Accurate and reliable traffic flow prediction is of significant importance for alleviating traffic congestion and serves as the foundation for deploying intelligent traffic management systems. Existing traffic flow prediction models often rely on data collected from fixed devices. However, these fixed devices cannot cover the entire traffic network due to high maintenance costs and other reasons. In response to this problem, this paper proposes a short-term traffic flow prediction CNNBILSTM-Attention model optimized based on the sparrow search algorithm. The model utilizes the sparrow search algorithm to search for global optimal advantages, thereby improving the accuracy of the traffic flow prediction model. A case study was conducted using ICV detection data obtained from the simulation of the Huangtian intersection in SUMO. The performance of various models, including LSTM, BiLSTM, CNNBiLSTM, CNNBiLSTM-ATT, and SSA-CNNBiLSTM-ATT, was compared at a 5% penetration rate of intelligent connected vehicles. Experimental calculations yielded the following results: the Mean Absolute Error (MAE) for the above models were 4.78, 4.75, 4.89, 4.69, and 4.66, respectively, while the Root Mean Square Error (RMSE) was 6.39, 6.45, 6.26, 6.20 and 5.97, respectively. Additionally, the predictive performance of the model was explored under different penetration rates of intelligent connected vehicles. The analysis revealed that the performance of SSA-CNNBiLSTM-ATT improved with increasing penetration rates of intelligent connected vehicles. The experimental results demonstrate that the performance of the SSA-CNNBILSTM-ATT model is significantly superior to existing models.
Traffic Flow Prediction Based on Sparrow Search Algorithm Optimized Neural Network Using Intelligent Connected Vehicle Data
2023-10-28
1544187 byte
Conference paper
Electronic Resource
English
European Patent Office | 2023
|Optimized Deep Neural Network Based Intelligent Decision Support System for Traffic State Prediction
Springer Verlag | 2023
|Intelligent traffic road network traffic flow prediction method based on Bayesian algorithm
European Patent Office | 2024
|Intelligent Traffic Flow Prediction Model Based on Deep Neural Network
Springer Verlag | 2023
|Unmanned Aerial Vehicle Path Planning Based on Disturbed Fluid and Sparrow Search Algorithm
Springer Verlag | 2022
|