Real-Time implementation of ALMS Unified Power Quality Conditioner (UPQC)

Abstract

In MODERN power systems, distribution utilities mandate the connected loads Keywords: ALMS; PI; VSI; UPQC; PQ. Abstract compliance with the strict power-quality standards. This is to improve the reliability of the distribution system to cater the needs of critical loads and sensitive automation systems. The major challenges to maintain good quality power are: 1) fundamental reactive power requirements of the connected loads; 2) voltage sags and swells at the point of common coupling (PCC) due to connection and disconnection of large industrial loads and reactive power compensating capacitors; and 3) voltage and/or current harmonic distortion due to the presence of nonlinear loads. These APFs can be classified as shunt APF, series APF, and hybrid APF. The combination of both series and shunt APFs, to mitigate almost all of the voltage and current- related PQ problems, is a unified power-quality conditioner (UPQC). Superior performance and the ability to mitigate almost all major PQ problems make UPQC the most attractive solution for PQ improvement despite its high cost, complex structure, and control. This thesis work suggests an adaptive LMS-NN (Least Mean Square Neural Network) control algorithm rule for the three phase two level Unified power quality conditioner (UPQC) system. The prime motto of this active power conditioning is to handle various power quality relevant issues such as harmonic mitigation (current and voltage), voltage regulation and its balancing, Power factor correction (PFC) and reactive power compensation in the power distribution network. Two control techniques (P-Q instantaneous power theory and ALMS-NN) are implemented to draw the reference source currents for both series and shunt active power filters (APF). Moreover, the voltage source inverter (VSI) of the UPQC is fired by using these reference currents and their performances are compared under both even and uneven loading conditions. The efficacy of the ANN controller is carried out by rigorous simulation study by applying MATLAB/SIMULINK tool in association with real-time implementation by RTDS (real time digital simulator).