Types of control systems
There are two principal types of control systems:
- Open-loop control system
- Closed-loop control system
Open-loop control system
Definition: An open-loop control system is one in which the control action is completely independent of the output. The controller acts on the input only and does not use feedback from the output to correct errors.
Such systems are also called non-feedback systems. In continuous control practice an open-loop arrangement means the output has no influence on the control action applied to the process.
Block-diagram representation: A typical block diagram shows the controller and the process in series with no feedback path from the output to the input.
Example (practical): A simple manual tank-level control where an operator adjusts a hand valve while observing the level in a transparent tube. The operator (human controller) does not measure the actual process variable electronically and there is no automatic feedback loop.
Characteristics and typical uses:
- Simple in construction and easy to understand.
- Economical to implement because no sensors, feedback elements or complex controllers are required.
- Easy to maintain due to fewer components and simpler wiring.
- Generally stable because absence of feedback avoids feedback-related oscillations.
- Useful when the output is difficult or expensive to measure or when the process is well known and disturbances are negligible.
Limitations:
- Inaccurate when there are process variations or external disturbances because the system cannot correct for errors automatically.
- Unreliable for tasks that require precise regulation or where the process parameters change with time.
- Cannot automatically correct for changes in the output; any correction must be made manually or by redesigning the controller.
Closed-loop control system
Definition: A closed-loop control system is one in which the control action depends on the output. The system measures the output, compares it with a reference, and uses the difference (error) to adjust the input so that the desired output is obtained.
A closed-loop system with automatic correction due to measured output is commonly referred to as a feedback control system or an automatic control system. By adding a feedback path to an open-loop system, it becomes a closed-loop system that can respond automatically to disturbances.
Block-diagram representation: The block diagram includes a feedback path from the output back to the summing point at the input where the error (reference minus feedback) is formed.
Example (practical): A level-control loop where the liquid level is measured by a transmitter, the transmitter signal is sent to a controller, and the controller adjusts a control valve to maintain the level at the setpoint.
Key concepts:
- Feedback: The information about the output that is returned to the input to reduce error. In regulation systems negative feedback is used to stabilise the system and reduce error.
- Error signal: The difference between the reference (setpoint) and the measured output; the controller acts on this error.
- Disturbance rejection: Ability of the system to maintain the desired output despite external disturbances; feedback improves disturbance rejection.
- Sensitivity: Feedback can reduce sensitivity to parameter variations in the plant, improving robustness.
Advantages:
- Higher accuracy since any deviation from the desired output produces corrective action through feedback.
- Improved performance in presence of nonlinearities and parameter changes because feedback corrects errors.
- Enables automation of processes, reducing the need for continuous human intervention.
- Sensitivity to plant parameter variations can be reduced, improving robustness.
- Can be designed to be less affected by certain types of measurement noise (with appropriate filtering and controller design).
Disadvantages and design challenges:
- Costlier due to additional sensors, transmitters, controllers and actuators.
- More complex to design and tune; stability analysis is required to prevent undesirable oscillations.
- Requires more maintenance because of additional components and signal conditioning electronics.
- Improper feedback design can lead to oscillatory or unstable responses; stability is a primary concern in closed-loop systems.
Comparison: open-loop versus closed-loop
- Accuracy: Closed-loop systems are generally more accurate because they correct errors; open-loop systems cannot correct errors automatically.
- Complexity and cost: Open-loop systems are simpler and cheaper; closed-loop systems are more complex and expensive.
- Maintenance: Open-loop systems require less maintenance; closed-loop systems require more due to sensors and controllers.
- Stability: Open-loop systems are free from feedback-induced instability; closed-loop systems require careful stability analysis to avoid oscillations.
- Disturbance rejection: Closed-loop systems handle disturbances much better than open-loop systems.
- Automation: Closed-loop systems are suitable for automation; open-loop is suitable only where manual or preset control suffices.
When to use which system
- Choose an open-loop system when the process is simple, well characterised, disturbances are negligible, and measurement of the output is difficult or expensive.
- Choose a closed-loop system when precision is required, external disturbances are significant, the process parameters vary, or automation and robustness are needed.
Practical notes for design (brief)
- In closed-loop design ensure negative feedback is used for regulation and perform stability analysis (Bode, Nyquist, Routh etc.) as required by the system complexity.
- Select sensors and transmitters according to required accuracy, response speed and environmental conditions.
- Consider cost versus performance trade-offs; in many industrial applications a combination of both (open sections with local closed loops) is used.
Summary: Open-loop systems offer simplicity and low cost but cannot correct for errors automatically. Closed-loop systems use feedback to improve accuracy, disturbance rejection and robustness, at the expense of greater complexity, cost and the need for stability considerations.