The closed-loop sensitivity function describes how the closed-loop error or sensitivity to certain disturbances and model variations depends on the open-loop loop gain. For a standard negative-feedback system, it is the reciprocal of one plus the loop transfer function.
Closed-Loop Sensitivity Function
The closed-loop sensitivity function describes how the closed-loop error or sensitivity to certain disturbances and model variations depends on the open-loop loop gain. For a standard negative-feedback system, it is the reciprocal of one plus the loop transfer function.
\(S(s) = 1/[1 + L(s)]\)
What this formula means
Scientific principle
Principle
Negative feedback reduces the effect of certain disturbances and parameter variations by comparing the output with the reference. The sensitivity function quantifies how strongly the closed-loop system responds to changes entering at specified locations in the feedback structure.
Derivation
1. Define the loop transfer function as L(s) = G(s)H(s).
2. For negative feedback, E(s) = R(s) − H(s)Y(s).
3. Since Y(s) = G(s)E(s), substitute to obtain E(s) = R(s) − H(s)G(s)E(s).
4. Rearrange: E(s)[1 + G(s)H(s)] = R(s).
5. Divide by R(s): E(s)/R(s) = 1/[1 + G(s)H(s)].
6. Define the sensitivity function as S(s) = 1/[1 + L(s)].
Final Result
S(s) = 1/[1 + L(s)]
Reference To Error Relation
E(s)/R(s) = S(s)
Plant Sensitivity Relation
For T(s) = L(s)/[1 + L(s)], the relative sensitivity of T to multiplicative changes in L is S(s), under the standard loop definition.
Variables and their meanings
| Symbol | Meaning | Unit |
|---|---|---|
S(s) | Closed-loop sensitivity function | Dimensionless |
L(s) | Open-loop loop transfer function G(s)H(s) | Dimensionless for a properly scaled loop |
G(s) | Forward-path transfer function | Depends on system signals |
H(s) | Feedback-path transfer function | Depends on system signals |
R(s) | Reference-input transform | Reference unit × s |
E(s) | Error-signal transform | Error unit × s |
SI units and conversions
Sensitivity Function
Dimensionless
Loop Gain
Dimensionless for a properly scaled loop
Dimensional Formula
Dimensionless
Conditions of use
The system must use the standard negative-feedback configuration.
The loop transfer function must be defined consistently as L(s) = G(s)H(s).
The system is assumed to be linear and time-invariant.
Transfer-function calculations assume zero initial conditions.
The disturbance or uncertainty location must be identified before interpreting a sensitivity response.
Limitations and assumptions
The sensitivity function does not have the same disturbance interpretation for every disturbance injection point.
A small sensitivity magnitude indicates reduced sensitivity in the corresponding defined relationship, not immunity to all disturbances.
High loop gain can improve low-frequency disturbance rejection but may reduce stability margins if poorly designed.
Unmodelled dynamics and delays can change the actual closed-loop response.
The equation alone does not establish stability.
Step-by-step worked examples
| Problem | Solution | Answer |
|---|---|---|
A negative-feedback system has loop gain L(s) = 9 at a particular frequency. Calculate its sensitivity magnitude at that frequency. |
| The sensitivity magnitude is 0.1 for this real positive loop-gain value. |
At a particular frequency, the loop transfer function is L(jω) = 4. Determine the sensitivity function at that frequency. |
| The sensitivity function is 0.2 at the specified frequency. |
Real-world applications
Feedback-controller robustness analysis
Disturbance-rejection design
Control-system model uncertainty analysis
Precision motion-control systems
Industrial process regulation
Formula Meaning
The closed-loop sensitivity function describes how the closed-loop error or sensitivity to certain disturbances and model variations depends on the open-loop loop gain. For a standard negative-feedback system, it is the reciprocal of one plus the loop transfer function.
Sources and further reading
No source has been supplied for this item yet. Treat the claim as needing editorial verification.