Study On An Efficient Pose Measurement Method For End Effector Of Mobile Manipulator Based On Binocular Stereovision

The mobile manipulators,with high flexibility and large working space,provide an ideal alternative to the high-quality manufacturing of large and complex components,such as wind blades,high-speed rail carbody,aerostructures.The high-precision acquisition of the pose of its end effector determines the processing quality,while the high-precision and wide-range measurement of the end effector remains challenging.However,among the available methods of large-scale spatial positioning,the single-site measurement has accumulated error and can be easily blocked by the work-piece;the multibase station measurement faces challenges of high cost and cumbersome deployment of station.With their respective shortcomings,neither of them is suitable for positioning of mobile manipulators.Hence this paper,based on binocular stereovision,proposes a low-cost large-scale spatial positioning method without laser block and accumulative error.This paper also explores its spatial positioning algorithm and targets' layout,including the following aspects:A large-scale positioning method for end effector of mobile manipulator based on binocular stereovision is proposed.This method is implemented through offline calibration and online measurement: distributed referential targets covering the whole measurement site,integrated binocular camera simultaneously measuring end effector and referential targets to obtain the absolute spatial pose of the end effector in a relative positioning manner.Two types of spatial positioning algorithms are discussed: the MLAT algorithm and the camera pose estimation algorithm.This paper also makes the iterative identical point algorithm,one of the latter algorithm,robust.The outliers with gross errors in the measurement are excluded in the robustness analysis.Simulation results show that the improved algorithm can effectively deal with data with gross errors,the robustness is significantly improved compared with the original algorithm,and the positioning accuracy are approximate to that of the original algorithm under ideal conditions.The error propagation model of the iterative identical point algorithm is discussed,then the rotation dilution of precision(RDOP)is used to represent the relation of measurement errors and estimation errors.According to the influencing parameters of RDOP,the relative layout of the targets is optimized,and the rectangular one is recognized as the best among various layouts.Finally,the overall layout of the target is further optimized,and it was determined that 33 cm is the most optimal one.The mobile measurement platform was built where the positioning experiment of the mobile manipulator's end effector and the experiment of 3D scanning of large work-piece were conducted.The results showed that the large-scale positioning method for mobile manipulator based on stereovision has no problems of light occlusion and cumulative error.