Dynamic Event Driven Robust Control Design with Uncertainty Compensator for Agent Misbehave in Autonomous AC Microgrid

In this paper an event based adaptive neural network (ANN) observer is designed, where selected neurons are trained by an adaptive law to estimate uncertainties in order to mitigate agent misbehave in the multiagent featured microgrid. Time dependent uncertainties can be directly compensated through sliding mode controller (SMC) but to compensate state dependent uncertainties and agent uncertainties, ANN algorithm is proposed. Novel triggering law is derived for neurons weight update in ANN observer. Event trigger consensus control in multiagent scenario allows all agents (DGs) to achieve desired reference value while reducing the control effort. Robust and fast response is guaranteed by designing integral sliding surface with proportional surface term in SMC reaching law. Proposed method is immune to system sensitivity against both state and time dependent uncertainties. The controller efficacy is demonstrated in MATLAB/Simulink platform.