Estudo e análise comparativa de algoritmos de planejamento de caminho para veículos autônomos

Abstract

The autonomous land vehicle can carry loads and people from an origin point to the destination using a system composed by its perception of the environment in which it is inserted, path planning and action. The perception covers several combined sensors able of promoting the environment perception as well as identifying objects and elements. When it comes to path planning, there are algorithms and strategies used in manoeuvres performing in the exact and safe way, following traffic rules. Regarding action, strategies of longitudinal and lateral control are present in order to move the vehicle. Specifically, the path planning is responsible for stable and comfortable user's displacement during path planning, manoeuvre and trajectory. In this work two algorithms of path planning applied to autonomous vehicles are presented, analysed, and set side by side. A map of a parking was defined representing a scenario to evaluate and compare both algorithms performances with respect to path planning and vehicles manoeuvres within parking slots.