Study Of Boundary Detection And Tracking Algorithm For Contaminated Region Based On Small Unmanned Helicopter

Contaminated plume is one kind of typical manifestations of environment pollutions, for example, air floating particles(puff, PM2.5), radioactive contamination, Chemical gas diffusion, water pollution(oil spills and red tide, etc), which is performed as different concentration and intensity distribution in specified region. At the same time, the boundary of the region will diffuse as well and also change over time drive by the external factors such as wind, water flow, geographical aspect, temperature, humidity changes. So it presents a uncertain form as feather on animal. Therefore, it is very important to identify, estimate or predict this contaminated region as quick as possible which is one key point in environmental monitoring as well as an essential basis of warning and quick decision. In order to reduce the damage caused by the accident, in this paper, a new algorithm that inspecting the environment just by one unmanned aerial vehicle(UAV) is proposed, once the environmental pollution occurs, the boundary of polluted area will be identified and tracked immediately.This paper includes the following four points:Firstly, an oil spills model is proposed and analyzed in detail. It is also simulated on the MATLAB platform.Secondly, since less information will be obtained by one UAV flying through the contaminated area, it is essential to use Spatial interpolation technique to map the whole. Kriging interpolation is adopted in reconstructing the environment map and then explore the effects of the flight trajectory of UAV to Kriging prediction from multiple perspectives. A good searching schema is put forward and proved through simulation.Thirdly, model predictive control(MPC) algorithm is used to make UAV Fly along the pollution boundary. As we know, MPC can handle control problem with constraints and provide a good tracking performance.Finally, with the help of geographic information system(GIS) to evaluate the environment further. GIS is a relatively professional geographic statistics software, the data can be transformed into a visual 3D image, so that the result is more intuitive and clear.