Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

An urban rail transit passenger flow volume prediction method based on A-LSTM

A technology of urban rail transit and forecasting method, which is applied in the field of urban rail transit passenger flow forecasting, can solve problems such as unobvious regularity, difficult prediction, and difficult prediction of rail transit passenger flow, and achieve the effect of improving prediction accuracy

Active Publication Date: 2019-04-05
CHONGQING UNIV OF POSTS & TELECOMM
View PDF6 Cites 33 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This uncertainty brings difficulties to rail transit passenger flow forecasting, especially the short-term passenger flow is affected by more random factors, the regularity is less obvious, and the uncertainty is stronger. The key to predicting harder

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • An urban rail transit passenger flow volume prediction method based on A-LSTM
  • An urban rail transit passenger flow volume prediction method based on A-LSTM
  • An urban rail transit passenger flow volume prediction method based on A-LSTM

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0048] This embodiment mainly preprocesses the time distribution characteristics in step S3 of the present invention, adopts a hierarchical clustering method to cluster and analyze the time distribution characteristics, and further defines different sample types:

[0049] Firstly, hierarchical clustering is carried out on the time distribution characteristics. For a batch of samples studied in the present invention, there are multiple observation indicators, and there are different degrees of similarity between the indicators, so a cluster analysis method is proposed. This method aggregates samples (or indicators) with a large degree of similarity into one category, among which, the closely related samples (or indicators) are aggregated into a small taxonomic unit, and the distantly related samples are aggregated into a large taxonomic unit, thus forming different division types, and obtaining A classification system from small to large, and finally presents the relationship be...

Embodiment 2

[0058] This embodiment is mainly to further illustrate the long-short-term memory neural network A-LSTM model based on the attention mechanism of the present invention:

[0059] Bring the clustered historical traffic data with spatial and temporal features obtained in the previous step into the A-LSTM model for training. in, image 3 For the specific structure of LSTM, it is assumed that the input data of historical passenger flow data and spatio-temporal features are: x=(x 1 , x 2 ,...,x T ), the sequence of LSTM to calculate the hidden layer vector is: h=(h 1 , h 2 ,...,h T ), the real passenger flow data is y=(y 1 ,y 2 ,...,y T ), the predicted value is iteratively obtained by the following equation:

[0060] h t =H(W xh x t +W hh h t-1 +Z t-1 +b h ) (1)

[0061] Among them, W xh Represents the weight matrix of hidden layer input, W hh Represents the weight matrix of the hidden layer state input, b h is the hidden layer bias vector; H is the hidden layer...

Embodiment 3

[0069] On the basis of the first two embodiments, as an optional method, the attention model used in the present invention is as follows Figure 4 As shown, the model requires n hidden layer states: h=(h 1 , h 2 ,..., h n ) and passenger flow data x=(x1 ,x 2 ,...,x n ); return vector z, return vector z can be understood as attention mechanism value Z t collection.

[0070] The A-LSTM model constructed by the present invention is as Figure 5 As shown, a layer of attention mechanism is added to the LSTM network to receive the hidden layer state and the input passenger flow data, so as to dynamically adjust the weight of the input passenger flow data.

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention belongs to the field of machine learning, and discloses an urban rail transit passenger flow volume prediction method based on A-LSTM. The method relates to three parts of time feature extraction, spatial feature extraction and prediction algorithm design. Wherein the time characteristics are mainly characterized in that clustering analysis is carried out on the week factors througha hierarchical clustering method, and relationships among the weeks are searched; Wherein the spatial characteristics refer to passenger flow distribution characteristics of different subway stations,and the spatial passenger flow distribution relation is searched by analyzing the different stations; The prediction algorithm is mainly based on an improved LSTM neural network, and an attention mechanism is added, so that the LSTM network pays more attention to input characteristics with greater influence on prediction by a model, and passenger flow data can be predicted more accurately.

Description

technical field [0001] The invention relates to an A-LSTM (Attention Long Short Term Memory)-based passenger flow prediction method for urban rail transit; it provides services for dispatching urban rail transit, and belongs to the intersecting field of rail transit and data mining. Background technique [0002] With the rapid development of the city, the continuous improvement of the total road network, and the rapid growth of the number of motor vehicles, traffic problems have become one of the most important issues in urban development and management. With the widening gap between the limited supply of road traffic capacity and the rapid growth of total traffic demand, the contradiction between traffic demand and traffic supply has become increasingly prominent. Although many cities have built urban rail transit systems, the peak hours in the morning and evening Traffic congestion is still one of the issues that urban people are very concerned about. Moreover, with the c...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06Q10/04G06Q50/30G06N3/04
CPCG06Q10/04G06N3/045G06Q50/40
Inventor 徐光侠王天羿黄海辉刘俊李伟凤胡梦潇
Owner CHONGQING UNIV OF POSTS & TELECOMM
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com