Research On Three-dimensional Navigation Method For Indoor Multi-layer Environment

The key to indoor navigation is multi-layer indoor positioning technology and multi-layer path planning.The positioning algorithm based on the Received Signal Strength Indication(RSSI)of the wireless network(WiFi)makes full use of the widely deployed wireless network infrastructure and becomes the starting point of the positioning technology in this thesis.At the same time,it is another research focus to propose a universal multi-layer path planning scheme.The thesis conducts a comprehensive research on the multi-layer navigation technology of RSSI based on WiFi.(1)Aiming at the long fingerprint collection time in the offline stage of positioning,a fingerprint library construction method based on improved kriging interpolation algorithm which is based on shuffled frog leaping algorithm is proposed.This thesis collects half of the sampling points and interpolates to build a complete location fingerprint library,and proves that the shuffled frog leaping algorithm is better than the least square method in optimizing the Kriging mutation function.On the premise of ensuring positioning accuracy,reduce the time to build a location fingerprint library by 50%.(2)Aiming at the indoor positioning in the multiple floors' s environment,an three-dimensional indoor positioning scheme based on affine clustering and trajectory tracking is proposed.In this thesis,affinity propagation clustering algorithm is applied to the offline process of multi-layer indoor positioning.Online positioning uses the coarse positioning based on class matching and fine positioning based on WKNN algorithm,and then the positioning trajectory is filtered using Kalman to obtain a trajectory closer to the real position.In the floor location process,a floor location algorithm that matches the cluster leader of each fingerprint library is proposed and verified.Finally,it is proved that after the introduction of affine propagation clustering algorithm,the average positioning accuracy is close,and the algorithm reduces the calculation amount by 50%,which greatly improves the portability of the algorithm to the mobile terminal.(3)Aiming at the multi-layer path planning,this thesis proposes a path planning algorithm based on multi-layer indoor environment.Compared with the method of dividing the waypoints on the map,a grid method for quickly constructing waypoints is proposed.At the same time,according to the "road network" principle,the cost between the connection points of each floor is stored in advance,a multi-floor path planning algorithm is proposed,and the path planning experiment is performed on the algorithm,indicating that the algorithm can effectively plan the multi-floor path.Finally,design a navigation interface,based on RSSI information collected on the real path without location information,the tests use the multi-layer indoor positioning algorithm and multi-layer path planning algorithm proposed in this thesis to perform positioning and path planning,and display,the results show that both can well applied to the three-dimensional navigation method in indoor multi-layer environment.