PID Controller

The PID Controller (Proportional-Integral-Derivative Controller) is an analog control node used to regulate a process and maintain a measured value as close as possible to a desired target value.

The controller continuously compares the Measured value with the Desired value and calculates an analog Output according to three control components:

• Proportional (Kp) — reacts to the current error.
• Integral (Ki) — reacts to the accumulated error over time.
• Derivative (Kd) — reacts to how quickly the error is changing.

The basic control error is:

Error = Desired value - Measured value

The controller output can be conceptually represented as:

Output = Proportional action + Integral action + Derivative action

The controller continuously compares the two input values:

Error = Desired value - Measured value

For example:

• Desired value = 22 °C
• Measured value = 20 °C

The resulting error is:

22 - 20 = +2 °C

The PID controller uses this error, together with the configured coefficients, to calculate the required output.

The proportional component reacts directly to the current error.

Conceptually:

P = Kp × Error

A larger error therefore produces a stronger proportional correction.

If the measured value approaches the desired value, the proportional contribution progressively decreases.

Increasing Kp generally makes the controller react more strongly to a difference between the setpoint and the measured value.

The integral component considers how long the error has been present.

Even when the instantaneous error is small, if it remains present for a long period the integral contribution progressively increases.

This allows the controller to compensate for a persistent residual error that cannot be completely removed by the proportional component alone.

Increasing Ki increases the correction applied to errors accumulated over time.

The derivative component considers how quickly the error is changing.

It can therefore react not only to the current difference from the setpoint, but also to the direction and speed at which the process is moving.

For example, if a temperature is rapidly approaching the setpoint, the derivative component can reduce the control action before the setpoint is exceeded.

Increasing Kd increases the influence of the rate of change of the error.

The derivative contribution can therefore help reduce overshoot and stabilize the regulation.

The Sampling Time defines how frequently the PID calculation is updated.

For example, with:

Sampling Time = 5 seconds

the controller updates its calculation every 5 seconds.

The appropriate sampling time depends on the controlled process.

A rapidly changing process may require a shorter sampling interval, while slow systems such as room temperature regulation generally do not require extremely fast calculations.

The Threshold defines a tolerance around the desired operating condition.

Small variations around the setpoint can therefore be considered within an acceptable error range, reducing unnecessary reactions to minor measurement fluctuations.

This is especially useful when the measured value is not perfectly stable.

The calculated PID output is constrained between the configured Minimum and Maximum values.

For example:

the output will always remain within:

0 ≤ Output ≤ 100

Even if the PID calculation produces a value outside this range, the resulting output is limited by the configured boundaries.

Consider a heating system where the PID output controls the heating power.

Assume:

The initial error is:

22 - 20 = +2 °C

The controller therefore generates an output according to the configured Kp, Ki and Kd coefficients.

As the room temperature approaches the desired value, the behaviour can be represented conceptually as follows:

Both the PID Controller and the Comparator / 2 Points Regulator can be used for regulation, but their behaviour is different.

The behaviour of the PID Controller can be checked directly using the Logic Module simulation tools.

During Offline Simulation, the input values can be modified manually while observing the values passing through the node connectors.

A typical test can be performed as follows:

• Connect a variable or simulated value to Measured value.
• Connect the required setpoint to Desired value.
• Configure Minimum, Maximum, Sampling Time, Threshold, Kp, Ki and Kd.
• Start the simulation.
• Modify the measured value.
• Observe how the Output changes over time.
• Gradually move the measured value toward and beyond the setpoint.
• Verify that the output remains between the configured Minimum and Maximum limits.

Online Simulation can also be used to observe the node while the logic scheme is running directly on the Thinknx server.

• Make sure that Measured value and Desired value use the same physical unit and scale.

• Configure Minimum and Maximum according to the valid range of the controlled actuator or destination value.

• Choose a Sampling Time suitable for the speed of the controlled process.

• A very short sampling time is usually unnecessary for slow processes such as room temperature regulation.

• Use the Threshold to avoid excessive reactions to insignificant fluctuations around the desired value.

• Increasing Kp makes the controller react more strongly to the current error.

• Increasing Ki increases the correction of errors that persist over time.

• Increasing Kd increases the response to rapid variations of the error.

• PID coefficients should be adjusted progressively while observing the behaviour of the controlled system.

• A common commissioning approach is to begin with the proportional component and then progressively introduce the integral and derivative contributions when required.

• Always verify the final configuration on the real installation, especially when controlling processes with significant inertia or delay.

• Logic Module
• Logic Module - Simulation management
• Comparator / 2 Points Regulator