Gas Leak Source Search Based On Olfactory And Visual Information Fusion And Mobile Robot

Harmful / toxic gas leakage affects the ecological environment,but also brings a lot of inconvenience to human life.Especially for industrial production,the leakage of some chemicals will cause explosion,fire or other accidents.In the destruction of the surrounding environment,but also cause serious economic losses.Therefore,it is of great significance to find and track the plume and leak source localization of gas leakage source.The main work in this paper is the search of gas leakage source,the further analysis of the task environment of the gas leakage based on the summary of past research studies,and the multi-sensor information fusion algorithm is applied to the search of gas leakage source.This paper focuses on the following aspects.First,several common plume models are analysed and compared,but as it is difficult to control the wind speed and the distribution of smell in an experimental environment,the plume model to the real environment was established with CFD software.Indoor ventilation environment plume model is built and gridded with the Gambit software.Besides,the fluid dynamic software,Fluent,is applied to simulate the process of fluid motion.The experimental environment has authenticity in practical application,and it is of practical significance to explore the source of leakage in the environment of smoke.According to the simulation results,the influence of the size of the leakage port,the position of the obstacle and the position of the leakage port on the concentration distribution of the CO gas in the indoor ventilation environment were analyzed.The results showed that the molecular flow of CO gas under a stable diffusion is like a smoke plume,while in the case of obstruction it will be formed along the obstacles to the formation of smoke group.The results of the experiment are conducive to the rapid detection of smoke and the location of suspected leakage sources and can increase search efficiency.Secondly,in the sense of olfactory sensing,the active lumen type laser gas sensing system is adopted.The system is used to obtain the information of the kind and concentration of the gas.In the aspect of the visual sensor,the fisheye lens is used to obtain the panoramic information of the experimental environment.Dynamic model of track robot is analysed.Then,the location of the leakage source is estimated by machine vision,machine smell,and wireless sensor respectively.The DS evidence theory is applied to the estimation of gas leakage source.The simulation results show that the evidence theory can be applied to the determination of gas / odor source.Finally,a new search strategy for leakage source search based on mobile robot is designed,which can complete all the processes of finding,tracking and locating the feathers at one time.The simulation experiment of Matlab shows the effectiveness of the strategy.