Zero-Velocity Detection Algorithm for Inertial Pedestrian Navigation Based on PSO-DBSCAN Clustering

In view of the problem that the zero-speed interval detection algorithm in pedestrian navigation has poor adaptability and inaccurate detection results when dealing with different motion states, a zero-speed detection algorithm based on particle swarm optimization DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering is proposed. This algorithm utilizes the DBSCAN clustering algorithm to cluster the two-dimensional array composed of the modulus values of acceleration and angular velocity processed by the inertial measurement unit (IMU) data. Meanwhile, the particle swarm algorithm is used to adjust the clustering parameters online to adapt to different motion states and obtain the zero-speed interval. The experimental results show that this algorithm improves the adaptability of zero-speed detection, effectively reduces the positioning error. In the experiment of variable-speed motion, the detection results obtained by this algorithm were applied to the correction of navigation and positioning solution. The final trajectory endpoint error was 2.3133 meters.Compared with the DBSCAN clustering algorithm with fixed parameters and the generalized likelihood ratio test algorithm, the results improved by 25.75% and 36.88% respectively.