A short-term passenger flow prediction method for urban rail transit based on deep learning

Abstract

The present invention relates to a kind of urban rail transit passenger flow forecasting method based on long-term short-term memory network, it is characterized in that, comprises steps as follows: Step 1, determine the input of improved spatio-temporal long-term short-term memory network model; k The x stations s with the highest spatial correlation 1 , s 2 …s x ; Step 3, calculate train by station s 1 , s 2 …s x Run to the site to be predicted s k Required time step 4, according to the inbound and outbound passenger flow of each station in the whole network described in step 1, the station s to be predicted is obtained k The sequence of outbound passenger flow and get the ratio of the site to be predicted s k early station s 1 , s 2 …s x In step 5 of the inbound passenger flow sequence, input the improved spatio-temporal long-term short-term memory network model and the output is the site to be predicted s k The outbound passenger flow; step 6, calculate the forecast performance index. The invention is used to predict the outbound passenger flow of the station, and improves the prediction accuracy.

Description

technical field [0001] The present invention relates to a method for predicting short-term passenger flow of urban rail transit based on deep learning, in particular to using an improved spatio-temporal long-term short-term memory network (Sp-LSTM) to deeply study historical passenger flow data of relevant stations in the entire network, thereby improving passenger flow prediction accuracy. The invention belongs to the technical field of urban rail transit passenger flow prediction. Background technique [0002] Since the 21st century, the construction of urban rail transit has entered a climax, and urban rail transit with large capacity, fast speed, safety and reliability plays an increasingly important role in modern cities. Passenger flow forecasting is of great significance in urban rail transit management. Accurate short-term passenger flow forecast is helpful for the operation department to arrange the personnel on duty on the platform in advance to avoid the occurre...

AI Technical Summary

Problems solved by technology

However, most forecasting models are based on shallow learning and single-station forecasting, ignoring the interaction between line-network stations under networked operations, and not taking into account the spatiotemporal nature of passenger flow data, resulting in large deviations between the forecast results and the actual situation. Difficulty guiding operating companies to make accurate decisions

Images