Perhaps OT but there are a lot of smart cookies in this NG.
I have a discrete representation of an analogue PI control algorithm. On the surface it looks very much like a velocity algorithm except for the integral part which is:
Ki*Ts*(E1+E2)
I was expecting to see:
Ki*Ts*E1
What effect will this have? And why has it been done like this?
On my basic level of thinking, this will act as to increase the integral action. What sort of system would this be required for?
Another common equation for PI controller would be....
u(i) = u(i-1) + Kp*(1+Ki*Ts)*(E1+E2)
This assumes that E1 and E2 are opposite sign (one is the process variable, the other is the process setpoint). Reverse the order of E1 and E2 to get reverse action. Often analog PI controllers are implemented such that the prop. gain compounds the integral component. Both these suffer from 'reset windup' when the process is in manual control, and transferring to auto would not be 'bumpless'.
As far as 'speed' vs. 'position', that is determined outside of these equations. If 'E1' is the process variable, this math can work for any 'process' such as position, speed, tank level, pH, whatever. If E1 comes from a 'speed' transducer, then you can control the speed of something with this. If E1 comes from a 'position' transducer, then it controls 'position' (all assuming that the output will have some affect on the sensed input through the 'process').
There is another form for PID where the 'D' stands for differential control. Here, the derivative of the process variable is used to provide another degree of stability. If the process variable being controlled is 'position', then the deriviative of that is 'speed'. But the derivative is usually calculated internally, not sensed with a separate transducer. (although a separate speed transducer is sometimes used)
Then again, there are cascade systems where the goal is control position. The output of the 'outer loop' PI controller is routed to an 'inner loop' PI controller as its setpoint, where the inner loop controls the speed of the positioner. For example, when position is 'way off', send signal to inner loop to command a high speed in a direction to get back on setpoint. As process approaches setpoint, signal from outer to inner loop commands a slower and slower speed until, when position is at setpoint, commanded speed is zero.
So, are you trying to control speed directly with a PI control, or are you mixing it up with the derivative of position ('speed') for a PID control?? Or maybe a cascade control configuration?
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.