I don't seem to understand how the pid function of matlab works without receiving an errorin input.

with the error e(t) defined as r(t) - y(t), we know that a PID controller is defined in the following manner : u(t) = Kp x e(t) + Kd x derivative of e(t) + Ki x Integral of e(t).

Matlab documentation states : "C = pid(Kp,Ki,Kd,Tf) creates a continuous-time PID controller with proportional, integral, and derivative gains Kp, Ki, and Kd and first-order derivative filter time constant Tf:"

C=Kp+Kis+Kds/(Tfs+1).

I dont seem to understand where the error comes into play with the pid function I quoted above. I am trying to create a pid controller for reference tracking, in order words I want a controller to make the error converge to zero and make tend y(t) tend to r(t) through the the control input. I have searched a lot but found nothing useful. The last time I did such, I did it through simulink and I had no problems placing the blocks logically to get the pid in function of the error.

A good explanation would be very useful thank you !

[Sam Chak]

Hi Mohammed Azeez

The "pid" function simply allows you to create PID controller in parallel form (as shown in the transfer function), if you know the values of the control gains , , and first-order derivative filter time constant . The input to the PID transfer function is the error .

Also, the PID controller is not a one-stop solution to every type of control problems. Before deciding to implement a PID controller, it is helpful to know the system order, system dynamics, stability analysis, as well as the magnitude of the disturbance (or load) that can have significant effects on the control system.

[Mohammed Azeez]

It's crystal clear now ! Thank you very much !!!

[Mohammed Azeez]

please if you don't mind could you check this question out ? https://www.mathworks.com/matlabcentral/answers/1572323-i-don-t-seem-to-understand-the-feedback-function-of-matlab