This study presents a closed-loop magnetically driven laser steering manipulator positioning system for endoscopic microsurgery.A multimagnetic field strength sensor circuit is embedded in a 16−mm diameter analog lase...This study presents a closed-loop magnetically driven laser steering manipulator positioning system for endoscopic microsurgery.A multimagnetic field strength sensor circuit is embedded in a 16−mm diameter analog laser steering manipulator.The magnetic field distribution of a cylindrical permanent magnet in three-dimensional space is first modeled using an integral model to overcome the large error induced by the magnetic dipole model when the sensor and permanent magnet are close.The integral in the model is then decomposed using the Gauss–Legendre quadrature to improve the computational efficiency of the formulation.Moreover,five commonly used global search optimization methods are compared.Then,the algorithm with the fastest computational rate among these five algorithms,the tree-seed algorithm,is fused with the Levenberg–Marquardt algorithm,which performs well in local search,to obtain a hybrid optimization algorithm.Finally,it is demonstrated through static and dynamic experiments that the system based on the hybrid algorithm can obtain satisfactory computational errors while maintaining a high computational rate.展开更多
基金supported by the Key Research and Development Projects in Guangxi,China(AB21076005)the Specialized Talents in Guangxi,China(AD18281018)+1 种基金the Innovation Project of GUET Graduate Education,China(2023YCXS129)the Innovation Project of GUET Graduate Education,China(2022YCXS153).
文摘This study presents a closed-loop magnetically driven laser steering manipulator positioning system for endoscopic microsurgery.A multimagnetic field strength sensor circuit is embedded in a 16−mm diameter analog laser steering manipulator.The magnetic field distribution of a cylindrical permanent magnet in three-dimensional space is first modeled using an integral model to overcome the large error induced by the magnetic dipole model when the sensor and permanent magnet are close.The integral in the model is then decomposed using the Gauss–Legendre quadrature to improve the computational efficiency of the formulation.Moreover,five commonly used global search optimization methods are compared.Then,the algorithm with the fastest computational rate among these five algorithms,the tree-seed algorithm,is fused with the Levenberg–Marquardt algorithm,which performs well in local search,to obtain a hybrid optimization algorithm.Finally,it is demonstrated through static and dynamic experiments that the system based on the hybrid algorithm can obtain satisfactory computational errors while maintaining a high computational rate.
