Implementation of Shunt Active Power Filter(SAPF) algorithms

Abstract

Improving power quality has been the major research topic in last few decades due to flooding of semiconductor and other non-linear devices. The power quality of any source is judged by the some indexes defined by international bodies such harmonics factor, telephonic interference level (TIF) etc. Using the different harmonic compensation schemes we must be able to meet those index limits. This is very important in reference to performance and economy of operation. Power filters are widely used in modern electrical distribution system to eliminate the harmonics associated with it. The active power filter (APF) is one of power filters which have better dynamic performance. The APF needs an accurate control algorithm that provides robust performance under source and load unbalances. The control methods are responsible for generating the reference currents which used to trigger the Voltage Source Inverters (VSI). Thus, compensation of harmonics depends largely on the algorithm adopted. For any Shunt APF system there is various way of implementing the control block whose output goes to gate of the voltage source Inverter. Further, the harmonic and frequency has been modeled to propose a new control strategy for the shunt Active Power Filter. The Least Mean Square (LMS) algorithm is the basic estimation algorithms which estimate the parameters based on only present data. The Recursive Least Square (RLS) algorithm uses recursive method to estimate the parameters using both present and past values. In this thesis, main aim is to implement a basic control algorithm in MATLAB-SIMULINK first then the harmonic and frequency estimation part is implemented in MATLAB. Further the estimation part is implemented using the Arduino Mega 2560 Microcontroller for experimental validation.