Search

Buildings are one of the largest consumers of electrical energy, making it important to develop different strategies to help to reduce electricity consumption. Building energy consumption forecasting strategies are widely used to support demand management decisions, but these strategies require large data sets to achieve an accurate electric consumption forecast, so they are not commonly used for buildings with a short history of record keeping. Based on this, the objective of this study is to determine, through continuous hourly electricity consumption forecasting strategies, the amount of data needed to achieve an accurate forecast. The proposed forecasting strategies were evaluated with Random Forest, eXtreme Gradient Boost, Convolutional Neural Network, and Temporal Convolutional Network algorithms using 4 years of electricity consumption data from two buildings located on the campus of the University of Valladolid. For performance evaluation, two scenarios were proposed for each of the proposed forecasting strategies. The results showed that for forecasting horizons of 1 week, it was possible to obtain a mean absolute percentage error (MAPE) below 7% for Building 1 and a MAPE below 10% for Building 2 with 6 months of data, while for a forecast horizon of 1 month, it was possible to obtain a MAPE below 10% for Building 1 and below 11% for Building 2 with 10 months of data. However, if the distribution of the data captured in the buildings does not undergo sudden changes, the decision tree algorithms obtain better results. However, if there are sudden changes, deep learning algorithms are a better choice. [ABSTRACT FROM AUTHOR]

This study introduces an improved hypocentral version of the space‐time Epidemic‐Type Aftershock Sequence (ETAS) model, entitled as the 3D‐finite source (3D‐FS) ETAS model, and applies it to the analysis of the Southern California earthquake catalog. By stochastic reconstruction, we are able to reconstruct the patterns of aftershock productivity density along the mainshock ruptures. Detailed analysis of the productivity patterns reveals that: (1) Directly triggered aftershocks make up 21% to 41% of all earthquakes within the mainshock rupture areas, and show significant spatial heterogeneity and temporal migrations; (2) Major aftershocks tend to locate in low productivity areas, at the edges of clusters formed by small aftershocks; (3) Large slip areas are depleted of aftershocks, over 60% of all productivity distributes in areas with slip less than 0.3 times of the maximum slip, and the trajectory of the productivity pattern on the fault plane demonstrates apparent compensation to coseismic slip. We relate the difference in triggering abilities of four mainshocks to the heat flow in corresponding regions. Simulation results suggest that the 3D‐FS ETAS model has apparent advantages of improving the performance of short‐term aftershock forecast. Moreover, the later aftershocks are more correlated with the locations of subsequent events than earlier aftershocks, suggesting that the migration of aftershocks is important for mitigating aftershock hazard. Plain Language Summary: How many aftershocks can be produced at each part along the rupture of a large earthquake is crucial for us to understand how aftershocks are generated and how the strain energy in the crust relaxes afterward. It is difficult to know whether an earthquake is an aftershock triggered by some previous earthquake or a background earthquake that has no ancestry since earthquake clusters not only overlap with each other but also with background seismicity in space and time. We proposed a new earthquake clustering model to disentangle such overlapping and quantify family trees among the earthquakes. Our study on the earthquake data in Southern California shows that aftershocks tend to occur at low slip areas in the mainshock rupture and that large aftershocks occur farther away near the boundary of such concentration of aftershocks. Moreover, this model improves the short‐term aftershock forecasting. Key Points: Large aftershocks occur in the edge of areas with high aftershock productivityAftershock productivity exhibit compensating patterns in association with the coseismic slipThe mainshock ruptures constrained from early aftershocks help improve aftershock forecasting [ABSTRACT FROM AUTHOR]

The sovereign debt crisis has increased the importance of monitoring budgetary execution. We employ real‐time data using a mixed data sampling (MiDaS) methodology to demonstrate how budgetary slippages can be detected early on. We show that in spite of using real‐time data, the year‐end forecast errors diminish significantly when incorporating intra‐annual information. Our results show the benefits of forecasting aggregates via subcomponents, in this case total government revenue and expenditure. Our methodology could significantly improve fiscal surveillance and could therefore be an important part of the European Commission's model toolkit. [ABSTRACT FROM AUTHOR]

