Design And Optimization Of Flat Top Pulse High Magnetic Field Systems

High magnetic field is widely applied in areas like scientific research, health care and industry as an experimental condition. Many interesting scientific discoveries benefit from high magnetic field. According to the pulse time, high magnetic field has two kinds: pulsed high magnetic field and continuous high magnetic field. Both fields have many important applications in various different areas. However, both of the two kinds of fields have certain limitations. Flat-top Pulsed High Magnetic Field (FPHMF) is a special kind of long pulse magnetic field that combines the high magnetic field and the continuous high field. FPHMF can supply a field as high as pulsed field with a field ripple as small as continuous field. And FPHMF costs much less than continuous field. Thus FPHMF has been favored by many scientists.Due to the high field level and long pulse time of FPHMF, the consumed power, temperature rise and stress of the FPHMF magnet can be very large. This is a great challenge to both the power supply system and the magnet. Due to the electric-heat coupling between the power supply and the magnet, the magnet and the power supply should be designed integratedly. This complicates the design of the FPHMF system. This paper mainly studied the integrated design of FPHMF system.Base on the existing equipment conditions of Wuhan National High Magnetic Field Center, Huazhong University of Science and Technology, the author established the mathematical models of the pulsed generator power supply, capacitor power supply and battery power supply. In addition the mathematical model of the magnet was established with electromagnetic, thermal and mechanical properties of the magnet system decoupled.According to the models of FPHMF, details of the design process and optimization methods of FPHMF system are presented. Taking the single-objective optimization of FPHMF as an example, the paper analyzed the application of the genetic algorithm in the optimization of FPHMF system. Enumeration method is adopted to validate the genetic algorithm.Single-objective optimization only pursues the highest magnetic field, with other parameters ignored. This makes the other performances of the power supply and magnet fairly poor. Thus multi-objective differential evolution optimization method is then applied, with most of the key parameters of the FPHMF system considered. Weighting method is used in the two-objective and three-objective optimizations of FPHMF. Process of the weighting method and pareto front of the FPHMF optimization are detailed.In the design of high-dimensional multi-objective optimization, a scoring system based on preference of the designer is adopted to classify the Pareto points of the optimization problem.Taking the single-coil FPHMF system powered by the generators12-pulse rectifie as an example, the design process and methods of multi-objective decision are presented.In order to make full use of the existing equipment at the WHMFC, the author combined the existing power supplies together to construct several new hybrid power supplies. Operating characteristics of the hybrid power supplies are discussed. The magnet can be single-coil and dual-coil. Thus there are many different combinations between the hybrid power supplies and the magnets. To find the optimized design of the various FPHMF systems, pareto-based multi-objective differential evolution algorithm was adopted. Finally, all the optimized designs are listed together. The design with the highest field level is modified and verified in detail.