Objectives
- To develop healthy and faulty models for Hydro-generators (Electromagnetics and Thermal).
- To deploy sensors and experimental set-up design for condition monitoring and fault diagnostic of HG.
- To forecast the residual time of fault occurrence or remaining useful life (RUL) of the deteriorating component.
- To propose CBM technique using machine learning algorithm.
Amrit Chapagain
MS By Research
Hydro-Himalaya Batch 2022
Expected Outcomes:
The goal of the project “Anomaly Detection and Condition Monitoring of Synchronous
Generators” is to transform our knowledge of and approaches to maintaining these essential
energy components through a comprehensive approach. The first major objective is to create
unique models that describe the operating characteristics of hydro-generators with a particular emphasis on both thermal and electromagnetic aspects. This project aims to create a
comprehensive framework that can quickly identify normal functionality and quickly detect and
characterize deviations or abnormalities within these generators by exploring the complexities of
both healthy and faulty states.
This project’s goal is to create a real-time monitoring system that can quickly and accurately
identify any anomalies and take proactive measures to prevent possible malfunctions or failures.
The project’s goal is to predict the amount of time left before a fault arises or to calculate the
remaining useful life (RUL) of deteriorating components for synchronous generators. This
predictive capability is expected to be an essential tool because it gives operators and
maintenance teams the insight, they need to plan proactive maintenance tasks, which reduces
downtime and maximizes operational efficiency.
The idea of a Condition-Based Maintenance (CBM) method seeks to anticipate malfunctions in
addition to detecting anomalies, allowing well-informed decisions to be made about maintenance
and repairs. It is projected that the CBM approach will result in an automated, sophisticated
system that can recognize patterns in data and gradually increase the accuracy of its fault
detection and remaining useful life prediction. This will ultimately lead to a change in
maintenance strategies from time-based and reactive to condition-based and proactive, leading to
a new era of longevity and reliability for synchronous generators.
Supervisors from host and Partner Universities
Bishal Silwal, PhD
Nils Jacob Johanssen, PhD