Studies Of Leapfrog ADI-FDTD And Its GPU Acceleration In Vacuum Electronic Device

The main components used in microwave generation and amplification are vacuum electronics devices and solid state devices.Since vacuum electronic devices can obtain higher power,higher efficiency,and better radiation resistance than solid-state devices,vacuum electronic devices have received more and more attention from domestic and foreign scholars in recent years,and its development has been rapid.With the development of computer technology,more and more vacuum electronic devices are designed and developed by using numerical simulation technology,which greatly reduces the development time and cost.Inside the vacuum electronics,the main algorithms used to solve the time-varying electromagnetic field are finite difference time domain(FDTD),alternating direction implicit finite-difference time-domain(ADI-FDTD)and leapfrog alternating direction implicit finite-difference time-domain(Leapfrog ADI-FDTD).Among them,Leapfrog ADI-FDTD is widely used in numerical simulation experiments due to its advantages of no computer time-consuming computing resources,simpler iterative equations,unconditional stability,and parallel execution of algorithms.Numerical simulation consumes a lot of computer resources and simulation time.At present,the most effective way to solve this problem is to use parallel computing methods.Based on the above research background and research requirements,this paper studies the Leapfrog ADI-FDTD algorithm and its GPU acceleration based on vacuum electronic device applications.The main work and innovations include:1.Firstly,described the application background of vacuum electronic devices.Secondly,introduced the algorithm of solving the field of vacuum electronic devices and the processing methods of boundary conditions.Finally,introduced the advantages and significances of GPU parallel computing;2.Compared the FDTD,the ADI-FDTD and the Leapfrog ADI-FDTD algorithms;Described the advantages of the Leapfrog ADI-FDTD algorithm;Deduced the FDTD algorithm,the ADI-FDTD algorithm and the Leapfrog ADI-FDTD algorithm in turn in three-dimensional Cartesian coordinates;3.Outlined the commonly used boundary processing methods.Deduced the Leapfrog ADI-FDTD algorithm with CPML absorption boundary in the three-dimensional rectangular coordinate system based on the Leapfrog ADI-FDTD algorithm,which reduces the influence of the reflected wave and reduces the simulation calculation area.From the perspective of the algorithm,shortened the calculation time;4.Developed Leapfrog ADI-FDTD serial program with CPML in 3D Cartesian coordinates.Simulated the electromagnetic field of the TE10 mode in the rectangular waveguide and compared with the theoretical value of the TE10 mode in the same model.Verified the electromagnetic field calculation;5.Combined with the Leapfrog ADI-FDTD algorithm containing CPML,a corresponding parallel acceleration algorithm was established.The parallel program was written by using the CUDA parallel computing platform,and parallel optimization was performed.The speedup reached 58.4,which greatly shortened the simulation calculation time.Verified the calculation results of the serial program and the parallel program,and verified the consistency of the calculation results of the serial program and the parallel program;...