Parameter Identification Method of 6-DOF Radiotherapy Beds Based on Combined Genetic Algorithm and Mini-max Optimization

LI Song,YANG Shiyi,ZHANG Fengfeng,SUN Lining

DOI: https://doi.org/10.3969/j.issn.1004-132X.2018.01.009

2018-01-01

Abstract:In order to conduct structural calibrations of the radiotherapy beds,and to improve their positioning accuracy,a new parameter identification method was proposed,which combined the genetic algorithm with mini-max optimization.Firstly,the calibration model of the radiotherapy bed was established based on the inverse kinematics.Then,the genetic algorithm and mini-max optimization methods,whose advantages were combined to identify 60 parameters,the residuals of the target function were reduced effectively.Finally,the laser tracker was taken to conduct two precision experiments,i.e.absolute positioning accuracy and repeated positioning accuracy experiments,for several groups of arbitrary poses,which were used to verify the results of parameter identification.The experimental results show that,for the positioning accuracy of calibrated radiotherapy beds,for the absolute positioning accuracy the position errors are less than 0.3 mm,and the attitude errors are less than 0.1°;for the repeated positioning accuracy,the position errors are less than 0.01 mm.Therefore,the parameter identification method may realize the calibrations of the radiotherapy beds,and the accuracy may meet the requirements.