Research On Key Technologies Of Visual Perception For Integration Of Loading And Tool Setting For Shaft Parts

Loading and tool setting are two important links before machining,the core of both lies in determining the relative position.Due to the difference in technical requirements,the automation studies of both are independent.However,the automation of a single process means that workers are still involved in the process preparation.To achieve a high level of automation,two links must be organically combined.The key technologies of visual perception for integration of loading and tool setting are studied by using industrial robots and several visual schemes.The research object of the scheme is shaft workpiece and lathe.The two links of loading and tool setting matching are optimized in the scheme.The verification experiment is also designed.The main contents are as following:In the loading,there are special requirements for clamping position of workpiece and depth into chuck.Therefore,the paper proposes a position-controlled loading method which can arbitrarily specify the clamping position of workpiece and locked position by chuck has high flexibility.In the process of loading,a method based on minimization of contour estimation function is also used to obtain the axis,and the feasibility of the method is effectively verified by experiments.In the tool setting,due to the poor absolute positioning accuracy of the industrial robot,the camera cannot reach the ideal position accurately,which makes pixel equivalents vary.Therefore,this paper proposes a dynamic calculation method of pixel equivalent for tool setting,and gives some measures to reduce errors of tool setting.The method also studies the ellipse of the end face of shaft workpiece from a general perspective.To realize the integration of loading and tool setting,the key is to solve the change of vision and accuracy.These changes can be quantified into corresponding constraints.According to the constraints,this paper completes the software and hardware design and related calibration of the scheme.The reprojection error of camera calibration is less than 0.5 pixels.The binocular ranging accuracy is 0.664 mm.In hand eye calibration,a method based on original consistency proposed in this paper is used to quickly screen the ideal hand eye matrix,and the hand eye calibration error is less than2 mm.Finally,the experiment is carried out according to the integrated scheme of loading and tool setting.The standard deviation of the workpiece axial direction is0.058 mm,and the standard deviation of the workpiece radial direction is 0.033 mm,which meets the requirements of semi-finishing machining.