smoothing algorithm
Posted by
Tom Eldredge
on 2001-06-22 07:02:05 UTC
Gentlemen,
Thanks to all of you who helped me decide which way to go on the point smoothing algorithm. A lot of ideas were shared which were too complex for me to tackle at this time, but this little routine does work, and I think it will serve my needs for now.
I have attached the point smoothing algorithm source code to this email. It takes input from a file which includes two characters, which indicate whether to accelerate and decelerate a move, plus the x and y coordinates of a point, and it smooths the points out. The input file, "cutfile.no1" , in my case is part of a file created by my control. It is a pre-proccessing control, so that the g codes are read in by the control, and a file is created which includes the results of a slope comparison routine, which compares the slope of the previous move with the present move and the next anticipated move. From this slope comparison and other considerations, the software makes a decision whether the present move should include acceleration and deceleration. In my case, I made a decision not to smooth out the points where there are radical changes in the slope, thus the two alpha characters, indicating Yes or No for acceleration and deceleration, respectively.
You can certainly strip all of the Y and N stuff out, and get down to the basic routine if you want to smooth all the points out. I have included a sample input file, and a basic routine, along with an exe program.
If it is of any value, feel free to use it. The more times it loops, the smoother the profile gets. It is set to loop 5 times.
It is a simple thing, and if someone wants to put it in the files section, have at it. I have never added anything to that goldmine before, so if someone thinks it should be added, go ahead. In fact, someone might want to trim it down first, to include just the points, and then save it.
Thanks again,
Tom Eldredge
[Non-text portions of this message have been removed]
Thanks to all of you who helped me decide which way to go on the point smoothing algorithm. A lot of ideas were shared which were too complex for me to tackle at this time, but this little routine does work, and I think it will serve my needs for now.
I have attached the point smoothing algorithm source code to this email. It takes input from a file which includes two characters, which indicate whether to accelerate and decelerate a move, plus the x and y coordinates of a point, and it smooths the points out. The input file, "cutfile.no1" , in my case is part of a file created by my control. It is a pre-proccessing control, so that the g codes are read in by the control, and a file is created which includes the results of a slope comparison routine, which compares the slope of the previous move with the present move and the next anticipated move. From this slope comparison and other considerations, the software makes a decision whether the present move should include acceleration and deceleration. In my case, I made a decision not to smooth out the points where there are radical changes in the slope, thus the two alpha characters, indicating Yes or No for acceleration and deceleration, respectively.
You can certainly strip all of the Y and N stuff out, and get down to the basic routine if you want to smooth all the points out. I have included a sample input file, and a basic routine, along with an exe program.
If it is of any value, feel free to use it. The more times it loops, the smoother the profile gets. It is set to loop 5 times.
It is a simple thing, and if someone wants to put it in the files section, have at it. I have never added anything to that goldmine before, so if someone thinks it should be added, go ahead. In fact, someone might want to trim it down first, to include just the points, and then save it.
Thanks again,
Tom Eldredge
[Non-text portions of this message have been removed]
Discussion Thread
Tom Eldredge
2001-06-22 07:02:05 UTC
smoothing algorithm
Tom Eldredge
2001-06-22 07:30:57 UTC
smoothing algorithm
Drew Rogge
2001-06-22 07:35:18 UTC
Re: [CAD_CAM_EDM_DRO] smoothing algorithm
wanliker@a...
2001-06-22 08:47:16 UTC
Re: [CAD_CAM_EDM_DRO] smoothing algorithm
Alan Marconett KM6VV
2001-06-22 11:11:19 UTC
Re: [CAD_CAM_EDM_DRO] smoothing algorithm
info.host@b...
2001-06-22 11:20:43 UTC
Re: [CAD_CAM_EDM_DRO] smoothing algorithm
Tom Eldredge
2001-06-23 06:23:36 UTC
smoothing algorithm