Bookmark and Share

Time Series Analysis - with a focus on Modelling and Forecasting in Energy Systems


Technical University of Denmark


General course objectives:
To give a hands on introduction to the statistical techniques, which are highly useful for modelling based on data observed from energy systems, as well as the use of these e.g. for control.

Learning objectives:
A student who has met the objectives of the course will be able to:
  • Achieve thorough understanding of maximum likelihood estimation techniques.
  • Formulate and apply non-parametric models using kernel functions and splines - with focus on solar and occupancy effects.
  • Formulate and apply time adaptive models.
  • Formulate and apply models for short-term forecasting in energy systems, e.g. for heat load in buildings, electrical power from PV and wind systems.
  • Application of statistical model selection techniques (F-test, likelihood-ratio tests, model validation).
  • Formulate and apply grey-box models - model identification - tests for model order and model validation, and advanced non-linear models.
  • Achieve understanding of model predictive control (MPC) - via applied examples on energy systems.
  • Achieve understanding of flexibility functions and indicies.

Contents:
Generally, one will need a self tuning model for each component in a system, which has only the complexity needed for the particular application. For example, a building with PV and a heat pump, one will need a model from weather forecasts and control variables to: power from the PV, load from the heat pump and the indoor temperature in the building. These, in combination with electricity prices, will in an MPC control be able to optimize the operation of the heat pump and shift the load to achieve the cheapest operation. There are many other applications of the data-driven models, e.g. performance assessment and fault-detection, these will also be presented with examples. The statistical techniques behind the models will be elaborated, with focus on non-linear models, both discrete (kernels and splines) and continuous (grey-box modelling with SDEs).

Back
Course organizer
Peder , Henrik
Place/Venue
Anker Engelunds Vej 1
City
2800 Kgs. Lyngby
Country
Denmark
Workload
2.5
Link
http://kurser.dtu.dk/course/02960