| With the development of social economy,indoor buildings are becoming larger and more complicated.Large shopping malls,hospitals,schools and other places have emerged one after another to provide people with rich and colorful functions.However,complex indoor buildings also bring great challenges to building fire safety management and evacuation in emergencies.Indoor personnel cannot choose the correct evacuation path at the first time,which is easy to cause evacuation congestion.High-precision positioning of personnel in densely populated places,obtaining the location information of personnel,conducting indoor traffic risk management and control,and positioning and evacuation guidance of indoor personnel in emergency situations will provide a strong guarantee for the safety of crowded places and the safety of life and property of indoor personnel.The existing positioning and path planning methods have the problems of poor positioning accuracy,slow positioning speed,slow path planning time and unreasonable planning path in densely populated places.However,the large flow of people in densely populated places and the complex indoor environment require relevant technologies to have the characteristics of hidden positioning,rapid positioning,accurate positioning,reasonable path and rapid planning,which urgently need the support of relevant efficient algorithms.This thesis aims to realize high-precision indoor personnel positioning,realize personnel data visualization,and carry out regional personnel monitoring and evacuation guidance,based on the project of improving independent innovation and social service ability of the safety discipline group of China University of Mining and Technology-Rapid Control of Fire Smoke and Personnel Evacuation Technology in Underground Space,with the help of Wi-Fi indoor positioning technology and A* path planning algorithm,personnel data visualization and evacuation path guidance in densely populated places are realized.The main work and contributions of this thesis are as follows :(1)The attenuation and fluctuation characteristics of Wi-Fi signals propagating in the horizontal and vertical directions are obtained.Aiming at the fluctuation of Wi-Fi signal,the filtering algorithm is analyzed and compared,and the Gaussian filtering is selected as the signal processing algorithm.The horizontal and vertical attenuation characteristics of Wi-Fi signal in indoor environment are studied and analyzed,and the horizontal and vertical attenuation curves and fitting parameters of Wi-Fi signal are obtained.(2)The off-line fingerprint database construction method based on attenuation characteristics and SL-EWKNN plane positioning algorithm are proposed.Using the attenuation characteristics of Wi-Fi signals in the horizontal direction,the offline fingerprint database construction method is proposed based on the attenuation formula,which reduces the early workload of the location fingerprint method while ensuring the accuracy of positioning.The indoor movement characteristics of personnel are studied,the probability function of personnel pace is proposed,and the personnel plane positioning algorithm based on pace reference(SL-EWKNN algorithm)is proposed,which improves the plane positioning accuracy and verifies it by field experiments.(3)The SAC-PWKNN stereo positioning algorithm is proposed.Using the horizontal and vertical attenuation characteristics of Wi-Fi signals,the attenuation offline fingerprint database is established by using Origin software to fit the attenuation curve.The hybrid offline fingerprint database is established by combining the position fingerprint method and Pearson correlation coefficient.The offline fingerprint database is processed by floor K-means clustering,and the offline fingerprint database is integrated.The SAC-PWKNN stereo positioning algorithm is proposed,and the stereo position information of indoor personnel is obtained and verified by field test.(4)The square fusion de-redundancy A* algorithm is proposed.The environment modeling and information storage methods of A* algorithm are compared and analyzed.Aiming at the problem of large storage capacity and slow operation of high-precision grid information,The grid map construction method of grid fusion is proposed,which greatly reduces the system information storage capacity and operation time,and improves the short-term work efficiency of the algorithm.The planned path is simplified,and the redundant node removal method is used to shorten the distance of path planning,which makes the path planning result more intuitive and convenient for evacuation guidance signal output.(5)The square fusion de-redundancy A* algorithm for safe evacuation path decision is constructed.Based on the Wi-Fi positioning results,the indoor personnel distribution heat map is generated by MATLAB software to realize the visualization of personnel data.With the help of the grid construction method of personnel positioning data and grid fusion,the indoor personnel clustering method based on grid map is proposed to classify indoor personnel reasonably and determine the evacuation starting point.Based on the heuristic function of the improved A* algorithm based on personnel distribution,the personnel distribution variable and the evacuation exit variable are introduced,and the square fusion de-redundancy A* algorithm for safe evacuation path decision is proposed to avoid the evacuation path pointing to the direction of personnel congestion during the evacuation process.(6)The passive positioning and safe evacuation path decision method and system are constructed.The personnel positioning method and path planning algorithm are integrated and analyzed.The field experiment of the system is carried out.The Pathfinder software is used to simulate the proposed decision-making method.At the same time,the influence of personnel compliance on evacuation efficiency is studied to verify and analyze the practical application of the decision-making method and system.This thesis includes 92 figures,16 tables and 90 references... |
PDF Full Text Request