Optimal control of energy storage systems with power converters: A catenary free tram case study
Author
Delgado, Endika
Term
4. term
Education
Publication year
2015
Submitted on
2015-06-03
Abstract
Dette speciale undersøger, hvordan man modellerer og styrer en sporvogn, der kører uden køreledninger og forsynes af en batteripakke. Målet er at udvikle to styringsstrategier, som kan sammenlignes. Den første følger en mere konservativ, industrinær tilgang. Den anden er modelprædiktiv regulering (MPC), hvor en model bruges til at forudsige den nære fremtid og vælge styresignaler. Arbejdet starter med at opbygge en matematisk model af sporvognen og går derefter videre til designet af regulatorerne. Den konservative strategi består af en række lineære regulatorer og anvender et udvidet Kalman-filter (EKF) som tilstandsestimator (en metode, der kombinerer målinger med modellen for at anslå de indre tilstande). For denne strategi præsenteres simulationsresultater. MPC-strategien beskrives og sættes op, dvs. problemformulering og konfiguration er på plads. MPC bruger modellen til at forudsige systemets udvikling og optimere styringen under givne begrænsninger. Der vises dog ingen resultater for MPC i dette arbejde.
This thesis examines how to model and control a tram that operates without overhead lines and is powered by a battery pack. The goal is to develop two control strategies that can be compared. The first follows a conservative, industry-style approach. The second is Model Predictive Control (MPC), which uses a model to predict near-future behavior and choose control actions. The work begins by building a mathematical model of the tram and then designing the controllers. The conservative strategy uses a series of linear controllers and an Extended Kalman Filter (EKF) as a state estimator (a method that fuses measurements with the model to infer internal states). Simulation results are presented for this strategy. The MPC strategy is described and set up, meaning the problem formulation and configuration are in place. MPC uses the model to predict system evolution and optimize control under given constraints. However, no results for MPC are reported in this thesis.
[This abstract was generated with the help of AI]
Documents