Research On Path Planning Of Fire Rescue Robot Based On Building Information Model

In recent years,urban buildings have become taller and taller,covering a wider area,and high-rise fires often occur.The speed of the firefighters' equipment is unable to keep up with the speed of the building's height growth,and high-level fire rescue is becoming more difficult.The scene of the fire is a very complex and unknown danger environment.Once people do not escape from the fire,the rapidly reduced visibility in the fire,as well as the high temperature and thick smoke,will become the roadblocks for the trapped people to escape.At the same time,the trapped person can easily give up the escape under the extremely tense psychological situation.Therefore,how to quickly reach the trapped person's position in a complex fire scene and bring it out of the fire is a serious challenge.In this paper,the key problems of robot path planning in fire environment are studied.In order to convert the relevant map information of the robot's surrounding environment into an abstract space that can be processed and recognized by the computer,the construction method of the environmental model is analyzed,and the environmental space model is constructed by the grid method.Considering speeding up the path search and quickly and accurately locating the trapped person,building information modeling(Building Information Modeling,BIM)is introduced to assist the robot in drawing,combined with the building material information in BIM,to predict the possibility of collapse at high temperatures Part,set this part as an obstacle in advance.The fire rescue robot can obtain the global map of the building interior before entering the fire site.The choice of path optimization algorithm is a very important part of path planning.This article discusses the relatively mature path planning algorithm,and discusses the theory of ant colony algorithm,analyzes the problems of large indoor obstacles,dense distribution,and rescue time is pressing.In view of the above situation,an improved algorithm is proposed,which uses the obstacle pretreatment method to obtain a feasible path,so that the algorithm can search around this path to avoid most obstacles in a stable manner at the early stage of operation;the new pheromone is used Update the rules,extract the shortest path in the current iteration,and compare it with the shortest path in the previous iteration,node by node,based on the node's "contribution",greatly improve the information of the nodes that exist in the two shortest paths at the same time Element content.And dynamically adjust the heuristic function according to the distribution of obstacles around the node,and adjust the pheromone volatilization rate according to whether there are extra turns in the path.It makes the algorithm avoid the dense area of obstacles while searching the optimal path quickly.For the same problem,the improved algorithm is compared with other algorithms of the same type.The simulation results show that the algorithm can find the optimal path stably and quickly.Finally,the types of escaped personnel are analyzed,and different escape methods and paths are selected for different types of escaped personnel.At the same time,for the unknown obstacles faced by the robot in the rescue process,this paper also studies the dynamic window strategy,continuously updates the environmental information through the setting of sub-goals,and converts the global path into multiple partial paths to complete.After the robot enters the fire site,the local path planning is combined with the "guide" of the global path,which makes the rescue more "directional".Finally,combined with a specific scenario,an environmental model is built with the help of BIM,and the fire dynamics simulation software(Fire Dynamics Simulator,FDS)simulates the temperature and CO concentration changes in key indoor areas.The robot uses the environmental model and FDS simulation data.Choose a safe and fast escape path to lead trapped people to escape from the fire scene.