The most complex innovation and development project in the history of HELB is completed

Loss optimization system seeks a place on the EU Market.

Innovation and development projects are one of the best ways to increase the competitiveness of small and medium-sized enterprises. However, it should not be overlooked that this is one of the most complex types of projects for entrepreneurs.

One such project was recently completed by our development team. It is a project called the Loss optimization system in advanced networks, co-financed by the European Regional Development Fund, which lasted for three years.

In collaboration with the Faculty of Electrical Engineering and Computing (FER), HEP-ODS, and Sedam IT, a system for continuous analysis of technical parameters of the distribution network was developed. Through the application of machine learning, the system estimates and predicts technical and non-technical losses in the network.

“Our collaboration and the flow of knowledge among partners are invaluable benefits”, says Kristina Mileta, a development engineer and project manager.

For the project, data were collected at pilot locations in Elektrodalmacija Split and Vinkovci, where measuring and monitoring devices were installed at a total of 21 transformer stations. Processed data included integrated geospatial information and pseudonymized data from settlement measuring points. Of course, teams faced numerous challenges during this process.

“The most significant challenge was addressing technical obstacles that no one knew existed. Of course, we should not overlook the comprehensive challenge of automated processing of a large amount of imperfect data,” explains Mileta.

The developed DINGO platform analyses available data, conducts power flow calculations, detects, classifies, and locates technical and non-technical losses, and predicts losses a day in advance. Measurements at the level of individual transformer station branches are monitored in almost real-time, enabling the establishment of an alarm system.

This platform offers functionalities such as loss procurement planning, reporting faults on transformer stations, detecting optimization possibilities on transformers to minimize losses, and many other functionalities for the development of advanced networks. Although the project has officially concluded, there are numerous plans for its future implementation into the existing system.

“Our long-term goal is to obtain a system, from a pile of technical documentation and measurements from millions of users, that will simply provide all the information needed for the secure supply of volatile energy,” explains Mileta, adding: “The next step is to recognize the needs of the international community and enter the European market”.

Although it is a long-term process with numerous challenges, DINGO already provides insight into the real state of the low-voltage network, enabling targeted projects where there is a priority need.
Therefore, it is believed that this project will contribute to the future of power engineering and smart grids, both in Croatia and the EU, and enable the reduction of losses in low-voltage networks.