Algorithm Recognizing Nonlinear-Load-Current States and Having Dynamic Iteration Step Size

摘要

An active power filter is a feasible solution to compensate harmonics, and harmonic currents detection is one of its key aspects. In order to explore a harmonic currents detection method with both high steady-state detection accuracy and short dynamic response time, two nonlinear load current (NLC) states, a steady and a change state, should be recognized accurately. Thus, a method should be researched for recognizing NLC states accurately. This paper derived a direct computational method, and a simple and an optimal iteration algorithm, for computing NLC average value. The direct computational method can be employed to compute its average value accurately when NLC is in steady states. Although the optimal iteration algorithm is more complex than the direct computational method, it provides an idea of recognizing NLC states. From this idea, an algorithm was presented for recognizing NLC states based on the optimal iteration algorithm. In this algorithm, a dynamic iteration step size (ISS) was introduced, and thus an algorithm was proposed for recognizing NLC states accurately based on the dynamic ISS. Finally, accuracy of the proposed algorithm is confirmed by simulation and experiment results.