>[!warning]
>This content has not been peer reviewed.
# Control and feedback — RST foundation
Classical **control theory** studies how systems use **feedback** to regulate
their behaviour: maintaining a setpoint, rejecting disturbances, and stabilising
otherwise unstable dynamics. In RST, the **identity loop**
\(I = \Omega \cdot \mu\) and homeostatic systems are naturally expressed in
control-theoretic terms: they sense deviations, allocate budget, and act to
keep key variables within tolerances.
---
## I. Classical side
- **Feedback loop:** A controller compares a reference signal (setpoint) to the
measured output and applies an action based on the **error**.
- **Open-loop vs closed-loop gain:** In linear systems, the **loop gain**
\(L(s)\) determines stability and response; closing the loop reshapes
dynamics and can suppress disturbances.
- **Sensitivity and robustness:** The **sensitivity function**
\(S(s) = 1 / (1 + L(s))\) quantifies how strongly disturbances affect the
output; small \(|S|\) over a bandwidth means good disturbance rejection
and robust tracking \([1,2]\).
---
## II. Mapping to RRT / RST
| Control concept | RRT / RST identity | Axiom / equation |
|:---|:---|:---|
| **Setpoint / reference** | Target identity / request \(I\) the system tries to maintain. | Resource Allocation Equation. |
| **Error signal** | Deviation between current \(\Omega \cdot \mu\) and requested \(I\); drives reallocation of budget. | Identity loop; A2, A3. |
| **Loop gain \(L\)** | Effective **signal-to-noise ratio** \(\eta = \Omega/N\) achieved by the controller. | \(\eta\) in \(\mu(\eta, n)\). |
| **Sensitivity \(S = 1/(1+L)\)** | Fraction of disturbances that survive; complements the useful fraction \(\mu\). | Noise fraction; \(\mu^n + \nu^n = 1\). |
| **Stability / homeostasis** | Regime where feedback keeps key observables in a high-\(\mu\) band despite noise. | Identity loop fixed point; RST field equation. |
In this reading, a **controller** is a configuration of relations that
continually adjusts \(\Omega\) to maintain \(I\) against variations in \(N\),
trading complexity for robustness. High loop gain corresponds to high
signal-to-noise \(\eta\) and thus high fidelity \(\mu\).
---
## Links
- **Foundation index:** [[../Foundation index]]
- **RRT axioms:** [[Relational Resolution Theory (RRT)]]
- **Applications Roadmap:** [[../../Applications Roadmap]]