Road network traffic time sequence characteristic data quality diagnosis and restoration method

Abstract

The invention provides a road network traffic time sequence characteristic data quality diagnosis and restoration method which comprises the following steps: S1, obtaining a regional road network topological structure, and constructing a traffic data prediction model; S2, acquiring historical traffic data, training a model and checking the prediction precision of the model; S3, carrying out the real-time detection and repair of the abnormal conditions of the data according to the data collected by a traffic detector and the traffic data after fusion production; and S4, performing incremental training on the prediction model so that the prediction precision of the model can be ensured. The method has the beneficial effects that the current day data is predicted based on the historical data,the mode of the road network data comprehensive production data is combined, the quality of the data is comprehensively judged, and finally reasonable and high-relevance high-quality data is continuously outputted.

Description

technical field [0001] The invention belongs to the field of traffic information processing, and relates to a method for diagnosing and repairing the quality of time-series feature data of road network traffic based on a graph neural network deep learning method. Background technique [0002] With the development of transportation technology, the use of various algorithms in traffic control has become more and more extensive, making traffic control intelligent and automated to a large extent. [0003] However, many algorithms can only be used in laboratories or certain pilot areas, and cannot be fully automated in large areas. The fundamental reason is that many algorithms such as KNN, CNN, and LSTN have very high data requirements. High, unqualified input data will often directly cause unqualified algorithm output results. In reality, the application of many traffic algorithms is often limited due to the fact that the detector cannot be guaranteed or the data produced does...

AI Technical Summary

Problems solved by technology

[0003] However, many algorithms can only be used in laboratories or certain pilot areas, and cannot be fully automated in large areas. The fundamental reason is that many algorithms such as KNN, CNN, and LSTN have very high data requirements. High, unqualified input data often directly cause unqualified algorithm output results

In reality, the application of many traffic algorithms is often limited due to the fact that the detector cannot be guaranteed or the data produced does not meet the actual situation.

[0004] In the existing research, people focus on the quality of the data, 1. The focus is mainly on the part of the missing data and ignore the quality inspection of the produced data; 2. For the missing part of the data, researchers usually use the method of patching 3. For the data of the entire road network, researchers usually only pay attention to the data quality feedback of a certain intersection / segment / lane, and lack of adjacent or similar Comprehensive quality comparison of intersection / road section / lane data

Images