This study aims to present a comprehensive steady-state analysis of parallel-connected self-excited induction generators (SEIGs) with hybrid excitation, addressing critical challenges in voltage stability and power quality for renewable energy applications.
The research uses a mathematical modeling approach based on the equivalent circuit model, transforming the hybrid excitation system into an equivalent star configuration to simplify analysis. The fixed-point iteration method (FPIM) is implemented to solve the system’s nonlinear equations through systematic convergence stages, requiring 250–300 iterations with O(n) computational complexity for a steady-state solution. The methodology integrates analysis of magnetizing characteristics, terminal voltage regulation and current distribution for parallel-connected SEIGs. This analytical framework is experimentally validated using a test setup of two SEIGs (2.2 kW and 5.5 kW) with hybrid excitation under steady-state conditions.
The hybrid excitation system improves voltage regulation from −8.4% to 0%, with 5.5 kW SEIG delivering 5,510 W while maintaining 50 Hz ± 0.2% frequency stability. Current distribution analysis shows 11.1 A from 5.5 kW and 4.8 A from 2.2 kW SEIG, with terminal voltage stabilizing at 415 V ± 2%. The system achieves a 40% reduction in neutral current compared to conventional configurations, with power factor optimization between 0.92 and 0.95.
Future research could explore the dynamic performance of SEIGs with hybrid excitation under transient conditions to further enhance system reliability, nonlinear magnetizing characteristics, voltage regulation, load sharing, frequency stability, power distribution and grid integration methodologies.
This study provides a novel contribution by integrating a hybrid excitation system with parallel-connected SEIGs, offering a detailed analysis of their steady-state behavior under various conditions. The findings present superior convergence over the Newton–Raphson method (500+ iterations) and binary search (400–450 iterations) while handling unbalanced loads up to 30% variation.
