LSTM based Intelligent Load Management in a Stand-Alone Microgrid

Abstract

In the recent era the increase in load in the distribution segment with limited generating unit makes the load management mandatory. The aggregation of load from the cluster in practical scenario is very difficult. Further, the intermittent generation from the renewable energy unit makes discontinuity in power supply to the load. Thus, this article proposes an intelligent load management technique through Long Short-Term Memory (LSTM) for the aggregation and control of thermostatic loads based on their temperature of operation and state to mitigate the frequency unbalance without compromising consumer's thermal comfort. The rigorous simulation is done in MATLAB Simulink to showcase the efficacy of the proposed control mechanism considering several load dynamics.