A Two-Stage Real-Time Path Planning: Application to the Overtaking Manuever

Abstract

This paper proposes a two-stage local path planning approach to deal with all kinds of scenarios (i.e. intersections, turns, roundabouts). The first stage carries out an off-line optimization , considering vehicle kinematics and road constraints. The second stage includes all dynamic obstacles in the scene, generating a continuous path in real-time. Human-like driving style is provided by evaluating the sharpness of the road bends and the available space among them, optimizing the drivable area. The proposed approach is validated on overtaking scenarios where real-time path planning generation plays a key role. Simulation and real results on an experimental automated platform provide encouraging results, generating real-time collision-free paths while maintaining the defined smoothness criteria. INDEX TERMS automated driving, path planning, overtaking, intelligent transportation systems

A Hybrid Planning Approach Based on MPC and Parametric Curves for Overtaking Maneuvers

“…Another tendency is the use of real-time optimization methods because they have a good capacity to solve the problem without programming specific cases or using an extensive dataset [ 14 ]. Another approach uses trajectory planning algorithms that are able to provide safe, human-like and comfortable trajectories by quintic Bézier curves [ 27 ].…”

“…Additionally, the total distance of the generated trajectory is fixed considering a vision horizon. Nevertheless, in these experiments the nominal trajectory was pre-computed offline, in a newer version of this method, the nominal trajectory is computed online [ 14 ].…”

“…A considerable number of works were executed in sampling-based approaches but general solutions using OCP have not been studied enough, especially the benefits of MPC with the currently available solvers (e.g., ACADO [ 13 ]). Other approaches are based on offline optimization but with an extra stage including dynamic obstacles [ 14 ]. A good review of the motion planning techniques for lane-change and obstacle avoidance for highway scenarios is presented in [ 15 ].…”

Platoon Merging Approach Based on Hybrid Trajectory Planning and CACC Strategies

“…For the overtaking maneuver, trajectory generation through parametric curves and Model Predictive Control (MPC) has shown good results [ 26 ]. Other recent approaches for overtaking use a two-stage planning method to tackle the overtaking maneuver, but using pre-computed optimal curves [ 27 ]. So far, in any of these examples, a mixed approach for optimal longitudinal and lateral functionalities has not been implemented for the two platoons merging.…”

A Holistic Safe Planner for Automated Driving Considering Interaction With Human Drivers

“…The main drawback of [16] [17] is the computational expense in iteratively generating multiple curves to find a collision-free path. An efficient trajectory generation method proposed in [18] uses a piecewise quartic Bézier curve for path smoothing, lane changing, and lane replanning. The curve maintains curvature continuity in the planned trajectory but relies on an offline database containing all potential turning angles and curve lengths, which is a drawback.…”