A novel cascade optimization method based on butterfly optimization algorithm

Abstract

The parameter optimization of a multi-component isotope separation cascade is a multi-dimensional problem, involving many variables such as cut for stages, feed stage and separation factors. In this paper, a novel metaheuristic algorithm called butterfly optimization algorithm (BOA) is applied for the first time in cascade optimization field. Different objective functions are introduced like D function and a combination of D function and total flow rates. BOA method solves the problem of a 20-stage square cascade optimization to separate Xenon isotopes with the aim of getting the greatest D value or the least flow rates. The optimization results show that the optimal cut is close to the sum of the concentrations of the lightest N′C isotopes. A 21-stage step cascade is optimized as well to obtain the best working conditions to separate 72Ge and the cut θ is consistent with the experiment results. The calculation results indicate that BOA shows good performance in terms of calculation stability, speed and convergence to global optimum compared with traditional optimization methods and other metaheuristic algorithm like TLBO.