Short-Term Household Load Forecasting Based on Attention Mechanism and CNN-ICPSO-LSTM

Accurate load forecasting forms a crucial foundation for implementing household demand response plans andoptimizing load scheduling. When dealing with short-term load data characterized by substantial fluctuations,a single prediction model is hard to capture temporal features effectively, resulting in diminished predictionaccuracy. In this study, a hybrid deep learning framework that integrates attention mechanism, convolution neuralnetwork (CNN), improved chaotic particle swarm optimization (ICPSO), and long short-term memory (LSTM), isproposed for short-term household load forecasting. Firstly, the CNN model is employed to extract features fromthe original data, enhancing the quality of data features. Subsequently, the moving average method is used for datapreprocessing, followed by the application of the LSTM network to predict the processed data. Moreover, the ICPSOalgorithm is introduced to optimize the parameters of LSTM, aimed at boosting the model’s running speed andaccuracy. Finally, the attention mechanism is employed to optimize the output value of LSTM, effectively addressinginformation loss in LSTM induced by lengthy sequences and further elevating prediction accuracy. According tothe numerical analysis, the accuracy and effectiveness of the proposed hybrid model have been verified. It canexplore data features adeptly, achieving superior prediction accuracy compared to other forecasting methods forthe household load exhibiting significant fluctuations across different seasons.

Seasonal Short-Term Load Forecasting for Power Systems Based on Modal Decomposition and Feature-Fusion Multi-Algorithm Hybrid Neural Network Model

To enhance the refinement of load decomposition in power systems and fully leverage seasonal change information to further improve prediction performance,this paper proposes a seasonal short-termload combination prediction model based on modal decomposition and a feature-fusion multi-algorithm hybrid neural network model.Specifically,the characteristics of load components are analyzed for different seasons,and the corresponding models are established.First,the improved complete ensemble empirical modal decomposition with adaptive noise(ICEEMDAN)method is employed to decompose the system load for all four seasons,and the new sequence is obtained through reconstruction based on the refined composite multiscale fuzzy entropy of each decomposition component.Second,the correlation between different decomposition components and different features is measured through the max-relevance and min-redundancy method to filter out the subset of features with strong correlation and low redundancy.Finally,different components of the load in different seasons are predicted separately using a bidirectional long-short-term memory network model based on a Bayesian optimization algorithm,with a prediction resolution of 15 min,and the predicted values are accumulated to obtain the final results.According to the experimental findings,the proposed method can successfully balance prediction accuracy and prediction time while offering a higher level of prediction accuracy than the current prediction methods.The results demonstrate that the proposedmethod can effectively address the load power variation induced by seasonal differences in different regions.

Short-term and imminent geomagnetic anomalies of the Wenchuan M_S8.0 earthquake and exploration on earthquake forecast

The diurnal variation of the geomagnetic vertical component is exhibited mainly by changes of phase and amplitude before strong earthquakes. Based on data recorded by the network of geomagnetic observatories in China for many years, the anomalous features of the appearance time of the minima of diurnal variations （i.e, low-point time） of the geo- magnetic vertical components and the variation of their spatial distribution （i.e, phenomena of low-point displacement） have been studied before the Wenchuan Ms8.0 earthquake. The strong aftershocks after two months＇ quiescence of M6 aftershocks of the Ms8.0 event were forecasted based on these studies. There are good correlativities between these geomagnetic anoma- lies and occurrences of earthquakes. It has been found that most earthquakes occur near the boundary line of sudden changes of the low-point time and generally within four days before or after the 27th or 41st day counting from the day of the appearance of the anomaly. In addition, the imminent anomalies in diurnal-variation amplitudes near the epicentral areas have also been studied before the Wenchuan earthquake.

Research status of earthquake forecasting in hydraulicfracturing induced earthquakes

In the new types of industrial activities including unconventional energy extraction associated with shale gas and hot dry rock,gas reservoir operations,CO2 geological storage,undergoing research on induced earthquake forecasting has become one of the forward positions of current seismology.As for the intense actual demand,the immature research on induced earthquake forecasting has already been applied in pre-assessment of site safety and seismic hazard and risk management.This work will review systematically recent advances in earthquake forecasting induced by hydraulic fracturing during industrial production from four aspects:earthquake occurrence probability,maximum expected magnitude forecasting,seismic risk analysis for engineering and social applications and key scientific problems.In terms of earthquake occurrence probability,we introduce statistical forecasting models such as an improved ETAS and non-stationary ETAS and physical forecasting models such as Seismogenic Index(SI)and hydro-mechanism nucleation.Research on maximum expected magnitude forecasting has experienced four stages of linear relationship with net injection volume of fluid,power exponential relationship and physical forecasting regarding fault parameters.For seismic risk analysis,we focus on probabilistic seismic hazard assessment and quantitative geological susceptibility model.Furthermore,this review is extended to key scientific problems that contain obtaining accurate fault scale and environmental stress state of reservoir,critical physical process of runaway rupture,complex mechanism of fault activation as well as physical mechanism and modeling of trailing effect.This work in understanding induced earthquake forecasting may contribute to unconventional energy development and production,seismic hazard mitigation,emergency management and scientific research as a reference.

