Paper Contents
Abstract
This research presents a predictive maintenance system for offshore wind farms using a neural network model implemented in MATLAB Simulink. The model utilizes vibration parameters as input features to predict the maintenance needs of wind turbine components. The dataset used for training the neural network has been mathematically generated, simulating the operational conditions of offshore wind farms. The model outputs a binary label, with a label of '0indicating no maintenance is required and a label of '1indicating the need for maintenance. To facilitate remote monitoring and real-time decision-making, the system is integrated with the ThingSpeak IoT platform. The predicted maintenance labels are sent to the ThingSpeak cloud server, making them accessible from any location via the platforms web interface. The model demonstrates exceptional performance, achieving an accuracy of over 99%, indicating its potential for efficient and proactive maintenance in offshore wind farm operations. This work provides a comprehensive solution to optimizing maintenance schedules, reducing unplanned downtimes, and improving the overall reliability of offshore wind turbines.
Copyright
Copyright © 2025 Pratibha. This is an open access article distributed under the Creative Commons Attribution License.
