Combination of Model-based Observer and Support Vector Machines for Fault Detection of Wind Turbines

International Journal of Automation and Computing, Jun 2014

Support vector machines and a Kalman-like observer are used for fault detection and isolation in a variable speed horizontal-axis wind turbine composed of three blades and a full converter. The support vector approach is data-based and is therefore robust to process knowledge. It is based on structural risk minimization which enhances generalization even with small training data set and it allows for process nonlinearity by using flexible kernels. In this work, a radial basis function is used as the kernel. Different parts of the process are investigated including actuators and sensors faults. With duplicated sensors, sensor faults in blade pitch positions, generator and rotor speeds can be detected. Faults of type stuck measurements can be detected in 2 sampling periods. The detection time of offset/scaled measurements depends on the severity of the fault and on the process dynamics when the fault occurs. The converter torque actuator fault can be detected within 2 sampling periods. Faults in the actuators of the pitch systems represents a higher difficulty for fault detection which is due to the fact that such faults only affect the transitory state (which is very fast) but not the final stationary state. Therefore, two methods are considered and compared for fault detection and isolation of this fault: support vector machines and a Kalman-like observer. Advantages and disadvantages of each method are discussed. On one hand, support vector machines training of transitory states would require a big amount of data in different situations, but the fault detection and isolation results are robust to variations in the input/operating point. On the other hand, the observer is model-based, and therefore does not require training, and it allows identification of the fault level, which is interesting for fault reconfiguration. But the observability of the system is ensured under specific conditions, related to the dynamics of the inputs and outputs. The whole fault detection and isolation scheme is evaluated using a wind turbine benchmark with a real sequence of wind speed.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

http://link.springer.com/content/pdf/10.1007%2Fs11633-014-0790-9.pdf

Combination of Model-based Observer and Support Vector Machines for Fault Detection of Wind Turbines

Regular Paper Special Issue on Recent Advances on Complex Systems Control, Modelling and Prediction Manuscript received February Combination of Model-based Observer and Support Vector Machines for Fault Detection of Wind Turbines Nassim Laouti 1 Sami Othman 1 Mazen Alamir 0 Nida Sheibat-Othman 1 0 Gipsa-lab/CNRS, University of Grenoble , Rue de la Houille Blanche, 38400 Saint Martin d H`eres , France 1 Universit ́e de Lyon , Universit ́e Lyon 1, CNRS, CPE Lyon, UMR 5007, LAGEP, F-69616 Villeurbanne , France Support vector machines and a Kalman-like observer are used for fault detection and isolation in a variable speed horizontalaxis wind turbine composed of three blades and a full converter. The support vector approach is data-based and is therefore robust to process knowledge. It is based on structural risk minimization which enhances generalization even with small training data set and it allows for process nonlinearity by using flexible kernels. In this work, a radial basis function is used as the kernel. Different parts of the process are investigated including actuators and sensors faults. With duplicated sensors, sensor faults in blade pitch positions, generator and rotor speeds can be detected. Faults of type stuck measurements can be detected in 2 sampling periods. The detection time of offset/scaled measurements depends on the severity of the fault and on the process dynamics when the fault occurs. The converter torque actuator fault can be detected within 2 sampling periods. Faults in the actuators of the pitch systems represents a higher difficulty for fault detection which is due to the fact that such faults only affect the transitory state (which is very fast) but not the final stationary state. Therefore, two methods are considered and compared for fault detection and isolation of this fault: support vector machines and a Kalman-like observer. Advantages and disadvantages of each method are discussed. On one hand, support vector machines training of transitory states would require a big amount of data in different situations, but the fault detection and isolation results are robust to variations in the input/operating point. On the other hand, the observer is model-based, and therefore does not require training, and it allows identification of the fault level, which is interesting for fault reconfiguration. But the observability of the system is ensured under specific conditions, related to the dynamics of the inputs and outputs. The whole fault detection and isolation scheme is evaluated using a wind turbine benchmark with a real sequence of wind speed. Fault detection and isolation; wind turbine; Kalman-like observer; support vector machines; data-based classification - With the widespread use of wind turbines (WTs) as renewable energy systems, it is now important to include control and supervision in the system design. Fault detection and isolation (FDI) of WTs allows reducing maintenance costs, which is particularly important for offshore WTs. Online supervision should suggest the best maintenance time as a function of fault occurrence and wind speed in order to reduce operation and maintenance costs. Early detection of faults allows also avoiding degradation of the material and other side effects. Furthermore, fault detection is essential for control reconfiguration in order to ensure optimal power in case of partial fault. Even though the wind turbine functionality might be similar to rotating machinery, it involves a number of difficulties ranging from a high variability in the wind speed, aggression by the environment, measurement difficulties due to noise and vibrations, besides the fact that wind turbines are supposed to run continuously for several years. For these reasons, the development of methods for FDI in WT is increasingly important. Similarly, a number of fault tolerant control (FTC) approaches are also being applied to WT, but this is out of the scope of this paper. FDI approaches can in general be classified as modelbased or data-based: On one hand, model-based methods require a comprehensive model of the system. On the other hand, success of data-based approaches is conditioned by the significance (amount and quality) of historical data and the mathematical method used to detect the patterns in data. However, training data is usually limited to some specific conditions that are typically normal, non faulty data, with limited variations of operating conditions. Limitations of both model-based and data-based approaches can be overcome by combining them in order to ensure optimal supervision. This represents the main idea of the present paper. Reviews of WT monitoring and fault diagnosis were proposed by [ 1 − 3 ]. Both data- and model-based approaches were reported. Among model-based approaches, observers were applied to monitoring several parts of wind turbines. Reference [ 4 ] proposes an unknown input observer to detect sensor faults around the WT drive train (...truncated)


This is a preview of a remote PDF: http://link.springer.com/content/pdf/10.1007%2Fs11633-014-0790-9.pdf

Nassim Laouti, Sami Othman, Mazen Alamir, Nida Sheibat-Othman. Combination of Model-based Observer and Support Vector Machines for Fault Detection of Wind Turbines, International Journal of Automation and Computing, 2014, pp. 274-287, Volume 11, Issue 3, DOI: 10.1007/s11633-014-0790-9