Research On Automatic Marking Technology Of New Roads

With the government paying more attention to the development of the transportation industry,the number of highway construction projects in China is increasing,the construction process is becoming increasingly complicated,and the related costs are also becoming increasingly high.Meanwhile the marking and construction of the road surface are also undergoing greater construction pressure,due to the relatively low level of intelligence of China’s road surface marking facilities at present.Therefore,it is urgent to enhance the degree of automation and intelligence of road marking machine.This thesis mainly focuses on the planning research of the path problem of the automatic road marking machine when the newly built road is marking the new line.In this paper,the A* Algorithm is used as the basic algorithm to study the route planning of the scriber.Firstly,the environment is modeled,and the environment model of the road which can be identified by the search algorithm is established by using the correlation modeling method.In order to provide the working environment for the following algorithms,this paper studies and analyzes several common modeling methods,and considers that A* Algorithm itself is a search algorithm based on grid traversal,and the actual working environment of the SCRIBER does not need to consider the problem of obstacle avoidance.After comprehensive analysis,this paper adopts the grid method to model the newly built road environment,the grain size of the grid and the Motion Direction of the scriber on the grid are defined.Then,on the basis of the established model,A* Algorithm is used to trace.Because A*Algorithm is only suitable for solving the optimal path between two points,it can not meet the requirements of the actual work of the SCRIBER,this paper proposes to improve A*Algorithm.Under the Global Open List,several sub-OPEN lists are set up to store each marker node separately,and the optimal path between multiple points is solved by nested circular call.In this paper,the path length is used as the benchmark of multi-point path planning,according to the marking instructions given by the remote terminal（edge line,lane line,Zebra Line,etc.）.The road automatic marking machine first models the environment by grid method,then through the received marking information,the improved algorithm is used for tracking,the improved A* Algorithm is simulated in Matlab.Due to the complexity of environment modeling and path-following Algorithm,and the large amount of computation and long time consuming,it can not meet the real-time requirement of scribing machine when scribing.Combined with the parallel processing of FPGA,A* Algorithm is transplanted to FPGA,which lays a foundation for fast and real-time path planning of Scriber.The whole FPGA structure of A* Algorithm is desinged.Thank to Verilog language,the function of each module is designed with specific code,and the test file is written on Quartus II 13.1 software,and the MODELSIM simulation software is called to simulate the logic time sequence of each program implemented by the A* Algorithm.The logic timing function of each Verilog program is verified,and the hardware design of the algorithm is completed.Finally,the improved algorithm is applied to the ROS intelligent mobile system,and the feasibility of the algorithm is proved in practice.At the same time,a road marking test system is provided,and some common new marking tests are carried out.The experimental results show that the road marking machine can trace according to the given node,thus realizing the new marking test,and laying a foundation for the road marking machine self-marking later.