Hello everyone,
I’m currently working on a control system project where I need to implement a state-space approach for designing a robust controller. While I have a decent understanding of classical control techniques like PID and root locus, I’m relatively new to state-space methods and would appreciate some guidance.
The system I’m working on is a multi-input, multi-output (MIMO) system with some nonlinearity. My main objectives are:
- State Estimation: I need to design an observer to estimate the full state vector. Would a Luenberger observer be a good choice, or should I consider a Kalman filter given that my system has process noise?
- Controller Design: I’m considering using pole placement or Linear Quadratic Regulator (LQR) for state feedback control. What are the trade-offs between these methods, particularly regarding robustness and implementation complexity?
- Discrete vs. Continuous Implementation: The system is digital, but the initial model is continuous. Would it be better to discretize the system first and then design the controller, or should I design the controller in continuous time and then discretize?
- Dealing with Model Uncertainty: My system parameters are not perfectly known. Are there recommended robust control techniques in the state-space framework that can handle such uncertainties effectively?
Any insights, references, or practical tips would be greatly appreciated! If anyone has faced similar challenges, I’d love to hear about your experiences.
I also checked this: https://state-space.ieeecss.org/d/4377-phd-position-on-algorithmic-fairness-for-salesforce-admin-data-driven-decision-making
Thanks in advance!
