- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
Performance ratio prediction of photovoltaic pumping system based on grey clustering and second curvelet neural network
摘要： Performance ratio is an important parameter of measuring the quality and efficiency of photovoltaic (PV) pumping system, which should be predicted correctly to provide guidance for regulating the measurements of reducing losses of every part, therefore this research proposed a prediction model of performance ratio of PV pumping system based on grey clustering and second curvelet neural network. The meaning of performance ratio is analyzed and the main affecting factors of the performance level for PV pumping system are also summarized. The second curvelet neural network is constructed combing the second curvelet transform and feed forward neural network, and the structure of second curvelet neural network is designed. The classification of training and testing samples are confirmed based on the improved grey clustering model, and the corresponding mathematical models are studied. The firefly algorithm is used to optimize the second curvelet neural network. The grey classifications of training samples are confirmed based on grey clustering, which are used to train the second curvelet neural network with different structure optimized by firefly with different parameters, and then the testing samples are used to carry out prediction analysis. Simulation results show that the second curvelet neural network has highest prediction precision and efficiency, which can correctly and efficiently predict the performance ratio of PV pumping system.
关键词： Performance ratio,PV pumping system,Second curvelet neural network,Grey clustering
更新于2025-09-11 14:15:04
-
[IEEE 2019 IEEE PES Innovative Smart Grid Technologies Conference - Latin America (ISGT Latin America) - Gramado, Brazil (2019.9.15-2019.9.18)] 2019 IEEE PES Innovative Smart Grid Technologies Conference - Latin America (ISGT Latin America) - An Adaptive Perturb and Observe Method with Clustering for Photovoltaic Module with Smart Bypass Diode under Partial Shading
摘要： The photovoltaic maximum power point tracking using perturb and observe method has a fixed step size, where a small step size has a slower time response and more accurate steady-state, while a large step size is the opposite. This work proposes an adaptive step size, proportional to the difference between actual and previous power sample, providing a fast time response and reducing the oscillations at steady-state. The oscillations are smaller with adaptive step size, but they are not annulled and the method presents loss by power oscillations. The clustering is used to eliminate this loss, setting the result of the simple average of the last five voltage samples. The enhanced method has been tested on a 72-cell photovoltaic module with a smart bypass diode per cell under partial shading. Modeling and simulation have been implemented using MATLAB/Simulink. The proposal obtained a faster time response and elimination of oscillations at steady-state.
关键词： Smart Bypass Diode,Adaptive Perturb and Observe with Clustering,Maximum Power Point Tracking,Photovoltaic Solar Module,Partial Shading
更新于2025-09-11 14:15:04
-
Short Term Prediction of Photovoltaic Power Based on FCM and CG-DBN Combination
摘要： Affected by many factors, the photovoltaic output power is characterized by nonlinearity, volatility and instability. Therefore, short-term forecasting models are required to have multiple inputs, levels, and categories. In order to solve the above problems and improve the accuracy of predictions, this paper proposes a combined model prediction method based on similar-day clustering and the use of Conjugate Gradient (CG) to improve Deep Belief Network (DBN). The initial method uses fuzzy C-Means Clustering Algorithm (FCM) to perform similar-day clustering on the original data according to the degree of membership. The CG-DBN prediction model is then designed according to the category, with the model ultimately being used to perform the short-term prediction of the PV output power. The proposed scheme uses data from Zhejiang Longyou power station for experimental analysis and verification, and the results were compared with the back propagation neural networks model, Support Vector Machine (SVM) model, and traditional deep belief network. The model’s predicted results are compared. Finally, it is concluded that, in the short-term PV power load forecasting, the prediction performance of the FCM and CG-DBN combination forecast model is better than the above three models and has strong feasibility in short-term PV power forecasting.
