基于手机传感器和互补滤波的行人航向解算

目前行人航迹推算逐渐成为室内定位研究与应用的热点。针对利用陀螺仪推算行人航向时存在较大累积误差的问题，本文提出了一种基于智能手机传感器的行人航向解算算法。该算法根据陀螺仪输出的角速度数据与手机传感器参数计算合适的阈值，实时调节PI调节器的误差补偿系数，对预处理后加速度计和磁力计数据解算的航向角进行补偿，并与陀螺仪数据互补滤波融合，得到融合后的航向角。试验基于低成本智能手机，分别在磁场强弱环境下采集手机传感器数据，对比分析本文算法与传统互补滤波算法及九轴数据融合算法在推算行人航向时的精度。试验结果表明，在室内磁干扰较强的环境下，本文算法与传统互补滤波算法、九轴数据融合算法相比定位精度分别提升了68.4%和65.9%，平均航向误差分别减小了3.4°和1.8°，验证了本文算法有较好的抗磁干扰性能，提高了行人航向角解算的可靠性。

Pedestrian determination based on mobile phone sensor and complementary filter

At present, pedestrian dead reckoning(PDR) has gradually become a hot spot in the research and application of indoor location. In view of the large cumulative error of the gyroscope in calculating the pedestrian heading, a pedestrian heading algorithm based on smart phone sensor is proposed. The algorithm adjusts the error compensation coefficient of the PI regulator in real time according to the angular velocity data output by the gyroscope and the parameters of the mobile phone sensor to compensate the heading angle of the preprocessed accelerometer and the magnetometer data. The gyroscope data is filtered by complementary filtering to obtain the combined heading angle.Based on low cost smart phone, we collect smart phone sensor data under the strong and weak environment of magnetic field, and compare the accuracy of the algorithm with the traditional complementary filtering algorithm and the nine axis data fusion algorithm in calculating pedestrian direction.Experimental results show that compared with the traditional complementary filtering algorithm and the nine-axis data fusion algorithm, the proposed algorithm improves the positioning accuracy by 68.4% and 65.9%, respectively, and the average heading error decreases by 3.4° and 1.8° in the indoor environment of strong magnetic interference,it verifies the proposed algorithm has better anti-magnetic interference performance and improves the reliability of pedestrian heading.