LQG Control with Matlab, 5–7 cr

Core content for 5 cr

• Quadratic norms of vectors, matrixces, functions and systems
• Quadratic non-norm performance indices
  
• Random variables & stochastic performance indices like variances of the system responses.
• Weighted Least Squares
  
• Use of suitable matrix factorizations, Lagrange multipliers
  
• State-feedback, Kalman predictor and filter
• Stationary LQG regulation and control
• Deterministic finite horizon LQ regulation 
  
• Parseval Theorems
  
• Exponential time-weighting
  
• Spectral factorization.
• Transfer function matrix methods. Frequency response studies. Principal gains and Hinf norm. Studies of robust stability using unstructured uncertainty, too.
• Introduction to DT Square Root Algorithms.

Complementary knowledge  for 5 cr

• Quadratic costs with polynomial time weighting
• Balanced realization of state space models
• LQ methods for model reduction
• LQ design of partial state feedback regulation
• LQ designs using Hamilton matrix

Specialist knowledge for 5 cr

• Use of Matlab tools ode45, dde23 (for differential equation models) &  fzero, fsolve (to solve algebraic equations) & fminbnd, fminsearch (to minimize and maximize algebraic functions).

Extension C for an extra 1 cr

• Introduction to Model Predictive Control of MIMO plants
• Generalized Predictive Control of SISO plants
• Constrained Model Predictive Control of MIMO plants

Extension D for an extra 1 cr

• DT Kalman Filter in Recursive Identification
• Morf-Kailath Algorithms for DT Kalman Filter
• Special Lyapunov and Riccati Algorithms