To recognize the tender shoots for high-quality tea and to determine picking points accurately quickly, this paper proposes a method of recognizing with improved YOLO-v3 deep convolutional neural network algorithm. This realizes end-to-end target detection recognition different postures shoots, considering both efficiency accuracy. At first, in order predict category position an image pyramid structure is used obtain characteristic map at scales. The residual block added downsampling part,...

Summary This article considers the parameter estimation for a special bilinear system with colored noise. Its input‐output representation is derived by eliminating state variables in system. Based on of system, multiinnovation generalized extended stochastic gradient (MI‐GESG) algorithm proposed using identification theory. Furthermore, decomposition‐based (ie, hierarchical multiinnovation) (H‐MI‐GESG) to enhance accuracy principle, and GESG presented comparison. Compared existing algorithms...

Lumbar spine segmentation is important to help doctors diagnose lumbar disc herniation (LDH) and patients' rehabilitation treatment. In order accurately segment the spine, a image algorithm based on improved Attention U-Net proposed. The U-Net, attention module multilevel feature map fusion adopted, two residual modules are introduced instead of original convolution blocks. hybrid loss function used for prediction during training process, finally, superposition process realized. this...

A drawback of fringe projection profilometry (FPP) is that it still a challenge to perform efficient and accurate high-resolution absolute phase recovery with only single measurement. This paper proposes single-model self-recovering method based on deep learning. The built Fringe Prediction Self-Recovering network converts image acquired by camera into four mode images. algorithm adopted obtain wrapped phases grades, realizing from shot. Low-cost dataset preparation realized the constructed...

ABSTRACT To simplify the soft gripper structure while improving its gripping range, load capacity, and radial load‐carrying Single‐axis Deformation (SD) grippers were designed, inspired by blooming behavior of petunias. Four SD finger structures designed analyzing anatomical petunia petals. The stress concentration resulting from deformation fingers with different was analyzed using numerical methods under same conditions. verify ability gripper, we conducted experiments on opening closing...

Dynamic obstacle avoidance is essential for unmanned surface vehicles (USVs) to achieve autonomous sailing. This paper presents a dynamic navigation ship domain (DNSD)-based approach USVs in compliance with COLREGs. Based on the detected information, can not only infer collision risk, but also plan local path trajectory make appropriate maneuvers. Firstly, analytical DNSD model established taking into account parameters, maneuverability, sailing speed, and encounter situations regarding...

This paper presents a weight optimization method for nonlinear model predictive controller (NMPC) based on the genetic algorithm (GA) ship trajectory tracking. The coefficients Q and R of objective function in NMPC are obtained via real-time instead trial error method, which improves efficiency accuracy controller. In addition, targeted improvements made to internal crossover operator, mutation rate, rate algorithm. simulation comparison tracking between with real-time-optimized one constant...

In order to eliminate driving dangers caused by tire surface bubbles, the detection method of bubble defects on surfaces based line lasers and machine vision is studied. Since it difficult recognize directly through images, laser scanning used obtain images. The filtering morphology are combined preprocess these gray centroid adopted extract center stripe, then algorithm determine positions proposed. According geometric characteristics coordinates starting points, ending rough vertices...

Abstract Bitter gourd, being perishable, requires timely harvesting. Delayed harvesting can result in a substantial reduction fruit quality. while premature leads to underdeveloped and decreased yields, the continuous flowering pattern bitter gourd underscores significance of accurately assessing growth ensuring for subsequent setting development. The current reliance on experience production personnel represents inefficiency. We present an improved real-time instance segmentation model...

ABSTRACT This paper proposes a robust discrete‐time integral fast terminal sliding mode predictive control (DIFTSMPC) scheme for DP ship trajectory tracking under the consideration of unmodeled dynamics, environmental disturbance, and input constraints. An improved single closed‐loop strategy is designed by adopting earth‐fixed reference velocity as virtual velocity. The nonlinear PID to achieve faster convergence. unknown dynamics disturbance are combined lumped uncertainty term estimated...

Pump as turbine (PAT) is one of the micro-hydro system components. It can be used alone a generator in remote areas without power supply, and hydraulic networks instead pressure-reducing valves. However, its optimization still remains an open research problem. One techniques rounding sharp edges at blade periphery. Existing studies are mostly based on prototype experiments to obtain effect. In order more intuitively analyze influence this structural internal flow PAT, paper uses CFD method...

This paper is devoted to the issue of computationally efficient and robust nonlinear model predictive control (NMPC) for ship dynamic positioning (DP) systems subjected input constraints unknown environmental disturbances. The Laguerre functions, typically applied linear systems, are introduced constrained NMPC design DP system reduce computational burden. unscented Kalman filter adopted estimate disturbances states; thus, disturbance estimates utilized as cancellation signal achieve...

Purpose This paper aims to present the mixed elastohydrodynamic lubrication (EHL) model and obtain leakage characteristics for skeleton reciprocating oil seal. Design/methodology/approach The consists of a finite element analysis contact pressure, fluid mechanics film, asperity deformation seal lip an iterative numerical simulation process. Findings Simulation results show that is in direct proportion roughness speed, inverse viscosity. Comparisons with experimental verify validity EHL...

To realize the automatic symptom recognition and classification of MR images improve accuracy efficiency diagnosis lumbar intervertebral disc herniation (LDH), a method for disease is proposed in this paper. The mainly includes three steps: preprocessing, target segmentation, classification. Preprocessing performed by noise reduction interference removal methods blurred images. contour poles are used to determine four points tail vertebra order reduce wrong segmentation vertebra. A based on...

Due to the complicated realization process, traditional three-dimensional (3D) reconstruction method of structured light gradually fails meet needs actual production and complex scenes. The combination fringe projection profilometry deep learning effectively improves situation. Classical neural network models have shown their unique advantages in field 3D reconstruction. Although existing methods been improved different aspects, they still problems data set low accuracy, so are difficult be...

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

With the rapid development of Industrial Internet Things (IIoT), prediction Remaining Useful Life (RUL) has played a crucial role in ensuring equipment reliability, enhancing production efficiency, and reducing maintenance costs. Particularly today's data-driven industrial environment, adoption effective RUL techniques is paramount. In response to this demand, we propose implement deep learning-based model. Our model begins by utilizing one-dimensional Convolutional Neural Network (CNN)...