Multisensor-Based Predictive Control for Autonomous Parking

This article formalizes, under a single common multisensor-based predictive control framework, five different types of parking maneuvers: perpendicular, diagonal for both forward and backward motions, parallel motions. Since, from practical point view, is usually not advisable, it addressed in this work. By moving the effort motion planning to control, tasks can be completely defined solely detected empty spots. Additionally, classical compromise between completeness computational efficiency when compared exploration-based path techniques eliminated. The results few individual cases are presented against state-of-the-art approach illustrate behavior performance proposed framework as well exhaustive simulations assess its convergence. As shown convergence analyses, allows us park virtually any sensible initial pose. Finally, real experimentation using robotized Renault ZOE shows validity robustness domain approach.