Feedback Loops and Controlled Plants

Hi nzur

It is as clear as the economic policy of Zimbabwe.

Is this related to servos or steppers?

If so google for Galil and have a look around.

I recall that you're interested in biomechanics, so I think this has something to do with that, perhaps something such as ankle angle. But, I'll give you a motor example since (I think) it's easier to understand. The principles are the same; it's only that bio systems are complex/messy/undefined.

So, let's say you have a motor where the speed is proportional to the applied voltage. Now you want it to run at 1000 rpm. There are two ways to do this.

1) You put a tachometer on the motor. Tachometers generate a voltage directly proportional to the speed (they're just little generators). You feed that signal from the tach back to a control circuit. That circuit compares the "feedback" voltage to a control voltage. If the two match, nothing happens. If the feedback is too low, the drive voltage to the motor is increased. If the feedback is too high, the drive voltage to the motor is decreased.

2) You learn, either by experience or from a calibration curve, the voltage that normally gives you the desired speed and you just put that in. If it's a little too high, you're overspeed. If it's a little too low, you're underspeed.

Your question is quite complicated in terms of biomechanics and requires a fairly extensive knowledge of A&P and control theory. Most of the people I've known who work in this area are post-docs and beyond. If you're interested, I can give you some links to follow to see what people do with this and what sort of resources are out there.

Thank you TVP45.

You always great answer for people.

In my view, it means that you have a certain control loop controlling, for example, a motor speed by means of applied power, at 1000 RPM. This is a closed loop while there's a speed measurement, and the controller will vary the power to achieve the set speed even if the load in the motor shaft varies. Cannot be simpler.

And, for this controller, there's an external input that tells it at what speed the motor should be run, depending on other process variables. What it does is to vary the closed loop setpoint, lets say, to 1200 RPM, due to any required change in the process. At this point on, the controller will seek the new value, until a new value is informed.

If all the plant fails but there's still power available, the motor will run around the set point.

Thanks for you opinion.

I think you probably close to the answer, yeah conceptually.

lets say, there is an inverted pendulum which is controlled by feedback scheme.

its set point(equilibrium point) is always zero.

when it goes with this situation, if there were some sort of perturbations..

THen, inverted pendulum is going to be controlled to go upright again.

caz, its set point is zero(upright).

but what if the controller is out of order suddenly while inverted pendulum is going back..

will this inverted pendulum still going back to its upright point(Set point) as if controller still works ?

---------------

your saying..

If all the plant fails but there's still power available, the motor will run around the set point.

---------------

pls, explain your saying one more time what this means..

"all the plant fails" ??

"but what if the controller is out of order suddenly while inverted pendulum is going back.. will this inverted pendulum still going back to its upright point(Set point) as if controller still works ?"

No. But this depends on the failure mode of the controller. Sometimes it might force the pendulum in one or other extreme directions and "stick" there, other times it could oscillate. If your controller is prone to failure you will need to build in a fail safe mechanism (not necessaruly mechanical) that will detect controller malfunction and return the pendulum to a safe or default position.

Nzur,

Your model is too simplistic for the question. In postural sway, the inverted pendulum is suitable for simple COP (center of pressure) studies, but is not sufficient for adding any sort of control theory.

Further, in biomechanics, you have to look at the type of control failure. People don't usually have massive strokes and remain standing. Control failure processes are often slow and chronic, e.g., Parkinson's, Meniere's, Gullain-Barré, etc. and each disease will screw up the control system in a different way. This is the basis for using postural studies for differential diagnoses.

I don't know your background, or your access to journal articles, but I would heartily recommend this article (and it's citations) as a starting point.

http://linkinghub.elsevier.com/retrieve/pii/002192909400113I

"the feedback loop and the controlled plant constitute an autonomous "closed-loop" system" ? The closed loop system is capable of controlling itself with no external inputs, outside of the ones inside the loop itself, be they temperature,speed,pressure,level,etc.The loop is autonomous and closed.It responds to changes in the variables automatically. In an open loop mode,a set value is plugged in, and is usually derived from a "best fit" scenario, but is incapable of making changes if the controlled process changes dynamically. An automobile is a good example.When the engine is cold, the Oxygen sensor does not function.It must get to about 600 degrees before it operates properly.The engine control computer uses a set of pre-defined variables to control the engine till the sensor comes on line. It then changes to closed loop mode, and the fuel,air mixture is automatically adjusted for changing conditions to maintain ideal fuel,air,and ignition timing, to name a few. Hope this helps------------------ HTRN

Taking an example from a speed controlled motor in industrial settings, a speed regulated section is in a simple form just a PI closed loop control. It regulates to zero speed error. However, if a change in reference occurs, there is a definite response that may be under or overdamped. By using feedforward control and controlling the rate of change of velocity reference we can estimate the torque required for the acceleration and minimize the transient.

In a biological system we often have many inputs (visual, auditory, pressure, joint angle, muscle tension, balance from the inner ear, etc) that can provide multiple feedback paths, and also provide information for feedforward. As an example a batter estimating the speed of a pitch to initiate the swing of the bat in anticipation of the strike.