PID control is a type of feedback control that adjusts a operation variable based on the difference between a set value and a measured value.

Taking temperature control as an example, if the set temperature is 200°C but the current temperature is 180°C, then operation is needed to raise the temperature. After that, as the temperature approaches the set value, such as 190°C and 195°C, operation amount needs to be changed to prevent the temperature from rising too high.

In PID control, the state of the control object is continuously monitored, and operation variables are determined to bring it closer to the target value.

Relationship between SV, PV, deviation, and MV

If SV is 200°C and PV is 180°C, the current temperature is 20°C lower than the target temperature. In PID control, this deviation is used to determine MV.

However, we don't just look at the current deviation. We also use how long the deviation has persisted and how it is changing. This difference is what leads to the respective roles of P, I, and D.

There are other methods besides PID control to bring the temperature closer to a target value. To understand why PID control is necessary, it's helpful to start with simpler control.

With ON/OFF control, the temperature tends to fluctuate easily.

The simplest method is to turn the heater on or off. The idea is to turn the heater on if the temperature is below the set temperature, and off if it's above it.

However, the effect of heating does not disappear the moment the heater is turned off. Because heat remains in the heater and the object being heated, the temperature may rise even after it has been turned off.

Conversely, even if the temperature drops and you turn on the heater, the temperature may not rise immediately. Therefore, the temperature may fluctuate around the set value.

In P control, the operation variable is changed according to the deviation.

In P control, operation variable is changed not only based on whether it is ON or OFF, but also according to the deviation from the target value.

The idea is to make larger operation if the value is far from the set value, and smaller operation if it is closer to the set value.

On the other hand, with P operation alone, a difference may remain between the set value and the measured value even after control has stabilized. This persistent difference is called "offset".

control is achieved by combining P, I, and D.

Therefore, we combine P action with I action and D action. I action works to reduce the continuous deviation, while D action takes into account the change in deviation.

By combining the functions of P.I.D., it is possible control not only approach the target value but also consider stability around the set value.

P stands for Proportional. A P action is an action that changes operation variable according to the current deviation.

If the difference between the set value and the measured value is large, increase operation amount; if the difference is small, decrease the operation amount.

What is a proportional band?

In temperature controllers, the term “proportional band” is sometimes used in the context of P-action settings. A proportional band refers to operation the range within which a quantity changes proportionally parameter.

When the proportional band becomes smaller, the change in operation variable for the same deviation becomes larger. Conversely, when the proportional band becomes larger, the change in operation variable becomes more gradual.

I stands for Integral. The I action changes the operation variable by considering how much the deviation has accumulated over time.

For example, suppose the set temperature is 200°C but it's stable at 198°C. The current deviation is only 2°C. However, if this 2°C deviation persists for a long time, a small deviation will continue to remain.

In I-action, operation variable is changed to account for the accumulation of such deviations. Therefore, it is used to reduce offset that may remain if only P-action is used.

What is integral time?

A key parameter related to I-action is "integration time." The integration time determines how quickly I-action affects a continuing deviation.

Increasing the intensity of the action isn't always a good thing. If the effect on control object is too strong, it can lead to large adjustments to operation input, potentially causing control instability.

D stands for Derivative. D-action changes the operation variable while considering how the deviation is changing.

For example, let's say the set temperature is 200°C and the current temperature is 190°C. Looking only at the current temperature, it is still 10°C lower, so heating is necessary.

However, if the temperature is rising very rapidly, continuing large operation may cause it to exceed the set temperature. Operation D adjusts operation amount to take such changes into consideration.

Therefore, it can sometimes work to suppress rapid changes.

Differential time is

A key parameter related to D-action is "derivative time." The effect of D-action on changes varies depending on the derivative time.

In actual setup, you need to consider characteristic of control object and how they combine with P and I.

PID uses different information.

To put it simply, the difference is that P shows the current deviation, I shows the accumulation of deviations, and D shows the change in deviation.

In PID control, the results of each action are combined to determine the final operation value. A key feature of PID control is that it combines three actions with different roles.

In PID control, the functions of P, I, and D are set according to the control object. In temperature controllers, proportional band, integral time, and derivative time are used as PID constant.

Even with equipment designed control the same 200°C, the way the temperature changes differs depending on whether it's a small heating tank, a large heating furnace, or equipment that heats air or liquid.

PID constant vary depending on characteristic of control system.

The appropriate PID constant depend on factors such as the heat capacity of control system, the rate at which the temperature changes, dead time, the capacity of the heater, the location of the sensor, load fluctuations, and disturbance.

Therefore, simply setting the PID constant that control well in one piece of equipment to another piece of equipment does not guarantee the same results.

Changing the PID constant also changes the process by which the temperature approaches the target value. By observing the change in temperature over time, you can evaluate the current control state.

Standing Up

This is the process of the PV approaching the vicinity of the SV after control has started.

over

This is a phenomenon where page views (PV) temporarily exceed site views (SV).

hunting

This describes a state where the PV (Perceptual Frequency) fluctuates periodically around the SV (Superficial Frequency).

Settling time

This is the time from when control is initiated until the PV settles into a stable state near the SV.

offset

This is the steady-state difference that remains between SV and PV even after control has stabilized.

By observing these control responses, you can clarify what you want to improve. From here on, the focus shifts from "what PID is" to the practical aspects of "how to adjust PID."

What is PID tuning? The basics of tuning from the perspective of temperature control waveforms.

PID constant can be determined not only by manually adjusting them while monitoring control response, but also automatically using functions such as controllers. This function is called auto tuning (AT).

In auto tuning, the PID constant are determined using the response of control system.

What is auto tuning? The mechanism and precautions for automatically adjusting PID constant.

When the temperature is unstable, it's easy to think that changing the PID constant will improve the situation. However, PID control is only one part of control loop.

Is the sensor measuring correctly?

If the type of temperature sensor, installation location, and insertion depth are not appropriate, the temperature of the object to be control may not be captured correctly.

Can the heaters and operation devices perform the necessary operation?

If the heater capacity is insufficient for the required heat quantity, adjusting the PID constant will not allow the temperature to rise at a sufficient rate. Furthermore, the operation equipment that reflects the MV (Motion Volume) calculated by the controller into the actual heater power also affects control result.

Is the heat capacity or dead time too large?

If the heat capacity of the equipment or object is large, it will take time for the temperature to change after operation. If the heater and sensor are far apart, it will also take time for the sensor to detect the effect of operation.

Are there significant disturbance or load fluctuations?

Opening and closing the door, feeding in material, changes in flow rate, and changes in ambient temperature all affect the PV. control When the target conditions themselves are changing significantly, it becomes difficult to consistently obtain the same response using only a fixed PID constant.

Depending on characteristic of the equipment, it may be difficult to obtain the desired response with a single PID control alone.

Cascade control

Cascade control combines multiple control loops. It is sometimes used when the heater and the location to be control are far apart.

What is cascade control? It's a mechanism that stabilizes temperature using two control loops.

Feedforward control

This approach involves detecting disturbance such as material input and flow rate changes, and operation the system before their effects appear in control results. It is not mutually exclusive with PID control and can be combined depending on the objective.

Z control

At CHINO, we have development "Z-control" as our unique approach to temperature control. Z-control is a temperature control technology that originated from research on the heat generation and temperature regulation of skunk cabbage.

What is Z-control? Chino's unique temperature control technology, born from research on skunk cabbage.