Fly-by-wire?

[DeeGee]

Reading the small print in my autopilot manual, I find there are all sorts of things it can do, even following a route if I input same.
That betokes a level of planning rather ahead of my time, so I thought to ask here - how can this work in practice? What happens if the next wp is somewhat to windward? I can see the AP holding course if given, but do they suddenly change course for you, when the next wp arrives?

Or are these really the domain of the stinkpot (no offence meant there)?


<hr width=100% size=1>Black Sugar - the sweetest of all

 

[qsiv]

They do work better for MoBo in some respects, and no they wont look at your boats polars and work out when to tack. They wont change course automatically on reaching a waypoint, but will ask you to push a button to confirm the change of course.

The sophisticated sailing a/p will however steer to optimum vmg - a very clever trick.

<hr width=100% size=1>

 

[DeeGee]

My ST4000+ allows for sailing to a WP, it will then allow for tide: or it allows for sailing upwind, holding an apparent wind ancle, but I don't think it will sail to get the best VMG.

I say this without actually using it much in this way - I insult the device by using it as a helmsman and telling it to sail on 265, with the GPS telling me that we are actually doing 278, or whatever...

<hr width=100% size=1>Black Sugar - the sweetest of all

 

[snowleopard]

you can put in a series of waypoints, link them together as a route, then ask the autopilot to follow the route.

when you reach a waypoint it will beep at you and you have to press a button to confirm it is safe to make a course change. the track followed will be a straight line between waypoints so it will steer to cancel out tides.

if a leg of the course is upwind it will try to steer that course but will fail and come to a halt. it will probably generate an 'off course' error so you'll have to steer manually for that leg.

the system works fine as long as you set the waypoints accurately and don't route yourself over rocks etc. first time i tried it i put waypoints close to buoys and forgot to allow for differences in chart datums resulting in a very near miss with a buoy. apart from that the route took us from chichester to inverness without touching the helm, using about 30 waypoints.

<hr width=100% size=1>

 

[qsiv]

I wasnt implying it wouldnt sail closehauled - rather that if you ask it to sail above close hauled, it will mimic a learner, and simply point at the waypoint - despite the fact the sails are flapping!

<hr width=100% size=1>

 

[DeeGee]

I think your reply was actually to SnowLeopard's? What I was asking was whether some of these AP's could be set to sail as close to the wind as THEY feel is appropriate, using VMG and presumably telling you to tack if neccy when shifted? That is what I understood your previous to imply, but I was probably wrong?

Clearly, with good software (nowt special actually) they could be told to sail c/h so as to maximise vmg, and, given the destination upwind, decide whether headed or not and advise putting in a tack. If they can't do that, maybe I could get a job doing it - far less complicated than engine control systems.

<hr width=100% size=1>Black Sugar - the sweetest of all

 

[qsiv]

Thats what my software does ...

<hr width=100% size=1>

 

[DeeGee]

What sort of autopilot are you using? Your posting earlier set me thinking, that my AP must be driven by NMEA/ST signals from all the transducers, and is using such algorithms, pretty simple, designed in by the makers. It should be possible to drive the AP from my laptop interpreting and using that same data with more clever adaptive algorithms. Only thing is that the 4000+ AP requires a manual intervention. When I set it to goto a wp, it requires me to press the Track button twice to confirm, I can't see any way round this.

<hr width=100% size=1>Black Sugar - the sweetest of all

 

[bedouin]

I think if you were trying to similate that with a computer you would be best using the AWA setting rather than the GPS setting. The ST4000+s interaction with GPS is a lot more sophisticated than with the Wind System.

<hr width=100% size=1>

 

[DeeGee]

Not quite sure what you mean? Do you mean telling the ap to hold a specific awa - rather than, say, steer to a specific course? Either of those should be OK, and one's own software would have to do more or less work depending on which you want to do. My problem is that any changes in objective/target for the ST4000+, whatever the target is, seems to need interaction. This may not be the case, which is why I said 'seems to', but I haven't found anything to suggest otherwise - but then, I haven't really looked. There don't seem to be sentences in NMEA languages which can change the target. I would be glad if someone can say 'oh, yes, you need to send ABCDEF signal, and this will change the objective to which the control head is working.

