Research on Passenger Flow Forecast of Hangzhou Metro Based on LSTM-SVR

Accurate short-term subway passenger traffic forecast is beneficial to active traffic control and travel planning of travellers. The paper proposes a long-term and short-term memory model-support vector regression (LSTM-SVR) combination model to simulate short-term subway passenger traffic prediction. LSTM is a variant of RNN. It has advantages in processing time series data. It has strong feature extraction ability. Support Vector Regression (SVR) is a regression algorithm based on SVM. It has high dimensional expression ability. In the case of data fitting, and the sensitivity of SVR kernel parameters, a chaotic particle swarm optimization strategy is proposed. A combined learning model based on LSTM-SVR was constructed, which was compared with long-short memory model (LSTM), support vector machine (SVR) and nearest neighbor regression (KNN). The simulation method was used to verify the proposed method. Effectiveness.