Subway short-term passenger flow forecasting method based on LS-SVM and real-time big data

Abstract

The invention discloses subway short-term passenger flow forecasting method based on LS-SVM and real-time big data. The method includes the following steps: (1) obtaining passenger in-out time and passenger out-in time, and obtaining passenger historical travel station OD coefficient; 2, adopt a Logit model to calculate that route selection probability of passengers involved in the route selectionof the ring line; 3, By using the historical OD coefficient to train the LS-SVM model to obtain the ratio coefficient of up-to-down route and up-to-down route. Using passenger flow data in and out ofthe station to train LS-SVM model to obtain the passenger entry and exit forecasts. 4, calculate that transfer ratio of the transfer station by adopting the OD coefficient and the Logit model, and carry out correction through the real-time video identification technology; Step 5, obtaining the distribution of passengers in the road network and updating the road network. The invention obtains useful real-time data through a screening system and corrects the prediction result, thereby obtaining the prediction result with higher accuracy and real-time performance.

Description

technical field [0001] The invention provides a short-term subway passenger flow prediction method based on LS-SVM and real-time big data, that is, provides a method for predicting passenger flow at subway stations and line sections based on AFC swiping card data and path selection probability, and belongs to the field of computer application technology. Background technique [0002] With the rapid development of urban subway construction, the subway network has become complex, and the routes available for travel are diversified. Short-term line passenger flow is very important to passenger route selection. If passengers can know the passenger flow of subway stations and lines in 5 minutes, 30 minutes, or even 60 minutes, they can adjust their travel plans in time according to the predicted passenger flow and avoid Congested lines can achieve the purpose of balancing passenger flow for the subway network. The short-term subway passenger flow forecast can promptly remind the...

AI Technical Summary

Problems solved by technology

[0011] Existing methods only predict based on historical passenger flow and real-time inbound passenger flow. The shortcoming is that there is no judgment on the change trend of real-time inbound passenger flow, which affects the effect of short-term passenger flow forecasting. There is no research on the route selection of the circular line. Neglecting the classification of subway passenger flow affects the forecast of station and line passenger flow

Images