Energy Management and Operational Planning of Networked Microgrids in a Stochastic Environment

This article introduces a two-stage linear model designed for the coordination of networked microgrids, aimed at optimizing energy management and enhancing profitability through proactive and corrective strategies. Initially, the first stage involves day-ahead hourly planning for microgrids, executed in a deterministic environment without accounting for uncertainties. Subsequently, the second stage addresses these uncertainties in real-time network operation through a stochastic programming approach. The model's objective function quantifies and incorporates the variations resulting from both the proactive and corrective phases. To handle uncertainties in wind and solar energy production as well as load demand, probability distribution functions derived from Monte Carlo simulations are utilized. From these, representative scenarios are chosen using a scenario reduction technique. Specifically, the K-means algorithm is employed for scenario clustering, with the Davies-Bouldin (DB) index facilitating automatic clustering. Additionally, load management is conducted via a demand response program. The proposed model stipulates that, within microgrids, only non-critical load levels can be modulated based on the network's economic benefit. This optimization model, formulated as mixed integer programming, is simulated and resolved in the GAMS software environment. The primary goal of this two-stage model is to achieve optimal energy management by balancing economic efficiency with robust network performance. The results obtained validate the model's effectiveness.