Tilbake til søkeresultatene


Hydropower Plants Operations’ Monitoring & Maintenance Recommendations through Digital Twins, empowered with eXt. Reality and eXplainable AI

Tildelt: kr 49 999

Hydropower is one of the renewable energy sources that have the highest conversion efficiency , contributing approximately 16%–17% of the total world electricity generation. Operations’ Monitoring and Maintenance Recommendations (M&M) are an important aspect for their efficient working and life expectancy, in the sense that ineffective processes (in M&M) can result into major problems: e.g., leading to electricity generation and revenues losses, producing threats to employees and public safety, etc. The M&M processes are usually data-driven, involving the analysis of data points, such as: daily fault reports, trends of operating parameters (retrieved from sensors), and system efficiency. We present the different stages involved in the maintenance activities of hydropower plants : diagnosis, planning, objectives selection, list maintenance activities, cost estimation, tender allocation, validation and implementation, and restoration of the system. The maintenance of hydropower plants can be categorized into two parts: (a) preventive maintenance, and (b) corrective maintenance. Preventive maintenance can be further classified into planned, conditional and predictive maintenance, whereas corrective maintenance is employed in emergency situations. Based on a recent literature review on the topic1, the preventive maintenance and operation optimization recommendations are provided through various condition-based monitoring techniques, such as: Fuzzy Logic, AHP, PSO, ANN, and SOM. Whereas, the usual technologies used for enabling remote monitoring and fault analysis are: SCADA, IoT, and cloud-based monitoring systems