<hr width=100% size=1>Black Sugar - the sweetest of all

 

[snowleopard]

i don't know of any autopilot that can sail to maximise vmg. certainly raytheon units can go to waypoint direct, follow a compass course or hold a constant angle to the wind. sounds like a good idea though.

i expect you could interface a laptop to the nmea or seatalk bus and control everything from there. i imagine that once the interface was set up the programming would be straightforward. (though i'm not sure i'd like to leave a computer running under windows in charge of my boat)

<hr width=100% size=1>

 

[qsiv]

Thats easy - B&G Hercules, but it really is only a performance boat thing. You need the add on performance processor, and to do it properly the gyro as well. The system then looks up optimum angles from the polars stored in the performance processor and then steers the boat to TRUE wind optimum angle - the true wind element is vital as it bypasses the acceleration related problem of systems steering to AWA with boats that accelerate rapidly.

I believe (but dont know for sure) that NKE do the same trick.

<hr width=100% size=1>

 

[bedouin]

If you are controlling this by an intelligent computer program, then you just need to simulate the AWA message. Set the Autohelm to maintain an AWA of say P 90, and then you can send any AWA you like from the computer to make it change course. Therefore no change of course as far as the ST4000+ is concerned - so no interaction.

In theory you could use the GPS messages instead - but the ST4000+ is more sophisticated in its use of GPS signals (e.g. it can take account of XTE and so on).

This is all theoretical of course - I have an st4000 but I haven't even connected it to the GPS yet; but if I were to try to write my own nav program that is the approach I would adopt

<hr width=100% size=1>

 

[stuartw]

As one of the stinkpot brigade, perhaps I can offer some suggestions.
Sophisticated systems can either use a wind vane input or input from the GPS in terms of cross track error.

Within the GPS, a sail plan is set up, which can have multiple waypoints. The GPS gives out amongst many others, a cross track error sentence (XTE) on the NMEA bus
This sentence contains total XTE offset in nautical miles and the direction of offset from intended track. L or R
This is fed to the Autopilot, which produces a rudder offset to counteract the XTE.
If tide or wind starts to shift the boat away from the indended track on the current leg, then the system will reduce this to nearly zero.
On reaching the next waypoint ( the closeness to it can also be defined), and you have to change tack, then it is better to assume manual control, change bearing, re-engage AP, and the XTE on the next leg will be automatically compensated for.

Most GPS's wil allow you to change waypoints on the fly, or add them or delete them. So in effect, if the next waypoint may not be reached, due to tide or wind conditions, then you can insert a new one, someway ahead of your present position, then delete the one you would not have reached. In other words the sail plan can be dynamic. An electronic input for waypoints, either Yeoman or other form of plotter can make this very quick. An additional waypoint can be put into the GPS library, then inserted into the current sail plan.

I have talked to quite a few yotties, who seemingly do not do sail plans with their GPS. Now I am aware that wind direction may change, and therefore your course may have to as well.

AP's which can change course at the waypoint, will require confirmation, just in case it turns into danger. Simpler ones, are better disengaged, change course manually, then re-engage AP. After all lads, when youre only doing 4 -5 knots, you've got all the time in the world to play



<hr width=100% size=1>

 

[DeeGee]

Looks like this is what I would like to achieve (I have an old gyro unit at home from model helicopter days)...

Cut from <A target="_blank" HREF=http://www.boatsyachtsmarinas.com>here</A>
Released 07 October 2003

Press release:

Raymarine Inc., the world leader in marine electronics for the recreational market, has introduced the SmartPilot product line, which includes a new selection of control heads for the sail and power market, and a powerful course computer with or without Raymarine precision gyro-enhanced technology for all boats up to 20,000 pounds.

"With the expansion of our autopilot line, we can now offer a full range of products with the best autopilot technology on the market today for owners of boats of all sizes. Our AST technology can now be enjoyed by the person in a 25 foot boat as well as one in a 45 foot boat," said Terry Carlson, President, Raymarine, Inc.

Joining Raymarine's SmartPilot product line is the new S1 and S1G (Gyro) course computers the new ST8001 Autopilot control head, and an all-new joystick control. The ST8001 features an ergonomic rotary control and oversized LCD. The rotary control offers precision course changes and convenient power steering control. For even greater control the optional Joystick provides effortless rudder control. The new SmartPilot S1G course computers with gyro-enhanced technology include Raymarine's AST (Advanced Steering Technology) software, which intelligently monitors the pitch and roll of the vessel and anticipates course changes. The AST software also enables rate gyro equipped SmartPilots to "AutoLearn" the boats steering characteristics, simplifying calibration and allowing the autopilot to constantly adapt to changing sea conditions. Control options and course computers can be mixed and matched according to the owner's selection of a specific control head and the computer that best suits the boat.

