A data-knowledge hybrid decision support system for wastewater treatment operations: the Acqua dei Corsari plant case study

This research introduces a Data-Knowledge Hybrid Decision Support System that combines embedded knowledge and data-driven approaches to address critical operational challenges in wastewater treatment plant management. The proposed methodology extends traditional decision support capabilities by combining fuzzy logic principles with advanced data augmentation techniques and deep neural networks, aiming to reduce discrepancies between theoretical recommendations and operational limitations in wastewater treatment facilities. In light of the escalating environmental regulations and growing operational complexities, this novel approach was implemented and validated at the “Acqua dei Corsari” wastewater treatment plant in Palermo, where it demonstrated the ability to handle real-world operational limitations while maintaining theoretical optimization principles. The system's distinctive architecture leverages fuzzy logic to encode expert knowledge, while employing a sophisticated data augmentation strategy to overcome the limited data availability. Results from extensive field testing validate not only the system's effectiveness, but also its unique ability to adapt to plant-specific constraints without compromising theoretical optimization principles. This study presents an approach to integrate domain expertise with machine learning techniques through a framework designed to bridge theoretical understanding with practical implementation constraints. This innovative approach provides a methodology for developing data-driven expert decision support systems, offering a practical solution for modern wastewater treatment plants operating under infrastructure limitations and strict regulatory requirements. The work represents a notable advance in intelligent water management systems, establishing a systematic path from domain knowledge to automated operational guidance.