Control Approaches for Tugboats

I'm taking the liberty of cross-posting this to sci.engr.control, because that group needs the traffic and because if there was anyone there that'd be the group to post to.

I'd break it down this way: for any maneuver that you you want to perform, you want to have a torque and a thrust. Solve for the lowest-cost way to attain that, and you're done. That leaves a lot hanging (just what is lowest cost?) but hopefully it provides you with some structure for your thinking.

If there's no speed disadvantage to pointing in a particular way (i.e. if the thruster is one of those tilting vane types, not a rotating nozzle), then your problem is just one of figuring out what's the lowest immediate cost.

One example (the one with the lowest engineering cost :), is to establish that

torque = d * y thrust = x

where your forward thruster is always generating a thrust equal to x * i

• y * k (where i and k are the unit vectors in the forward and side directions), and your rear thruster is always generating a thrust equal to x * i - y * k.

Now your math is easy, and you can proceed to getting the rest of the system working right. Once you're not crashing into barges or squishing pleasure craft up against the sides of canals, then you can think of ways to do this that use less fuel and put less wear and tear on the thrusters.

This system is actually a rotating nozzle ...

Oh fun. That means that you've got a problem with more than average nonlinearity. How fast can you rotate a nozzle compared to the dynamics of the boat? If the answer is 'not too', then you're in for all sorts of work trying to (a) position the nozzles so that they will have the best chance of getting to where they're needed the most quickly, and (b) dealing with stability et all when you cannot treat their dynamics in a linear fashion.

I know that thrusters have been used on boats for years, so there should be paper after paper in the literature someplace. Just a web search with the keywords boat, thruster, steering may do it.

Unless those thrusters can transit _fast_ though, your problem cannot be pulled fully into the linear domain.

Hi,

I find it hard to understand what the system looks like. What's the size of the boat, where are the thrusters mounted, how do they adjust a position and how are they driven? Do you have a drawing? Are sensors available for position, direction or something else?

-- pt