基于的贝叶斯理论的多系统定位融合算法

针对传统多系统融合定位中协作性较差,自适应性不足的问题。为了多系统定位达到更好的效果,提高多系统协同定位算法中信息融合的高效性、场景间切换的适应性,本文对传统的多系统定位融合算法进行了改进。该算法采用贝叶斯理论,多系统观测数据融合输入,建立贝叶斯概率观测模型,对多系统间定位信息直接交互,通过扩展卡尔曼滤波理论估计定位信息。在此基础上,利用各系统滤波新息和方差对场景间切换时系统概率进行实时更新,将估计结果以系统概率加权方式融合输出；仿真结果表明,在相同观测条件下,本算法与传统定位算法相比,具有更好的稳定性及自适应性。

Bayesian Multi-System Positioning Fusion Algorithms

Aiming at the problem of poor collaboration and inadequate adaptability in traditional multi-system fusion positioning. In order to achieve better results in multi-system localization, improve the efficiency of information fusion and the adaptability of scene switching in multi-system cooperative localization algorithm, this paper improves the traditional multi-system localization fusion algorithm. The algorithm used Bayesian theory, multi-system observation data fusion input, establishes Bayesian probabilistic observation model, directly interacted with positioning information among multiple systems, and estimated positioning information by extended Kalman filter theory. On this basis, the system probability was updated in real time by using the filtering innovation and variance of each system, and the estimated results are fused and output in the way of system probability weighting. The simulation results show that the algorithm has better stability and adaptability than the traditional location algorithm under the same observation conditions.