Answers are grounded in this paper and cite their sources.
Comprehensive and accurate understanding of the interactive traffic environment facilitates reasonable motion planning behaviors for autonomous driving. This paper presents an overall risk assessment method for the host vehicle to achieve practical motion planning considering uncertainties of the predictive driving behaviors for the surrounding vehicles. A social temporal convolutional long short-term memory network is constructed to capture the interactive characteristics among the host and the surrounding vehicles, thus predicting the statistical distribution characteristics of the trajectory uncertainty in the prediction horizon. Then a two-dimensional Gaussion distribution-based dynamic risk assessment associated with its soft update method is developed to comprehensively quantify the driving risk spatially and temporally through constructing the occupancy map based on the multi-modal distribution of the predictive trajectories for the surrounding vehicles. Then the optimal motion is determined by minimizing the multi-objective function of the alternative potential driving behaviors including straight driving and lane changing. The effectiveness of the proposed algorithm is verified under the driving scenarios extracted from the NGSIM dataset. Results show that the proposed method can provide satisfactory motion planning results while minimizing the driving risk.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: