Perception & Fusion of Multidimensional Information & Cooperative Decision-making for Intelligent Diagnosis of Wind Turbine Critical Parts

This is a university-dominated, industry-assisted and competence-complementary collaborative project between Norway and China with focus on structural health/condition monitoring and fault detection/diagnosis for onshore and offshore wind turbine components using digital technologies. During the period of 2020-2021, two PhD candidates were hired at NTNU, working with the Norwegian partners EDR & MEDESO AS (EDR) and SAFETEC NORDIC AS (SAFETEC). One PhD is working on the development of a digital twin for wind turbine drivetrain and as the first step, a methodology is developed and validated by numerical simulations to estimate the gearbox loads using vibration data, for the purpose of online fatigue monitoring. The other PhD is working on the application of machine-learning methods (Convolutional Neural Network-CNN) for gearbox bearing damage and fault detection and diagnosis. Results based on the numerical simulations indicate a good accuracy for damage detection and characterization using the CNN method. Regular meetings and seminars with the participation of Norwegian and Chinese partners were carried out and the future cooperation on the use and analysis of the operational data for an onshore wind farm in China was also discussed.

This is a collaborative project between Norway and China. The Norwegian team is led by Norwegian University of Science and Technology (NTNU), in collaboration with two monitoring and digital service companies, EDR & MEDESO AS (EDR) and SAFETEC NORDIC AS (SAFETEC). The Chinese team is led by Hunan University (HNU), in collaboration with Central South University (CSU), Hunan University of Science and Technology (HUST) and two leading wind power companies XEMC WINDPOWER (XEMC) and GOLDWIND TECHNOLOGY (GOLDWIND). The goal is to use digital and intelligent solutions for structural health/condition monitoring to address the issues of high failure rate and high operation/maintenance cost of wind turbines. To achieve this goal, a "perception-transmission-processing-utilization-control" digital solution will be proposed for operation/maintenance of wind turbine critical components and an advanced visualization platform, which supports data exchange and sharing among various parts, for wind turbine monitoring and fault diagnosis will be developed and demonstrated. The main activities and novelties are summarized as below: - A methodological system for global information perception via measurements is proposed for the full life circle information of the wind turbine critical components (blades, bearings and gearbox). - A framework combining deep fusion and automatic acquisition is proposed to fuse multi-source heterogeneous data and acquire field knowledge. - Intelligent fault detection/diagnosis methods for wind turbine critical components are proposed and applied using simulated data from numerical models and measurement data. - A strategy for hierarchical early warning and fault tracing is proposed based on coordinated decision-making using digital twin technologies and demonstrated in actual wind farms.

Publications from Cristin

A data-driven approach for fault diagnosis of drivetrain system in a spar-type floating wind turbine based on the multi-point acceleration measurements

Digital twin based virtual sensor for online fatigue damage monitoring in offshore wind turbine drivetrains

Modelling of wind turbine gear stages for Digital Twin and real-time virtual sensing using bond graphs

Fault detection of offshore wind turbine drivetrains in different environmental conditions through optimal selection of vibration measurements

Estimation of Wind Turbine Gearbox Loads for Online Fatigue Monitoring Using Inverse Methods

1

No publications found

No publications found

No publications found

No publications found