Unlock instant, AI-driven research and patent intelligence for your innovation.

A Filling Method for Missing Data in Urban Traffic Based on Tensor Decomposition

A technology for missing data and urban traffic, applied in the field of intelligent transportation, which can solve problems such as complex calculations, poor interpretation, and unfavorable understanding models

Active Publication Date: 2020-10-30
SUN YAT SEN UNIV
View PDF9 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although this type of algorithm has a good filling effect, it also has the disadvantages of complex calculation, many filling model parameters, and poor interpretation, which is not conducive to other users' understanding of the model and reproduction of the model, which also limits the promotion of related applications to a certain extent.

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
  • A Filling Method for Missing Data in Urban Traffic Based on Tensor Decomposition
  • A Filling Method for Missing Data in Urban Traffic Based on Tensor Decomposition
  • A Filling Method for Missing Data in Urban Traffic Based on Tensor Decomposition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0062] Such as figure 1 Shown, the method provided by the invention comprises the following steps:

[0063] 1. Build tensors

[0064] The vehicle speed data collected by the sensor, as shown in Table 1 (the time period refers to the division of the time of the day at intervals of 10 minutes), is visualized for data visualization, and the change rule of its time and space characteristics is preliminarily explored. From Figure 2-Figure 4 According to the observation results in , the average vehicle speed has obvious changes in different road sections, different dates and different time periods: ① There is a large difference between the average speed of the congestion-prone road section and the congestion-prone road section; The average vehicle speed varies greatly; ③ the average vehicle speed of the morning peak, evening peak and flat peak varies greatly; ④ the change of the average vehicle speed has no obvious periodicity. Therefore, according to the results of the prelimina...

Embodiment 2

[0093] In order to test the missing data filling effect of the method proposed by the present invention, 214 road sections in Guangzhou were selected, and 1,855,589 pieces of vehicle speed detection data from August 1, 2016 to September 30, 2016 (61 days in total) were tested, and the following experiments were set Scenario: ①Test the filling effect of scattered data missing and continuous data missing under different data missing rates, and compare the effect with the commonly used neural network filling method; ②Test the interpretability of the filling model, that is, the spatiotemporal variation pattern of vehicle speed data mining effect.

[0094] The mean absolute error MAE is introduced as the filling effect evaluation index, and the calculation method is shown in formula (6).

[0095]

[0096] Among them, N is the number of filled data, y i is the padding value, z i is the real value.

[0097] The error calculation result is as Figure 9-Figure 10 As shown, it ca...

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 relates to an urban traffic missing data filling method based on tensor decomposition. The method includes the following steps that 1, the tensor, based on the road section, the date andthe time dimension, of urban traffic data is constructed; 2, missing data is prefilled, and initialization of the missing data is completed; 3, the missing data obtained through pre-filling is subjected to truncated singular value decomposition, and the left singular vectors of the road section, date and time dimension of the missing data are obtained through mining; 4, by using the left singularvectors of the road section, the date and the time dimension, the core tensor is obtained through calculation; 5, a missing data filling model is constructed, the left singular vectors of the road section, the date and the time dimension and the core tensor are input so as to train the missing data filling model, through the combination of the optimization algorithm, the missing data filling model is constantly optimized, and after optimization, filling of the missing data is achieved through the missing data filling model.

Description

technical field [0001] The invention relates to the technical field of intelligent transportation, and more specifically, to a method for filling missing data in urban traffic based on tensor decomposition. Background technique [0002] In the era of big data, data-driven analysis methods have brought countless dividends to people. Taking the transportation industry as an example, by collecting and analyzing vehicle speed data, city managers can grasp the operation status of road sections. However, in the process of data collection, the inevitable problem is the lack of data, which directly hinders the city managers from grasping the operation of the road section, and thus cannot make effective management decisions. But fortunately, urban traffic data often have strong spatio-temporal regularities, such as morning and evening peaks often appear on weekdays. Therefore, certain methods can be used to mine laws from historical data to fill in missing data. [0003] How to fu...

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 Patents(China)
IPC IPC(8): G06F16/21G06Q10/06G06Q50/30
CPCG06Q10/067G06Q50/30G06F16/211
Inventor 何兆成钟嘉明
Owner SUN YAT SEN UNIV