Evaluation of numerical earthquake forecasting models

Evaluation of numerical earthquake forecasting models needs to consider two issues of equal importance:the application scenario of the simulation,and the complexity of the model.Criterion of the evaluation-based model selection faces some interesting problems in need of discussion.

On stress-forecasting strategy of earthquakes from stress buildup,stress shadow and stress transfer(SSS) based on numerical approach

Global Positioning System （GPS） and Interferometric Synthetic Aperture Radar （InSAR）, used for monitoring crust deformation, are found to be very promising in earthquake prediction subject to stress-forecasting. However, it is recognized that unless we can give reasonable explanations of these curious precursory phenomena that continue to be serendipitously observed from time to time, such high technology of GPS or InSAR is difficult to be efficiently used. Therefore, a proper model revealing the relation between earthquake evolution and stress variation, such as the phenomena of stress buildup, stress shadow and stress transfer （SSS）, is crucial to the GPS or InSAR based earthquake prediction. Here we address this question through a numerical approach of earthquake development using an intuitive physical model with a map-like configuration of discontinuous fault system. The simulation provides a physical basis for the principle of stress-forecasting of earthquakes based on SSS and for the application of GPS or InSAR in earthquake prediction. The observed SSS associated phenomena with images of stress distribution during the failure process can be continuously simulated. It is shown that the SSS are better indicators of earthquake precursors than that of seismic foreshocks, suggesting a predictability of earthquakes based on stress-forecasting strategy.

Loss forecasting of earthquake fire based on radial basis function network

According to complexity and multiplicity of the post-earthquake fire, the loss forecasting model of earthquake fire is established by using radial basis function neural network with adaptability, self-learning and fault-tolerant based on the historical information. The applicability and validity of the model is manifested through testing and discussion. A simple and available method is provided for the prediction of losses of other natural disaster.

A Review on the Research Progress in Operational Earthquake Forecasting(OEF)in the World

In this paper,the research progress of the Operational Earthquake Forecasting( OEF) is introduced from the major areas of concern,the concept of probability gain,hybrid model development,and the application to earthquake disaster reduction. Due to the development of OEF based on the global "Collaboratory for the Study of Earthquake Predictability( CSEP) " plan,it provides a significant technical foundation for earthquake forecast modeling and a practical foundation for solving the actual problems in earthquake preparedness and disaster mitigation. Therefore, related research and technical ideas provide inspirational and referential significance for earthquake forecasting/prediction.

Discussion on Earthquake Forecasting and Early Warning

Through analysis of natural and social attributes of earthquake forecasting,the relationship between the natural and social attributes of earthquake forecasting(early warning)has been discussed.Regarding the natural attributes of earthquake forecasting,it only attempts to forecast the magnitude,location and occurrence time of future earthquake based on the analysis of observational data and relevant theories and taking into consideration the present understanding of seismogeny and earthquake generation.It need not consider the consequences an earthquake forecast involves,and its purpose is to check out the level of scientific understanding of earthquakes.In respect of the social aspect of earthquake forecasting,people also focus on the consequence that the forecasting involves,in addition to its natural aspect,such as the uncertainty of earthquake prediction itself,the impact of earthquake prediction,and the earthquake resistant capability of structures(buildings),lifeline works,etc.In a word,it highlights the risk of earthquake forecasting and tries to mitigate the earthquake hazard as much as possible.In this paper,the authors also discuss the scientific and social challenges faced in earthquake prediction and analyze preliminarily the meanings and content of earthquake early warning.

SeisGuard: A Software Platform to Establish Automatically an Earthquake Forecasting Model

SeisGuard, a system for analyzing earthquake precursory data, is a software platform to search for earthquake precursory information by processing geophysical data from different sources to establish automatically an earthquake forecasting model. The main function of this system is to analyze and process the deformation, fluid, electromagnetic and other geophysical field observing data from ground-based observation, as well as space-based observation. Combined station and earthquake distributions, geological structure and other information, this system can provide a basic software platform for earthquake forecasting research based on spatiotemporal fusion. The hierarchical station tree for data sifting and the interaction mode have been innovatively developed in this SeisGuard system to improve users’ working efficiency. The data storage framework designed according to the characteristics of different time series can unify the interfaces of different data sources, provide the support of data flow, simplify the management and usage of data, and provide foundation for analysis of big data. The final aim of this development is to establish an effective earthquake forecasting model combined all available information from ground-based observations to space-based observations.

