Adaptive Switching Frequency Selection Based on Two-Step Efficiency Optimization for Grid-Connected Inverters

Abstract

A two-step efficiency optimization of a three-phase voltage source inverter with series R damped LCL filter is proposed. The first optimization step aims to find the optimal parameters of the series R damped LCL filter at rated load conditions. The objective function of step-one is to minimize the values of the passive elements of the filter-stage. Afterward, the magnetic cores of inductors are thoroughly designed using soft magnetic powder materials. This design is proposed to specify the smallest core-size required based on the value of the inductance found in step-one of the optimization. The second optimization step aims to optimally select the switching frequency of the inverter-stage considering load variations. This optimization is based on the optimized filter-stage parameters. The proposed objective function of step-two aims to minimize the power losses in the inverter-stage and filter-stage while limiting the harmonic-levels at the point of common coupling. Detailed loss-formulation including filter-stage core and copper losses is thoroughly presented. Teaching-learning-based optimization algorithm is adopted to perform the optimization problem. Experiment and simulation are done to justify the proposed optimization procedure.

Keywords

• Voltage source inverter (VSI)
• passive filters
• teaching-learning-based optimization (TLBO)
• core losses