Feedback Systems¶
Åström, K. J., & Murray, R. M. (2008). Feedback Systems: An Introduction for Scientists and Engineers. Princeton University Press.
Cited by¶
34 citations across 34 artifacts.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Circuit Breaker
This sourceCanonical feedback-control text: develops continuous regulation toward a setpoint (PID) versus discrete switched action, and treats relay feedback with hysteresis as the standard remedy for chattering.
- Controllability
- T2 — Controllability versus stability / unintended consequences
This sourceIntroductory control text covering feedback, stability, time delays, and nonlinear effects across physical/biological/economic systems.
- T2 — Controllability versus stability / unintended consequences
- Discrepancy-Driven Correction
- In engineering control theory the setpoint is compared to the plant output, the gap is the error signal, the controller's gain function maps error to actuation, and the loop is the closed-loop control system.
This sourceSetpoint–error–actuation loop, PID control, gain-and-stability and the integral term for steady-state error.
- In engineering control theory the setpoint is compared to the plant output, the gap is the error signal, the controller's gain function maps error to actuation, and the loop is the closed-loop control system.
- Homeostasis
- homeostasis sits at the intersection of three companion abstractions (#390 observability, #391 controllability, #387 requisite variety); (3) homeostasis delivers a resilience strategy against perturbation — as long as disturbances fit within the mechanism's envelope (correction authority, response speed, variety coverage), the system maintains essential variables within tolerance; outside the envelope, homeostasis fails catastrophically (hypothermia, runaway inflation, cascading cellular stress) —
This sourceCanonical feedback-control text: develops continuous regulation toward a setpoint (PID) versus discrete switched action, and treats relay feedback with hysteresis as the standard remedy for chattering.
- homeostasis sits at the intersection of three companion abstractions (#390 observability, #391 controllability, #387 requisite variety); (3) homeostasis delivers a resilience strategy against perturbation — as long as disturbances fit within the mechanism's envelope (correction authority, response speed, variety coverage), the system maintains essential variables within tolerance; outside the envelope, homeostasis fails catastrophically (hypothermia, runaway inflation, cascading cellular stress) —
- Measurement and Disturbance
- Control systems: Monitoring a system's state through sensors and actuators necessarily couples the system to the controller; the control action itself disturbs the plant, requiring careful design to minimize overshoot and oscillation, as Åström and Murray (2008) develop in their foundational treatment of feedback systems.
This sourceCanonical feedback-control text: develops continuous regulation toward a setpoint (PID) versus discrete switched action, and treats relay feedback with hysteresis as the standard remedy for chattering.
- Control systems: Monitoring a system's state through sensors and actuators necessarily couples the system to the controller; the control action itself disturbs the plant, requiring careful design to minimize overshoot and oscillation, as Åström and Murray (2008) develop in their foundational treatment of feedback systems.
- Monitoring
- The feedback-loop structure builds on monitoring but is not identical to it.
This sourceCanonical feedback-control text: develops continuous regulation toward a setpoint (PID) versus discrete switched action, and treats relay feedback with hysteresis as the standard remedy for chattering.
- The feedback-loop structure builds on monitoring but is not identical to it.
- Non-Stationary Objective
- In control theory, the explicit distinction between regulator problems (constant set-point) and tracking problems (time-varying set-point) turns on the loop time-constant relative to the reference's rate of change.
This sourceStandard control-theory text covering the regulator (constant set-point) versus tracking (time-varying reference) distinction and how loop bandwidth relative to the reference's rate of change governs tracking error.
- In control theory, the explicit distinction between regulator problems (constant set-point) and tracking problems (time-varying set-point) turns on the loop time-constant relative to the reference's rate of change.
- Stability
- Dynamical systems and control engineering (canonical): Lyapunov stability of equilibria; stable versus unstable fixed points; closed-loop stability via gain and phase margins; the PID controller exists to induce stability where it is absent.
This sourceStandard reference on closed-loop stability, gain and phase margins, and the PID controller's stabilizing role.
- Dynamical systems and control engineering (canonical): Lyapunov stability of equilibria; stable versus unstable fixed points; closed-loop stability via gain and phase margins; the PID controller exists to induce stability where it is absent.