Short‐term wind‐power forecasting methods like neural networks are trained by empirical risk minimization. The local optimum and overfitting problem is likely to occur in the model‐training stage, leading to the poor ability of reasoning and generalization in the prediction stage. To solve the problem, a model of short‐term wind power forecasting is proposed based on 2‐stage feature selection and a supervised random forest in the paper. First, in data preprocessing, some redundant features can be removed by a variable importance measure method and intimate samples can be selected based on relevant analysis, so that the efficiency of model training and the correlation degree between input and output samples can be enhanced. Second, an improved supervised random forest (RF) methodology is proposed to compose a new RF based on evaluating the performance of each decision tree and restructuring the decision trees. A new index of external validation in correlation with wind speed in numerical weather prediction has been proposed to overcome the shortcomings of the internal validation index that seriously depends on the training samples. The simulation examples have verified the rationality and feasibility of the improvement. Case studies of measured data from a wind farm have shown that the proposed model has a better performance than the original RF, back propagation neural network, Bayesian network, and support vector machine, in aspects of ensuring accuracy, efficiency, and robustness, and especially if there is high rate of noisy data and wind power curtailment duration in the historical data. [ABSTRACT FROM AUTHOR]

Distributed generation has many potential benefits including use of renewable resources, increased customer participation, and decreased losses. However, as the penetration of distributed renewable energy sources increases, the technical challenges of integrating these resources into the power system increase as well. One such challenge is the rapid variation of voltages along distribution feeders in response to photovoltaic (PV) output fluctuations, and the reactive power capability of PV inverters can be used to address this challenge. A method of achieving optimal expected performance with respect to a figure of merit of interest to the distribution system operator while maintaining appropriate system voltage magnitudes and considering the uncertainty of PV power injections is proposed. The method utilises reactive power injection both to improve system performance and to compensate for variations in active power injection. It requires infrequent communication between the distribution system operator and the PV inverters and bases its decisions on short‐term forecasts, formulating voltage magnitude requirements as chance constraints. The proposed method is validated using the IEEE 123‐node radial distribution test feeder and shown to improve the distribution system performance (with respect to existing methods) and maintain suitable voltages. [ABSTRACT FROM AUTHOR]

In liberalized electricity markets, the electricity generation companies usually manage their production by developing hourly bids that are sent to the day-ahead market. As the prices at which the energy will be purchased are unknown until the end of the bidding process, forecasting of spot prices has become an essential element in electricity management strategies. In this article, we apply forecasting factor models to the market framework in Spain and Portugal and study their performance. Although their goodness of fit is similar to that of autoregressive integrated moving average models, they are easier to implement. The second part of the paper uses the spot-price forecasting model to generate inputs for a stochastic programming model, which is then used to determine the company's optimal generation bid. The resulting optimal bidding curves are presented and analyzed in the context of the Iberian day-ahead electricity market. [ABSTRACT FROM AUTHOR]

Travel time is a good operational measure of the effectiveness of transportation systems. The ability to accurately predict motorway and arterial travel times is a critical component for many intelligent transportation systems (ITS) applications. Advanced traffic data collection systems using inductive loop detectors and video cameras have been installed, particularly for motorway networks. An inductive loop can provide traffic flow at its location. Video cameras with image-processing software, e.g. Automatic Number Plate Recognition (ANPR) software, are able to provide travel time of a road section. This research developed a dynamic linear model (DLM) model to forecast short-term travel time using both loop and ANPR data. The DLM approach was tested on three motorway sections in southern England. Overall, the model produced good prediction results, albeit large prediction errors occurred at congested traffic conditions due to the dynamic nature of traffic. This result indicated advantages of use of the both data sources. Copyright © 2008 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

This paper uses the probit model to examine whether leading indicator information could be used for the purpose of predicting short-term shifts in demand for business travel by air to and from the UK. Leading indicators considered include measures of business expectations, availability of funds for corporate travel and some well-known macroeconomic indicators. The model performance is evaluated on in- and out-of-sample basis, as well as against a linear leading indicator model, which is used to mimic the current forecasting practice in the air transport industry. The estimated probit model is shown to provide timely predictions of the early 1980s and 1990s industry recessions and is shown to be more accurate than the benchmark linear model. Copyright © 2005 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

