- No sooner did I let the post get out than I realized that

I was dead wrong. Please disregard my comment. The

calculation is correct as stated in the document.>

Sorry,

> deltaDistance = (leftCounts+rightCounts)/2.0 * distancePerCount

>

Gary - The document is correct. The calculation is intended to determine the

distance the center of the robot moves, which, for small movements, is

the average of the left and right encoder counts multipled by the

distance a wheel travels for each count:

deltaDistance = averageEncoderCounts * distancePerCount

where

averageEncoderCounts = (leftCounts + rightCounts)/2.0

and

deltaDistance = ((leftCounts+rightCounts)/2.0) * distancePerCount

or

deltaDistance = (leftCounts+rightCounts)/2.0 * distancePerCount

For example, if the left encoder increments by 2 and the right encoder

increments by 3, the center of the robot will move the average of the

distances the two wheels traverse, which is 2.5 * distancePerCount.

RidgeSoft Support

--- In intellibrain@yahoogroups.com, "G.W. Lucas" <gwlucas@...> wrote:

>

> I have a comment on the article at the RidgeSoft web site "Enabling

> Your Robot to Keep Track of Its Position". I also have a request

> for the membership of this discussion group in general.

>

> First, the material in the article is excellent and it is written

> very clearly, but there is a small glitch in one of the example

> calculations. Page 5 gives the text

>

> deltaDistance = (leftCounts+rightCounts)/2.0 * distancePerCount

>

> the way a Java would interpret the order of operations, it isn't

> quite right and should be restated as

>

> deltaDistance = (leftCounts+rightCounts) / (2.0*distancePerCount)

>

> The actual code snippets offered in the document all look okay.

>

>

> Second, I have a request. There is a Java applet at the Rossum

> Project web site that allows a user to interactively play with

> different combinations of parameters for differential drive system

> like the one the IntelliBrain-Bot uses. If anyone has a Java-enabled

> browser and feels inclined to do so, I'd like to get some feedback

> on whether the applet agrees with your own experience in working

> with the IntelliBrain-Bot. I'm always interested in how well things

> like the applet square with reality... The URL for the applet is

>

> http://rossum.sourceforge.net/tools/MotionApplet/

>

> the source code for the applet is available at the site as is a

> paper that talks a bit about the logic used in the applet at

>

> http://rossum.sourceforge.net/papers/DiffSteer/

>

>

> Thanks.

>

> Gary

>