Javascript must be enabled for the correct page display

Output Prediction for non Demand-Driven Power Systems using Neural Networks

Siljee, B.I.J. (2003) Output Prediction for non Demand-Driven Power Systems using Neural Networks. Master's Thesis / Essay, Computing Science.

Infor_Ma_2003_BIJSiljee.CV.pdf - Published Version

Download (2MB) | Preview


In the Dutch liberalized electricity market, parties involved in the use of the electricity network need to send their planned electricity activities to the national grid administrator each day in advance. When a market party creates imbalance by deviating from the planned activity, it has to pay imbalance costs. The output of non demand-driven power systems is irregular and depends on different stochastic factors. To minimize the producer's imbalance costs, the power output has to be accurately predicted.This thesis investigates the output prediction of non demand-driven power systems using neural networks. Models based on the multilayer perceptron neural network are used for this purpose. The results obtained when building models based on real-life data, as well as the simulation of the models in use are described. A detailed analysis of the influence of various parameters in neural network based modeling is given, like input relevance and the optimal division of one year for separate models.It is shown that ensembles of these neural networks are able to generalize from real-life dataand achieve a reasonably well performance in predicting the power output.

Item Type: Thesis (Master's Thesis / Essay)
Degree programme: Computing Science
Thesis type: Master's Thesis / Essay
Language: English
Date Deposited: 15 Feb 2018 07:29
Last Modified: 15 Feb 2018 07:29

Actions (login required)

View Item View Item