基于图卷积神经网络的城市交通态势预测算法
|
|
闫旭,范晓亮,郑传潘,臧彧,王程,程明,陈龙彪
|
Urban traffic flow prediction algorithm based on graph convolutional neural networks
|
|
Xu YAN,Xiao-liang FAN,Chuan-pan ZHENG,Yu ZANG,Cheng WANG,Ming CHENG,Long-biao CHEN
|
|
| 表 3 基于PEMSD7（S）和PEMSD7（L）的周末数据采用不同方法训练模型的交通流预测准确度结果 |
| Tab.3 Traffic prediction performance comparison of different approaches for model training based on weekends data of PeMSD7（S） and PeMSD7（L） |
|
| 算法 | MAE（15/30/45 min） | | MAPE（15/30/45 min） | | RMSE（15/30/45 min） | | PeMSD7（L） | PeMSD7（S） | PeMSD7（L） | PeMSD7（L） | PeMSD7（S） | PeMSD7（L） | | ARIMA | 2.511/2.778/3.019 | 2.124/2.350/2.546 | | 5.773/6.285/6.761 | 4.824/5.259/5.646 | | 6.498/6.861/7.212 | 5.768/6.075/6.361 | | SVR | 4.157/4.562/4.825 | 3.536/3.890/4.135 | 11.289/12.053/12.542 | 8.832/9.442/9.862 | 8.984/9.395/9.662 | 7.829/8.239/8.516 | | CNN | 3.502/3.863/3.976 | 2.863/2.930/3.093 | 8.040/9.133/9.663 | 6.652/6.807/7.295 | 6.506/7.391/7.694 | 5.727/5.887/6.216 | | LSTM | 3.254/3.359/3.457 | 2.743/2.753/2.768 | 7.522/7.794/8.036 | 6.373/6.417/6.478 | 6.460/6.725/6.958 | 5.854/5.900/5.960 | | STGCN | 1.530/2.122/2.527 | 1.322/1.759/2.057 | 3.185/4.577/5.486 | 2.896/4.077/4.787 | 3.249/4.691/5.569 | 3.006/4.271/5.011 | | FFR-STGCN | 1.486/2.045/2.428 | 1.310/1.741/2.029 | 3.108/4.469/5.339 | 2.855/4.119/4.919 | 3.167/4.484/5.254 | 2.992/4.214/4.891 |
|
|
|