Optimal Current Harmonic Extractor Based on Unified ADALINEs for Shunt Active Power Filters
In this study, A unified ADALINE control algorithm for shunt APFs is proposed in this paper. The novelty of the proposed control structure lies in its control universality, estimation optimality, and reduced computational burden. The control scheme does the following functions: frequency estimation, synchronization, and current harmonic extraction. Each function is accomplished using one ADALINE. To achieve the best dynamic performance, the learning rates of all the three ADALINEs are optimized using PSO technique. Furthermore, an adaptive learning rate for frequency estimation is proposed to improve the dynamic performance of the frequency-ADALINE. The proposed unified ADALINE structure shares the same input vector, and hence it is more computationally efficient. The superiority of the proposed control scheme has been validated experimentally with distorted supply condition under frequency and load disturbances. The proposed control adapts the change in the system parameters (frequency and load) effectively within a cycle due to its neural nature.