基于场景检测的城市环境GNSS/INS组合定位方法研究

城市复杂环境下,采用GNSS/INS组合定位方法能有效提升GNSS信号部分或全部被遮挡情况下的车辆连续定位能力。GNSS信号受遮挡的程度具有随机性和突变性,给GNSS/INS组合算法的架构设计和参数设置带来了挑战。基于此,设计了一套基于场景检测的GNSS/INS组合定位策略与方法,提出了以多星GNSS观测量的八类特征为输入,利用支持向量机分类思想将定位场景分为室外开阔、室外遮挡和室内三种类型,并建立了与之相适应的组合滤波量测误差估计模型,进而实现了GNSS/INS组合定位。实测数据验证表明,提出的方法能够有效提升车辆组合定位精度。

Research on GNSS/INS Integrated Positioning Method for Urban Environment Based on Context Aware

In complex urban environment,GNSS/INS combined positioning can effectively improve the continuous vehicle positioning service capability when GNSS signals are partially or completely blocked.However the random and abrupt occlusion of GNSS signals brings challenges to the architecture design and parameter setting of the GNSS/INS algorithm.This paper mainly proposes a set of GNSS/INS integration positioning strategy and method based on context aware technology,which takes eight features of multi-satellite GNSS observations as input of support vector machine classifier to divide the positioning environment into three types,open outdoor,occlusion outdoor and indoor contexts,and establishes a corresponding measurement error estimation model of integration filter to achieve GNSS/INS integrated positioning.The verification result shows that the proposed method can effectively improve the vehicle integration positioning accuracy.