PIDs and Air Brakes

[pallen]

I am currently working on a project that uses an air brake controlled via proportioning valve to provide load to a motor driven D.U.T.   Load feedback is acquired through a torque cell.

Initially, I was unable to achieve satisfactory control using just the PID.  The system seemed too slow to react to changes in load.  The oscillations were very bad regardless of how low I set the Proportional Gain. 

Looking at other NI tools, I found the Update Rate Limiter VI and implemented it.  With this I was able to achieve a much better steady-state.  But the limited rate meant that it took too long to achieve the desired setpoint.

At this stage I have implemented a "Set Point Ramp" that will limits the rate at which the setpoint will rise or fall.  I combined this with two Rate Limit settings.  One for fast rise/fall.  One for steady state operation. 

While this is the best result I have achieved so far, the rise and fall graphs are not pretty, nor are they linear at all. 

I'm still looking into all my options.  But I'm curious if any other developers that have worked with air brakes and PID control would have any tips for me. 

My greatest concern with this equipment is that when the brake overshoots, the torque rises very fast.  Even with "safety" code that shuts things down if a line is crossed, I have concern for my D.U.T.  

Thanks. 

[ssk]

In the past I have implemented similar solutions but not relied on the PC for the closed loop control. Instead I use a stand-alone, dedicated PID controller and write the setpoint via the PC (RS232/485 or analogue).

This frees up the PC resources and provides a quick PID loop.

ssk

[Michelle_N]

It seems strange that PID could not keep up - how fast is your system updating?

[pallen]

Hello Michelle,

Thanks for your response.

I don't think the problem is with the PID being unable to keep up.  Although my Process Variable is one of 39 Strain measurements made on that machine.  Since I need to monitor and record all of those signals during the test, I'm not if there's a way to separate that one signal from the Task I'm running for all the others.

To me it would appear the control process is complicated by the delay between the proportioning valve and the air brake.  The brake itself is also non-linear with respect to voltage applied to the proportioning valve. 

Without the Update Limiter in my PID loop, I have uncontrollable oscillations right from the start.  The trick is finding a way to both achieve my desired load quickly, and then hold it within 2% of full load.

Currently, with a combination of a Setpoint Ramp and multiple Update Limit Factors; I'm pretty close to having an acceptable solution.  I would still like to see a more linear rise to the load.  But it's maintaining a steady state that's key in this equipment.  I was just curious if others had any "tricks" they've learned from dealing with other similar braking systems.        

[Michelle_N]

It sounds like you have a controls back ground, have you thought of using LabVIEW Control Design and Simulation Module? This will give you much greater control than you are getting with PID but it also requires a better knowledge of your system and control theory.

[pallen]

Already working with it.  🙂



I've been pouring over the documentation and working out a series of tests to go through. 

What I have now has been good enough for the first phase.  But for the next phase I'll need to achieve precise and stable loads in excess of 10,000 in.lbs as quickly as possible.

Message Edited by pallen on 10-24-2008 11:32 PM

[catchvsystems@sify.com]

Hi Mr Pallan ,

I am struggling ramp rate based pressure setting in y brake application.

In my case I am using SMC proportional control valve for brake pressure. One more thing is Air to oil pressure. For this air to oil conversion  mechanical setup is ready with brake chamber.  It is working fine.

I arrived a linear equation based on the applied  voltage to  oil pressure. 

This system works fine with this arrangement.

But customer need ramp rate for  this application.

 Since my derived linear equation working fine , I am not using PID.

Can you please help me with your experience ? How to achive ramp rate with or without PID?

[GerdW]

Hi catch,

---

catchvsystems@sify.com wrote:

But customer need ramp rate for  this application.

 Since my derived linear equation working fine , I am not using PID.

---

You can use the "PID Rate limiter" function also without using a PID controller…

What have you tried and where are you stuck?

(You know this thread is 16 years old?)