In this study, an event-based control strategy is proposed for second-order disturbed multi-agent systems (MASs) to achieve predefined-time practical consensus. comparison the finite/fixed-time control, consensus (PTC) can be achieved in a completely pre-specified time. For consensus, global information of system not required calculate estimate convergence time, including initial states and communication topology. addition, singular problem loop are also avoided. Moreover, developed obtain...

This paper presents a data-driven economic model predictive control (EMPC) framework for maximizing the output power of nonlinear wave energy converter (WEC). The linear representation WEC is realized by Koopman operator theory, and Extended Dynamic Mode Decomposition (EDMD) used to approximate infinite-dimensional finite-dimensional model. lifted plays role in designing controllers system using model-based controller strategies. EMPC method adopted realize optimization process WEC. designed...

Light propagates in water with certain attenuation, which causes quality problems, such as color cast, low contrast, and illumination, underwater images. Moreover, it is generally difficult to obtain a large number of real-world images the corresponding ground truth convoluted environments. To address these two this article recommends an image enhancement scheme based on few-shot learning multi-color space, called UIE-FSMC. Specifically, for first time, we propose lightweight network...

In this article, a reinforcement learning (RL)-based strategy for unmanned surface vehicle (USV) path following control is developed. The proposed method learns integrated guidance and heading policy, which directly maps the USV's navigation states to motor commands. By introducing twin-critic design an integral compensator conventional deep deterministic policy gradient (DDPG) algorithm, tracking accuracy robustness of controller can be significantly improved. Moreover, pretrained neural...

This article considers the exponential output synchronization of complex networks with coupling (CNOC) based on intermittent event-triggered control (IE-TC) strategy. Unlike existing strategies, proposed IE-TC strategy is instead time-triggered during intervals. Moreover, average rate adopted for control, which makes results less conservative. A new Lyapunov function to simplify proof, does not require traditional mathematical induction method control. Based and inequality technique,...

The data are rapidly expanding nowadays, which makes it very difficult to analyze valuable information from big data. Most of the existing mining algorithms deal with problems at large time and space costs. This paper focuses on sampling problem puts forward an efficient heuristic Cluster Sampling Arithmetic, called CSA. Many former researchers adopted random method extract early sample set original then made a variety different processing in order obtain corresponding minimum set, is...

Feature matching is a crucial technique to estimate robot pose in the light of natural landmarks for vision navigation system. It difficult present feature approaches balance accuracy, robustness and efficiency vision-based complex workspaces. A multi-stage KNN-TF-BF-EN-RANSAC (KTBER) refinement proposed combine with adaptive ORB extraction method, order develop novel KTBER_AORB model that can outperform state-of-the-art techniques. Firstly, (AORB) features are extracted Laplacian...

Localization in unknown environments is an essential requirement for vision navigation of robotic vehicles intelligent transportation systems. However, moving objects dynamic scenarios usually bring about great difficulty robot localization, because motion estimation disturbed by increasing feature outliers caused objects. In order to improve the accuracy and robustness a novel saliency-induced object detection (SMOD) approach proposed filter out RGB-D-based simultaneous localization mapping...

Pose estimation is a crucial technique to achieve robot localization using natural features for an RGB-D vision navigation system. However, it still suffers from the impoverishment of robustness and accuracy in complex workspaces. A robust real-time pose approach hybrid point twin line reprojection proposed allow accurate vision-based under various challenging scenarios including low-texture, low-structure, occlusion, viewpoint illumination changes. Firstly, improved adaptive threshold...

The aim of the study was to assess ability contrast-enhanced computed tomography (CECT) differentiate aggressive from nonaggressive solid pseudopapillary neoplasms (SPNs).Forty treatment-naive patients with pathologically proven pancreatic SPNs were included. Imaging characteristics determined by consensus 3 radiologists blinded histopathologic aggressiveness. All underwent 4-phase CECT using a protocol. regions interest tumor and normal pancreas documented on all phases. Lymph nodes...

The marine aquaculture fishing industry has caused many problems such as low work efficiency and hidden safety hazards due to automation. Research on robots that can replace humans for autonomous operations excellent significance prospects. Aiming at the of uneven lighting, poor visibility, magazines in underwater environment, robot quickly autonomously detect products other targets. This paper implements an image enhancement algorithm be deployed a small AI system object detection based...

