¶ Overview
One or more PID (proportional-integral-differential) controllers. Each controller is specified by a different index: 1, 2, 3, ...
PID controllers are used to control a system using feedback: measure an error and use this to apply a correction. A common use case is operating an actuator with an encoder, such as a robot wheel, to move a specified amount. For an actuator, you can determine the error by comparing where you want to be with where the encoder says you are. Feed the measured error into compute pid and scale the resulting correction to determine a drive signal. See Recommendations for using the PID library for more information.
¶ Summary of Blocks
For each block, there is a short description entry and a detailed block and component description. You can click on block pictures in the short description table to access the details.
| Short Description | |
|---|---|
 |
Compute the next PID correction for the specified PID loop. |
 |
Reset the specified PID loop. |
 |
Limit the range of a value, typically a drive signal. |
 |
Apply the sign of one value to another value. |
¶ Block Descriptions
¶ compute pid
Compute the next PID correction for the specified PID loop. The inputs are as follows:
index: index of the PID loop: 1, 2, 3, ...error: error to correctpCoeff: proportional coefficient (corr/error)iCoeff: integral coefficient (corr=msec/error)dCoeff: derivitive coefficient (corr/error-msec)maxIntegral: maximum absolute value of the integrated error; ignored if 0
¶ reset pid
Reset the specified PID loop. It may be useful to reset a PID before commanding a new move.
¶ constrain value
Limit the range of a value. This is provided to constrain the drive signal to useful values. The algorithm is as follows (where |value| means the absolute value of value):
- If
|value| < deadband: return 0.
This can keep the system from jittering or whining after a move. - If
|value| < minimum: returnminimumwith the sign ofvalue.
This prevents applying a correction that is too small to do anything useful.
For example the XRP wheel motors cannot reliably move when the effort is less than about 200. - If
|value| > maximum: returnmaximumwith the sign ofvalue.
This prevents demanding more than the system can handle.
For example the XRP wheel motors have a maximum "effort" of 1023.
¶ apply sign to value
Apply the sign of the first argument to the second argument:
- If the first argument is positive or 0, return the second argument unchanged.
- If the first argument is negative, return the second argument * -1.
¶ Recommendations for using the PID library
- In order to allow tuning flexibility, use a
pCoeffsignificantly larger than 1 (e.g. 100 or 1000),
then divide the resulting correction by a suitable factor to get your drive signal.
This helps compensate for the lack of floating=point values in microBlocks. - Limit the resulting drive signal to reasonable values using
pid_constrainValue. - A non-zero
iCoeffcan easily lead to oscillations or instability.
To make the system more stable you can try either or both of the following:- Use a non-zero
dCoeff(this is very common when specifying a non-zeroiCoeff). - Specify a non-zero value for
maxIntegral.
- Use a non-zero
- If you can predict the value of the drive signal, use the prediction as a "feedfoward" signal, as explained here.
Done well, this should greatly reduce the size of the PID corrections, which should make your system more responsive and stable.- Compute the predicted drive signal.
- Compute the PID correction and scale it as usual (see the first recommendation) to produce a PID drive signal correction.
- Compute drive signal = predicted drive signal + PID drive signal correction.
- Limit the resulting drive signal to a reasonable value using constrain value, as usual.