室内环境下基于图优化的UWB定位方法

Ultra wide-band positioning method based on general graph optimization in indoor environment

针对室内三维环境中对定位结果准确性的要求,提出了一种基于图优化的室内超宽带(UWB)定位算法。首先,利用最小二乘法对UWB 技术测量的距离进行解算,得到第一次定位初值。然后,将第一次定位初值代入泰勒级数展开算法,得到更可靠的第二次定位初值。最后,建立包括定位残差和位移残差两项的目标函数,通过基于图优化的算法,得到目标函数的最小值,计算出优化之后的最终定位结果。通过样机实验验证定位精度,实验结果表明平均水平定位误差小于10cm,同时平均高度定位误差也比相同实验条件下使用Chan算法和改进的卡尔曼滤波方法分别低了77%和61%。

An indoor positioning algorithm based on graph optimization is proposed to meet the requirements for the accuracy of positioning results in three-dimensional environments. First, the weighted least squares method is used to calculate the first initial value using the distance measured by Ultra Wide-band (UWB) technology. Then, the first initial value is substituted into the Taylor series expansion algorithm to obtain a more reliable positioning initial value. Finally, an objective function including positioning residual and smooth residual is established. The minimum value of the objective function is obtained by the proposed algorithm based on graph optimization, and the final positioning result after optimization is calculated. The prototype experiments are carried out to verify the positioning accuracy of the algorithm, which show that the average horizontal positioning error of the algorithm is less than 10 cm, and the average altitudinal positioning error is 77% lower than that of Chan algorithm and 61% lower than that of an improved KF algorithm under the same experimental conditions.