Development of a Decision Support System Integrated with a Distribution Management System for Solar Energy Systems and Electric Vehicle Fleets

In the project titled "Development of a Decision Support System Integrated with a Distribution Management System for Solar Energy Systems and Electric Vehicle Fleets," a hybrid model based on Gaussian Distribution and supported by a Kalman Filter was developed for solar energy generation forecasting. The model enables generation forecasts over a wide geographical area using a limited number of measurement points. This approach provides a scientific method for achieving accurate estimation while reducing sensor costs, and its results were presented at the SIU 2024 Conference.

Probabilistic Load Flow

To manage the uncertainties introduced by renewable energy sources, particularly rooftop solar energy, in distribution networks, a Monte Carlo simulation-based Probabilistic Load Flow algorithm integrated with the generation forecasting method was developed. Going beyond deterministic approaches, this method provides distribution system operators with risk-based analysis capabilities and enables voltage violations to be predicted with high accuracy. The aforementioned generation forecasting method was presented at the SIU 2024 Conference.

Network Reconfiguration

A brute-force-based approach was developed for the network reconfiguration problem. More importantly, an original preprocessing method that reduces the solution space of this problem was introduced to the literature. Based on the identification of independent loops and the elimination of unsuitable topologies, this method significantly reduces optimization times and paves the way for real-time applications. This work was presented to the scientific community at the ISGT-Europe 2024 Conference.

Fleet Management and Smart Charging

The Mixed-Integer Linear Programming (MILP)-based algorithm developed for fleet management aims to reduce charging costs while addressing "service continuity" as a constraint, thereby filling a gap in the literature. Dynamic route optimization and vehicle replacement strategies, which are activated when the charging schedule is insufficient, introduce a new scientific approach to improving operational efficiency. The knowledge and expertise gained through the development of the smart charging algorithm contributed to our participation as a partner in the HORIZON-funded SOLAR-MOVE project.

ODTÜGÜÇ Software

The ODTÜGÜÇ software, which integrates all the developed PVDS algorithms and provides visualization capabilities, has been incorporated into our scientific infrastructure as a powerful R&D tool for validating academic studies in a virtual environment and conducting scenario analyses.

Objectives and Validation

The initial objectives of the project have been largely achieved. Due to delays encountered in the development and procurement of field monitoring units during the project, a strategic change was made in the data collection methodology. Instead of relying on field hardware, real-world data were collected using the existing laboratory infrastructure at the Ayaslı Research Center, Raspberry Pi-based data acquisition gateways, and the established OCPP server, enabling validation of the developed algorithms. This alternative approach allowed the scientific outputs of the project to be obtained without disruption and, in fact, within a more controlled environment.