Region of convergence by parameter sensitivity constrained genetic algorithm-based optimization for coordinated load frequency control in multi-source distributed hybrid power system

The imbalances created by generation and load can indeed result in frequency variations demanding a load frequency control (LFC). The LFC becomes even more complicated with the multi-source distributed hybrid power system (HPS), where designing optimal controller parameters by identifying the region of convergence (RoC) is critical as it guarantees the HPS's stability. Hence, this paper aims to address the frequency regulation (FR) issue in a state-of-the-art modeled multi-source distributed HPS with a coordinated control approach. The modeled HPS includes a reheat thermal power system (RTPS), wind turbine generator (WTG), fuel cell (FC) stack, battery energy storage system (BESS), diesel engine generator (DEG), and various controllers. A novel method called parameter sensitivity algorithm (PSA) is proposed to obtain the RoC of the controller gains and is further optimized using a constrained genetic algorithm (GA). To demonstrate coordinated control effectiveness, a hybrid objective function for optimization is formulated as constrained GA's fitness function using the integral time absolute error (ITAE) and integral absolute error (IAE). Furthermore, a comparative assessment for PI, PID with filter (PIDN), cascaded PI-PD with filter (PIPDN) controller topologies is carried out using error index-based performance metrics to ascertain the preeminence of the hybrid objective function over existing objective functions. The investigation results validate that the suggested optimized controllers provide enhanced FR and are robust to uncertainties. The practical feasibility of the proposed methodology is reinforced using dSPACE hardware in loop testing.