- Withdrawal Rebound
- In engineering control it is integrator wind-up: an integral controller wound up against a saturated actuator overshoots sharply when the disturbance is removed.
This sourceDefines the PI controller u(t)=Kp e(t)+Ki∫e dτ, integrator wind-up against a saturated actuator, and the resulting overshoot when the disturbance is removed.
- In engineering control it is integrator wind-up: an integral controller wound up against a saturated actuator overshoots sharply when the disturbance is removed.
Domain-specific¶
Mechanisms¶
- Actively Shaped Field Control
- The guarding discipline is to respect the loop's stability margin
This sourceRelates robust feedback design to adequate stability margins and to limiting loop bandwidth where sensor noise and uncertainty dominate.
- The guarding discipline is to respect the loop's stability margin
- Actuator Saturation Alarm
- Its strength is diagnostic honesty: it is the cheapest way to break the "track harder" misdiagnosis, and pairing it with a windup freeze also protects the loop from the classic post-saturation lurch known as integral windup.
This sourceExplains integrator windup during actuator saturation and the need for anti-windup protection that limits or suspends integral accumulation.
- Its strength is diagnostic honesty: it is the cheapest way to break the "track harder" misdiagnosis, and pairing it with a windup freeze also protects the loop from the classic post-saturation lurch known as integral windup.
- Adaptive Reconstraint Protocol
- It misleads through oscillation — a trigger with too little hysteresis makes the constraint chatter on and off, and the flapping is often worse than either steady state.
This sourceExplains that feedback can generate dynamic instability and sustained oscillation.
- It misleads through oscillation — a trigger with too little hysteresis makes the constraint chatter on and off, and the flapping is often worse than either steady state.
- Adaptive Sampling-Rate Controller
- The guarding discipline is to damp the loop with hysteresis
This sourceExplains hysteresis as a control nonlinearity and presents feedback-design methods for damping oscillation and operating within bounded control regimes.
- The guarding discipline is to damp the loop with hysteresis
- Automatic Control Trigger
- Its characteristic pathology is the limit cycle: with too narrow a band or too aggressive an authority, the loop settles into a self-sustained oscillation, hunting on and off around the setpoint even in the absence of any real disturbance.
This sourceExplains that on-off feedback can overreact because a small error commands the full actuation range, producing oscillation around the reference, and defines the corresponding steady oscillatory behavior as a limit cycle.
- Its characteristic pathology is the limit cycle: with too narrow a band or too aggressive an authority, the loop settles into a self-sustained oscillation, hunting on and off around the setpoint even in the absence of any real disturbance.
- Control Knob
- Turn a sensitive knob in response to short-term noise and the system hunts — the operator chases their own lagged effect up and down.
This sourceExplains that improperly used feedback can produce instability and couple measurement noise into the system dynamics.
- Turn a sensitive knob in response to short-term noise and the system hunts — the operator chases their own lagged effect up and down.
- Control Loop Damping
- Its strength is precision: when the loop is instrumentable and its dynamics are roughly known and stationary, tuning gain, anticipation, and filtering can turn a hunting system into one that settles in a single approach, and the step response gives an objective read on whether you have hit critical damping.
This sourceShows how anticipatory derivative action and filtering can damp oscillation, and how a second-order step response changes at critical damping.
- Its strength is precision: when the loop is instrumentable and its dynamics are roughly known and stationary, tuning gain, anticipation, and filtering can turn a hunting system into one that settles in a single approach, and the step response gives an objective read on whether you have hit critical damping.
- Control Loop Tuning
- Crank the gain to chase every wiggle and you spend your stability margin — the buffer of gain and phase the loop needs before feedback turns to oscillation — until the loop rings or goes unstable outright.
This sourceDefines gain, phase, and stability margins as robustness buffers and shows that a low stability margin can yield highly oscillatory closed-loop behavior.
- Crank the gain to chase every wiggle and you spend your stability margin — the buffer of gain and phase the loop needs before feedback turns to oscillation — until the loop rings or goes unstable outright.
