Auto-Tune Sequence - Siemens SIMATIC S7-200 System Manual

S7-200 Programmable Controller System Manual

Auto-Tune Sequence

The auto-tuning sequence begins after the hysteresis and deviation values have been
determined. The tuning process begins when the initial output step is applied to the loop output.
This change in output value should cause a corresponding change in the value of the process
variable. When the output change drives the PV away from setpoint far enough to exceed the
hysteresis boundary a zero-crossing event is detected by the auto-tuner. Upon each zero crossing
event the auto-tuner drives the output in the opposite direction.
The tuner continues to sample the PV and waits for the next zero crossing event. A total of twelve
zero-crossings are required to complete the sequence. The magnitude of the observed
peak-to-peak PV values (peak error) and the rate at which zero-crossings occur are directly
related to the dynamics of the process.
Early in the auto-tuning process, the output step value is proportionally adjusted once to induce
subsequent peak-to-peak swings of the PV to more closely match the desired deviation amount.
Once the adjustment is made, the new output step amount is written into the Actual Step Size field
(ASTEP) of the loop table.
The auto-tuning sequence will be terminated with an error, if the time between zero crossings
exceeds the zero crossing watchdog interval time. The default value for the zero crossing
watchdog interval time is two hours.
Figure 15-1 shows the output and
process variable behaviors during an
auto-tuning sequence on a direct acting
loop. The PID Tuning Control Panel was
used to initiate and monitor the tuning
sequence.
Notice how the auto-tuner switches the
output to cause the process (as
evidenced by the PV value) to undergo
small oscillations. The frequency and the
amplitude of the PV oscillations are
indicative of the process gain and natural
frequency.
Based upon the information collected about the frequency and gain of the process during the
auto-tune process, the ultimate gain and ultimate frequency values are calculated. From these
values the suggested values for gain (loop gain), reset (integral time) and rate (derivative time) are
calculated.
Tip
Your loop type determines which tuning values are calculated by the auto-tuner. For example,
for a PI loop, the auto-tuner will calculate gain and integral time values, but the suggested
derivative time will be 0.0 (no derivative action).
Once the auto-tune sequence has completed, the output of the loop is returned to its initial value.
The next time the loop is executed, the normal PID calculation will be performed.
386
Figure 15-1 Auto-Tuning Sequence on a Direct Acting Loop