Project Chrono Vehicle for Automotive Control Research

[Trevor Vidano]

Hi Project Chrono,

tl;dr:

• Can you help me make the strongest argument for the credibility of using Project Chrono Vehicle in researching automotive control?
• Do you have validation studies, or sources, for the various models in the Project Chrono Vehicle database?
• Can you port the vehicles implemented in NADS to Project Chrono? 

Dan Negrut recently gave a seminar at UC Davis, which inspired me to take another look at Project Chrono (I had come across it back in version 6 or so). I currently am researching front steering control of autonomous ground vehicles. The research standard (what appears most often in literature) appears to be to use CarSim for this. I have been told that the industry standard (what many OEMs use) is Adams/Car. 

In research, open-source projects are immensely useful because they enable free exchange and repeatability of research efforts. Unfortunately, CarSim is rather expensive ($6000/year or $14000 for an indefinite license). I have not looked into Adams/Car pricing. 

I would like to argue that Project Chrono Vehicle should become the new research standard for automotive control research. However, this requires arguing that Project Chrono's credibility is at or beyond that of CarSim's. 

I have studied multi-body dynamics a little bit (a Graduate class taught by Prof. Haug), and I am quite convinced of Project Chrono's credibility because of the validation studies you have online, the funding sources, the list of current and past users, the main developers, as well as the extensive capability of the codebase. Could you help me with making a strong argument for this?

When doing automotive control research, I understand that the gold standard is to have an experimentally validated model of an existing vehicle. However, in most literature using CarSim as a virtual vehicle, it appears that a generic vehicle (one that does not exist in real life but behaves like a realistic vehicle) is sufficient for the validation of control algorithms. I am only aware of the validation study comparing Project Chrono Vehicle to Adams/Car, and I am guessing that the vehicle model used in that study is the Generic Wheeled Vehicle. However, I can't be sure. I also am assuming that the validation study is quite old, but that it is still valid because I am assuming new releases make improvements. There are numerous other vehicle models in the repository, but I am not aware of their sources or if any of them are validated. If possible could you share what you can about the source and validation work (if any) of these vehicle models?

I have also considered implementing the vehicles developed for the National Advanced Driving Simulator (NADS) in Project Chrono. However, checking that the constructed model matches the original NADS implementation is mostly eyeballing (or guesswork) without the raw data from the NADS experiments.  Since some of the Project Chrono developers have a connection to NADS (I think?) is it possible to get this data or port those models to Project Chrono?

Thank you,

Trevor 

PhD Candidate at UC Davis

[Trevor Vidano]

My apologies. You can ignore the second question as you Radu Serban previously answered this for me (never got the notification).