I am new to PID applications, and am currently working a project implementing PID control to Maintain Constant Pressure of Power Steering Gear by Varying Load of the cylinder (Varying the Cylinder Pressure by using Proportional Valve) Connecting to the Shutter shaft of the Gear. First of all i will explain my Rig Setup:
1) It consist of Hydro motor which rotate the input Shaft(Valve Shaft) of the gear.
2) Shutter Shaft (Output shaft) Is connected to Cylinder Piston Via Pitman Arm (Rotating motion is converted in to Linear Motion).
3) Proportional Valve is Connected to the Cylinder Line for Applying load by Varying the Pressure.
4) Seperate Pressure Relief Valve for Cylinder forward and Reverse Direction.
5) Main Relief valve.
Rotate the input shaft by using Hydro motor in CW and CCW direction,and Apply the opposing load in Output shaft by varying Cylinder pressure by using proportional valve to maintain the Gear Pressure at desire value.
To maintain Constant pressure am implementing PI COntroller where my PV is the Gear Oil Pressure(+/10 vdc) and my Output is Proportional Valve (+/- 10vdc), I have tuned my System by using ZN method and my system works well for particular set point, If i disturb my system by increase the Setpoint or varying the Cylinder relief pressure the response of the system become overshoot beyond the setpoint, I have to Control my system to work at any Relief pressure at Desired Set point. So kindly Suggest me some Controlling methods.
Are you saying that the controller oscillates but eventually settles down to the correct pressure that matches the setpoint, or does the pressure just keep oscillating ? ZN tuning tends to produce oscillatory results for even a basic system. Without knowing the detail of your system, it may be that you have an integrating process and that is making it even more unstable / oscillatory. If you applied a voltage step (in open loop, no PI control) to the servovalve would the cylinder pressure just keep going rising ? Even if that isn't the case, it might be worth trying a proportional only controller anyway, to see if that results in a stable controller - integral action tends push a control loop closer to instabilty (you may need to consider using a PD or PID controller to get both speed of response and stability).
If you have a more fundamental problem, like the pressure doesn't settle to the right value at all, then it could be something simple like you don't have the sign of the control action correct. Or it could be something like the control loop inteacting with the pressure relief valve.
Hope this helps,