Available parking space occupancy change characteristics and short-term forecasting model

Based on an available parking space occupancy （APSO） survey conducted in Nanjing, China, an APSO forecasting model is proposed. The APSO survey results indicate that the time series of APSO with different time-sections are periodical and self-similar, and the fluctuation of the APSO increases with the decrease in time-sections. Taking the short-time change behavior into account, an APSO forecasting model combined wavelet analysis and a weighted Markov chain is presented. In this model, an original APSO time series is first decomposed by wavelet analysis, and the results include low frequency signals representing the basic trends of APSO and several high frequency signals representing disturbances of the APSO. Then different Markov models are used to forecast the changes of low and high frequency signals, respectively. Finally, integrating the predicted results induces the final forecasted APSO. A case study verifies the applicability of the proposed model. The comparisons between measured and forecasted results show that the model is a competent model and its accuracy relies on real-time update of the APSO database.

Comparison of the City Water Consumption Short-Term Forecasting Methods

There are a lot of methods in city water consumption short-term forecasting both inside and outside the country. But among these methods there exist many advantages and shortcomings in model establishing, solving and predicting accuracy, speed, applicability. This article draws lessons from other realm mature methods after many years′ study. It′s systematically studied and compared to predict the water consumption in accuracy, speed, effect and applicability among the time series triangle function method, artificial neural network method, gray system theories method, wavelet analytical method.

Rapid afterslip and short-term viscoelastic relaxation following the 2008 M_W7.9 Wenchuan earthquake

Significant postseismic deformation of the 2008 M W 7.9 Wenchuan earthquake has been observed from GPS data of the first 14 days after the earthquake. The possible mechanisms for the rapid postseismic deformation are assumed to be afterslip on the earthquake rupture plane and viscoelastic relaxation of coseismiclly stress change in the lower crust or upper mantle. We firstly use the constrained least squares method to find an afterslip model which can fit the GPS data best. The afterslip model can explain near-field data very well but shows considerable discrepancies in fitting far-field data. To estimate the effect due to the viscoelastic relaxation in the lower crust, we then ignore the contribution from the afterslip and attempt to invert the viscosity structure beneath the Longmenshan fault where the Wenchuan earthquake occurred from the postseismic deformation data. For this purpose, we use a viscoelastic model with a 2D geometry based on the geological and seismological observations and the coseismic slip distribution derived from the coseismic GPS and InSAR data. By means of a grid search we find that the optimum viscosity is 9×10 18 Pa·s for the middle-lower crust in the Chengdu Basin, 4×10 17 Pa·s for the middle-lower crust in the Chuanxi Plateau and 7×10 17 Pa·s for the low velocity zone in the Chuanxi plateau. The viscoelastic model explains the postseismic deformation observed in the far-field satisfactorily, but it is considerably worse than the afterslip model in fitting the near-fault data. It suggests therefore a hybrid model including both afterslip and relaxation effects. Since the viscoelastic model produces mainly the far-field surface deformation and has fewer degree of freedoms （three viscosity parameters） than the afterslip model with a huge number of source parameters, we fix the viscositiy structure as obtained before but redetermine the afterslip distribution using the residual data from the viscoelastic modeling. The redetermined afterslip distribution becomes physically more reasonable; it is more localized and exhibits a pattern spatially complementary with the coseismic rupture distribution. We conclude that the aseismic fault slip is responsible for the near-fault postseismic deformation, whereas the viscoelastic stress relaxation might be the major cause for the far-field postseismic deformation.

A forecasting model for wave heights based on a long short-term memory neural network

To explore new operational forecasting methods of waves,a forecasting model for wave heights at three stations in the Bohai Sea has been developed.This model is based on long short-term memory(LSTM)neural network with sea surface wind and wave heights as training samples.The prediction performance of the model is evaluated,and the error analysis shows that when using the same set of numerically predicted sea surface wind as input,the prediction error produced by the proposed LSTM model at Sta.N01 is 20%,18%and 23%lower than the conventional numerical wave models in terms of the total root mean square error(RMSE),scatter index(SI)and mean absolute error(MAE),respectively.Particularly,for significant wave height in the range of 3–5 m,the prediction accuracy of the LSTM model is improved the most remarkably,with RMSE,SI and MAE all decreasing by 24%.It is also evident that the numbers of hidden neurons,the numbers of buoys used and the time length of training samples all have impact on the prediction accuracy.However,the prediction does not necessary improve with the increase of number of hidden neurons or number of buoys used.The experiment trained by data with the longest time length is found to perform the best overall compared to other experiments with a shorter time length for training.Overall,long short-term memory neural network was proved to be a very promising method for future development and applications in wave forecasting.

