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

Data Noise Point Detection Method Based on Average Offset Shift

A technology of average offset and detection method, applied in traffic flow detection, traffic control system of road vehicles, instruments, etc., can solve problems such as poor performance, poor detection, large difference in traffic flow data, etc. , to achieve the effect of promoting more precise processing

Active Publication Date: 2020-10-16
ZHEJIANG GONGSHANG UNIVERSITY
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Clustering: suitable for small differences in data, but large differences in traffic flow data, the effect of cluster analysis is poor
Low-order polynomial sliding fitting method: This method uses statistical methods to judge outliers, but the traffic flow is highly abrupt, and this method cannot detect sudden noise very well
However, these noise-solving methods cannot have good performance against the noise in traffic flow data, which affects the prediction performance of machine learning algorithms on traffic flow.
[0005] According to the analysis of traffic flow data, it is found that traffic flow data is easily affected by the external environment, and traffic control, bad weather, holidays, etc. can easily cause sudden changes in traffic flow data

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
  • Data Noise Point Detection Method Based on Average Offset Shift
  • Data Noise Point Detection Method Based on Average Offset Shift
  • Data Noise Point Detection Method Based on Average Offset Shift

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0010] The present invention will be further described below in conjunction with accompanying drawing

[0011] Such as figure 1 Shown, the present invention comprises the following steps:

[0012] Step 1: Assume that the real data of traffic flow is T, and the period is N, T i Indicates the traffic flow data corresponding to the i-th moment in a certain period. Assuming that there are M periods in total, the historical value of the traffic flow at this moment is H={T i1 , T i2 ,...,T ij}, j=1,...M. T ij Indicates the i-th traffic flow data in the j-th cycle. First calculate the average value at a certain time in the cycle, and then calculate the offset D at that time i As follows, i=1,...N.

[0013]

[0014] Step 2: sort the offset D and record it as D′, and there are N pieces of single-period traffic flow data. Filter out the smaller and larger 1 / 5 data, and average the remaining 3 / 5 data. In this way, the data with too large or too small offset can be filtered ...

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 discloses a traffic flow data noise point detection method based on average offset translation. Noise judgment is performed by mainly combining historical data and adjacent data so as tofacilitate subsequence corresponding processing, wherein the historical data refers to an average value of traffic flow data at a certain moment in the last several periods, and the adjacent data refers to traffic flow data around a point to be judged. The offset is calculated by combining the historical data and the adjacent data, then the historical data is fit by using the data within the period and the offset, and finally whether the data is noise data or not is judged according to the deviation between the fit data and the historical data. The data noise point detection method can position the noise existing in the traffic flow data more accurately and promote more accurate processing for intelligent transportation.

Description

technical field [0001] The invention belongs to the technical field of traffic flow data noise processing, and in particular relates to a traffic data noise point detection method based on average offset translation. Background technique [0002] With the maturity of computer hardware and software, artificial intelligence technology has gradually attracted social attention. The application of artificial intelligence in the field of transportation, that is, intelligent transportation, has a vital impact on people's lives. Traffic intelligence solves many difficult problems in people's travel, such as traffic flow forecasting can effectively avoid traffic jams. Through the deep learning method, the historical traffic flow change law is learned, and the weight value learned is used to predict the traffic flow situation at a certain moment in the future, which is convenient for early decision-making to avoid traffic congestion. [0003] Deep learning has good performance for t...

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): G08G1/01G08G1/065
CPCG08G1/0129G08G1/065
Inventor 杨柏林田彦林贤煊孙书林张凯丽
Owner ZHEJIANG GONGSHANG UNIVERSITY