EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

4271. Integrating multicriteria analysis with advanced data analysis methods

Invited abstract in session WB-47: MCDA in energy, stream Multiple Criteria Decision Analysis.

Wednesday, 10:30-12:00
Room: 50 (building: 324)

Authors (first author is the speaker)

1. Janusz Granat
Institute of Control and Computation Engineering, Warsaw University of Technology

Abstract

Data-driven multicriteria analysis is a challenging research area integrating data analysis from two sources: first, raw data used for development of a model or a set of alternatives, and second, the data provided by the model simulation or its interactive analysis. The presentation focuses on the second issue.

First stage of interactive multicriteria decision analysis generates several sets of solutions. The corresponding data can be then analyzed for supporting selection of a final decision, as well as providing information guiding specification of parameters of the user preferences in further interactions. Applying data analysis in the process of multicriteria analysis significantly improves the decision-making process and reduces the required computation time. Use of the cluster-based surrogate in discrete decision-space helps in reducing the discrete decision space dimension, and thus the computational resources needed for solving the underlying optimization tasks.

The effectiveness of the approach is illustrated by analysis of an energy model supporting decisions of a system operator of an onshore wind farm equipped with a hydrogen energy storage. The objective is to maximize the cumulative profit by deciding on: (1) a constant hourly amount of electricity to be delivered, (2) configuration of the storage system components, and (3) control of flows to/from the storage to deal with the variable renewable energy generation.

Keywords

Status: accepted

Back to the list of papers