Computer-aided detection (CAD) of lobulation can help radiologists to diagnose/detect lung diseases easily and accurately. Compared CAD nodule other lesions, remained an unexplored problem due very complex varying nature lobulation. Thus, many state-of-the-art methods could not detect successfully. Hence, we revisited classical with the capability extracting undulated characteristics designed a sliding window based framework for in this paper. Under framework, investigated three categories...

Pancreas segmentation is an essential predetermination for effective treatment of pancreac organs. To boost the capability pancreas segmentation, this paper proposes a method combining superpixel and region merging. Firstly, medical proposed to preprocess CT image ameliorate verge features in images. Then, uses New Maximal Similarity-Based Region Merging (NMSRM) way obtain results on preprocessed We evaluated it NIH dataset compared with latest methods, where Dice-Sørensen coefficient (DSC) 94%.

This paper investigates the autonomous landing guidance and control for cooperative unmanned aerial vehicle-unmanned surface vehicle (UAV-USV) systems. A deep learning-based visual positioning architecture is developed to estimate position of pad on USV from images captured by UAV’s onboard camera. By introducing spatial softmax operation into conventional convolutional networks, accuracy computational efficiency are improved. Then, a three-stage strategy with safety protocols proposed,...

In recent years, computer-based testing has become an effective approach to evaluate students' learning level.In our previous work, a personalized genetic algorithm (PGA) for test sheet assembling was proposed.In this paper, improved named PGAC which makes improvement in the crossover process of PGA is presented.Considering item distribution, incorporated distribution (IGAID) based on presented assemble simulation sheets have good knowledge hierarchy each student.Experiments and comparison...

The serious eco-environmental hazard affecting hilly city have appeared. we should input more investment into environmental protection, and develop green technology recycling economy. In order to guarantee ecological security, the countermeasures for infrastructure construction are put forward, establishing space system including suburban forest, farmland space, recovering wetland ecosystem, optimizing industrial structure urban spatial distribution, developing economy, monitoring network...

Feature matching technology is one important part for a vision navigation system to estimate robot pose according natural landmarks. However, it difficult the existing techniques solve dilemma between accuracy, and efficiency, as well robustness under complex workspace involving non-uniform illumination texture-less. To address this issue, paper proposes novel feature algorithm that robust, relatively fast in workspaces. The main idea of proposed method multi-stage module consisting K...

The open-up data mining system is one of developing directions the present system. In this paper we have put forward a based on Web Services and given an applied example basis analysis As stronger sealing their joints are simple, they make integration more flexible, heightening working

Robot localization with an RGB-D sensor has attracted substantial research attention lately. However, its robustness and accuracy are still poor in complex environments. To alleviate this issue, paper presents a robust visual odometry (VO) based on point line features to allow accurate scenarios including illumination change, low-texture occlusion. Firstly, enhance the of localization, we analyze performance for different feature extraction, add EDLine our VO system. Second, lines descripted...

With fast development and deep appliance of the Internet, problem mass image data storage stand out, so low management efficiency, ability high cost traditional framework has appeared. The appearance Hadoop provides a new thought. However, itself is not suit for handle small files. This paper puts forward files based on Hadoop, solved internal bottleneck NameNode when are excessive through classification algorithm preprocessing module lead-in efficiency first-level index mechanism. test...

Alzheimer's disease (AD), the most common form of dementia, is a progressive neurodegenerative disorder. Mild cognitive impairment (MCI) due to AD refers symptomatic predementia phase AD. Since not all MCI caused by AD, identifying phenotypic or endophenotypic characteristics persons with who develop dementia could provide important prognostic information. Here, we explore combined prediction performance Pittsburgh Compound-B positron emission tomography (11 C-PIB-PET) for amyloid imaging,...

The traditional search engine is unable to correct for the magnanimous information in Deep Web hides. database's classification key step which integrates with database and retrieves. This article has proposed one kind of based on machine learning's web database. experiment indicated that after this taxonomic approach undergoes few sample training, it can achieve very good classified effect, along training sample's increase, classifier's performance maintains stable rate accuracy recalling...