In order to accurately predict short-term traffic flow, alleviate traffic congestion and improve traffic operation efficiency, a short-term traffic flow prediction method based on cnn-xgboost is proposed. Combined with the temporal and spatial correlation of short-term traffic flow data, the historical data of this section and adjacent sections are taken as input for prediction. This paper uses convolutional neural networks (CNN) to extract features to reduce data redundancy. An xgboost model with parameters optimized by Drosophila algorithm is proposed for traffic flow prediction. The results show that CNN can effectively extract the traffic flow data under the combination of time and space; compared with SVR, LSTM and other models, the traffic flow prediction error of the improved xgboost model is significantly reduced.
Short-term Traffic Flow Prediction Based on Time-space Characteristics
2020-09-01
696680 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Short-term traffic flow prediction method based on space-time convolutional network
Europäisches Patentamt | 2023
|Short-term traffic flow prediction based on fuzzy time series
Kraftfahrwesen | 2002
|Short-Term Traffic Flow Prediction Based on Fuzzy Time Series
SAE Technical Papers | 2002
|Road network short-term traffic flow prediction method based on deep space-time residual network
Europäisches Patentamt | 2021
|Event-Based Short-Term Traffic Flow Prediction Model
British Library Conference Proceedings | 1995
|