Rapid computation of optimal control for vehicles

This paper describes a rapid computational method of optimal control for vehicles and the software realizing it. The optimal control problem is solved with the dynamic programming technique. A traditional engineering simulator is used to predict the vehicle’s performance and economy. An objective function is suggested that permits high-speed computation. Results show the dependence of optimal fuel consumption on average speed for various vehicle masses in a number of situations: acceleration from rest to cruising speed, driving between stop signs and driving on hilly terrain. Findings include that one can travel up to 20 to 80% (depending on the situation) faster than the most economical average speed without increasing fuel consumption more than an equal amount.