| The Extended Interaction Oscillator(EIO)is a new type of vacuum electronic device with high power and wide bandwidth,which is widely used in millimeter wave to terahertz band.The global optimization of vacuum electronic devices has been developed as a result of the deepening of theoretical and practical research,and the problem of how to realize the co-transition of multiple targets with high frequency,wide bandwidth,and high power has become the focus of research on high frequency band microwave devices.In this paper,based on the multi-objective global optimization algorithm,combined with the operating characteristics of the extended interaction oscillator,a theoretical study and simulation of the multiobjective co-transition problem of oscillator output power,structure length and bandwidth are carried out,and the following aspects are mainly accomplished.(1)Based on the one-dimensional electronic disk model and the basic principle of self-excitation saturation of the resonant cavity,a non-linear self-consistent calculation program suitable for extended interaction oscillators is obtained.The theoretical simulation results show that the period length and its corresponding synchronous voltage are the fundamental prerequisites for the beam-wave interaction.The applicable current,characteristic impedance and external quality factor are essential for resonator energy building up.This requires that when the beam-wave synchronous interaction occurs,the energy released by the electron beam is greater than that of coupled output.Moreover,intense beam-wave interaction and few coupling outputs will shorten start-up time.The maximum output power does not depend on the electrons slowing down to saturation in a certain gap,but on the electrons releasing energy in this gap to make the electrons bunching more effectively.In order to release energy better in the next or several gaps,the best combination of the electrons interacting in all gaps is realized.It is noteworthy that the optimal values of different targets,such as output power,efficiency,start-up time,and bandwidth for extended interaction oscillator are mutually constrained.Therefore,trying to solve the co-transition of multiple targets is a key issue for further research.(2)A one-dimensional nonlinear self-consistent procedure based on the extended interaction oscillator is combined with Non-dominated Sorting Genetic Algorithms ?(NSGA-?),Multi-Objective Evolutionary Algorithm based on Decomposition(MOEA/D),Pareto Envelope-based Selection Algorithm ?(PESA?),Multi-Objective Particle Swarm Optimization(MOPSO)to obtain an automatic EIO optimization procedure for multiple objectives.The program equates EIO optimization as a multi-objective optimization problem,completes the construction of decision vectors,solution vectors,and realizes a global optimization algorithm with the ability to handle multi-dimensional and complex problems of EIO.By this method,the optimization procedure of a 95 GHz EIO example is presented,and the Pareto set with the best target points of output power,interaction structure length,and bandwidth is obtained to elaborate the characteristics of structural and electrical parameters distribution during the co-transition of multiple targets.(3)Based on the results of the multi-objective algorithm EIO optimization procedure,the corresponding three-dimensional electromagnetic simulation model is established.The results show that the simulation conclusions of the CST-PIC solver agree well with the results of the multi-objective algorithm-based EIO optimization procedure,and the automatic optimization procedure gives a set of Pareto optimal solutions,replacing the previous single optimal solution to effectively solve the cotransition problem of multiple objectives,and also explains the constraint relationship between different parameters and objectives,especially the number of gaps,appearance quality factor and power and bandwidth,providing designers with more choices in practical applications. |
PDF Full Text Request