Role of controllers in web guiding systems

[Music] The controller is the central processor that takes the sensor input, computes the corrective action and sends that information to the actuator. Nowadays, the controllers also include human machine interface like an operator interface. Previously, the controller could be standalone. It didn't really have an interface.

So the controller could be analog in the sense of electrical analog or pneumatic analog controllers. So basically controller is taking the sensor signal and then making the necessary computation so that the actuator can be positioned at the desired location. In terms of terminology, gain is one of the most common things that you're going to hear in controllers. That's basically saying how quickly or what kind of a dynamic response you need.

gain is going to do that. Other things that you're going to see is operating voltage, power consumption, whether you have a operator interface or not. And then whether this is a controller for a servo motor or stepper motor, they have drives or drivers for it. How many sensor inputs you have?

Does it have Ethernet connectivity? Does it have remote control and stuff like that? In terms of the control structure, most control systems have webguide control systems have this kind of a structure where you have a fixed game proportional control. You really don't need more than a proportional control for a web guide because there's integrator built into it.

But usually you have something like you have a motor, it might have a current loop, it might also have a velocity loop with a tachometer. Then you have a guide structure which has its own transmission ratio. And there's the web dynamic which is unknown. Web dynamics means that if you move the guide 1 mm, how much is the web going to move?

That really depends upon transport conditions, the stiffness of the web, tension and all those kind of things. And then finally, you have a sensor that measures the edge position and then it sends that to a position controller. So this is a pretty simple architecture for most webguide controllers. They are fixed gain and most often they are d-tuned because of the stability and all the other reasons.

Most web guides their controller is kind of d-tuned for the conditions. If you want to get the best out of it, you would need to retune them. The tuning has to be based on the optimal performance because the web dynamics is unknown. Most often DC motors or DC servo motors or stepper motors are used in this kind of control structure.

There are some other advanced control technologies like adaptive control where the controller can adapt or learn on the fly and tuning may not be required. When we say learning on the fly, it means that it adapts to sensor gain changes or the dynamics of the web. In our case, we have a similar structure as the one I showed first. It's still a fixed game controller, but with some motion control aspects built into it in terms of curving the position and having trajectories for velocity, we can increase the stability of the controller and provide a pretty aggressive output performance.

You can have a current loop, a position loop, if you have a encoder, and then a position of the actuator here. And finally, the web position, which includes the web dynamic. Just to give you an idea, we have a lot of different things that we can do with the controller. But the dynamics is basically if you move the this is showing a step response, openloop step response for a web.

This was like a non-woven web that we had at different speeds and see how it behaves. And you can see that when you have a step, even though it's open loop, it's trying to get there. The openloop dynamics is different based on how fast you're running. So faster you're running, it gets to that desired location uh as fast as possible.

But the slower you're running, it takes longer. Again, this is an openloop response. This is the final part of the whole web guiding, which is the dynamics of the web. Now if we add a controller to it then we can have a much better response and actually push the web guide.

In this case at the two different speeds that we were running at this was the reference change and this was the actual response of the web guide at the sensor. And then when we have another sensor installed once span downstream you can see how long it takes for that to go through. When we have a step response, we were able to get up to about 170 mm/s or like 7 in per second. This is about close to 70% improvement over an openloop response.

And again, this is closed loop. That means you're actually actively guiding the web. So with a proper control structure design, you can get a high bandwidth system close to 6 hertz or something like that and then even get a well damped system. So you can actually have an aggressive correction if you need to.

In terms of the characteristics of a good webg guiding system, it should have the ability to attenuate disturbances, easy to tune. Obviously, it needs to be stable, has good processing power so that it can process multiple sensors, have industrial Ethernet connectivity. These are for advanced functionalities. And then smart intelligence for industry 4.0.

[Music]