The deployment of smart grids and renewable energy dispatch centers motivates the development of forecasting techniques that take advantage of near real‐time measurements collected from geographically distributed sensors. This paper describes a forecasting methodology that explores a set of different sparse structures for the vector autoregression (VAR) model using the least absolute shrinkage and selection operator (LASSO) framework. The alternating direction method of multipliers is applied to fit the different LASSO‐VAR variants and create a scalable forecasting method supported by parallel computing and fast convergence, which can be used by system operators and renewable power plant operators. A test case with 66 wind power plants is used to show the improvement in forecasting skill from exploring distributed sparse structures. The proposed solution outperformed the conventional autoregressive and vector autoregressive models, as well as a sparse VAR model from the state of the art. Copyright © 2016 John Wiley & Sons, Ltd.

Forecasting the hourly spot price of day-ahead and intraday markets is particularly challenging in electric power systems characterized by high installed capacity of renewable energy technologies. In particular, periods with low and high price levels are difficult to predict due to a limited number of representative cases in the historical dataset, which leads to forecast bias problems and wide forecast intervals. Moreover, these markets also require the inclusion of multiple explanatory variables, which increases the complexity of the model without guaranteeing a forecasting skill improvement. This paper explores information from daily futures contract trading and forecast of the daily average spot price to correct point and probabilistic forecasting bias. It also shows that an adequate choice of explanatory variables and use of simple models like linear quantile regression can lead to highly accurate spot price point and probabilistic forecasts. In terms of point forecast, the mean absolute error was 3.03 e/MWh for day-ahead market and a maximum value of 2.53 e/MWh was obtained for intraday session 6. The probabilistic forecast results show sharp forecast intervals and deviations from perfect calibration below 7% for all market sessions.

Urban wastewater sector is being pushed to optimize processes in order to reduce energy consumption without compromising its quality standards. Energy costs can represent a significant share of the global operational costs (between 50% and 60%) in an intensive energy consumer. Pumping is the largest consumer of electrical energy in a wastewater treatment plant. Thus, the optimal control of pump units can help the utilities to decrease operational costs. This work describes an innovative predictive control policy for wastewater variable-frequency pumps that minimize electrical energy consumption, considering uncertainty forecasts for wastewater intake rate and information collected by sensors accessible through the Supervisory Control and Data Acquisition system. The proposed control method combines statistical learning (regression and predictive models) and deep reinforcement learning (Proximal Policy Optimization). The following main original contributions are produced: i) model-free and data-driven predictive control;ii) control philosophy focused on operating the tank with a variable wastewater setpoint level; iii) use of supervised learning to generate synthetic data for pre-training the reinforcement learning policy, without the need to physically interact with the system.The results for a real case-study during 90 days show a 16.7% decrease in electrical energy consumption while still achieving a 97% reduction in the number of alarms (tank level above 7.2 meters) when compared with the current operating scenario (operating with a fixed set-point level). The numerical analysis showed that the proposed data-driven method is able to explore the trade-off between number of alarms and consumption minimization, offering different options to decision-makers.

A recent research trend is driven to increase the monitoring and control capabilities of low voltage networks. This paper describes a probabilistic forecasting methodology based on kernel density estimation and that makes use of distributed computing techniques to create a highly scalable forecasting system for LV networks. The results show that the proposed algorithm outperforms three benchmark models (one for point forecast and two for probabilistic forecasts) and demonstrate the applicability of the distributed in-memory computing solution for a practical operational scenario. The ultimate goal is to integrate information about net-load forecasts in power flow optimization frameworks for low voltage networks in order to solve technical constraints with the available home energy management system flexibility.