Road network matching method for bicycle track data with noise and unknown parameters

Abstract

The invention relates to a road network matching method for bicycle track data with noise and unknown parameters, and the method comprises the steps: collecting a track point sequence of a riding track line of a bicycle and a neighbor road segment set formed by all road segments in a distance threshold S range of the riding track line in a road network; constructing a matching probability network T-R-N between the trajectory and the road network; constructing a random event probability model mapped from a single track point to a road section, and calculating a node event occurrence probability; constructing a road section mapping condition probability model of front and rear track point pairs of the bicycle, and calculating a transition probability between node events; constructing a Markov chain between the bicycle track and the road section, and obtaining an optimal matching result of the track line according to the Markov chain between the bicycle track and the road section with the maximum combination probability. The scheme of the invention is a solution which fully considers the actual noise condition of the current bicycle track data, and can break through the actual problem that the application of the riding track data is restricted by the quality of the shared bicycle data in a real system.

Description

technical field [0001] The invention relates to a bicycle track data road network matching technology, in particular to a road network matching method for bicycle track data under the condition of relatively large noise and unclear error parameters. Background technique [0002] Under the trend of advocating green transportation in the whole society, the rise of shared bicycles not only promotes the development of slow traffic, but also provides a large amount of bicycle riding trajectory data for many different applications in the field of traffic management. Since trajectory data and urban road network data have different sources, building spatial association between the two is an inevitable condition for using trajectory data. Therefore, it is of great significance to accurately match bicycle trajectory data to road network. [0003] In the field of transportation, the current research results of trajectory data matching are mainly aimed at traffic subjects such as automo...

AI Technical Summary

Problems solved by technology

The bicycle trajectory data is affected by the subjective factors of riders, the differences of positioning equipment, and the superimposition of various external interference factors, resulting in more complex and more biased data noise.

In addition, the currently available cycling track data does not come from professional GPS equipment, but from the GPS positioning data of cyclists’ mobile phones collected and stored by the shared bicycle operation platform, which makes the noise characteristics of the positioning data more unclear. Due to the limitation that the "true value" data of the riding process is difficult to obtain, there is no condition for correcting the trajectory data before performing road network matching

In this case, the current vehicle trajectory data matching methods are not applicable, and it is necessary to study the appropriate trajectory data road network matching method according to the characteristics of riding characteristics and trajectory data

Images