- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
[IEEE 2019 International Conference on Smart Energy Systems and Technologies (SEST) - Porto, Portugal (2019.9.9-2019.9.11)] 2019 International Conference on Smart Energy Systems and Technologies (SEST) - Machine Learning Algorithms in Forecasting of Photovoltaic Power Generation
摘要： Due to the intrinsic intermittency and stochastic nature of solar power, accurate forecasting of the photovoltaic (PV) generation is crucial for the operation and planning of PV-intensive power systems. Several PV forecasting methods based on machine learning algorithms have recently emerged, but a complete assessment of their performance on a common framework is still missing from the literature. In this paper, a comprehensive comparative analysis is performed, evaluating ten recent neural networks and intelligent algorithms of the literature in short-term PV forecasting. All methods are properly fine-tuned and assessed on a one-year dataset of a 406 MWp PV plant in the UK. Furthermore, a new hybrid prediction strategy is proposed and evaluated, derived as an aggregation of the most well-performing forecasting models. Simulation results in MATLAB show that the season of the year affects the accuracy of all methods, the proposed hybrid one performing most favorably overall.
关键词： intelligent algorithms,photovoltaic,machine learning,Forecasting,neural networks
更新于2025-09-11 14:15:04
-
[IEEE 2019 IEEE International Conference on Electrical, Computer and Communication Technologies (ICECCT) - Coimbatore, India (2019.2.20-2019.2.22)] 2019 IEEE International Conference on Electrical, Computer and Communication Technologies (ICECCT) - Improved Fault Detection and Location Scheme for Photovoltaic Systems
摘要： Photovoltaic is encountering a quick innovation development since a decade ago. Yet, strange conditions, for example, shortcomings, low irradiance and so forth it influence the yield of PV framework. To enhance the execution of and productivity of PV framework, it is important to create enhanced blame location procedures. This paper for the most part centres around recognition conspire for LL and LG blames in the PV cluster. Such blames stay undetected under irradiance conditions, especially, when a most extreme power point following calculation is in administration. In the event that these shortcomings are undetected, there is extensively loss of yield of PV framework, in the event that these issues are not recognized for longer time, it might harm the board and conceivably cause fire dangers. The exhibited blame identification conspire utilizes Multi-Resolution Signal Decomposition (MSD) procedure and two machine learning calculations to be specific Fuzzy Logic and K-Nearest Neighbor (KNN) to group the blame and decide its area. Reenactment results confirm the exactness, unwavering quality and versatility of the exhibited plan.
关键词： K-Nearest Neighbour (KNN) algorithm,Fuzzy logic,MSD,Machine Learning algorithm,Fault detection
更新于2025-09-11 14:15:04
-
Weighting Factor Selection of the Ensemble Model for Improving Forecast Accuracy of Photovoltaic Generating Resources
摘要： Among renewable energy sources, solar power is rapidly growing as a major power source for future power systems. However, solar power has uncertainty due to the effects of weather factors, and if the penetration rate of solar power in the future increases, it could reduce the reliability of the power system. A study of accurate solar power forecasting should be done to improve the stability of the power system operation. Using the empirical data from solar power plants in South Korea, the short-term forecasting of solar power outputs were carried out for 2016. We performed solar power forecasting with the support vector regression (SVR) model, the na?ve Bayes classifier (NBC), and the hourly regression model. We proposed the ensemble method including the selection of weighting factors for each model to improve forecasting accuracy. The forecasted solar power generation error was indicated using normalized mean absolute error (NMAE) compared to the plant capacity. For the ensemble method, the results of each forecasting model were weighted with the reciprocal of the standard deviation of the forecast error, thus improving the solar power outputs forecast accuracy.
关键词： support vector regression,na?ve Bayes classifier,solar power forecasting,machine learning,ensemble,day ahead power forecasting
更新于2025-09-11 14:15:04
-
A plasmonic ellipse resonator possessing hybrid modes for ultracompact chipscale application
摘要： Data mining methods based on machine learning play an increasingly important role in drug design and discovery. In the current work, eight machine learning methods including decision trees, k- Nearest neighbor, support vector machines, random forests, extremely randomized trees, AdaBoost, gradient boosting trees, and XGBoost were evaluated comprehensively through a case study of ACC inhibitor data sets. Internal and external data sets were employed for cross- validation of the eight machine learning methods. Results showed that the extremely randomized trees model performed best and was adopted as the first step of virtual screening. Together with structure- based virtual screening in the second step, this combined strategy obtained desirable results. This work indicates that the combination of machine learning methods with traditional structure- based virtual screening can effectively strengthen the ability in finding potential hits from large compound database for a given target.
关键词： molecular docking,machine learning,extremely randomized trees,ACC inhibitors
更新于2025-09-11 14:15:04
-
Exploring materials band structure space with unsupervised machine learning
摘要： An unsupervised machine learning algorithm is applied for the first time to explore the space of materials electronic band structures. T-student stochastic neighbor embedding (t-SNE), a state of the art algorithm for visualization of high dimensional data, is applied on feature spaces constructed by extracting electronic fingerprints straight from Brillouin zone of the materials. Different spaces are designed and mapped to lower dimensions allowing to analyze and explore this previously uncharted band structure space for thousands of materials at once. In all cases analyzed machine learning was able to learn and cluster the materials depending on the features involved. t-SNE promises to be a extremely useful tool for exploring the materials space.
