In order to detect outliers in hydrological time series data for improving quality and decision-making related design, operation, management of water resources, this research develops a outlier detection method hydrologic that can be used identify deviate from historical patterns. The first built forecasting model on the history then it predict future values. Anomalies are assumed take place if observed values fall outside given prediction confidence interval ( PCI ), which calculated by...

With the development of hyperspectral remote sensing technology, spectral resolution image data becomes denser, which results in large number bands, high correlation between neighboring and redundancy. It is necessary to reduce these bands before further analysis, such as land cover classification target detection. Aiming at task, this paper proposes an effective band selection method from novel perspective shape similarity analysis with key points extraction thus retains physical...

Ensuring the quality of hydrological data has become a key issue in field hydrology. Based on characteristics data, this paper proposes data-driven control method for data. For continuous time series two combined forecasting models and one statistical model are constructed from horizontal, vertical, perspectives three provide confidence intervals. Set suspicious level based number intervals violations, suggested values missing discrete with large time-space difference, similar weight...

The accurate and timely estimation of river discharge plays an important role in hydrological modeling, especially for avoiding the consequences flood events. majority existing work on hydrologic prediction focuses modeling inherent physical process specific basins, while geographic-connections between rivers are largely ignored. Geographically connected provide rich spatial information that can be used to predict amounts. In this paper, we study a novel problem exploiting both temporal...

Abstract Run‐time monitoring is an important technique to detect erroneous run‐time behaviors. Several techniques have been proposed automatically generate monitors from specification languages check temporal and real‐time properties. However, of probabilistic properties still requires manual generation. To overcome this problem, we define a formal property language called Probabilistic Timed Property Sequence Chart (PTPSC). PTPSC timed extension the existing scenario‐based formalism (PSC)....

Abstract Water level prediction of small- and medium-sized rivers plays an important role in water resource management flood control. Such a is concentrated the season because frequent occurrence disasters plain area. Moreover, mountainous areas suddenly rises falls, slope steep. Thus, establishing hydrological model for with high accuracy different topographic features, that is, plains mountains, urgent problem. A method based on ASCS_LSTM_ATT proposed to solve this First, parameters are...

Abstract Hydrological time series data is stochastic and complex, the importance of its historical features different. A single model difficult to overcome own limitations when dealing with hydrological prediction problems, accuracy a can be further improved. According characteristics data, CNN-BiLSTM water level method attention mechanism proposed. In this paper, CNN extracts spatial BiLSTM learns period by combining past future sequence information, introduced focus salient in sequence....

Multivariate time series (MTS) datasets are very common in various financial, multimedia, and hydrological fields. In this paper, a dimension-combination method is proposed to search similar sequences for MTS. Firstly, the similarity of single-dimension calculated; then overall MTS obtained by synthesizing each based on weighted BORDA voting method. The could use existing searching Several experiments, which used classification accuracy as measure, were performed six from UCI KDD Archive...

A novel extreme rainfall prediction model combined with data mining is proposed in this paper. Because of the special nature hydrological data, our uses clustering method to group next year's average and then establish a hybrid based on building neural networks for each group. Furthermore discriminant analysis employed classify sample's coming corresponding BP network chosen obtain prediction. We also take impact discrimination error into account computation precipitation predictive value...

Similarity analysis of small- and medium-sized watersheds mainly depends on manual work, there is no complete automated method. In order to solve this problem, we propose a similarity method based clustering ensemble model. First, the iterative construction algorithm with weighted random sampling (WRS-CCE) proposed get great collectives. Then, combine spectral fuzzy C-means design consensus function for watershed data sets. Finally, carried out according results. Experiments show that model...

Large amount of hydrological data set is a kind big data, which has much hidden and potentially useful knowledge. It necessary to extract these knowledge from set, can provide more valuable information be for future forecasting. Data mining based on time series widely used currently. There are some techniques anomaly. However, most cannot suit unstable such as set. Some important problems high fitting error after dimension reduction low accuracy results. In this work we propose new idea...

Based on the analyses of a genetic algorithm’s properties and shortages, novel algorithm (GTGA) is proposed with analogies to concept method gene therapy theory. The core GTGA lies construction pool operator. pool, which first created according prior knowledge then updated posterior knowledge, contains eminent genes, morbid genes their character istic information as well. operator consists an insertion operation that inserts into individual removing removes from individual. methods creation...

AbstractDue to the high spectral resolution, hyperspectral images (HSI) have been widely used in land cover classification and material identification. Band selection is one of necessary preprocessings reduce data volume redundancy therein for subsequent analysis. Aiming at speeding up search-based band process, this letter proposes a new technique from perspective curve shape similarity. Through newly defined measure subset discriminativeness (BSD), class-specific important bands (IBs) are...

Large amount of hydrological data set is a kind big data, which has much hidden and potentially useful knowledge. Hydrological prediction important for the state flood control drought relief. How to forecast accurately timely with becomes challenge. There are some forecasting techniques used widely. However, they limited by their adaptability, volume feature. The most problems high time consumption, low accuracy bad adaptability prediction. In this paper, new approach based on an integration...

Anomalous patterns are common phenomena in time series datasets. The presence of anomalous hydrological data may represent some hydrometeorological events that significantly different from others and induce a bias the decision-making process related to design, operation management water resources. Hence, it is necessary extract those “anomalous” knowledge can provide valuable useful information for future analysis forecasting data. This paper focuses on problem detecting data, proposes an...

Standing the perspective of data mining and using basic principles artificial neural network to establish a average extreme rainfall prediction model which is based on BP netwok.This only use precipitation indexes as factors predict in coming year.The combined with stepwise regression select input vectors used bayesian regularization method further improve generalization ability, thereby increased forecast accuracy trend rainfall.It proved that indeed valid reliable by experimenting many...

Solar Power Tower (SPT) plant is a hugeous and complicated system, thus there have not been relative research productions on the record in aspect of developing knowledge base its Fault Diagnosis System (FDS) whole world yet. In this paper, modular hierarchical FDS designed developed to use SPT plants according characteristics structure operation plants. This consists Main Control Module, Concentrator Subsystem, Receiver Heat Storage Generating Subsystem Assistant Subsystem. Each subsystem...

There has been large amount of hydrological data collected by various sensors during the last years and how to discover hidden knowledge among these caused more attention from diverse fields, such as hydrologist researchers mining. This paper deals with similarity mining time series concentrates itself on analysis multivariate (MTS). A novel measure put forward, which is based well-known BORDA count in multiple classifier system. Firstly, dimension reduction adaptively conducted according...

Current algorithms that utilize water index to extract information from high resolution remote sensing image are inadequate in it is difficult determine the optimal thresholds, result of boundary not satisfactory and prone error. We propose a new algorithm which combines segmentation based on MRF model with Normalized Difference Water Index (NDWI) for extracting information. represent pixels as random variables introduce hybrid feature space energy function these variables, then minimize by...

Precipitation images play an important role in meteorological forecasting and flood forecasting, but how to characterize precipitation conduct rainfall similarity analysis is challenging meaningful work. This paper proposes a research method based on deep learning by using images. The algorithm first extracts regional precipitation, distribution, center of the defines measures, respectively. Additionally, ensemble weighting Normalized Discounted Cumulative Gain-Improved Particle Swarm...

Most of the time series data mining tasks attempt to discover patterns that appear frequently. Abnormal is often ignored as noise. There are some techniques based on extract anomaly. However, most these cannot suit big unstable existing in various fields. Their key problems high fitting error after dimension reduction and low accuracy results. This paper studies an approach abnormal hydrological field. The authors propose a new idea solve problem anomaly series. They Feature Points Symbolic...