Tuning PID (triple mode) controller

Tuning a temperature controller involves setting proportional, integral, and derivative values ​​to get the best possible control of a particular process. If the controller does not include an auto-tuning algorithm, or if the auto-tuning algorithm does not provide adequate control for a particular application, the device must be tuned using trial and error.

The following are the standard tuning steps for the temperature controller. Other tuning steps can also be used, but similar trial and error methods are used. Please note that if the controller uses a mechanical relay (rather than a solid state relay), it should start with a longer cycle time (20 seconds).

You may need to use the following definition:

Cycle time - Also known as the duty cycle, is the total length of time it takes the controller to complete an on-off cycle. Example: For a cycle time of 20 seconds, the 10 second on time and the 10 second off time represent 50% of the power output. In the proportional band, the controller will cycle on and off. Proportional Band – The temperature range expressed in % of full scale or degrees. The proportional action of the controller occurs within this range. The wider the proportional band, the larger the area surrounding the setpoint where proportional effects occur. Sometimes expressed in terms of gain, the gain is the inverse of the proportional band. Integration, also called reset, is a function of adjusting the proportional bandwidth according to the set value to compensate for the offset (fixed deviation) from the set value, that is, it adjusts the control temperature to the set value after the system stabilizes. value. Differential, also known as speed, senses the rate at which the system temperature rises or falls and automatically adjusts the proportional band to minimize undershoot or overshoot.

The PID (triple mode) controller can have excellent control stability if it is correctly set and used. By carefully following these instructions, the operator can achieve the fastest response time and minimal overshoot. The information for tuning this three-mode controller may be different from other controller tuning steps. For the main output, the auto-tuning function is usually used to eliminate the need to use this manual tuning step, but the auto-tuning value can be adjusted when needed.

After the controller is installed and wired: Apply power to the controller. Disable the controller output if possible. For time proportional main output, set the cycle time. Enter the following value: CYCLE TIME 1

5 seconds (displayed only when the output is time proportional output. For systems with very fast response times, a shorter cycle time may be required.)

Then select the following parameters:

PR BAND 1 ______ 5% (PB)
RESET 1 ________ 0 R/M (disconnect reset function)
RESET 2 ________ 0 R/M
RATE 1 _________ 0 MIN (disconnect reset function)
RATE 2 _________ 0 MIN

Note

In the device with double three-mode output, the main setting parameters and auxiliary setting parameters are independently set and must be set separately. The steps used in this section are for the main "heating" output. A similar procedure can be used for the main "cooling" output or the secondary "cooling" output.

A. Set output of heating control Enable output and start the process. The process should be run at the set point and the desired heat input will be used to stabilize the temperature.

In the case of a rate and reset disconnect, the temperature will stabilize and there will be a steady-state deviation or a fixed deviation between the setpoint and the actual temperature. By observing the measured values ​​on the display, pay close attention to whether there is a regular cycle or oscillation at this temperature. (The oscillation can last up to 30 minutes.)

Figure 1: Temperature oscillation

If the temperature does not oscillate regularly, divide PB by 2 (see Figure 1). Let the process stabilize, and then check for temperature oscillations. If there is still no oscillation, divide PB by two. Repeat this until you get a loop or oscillation. Go to step 5.
If the oscillation is observed immediately, multiply PB by two. Observe the resulting temperature for a few minutes. If oscillation continues, multiply PB by a factor of 2 until the oscillation stops. At this point, PB is very close to its critical setting. Carefully increase or decrease the PB setting until there is just a cycle or oscillation in the temperature log.
If process temperature does not oscillate even at the minimum PB setting of 1%, skip steps 6 to 11 below and go to section B. At the "critical" BP setting that has been reached, read the steady-state deviation, or fixed deviation, between the setpoint and the actual temperature. (Because there is a little temperature cycling, use the average temperature.) Measure the oscillation time between adjacent peaks or troughs in units of minutes (see Figure 2). This type of measurement is easiest to perform using a chart recorder, but readings can be taken every minute to keep track of time. At this time, increase the PB setting until the temperature deviation (or fixed deviation) increases by 65%.

The desired final temperature deviation can be calculated by multiplying the initial temperature deviation obtained under the “Critical” BP setting by 1.65 or using the convenient Line Graph I (see Figure 4). Tried PB control settings several times with trial and error until the desired final temperature deviation was obtained.

At this point you have completed all the measurements required to get the best performance from the controller. Just make two more adjustments - Rate and Reset. Using the oscillation time measured in step 7, calculate the reset value in the following way, in units of repetitions per minute.

