Fault diagnosis of wind turbine with SCADA alarms based multidimensional information processing method

Qiu, Yingning and Feng, Yanhui and Infield, David (2020) Fault diagnosis of wind turbine with SCADA alarms based multidimensional information processing method. Renewable Energy, 145. pp. 1923-1931. ISSN 0960-1481

[img] Text (Qiu-etal-RE-2019-Fault-diagnosis-of-wind-turbine-with-SCADA-alarms-based-multidimensional)
Qiu_etal_RE_2019_Fault_diagnosis_of_wind_turbine_with_SCADA_alarms_based_multidimensional.pdf
Accepted Author Manuscript
Restricted to Repository staff only until 24 July 2020.
License: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 logo

Download (944kB) | Request a copy from the Strathclyde author

    Abstract

    This paper presents a first attempt to use Dempster-Shafer (D-S) evidence theory for the fault diagnosis of wind turbine (WT) on SCADA alarm data. As two important elements in D-S evidence theory, identification framework (IF) and Basic Probability Assignment (BPA) are derived from WT maintenance records and SCADA alarm data. A procedure of multi-dimensional information fusion for WT fault diagnosis is presented. The diagnosis accuracy using BPAs obtained from a sample WT and from the wind farm are compared and evaluated. The result shows that D-S evidence theory as a multidimensional information processing method is useful for WT fault diagnosis. Compared to previous SCADA alarms processing methods, the approach proposed predominates at aspects of simple calculation, superior capability on dealing with large volume of alarms through quantifying fault probabilities. It has the advantages of being easy to perform, low cost and explainable, which make it ideal for online application. A self-BPA-generating procedure for future online application with this approach is also provided in this paper. It is concluded that D-S evidence theory applied to SCADA alarm analysis is a valuable approach to intelligent wind farm management.