what are the specific critics forms of the DWA's cost function,when scoring trajectories?

asked 2015-06-13 05:27:54 -0600

jxl gravatar image


i learned the theory of DWA, when i looked at the sources of the navigation,i found out dwa_local_planner package have a lot connection with the base_local_planner package. In the line 319 of the dwa_planner.cpp,it calls the SimpleScoredSamplingPlanner::findBestTrajectory(.) to found the best trajectory ,when i looked into the simple_scored_sampling_planner.cpp,there is one question confuses me .

Q: what are the specific cost function forms of the DWA? I just saw in the constructor of the DWAPlanner ,lines 167~179, dwa_planner.cpp,the ''critics'' was set up, the " generator_list" was set up ,and then class SimpleScoredSamplingPlanner of object "scored_sampling_planner_ "was instantiated.when i looked into the obstacle_cost_function.cpp, oscillation_cost_function.cpp,map_grid_cost_function.cpp,i can understand how the critics oscillation_costs_and the obstacle_costs_scores trajectories,but when i don't figure out how the global goal critics path_costs_&alignment_costs_and the local goal critics goal_costs_& goal_front_costs_scores trajectories. In the map_grid.cpp, MapGrid::setTargetCells (.) aimed at global goal critics ,MapGrid::setLocalGoal(.)aimed at local goal critics,what are the principles of the two cost functions respectively ??

Any help will be appreciated much,thanks a lot :)

edit retag flag offensive close merge delete


I'm not sure whether i am right ,the both take wavefront propagation algorithm .

jxl gravatar image jxl  ( 2015-06-19 03:05:38 -0600 )edit