Optimization of Velocity Ramps with Survival Analysis for Intersection Merge-Ins

Abstract

We consider the problem of correct motion planning for T-intersection merge-ins of arbitrary geometry and vehicle density. A merge-in support system has to estimate the chances that a gap between two consecutive vehicles can be taken successfully. In contrast to previous models based on heuristic gap size rules, we present an approach which optimizes the integral risk of the situation using parametrized velocity ramps. It accounts for the risks from curves and all involved vehicles (front and rear on all paths) with a so-called survival analysis. For comparison, we also introduce a specially designed extension of the Intelligent Driver Model (IDM) for entering intersections. We show in a quantitative statistical evaluation that the survival method provides advantages in terms of lower absolute risk (i.e., no crash happens) and better risk-utility tradeoff (i.e., making better use of appearing gaps). Furthermore, our approach generalizes to more complex situations with additional risk sources.

A Risk and Comfort Optimizing Motion Planning Scheme for Merging Scenarios

“…As for yielding the target state is static, a solely jerk optimal trajectory is planned as described in [16]. Once x ego = x(t f ), the merging problem is simplified to the problem formulation descirbed by Puphal et al [3] and can be solved accordingly.…”

A Foresighted Driver Model derived from Integral Expected Risk

“…A full-blown optimization of the ego-vehicle predicted trajectory space is also feasible, and has been studied in numerous previous publications related to risk estimation and risk maps with survival theory, see e.g. [2]. Here, however, we introduced a simple and computationally cheap model which nevertheless is able to capture risks that occur during normal driving situations.…”

“…the risk of the own predicted behavior within the predicted scene. By systematic variation of the own behavior, appropriate safe actions can be found as the result of an optimization process, see [2]- [5] for behavior architectures which follow this scheme.…”

“…FE stands for "First Event", meaning that PFE,e is the probability that event e takes place for the first time at t + s if we start at t. We will come back to this formulation later in sections IV-B and V-A 2. Ie are the risk indicators resp.…”

Development of a Cooperative On-Demand Intersection Assistant

“…Therefore significant effort is invested in the research of assistance systems which support the driver in such left-turn scenarios with crossing traffic (6) and also for the less challenging task of only oncoming traffic (7), (8) . To further improve these systems, in particular the participants' manoeuver and intention prediction is investigated (9), (10), (11) . These systems provide collision avoidance functionality, which prevents safety-critical situations.…”