Multi objective operation optimization method for flexible resources in industrial parks considering multiple uncertainties of source and load

Abstract

With the large-scale integration of distributed renewable energy and the increasing volatility of electric and gas loads, traditional deterministic models struggle to accurately characterize the dynamic response behavior of park-level energy systems. To address this issue, this paper proposes a multi-objective operation optimization methodology for park-level flexible resources that accounts for multiple source–load uncertainties. First, an empirical distribution is constructed based on historical source–load forecast error data, and a probability ambiguity set containing the true distribution is formulated using the Wasserstein distance to characterize the multiple uncertainties in both source and load. Second, a min–max distributionally robust optimization model is established with operational cost and carbon emissions as objectives, where the minimization problem corresponds to the search for the optimal operation scheme, and the maximization problem identifies the worst-case probability distribution within the ambiguity set. Furthermore, the model is transformed into a finite-dimensional convex optimization form using strong duality theory, and a weighted fuzzy membership degree method is introduced to achieve an equivalent solution. Simulation results demonstrate that the proposed methodology can effectively balance economy, environmental sustainability, and robustness, providing a credible decision-making tool for the optimal dispatch of flexible resources in park-level systems under data-limited scenarios.