Use actual depth vbs expectations #113
Conversation
Signed-off-by: Arjo Chakravarty <[email protected]>
Signed-off-by: Arjo Chakravarty <[email protected]>
Signed-off-by: Arjo Chakravarty <[email protected]>
Signed-off-by: Arjo Chakravarty <[email protected]>
| this->CheckRange(26000, {16.95, 3.13, -5.16, 0.00, 0.00, 1.66}, {15.0, 3.14}); | ||
| this->CheckRange(28000, {16.65, 5.57, -6.79, -0.00, 0.02, 2.17}, {15.0, 3.14}); | ||
| ignition::math::Vector3d maxVel(0, 0, 0); | ||
| for (int i = 1; i <= this->tethysPoses.size(); i ++) |
There was a problem hiding this comment.
How about not checking every single pose? It takes longer with little benefit. I think that every 1000~2000 iterations should be enough to catch issues.
There was a problem hiding this comment.
Ping @arjo129 , I think this comment is especially important now that the number of iterations was increased
There was a problem hiding this comment.
We can address this later
Signed-off-by: Arjo Chakravarty <[email protected]>
…_expectations Signed-off-by: Arjo Chakravarty <[email protected]>
Signed-off-by: Arjo Chakravarty <[email protected]>
…ctations Signed-off-by: Arjo Chakravarty <[email protected]>
…lrauv into arjo/fix/testdepthvbs_expectations Signed-off-by: Arjo Chakravarty <[email protected]>
Signed-off-by: Arjo Chakravarty <[email protected]>
Signed-off-by: Arjo Chakravarty <[email protected]>
Signed-off-by: Arjo Chakravarty <[email protected]>
| this->CheckRange(26000, {16.95, 3.13, -5.16, 0.00, 0.00, 1.66}, {15.0, 3.14}); | ||
| this->CheckRange(28000, {16.65, 5.57, -6.79, -0.00, 0.02, 2.17}, {15.0, 3.14}); | ||
| ignition::math::Vector3d maxVel(0, 0, 0); | ||
| for (int i = 1; i <= this->tethysPoses.size(); i ++) |
There was a problem hiding this comment.
Ping @arjo129 , I think this comment is especially important now that the number of iterations was increased
| ignition::math::Vector3d maxVel(0, 0, 0); | ||
| for (int i = 1; i <= this->tethysPoses.size(); i ++) | ||
| { | ||
| // Vehicle roll should be constant |
There was a problem hiding this comment.
Should we check other quantities that should also be constant? i.e. X, Y and yaw?
There was a problem hiding this comment.
X, Y and Yaw are not constant. This was also found in real life data. There is a significant amount of translation along the X axis and yawing.
There was a problem hiding this comment.
Do we know how much change we should expect on these dimensions? Is there a point when it's too much?
|
I get the concern over excessive translation. We don't have such data. As I
said earlier it is ok for the vehicle to translate and spin (significantly
as normally the pitchmass is used to prevent the vehicle from excessive
pitching and thus translation). If you look at the end of the test we do
check that the overall vehicle velocity reduces. If we really want we can
run the test for longer till the velocity approaches zero. It doesn't make
sense for us to set upper bounds on translation as there is no clear amount
and we will have to keep changing it. The test pitchmass + depth VBS
mission however should have some bounds on translation.
…On Wed, Dec 22, 2021, 11:37 AM Louise Poubel ***@***.***> wrote:
***@***.**** commented on this pull request.
------------------------------
In lrauv_ignition_plugins/test/test_mission_depth_vbs.cc
<#113 (comment)>:
> - this->CheckRange(12000, {14.82, 8.71, -15.58, 0.00, 0.04, 2.55}, {15.0, 3.14});
- this->CheckRange(14000, {10.69, 10.22, -16.95, -0.00, -0.03, -2.80}, {15.0, 3.14});
- this->CheckRange(16000, {7.85, 8.85, -16.07, -0.00, -0.01, -2.13}, {15.0, 3.14});
- this->CheckRange(18000, {6.48, 6.29, -13.84, 0.00, -0.07, -1.59}, {15.0, 3.14});
- this->CheckRange(20000, {6.99, 2.26, -10.73, 0.00, -0.17, -0.93}, {15.0, 3.14});
- this->CheckRange(22000, {11.22, -0.97, -7.71, -0.00, -0.10, 0.06}, {15.0, 3.14});
- this->CheckRange(24000, {15.47, 0.32, -5.62, -0.00, 0.04, 0.98}, {15.0, 3.14});
- this->CheckRange(26000, {16.95, 3.13, -5.16, 0.00, 0.00, 1.66}, {15.0, 3.14});
- this->CheckRange(28000, {16.65, 5.57, -6.79, -0.00, 0.02, 2.17}, {15.0, 3.14});
+ int maxIterations{28000};
+ ASSERT_LT(maxIterations, this->tethysPoses.size());
+
+ ignition::math::Vector3d maxVel(0, 0, 0);
+ for (int i = 1; i <= this->tethysPoses.size(); i ++)
+ {
+ // Vehicle roll should be constant
Do we know how much change we should expect on these dimensions? Is there
a point when it's too much?
—
Reply to this email directly, view it on GitHub
<#113 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAEEMQDYUMV5SIMSVAGUV7LUSFBV3ANCNFSM5JYUPS3A>
.
Triage notifications on the go with GitHub Mobile for iOS
<https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675>
or Android
<https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
I see, ok, so it sounds like it's difficult (impossible?) to use this test to detect regressions on our dynamics if it introduces extra drifting, and we have other tests that can be used to check that instead 👍 |
|
I'll merge this so the test becomes stable for other branches |
This test is a rewrite of the testDepthVBS test. The key change is instead of a trajectory, we check that the vehicle eventually reaches a depth of 10m and its depth rate eventually reaches zero. It also checks that the velocity is reducing slowly. I expect this to be a ❌ till the upstream changes in the controller are merged and a new docker image is generated,