- De-Escalation Pathway
- Its failure mode is timing: de-escalate too early and the problem re-erupts with no one senior watching; leave the step-down threshold equal to the escalation trigger and authority flaps up and down — which is why a deliberate gap, hysteresis, between escalate and de-escalate matters
This sourceShows that hysteresis uses separated switching thresholds rather than a single shared boundary.
- Its failure mode is timing: de-escalate too early and the problem re-erupts with no one senior watching; leave the step-down threshold equal to the escalation trigger and authority flaps up and down — which is why a deliberate gap, hysteresis, between escalate and de-escalate matters
- Dispatch Rescheduling System
- The discipline that keeps it honest is hysteresis — requiring a change to clear a margin before it reverses course — plus a firm lock window and a standing human override, so the board moves when it must and holds when it should.
This sourceShows hysteresis modifying on–off control so that a small change around a switching boundary does not immediately reverse the controller’s state.
- The discipline that keeps it honest is hysteresis — requiring a change to clear a margin before it reverses course — plus a firm lock window and a standing human override, so the board moves when it must and holds when it should.
- Feedback-Gain Reduction
- Reduce the gain too far and you don't just tame the runaway, you deaden the legitimate responsiveness that shared the same loop; reduce it too little and the pathway still creeps past the stability line, only slower.
This sourceExplains that excessive feedback gain can drive oscillatory behavior in a closed-loop system.
- Reduce the gain too far and you don't just tame the runaway, you deaden the legitimate responsiveness that shared the same loop; reduce it too little and the pathway still creeps past the stability line, only slower.
- Feedforward Brake Circuit or Protocol
- This is textbook feedforward control,
This sourceDefines feedforward control as acting from reference or measured-disturbance information without waiting for output error.
- This is textbook feedforward control,
- Gain-Collapse Test
- The behavior it detects is control-systems saturation — an element run out of its range produces no further output for more input, so its incremental gain goes to zero.
This sourceDefines actuator saturation as an input limit beyond which greater requested input produces no additional output, reducing incremental gain to zero in the saturated region.
- The behavior it detects is control-systems saturation — an element run out of its range produces no further output for more input, so its incremental gain goes to zero.
- Homeostatic Adjustment Protocol
- The guarding discipline is to build in a deadband and hysteresis
This sourceShows that feedback delay constrains stabilizing controller gain and that damping and gain shape closed-loop response.
- The guarding discipline is to build in a deadband and hysteresis
- Hysteresis or Deadband Window
- Its strength is that it is the cheapest cure for chatter: a loop that oscillates only because it keeps chasing sub-threshold reversals is stabilised outright, and the two-edged behaviour that makes this work is textbook hysteresis
This sourceShows how hysteresis keeps noisy threshold crossings from causing jitter by making switching depend on prior input.
- Its strength is that it is the cheapest cure for chatter: a loop that oscillates only because it keeps chasing sub-threshold reversals is stabilised outright, and the two-edged behaviour that makes this work is textbook hysteresis
- Motor-Command Echo Routing
- This is the transport-delay problem familiar from control theory, where an unaccounted lag between a command and its observed effect destabilizes the loop.
This sourceExplains transport delay as command-to-response lag whose phase effect can destabilize a feedback loop.
- This is the transport-delay problem familiar from control theory, where an unaccounted lag between a command and its observed effect destabilizes the loop.
- Net-Flow Lever Adjustment
- A policy that reacts only when the level hits a threshold, and then moves a big lever, behaves like bang-bang control — it slams fully on
This sourceShows that thresholded on–off control switches between maximum and minimum action and often oscillates because small errors drive the actuator through its full range.
- A policy that reacts only when the level hits a threshold, and then moves a big lever, behaves like bang-bang control — it slams fully on
- Phase Margin or Dead-Time Test
- The margin is only as good as the characterization: a linear test on a nonlinear loop
This sourceExplains that a stability margin depends on the system characterization and that linearization of a nonlinear system is local to a chosen operating point.
- The margin is only as good as the characterization: a linear test on a nonlinear loop
- Proportional Trim Adjustment
- It leaves a steady-state offset — a persistent small error it never fully removes, because the trim goes to zero exactly when the error does, so it settles slightly off target unless helped by additional control terms.
