Comparative Performance of Neuro-Fuzzy PSS Architectures with Adaptive Input Link Weights and Nonlinear Functions

Ramirez-Gonzalez, M.; Malik, O.P.

doi:10.3182/20080706-5-kr-1001.02358

Cited by 1 publication

(3 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In (11), d can be realised through non-linear scaling [35] since the adjustment of the relative sensitivity of areas within their operating intervals represents a simple alternative to find the fuzzy sets that best fit the linguistic terms d is associated with. By affecting d through a non-linear factor ρ to change its membership function distribution according to the following function F: .…”

Section: Fuzzy Control Algorithmmentioning

confidence: 99%

“…However, K o is determined by the physical limits of the controller output, which were already set to ±0.1 in this study. As for the parameters of fuzzy membership functions and singletons, they can be adjusted indirectly in a relatively simple way by using an associated non‐linear factor [35]. In this study, only K e , K Δ e , and the non‐linear factor ρ to change the membership function distribution for d are selected to be tuned.…”

Section: Fpodc Designmentioning

confidence: 99%

“…On the other hand, d denotes a normalised variable computed as in the following equations:

y = K_{e} \cdot normalΔ v_{6} + K_{normalΔ e} \cdot normalΔ \dot{v_{6}}

d = (||, y)

where Δ v 6 and

normalΔ \dot{v_{6}}

are the voltage deviation at bus 6 and its rate of change, respectively, and K e and K Δ e are the associated scaling factors. The sign of the control action is determined from the following expression:

s_{u} = ({, \begin{matrix} 1em4pt \\ 1 & if y \geq 0 \\ - 1 & otherwise \end{matrix})

In (11), d can be realised through non‐linear scaling [35] since the adjustment of the relative sensitivity of areas within their operating intervals represents a simple alternative to find the fuzzy sets that best fit the linguistic terms d is associated with. By affecting d through a non‐linear factor ρ to change its membership function distribution according to the following function F :

F false(d, ρ false) = d^{ρ}

the fuzzy controller output with d = 0.7, as a function of Δ v 6 and

normalΔ \dot{v_{6}}

scaled within the range {−1,1}, is illustrated in Fig.…”

Section: Fpodc Designmentioning

confidence: 99%

See 2 more Smart Citations

Conventional and fuzzy PODCs for DFIG‐based wind farms and their impact on inter‐area and torsional oscillation damping

Ramirez-Gonzalez

Malik

Castellanos

et al. 2017

IET Renewable Power Generation

View full text Add to dashboard Cite

Contribution to the damping of inter-area and torsional oscillation modes, in doubly fed induction generators (DFIG) based wind farms, by power oscillation damping controllers (PODCs) based on two conventional structures and a fuzzy control strategy is investigated in this study. In this regard, a PODC with no lead/lag compensation is designed first and then a PODC with one lead/lag block is developed using eigenvalue techniques and the application of an iterative process based on the bat optimisation algorithm (BOA). Moreover, a fuzzy PODC, based on a simple fuzzy controller and tuned with the BOA according to the system transient response under a critical perturbation, is also designed. Comparative performance of the three PODCs is evaluated on a multi-machine power system. It is observed that all three PODCs can contribute to improving the damping of inter-area oscillations. However, eigenvalue analysis and non-linear time domain simulations reveal that each of them may also impact to a lesser or greater extent the shaft torsional oscillation mode damping. Their relative impact in this regard is also investigated.

show abstract

Section: Fuzzy Control Algorithmmentioning

confidence: 99%

Section: Fpodc Designmentioning

confidence: 99%

“…On the other hand, d denotes a normalised variable computed as in the following equations: