Graph Convolution Networks for Probabilistic Modeling of Driving Acceleration
Graph Convolution Networks for Probabilistic Modeling of Driving Acceleration
复制标题
DOI：
10.1109/itsc45102.2020.9294294
复制
发表时间：
2019-11
期刊：
影响因子：
--
通讯作者：
Jianyu Su;P. Beling;Rui Guo;Kyungtae Han
中科院分区：
文献类型：
--
作者：
Jianyu Su;P. Beling;Rui Guo;Kyungtae Han
对自驾车周围交通进行建模和预测的能力对自动驾驶和智能驾驶辅助系统至关重要。加速度预测是交通量预测的重要组成部分。本文提出了解决加速度预测问题的新方法。通过用图模型表示车辆之间的空间关系，建立了广义加速度预测框架。本文研究了所提出的图卷积网络的有效性，该网络运算在预测高速公路上行驶的车辆的加速度分布的图上。我们通过集成递归神经网络进一步研究预测改进，以解决交通数据中固有的时间复杂性。综合性能指标的模拟结果支持我们提出的网络在生成预测范围内的真实轨迹方面优于最先进的方法。
The ability to model and predict ego-vehicle’s surrounding traffic is crucial for autonomous pilots and intelligent driver-assistance systems. Acceleration prediction is important as one of the major components of traffic prediction. This paper proposes novel approaches to the acceleration prediction problem. By representing spatial relationships between vehicles with a graph model, we build a generalized acceleration prediction framework. This paper studies the effectiveness of proposed Graph Convolution Networks, which operate on graphs predicting the acceleration distribution for vehicles driving on highways. We further investigate prediction improvement through integrating of Recurrent Neural Networks to disentangle the temporal complexity inherent in the traffic data. Results from simulation with comprehensive performance metrics support that our proposed networks outperform state-of-the-art methods in generating realistic trajectories over a prediction horizon.