Research On The Method Of Detection And Location For Indoor Fire Warning Point Based On Mobile Robot

Fire detection is one of the most important methods to prevent the fire.At present,the commonly used fire detection devices are fixed point detection,so the detection range and detection capabilities have certain limitations.Aiming at the problem,this paper introduces the mobile robot into the fire detection and location,and studies the method of intelligent detection and location.The main contents of this thesis are as follows:First of all,a dual sensor smoke measurement method is proposed and the K/N rule is used to fuse the sensor data based on the analysis of smoke diffusion model and the method of simulating the human nose to the fire alarm point;further,according to imitate the human nasal search process alarm point,the paper puts forward the humanoid nose fire point detection mechanism and positioning method and gives the corresponding work flow.Secondly,the paper designs a humanoid nose mechanism of a multiposition based on the working mechanism of humanoid nose,the mechanism includes a double nose pipeline freedom,torsional freedom,pitching degrees of freedom;at the same time,the theoretical calculation and analysis of the design of the nasal ANSYS fluid based structure,and gives the sensor placement criteria.Thirdly,considering the uncertainty of sensor detection and localization,the robot motion planning method based on the idea of inverse tracking is studied.The establishment of smoke diffusion model,which by the extended Kalman filter to realize motion parameter error of the suspected fire alarm point minimum estimation;Inverse tracking based on the principle of the motion parameters of the trajectory of the robot.Mobile robot on suspected fire point positioning process trajectory and azimuth deviation through simulation experiments,verified the extended Kalman filter method based on the idea of inverse tracking feasibility of suspected fire point positioning.Then,the dynamic path planning method for the robot to locate the fire point in the known indoor environment is studied.A path planning strategy based on gridmap and fuzzy logic is proposed.First,considering of the environment of the obstacles,the fuzzy inference programming of robot path is established based on the geometric relation of obstacles,suspected fire alarm point and robot;Then,considering the existence of smoke gathered dead zone environment,the classification positioning strategy of the suspected fire alarm points,dead areas and mutiple dead areas is put forward.Fuzzy reasoning the rules of mobile robot path is established.Much simulation of path planning of the robot is conducted for a variety of environments.The simulation results show that the robot can successfully achieve the positioning of the suspected fire point if there exists an obstacle.When there is smoke dead angle,the robot takes a long time to locate the fire alarm,but it can also be positioned to the fire point finally.At last,a prototype of a multi position nose like human nose detection device is developed,and the device is installed on the mobile robot 2.The experimental results show that the smoke concentration 2400 ppm as smoke source,the maximum capacity of the fire point positioning distance robot 7.5m;When initial azimuth angle and the fire point is not same,location is approximately 1.8 s maximum time difference;No matter whether there are obstacles,the robot can navigate to the fire point.It is limited to the test condition,and the location of the fire alarm has not been carried out at the place which do not have smoke.