| Two topics have been studied in this paper. The first one is the reconstruction algorithm in electromagnetic imaging. The second is the design and realization of the array for focused magnetic field technology. Optimization works on both topics are based on improved Genetic Algorithm (GA).Strong nonlinear and complex characteristics must be concerned in the reaction between imaging object and electromagnetic field. Applications of electromagnetic imaging much depend upon the reconstruction algorithms' developments. There are many problems need to be handled, such as: low efficiency, uncomfortable precision, hard to use known information, ill-posed, inverse-problem, etc. Study of the new methods to achieve reconstruction calculation are important works for both research and application of electromagnetic imaging.Focused magnetic field technology, especially the technology in relatively low frequency is an absolute new idea in many applications. Apply this technology in electromagnetic imaging could get ideal focused magnetic field distribution in thegiven domain, and in this domain the magnetic field is strong enough to ignore the effects due to the magnetic field out of the boundary. In this way, calculation should be performed only in the given domain. Otherwise, if the focused domain is small enough, the imaging object in the given domain could be considered as uniform stratified biological tissue. It should reduce the reconstruction time remarkably and get better spatial resolution. Moreover, it has been proved by many clinical researches that Transcranial Magnetic Stimulation (TMS) could be a new solution and hope for psychotic therapy. Applying focused magnetic field technology in this field should significantly improve the selection capability to nerve when stimulation is in execution.Main research works and innovations in this paper are introduced as follows:1. Improvement of Genetic AlgorithmGenetic Algorithm (GA) is widely used as a kind of global optimization method. In order to achieve the balance between its convergence speed and efficiency, an improved GA which is called as SRAGA (Self-Reparative & Adaptive GA) is brought forward in this paper.Haiming Distance of each individual is considered in the adaptive operators of crossover, mutation and the method to get the first generation.In order to enhance generation's diversity, when GA is running, missed gene detection and reparation policy is applied to find and repair missed gene in the generation. This policy has strong capability to avoid GA convergence in a local best value, and improve the probability to convergence to the best answer of whole value space.Elite technique is used in the select operator of the improved GA. In the improved GA, the select operator copies the best individual directly to next generation instead randomly selects individuals to be copied as it does in a simple GA. So the phenomena of losing best individual during optimization process can be avoided and the speed of convergence can be improved.Missed gene detection and reparation policy together with dynamic crossover and mutation operators which are based on Haiming Distance, can prevent the appearance of "super individual" efficiently. The Elite technique ensures the direction of optimization is always towards better answer of whole value space.2. Performance evaluation of improved GAIn this paper, five test functions that can express the common requests have been chosen to accomplish an impersonal and comprehensive performance evaluation of SRAGA.As a contrast algorithm, this paper only canceled missed gene detection and reparation policy together with dynamic crossover and mutation operators of SRAGA. Both of contrast algorithm and SRAGA are forced to start from same generation in every comparison. The evaluation results show SRAGA has better performance in both speed and efficiency. It is shown the success of improvement policies that applied to GA. It is also shown that SRAGA not only has satisfied efficiency and also has enhanced the capability to converge to the best answer.3. Applying SRAGA in electromagnetic imaging for biological tissueIn order to solve the problem in electromagnetic imaging, such as ill-posed, inverse-problem, etc. The feasibility of applying GA in this filed has been studied in this paper. By defining an objective function, reconstruction is converted to be an optimization process of searching an optimum parameter configuration. Finally, SRAGA has been applied ia conductivity reconstruction of the edema in head, and got satisfied results.4. Applying SRAGA in focused magnetic field technologySchemes of coil array are presented in this paper to get the electromagnetic field distribution with ideal focusing capability. Above ten small coils excited by different current are employed in the design, and the total field distribution is summation of the field of each single coil. Three types of coil array are designed in semi-spherical, plane and torus shape. Then SRAGA is applied in the optimizationof both value and phase of the current infused into each coil. Base on the calculated results of the optimized current configurations, ideal focusing capability is presented as contour lines and 3-D mesh charts of magnitude of both magnetic and electric field within the calculation area. It is shown that the coil arrays have good capability to establish a focused magnetic field distribution in a given area. In addition, it is also shown that the coil arrays have the capability to focus on two or more targets simultaneously, and the capability to focus on arbitrary point. So it can be done by adjust the exciting current to get given performance of the focused magnetic field distribution.5. Realization of 3-dimensions magnetic measurement equipment and focused magnetic field designIn order to test the real performance of the focused magnetic field design, a coil array has been established, which is based on the semi-spherical model. Further more, a 3-dimensions magnetic field distribution measurement equipment with enough spatial resolution has also been developed, in which Hall components have been used as the sensors. Then magnetic flux density that is founded by the array has been measured in 3-dimensions by the measurement equipment.Good agreement is observed between the measured results and the calculated results. It is shown that the 3-dimensions magnetic field measurement equipment work properly. It is also shown that the coil arrays designed above have the good performance and realizable capabilities. |
PDF Full Text Request