I have found the solution to my problem to be deleting the static declaration of some variables in my cost function. Maybe memory directions got too far in time for the numeric computation to be efficient.
Carlos Tampier Ingeniero de Desarrollo Advanced Mining Technology Center (AMTC) Universidad de Chile Av. Tupper 2007, 837-0451 Santiago, Chile Tel: +56-2-2977 1019 E-mail: [email protected] On Thu, Oct 29, 2015 at 9:41 AM, Carlos Tampier < [email protected]> wrote: > Hello, > > I'm currently using the COBYLA algorithm to solve an optimized trajectory > planning use a vehicle's kinematics (basically finding a trajectory > generated from the vehicle's kinematics and that tries to maximize speed > while staying away from obstacles). I'm calling the nlopt optimization at a > fixed rate (e.g. 2 Hz) and I've found it to take considerably more time to > find the optimal variables after a number of calls (e.g. the first couple > of calls take 10ms to finish and after 100 calls the optimization time is > around 100ms. Is there anything that could be going wrong for this to > happen? > I have run the simulation for over 10 hours while giving nlopt a maximum > time of 300ms per optimization and the computer didn't crash or anything, > but the optimization time was always around those 300ms. > Also I've stopped the trajectory planner and restarted it with the > simulation still going and I get back to the 10ms for the first iterations, > so I doubt it's a matter of my vehicle finding more complex obstacles (the > obstacles are actually the walls of a tunnel the car is moving through). > > Any comments and ideas are appreciated. > Regards, > > Carlos Tampier > Ingeniero de Desarrollo > Advanced Mining Technology Center (AMTC) > Universidad de Chile > Av. Tupper 2007, 837-0451 Santiago, Chile > > Tel: +56-2-2977 1019 > E-mail: [email protected] >
_______________________________________________ NLopt-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/nlopt-discuss
