Indoor fire map fusion using extended particle filter for UWB/LiDAR

Aiming at the fact that it is difficult for the command center to get the indoor fire scene map instantly when firefighters rescue the fire scene, a real-time measurement and map fusion method using extended particle filter (EPF) fusion of ultra-wideband (UWB) and Light detection and ranging (LiDAR) is proposed. The UWB and LiDAR observations are used to update the weights of the weighted particles, and the optimal estimates and variances are passed to the extended Kalman filter (EKF) process so that the filtered particles converge to the posterior probability distribution, thus improving the filtering accuracy and mitigating the problem of particle degradation that occurs when traditional particle filter (PF) is performed for data fusion. By collecting UWB and LiDAR data, the extended Kalman prediction is performed on the initial particle, and the particle weight is updated according to the position distance between the UWB/LiDAR data and the particle, the particles are re-obtained using resampling to output the optimal pose map of the iteratively fused UWB and LiDAR. Experimental simulation results show that compared with the PF algorithm, the method proposed in this paper reduces the error of the constructed maps by 54.5%, and the map accuracy is significantly improved.