A Method of Improving Map Matching Outliers Based on Machine Learning Algorithms

Abstract

A method for improving map matching abnormal points based on machine learning methods, obtaining the projected coordinate point data of one or more taxis as sample data, and then filtering out matching normal and abnormal projected points; for each matching abnormal The projected coordinate points of the outliers are selected from the matching data set to select the candidate data set of abnormal points; the temporal similarity and spatial similarity between each abnormal point and the projected points in the candidate data set of abnormal points are calculated to obtain the spatiotemporal similarity Set, and then calculate the mean value of spatio-temporal similarity according to temporal similarity and spatial similarity; compare the size of spatio-temporal similarity and the mean value of spatio-temporal similarity, obtain the number of spatio-temporal similarity greater than the mean value of spatio-temporal similarity, and combine As the final candidate data set; in the final candidate data set, the improved projection coordinate points are calculated by using the knn algorithm. The invention has the advantages of lower calculation amount and higher accuracy.

Description

technical field [0001] The invention relates to the technical field of map matching, in particular to a method for improving abnormal points in map matching based on machine learning algorithms. Background technique [0002] Map matching is a technology that corrects navigation positioning errors through application software. The map matching algorithm outputs the projection vector of each GPS point, thus projecting the GPS point into the traffic road network. Because the GPS equipment and the sampling interval are too long, the GPS measurement error will cause the timing error of the projection point. At present, although the timing error of the matching result can be improved by ordinary knn, the effect is poor. Contents of the invention [0003] Aiming at the problems existing in the prior art, the purpose of the present invention is to provide a method for improving map matching abnormal points based on machine learning algorithm. On the basis of ordinary knn, by int...

AI Technical Summary

Problems solved by technology

Due to GPS equipment and too long sampling interval will cause GPS measurement error, resulting in timing error of projected points

At present, although the timing error of the matching result can be improved by ordinary knn, the effect is poor

Images