine these deep representations to make prediction. Furthermore, with the advantage of transfer learning, our deep learning model can work well on protein sumoylation site prediction as well after fine-tuning. On the high-quality annotated database Swiss-Prot, our model outperformed several well-known ubiquitylation and sumoylation site prediction tools. Our code is freely available at https://github.com/ruiwcoding/DeepUbiSumoPre.

Background Lysine succinylation is a new kind of post-translational modification which plays a key role in protein conformation regulation and cellular function control. To understand the mechanism of succinylation profoundly, it is necessary to identify succinylation sites in proteins accurately. However, traditional methods, experimental approaches, are labor-intensive and time-consuming. Computational prediction methods have been proposed recent years, and they are popular because of their convenience and high speed. In this study, we developed a new method to predict succinylation sites in protein combining multiple features, including amino acid composition, binary encoding, physicochemical property and grey pseudo amino acid composition, with a feature selection scheme (information gain). And then, it was trained using SVM (Support Vector Machine) and an ensemble learning algorithm. Results The performance of this method was measured with an accuracy of 89.14% and a MCC (Matthew Correlation Coefficient) of 0.79 using 10-fold cross validation on training dataset and an accuracy of 84.5% and a MCC of 0.2 on independent dataset. Conclusions The conclusions made from this study can help to understand more of the succinylation mechanism. These results suggest that our method was very promising for predicting succinylation sites. The source code and data of this paper are freely available athttps://github.com/ningq669/PSuccE. Electronic supplementary material The online version of this article (10.1186/s12859-018-2249-4) contains supplementary material, which is available to authorized users.

We propose to make use of the hydraulic reservoir of a floating barge hydrostatic wind turbine (HWT) to suppress the pitch and roll motions of the barge by making the reservoir into a shape of an annular rectangular to serve as a bidirectional tuned liquid column damper (BTLCD). This means we have made a barge-motion damper with negligible extra costs as an HWT needs a reservoir for fluid storage anyway. The barge HWT simulation model is transformed from the NREL (National Renewable Energy Laboratory) 5-MW geared equipped ITI Energy barge wind turbine model within the FAST (Fatigue, Aerodynamics, Structures, and Turbulence) code by replacing its drivetrain with a hydrostatic transmission drivetrain and incorporating the coupled dynamics of the barge-reservoir system. We use two simplified turbine-reservoir models to optimise the parameters of the BTLCD reservoir, which describe the pitch and roll motions of the turbine-reservoir system respectively. Simulation results based on the transformed NREL 5-MW barge HWT model show that the optimal BTLCD reservoir is very effective in mitigating pitch and roll motions of the barge under realistic wind and wave excitations, which reduces the tower load and improves the power quality.

In recent years, multilabel classification has attracted significant attention in multimedia annotation. However, most of the multilabel classification methods focus only on the inherent correlations existing among multiple labels and concepts and ignore the relevance between features and the target concepts. To obtain more robust multilabel classification results, we propose a new multilabel classification method aiming to capture the correlations among multiple concepts by leveraging hypergraph that is proved to be beneficial for relational learning. Moreover, we consider mining feature-concept relevance, which is often overlooked by many multilabel learning algorithms. To better show the feature-concept relevance, we impose a sparsity constraint on the proposed method. We compare the proposed method with several other multilabel classification methods and evaluate the classification performance by mean average precision on several data sets. The experimental results show that the proposed method outperforms the state-of-the-art methods.