关键词： Depth belief network,Photovoltaic short-term forecast,Similar day clustering,Combined forecasting model
更新于2025-09-11 14:15:04
-
Identification of tea varieties by mid‐infrared diffuse reflectance spectroscopy coupled with a possibilistic fuzzy c‐means clustering with a fuzzy covariance matrix
摘要： Mid-infrared diffuse reflectance spectroscopy was used to rapidly and nondestructively identify tea varieties together with the proposed possibilistic fuzzy c-means (PFCM) clustering with a fuzzy covariance matrix. The mid-infrared diffuse reflectance spectra of 96 tea samples with three different varieties (Emeishan Maofeng, Level 1, and Level 6 Leshan trimeresurus) were acquired using the FTIR-7600 infrared spectrometer. First, multiplicative scatter correction was implemented to pretreat the spectral data. Second, principal component analysis was employed to compress the mid-infrared diffuse reflectance spectral data after preprocessing. Third, linear discriminant analysis was utilized for extracting the identification information required by the fuzzy clustering algorithms. Ultimately, the fuzzy c-means (FCM) clustering, the allied fuzzy c-means (AFCM) clustering, the PFCM clustering, and the PFCM clustering with a fuzzy covariance matrix were used to cluster the processed spectral data, respectively. The highest identification accuracy of the PFCM clustering with a fuzzy covariance matrix reached at 100% compared with those of FCM (96.7%), AFCM (94.9%), PFCM (96.3%), and partial least squares discrimination analysis (PLS-DA) algorithm (33.3%). It is sufficiently demonstrated that the mid-infrared diffuse reflectance spectroscopy coupled with the PFCM clustering with a fuzzy covariance matrix was a valid method for identifying tea varieties.
关键词： possibilistic fuzzy c-means clustering,tea varieties,Mid-infrared diffuse reflectance spectroscopy,fuzzy covariance matrix,nondestructive detection
更新于2025-09-11 14:15:04
-
[IEEE 2018 25th IEEE International Conference on Image Processing (ICIP) - Athens, Greece (2018.10.7-2018.10.10)] 2018 25th IEEE International Conference on Image Processing (ICIP) - A Perception-Based Framework for Wide Color Gamut Content Selection
摘要： Considering the content dependence of the perceived quality, selection of source content can significantly influence the results of studies related to Quality of Experience. In this paper, we propose an automated content selection method towards wide color gamut stimuli. The framework enables to objectively characterize the content according to its perceptual properties and thus allows to select a representative, diverse, and challenging subsets for various studies. Experimental results validate the reliability and robustness of the proposed framework.
关键词： quality of experience,color gamut mapping,clustering,Wide color gamut,content selection
更新于2025-09-11 14:15:04
-
Vortex Dynamics and Optical Vortices || Dynamical Particle Motions in Vortex Flows
摘要： Circular vortex flows generate interesting self-organizing phenomena of particle motions, that is, particle clustering and classification phenomena. These phenomena result from interaction between vortex dynamics and relaxation of particle velocity due to drag. This chapter introduces particle clustering in stirred vessels and particle classification in Taylor vortex flow based on our previous research works. The first part of this chapter demonstrates and explains a third category of solid-liquid separation physics whereby particles spontaneously localize or cluster into small regions of fluids by taking the clustering phenomena in stirred vessels as an example. The second part of this chapter discusses particle classification phenomena due to shear-induced migration. Finally, this chapter discusses about process intensification utilizing these self-organizing phenomena of particle motions in vortex flows.
关键词： particle clustering,solid-liquid flow,particle classification,chaotic-mixing field,process intensification
更新于2025-09-10 09:29:36
-
[IEEE 2018 International Conference on Machine Learning and Cybernetics (ICMLC) - Chengdu, China (2018.7.15-2018.7.18)] 2018 International Conference on Machine Learning and Cybernetics (ICMLC) - Image Segmentation Algorithm Based On Clustering
摘要： Image segmentation plays an important role in image processing. Image segmentation algorithms have been proposed as early as the last century, and constantly find and optimize various algorithms. The quality of the image segmentation algorithm determines the result of image analysis and image understanding. The principle, advantages and disadvantages of image segmentation algorithms are briefly introduced in this paper. The variety of image segmentation algorithms is determined by the complexity of the image itself. In recent years, scholars continue to improve a variety of image segmentation algorithms, the paper introduces the improvement of fuzzy C-means algorithm and mean-shift algorithm. The fuzzy C-means algorithm does not consider the spatial information of the image. Put forward an fuzzy C-means algorithm based on membership correction is proposed, taking into account the high correlation of pixels in image segmentation. The mean shift algorithm converges slowly, and mean shift algorithm based on conjugate gradient method is proposed to improve the convergence speed of the algorithm.
