Artificial intelligence based nonlinear control of hybrid DC microgrid for dynamic stability and bidirectional power flow

Conventional power generation resources are depleting rapidly and world power sector has been moving towards renewable energy sources (RESs). In this paper nonlinear control for renewable energy and hybrid energy storage system (HESS) based DC microgrid (DCMG) has been presented. PV and wind energy being the renewable sources whereas fuel cell, battery and ultracapacitor constitute the HESS. The global mathematical model of the said system has been presented. Datasets of varying solar irradiance and temperature have been trained by Artificial Neural Network for the reference voltage generation of PV. Integral terminal sliding mode controller (ITSMC) has been proposed for the output DC bus voltage regulation. Lyapunov stability criterion has ensured the overall stability of the system. A comparison of ITSMC with SMC and Lyapunov redesign controller has also been presented. Grid connected battery electric vehicle charger with grid to vehicle (G2V) and vehicle to grid modes (V2G) has been presented being an application of DCMG. The proposed system has been validated by using MATLAB/Simulink (2020b). Moreover, the hardware in loop setup has been used to observe the real time applicability of the proposed controller.

Automated summary

This paper presents a nonlinear control method for renewable energy and hybrid energy storage system (HESS) based DC microgrid (DCMG). The renewable sources are PV and wind energy, while the HESS includes fuel cell, battery, and ultracapacitor. The global mathematical model of the system is presented, and Artificial Neural Network is used to train datasets of varying solar irradiance and temperature for PV reference voltage generation. An integral terminal sliding mode controller (ITSMC) is proposed for the regulation of output DC bus voltage, and the Lyapunov stability criterion ensures the stability of the system. A comparison of ITSMC with other controllers is conducted, and an application of DCMG is presented for a grid connected battery electric vehicle charger with G2V and V2G modes. The proposed system is validated using MATLAB/Simulink (2020b), and the real-time applicability of the proposed controller is observed using a hardware-in-loop setup.