CL-FDAPF trajectory planner and FO-LADRC motion controller for autonomous sweeper vehicle

Abstract

Aiming at keeping safe in time and addressing disturbance of uncertainty, an closed loop forward simulation filtering double-layer artificial potential field (CL-FDAPF) trajectory planner and first order linear active disturbance rejective control (FO-LADRC) motion controller are proposed for autonomous sweeper vehicle. Firstly, the double-layer artificial potential field, which consists of traditional potential cost layer and safe level layer, is adopted here to keep planning realtime, meet safe limitations and satisfy operational requirements, and the postprocessing of mean filtering and closed loop forward simulation is for vehicle dynamic constraints. Secondly, it is worth developing active disturbance rejection control strategy, which has the ability to accommodate uncertainty, since an accurate mathematical model of autonomous sweeper vehicle is unavailable as there being inevitable uncertainties in the system state observation and unavoidable environmental disturbances. Thirdly, several typical scenarios are designed in order to verify the real-time and reliability of the proposed algorithm. The results illustrate that the CL-FDAPF planner has highly real-time and stability as the peak time less than 0.045 s and mean time being about 0.02 s in 1000 cycles, and FO-LADRC controller has robust both at uncertainty of wheelbase and steering ratio, since the FO-LADRC have smaller lateral errors compared with two existing methods.