关键词： Fuzzy C-means algorithm,Clustering,Image segmentation,Mean shift algorithm
更新于2025-09-10 09:29:36
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia, Spain (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - The Influence of Sampling Methods on Pixel-Wise Hyperspectral Image Classification with 3D Convolutional Neural Networks
摘要： Supervised image classification is one of the essential techniques for generating semantic maps from remotely sensed images. The lack of labeled ground truth datasets, due to the inherent time effort and cost involved in collecting training samples, has led to the practice of training and validating new classifiers within a single image. In line with that, the dominant approach for the division of the available ground truth into disjoint training and test sets is random sampling. This paper discusses the problems that arise when this strategy is adopted in conjunction with spectral-spatial and pixel-wise classifiers such as 3D Convolutional Neural Networks (3D CNN). It is shown that a random sampling scheme leads to a violation of the independence assumption and to the illusion that global knowledge is extracted from the training set. To tackle this issue, two improved sampling strategies based on the Density-Based Clustering Algorithm (DBSCAN) are proposed. They minimize the violation of the train and test samples independence assumption and thus ensure an honest estimation of the generalization capabilities of the classifier.
关键词： DBSCAN,clustering,sampling strategies,Convolutional Neural Networks (CNNs),deep learning,Hyperspectral image classification
更新于2025-09-10 09:29:36
-
[IEEE 2018 International Russian Automation Conference (RusAutoCon) - Sochi (2018.9.9-2018.9.16)] 2018 International Russian Automation Conference (RusAutoCon) - Object Hierarchy in a Digital Image
摘要： The paper describes a model of binary hierarchical clustering of image pixels for object detection. In the model, a hierarchical sequence (a hierarchy) of pixel clusters is obtained adaptively to an image by iterative merging of pixel sets. Clustering of pixels depending on the number of clusters is given by a hierarchy of piecewise-constant approximations of the image and is described by a convex sequence of corresponding values of the total quadratic error, which is minimized for a given number of clusters. Due to the convexity property, the pixel clusters and their colors in the image are ordered by the absolute value of the increment of the total squared error accompanied by the dividing of cluster in two parts. For the hierarchy of pixel clusters, the problem of unambiguous assignment of image points to detected objects is formalized. In this case, the output of object detection is a sequence of object associations that incrementally reveal or disappear on a certain background. Objects are detected in accordance with the threshold value of the number of pixels in the cluster, or the threshold for the increment of the total squared error, or by other pixel cluster attributes that have a sense of a quantitative measure. The hierarchy of pixel clusters and the hierarchy of object associations are encoded with "pixel rating" stereo pair and "object rating" stereo pair. The pilot experimental results are demonstrated.
关键词： digital image,minimization,Ward’s pixel clustering,piecewise constant approximations,standard deviation,K-means method
更新于2025-09-10 09:29:36
-
Inversion of stellar spectral radiative properties based on multiple star catalogues
摘要： The spectral ?ux density of stars can indicate their atmospheric physical properties. A detector can obtain any band ?ux density at the design stage. However, the band ?ux density is con?rmed and ?xed in the process of operation because of the restriction of ?lters. Other band ?ux densities cannot be obtained through the same detector. In this study, a computational model of stellar spectral ?ux density is established based on basic physical parameters which are e?ective temperature and angular parameter. The stochastic particle swarm optimization algorithm is adopted to address this issue with appropriately chosen values of the algorithm parameters. Four star catalogues are studied and consist of the Large Sky Area Multi-Object Fibre Spectroscopic Telescope (LAMOST), Wide-?eld Infrared Survey Explorer (WISE), Midcourse Space Experiment (MSX), and Two Micron All Sky Survey (2MASS). The given ?ux densities from catalogues are input parameters. Stellar e?ective temperatures and angular parameters are inverted using the given ?ux densities according to SPSO algorithm. Then the ?ux density is calculated according to Planck’s law on the basis of stellar e?ective temperatures and angular parameters. The calculated ?ux density is compared with the given value from catalogues. It is found that the inversion results are in good agreement for all bands of the MSX and 2MASS catalogues, whereas they do not agree well in some bands of the LAMOST and WISE catalogues. Based on the results, data from the MSX and 2MASS catalogues can be used to calculate the spectral ?ux density at di?erent wavelengths of given wavelength ranges. The stellar ?ux density is obtained and can provide data support and an e?ective reference for detection and recognition of stars.
关键词： galaxy clustering,semi-analytic modeling,stars,absorption and radiation processes
更新于2025-09-10 09:29:36