Multi-Objective Approach for Optimum Inverter Switching Frequency Selection Using Parabolic Carrier-Based PWM in Solar Microgrids

This paper introduces a systematic weighted multi-objective approach for selecting the switching frequency (fs) of a parabolic-carrier-based pulse-width-modulated (PC-PWM) inverter in a solar microgrid. The MATLAB (Version 23.2.0) /Simulink (Version 23.2) model incorporates a 10 kW PV array, an incremental-conductance MPPT boost converter, a regulated 700 V DC link equipped with a 4700 µF capacitor, a 15 kVA three-level neutral-point-clamped (NPC) inverter, an LCL output filter, and an 8 kW nonlinear three-phase rectifier load. The inverter fs varies from 2 to 20 kHz in 1 kHz increments under otherwise fixed nominal operating conditions. For each fs, voltage total harmonic distortion (THDV), current total harmonic distortion (THDI), inverter loss, and efficiency are evaluated and integrated into a normalized weighted objective function. FFT-based THD is evaluated over the last five fundamental cycles using a 1 µs discrete simulation step and harmonic orders up to the 39th. For the stated model, candidate-frequency set, normalization, and baseline weights, 16 kHz is the best evaluated point, with THDV = 1.44%, THDI = 2.28%, inverter switching loss = 353 W, and inverter efficiency = 96.78%. The added weight-sensitivity analysis demonstrated that this optimum depends on the selected engineering priorities: stronger harmonic weighting shifts the optimum upward, whereas stronger loss weighting shifts it downward. Consequently, the proposed procedure is presented as an offline design method rather than a universal or adaptive optimal solution.