Why the controller must be tuned?
We should tune a controller so that it can match the characteristics of the control equipment to the process. So because of the tuning, it would respond quickly to the errors and the system would be stable too. The controller tuning is done to control the way in which the final control element would respond to the change in error. So by varying the controller gain the controller input will change and because of this, the output of the controller would change too. So this would do the adjustment of the final control element to cancel the error but it won’t cause any system instability. So by doing the controller tuning the feedback controller parameters can be varied to get proper output.
- What is the importance of a controller in process control?
- What is cascade control?
- What is a feedback control system and what are its types?
- What is an automatic control system?
What are the factors that must be done for the controller tuning?
If the process is fast then the controller must be tuned fast, slow or fast in a controller refers to reset.
- Before tuning the model structure must be defined as the first order, linear, etc
- There would always be errors for the linear models due to the non-linear plant and also because of its change in operation. So based on this, the tuning must be done, the tuning should be done by considering the chance of errors in the process. So if the dynamics of the process changes then the controller will be capable to handle the process
- The disturbance must be defined with their behavior
- The measured variable must be selected for the process
For the controller
- The controlled structure must be defined like if it is PI or PID
- Tuning constant is the proper way to find out the optimum values of the tuning parameter
- In order to get the proper controller performance, to get the proper controlled variable performance the IAE of the controlled variable must be selected
- In order to get the proper manipulated variable behavior, the manipulative variable must not have variation outside the defined limit.
What is the need for the controller tuning and how should it be done?
In order to vary the control parameters, the control loop is tuned the parameters are integral gain, derivative gain, etc. So all these control parameters are adjusted in order to get the required control response. If the controller parameters are not selected properly then the controlled process input will be unstable. The tuning would vary from trial and error attempt, in order to determine the proper control parameters the tuning will be done according to the process model. The change in the process or set point is entirely dependent upon the application, certain processes must not do the overshoot of the process variable from the set-point. Some process should reduce the energy which is expanded while reaching the set point. So in order to attain good stability, the process should be stable for any combination of the process condition and set points.
Loop tuning is really not simple, because of the process response time, in order to change the setpoint or to achieve a proper output it would take minutes or hours. Certain processes are non-linear, so in this process, parameters could work better in good load conditions but it won’t be the same during the starting or low load condition so this process must be tuned according to it. There are many types of controller tuning we must select the tuning method according to the process model. We must check if the loop can be tuned offline or online while selecting the tuning method and also need to consider the system speed.
What are the factors that must be considered while doing the PID tuning?
We must control four variable while doing the PID controller tuning
Rise time – it is the required time for the initial output of the system to reach more than ninety percent of its required output.
Overshoot – It is the amount of the initial response which exceeds the set-point value.
Resolving time – It is the time taken by the system to reach the set point value
Steady state error – It is the measured difference between the system output and set-point value
We must tune three parameters in a PID controller in order to achieve the process requirements
Proportional term (P)
This is the amount added to the output according to the error, so the tuning parameters for the proportional terms are proportional gain, proportional band, and the conversion between the P gain and P band.
Integral term (I)
This is the amount added to the output according to the sum of errors. So the parameters that must be tuned will be the time constant, inverse time constant, reset rate, and conversion between the time and reset rate.
Derivative term (D)
It is the amount that is subtracted from the error according to the rate of change of error. The required tuning parameters for the derivative terms are time constant and the derivative gain.
How to do the manual tuning of the controller?
We can choose manual tuning when the system must remain online during the tuning process, manual tuning procedure.
- We must set the integral gain and derivative gain to zero
- The proportional gain should be increased till the output loop is not stable
- In order to get one half of this value the proportional gain must be reduced so that a quarter-wave decay can be obtained
- The integral gain must be increased to control the offset so that the system can resolve in proper time.
If the integral gain is increased really high then it will cause system instability, if so the derivative gain should be varied till the system resolves to its set point.
How to do the cascade loop tuning
Cascade loop tuning
Slave controller tuning
The close loop tuning method put the slave controller in auto mode and for the open-loop method put it in manual mode after that apply the tuning procedure for the slave controller.
In this tuning, the slave controller must be in CAS mode and the master must be in the auto mode and this must be done for the closed-loop tuning. In the open-loop method put the master in the manual mode. Apply the tuning procedure for the master controller. If the slave controller is not operating as the requirement of the master controller then there is no need for doing master setting the set point of the slave, in this case, the slave needs to be fixed.
What are the types of controller tuning?
Open loop tuning
This type of tuning is done in an open-loop, isolated from the process, while doing the open-loop test it will disturb the process so the operator must be alert while doing the open-loop tuning. There are certain types of open-loop tuning they are single pulse test and double pulse test and these tests are done in conjunction with the software tools.
If we do the double pulse test properly then the process variable would reach the setpoint and also the test time will be shortened. This advantage can’t be seen in a single pulse test but it would help to reach the process variable to the normal value and thus it would help to prevent the deviation from the normal process condition.
Single pulse and double pulse test
In the single pulse test, the controller output is stepped up and when the measured value shows a proper value, then the controller output will have its original value. . The process would only reach the steady-state after the return step. The double test is a two-pulse test done in opposite directions, the second pulse will be implemented really soon when the first pulse gets proper response from the process. In this type of test, we would get data both above and below the design level of operation
Process reaction curve technique
This is an open-loop tuning method it would react when a step input is applied to the system. At first, we need to apply some control outputs to the system manually and should record the response curve. Then we need to check the dead time, slope, rise time of the curve and these values should be given to P and I, D equations so that we could get the gain values of PID terms.
Closed loop tuning (Ziegler Nichols method)
The first step is to know the control loop dynamic behavior. We should estimate the controller tuning parameters which creates a required response for the dynamic characteristics determined in the first step. We should match the controller to the other elements’ personalities in the loop. The process dynamic characteristics are determined by the gain of the proportional controller and also according to the loop oscillation period.
In this method, the controller should be placed automatically with low gain and there shouldn’t be any reset or derivative. After that increase the gain slowly and when the oscillation starts make a slight change in the setpoint. Then the gain should be adjusted so that the oscillations can continue with a constant amplitude. The period and the gain should be noted, the ultimate gain is the gain in which the oscillation takes place with a constant amplitude.
Trial and error method
It is a simple PID tuning method and it can be done while working on the system or controller. In this type at first, we have to set the integral gain and derivative gain to zero, and then we should increase the proportional gain till the system is not stable. So during the system oscillation adjust the integral gain so that the system will be stable and also adjust the derivative parameter to get the proper speed.