融合多车风格感知与交互特征的换道行为预测
|
|
韩泽一,王文璇,王元庆
|
Prediction of lane-changing behavior integrating multi-vehicle style perception and interaction features
|
|
Zeyi HAN,Wenxuan WANG,Yuanqing WANG
|
|
| 表 2 不考虑驾驶风格的LSTM和2-CNN模型预测结果 |
| Tab.2 Prediction results of LSTM and 2-CNN models without considering driving style |
|
| T/s | Acc/% | | P/% | | R/% | | F1/% | | LSTM | 2-CNN | | LSTM | 2-CNN | | LSTM | 2-CNN | | LSTM | 2-CNN | | 1.0 | 98.81 | 99.21 | | 98.21 | 99.50 | | 96.07 | 96.71 | | 97.13 | 98.09 | | 1.5 | 98.92 | 99.06 | | 97.15 | 97.15 | | 97.85 | 98.58 | | 97.50 | 97.86 | | 2.0 | 98.60 | 98.64 | | 96.69 | 96.17 | | 96.17 | 96.93 | | 96.42 | 96.55 | | 2.5 | 98.62 | 98.40 | | 96.46 | 97.54 | | 95.88 | 93.47 | | 96.16 | 95.46 | | 3.0 | 98.26 | 98.62 | | 94.33 | 95.31 | | 94.33 | 95.70 | | 94.32 | 95.50 | | 3.5 | 98.11 | 98.52 | | 93.66 | 94.98 | | 92.91 | 94.53 | | 93.28 | 94.75 | | 4.0 | 98.55 | 98.47 | | 93.68 | 95.06 | | 95.98 | 93.82 | | 94.82 | 94.43 | | 4.5 | 98.10 | 98.42 | | 96.63 | 98.22 | | 88.88 | 89.83 | | 92.58 | 93.84 | | 5.0 | 98.45 | 98.91 | | 94.09 | 97.50 | | 91.61 | 92.37 | | 92.82 | 94.86 |
|
|
|