Real-time capable path planning for energy management systems in future vehicle architectures

摘要

In this paper an energy optimal path planning andudvelocity profile generation for our highly maneuverableudRobotic Electric Vehicle research platform ROboMObil isudpresented. The ROMO is a development of the GermanudAerospace Center’s Robotics and Mechatronics Center to copeudwith several research topics, like energy efficient, autonomousudor remote controlled driving for future (electro-) mobilityudapplications. The main task of the proposed algorithms is toudcalculate an energy optimal trajectory in a real-time capableudway. It is designed to incorporate data from actual trafficudsituations (e.g. oncoming traffic) or changed conditions (e.g.udsnowy conditions). The resulting trajectory is then fed forwardudto a lower level time independent path following control [2] thatudcalculates the motion demands for our energy optimal controludallocation. This in turn distributes the demand to the actuatorsudof the over-actuated vehicle. We show a numerical reliable wayudto formulate the energy optimal path planning optimizationudobjective, which is able to provide a consistent replanningudfeature considering the actual vehicle states. Besides this,uddifferent types of optimization methods are evaluated for theirudreal-time capabilities. The velocity profile will be calculatedudafterwards and the generation of the profile is also enabled toudhandle dynamic replanning. Finally, we show severaludexperimental results, using a virtual road definition and testsudon a commercial real-time platform.