A Universal Map Matching Framework Based on Dynamic System and Optimal Control

Abstract

Map matching is the process of matching a time sequence of location coordinate points into a digital road map. Currently there are different map matching algorithms designed for different types of location data with different frequencies, different accuracies, and different types of additional information that can be used to improve the results. However, there is not yet a universal map matching framework that considers all the map matching criteria and covers all these different scenarios. This paper presents a batch map matching framework that works well for high-frequency, high-accuracy location data and low-frequency low-accuracy location data, and it can consider different types of additional data for data fusion. The map matching problem is formulated as an optimal control problem for a discrete-time domain dynamic system, and the problem can be solved using dynamic programming. This framework will greatly benefit the map matching application development, as the core building blocks for the algorithm are the same for all types of data.