Discrete Non-Integer Order Differentiator Models Using Moth-Flame Optimization Algorithm
Discrete rational approximation models to the non-integer order differentiator sλ, where λ ε (0, 1), using Moth-Flame Optimization (MFO) algorithm is proposed in this paper. The proposed metaheuristic optimization approach used to design the discrete non-integer order differentiators (DNODs) does not employ any s-to-z domain mapping function to perform the discretization operation. Frequency domain characteristics of DNODs, solution reliability, and algorithm convergence performances are investigated among MFO and an advanced evolutionary algorithm called Particle Swarm Optimization with adaptive inertia weight (PSO-w). Results demonstrate the effectiveness of MFO in outperforming PSO-w in solving this non-linear and multimodal optimization problem. The proposed DNODs also exhibit better performance in comparison with the designs based on techniques such as Nelder-Mead Simplex algorithm and Cuckoo Search Algorithm published in recent literature.