Application of Statistical Parameters to Analyse the Performance of PWM Techniques in 3-Level Inverter-Based Compensator for Power Quality Improvement

This paper evaluates the performance of various PWM techniques for a 3-level inverter-based shunt active power filter by statistical parameters. It is well stated in the literature that the uses of power electronics-based devices create several power quality problems, such as poor power factor, harmonics in source current, poor voltage regulation, etc. Shunt active filters effectively alleviate these issues. A 3-level inverter, as a compensator depletes the harmonics in a single-phase grid-tied system. Level-shift, phase-shift, and hybrid PWM techniques generate switching pulses for 3-level inverter operation. These PWM techniques are compared for harmonic distortion at the source current, input power factor, and active filtering efficiency by statistical parameters such as mean squared error, root-mean-squared error, mean absolute deviation, mean absolute percentage error, standard deviation, variance, mean absolute error, and mean relative error. The above statistical parameters have been used to determine the optimum PWM technique for shunt active power filter operation. It is revealed from the results that for the PS PWM technique, the MSE is found to be less, i.e., 0.049441, and also has a low THD, i.e., 1.22% in source current. The behaviour of these PWM schemes is analysed through MATLAB/Simulink 2021b software. This paper contributes extensive information on PWM schemes for 3-level inverter-based compensator. It will help to comprehend various aspects of the power quality of a single-phase distribution system.