An orthogonal-array-based particle swarm optimizer with nonlinear time-varying evolution

A PSO method with nonlinear time-varying evolution based on neural network for design of optimal harmonic filters

“…Based on the concept presented by Ratnaweera et al (2004) and Chatterjee and Siarry (2006), a PSO method is proposed (Ko et al, 2007), in which nonlinear time-varying evolution (PSO-NTVE) is performed. The mathematical expressions in PSO-NTVE are given as shown in Eqs.…”

“…Nowadays, PSO has been developed to be real competitors with other well-established techniques for population-based evolutionary computation (Eberhart & Shi, 2001;Elbeltagi, Hegazy, & Grierson, 2005;Gaing, 2004;Ho, Yang, Ni, & Wong, 2006). A PSO-NTVE method based on orthogonal array experiments has been presented by Ko, Chang, and Wu (2007). However, the degrees of the evolutionary functions are discrete.…”

Simple adaptive cognition for PSO

“…Similar changes to φ g and φ p have been proposed in the literature, typically favoring a linear increase or decrease of these coefficients over time, bounded by a maximum number of iterations [26], [27]. Nonlinear versions based on the same idea have also been proposed [14]. Significant improvements have been reported using these methods, but the complexity of finding useful settings for all of the new parameters introduced is considerable.…”

The minimization of voltage sag effect for specially connected transformers with a sensitive load and distributed generation systems

“…A PSO-NTVE method had been proposed wherein the optimal combination of inertia weights and acceleration coefficients is determined by orthogonal array experiments. Then, by utilizing five widely used benchmarks, the performances of this method and three other PSO methods are evaluated and compared [24]. This optimization method allows system equipment parameters to be determined over a wide range.…”

“…Node admittance matrix models are used. Voltage sag planning is analyzed using the PSO-NTVE method, which is effective for population based evolutionary computation [23,24]. A PSO-NTVE method had been proposed wherein the optimal combination of inertia weights and acceleration coefficients is determined by orthogonal array experiments.…”