Dynamic Positioning with single jet

Hi,
I develop systems for marine geological survey company. We do a lot of bottom sampling where we winch down equipment on the bottom and do measuring for a couple of minutes. We normally stay at a site for maybe 30 minutes. The boat is a 9x2.5 boat, 4000 kg with 300Hp diesel and a hamilton jet.

We use either anchor or "manual dynamic positioning". The "manual dynamic positioning" is done by putting the stern to the wind/current and have quite low engine speed and use the bucket in reverse. By adjusting the steering and bucket we can maintain a position within a couple of meters, we have done it in currents up to 2 knots without issues but it is very demanding to do this manually! The heading of the boat is of no interest (as long as it doesn't start to rotate), the focus is to keep the position of the stern fixed.

We have autopilot on the boat so there is a pump that I can use and we have accurate heading and position (RTK with ±2 cm accuracy). I plan to use a big servo that manually is attached to the bucket lever when you need to activate dynamic positioning.

Can this be done?

The issue for me is the control algorithm, any leads on how I should approach that problem?

Hardware and embedded software is no problem and I can run a control loop at 10Hz (that is the update rate we get on heading and position) where I adjust steering and bucket angle.

Cheers
Hakan

 

what important heading or position and how will your software work that out?

 

Hi,
position and heading is sent over NMEA 0183 with 10Hz.
I will adjust the position from the GPS using heading since it is mounted on the bow of the boat and I need the position of the stern.

The NMEA0183 data will be processed in a custom made PCB with an embedded controller and some drivers so I can control the speed and direction of the hydraulic pump for the steering and adjust the position of the servo that controls the bucket.

I will probably try some simple PID controller that will calculate a force based on the position error and the I-part will handle wind and current.

The direction that should be used for the force could be more tricky....

/hakan

 

sure
but when its out of position and heading what will happen?

 

How are you getting your heading data? Your autopilot doesn't care if you're underway or not, if your heading data is fast enough your A/P might be able to "hold course." You may need to reverse the wires to the pump so it pumps backwards to account for you being in reverse.
If the heading data is too slow this probably won't work, I would think response time needs to be fast for this to work and keep you on station. You might have to play with your A/P settings to get the response time and speed right, and it may only work under certain conditions for a given set of settings.

Some Harbor Pilots are using a very accurate heading system by mounting 2 GPS antenna's a couple of meters apart. The 2 different signals compute the heading in very near real time. a flux gate compass may work well enough for this too. I would think GPS heading information would be useless as there is no way on and therefore no way to compute accurate heading data.

 

i'm still trying to get to grips with holding position and heading with one source of thrust at one end of the vessel

 

I've operated many DP vessels from 1990 to 2011.
I can't imagine how a single thruster could DP.
However, a slim possibility exists to do it, if the computer could recognize "the sweet spot" and use it.
A heading exactly/proportionately dividing the wind and current forces so there is no tendency to fall off course or slide to either side, is called the "sweet spot".

In ship handling, the sweet spot is used to deliberately slide (crab) into a position, or follow a course (as in walking a diver) slowly and carefully controlled. Or in landing a barge alongside another vessel offshore. The sweet spot is constantly changing a few degrees because of wind gusts ect.

If the computer could detect the sweet spot and delicately steer/correct so the correct headings were maintained (constantly adjusting), then speed could be controlled to only and precisely compensate for drift.
If it could do all of that, then it could DP in appropriate weather with some necessary wind and current.
Wind and current need to be of sufficient magnitude that the vessel can use enough thrust to steer without head reaching.
DP isn't static. It's not PARK. It's HOVER.
We consider a vessel is DPing well, if we aren't oscillating more than 4 feet port to starboard and 8 feet fore and aft.

Oilfield boat captains are too enamored of working only stern into the wind and current. They're missing some skills. TALK TO A 4 POINT/DP DIVE BOAT CAPTAIN. And not all of THOSE are possessing all the boat handling skills. Someone with tug and dive boat and OSV experience is pretty much well rounded.
Stern in to weather, the boat weathervanes to NEAR the sweet spot, but not precisely enough. Put your bow into weather, and the sweet spot is apparent when you find it.
goodluck

 

You can buy a system off the shelf from either Minn Kota or Motorguide that will hold position to within 5' even in varying wind and current

It can work off of magnetic north, or GPS positioning.

These systems are on small electric trolling motors, but the software is all there, all you need to do adapt it to control a larger drive. These are bow-mount systems and work well.

 

The target here is to keep the stern at a precis position, heading will not be controlled. With those constrains it is possible to do this with a single waterjet.
If they can do this with a trolling motor at the bow we can do the same with a waterjet at the stern. Both motors can produce a force vector in either direction.

The question was about the control algorithm.

I would call this dynamic positioning (of the stern) but heading of the vessel isn't controlled. I know how a real DP system works, we have a Kongsberg system on our ship (40x12 m), and there we use 4 thrusters (fixed pitch) and two propellers (variable pitch) and it works really fine.

As I stated earlier in the thread we do this manually today so it is possible.

We use a gps compass to get the heading and the rate of the heading update is 10Hz.
It uses moving base rtk to calculate the position differance between two antennas with a separation of 0.5 m.

We have a RTK GNSS receiver at the bow of the boat and I will use heading info to calculate the position of the stern.
Heading info will also be needed to calculate what direction the waterjet must point at to correct the the position error.

Cheers
Hakan

 

good luck