參數(shù)優(yōu)化的圖卷積門(mén)控循環(huán)網(wǎng)絡(luò)地鐵客流預(yù)測(cè)

打開(kāi)文本圖片集

中圖分類號(hào)：U293.5 文獻(xiàn)標(biāo)志碼：A

本文引用格式：，李露，.參數(shù)優(yōu)化的圖卷積門(mén)控循環(huán)網(wǎng)絡(luò)地鐵客流預(yù)測(cè)[J].華東交通大學(xué)學(xué)報(bào)，2025，42（（3））：77-

Parameter Optimization of Graph Convolution Gated Recurrent Neural Network for Subway Passenger Flow Prediction

Zhang Yang，Li Lubin， Chen Yanling （SchoolofTransportation，F(xiàn)ujianUniversityofTechnology，F(xiàn)uzhou35o118，China）

Abstract：Fully exploiting the spatial correlation of passenger flow between related stations in the subway network has a positive effect on the improvement of subway passenger flow prediction accuracy. Capturing and quantifying spatial paterns in passenger flow data is dificult due to the difficulty of learning and transferring spatial correlations between metro stations.An improved graph-convolution gated recurrent neural network （GCGRU） metro passenger flow prediction model was proposed to enhance the model’s ability to handle different data types by integrating multivariate spatio-temporal data.The spider wasp optimisation （SWO） algorithm based on Tent chaotic mapping and Levy flight disturbance strategy was used to dynamically adjust the model structural parameters inorder to optimize the hidden layer structure of the gated recurrent neural network.The experimental results show thatthe prediction accuracy of the model is significantly higher on weekdays than on weekends，and the root mean square error， mean absolute eror，and mean absolute percentage error are reduced by 13 percentage points，12 percentage points，and O.O8 percentage points，respectively，during weekdays compared to wekends.Dynamic optimization ofthe hidden structure of gated recurrent networks can lead to beter convergence of the prediction model and higher prediction accuracy.

Key Words： gated recurent neural network; graph convolution operation; attention mechanism; Levy flight disturbance; subway passenger flow prediction

Citation format： ZHANG Y，LI L B，CHENYL.Parameter optimization of graph convolution gated recurrent neural network for subway passenger flow prediction[J].Journal ofEast China Jiaotong University，2O25，42（3）： 77-86.

隨著我國(guó)城市化進(jìn)程的加快，城市人口逐年增長(zhǎng)，地鐵面臨著越來(lái)越大的客流壓力，合理的列車(chē)調(diào)度及客流管控策略受到地鐵運(yùn)營(yíng)方的重視。（剩余12413字）