Long-term scenario generation of heterogeneous random variables based on Markov chains for smart road tunnels planning

This paper presents a methodology for the synthetic generation of heterogeneous random variables that affect the decision-making process in planning electric components and Distributed Energy Resources (DERs) for smart road tunnels. To ease the task of the decision maker, the methodology is developed in a generalized form in order to tackle variables of different sources and extractions (e.g., electrical, weather and economic variables). The methodology allows the generation of long-term synthetic scenarios of these heterogeneous variables through the development of Markov Chains (MCs), assuming that each variable can be represented by states with memory. The proposed methodology is applied and tested on actual data collected at the locations of two smart road tunnels in Italy, that are under evaluation to be equipped with Photovoltaic (PV) and wind generation systems and with a Battery Energy Storage System (BESS). Numerical experiments confirm the validity of the proposal for the synthetic generation of variables like solar irradiance, wind speed and electricity prices, as confirmed through the evaluation of their relevant statistics in a twenty-year period.