Research On Mission Planning For Mobile Visual Measurement Robotics

With the continuous development of the national defense industry and the continuous improvement of the level of science and technology,the intelligent assembly technology for large aerospace products has been developed,mobile measurement robots play a key role in the intelligent assembly process.Visual measurement and path planning are two important research directions in mobile measurement robot mission planning system.For the object to be assembled,we need to determine the optimal observation position of the robot at which the measurement produces the least error,allowing the robot to guide the assembly process better.In this paper,a method of determining the optimal position of the robot is proposed by using the monocular vision which is used as the measurement scheme.The purpose of path planning is generally to determine the robot's shortest path in the working environment.We improve the A star algorithm,and a path planning method based on improved A star algorithm is proposed.Three different map construction methods are proposed based on the grid method,in these method,we consider factors such as marking point,measured object and observation position.The following research work is carried out in this paper:Firstly,a monocular vision measurement model and a best observation position determination model are established.According to the principle of monocular vision imaging to determine the measurement scheme,based on this,two-dimensional and three-dimensional visual measurement model are established.The optimal observation position determination model is established for different working conditions,so we can determine the best observation position.Secondly,the path planning method of mobile vision measurement robot is studied.The implementation process of A star algorithm is described.We do simulation of path planning method based on A star algorithm in raster map.When we use traditional A star algorithm,there maybe some collision problems,we improve the A star algorithm,so that it can avoid collision problems.Thirdly,the method of constructing the map of mobile vision measurement robot is studied.Different from the traditional map construction,we add mark points and the measured objects to this map.Based on this,three different map construction methods are simulated,the first method is considering mark points separately,the second is considering mark points and measured objects,the third is considering mark points,measured objects and best observation position.And analyze the advantages and applications of different methods.Finally,verify the best observation position measurement model and path planning algorithm through experiments,and at the same time verify the advantages of mobile measurement.We build the appropriate experimental environment for different experiments,and use robot's CCD camera and the global camera to measure objects.Then,analyze measurement error.According to the conclusions drawn to verify the theory proposed in this paper.