Control loop performance advisory tool

A stiction control system for compensating for stiction of an actuator of a controllable pneumatic valve, comprising:

• a proportional-integral-derivative (PID) controller configured to receive set-point values (SP) at a PID controller input and output control signals at a PID controller output (OP);
• a first switch connected to the PID controller output;
• a stiction compensator switchably connected to the PID controller output by the first switch and configured to receive the control signals;
• a controllable pneumatic valve input connected to the stiction compensator and switchably connected to the PID controller by the first switch;
• a vibration measurement device connected to an output of the controllable pneumatic valve, the vibration measurement device configured to generate electrical signals based on an oscillation of an actuator of the controllable pneumatic valve;
• a process measurement device connected at an input to the output of the controllable pneumatic valve and configured to output a process variable (PV) at a process output;
• a feedback loop connected between the process output and the PID controller input;
• a computing device connected to the stiction compensator, the computing device including circuitry and program instructions stored therein that, when executed by one or more processors, cause the one or more processors apply an artificial intelligence algorithm to the electrical signals of oscillation of the actuator of the controllable pneumatic valve, wherein the artificial intelligence algorithm generates sets of stiction reduction values;
• and
• wherein the stiction compensator is configured to modify the control signals based on the sets of stiction reduction values,
• wherein the computing device is further configured to:
• build a digital twin model of the controllable pneumatic valve;
• receive at least one current segment of electrical signals based on an oscillation of the controllable pneumatic valve;
• monitor the operation of the controllable pneumatic valve by comparing the at least one current segment to the digital twin model;
• detect stiction when an anomaly is detected between the at least one current segment and the digital twin model;
• determine a severity of the stiction;
• compensate for the severity of the stiction by generating the sets of stiction reduction values;
• transmit the sets of stiction reduction values to the stiction compensator;
• and
• display the digital twin model of the controllable pneumatic valve with a representation of the severity of the stiction;
• and
• compare the at least one current segment to the digital twin model by:
• applying differential evolution to the at least one current segment and/or the digital twin model of the at least one controllable pneumatic valve;
• identifying a best proportional-integral-derivative (PID) outer loop control gain and/or a PID inner loop control gain which reduces the severity of the stiction;
• and
• compensate for the stiction by applying the best PID outer loop control gain to a PID outer loop that controls the actuator;
• wherein the computer device is further configured to:
• determine which of the control signals based on a finite impulse response filter parameters, the control signals based on the non-linear inverse model, the PID outer loop control gain and the PID inner loop control gain has a greatest effect on reducing the stiction of the digital twin model;
• and
• generate instructions for the stiction control device to apply the control signals, the PID outer loop control gain or the PID inner loop control gain which have the greatest effect on reducing the stiction of the digital twin model to the actuator of the at least one control valve.