Research On Error Compensation Technology Based On A3 Parallel Spindle Head

In the thesis,the key technology of A3 parallel spindle head error compensation is studied in close combination with the national fund.Taking a 3-RPS parallel spindle head with one translation and two rotations as the research object.The optimization of the measurement configuration,the error identification method,the error compensation strategy,and the calibration method during the kinematics calibration process of this mechanism are studied.The main research results are as follows:Based on the kinematics analysis and geometric error modeling of the parallel power head,the sensitivity of the geometric error source of the workspace and the condition numbers of the identification matrix are used as indicators to derive a measurement configuration optimization algorithm based on an improved particle swarm optimization algorithm.Simulation results show that this algorithm can optimize the configuration trajectory and measurement points,and improve the identification efficiency.The laser tracker is used as a measurement tool to establish the mapping relationship between the geometric error source of the mechanism and the distance measurement error.Based on the pose error,a linear regression model for the identification of general geometric error parameters is derived.An improved PSO-Liu geometric error identification method for parallel spindle head calibration was used for identification,and the validity of the method was verified using the pose error residual ratio as an evaluation index.Taking the A3 parallel spindle head prototype independently developed by Tianjin University as the object,the coarse and fine calibration geometric error compensation strategy is adopted.In the initial configuration,the A3 parallel power head is subjected to branch chain drive shaft pitch error compensation and zero point calibration.In the working space,a laser tracker is used to identify and compensate the position error of the end of the mechanism.Experimental results show that the identification algorithm used is robust,efficient,and effective in error compensation strategies.The research results are applied to the kinematic calibration of hybrid machine tools with parallel spindle heads as the core.