Optimal current harmonic extractor based on unified ADALINEs for shunt active power filters

Adaptive linear neuron (ADALINE) is widely used in parameter estimation due to its algorithmic simplicity and parallel computing nature. One of the most popular training schemes for ADALINE is the least-mean-squared (LMS) rule, which can be implemented online to reduce the computation and storage requirements greatly. In this paper, an optimal current harmonic extractor based on unified ADALINEs for the shunt active power filter (APF) is proposed to achieve a better dynamic performance and reduced computation burden. The proposed control algorithm consists of three ADALINEs. Two ADALINEs are used for frequency estimation and supply voltage synchronization, while the third ADALINE is used to extract the fundamental active component of the load current. The main factor that affects the estimation speed and accuracy is the learning rate involved in LMS weight-update rule. Generally, this learning rate is selected by trial and error. In this paper, the learning rate of each ADALINE is tuned using particle swarm optimization to achieve the best dynamic performance. Furthermore, an adaptive learning rate for the frequency-ADALINE is proposed to enhance the estimation speed. The proposed ADALINE-based control structure is validated with a detailed experimental study.