Realization Of Feedforward Function In LLRF System For Electron LINAC Based On FPGA

Thorium Molten Salt Reactor(TMSR) project is one of the strategic pilot technolog of Chinese Academy of Sciences, which goal is to research and develop the fourth generation fission reactor nuclear power system, solve the problem of thorium-uranium fuel cycle and significant technical challenges of Thorium Molten Salt Reactor. In this project,a Photo-Neutron Source, which is driven by a 15 MeV LINAC, was developed and built for the measurement of thorium-uranium fuel cycle data,and it located in Shanghai Institute of Applied Physics, Chinese Academy of Science. In this context, in order to ensure the LINAC can provide stable and high quality beam current, it developed a Low-Level Radio Frequency control system which based on FPGA.In this 15 MeV electron accelerator, the exist of transient of beam loading effect made the head of the bunch has highier energy gain and finally led to energy spread.Through compensating this effect, we can ensure the stability of cavity chamber pressure, reduce the energy spread, improve the power utilization and transmission efficiency of the beam. For the sake of reducing the beam loading effect, the mainly work in this thesis is to remove the needless energy gain of the head of the bunch. On this basis, this thesis first research the relevance theory of transient beam loading effect and to explore the feasibility of the method of beam injection in advance in15 MeV electron accelerator by using MATLAB to do the simulation. And, we found that even had chosen the best control curve, we couldn’t eliminate all the energy gain caused by transient beam loading effect.In order to get better control performance, this thesis mainly research the cutting-edge methods that used the low-level radio frequency(LLRF) control system to do the beam loading compensation. And then "the LLRF feedforward control methods" was decided to use in 15 MeV linac. In this method, we designed and built the feedforward module and added it into the original FPGA modules. By adding aconcave feedforward waveform on the original waveform, then setting it to the head of the RF field, we can weaken the forepart of RF field, so that the energy gain of the head of bunch will get a corresponding decrease when beam go through the cavities,the energy distribution will be more balance and the energy spread is brought under effective control. Comparing with the traditional methods, the LLRF feedforward control methods has highier control accuracy and better control performance, what’s more, it is suitable for the situation of big bunch and strong beam current. The actual test shows that the feedforward function of LLRF control system works well, the energy spread had been suppressed to an ideal level and the transmission efficiency of the beam current had also been significantly improved.