Pre-sampled data based prediction control for active power filters

A prediction control based on pre-sampled data from previous fundamental cycle for active power filters is presented in this paper. In which the reference current is calculated by the predicted current values in few steps ahead. In this way the active filter current phase delay can be effectively compensated. In addition the detection of harmonic components of load current, design of conventional proportional–integral (PI) current controller and hardware implementation for active filters are also discussed. The proposed prediction method is simple and easy for practical application. Tests were carried out for active filter with and without the phase delay compensation. Results show that the proposed prediction method is effective in compensation the phase delay of filter current.

► A prediction control based on pre-sampled data for active power filters is presented in this paper.
► The pre-sampled data can be obtained from sampled data in previous fundamental cycle.
► The reference current is calculated by the predicted current values in few steps ahead.
► The active filter current phase delay can be effectively compensated.
► The proposed prediction method is simple and easy for practical application.