Theory Study and Application of the BP-ANN Method for Power Grid Short-Term Load Forecasting

Aiming at the low accuracy problem of power system short-term load forecasting by traditional methods, a back-propagation artificial neural network （BP-ANN） based method for short-term load forecasting is presented in this paper. The forecast points are related to prophase adjacent data as well as the periodical long-term historical load data. Then the short-term load forecasting model of Shanxi Power Grid （China） based on BP-ANN method and correlation analysis is established. The simulation model matches well with practical power system load, indicating the BP-ANN method is simple and with higher precision and practicality.

Myths about Earthquakes:Quo vadis?

We are living in a world of numbers and calculations with enormous amount of pretty fast user-friendly software ready for an automatic output that may lead to a discovery or,alternatively,mislead to a deceptive conclusion,erroneous claims and predictions.As a matter of fact,nowadays,Science can disclose Natural Hazards,assess Risks,and deliver the state-of-the-art Knowledge of looming disaster in advance catastrophes along with useful Recommendations on the level of risks for decision making regarding engineering design,insurance,and emergency management.

The Collaboratory for the Study of Earthquake Predictability in China:Experiment Design and Preliminary Results of CSEP2.0

Since the inaugural international collaboration under the framework of the Collaboratory for the Study of Earthquake Predictability(CSEP)in 2007,numerous forecast models have been developed and operated for earthquake forecasting experiments across CSEP testing centers(Schorlemmer et al.,2018).Over more than a decade,efforts to compare forecasts with observed earthquakes using numerous statistical test methods and insights into earthquake predictability,which have become a highlight of the CSEP platform.

The Influence of Air Pollution Concentrations on Solar Irradiance Forecasting Using CNN-LSTM-mRMR Feature Extraction

Maintaining a steady power supply requires accurate forecasting of solar irradiance,since clean energy resources do not provide steady power.The existing forecasting studies have examined the limited effects of weather conditions on solar radiation such as temperature and precipitation utilizing convolutional neural network(CNN),but no comprehensive study has been conducted on concentrations of air pollutants along with weather conditions.This paper proposes a hybrid approach based on deep learning,expanding the feature set by adding new air pollution concentrations,and ranking these features to select and reduce their size to improve efficiency.In order to improve the accuracy of feature selection,a maximum-dependency and minimum-redundancy(mRMR)criterion is applied to the constructed feature space to identify and rank the features.The combination of air pollution data with weather conditions data has enabled the prediction of solar irradiance with a higher accuracy.An evaluation of the proposed approach is conducted in Istanbul over 12 months for 43791 discrete times,with the main purpose of analyzing air data,including particular matter(PM10 and PM25),carbon monoxide(CO),nitric oxide(NOX),nitrogen dioxide(NO_(2)),ozone(O₃),sulfur dioxide(SO_(2))using a CNN,a long short-term memory network(LSTM),and MRMR feature extraction.Compared with the benchmark models with root mean square error(RMSE)results of 76.2,60.3,41.3,32.4,there is a significant improvement with the RMSE result of 5.536.This hybrid model presented here offers high prediction accuracy,a wider feature set,and a novel approach based on air concentrations combined with weather conditions for solar irradiance prediction.

A hybrid model for short-term rainstorm forecasting based on a back-propagation neural network and synoptic diagnosis

Rainstorms are one of the most important types of natural disaster in China.In order to enhance the ability to forecast rainstorms in the short term,this paper explores how to combine a back-propagation neural network(BPNN)with synoptic diagnosis for predicting rainstorms,and analyzes the hit rates of rainstorms for the above two methods using the county of Tianquan as a case study.Results showed that the traditional synoptic diagnosis method still has an important referential meaning for most rainstorm types through synoptic typing and statistics of physical quantities based on historical cases,and the threat score(TS)of rainstorms was more than 0.75.However,the accuracy for two rainstorm types influenced by low-level easterly inverted troughs was less than 40%.The BPNN method efficiently forecasted these two rainstorm types;the TS and equitable threat score(ETS)of rainstorms were 0.80 and 0.79,respectively.The TS and ETS of the hybrid model that combined the BPNN and synoptic diagnosis methods exceeded the forecast score of multi-numerical simulations over the Sichuan Basin without exception.This kind of hybrid model enhanced the forecasting accuracy of rainstorms.The findings of this study provide certain reference value for the future development of refined forecast models with local features.

Application test of matter element analysis in earthquake forecast

Calculation by means of the previous indices of the seismic activity can have the matter element analysis possess the forecast function. Readjusting repeatedly the grade limit value of every index can maximize the historical fitting ratio of the calculated and actual grade of the annual maximum magnitude, whose result is relatively ideal.