In this paper forecasting the production volume of the Swedish and Finnish metal and engineering industries one quarter ahead is discussed. A practical way of making use of the predictive information in the answers of the quarterly business survey is presented which is based on the application of the Kalman filter. It is found that the most informative questions of the business survey from the point of view of forecasting are different in the two countries. However, for both Sweden and Finland, the improvement in prediction accuracy after taking account of relevant business survey information is significant when the precision of autoprojective forecasts is used as a baseline. [ABSTRACT FROM AUTHOR]

To achieve further improvements in quantitative predictability, a chaos‐based adaptive forecasting method for the critical frequency of the F2 layer (foF2) is proposed for the development of an ionospheric forecasting technique for one hour ahead. This method has three new characteristics. (1) It is based on Volterra filters and it has a simplified structure with easy implementation. (2) Based only on past measured data, it can forecast foF2 values without the requirement for past or forecast values of any solar and geomagnetic indices. (3) It can achieve high forecast accuracy with a small training dataset of 27 days (one solar rotation period). Diurnal, seasonal, and annual comparisons of measured and forecasted foF2 values are presented to illustrate the applicability and suitability of the proposed method. Statistical results reveal that the foF2 values calculated using the proposed model are consistent with the trend of measurements irrespective of whether geomagnetic conditions are quiet or disturbed. The average RMSE and RRMSE values were 0.86 MHz and 17.36%, respectively, when using measured data from periods of past 27 days during 2008–2015. The proposed method has potential to forecast other ionospheric characteristic parameters, and that it could achieve satisfactory regional or global 1‐hr forecasting if combined with a spatial reconstruction technique. Plain Language Summary: The critical frequency of F2 layer of ionosphere (foF2) is a very important characteristic parameter in various civil and military applications. To achieve further improvements in quantitative predictability, a chaos‐based adaptive forecasting method for foF2 is proposed for the development of an ionospheric forecasting technique for 1 hr ahead. This method has three advantages: simplified structure with easy implementation, high forecast accuracy with a small training dataset, and forecast without the requirement for any solar and geomagnetic indices. Comparisons between the measured and forecasting values of foF2 have shown the usability and appropriateness of the proposed method. Statistical results reveal that the foF2 calculated by the proposed model is consistent with the trend of the measurements whenever quiet and disturbed geomagnetic conditions. The average RMSE and RRMSE are 0.86 MHz and 17.36% using the past 27 days measured data during the year of 2008–2015. Moreover, the analysis results have proved the proposed method has the potential to forecasting any other ionospheric characteristic parameters and can achieve 1‐hr forecasting for global by combining spatial reconstruction technique. Key Points: Chaos and nonlinear dynamic theory have been applied to forecast foF2 short‐term variation valueThe forecasting procedure is achieved by chaotic space reconstruction and Volterra series adaptive filter techniqueThe proposed method has good accuracy in terms of forecasting values of foF2 with 27‐day training data [ABSTRACT FROM AUTHOR]

The increase in floods and flash floods over the last decades has motivated researchers to develop improved methodologies for flood risk prevention and warning. Flood forecasting models available today have evolved technologically but are subject to limitations due to the lack of data and limited community participation. This paper presents the Hydrological Alert Model with Participatory Basis (HAMPB) model, an approach for integrating water level data reported by citizens, which has the advantage of being inexpensive and potentially highly available, with traditional data to improve flood forecasting. The model assimilates spatiotemporal water levels measured in the field when they are available through a real‐time estimator. We added random perturbations of up to |10| and |15| cm to those data using the Monte Carlo Method to mimic the uncertainty in citizen science data collection. Applying the HAMPB model for urban nested‐scale catchments (0.11 km2 ≤ Area ≤ 21.84 km2) in Brazil shows: (a) significant improvements in flood simulations when field data was assimilated even considering the volunteered data uncertainty; (b) capability to update simulations in more than one point in the semi‐distributed hydrological model by a regionalization method; and (c) flood hazard indexes and their uncertainties show better estimations using field data for updating. [ABSTRACT FROM AUTHOR]