关键词： Fermiology,Data visualization,Band structure,Unsupervised machine learning,Data mining,Materials informatics,High throughput materials calculations
更新于2025-09-11 14:15:04
-
Image-Based Visibility Estimation Algorithm for Intelligent Transportation Systems
摘要： Posted road speed limits contribute to the safety of driving, yet when certain driving conditions occur, such as fog or severe darkness, they become less meaningful to the drivers. To overcome this limitation, there is a need for adaptive speed limits system to improve road safety under varying driving conditions. In that vein, a visibility range estimation algorithm for real-time adaptive speed limits control in intelligent transportation systems is proposed in this paper. The information required to specify the speed limit is captured via a road side unit that collects environmental data and captures road images, which are then analyzed locally or on the cloud. The proposed analysis is performed using two image processing algorithms, namely, the improved dark channel prior (DCP) and weighted image entropy (WIE), and the support vector machine (SVM) classi?er is used to produce a visibility indicator in real-time. Results obtained from the analysis of various parts of highways in Canada, provided by the Ministry of Transportation of Ontario (MTO), show that the proposed technique can generate credible visibility indicators to motorists. The analytical results corroborated by extensive ?eld measurements con?rmed the advantage of the proposed system when compared to other visibility estimation methods such as the conventional DCP and WIE, where the proposed system results exhibit about 25% accuracy enhancement over the other considered techniques. Moreover, the proposed DCP is about 26% faster than the conventional DCP. The obtained promising results potentiate the integration of the proposed technique in real-life scenarios.
关键词： image processing,dark channel prior,intelligent transportation system,SVM,Visibility,smart cities,entropy,machine learning
更新于2025-09-11 14:15:04
-
[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 - Machine Learning Framework for Mapping of Mississippi River Levees and Damage Assessment Using Terrasar-X Data
摘要： Earthen levees protect large areas of populated and cultivated land in the United States. Unstable slope conditions can lead to slump slides, which weaken the levees and increase the likelihood of failure during floods. Such slides could lead to further erosion and through seepage during high water events. Currently, levee inspections are performed infrequently and some of the problem areas are not visible. There is a need to develop cost-effective large-scale methods of screening levees in a timely manner. Sensing the condition of levees remotely can help levee managers to focus and prioritize their inspection and maintenance activities. This paper presents results of applying the TerraSAR-X synthetic aperture radar data to detect vulnerabilities on Francis, Mississippi river levees. In this study, texture features were computed using the discrete wavelet transform (DWT), and both supervised and unsupervised classifiers were tested. The supervised method tested is the SVM (support vector machine), and the unsupervised one is the RXD (Reed-Xiaoli Detector). Both algorithms achieved high accuracy detection of potential slump slides in the test area.
关键词： Levee classification,remote sensing,synthetic aperture radar (SAR),machine learning
更新于2025-09-11 14:15:04
-
[IEEE 2018 20th International Conference on Transparent Optical Networks (ICTON) - Bucharest (2018.7.1-2018.7.5)] 2018 20th International Conference on Transparent Optical Networks (ICTON) - Examples of Machine Learning Algorithms for Optical Network Control and Management
摘要： Machine learning (ML) offers a great variety of algorithms that can be used in the context of optical networks. In particular, ML algorithms might be applied for classification and to detect patterns, among others. Both, can help to facilitate improving its performance, as well as to understand the behavior of optical networks. In this paper, we review two of these ML algorithms, one for classification and the other for clustering. Illustrative examples of the application of such supervised and unsupervised ML algorithms applied to optical networks are presented.
关键词： support vector machine,machine learning,data visualization,clustering
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 - Automatic Recognition of Oil Industry Facilities Based on Deep Learning
摘要： Effectively monitoring the real-time position and status of oil facilities (mainly well-site) in oil field is very important for the safety production. Considering the low efficiency of traditional visual interpretation method and the high demands of preset feature for machine learning method, one of the object detection methods in Deep learning (YOLOv2) was introduced to recognize oil industry facilities automatically. After establishing the dataset of oil facility samples, 90 percent of samples are used for model training while 10 percent are for validating. Comparing with the results extracted by machine learning (Adaboost model based on Haar-like), YOLOv2 recognition results of oil facilities indicated that: Deep learning improve the recognition efficiency and accuracy of oil facilities. The accuracy can be as high as 92% while the error rate and omission rate can be maintained in a low level. At the same time, the constructed model was applied in an oilfield in eastern part of China, and the result shows that the model can identify most of the oilfield facilities correctly with only 4% omission rate, which is much lower comparing with manual interpretation. However, the 11% error rate, caused by insufficient sample types and sample quantities, is relatively high especially in city area.
关键词： machine learning,petroleum industry facility,deep learning
更新于2025-09-10 09:29:36
-
Machine learning analysis of extreme events in optical fibre modulation instability
摘要： A central research area in nonlinear science is the study of instabilities that drive extreme events. Unfortunately, techniques for measuring such phenomena often provide only partial characterisation. For example, real-time studies of instabilities in nonlinear optics frequently use only spectral data, limiting knowledge of associated temporal properties. Here, we show how machine learning can overcome this restriction to study time-domain properties of optical fibre modulation instability based only on spectral intensity measurements. Specifically, a supervised neural network is trained to correlate the spectral and temporal properties of modulation instability using simulations, and then applied to analyse high dynamic range experimental spectra to yield the probability distribution for the highest temporal peaks in the instability field. We also use unsupervised learning to classify noisy modulation instability spectra into subsets associated with distinct temporal dynamic structures. These results open novel perspectives in all systems exhibiting instability where direct time-domain observations are difficult.
关键词： machine learning,extreme events,optical fibre modulation instability,unsupervised learning,supervised neural network
更新于2025-09-10 09:29:36