Bp 122 en

Z DCEwiki
Verze z 26. 8. 2014, 22:57, kterou vytvořil Sturcmar (diskuse | příspěvky) (Zobrazení PDF náhledů v rámečkách.)
(rozdíl) ← Starší verze | zobrazit aktuální verzi (rozdíl) | Novější verze → (rozdíl)
Skočit na navigaci Skočit na vyhledávání

Predictive control of power balance in electric grid

Author: Ondřej Malík

Bakalářské práce 2007

Download thesis in PDF

The problem of optimal utilization of regulation reserves (provided by the Ancillary Services) from the Transmission System Operator point of view received little attention to date. The purpose of this thesis is to create an algorithm which will find a schedule of Ancillary Services activation so that their regulation energy covers the predicted deviation between the power production and the power demand (the System Deviation prediction) with minimal expense for the regulation energy. It is therefore a specific regulation problem which utlilizes methods of optimal scheduling. This problem has some uncommon properties resulting from technical limitations of generating units which provide the Ancillary Services and from agreements which specify how the services should be utilized. The scheduling algorithm must take these into account, moreover, inclusion of services with fast reaction times requires a fast sampling rate, which puts high requirements on algorithm efficiency.

The proposed algorithm is based on Mixed Integer Linear Programming, as integer variables allow for modelling of some properties that would be hard to model by using continuous variables. Included in the algorithm are typical properties of generating units such as ramp rate limitations, minimal time the reserve power of generating unit must be activated or deactivated, startup costs and costs of regulation energy, along with special penalizations for reactivating the reserve power of generating unit too early after its deactivation and deactivation of reserve power too early after its activation or for number of reserve power activations. The results show that the algorithm allow for efficiently influencing the way the services are activated while respecting the services properties such as constant ramp rates, startup delay and other. Performed tests indicate that the algorithm is applicable even for problems with size which is close to the size of problems dealt with when controlling the Czech Transmission System and therefore may serve as a decision support tool for Transmission System Operator dispatchers.


Bp 2007 malik ondrej.pdf