The new Smart Pilot line up seamlessly integrates with Raymarine radar, charting and echo sounders systems. The Smart Pilot course computers and control heads were designed for easy installation with single a connection to the display and all other interfacing cables installed below decks at the SmartPilot course computer.


<hr width=100% size=1>Black Sugar - the sweetest of all

 

[qsiv]

The only problem with this is the concept of steering to AWA. For slower, crusing boats, this works fine, but not lighter boats.

Assume steady wind direction, and assume you are reaching. Now take a gust, the wind will appear to move aft, so the pilot will puth the helm down to try and bring the wind fwd again. Now boatspeed builds - quickly, so wind comes fwd rapidly and the pilot will bear away, maybe by 20 degrees to try and return to your requested AWA. Bad behaviour on a reach - dangerous if you were already running deep, as you will gybe. Compare this to a good helmsman, who will bear away slightly at the first sniff of the gust (at the very least the helm will be UP - to counteract the weather helm that the gust will induce), and then slowly come back to the original course as the gust steadies and decreases. I 'fool' my A/P by feeding TWA on the AWA sentence, but it's a poor solution! Better are the newer system (certainly B&G, possibly NKE, and I think now Raytheon) that can be asked to steer to TWA, which doesnt change due to boat dynamics. There used to be a school of thought that wind veered in gusts, but new met office research tends to discount it - but it is still the perceived wisdom! Steering to TWA is all well and good - but it does require really good sensors and software to extract from the available on board data. The difficult elements to remove are the effects of masthead movem,ent, and upwash from the mainsail, but the good systems do just that.

<hr width=100% size=1>

 

[DeeGee]

Hmm. Have had a think overnight, and realised that I have a hand-held remote, which is great in the rain, and it is able to give nudges of 1deg, 10deg, and tack commands, on the SeaTalk bus. That is only one step away from NMEA control. I would want to control heading, rather than an indirect like XTE or AWA (a target not a control parameter) . If the winter circumscribes sailing, then I may get an opportunity for some experimentation.

I am convinced that I can do better than the current control, even if not the gyro-controlled feed-forward systems being announced right now (see bym news). The AP has to use control algorithms which are all things to all men, so using tuning algorithms one's specific craft should be possible. Watch this space....

<hr width=100% size=1>Black Sugar - the sweetest of all

 

[Twister_Ken]

XTE sometimes a good thing.

But bear in mind that sometimes you do't want to correct XTE - the classic example being a 12 hour x-tide passage where steernig to stay on the rhumb line means sailing further through the water than letting the tide take you left, then right (or vice-versa)

<hr width=100% size=1><A target="_blank" HREF=http://www.writeforweb.com/twister1>Let's Twist Again</A>

 

[bedouin]

I tend to agree - I haven't even bothered to connect my Wind instrument up to the AH yet as I am not sure I really want to use it anyway.

The way the RayMarine units work they always steer to the built in compass and correct the required compass course by comparison with the AWA, or GPS data about once a minute. This means that there is no way they can respond well in anything other than steady conditions.

My point was not actually to use the Wind Instruments to steer, but rather to use a navigational computer to simulate AWA NMEA sentences and control the AH that way.

<hr width=100% size=1>

 

[stuartw]

No, No, No!

That's the classic thinking, where you would have to point against a tidal stream, thus describing a banana shaped course.
Lets say you a going from IOW to Cherbourg. What YOU do, is that you do it on the ebb tide, which is flowing E-W. Your initial bearing would probably be heading for say Le Havre. The tide would carry you West, and hopefully end up near Cherbourg.
What I would do, is to fix a series of waypoints on a straight line, perhaps 6 say. The XTE changes would mean that the boat would probably have constant but a small amount of port rudder. Now even if you were tacking into say an South easterly wind, compounding the tidal effect, you should arrive at each waypoint as set out.
Now I can imagine the argument I am going to get with this, is the fact that the constant rudder offset will slow you down. True. But counter that with the fact you are taking a more direct route.
Now of course what I imagine one would have to do if the wind was veering, is to readjust sails. I am sure the pureists will argue that to trim the sails correctly, and get the maximum speed through the water, will give you the shortest passage time. I am not so sure.

<hr width=100% size=1>