Vis enkel innførsel

dc.contributor.advisorGravdahl, Jan Tommynb_NO
dc.contributor.advisorUddin, Nurnb_NO
dc.contributor.authorOvervåg, Thomas Ferstadnb_NO
dc.date.accessioned2014-12-19T14:06:32Z
dc.date.available2014-12-19T14:06:32Z
dc.date.created2013-04-15nb_NO
dc.date.issued2013nb_NO
dc.identifier616147nb_NO
dc.identifierntnudaim:8656nb_NO
dc.identifier.urihttp://hdl.handle.net/11250/260785
dc.description.abstractThe work presented in this thesis examines the possibilities of having compressors running in an optimal manner which can result in energy savings. This research looks at how a compressor operates and the problems which can occur when connecting several machines, both in series and parallel. It is mainly focused on using Model Predictive Control (MPC), as a setup for controlling each compressor to a fixed operating point on the characteristic with relation to mass flow and pressure. Constraints are set on the controller outputs to help minimize the area of operation. Several types of efficiency are investigated in detail and this has helped to create a continuous efficiency island definition. These continuous curves are assumed to be equal to the characteristic plotted from tests and measurements in the lab. The continuous efficiency definition enables the possibility of calculating an operating point for each compressor. This is done using a Quadric Programming setup, which is calculated explicitly beforehand. Together with this and the implementation of MPC, the result is a load sharing scheme used here on two compressors connected in parallel. The control inputs to the plant being the impeller rotational speeds and the outflow from the throttle valve, while the measurements are in the form of plenum pressure and mass flow from each machine.nb_NO
dc.languageengnb_NO
dc.publisherInstitutt for teknisk kybernetikknb_NO
dc.titleCentrifugal Compressor Load Sharing with the use of MPCnb_NO
dc.typeMaster thesisnb_NO
dc.source.pagenumber130nb_NO
dc.contributor.departmentNorges teknisk-naturvitenskapelige universitet, Fakultet for informasjonsteknologi, matematikk og elektroteknikk, Institutt for teknisk kybernetikknb_NO


Tilhørende fil(er)

Thumbnail
Thumbnail
Thumbnail

Denne innførselen finnes i følgende samling(er)

Vis enkel innførsel