Multi-stage and Multi-objective Dynamic Optimization-Based Scheduling of Uncertain Hybrid Renewable Energy Systems with Real-Time Energy Compensation

Majority of the existing works on scheduling of hybrid renewable energy systems focus on offline methods that generally neglect forecasting errors of both supply and demand sides. The highly intermittent nature of renewable energy sources and the consumer demands can make this offline strategy highly challenging and inaccurate. Although various stochastic optimization-based techniques are discussed in literature to deal with these uncertainties, these methods generally work either on probabilistic model-based approaches that reflect the possible uncertainties or on uncertain parameter forecasting-based approaches, both of which can often become less realistic. In this work, a multi-stage real-time scheduling of a hybrid renewable energy based micro grid in the presence of uncertainties is implemented on to a wood pellet manufacturing industry based in Singapore that depends on the hybrid renewable energy sources for meeting its energy requirements. The two main objectives in the formulated optimization problem are to minimize the operational cost and carbon emissions associated with the energy generation. The incorporation of a multi-stage approach in the scheduling algorithm makes sure that the latest information of the uncertain parameters are updated which guarantees a better accuracy as compared to the conventional single-stage approach.