Georges Kariniotakis

Wind power forecasting using advanced neural networks models

In this paper, an advanced model, based on recurrent high order neural networks, is developed for the prediction of the power output profile of a wind park. This model outperforms simple methods like persistence, as well as classical methods in the literature. The architecture of a forecasting model is optimised automatically by a new algorithm, that substitutes the usually applied trial-and-error method. Finally, the online implementation of the developed model into an advanced control system for the optimal operation and management of a real autonomous wind-diesel power system, is presented.

Standardizing the Performance Evaluation of Short-Term Wind Power Prediction Models

Short-term wind power prediction is a primary requirement for efficient large-scale integration of wind generation in power systems and electricity markets. The choice of an appropriate prediction model among the numerous available models is not trivial, and has to be based on an objective evaluation of model performance. This paper proposes a standardized protocol for the evaluation of short-term windpower prediction systems. A number of reference prediction models are also described, and their use for performance comparison is analysed. The use of the protocol is demonstrated, using results from both on-shore and offshore wind farms. The work was developed in the frame of the Anemos project (EU R&D project) where the protocol has been used to evaluate more than 10 prediction systems.

Conditional Prediction Intervals of Wind Power Generation

A generic method for the providing of prediction intervals of wind power generation is described. Prediction intervals complement the more common wind power point forecasts, by giving a range of potential outcomes for a given probability, their so-called nominal coverage rate. Ideally they inform of the situation-specific uncertainty of point forecasts. In order to avoid a restrictive assumption on the shape of forecast error distributions, focus is given to an empirical and nonparametric approach named adapted resampling. This approach employs a fuzzy inference model that permits to integrate expertise on the characteristics of prediction errors for providing conditional interval forecasts. By simultaneously generating prediction intervals with various nominal coverage rates, one obtains full predictive distributions of wind generation. Adapted resampling is applied here to the case of an onshore Danish wind farm, for which three point forecasting methods are considered as input. The probabilistic forecasts generated are evaluated based on their reliability and sharpness, while compared to forecasts based on quantile regression and the climatology benchmark. The operational application of adapted resampling to the case of a large number of wind farms in Europe and Australia among others is finally discussed.

Wind power forecasting using fuzzy neural networks enhanced with on-line prediction risk assessment

The paper presents an advanced wind forecasting system that uses on-line SCADA measurements, as well as numerical weather predictions (NWP) as input, to predict the power production of wind parks 48 hours ahead. The prediction system integrates models based on adaptive fuzzy-neural networks configured either for short-term (1-10 hours) or long-term (1-48 hours) forecasting. The paper presents detailed one-year evaluation results of the models on the case study of Ireland, where the output of several wind farms is predicted using HIRLAM meteorological forecasts as input. A method for the online estimation of confidence intervals of the forecasts is developed together with an appropriate index for assessing online the risk due to the inaccuracy of the numerical weather predictions.

A Stochastic Dynamic Programming Model for Optimal Use of Local Energy Resources in a Market Environment

The unbundling of power systems and the emergence of electricity markets favor the deployment of distributed generation in electricity networks. Microgrids are low-voltage distribution networks comprising micro-generation, storage devices and controllable loads that can operate interconnected or isolated from the main distribution grid as a controlled entity. This paper proposes a new method, based on stochastic optimization, to optimize the operation of a microgrid. This involves optimization of the production of the local micro- sources and storage and the exchange with the main distribution grid subject to market conditions. The results of the stochastic approach, and a comparison with a simpler deterministic one are presented using as case study a typical microgrid.

Energy management and control of island power systems with increased penetration from renewable sources

Penetration of renewable energy sources in isolated and weakly interconnected power systems can be increased in a secure and reliable way, if advanced control tools are available to the operators of these systems. In this paper the functions of MORE CARE are described. This is an advanced control software system aiming to optimize the operation of isolated and weakly interconnected systems by increasing the share of wind energy and other renewable forms, taking into account pumped hydro storage facilities and providing advanced on-line security functions, both in preventive and corrective mode. The main features of the control system comprise advanced software modules for load and wind power forecasting, unit commitment and economic dispatch of the conventional and renewable units and on-line security assessment capabilities integrated in a friendly man-machine environment. In this way, penetration of renewable energy sources in isolated systems can be increased in a secure and reliable way. Pilot installations of advanced control functions are foreseen on the islands of Crete, Ireland and Madeira.

An advanced control system for the optimal operation and management of medium size power systems with a large penetration from renewable power sources

An advanced control system for the optimal operation and management of autonomous wind-diesel systems is presented. This system minimises the production costs through an on-line optimal scheduling of the power units, which takes into account the technical constraints of the diesel units, as well as short-term forecasts of the load and renewable resources. The power system security is maximised through on-line security assessment modules, which enable the power system to withstand sudden changes in the production of the renewable sources. The control system was evaluated using data from the island of Lemnos, where it has been installed and operated since January 1995.

Operation and control of island systems-the Crete case

In this paper, an advanced control system for the optimal operation and management of isolated power systems with increased renewable power integration is presented. The control system minimises the production costs through on-line optimal scheduling of the power units, taking into account short-term forecasts of the load and the renewable resources. The power system security is supervised via on-line security assessment modules, which emulate the power system frequency changes caused by pre-selected disturbances. For each of the above functions, a number of techniques have been applied, both conventional and AI based. The system has been installed in the dispatch center of Crete since June 1999, and is under evaluation.

A local energy management system for solar integration and improved security of supply: The Nice Grid project

This paper describes Nice Grid, a demonstration project part of the European initiative Grid4EU. The project aims at developing a smart solar neighbourhood in the urban area of the city of Nice, France. The four year project started in November 2011 and will test the suitability of recent developments in distribution networks management for facilitating the connection of distributed renewable generators, improving the security of supply and let customers and other actors to provide network services. The idea behind Nice Grid is to combine controllable distributed electricity and thermal storage devices with forecasts of solar power production and load in a local energy management system. The paper, which represents a useful reference for the project, presents also a detailed overview of relevant European demonstration projects on Smart Grid.

Preliminary results from the More Advanced Control Advice Project for secure operation of isolated power systems with increased renewable energy penetration and storage

In this paper, preliminary results from MORE CARE, a European R&D project financed within the Energy Program are described. This project has, as its main objective, the development of an advanced control software system, aiming to optimize the overall performance of isolated and weakly interconnected systems in liberalized market environments by increasing the share of wind energy and other renewable forms, including advanced online security functions. The main features of the control system comprise advanced software modules for load and wind power forecasting, unit commitment and economic dispatch of the conventional and renewable units and online security assessment capabilities integrated in a friendly man-machine environment. Pilot installations of advanced control functions are foreseen on the islands of Crete, Ireland and Madeira.