Fig. 4: Impact of modified constraints on the particle topology optimization process.
From: Inverse-design topology optimization of magnonic devices using level-set method
a Contribution of the additional constraints to the total objective function value during optimization. b Evolution of the learning rate α throughout the optimization process. c Final design of the nanoparticle obtained using the decreasing learning rate technique and the modified objective function (Eq. (9)) with the following weights of the constraints: ζ = 10, ξ = 0.05, ν = 1.
