Development Of Free Electron Laser Optimization Software Based On Genetic Algorithm

Development of high level application software played an important role during the construction and operation of free electron laser(FEL). Excellent high level application can provide convenience for the runtime debugging and the parameter optimization of the accelerator. The XAL designed by Oak Ridge National laboratory for SNS showed dominance after comparing its functions to those of the other software around the world. It was upgraded to Open XAL as many laboratories cooperated to make it an open and general application software for accelerators physics. Though it seemed not an easy work to make the best of Open XAL for FEL application, we still found its virtual accelerator quite universal based on our intensive study. Transplant and development of the virtual accelerator, including the study on the framework of Open XAL, writing of the configure files and debugging of Open XAL applications, were presented in the first half of this thesis.FEL is the fourth generation light source, optimization of FEL facilities is of great significance to achieve radiations of high-quality, e.g. higher brilliance, purer spectrum. Optimization for radiation power of FEL were discussed in the latter part. An optimization application code based on real-coded genetic algorithm for FEL were designed, within which some appropriate operators have been chosen and ameliorated. In addition, a friendly Graphical User Interface was made by Java Swing. The virtual accelerator will enable an excellent output of FEL before the FEL was finally constructed. It would be very cheering to combine virtual accelerator and optimization application software as it would be helpful to ensure the result of optimization close to reality by simulation.The work presented in this thesis were part of the high level application software for Shanghai Soft X-ray Free Electron Laser and Dalian Coherent Light Source, which will make off-line debugging and optimization of FEL super convenient. By transplanting Open XAL, its virtual accelerator was made available for FEL. Meanwhile, a universal optimization software for FEL was designed based on genetic algorithm.