Performance improvement of the shunt active power filter using a novel adaptive filtering approach

Abstract

This paper introduces an efficient control approach to enhance harmonics mitigation performance of the Shunt Active Power Filter(SAPF). This approach is based on adaptive filters favored by their built-in automatic parameters adaptation capability. The proposed filter, which uses a Variable Leaky Least Mean Square (VLLMS) adaptation, is applied with a modified instantaneous power PQ theory. This enhances its dynamic performance over the use of conventional time-invariant filters, and overcomes its limitations in the presence of non-sinusoidal voltage conditions. The studied SAPF model is simulated in MATLAB/SIMULINK with combinations of non-linear and unbalanced loads. Simulation results indicate a significant improvement in the transient response to less than a quarter of a cycle in its worst case, on top of additional improvements in the steady-state THD value. Moreover, the algorithm's computation efficiency evaluated in an embedded hardware prototype has given a fast execution time of 1.4µs .