Implementation of Kalman Filter, Extended Kalman Filter and Moving Horizon Estimation to the stirred tank mixing process. This repository uses the same system as the one used in Implementation and comparison of Advanced process control to stirred tank mixing process.
The Continuously stirred mixing tank in the figure has two time-varying inlets F1(t) and F2(t) with different density. The density of both the inlets is constant and is given by rho1 and rho2. It is assumed that the tank is well mixed so that outlet F(t) has the same density as the density in the tank i.e. rho(t). The volume of the tank occupied by the liquid is V(t) and the corresponding height is h(t) with the surface area S.
- KF :
KF_3.mMain File.
- EKF:
EKF1.m Main file.
lin1.m To compute Jacobian at different states.
- MHE : For further information about MHE and the code flow kindly refer to the MHE report.
MHE_3.mMain File.
MHE_costTo calculate the total cost at every time step.
lin1.mTo compute Jacobian at different states.
mhe.mTo perform constrained non-linear optimization for state estimation.
sys.mTo calculate one step ahed prediction.
Code Flow Diagram for MHE:
