كنفرانس بين المللی مهندسی صنايع و سيستم ها

صفحه اصلی / دوازدهمین كنفرانس بين المللی مهندسی صنايع و سيستم ها

Three Objective Data Driven Robust Optimization of a Forest Biomass to Electricity Supply Chain Network Integrating Electricity Demand Forecasting

نویسندگان :

Afsaneh Taleb Beydokhti¹ Seyed Hamid Reza Pasandideh²

1- دانشگاه خوارزمی - دانشکده فنی - گروه مهندسی صنایع 2- دانشگاه خوارزمی - دانشکده فنی - گروه مهندسی صنایع

کلمات کلیدی :

Biomass،sustainable،optimization،Robust،electricity demand،multi-objective

چکیده :

The transition towards low-carbon energy systems has increasingly drawn attention to reliable renewable energy sources. Among these, forest biomass stands out as a promising option due to its availability, storability, and dispatchable power generation capability. However, the performance and stability of biomass-based power generation systems critically depend on accurate electricity demand forecasting and optimal supply chain design. This study presents an integrated, data-driven, and robust framework for designing a forest biomass-to-electricity supply chain network. Within this framework, we first compare two forecasting models, ARIMA and Holt, selecting the Holt model due to its superior accuracy in predicting electricity demand (based on RMSE, MAE, and MAPE metrics). Subsequently, electricity demand data is generated by combining the Holt model with the Exponentially Weighted Bootstrap (EWB) method, and a data-driven uncertainty set is constructed using the Support Vector Data Description (SVDD) approach. Finally, the network design problem is formulated as a multi-objective robust Integer Linear Programming (MILP) model, simultaneously optimizing economic (total system cost), social (job creation), and environmental (CO₂ emission reduction) objectives. The proposed model is solved using the weighted sum method, evaluating the trade-offs between different objectives. The findings of this research provide a reliable quantitative tool for strategic decision-making in power generation planning.