how can I do an antiwindup, on my state feedback control (with LQR, intergral action and an observer)?

I think with a saturation in integral action, before of the process, I'm not sure.

Answer 1

[Atharva]

Hey Studentro,

To implement anti-windup in a state feedback control system with LQR (Linear Quadratic Regulator), integral action, and an observer, you can follow these general steps:

1. Design the LQR controller: Begin by designing the state feedback controller using LQR. Determine the desired weighting matrices (Q and R) to define the performance and control effort criteria for the system. Use the lqr function in MATLAB to compute the optimal gain matrix.
2. Design the observer: Design an observer (also known as a state estimator) to estimate the system states. You can use techniques like Luenberger observer or Kalman filter. Calculate the observer gain matrix using the place function or by solving the observer pole placement problem.
3. Implement the control system: Combine the LQR controller and the observer to form the full control system. The control law can be formulated as:

• u = -K * x_hat + Ki * integral_error where u is the control input, x_hat is the estimated state from the observer, K is the LQR gain matrix, and Ki is the integral gain.

4.Implement integral action: To incorporate integral action, introduce an integrator block in the control loop. You can create an additional state (integral_error) that accumulates the integral of the tracking error. Update the integral_error at each time step using the error between the desired and estimated states.

5.Anti-windup mechanism: To prevent windup, you need to limit the integration term when the control effort saturates. This ensures that the integral action does not cause the system to exceed its physical limits. One way to implement anti-windup is by using a saturation block on the integrator output. The saturation block should have upper and lower bounds that represent the saturation limits of the actuator.

6.Simulate and tune: Simulate the system with the implemented control, observer, and anti-windup mechanism. Tune the gains of the LQR controller, observer, and integral action to achieve desired performance and stability. You may need to experiment with different gain values and observe the system's response to refine the control parameters.

These implementation details may vary depending on the specific system and control requirements. The steps provided here offer a general guideline for incorporating anti-windup in a state feedback control system with LQR, integral action, and an observer.