基于改进遗传算法的磁靶向聚焦线圈阵列优化设计

发布时间：2018-11-03 21:40

【摘要】：磁性载药颗粒的靶向治疗是近年来肿瘤治疗中受到国内外广泛关注的新方法,由于其非侵入性和高靶向性的特点,具有广泛的用途以及潜在的巨大市场需求。外磁场的特性是治疗效果的关键影响因素。相比于传统的由永磁体产生外磁场的方式,应用线圈阵列形成外磁场具有灵活简便、可操作性强等优点,可以根据线圈阵列的排布及所加电流激励的不同而改变靶向位置、磁场强度及聚焦范围。论文设计了基于改进遗传算法的磁靶向聚焦线圈阵列,并对其聚焦性能进行全面的优化。基于电磁场及三维电磁场有限元理论,对单线圈模型的结构、参数进行设计。在此基础上,提出三种多线圈阵列,对比其聚焦效果并结合实际应用中的可行性,选择平面形线圈阵列为更优。然后研究了遗传算法的基本原理和基本理论,并对其进行改进,应用改进的遗传算法对线圈阵列中各子线圈电流组合进行求解。通过仿真计算、空间电场实测及实物试验验证其聚焦性能。主要的研究工作包括:①设计具有更优聚焦性能的单线圈参数及结构。在深入研究并掌握电磁场基本理论及三维有限元分析方法的基础上,分析匝数、内径及绕制方式对单线圈磁场磁感应强度分布的影响,分别对不同参数下线圈轴向、径向磁感应分布进行分析,综合选择磁感应强度高、靶向位置聚焦深度大、聚焦性能好的单线圈参数设计方案,最终确定了匝数n=6、内径d=2cm、以多层绕法的一种变形绕法绕制的单线圈方案为最优。②设计具有更优聚焦性能的多线圈阵列方案。基于已选择的单线圈参数,设计了半球形、平面形、圆环形三种多线圈阵列组合,对比分析相同激励下三种线圈阵列的聚焦性能,通过仿真计算,结合实际应用中的可行性,最终选择平面形线圈阵列为最优方案,并实现了多层平面形线圈阵列的非中心点聚焦及多点聚焦。③应用改进的遗传算法求解线圈阵列电流组合。研究并分析标准遗传算法的基本原理、概念以及求解流程,在此基础上,对算法的优化和改进从初始种群产生、基因丢失检测及修复策略、自适应交叉和变异算子、最佳个体保护等方面进行。通过四组测试函数,对其性能进行全面的测试与评估。基于改进的遗传算法对线圈阵列子线圈电流组合进行求解,得到磁感应强度高、梯度大,聚焦性能好且能量利用率高的电流组合。④根据设计出的线圈阵列方案,搭建了线圈阵列模型,应用空间磁场测量平台对靶向位置磁感应强度分布进行测量,测量结果验证了其聚焦性能,并进行实物试验,观察聚焦效果。
[Abstract]:The targeted therapy of magnetic drug-loaded granules is a new method which has attracted wide attention in recent years. Because of its non-invasive and high-targeting characteristics, it has wide application and potential huge market demand. The characteristic of external magnetic field is the key factor of therapeutic effect. Compared with the traditional way of producing external magnetic field by permanent magnet, the application of coil array to form external magnetic field has the advantages of flexibility, simplicity, maneuverability and so on. The target position can be changed according to the arrangement of the coil array and the different excitation of the current applied. Magnetic field intensity and focus range. A magnetic target focusing coil array based on improved genetic algorithm is designed and its focusing performance is optimized. Based on the finite element theory of electromagnetic field and three dimensional electromagnetic field, the structure and parameters of single coil model are designed. On this basis, three kinds of multi-coil arrays are proposed. Compared with the focusing effect and the feasibility of practical application, the planar coil array is selected as the best. Then the basic principle and basic theory of genetic algorithm are studied and improved. The improved genetic algorithm is applied to solve the current combinations of subcoils in the coil array. The focusing performance is verified by simulation calculation, space electric field measurement and physical experiment. The main research work includes: 1 designing single coil parameters and structure with better focusing performance. On the basis of deeply studying and mastering the basic theory of electromagnetic field and three-dimensional finite element analysis method, the influences of turn number, inner diameter and winding mode on the magnetic induction intensity distribution of single coil magnetic field are analyzed, respectively for the axial direction of the coil under different parameters. Based on the analysis of radial magnetic induction distribution, a single coil parameter design scheme with high magnetic induction intensity, high focusing depth at target position and good focusing performance is selected. Finally, the turn number n ~ (6) and the inner diameter (D ~ (2) cm) are determined. The single coil scheme of multi-layer winding method is the best. 2 A multi-coil array scheme with better focusing performance is designed. Based on the selected single coil parameters, three kinds of multi-coil arrays, hemispherical, planar and annular, are designed. The focusing performance of the three kinds of coil arrays under the same excitation is compared and analyzed, and the feasibility of practical application is combined with the simulation calculation. Finally, the planar coil array is selected as the optimal scheme, and the noncentral and multi-point focusing of the multi-layer planar coil array is realized. 3 the improved genetic algorithm is applied to solve the coil array current combination. The basic principle, concept and solution flow of the standard genetic algorithm are studied and analyzed. On this basis, the optimization and improvement of the algorithm are derived from the initial population, gene loss detection and repair strategy, adaptive crossover and mutation operators. The best individual protection is carried out. Through four groups of test functions, the performance of the comprehensive test and evaluation. Based on the improved genetic algorithm, the current combination of coil array subcoils is solved, and the current combination with high magnetic induction intensity, high gradient, good focusing performance and high energy efficiency is obtained. The coil array model was built, and the magnetic induction intensity distribution of the target position was measured by using the space magnetic field measurement platform. The results verified its focusing performance, and the physical experiment was carried out to observe the focusing effect.
【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TH789

【相似文献】