Font Size: a A A

The Research Of The Electromagnetic Tracking System

Posted on:2014-06-28Degree:MasterType:Thesis
Country:ChinaCandidate:C QinFull Text:PDF
GTID:2252330422463119Subject:Power system and its automation
Abstract/Summary:PDF Full Text Request
Electromagnetic tracking is a method of using magnetic field information for thelocation and spatial attitude of objects, which has important applications in bio-medicine,virtual reality, etc. Especially electromagnetic tracking systems prove to serve as real-timenavigation of medical devices in the image guided surgery systems. With the help, doctorscan try to avoid the danger zone and reduce surgical trauma and complicating disease.Compared with the mechanical tracking technology and ultrasonic ranging technology, itsexcellency resides in that it’s convenient to use, highly precise and without line-of-sightbarrier. In order to research deeply the principles and algorithms of the the electromagnetictracking systems, all the work such as electromagnetic tracking systems modeling andsimulation analysis, constructing the experimental platform, error analysis and algorithmvalidation are carried out, The main research content and progress are as follows:The basic working principles of electromagnetic tracking systems are investigated.This article uses the Comsol multiphysic software to create a physical model of thetransmitter coil and the induction coil, simulates the magnetic flux density and the inducedelectromotive force in the algorithm, and analyses the theoretical error of the algorithms..Simulation results show that the magnetic dipole model calculating magnetic fielddistribution and the simulation magnetic field distribution are conformable. When thedistance between spot point and source point is20times further than the ring coil’sdimensions, The relative error is less than1%, As same as the simulation results of theinduction electromotive force. Hence, the magnetic dipole model can be applied to theelectromagnetic tracking systems.The transmitter coil and the induction coil are wrapped, and the experimental testplatform are built. High-speed data acquisition card real-time access the waveform data ofthe induction coil’s induced electromotive force. To reduce the noise, using Matlab softwareimplements digital lock-in with the the induction signal data and reference signal data, thenthe amplitude of the induction coil’s induced electromotive force is obtained.High-frequency magnetic field induction voltage in space and the noise through the signalacquisition channels are analysed with experimental test. It proves that the interference signal has little effect on the measurement results. some experiments measuring theinduction electromotive force are made verifying the test data can match the theoreticalresults well with the data processing of Normalization.The electromagnetic tracking positioning algorithm firstly lists the magnetic fieldequations of the magnetic dipole model basing on magnetic field coupling between thesource and sensor, and then uses optimization algorithm iteratively solving out the positionand orientation parameters. In this paper, both of the Gauss-Newton algorithm and geneticalgorithm are used to solve the equations. The solving results show that the Gauss-Newtonalgorithm calculates fast, has nice solution accuracy, but the rapidity of the algorithmconvergence highly depends on the selection of the initial point. Genetic algorithm is lessdemanding on the initial conditions, and it can easily achieve global convergence, but thecomputed speed is slower, the solution accuracy is not as good as the Gauss-Newtonalgorithm. In the Practical applications, using genetic algorithm for the initial values firstly,and then accurately solve the equations can guarantee that algorithm does not occur in thecase of not converge and the optimum solution can be got.Limited to the experimental conditions and work progress, This thesis starts frominvestigate the principle of electromagnetic tracking systems, and does some work aboutmeasurement and equation solving problems. There are still many problem need to be solve,such as the theoretical model refinement correction, reducing the error between the modelcalculated data and experimental data, producing a flexible mechanical structure whichhas scales and rotates to measure the angle. These require to be improved in the future.
Keywords/Search Tags:Electromagnetic tracking system, magnetic dipole, digital lock-in, solution ofnonlinear equations, Gauss-Newton Algorithm, Genetic Algorithm
PDF Full Text Request
Related items