V-RTK: A velocity-constrained RTK algorithm to improve position accuracy of low-cost receiver in urban environments

Abstract

Listen

Translate article

In harsh signal environments, such as urban canyon, overpass and so on, global navigation satellite system (GNSS) navigation will be influenced by insufficient satellites, signal degradation, and bad quality which lead to the lack of redundancy and re-initialization of the ambiguities, especially for the low cost real-time kinematic (RTK) equipment. To address these issues, an velocity-constrained RTK method (V-RTK) for low-cost receiver is proposed. Instead of being initialized with the standard point positioning (SPP) result, the predicted position parameters are calculated with the velocity obtained from GNSS observation or other sensors, so that the historical filter message can be delivered to the next epoch both from the ambiguity domain and the position domain to maintain the accuracy of historical epochs. The performance of the proposed method has been assessed with two kinematic vehicular datasets collected in an urban environment, which lasts about 3 h. The results manifest that the positioning trajectory is smoother and location jumps are significantly reduced after applying the velocity constraint. And the accuracy of each dimension is improved by 10%∼40%. Besides, an in-field experiment is conducted to compare the proposed method with the PVA method, which shows a better performance in our proposed method for making full use of observations.