GPS and lean angle

[Paul M]

Hi All
I have a project on at the moment and I would some input please.
I am trying to find a good methord of finding lean angle on a road bike.

Many Thank

Paul

[Holmz]

Hi All
I have a project on at the moment and I would some input please.
I am trying to find a good methord of finding lean angle on a road bike.

Many Thanks

Paul

Paul,

GPS only gives you position and rate.
Differential-GPS can be used to do this with 2 GPS units fore and aft to get pitch in a differential-GPS setup, to get roll angle you need the GPS units port and starboard - but the bikes I like are narrow.

The other way is use a gyro to get the lean angle.
I have a unit that gives roll, pitch and yaw - but also need to work on the external-maths so a .dll plugin can be used to condition the data.
This has been a slow process, but it is what I am working on.

Is this for a race bike or is it for a road bike?
Is it a personal project or other sort of project?
Where are you located?
I am looking for a rider to carry the gear to do this so I can work out the data - but not have a lot of luck with the local crowd.

[Paul M]

Hi
I am in Brisbane, the project is for a Government dept.

The Project is being used as driver training and BLACK SPOT detection with the hope of reducing bike accidents

Maybe there is room for the project to be expanded to a national level by early 2010

I would happy to work with you

Paul

[Holmz]

Paul - that took me by surprise.
I was expecting either:
1) Race team
2) A grad student research project
3) Motorcycle company
4) Keen individual

I am PM'ing my contact details.

[toby]

I have been looking at lean angle on bikes for years and I don't believe there is a perfect answer as yet from either gyros, accelerometers or GPS or even a combination. You can get pretty close in post processing but then you still have no proper baseline to work on unless you use a dual antenna GPS system with the antennas about 1 meter apart across the bike. What I have always been striving to achieve is a 0-5 volt output proportional to lean. If someone has this solution at a realistic cost, I would be very interested to know about it!

[DarrenR]

The ultimate way to do it is with two laser ride height sensors, one mounted each side and comparing the difference. this gives angle to the road surface at least. They are very expensive tho... Inferred height sensors would work as well and are inexpensive tho not as accurate and effected be sunlight and wet roads.

Another way is to use ground speed and heading rate of change (or yaw rate). A given corner radius and speed will require a given lean angle. Factoring in gyroscopic forces, this should give a pretty good idea of the angle from point of contact to the ground and the centre of gravity of the bike and rider. Probably as good as you can get from just a GPS receiver.

[Holmz]

I have his pretty much sorted now... (I am using an IMU and GPS).
I am looking to wire-in wheel speed and steering position, but there never seems to be enough time for everything.

I have just ordered a GoPro camera so I can attach the video with data... Or maybe Data onto top of video describes it better.
What can I use for the overlay to make a YouTube of it? "TrackVision"?

[DarrenR]

We trialled an IMU (MicroStrain 3DM-GX1) on a bike with video, and while the internally generated roll angle we got from it was pretty good most of the time, it got lost quite a few times and would track back on the straights. Comparing to the angle of the horizon on the video is a very good way to tell if it was correct!
Our test could have been better since we had limited time, so not conclusive.

In i2 Pro under the file menu, 'Generate Video' is made to export data to TrackVision, so yes use that to overlay your data on your video.

[Polux RSV]

Holmz, wich kind of IMU do you use ? Is it ok with eavy acceleration, braking, wheelies, etc?

Regards

Angelo

[Holmz]

Holmz, wich kind of IMU do you use ? Is it ok with eavy acceleration, braking, wheelies, etc?

Regards

Angelo

Angelo,

See Darren's post above.
I set my other computer up as dual boot, so I lost track of this post.

I have a +/- 10G with 1080 degrees/second on the roll pitch yaw rate.
If you have a lot of shock/vibration, then you be clipping some of the time.
If you are using RPY (roll, pitch, yaw) the IMU determines those, and does some filtering.
The other way to use the raw measurements which can allow for better conditional issues.
I do not believe that this can be easily done in the ACL itself, but maybe it can...

Just got a camera a month ago, and have it charging.
Found an error in my code and have it running currently in the other room, but it is looking like it is converging finally. Doing this in MATLAB, because the young student I had working with me last June was using MATLAB.
I'll have to convert this to a .DLL at some point as MATLAB can be a it slow.
I am working on a Kalman filter to keep the IMU trained with the GPS data.
The GPS jumps around as satellites come and go in the view, which happens quickly around hills, bridges etc.
So additional logic for that is needed.

The are other options for IMUs, and ideally any s/w processing would be best either as a DLL that is called from the i2, or as a built in feature within i2.
I am working on the code to get the DLL approach.

Progress is always slow, and I might have another business trip in April.