Chapter 7: Closed-Loop CCD and Information Availability

Designing systems that can actually be implemented¶

Closed-loop CCD treats sensing, estimation, prediction, sampling, computation, and controller structure as design resources. Information availability can change both achievable performance and the preferred physical plant.

Learning objectives¶

After completing this chapter, you should be able to:

1. explain and apply physical plant;

2. explain and apply sensors;

3. explain and apply state estimator;

4. explain and apply prediction horizon;

5. formulate and verify the chapter methods on an active suspension under full preview, finite-horizon MPC, reactive feedback, and noisy estimation.

Mathematical lens¶

The recurring quantities are $p$, controller and estimator $c$, information architecture $a_I$, and $\mathcal I_t$:

Running example¶

The recurring example is an active suspension under full preview, finite-horizon MPC, reactive feedback, and noisy estimation. Retaining one system prevents apparent improvements from being caused by changed physics, information, loads, or metrics.

Recommended workflow¶

1. compute OLOC limit.

2. restrict information.

3. design estimator.

4. select controller.

5. validate real time.

Chapter map¶

1. Fixed-Structure Controller Co-Design

2. PID and State-Feedback CCD

3. LQR and Robust-Controller CCD

4. State Estimation and Kalman Filtering

5. Model Predictive Control

6. Preview and Prediction Horizons

7. Sampling, Computational Delay, and Model Fidelity

8. Controller Architecture Enumeration

9. Bridging OLOC and Closed-Loop Control

10. A Staged Plant-and-Controller Development Process