We applied flatsim simulation model to an actual model ,by replacing the flatscan source with relal lidar scan data ,which is published as flatscan source in the “flatsim.app.json”,and we replaced the diff_base_state to control our vehicle too。
It seems that the navigation works well after we build the map by the “logmapping” module.But there’s one problem still confused us, wish you can be helpful:
After the vehicle has arrived the desired_goal & made the direction correct, the vehicle was still moving and turn around。
It looks like that it didn‘t think itself has finish the goal.
And,mostly important issue is that the vehicle will somehow speed up ,and then lost itself.
In reality, our vehicle didn’t work as well as simulation do. Suddenly speed up can be very dangerous, obviously we can not apply this in production.
As we has made a very good job in ROS-SLAM using the same vehicle, and now we don’t know where could the problem probably happen.
Hi Jad_Keryakos，
I’m very glad for your suggestion.It helps me a lot.
Since I printed the linearSpeed and angularSpeed values.I found the speed is within a reasonable range after the vehicle has arrived the goal,linearSpeed and angularSpeed usually lower than 0.1 m/s or 0.1rad/s. It seems that the suddenly speed up turn around is something wrong in our vehicle driver.
To compare the vehicle moving state,I reconfigure the “flatsim” module to the initialized values.Only thing changed is that I copied the edge component
to publish to our vehicle.
All I did is to prove that “After the vehicle has arrived the desired_goal & made the direction correct, the vehicle was still moving and turn around。”
is not caused by our wrong config.Things has proved that even in the official simulation example “flatsim”,The vehicle can’t stay static when it has already arrived the goal and made the direction correct. We have tried some found relate-like params, like “arrived_position_thresholds”/“waypoints:radis” etc.It didn’t work.
By the way ,the above situations happens when we configured the “desired_behavior” as the “waypoint”,and just give one goal at a time.Things happend when it has arrrived the goal and didn’t get the next waypoint.In the initialized mode “random_walk” with continuous random goal thing all going well.
That’s all my situation.
Hope for you next wonderful advice!
Best Regards.