This sourcePure proportional control can leave steady-state error when nonzero input is needed at the target; integral action can eliminate the offset.
- It leaves a steady-state offset — a persistent small error it never fully removes, because the trim goes to zero exactly when the error does, so it settles slightly off target unless helped by additional control terms.
- Queue-Based Feedback Controller
- Its signature failure mode is oscillation and thrash: gains set too hot, a deadband too narrow, or a sensing-to-action lag too long, and the loop scales up and down in cycles that destabilize schedules, systems, and people — a textbook control-theory pathology that hysteresis
This sourceExplains that simple on-off feedback often oscillates because small errors trigger full-range corrective action.
- Its signature failure mode is oscillation and thrash: gains set too hot, a deadband too narrow, or a sensing-to-action lag too long, and the loop scales up and down in cycles that destabilize schedules, systems, and people — a textbook control-theory pathology that hysteresis
- Rate-of-Change Limit
- Its strength is cheap, robust, model-free protection: because it only bounds rate, it works on any actuator without knowing the plant, and it directly removes the case where a violent correction is itself the oscillation's engine.
This sourceA slew-rate limit bounds how quickly an actuator command can change.
- Its strength is cheap, robust, model-free protection: because it only bounds rate, it works on any actuator without knowing the plant, and it directly removes the case where a violent correction is itself the oscillation's engine.
- Thermostat Control
- - On/off (bang-bang) actuation. The output is binary — heat or no heat — rather than a proportioned dose.
This sourceDescribes on-off control as switching the actuator between discrete states rather than continuously proportioning its output.
- - On/off (bang-bang) actuation. The output is binary — heat or no heat — rather than a proportioned dose.
- Threshold Policy Rule
- A deadband between the release-on and release-off lines keeps the gates from chattering open and shut around a single level.
This sourceExplains hysteresis as distinct on and off thresholds that prevent rapid switching around one boundary.
- A deadband between the release-on and release-off lines keeps the gates from chattering open and shut around a single level.
- Threshold-Triggered Refresh Controller
- A hysteresis deadband between the "start" and "stop" setpoints keeps the valve from chattering on and off around a single level
This sourceUses distinct start and stop thresholds in a hysteresis band to prevent chattering near a single setpoint.
- A hysteresis deadband between the "start" and "stop" setpoints keeps the valve from chattering on and off around a single level
Verification¶
Does it exist? Confirmed. This work's DOI resolves to a registered record, which fixes its identity. That is all it fixes.
Does it back the claim? Recorded for 1 of 34 citations of this work.
Support is checked per citation rather than per work — the same source can be cited soundly in one article and wrongly in another. Per-citation recording began recently, so a citation with no recorded check is a gap in the record rather than evidence it went unchecked.
See how references were verified.
Links previously used in the corpus¶
Before the registry existed this work was also linked 15 other ways.
- https://authors.library.caltech.edu/records/yzs24-xsx88 ×4
- https://doi.org/10.2307/j.ctvcm4gdk ×4
- https://fbsbook.org/ ×4
- https://press.princeton.edu/books/hardcover/9780691135762/feedback-systems ×3
- https://press.princeton.edu/books/hardcover/9780691193984/feedback-systems ×2
- https://www.cds.caltech.edu/~murray/FBS/First_Edition.html ×2
- https://www.cds.caltech.edu/~murray/books/AM08/pdf/am08-complete_02Dec08.pdf ×2
- https://www.cds.caltech.edu/~murray/books/AM08/pdf/am08-complete_22Feb09.pdf ×2
- https://assets.press.princeton.edu/catalogs/math10.pdf ×1
- https://assets.press.princeton.edu/catalogs/phys09.pdf ×1
- https://doi.org/10.1515/9781400828739 ×1
- https://obnb.uk/p14524555-feedback-systems-an-introduction-for-scientists-and-engineers ×1
- https://www.cds.caltech.edu/~murray/FBS/Version_2.9d.html ×1
- https://www.cds.caltech.edu/~murray/amwiki/First_Edition.html ×1
- https://www.cds.caltech.edu/~murray/amwiki/index.php/Main_Page ×1
Registry ID ref:4ff016f919a2 · see in the full table