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

Bus arrival time prediction method

A technology of time prediction and bus, which is applied in the field of intelligent transportation systems, can solve the problems of not being able to use the long-term and short-term variation characteristics of time series, and achieve the effect of ingenious design concept, good application environment and accurate prediction results

Inactive Publication Date: 2017-07-21
QINGDAO UNIV
View PDF6 Cites 30 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] In the prior art, the Kalman filtering model adjusts the static prediction results according to historical data and real-time data, which has high prediction accuracy. Insufficiency of the static prediction model, design a bus arrival time prediction method, the method is based on LSTM (long short-term memory recurrent neural network) and Kalman filter bus arrival time prediction model, fully consider the impact of various factors, Maximizing the Prediction Accuracy of Vehicle Arrival Times

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
  • Bus arrival time prediction method
  • Bus arrival time prediction method
  • Bus arrival time prediction method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0036] This embodiment provides a method for predicting the arrival time of buses. First, the static prediction model is used to predict the time that the bus takes to arrive at each station from the departure station, and then the dynamic adjustment model is used to adjust the time according to the observation time of the bus arrival at each station. Static prediction time is dynamically adjusted, specifically including the following steps:

[0037] (1), collect historical data:

[0038] Collection collects the historical data of the bus line by the bus GPS system, the historical data includes the number of trips of each bus, the number of stops, the time and speed of arriving at each station, and the number of trips of the bus is the number of trips of the bus from the origination The process of departing from the station and arriving at the departure station again becomes a train trip;

[0039] (2), historical data conversion:

[0040] Convert the historical data obtained...

Embodiment 2

[0082] The prediction process of this embodiment is the same as embodiment 1. The GPS data of No. 468 bus in Qingdao City from August 3 to August 28 is used to predict the time of the 19 stations that the bus passes through. The data on August 21 is used as a training set, and the data from August 24 to August 28 is used as a test set. Since the operation of buses on weekdays and Saturdays and Sundays is different, this embodiment does not use Saturdays, Saturdays, and Sundays. Sunday’s data is used for training and forecasting, and the division rules of the time period are as follows: from 6:00 am to 21:00 pm, every 30 minutes is divided into a section, and a bus will pass through 19 stops; the final method adopted in this embodiment It is a long-short-term memory recursive neural network with 3 nodes in the input layer, 9 nodes in the hidden layer, and 19 nodes in the output layer. All samples in the training set and test set have been normalized to obtain the bus arrival tim...

Embodiment 3

[0088] The prediction process is the same as in Example 1, and the mean square error and average absolute error of each site of the present embodiment and a single LSTM model are as follows figure 2 and image 3 As shown, it can be seen that the mean square error and mean absolute error of the time prediction value adjusted by the Kalman filter model are smaller than the mean square error and mean absolute error of the time predicted by the LSTM model, indicating that the original time baseline is dynamically adjusted by the Kalman filter The prediction accuracy is improved and the prediction bias is reduced.

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 technical field of an intelligent traffic system, and relates to a bus arrival time prediction method which includes two parts of static prediction and dynamic adjustment. Firstly the duration time of a bus from a starting station to each station is predicted by using a static prediction model, and then the static prediction time is dynamically adjusted by using a dynamic adjustment model according to observation time of the bus arriving at each station, wherein the process concretely comprises the steps of acquiring historical data, converting the historical data, normalizing the conversion data, selecting network node number, determining a basic time sequence 1, starting dynamic adjustment, determining a basic time sequence 2, calculating the final prediction value and evaluating the prediction result. According to the method, prediction is performed by the mode of combining the static prediction model and the dynamic adjustment model so that the prediction accuracy can be effectively enhanced. The method is novel and unique, ingenious in design concept, accurate in prediction result, great in application environment and wide in market prospect.

Description

[0001] Technical field: [0002] The invention belongs to the technical field of intelligent transportation systems, and relates to a bus arrival time prediction method, in particular to a bus arrival time prediction method, which is used for intelligent bus system scheduling and bus operation status understanding. [0003] Background technique: [0004] With the development of the national economy, the number of private cars is increasing, which brings huge pressure on urban traffic. Traffic problems have become a huge problem facing urban development. Therefore, the construction of intelligent transportation systems is the main task of urban traffic construction. In intelligent transportation In the system, the development of the public transportation system is the main means to alleviate the pressure of urban traffic. The public transportation system still occupies a dominant position in the construction of the intelligent transportation system. The public transportation syst...

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
IPC IPC(8): G08G1/01
CPCG08G1/0112G08G1/0129G08G1/0137
Inventor 孙仁诚邵峰晶范光鹏隋毅孙颢冬
Owner QINGDAO UNIV
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