A High-Accuracy 3-D Ultrawideband Positioning Approach Using Clustering Optimization Algorithm and Error-Correction Model

The high-accuracy positioning technology has attracted significant attention in 3-dimensional (3-D) indoor environment. Because of the unknown disturbance and random error, the positioning precision of UWB is severely compromised by the complex indoor environment. In this research, a high-accuracy 3-D UWB positioning approach is developed based on the clustering optimization algorithm and error-correction model. Using the particle swarm optimization (PSO) K-means clustering algorithm, the positioning space is divided into different clustering zones. Then, the parameters (predicted error P, process noise Q and measurement noise R) of Kalman Filter (KF) function are optimized by the PSO algorithm with linear decreasing inertia weight strategy for different clustering zones. In addition, the error-correction model is established to reduce the random error. According to the results of indoor positioning experiments, the maximum mean error in axis of the developed approach for experiment 1/2 is 7.89/8.12 cm, while that of the traditional positioning method is 11.76/8.31 cm. The root mean square error of the developed approach in experiment 1/2 is 15.55/14.69 cm, where those of the traditional approach is 21.28/16.08 cm. The mean absolute error of the developed approach in experiment 1/2 is 6.86/6.02 cm, where those of the traditional positioning approach is 9.96/6.26 cm. Therefore, the developed approach can be used to improve the 3-D UWB positioning accuracy, and it is valuable in the smart factory, underground space, and logistics storage in indoor environment.