Research On Route Planning Of UAV Bridge Disease Detection

The construction of bridges is an important matter related to national economy and people’s livelihood.Regular inspection of bridge diseases is an important guarantee for the long-term normal operation of bridges.Traditional bridge disease detection relies on inspection vehicles,but this method is time-consuming and labor-intensive,and the detection process is also constrained by conditions.This defect is particularly obvious in the detection of large-span bridges.With the development of science and technology,the application of UAVs is becoming more and more extensive.The UAV is light,compact,and easy to operate.It is equipped with a gimbal and high-definition camera to shoot or record bridges,and then transmits the information back to the bridge disease detection system for detection.At the same time,it also has certain defects: operation dead zone,small battery capacity,limited battery life,and signal loss.In order to complete the task,reasonable route planning is very important for UAV flight.This paper takes bridge disease detection as the task background and studies the route planning of the bridge disease detection UAV.The main research work is as follows:(1)Analyze the relevant factors of UAV route planning: environment model construction,bridge model,constraints,detection target points and route planning system,etc.Select the grid method to build the environment model,use the key points of coordinate transformation to simplify the bridge model;analyze the constraints of UAV route planning,including task constraints and self-restrictions;the selection of UAV route planning target points must consider constraints and inspection requirements;The way of expressing the track uses the connection of navigation points and the selection of the route planning algorithm type according to the detection task.(2)Study the route planning algorithm of the UAV under the bridge general inspection task.In order to complete the general inspection task of the bridge,the UAV needs to traverse the bridge and design a full coverage planning process.First: Use the UAV to obtain bridge information and build a bridge model.Second: In the algorithm,the grid full coverage algorithm is selected,and the algorithm process is designed: the bridge environment is first modeled in the grid environment,and then the model is decomposed,and the decomposed subregions are passed through the comb traversal algorithm based on the dichotomy Realize the traversal,and finally connect the traversed areas to get the full coverage of the drone during the bridge survey.(3)When planning the UAV route for special inspection,by comparing several commonly used planning algorithms,the rapid expansion random tree(RRT)algorithm is used for planning.Aiming at the parameter adjustment problem of this algorithm,through experimental research,the selection basis of step size and selection probability in RRT algorithm is summarized.By deleting the redundant nodes in the random tree expansion process,the algorithm is improved and its practicability is increased.Finally,the comparison and analysis of the experimental results verify the value of the algorithm research,which lays the foundation for the route planning in the background of bridge inspection tasks.(4)Taking a PC continuous rigid frame bridge model as an example,under the application background of two kinds of bridge disease detection(conventional census and special detection),a simulation experiment was carried out on the UAV’s route planning.First,the simulation of the full-coverage route planning of the UAV bridge under the census task was done.In order to facilitate the planning and simulation of the RRT algorithm under the special bridge inspection task,a GUI interface is designed on MATLAB,and the route planning under the special bridge inspection is carried out.Through the simulation results,combined with the UAV’s own characteristics to analyze the rationality of the route,it is believed that the planned route can meet the flight requirements of the drone,and an immediate and reasonable drone route planning.