Nonlinear Model Predictive Control in JModelica.org
(2015)Department of Automatic Control
 Abstract
 In this thesis, a stronger support for Model Predictive Control (MPC) in JModelica.org has been implemented. JModelica.org is an opensource software for simulation and optimization of systems described by Modelica models. MPC is an optimizationbased control strategy where one formulates an Optimal Control Problem (OCP) to describe the aim of the controller. At discrete time points the state of the system is estimated and the OCP is solved to find the optimal input to apply to the system. The main goal of this thesis has been to make the time it takes to obtain the optimal input as short as possible and also streamlining the setup of MPC in JModelica.org. This has been done by implementing an MPC class, which utilizes the fact that the... (More)
 In this thesis, a stronger support for Model Predictive Control (MPC) in JModelica.org has been implemented. JModelica.org is an opensource software for simulation and optimization of systems described by Modelica models. MPC is an optimizationbased control strategy where one formulates an Optimal Control Problem (OCP) to describe the aim of the controller. At discrete time points the state of the system is estimated and the OCP is solved to find the optimal input to apply to the system. The main goal of this thesis has been to make the time it takes to obtain the optimal input as short as possible and also streamlining the setup of MPC in JModelica.org. This has been done by implementing an MPC class, which utilizes the fact that the structure of the OCP is the same in each consecutive sample for efficiency. Two different benchmarks, one on a smaller problem and one on a larger problem, shows that by using the new MPC framework we obtain similar results as before, but considerably faster. The total average computation time for one sample is decreased by almost 60% for the large problem and by almost 90% for the smaller problem. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/studentpapers/record/7760729
 author
 Axelsson, Magdalena
 supervisor

 Fredrik Magnusson ^{LU}
 Anders Rantzer ^{LU}
 organization
 year
 2015
 type
 H3  Professional qualifications (4 Years  )
 subject
 ISSN
 02805316
 other publication id
 ISRN LUTFD2/TFRT5987SE
 language
 English
 id
 7760729
 date added to LUP
 20150818 09:21:02
 date last changed
 20150818 09:21:02
@misc{7760729, abstract = {In this thesis, a stronger support for Model Predictive Control (MPC) in JModelica.org has been implemented. JModelica.org is an opensource software for simulation and optimization of systems described by Modelica models. MPC is an optimizationbased control strategy where one formulates an Optimal Control Problem (OCP) to describe the aim of the controller. At discrete time points the state of the system is estimated and the OCP is solved to find the optimal input to apply to the system. The main goal of this thesis has been to make the time it takes to obtain the optimal input as short as possible and also streamlining the setup of MPC in JModelica.org. This has been done by implementing an MPC class, which utilizes the fact that the structure of the OCP is the same in each consecutive sample for efficiency. Two different benchmarks, one on a smaller problem and one on a larger problem, shows that by using the new MPC framework we obtain similar results as before, but considerably faster. The total average computation time for one sample is decreased by almost 60% for the large problem and by almost 90% for the smaller problem.}, author = {Axelsson, Magdalena}, issn = {02805316}, language = {eng}, note = {Student Paper}, title = {Nonlinear Model Predictive Control in JModelica.org}, year = {2015}, }