Enter this value for RESET 1. Then use the oscillation time measured in step 7 to calculate the value of the rate in the following way, divided into units.

Enter this value into RATE 1. If overshoot occurs, it can be eliminated by reducing the reset time. When you change the reset value, you should also make corresponding changes to the rate adjustment so that the rate value is equal to:

That is: If reset = 2 R/M, rate = 0.08 minutes If you want to get the correct balance between the "response time" and "set time" of the system disturbance, you may need to make multiple setpoint changes and follow The resulting reset and rate control time adjustments. Rapid response is often accompanied by a large overshoot, and the time required for the process to "stabilize" is also shorter. In contrast, if the response is slow, the process tends to slowly slide to the final value, with little or no overshoot. It is up to the system's requirements to decide what action to take. When a satisfactory timing is achieved, the cycle time should be increased to save the life of the contactor (applicable to a time-proportional output device (TPRI)). If the measured value does not oscillate due to the load cycle, the cycle time should be increased as much as possible. Go to section C. B. Setting procedure when no oscillation is observed At the minimum PB setting, the steady-state deviation, or fixed deviation, between the setpoint and the actual temperature is measured. Increase the PB setting until the temperature deviation (fixed deviation) increases by 65%. Line graph I (see Figure 4) provides an easy way to calculate the desired final temperature deviation. Set RESET 1 to a higher value (10 R/M). Set RATE 1 to a corresponding value (0.02 minutes). At this time, the measured value should be stable at the set temperature due to the reset action. Since we cannot determine the critical oscillation time, trial and error must be used to determine the optimal settings for reset and rate adjustment. After the temperature stabilizes at the set point, the set temperature setting is increased by 10 degrees. Observe the overshoot that accompanies the actual temperature rise. Then return the setting of the set temperature to its initial value, and observe the overshoot that accompanies the actual temperature rise. Overshooting indicates that the reset and/or rate values ​​are set too high. An over-damped response (without overshoot) indicates that the reset and/or rate values ​​are set too low. See Figure 7. When you need to improve performance, change one setting parameter at a time and observe the effect of this parameter on performance when the setting value changes. Increment the parameters until the performance is optimized. When a satisfactory timing is achieved, the cycle time should be increased to save the life of the contactor (applicable to a time-proportional output device (TPRI)). Increase the cycle time as much as possible without causing the measured value to oscillate due to load cycling.

Figure 7: Setting the reset and/or rate C. Tuning the main output of the cooling control

Use the same procedure as heating. The process should be run at a set point, requiring cooling control before the temperature stabilizes.

D. Simplified tuning steps for the PID controller

The following steps are graphical methods for analyzing the response curve of a process to a step input. It is much easier to read process variables (PV) using the long map recorder.

From the cold start (PV at room temperature), the process is powered at maximum power when the controller is not in the loop (ie, open loop). Record this start time. After some delay (let the heat reach the sensor), the PV will begin to rise. After a delay, PV will reach the maximum rate of change (slope). Record the time when this maximum slope occurs and the PV at that time. Record the maximum slope in degrees/minutes. Turn off the system power. A line drawn from the maximum slope point back to the ambient temperature axis gives the total system delay Td (see Figure 8). You can also use the following formula to get the delay: Td = time when the maximum slope is reached - (PV at maximum slope - ambient temperature) / maximum slope s

Apply the following formula to get PID parameters:

Ratio Range = Td x Maximum Slope x 100/Range = % of Range
Reset = 0.4 / Td = Minor / Min Rate = 0.4 x Td = Minute Restart the system, bring the process to the setpoint while the controller is in the loop, and observe the response. If the response is too large or oscillates, change the PID parameters in the following direction (slightly change, change one parameter at a time, and observe the process response): Widen the proportional band, lower the reset value, and increase the rate value.

Example: The chart record in Figure 8 was obtained when the furnace was powered at maximum power. The chart scale is 10 F/cm and 5 points/cm. The controller range is 100 ~ 600 F, or 500 F range.

Example: The chart record in Figure 8 was obtained when the furnace was powered at maximum power. The chart scale is 10°F/cm and 5 cents/cm. The controller range is 100 ~ 600 °F, or 500 °F range.

Maximum slope = 18°F/5min
= 3.6°F/minute delay = Td = approximately 7 points proportional band = 7 minutes x 3.6°F/minute x 100/500°F = 5%.
Reset = 0.4 / 7 points = 0.06 beats/minute rate = 0.4 x 7 points = 2.8 points

Figure 8: System delay