A Decentralized Cooperative Localization Method for Firefighters Based on Inertial Navigation and Ultrawideband Ranging

Location-based service is vital for firefighters. To reduce the deviation of the inertial navigation system (INS) and to suppress the non-line-of-sight (NLOS) error of the ultrawideband (UWB) positioning, a multiperson decentralized cooperative localization method is proposed in this study. Every firefighter has a foot-mounted inertial navigation system, which uses extended kalman filter (EKF) to estimate firefighter’s attitude and position. Furthermore, a group of firefighters construct a decentralized cooperative localization system. To suppress the accumulation error of INS and estimate the optimal position for every firefighter according to the relative ranging measured by UWB, a combined filter which fused particle filter (PF) and EKF is proposed. The cooperative localization method uses firefighters as mobile anchors and sets three fixed anchors. To suppress the NLOS error, the combined filter adaptively adjusts the observation vector according to UWB ranging. The experimental results show, the average and maximum positioning errors of the decentralized cooperative localization method have been suppressed by 60.50% to 61.77% and 59.61% to 65.83% compared with the conventional INS method, respectively. Therefore, the proposed method improves the positioning accuracy and has good robustness to environment.