Nonlinear Model Predictive Control and Estimation applied to Selective Catalytic Reduction
Aaltonen, Oscar (2022)
Aaltonen, Oscar
2022
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi-fe2022091559140
https://urn.fi/URN:NBN:fi-fe2022091559140
Tiivistelmä
Nonlinear Model Predictive Control (NMPC) is an advanced optimization-based control method for both linear and nonlinear dynamical systems. In this thesis, a NMPC software is developed in Matlab to control a Selective Catalytic Reduction (SCR) process, which is a process to reduce nitrogen oxide emissions from diesel and gas engines using ammonia or a urea solution. The SCR model that is used in this thesis is modeled as a state space model consisting of three nonlinear ordinary differential equations. A simplified nonlinear version of this model is used in the NMPC as a prediction model. State estimation is used to estimate missing measurements from the SCR process; a Moving Horizon Estimator (MHE) is implemented in Matlab for this purpose. Since no theory is available for this kind of nonlinear output feedback MPC, the results of the control and estimation are presented through simulation. The simulations show that the SCR can be controlled with only a few measurements using MHE and NMPC. A major advantage with NMPC is that the ammonia slip can also be controlled. Some mathematical results of NMPC combined with nonlinear MHE are discussed and MHE convergence for a linear detectable plant is proved, slightly improving the corresponding results in the research literature.
Kokoelmat
- 111 Matematiikka [39]