SLIIT Researchers Develop AI Forecasting System for Biomass Power

SLIIT researchers have developed an artificial intelligence framework to forecast power requirements in biomass-fuelled steam turbines used in palm oil production, according to the institute’s announcement.

The research examines how factories can plan energy generation using agricultural by-products such as fibres, shells and husks. SLIIT links more accurate forecasting to better management of biomass resources and industrial energy use.

According to SLIIT, the study was published in Energy Conversion and Management and used an eight-year industrial dataset. The research team reports approximately 25% greater forecasting accuracy than benchmark methods.

Palm oil factories use agricultural residues as fuel for steam turbines. Overestimating energy requirements can waste biomass, while underestimating them can increase reliance on electricity from the grid, the institute explains.

The forecasting system learns from industrial operating patterns to help operators anticipate power requirements. SLIIT presents the approach as a way to support fuel planning, cost control and production decisions.

Himaya Perera, a SLIIT Electrical and Electronic Engineering graduate, led the research. According to the institute, she is pursuing a Master of Engineering in Electronics at La Trobe University in Australia.

The team includes SLIIT’s Eranga Wijesinghe and Bhagya Nathali Silva, together with Shalitha Jayasekara and Honnyong Cha of Kyungpook National University in South Korea.

SLIIT links the research to wider efforts to improve industrial energy efficiency and the use of renewable resources. The institute identifies potential applications beyond the sector examined in the study.

The work focuses on predicting requirements for biomass-based power generation. Its stated purpose is to give plant operators information that could help match turbine output with demand and improve the management of available fuel.