search

henryhcliu / udmc_carla

This repository achieves an optimization-based unified decision-making and control for urban autonomous driving in Carla simulator.
7 stars 0 forks source link

No file？ #1

Open Happy-Yasuo opened 10 months ago

Happy-Yasuo commented 10 months ago

emmm, currently there are no files in the program....

henryhcliu commented 10 months ago

Thank you for your feedback. We apologize for any inconvenience caused. Please note that this repository is currently intended for advertisement purposes, showcasing our latest work. We will release the latest version of the code as soon as possible in due course. Your attention and understanding are greatly appreciated.

Gaochengzhi commented 7 months ago

Is real-time execution of such large-scale AV feasible? Does this system operate asynchronously? I am doing a similar project.

henryhcliu commented 6 months ago

Is real-time execution of such large-scale AV feasible? Does this system operate asynchronously? I am doing a similar project.

This framework/system operates synchronously with the simulator. It allows us to control the vehicles in real-time, and the control performance is visually showcased in the simulator's spectator view simultaneously. Could you please provide more details about your intriguing project?

Gaochengzhi commented 6 months ago

I am doing a Cooperative Driving Framework in Carla, but I find that the way OpenCDA does (a) Synchronous Blocking Update casues the runtime grow as the agent number increase, so I choose to do it in (c) Asynchronous Blocking Update

AsynchronousVsSynchronous@1x

The work will be finished in the coming days, by the time I will update my github https://github.com/Gaochengzhi/Carla_Truck_Platoon/tree/main Simulation architecturse vehicle implemntation

Looking forward to commuicate with you more on Carla. My wehcat is Gaocz1999wechat 😁

henryhcliu commented 5 months ago

I am doing a Cooperative Driving Framework in Carla, but I find that the way OpenCDA does (a) Synchronous Blocking Update casues the runtime grow as the agent number increase, so I choose to do it in (c) Asynchronous Blocking Update

AsynchronousVsSynchronous@1x

The work will be finished in the coming days, by the time I will update my github https://github.com/Gaochengzhi/Carla_Truck_Platoon/tree/main Simulation architecturse vehicle implemntation

Looking forward to commuicate with you more on Carla. My wehcat is Gaocz1999wechat 😁

It appears that OpenCDA currently achieves cooperative driving by employing separate algorithms for each vehicle, functioning as a prototype to assess and validate the feasibility of the generalized framework. While it is true that the initial synchronous blocking update is less efficient when compared to your asynchronous blocking update using multi-threads, I believe there is still room for optimization. In my simulation settings, all the surrounding vehicles are generated and operated using the CARLA official APIs, which have a high probability of parallel updating, thereby enabling higher engineering efficiency.

By the way, I have added your account for further discussion about the simulation. Thank you for the impressive and detailed feedback about your excellent work. =D