Automatic generation control analysis of power system with nonlinearities and electric vehicle aggregators with time-varying delay implementing a novel control strategy Nimai Charan PATEL1∗,, Binod Kumar

Abstract

Automatic generation control (AGC) also known as load frequency control plays a vital role in interconnectedpower system for frequency regulation. Electric vehicles (EVs) with battery as the storage device can participate infrequency regulation service. In practice, a large number of EVs are aggregated as a single unit called EV aggregatorfor participation in frequency regulation service in AGC system. Participation of EV aggregators in AGC systemfor frequency regulation will be encouraged in near future because EVs have less environmental pollution than theconventional vehicles. However, participation of EV aggregators in AGC system may introduce time delay whichdegrades the dynamic performance of the power system and even may lead to system instability. Further, generationrate constraint (GRC) and governor dead band (GDB) of synchronous generator introduce nonlinearity in the system,which adversely affects its dynamic performance. Hence, selection of an appropriate control strategy is essential forperformance enrichment of the power system. In this work, a PID-fuzzy-PID (PID-FPID) controller optimally designedby hybridizing particle swarm optimization (PSO) and modified sine cosine algorithm (MSCA) is proposed and a novelattempt is made to improve the frequency stability of a two-equal-area interconnected thermal power system with GDBand GRC incorporating an EV aggregators with time-varying delay in each area. Integral time absolute error has beenchosen as the objective function to optimally design the controller parameters. Dynamic performance of the system isalso investigated with proportional-integral-derivative (PID) controller optimally designed by PSO, sine cosine algorithm(SCA), MSCA, and hybrid PSO-MSCA. The efficacy and supremacy of the proposed hybrid PSO-MSCA-based PIDFPID control strategy is established by contrasting the results with the others. Finally, the robustness of the proposedcontrol strategy is validated by (i) including two EV aggregators with time-varying delay in each area, (ii) applying alarge disturbance of 0.4 p.u. in area-1, and (iii) applying a random load in area-1.