融合动态风险图与多变量注意力机制的车辆轨迹预测模型
|
|
陈文强,冯琳越,王东丹,顾玉磊,赵轩
|
Vehicle trajectory prediction model integrating dynamic risk map and multivariate attention mechanism
|
|
Wenqiang CHEN,Linyue FENG,Dongdan WANG,Yulei GU,Xuan ZHAO
|
|
| 表 1 不同模型的均方根误差对比 |
| Tab.1 Root mean square error comparison of different models |
|
| 数据集 | tp/s | RMSE/m | | CS-LSTM | MHA-LSTM | STDAN | iNATran | STGAMT | GRIP | GSTCN | RGMA | | NGSIM | 1 | 0.61 | 0.41 | 0.42 | 0.41 | 0.21 | 0.37 | 0.44 | 0.16 | | NGSIM | 2 | 1.27 | 1.01 | 1.01 | 1 | 0.78 | 0.86 | 0.83 | 0.60 | | NGSIM | 3 | 2.09 | 1.74 | 1.69 | 1.7 | 1.49 | 1.45 | 1.33 | 1.17 | | NGSIM | 4 | 3.1 | 2.67 | 2.56 | 2.57 | 2.4 | 2.21 | 2.01 | 1.88 | | NGSIM | 5 | 4.37 | 3.83 | 3.67 | 3.66 | 3.57 | 3.16 | 2.98 | 2.73 | | NGSIM | 平均值 | 2.29 | 1.93 | 1.87 | 1.87 | 1.69 | 1.61 | 1.52 | 1.31 | | HighD | 1 | 0.22 | 0.06 | 0.19 | 0.04 | 0.07 | — | — | 0.05 | | HighD | 2 | 0.61 | 0.09 | 0.27 | 0.05 | 0.19 | — | — | 0.14 | | HighD | 3 | 1.24 | 0.24 | 0.48 | 0.21 | 0.32 | — | — | 0.27 | | HighD | 4 | 2.1 | 0.59 | 0.91 | 0.54 | 0.61 | — | — | 0.47 | | HighD | 5 | 3.27 | 1.18 | 1.66 | 1.11 | 1.14 | — | — | 0.80 | | HighD | 平均值 | 1.49 | 0.43 | 0.70 | 0.39 | 0.47 | — | — | 0.35 |
|
|
|