Accuracy of succeeded waypoint

53 views
Skip to first unread message

ricardo...@gmail.com

unread,
Oct 1, 2018, 7:36:07 AM10/1/18
to LSTS Toolchain
Hi everyone.

We have an autonomous surface(sea) vehicle that has 3.5 meter length and travels with 2 m/s on the sea. We are testing it with giving different plans and routes. Unfortunately, when our vehicle was trying to reach a way point (Goto maneuver), it couldn't find the point because of precision.

For example,

In figure1, our vehicle is passing the point with 1.13 meter. However, we are on the sea and there are to many factors that affect the exact location of the vehicle. Our vehicle has to pass point less than 1 meter distance to achieve the target. However it is impossible to achieve.

As you see in figure 2, our vehicle found the way point with 0.83 meter distance from it. When we are on the sea, only one of  our tests is succeed. This is our main problem right now. Because, there are two main factor affect the location, GPS and sea current. By the way, we are using U-blox neo-m8p gps module.

Thus, we want to know that, how we can increase the radius of the target or whatever it is. We want to found the target point even if our vehicle is 5 or 10 meter away from it. We are waiting for a quick help if you can because we are stuck. We didn't found the solution and progress. Also, I attached our vehicle's nvcl config file.

Thank you everyone .
figure1.jpeg
figure2.jpeg
k1.nvcl

José Braga

unread,
Oct 1, 2018, 9:32:30 AM10/1/18
to LSTS Toolchain
Hi,

Can you share a 'Config.ini' file from any of the log files generated?
The arrival is dependent on ETA and not a hard-constrained distance.

If memory serves me right you should look into:

[General]
Time Of Arrival Factor = 5.0   // 5s is the most common value.

Kind Regards,
José



--
www.github.com/LSTS/
www.lsts.pt
---
You received this message because you are subscribed to the Google Groups "LSTS Toolchain" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lsts-toolchai...@googlegroups.com.
Visit this group at https://groups.google.com/group/lsts-toolchain.
To view this discussion on the web visit https://groups.google.com/d/msgid/lsts-toolchain/b4488c08-6f87-4515-b9d4-31e71017dd1c%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

ricardo...@gmail.com

unread,
Oct 1, 2018, 10:48:03 AM10/1/18
to LSTS Toolchain
Do you mean the parameter below? There is not any matching parameter in our .ini file from that. I'll attach .ini file for you to take a look.
If you are right, our value is lower. Maybe  if we make this value 5, it can work. What do you say ?


[Control.Path.VectorField]
Position Jump Time Factor               = 0.5

Regards.



1 Ekim 2018 Pazartesi 16:32:30 UTC+3 tarihinde José Braga yazdı:
k1.ini

Kristoffer Gryte

unread,
Oct 1, 2018, 12:06:44 PM10/1/18
to LSTS Toolchain
Hi,

Jose is reffering to a different variable that can be put in the [General] section of your config, however; it defaults to 5 if it is not explicitly set.

The "acceptance" of a Goto is related to the getEta function in PathController, which depends on this time of arrival factor, but also on the speed. I noticed that you have set "ETA Minimum Speed= 0.1", which means that if your speed is lower than this, the "radius of acceptance" is 0.1*time_factor (aka 0.1*5 = 0.5 m). So you could try to increase "ETA Minimum Speed", but if your speed is 2 m/s also when you are close to the Goto, this is probably not the cause. (I notice that you have defined it twice, so be sure to change both (or remove one) if you decide to change it)

Hope this helps,
Kristoffer

João Teixeira

unread,
Oct 1, 2018, 12:06:53 PM10/1/18
to lsts-to...@googlegroups.com
Hi, 

Please provide a log where these events occurred.

best regards,

João Teixeira

ricardo...@gmail.com

unread,
Oct 3, 2018, 2:49:23 AM10/3/18
to LSTS Toolchain
Ok, then. I will try to add new variable under the [General] section and remove one of the "ETA Minimum speed. Then, I will increase the ETA variable. Hope it'll work.

Thanks for your help,
Regards


1 Ekim 2018 Pazartesi 19:06:44 UTC+3 tarihinde Kristoffer Gryte yazdı:

José Braga

unread,
Oct 3, 2018, 5:29:18 AM10/3/18
to LSTS Toolchain
Hi,


one thing that **may** cause this is if the proper speed variables are not filled in:

if "Course Control" is true in the path controller task, then m_estate.vx and m_estate.vy (EstimatedState message) need to be filled in. If false, then m_estate.u is used.
My suggestion is to make sure the speed is set on both variables, or, if you are unable to compute one of those, to set course control to the appropriate setting.

if speed is correct on those variables already, just ignore me.

Regards,
José Braga




ricardo...@gmail.com

unread,
Oct 3, 2018, 9:12:35 AM10/3/18
to LSTS Toolchain
Hi guys,

Now, we made a test for increasing ETA Minimum Speed on land and I think it will be succeed. I can set the radius of acceptance for whatever I want. However, I will say the final result after we make sea test. If it will be fail, we will try to Jose's speed variable suggest. We noted that. One of my friend is looking it.

Thus, Kristoffer and José, thanks a lot for your help. You made great help for us. I think it will work and you will save our lives :) There was no chance for us to notice that variable changing.

I will be informed you for succession of result of sea tests. Thank you again,

Best Regards.


3 Ekim 2018 Çarşamba 12:29:18 UTC+3 tarihinde José Braga yazdı:
Reply all
Reply to author
Forward
0 new messages