Sandra Rothe

Application of diagnosis and prognosis to wind turbine system based on fatigue load

Fatigue damage in wind turbine structures is mainly induced by fluctuating loads strongly affecting the structural response of the system. The examination of fatigue damage, classification of the system state, prediction of remaining lifetime as well as the extension of maintenance interval become a challenge in structural health monitoring of wind turbine systems mainly due to offshore application. This contribution focuses on the structural load analysis in terms of maintenance intervals as well as service lifetime extensions. To postpone the point in time at which the system becomes nonfunctional, the structural load has to be mitigated while the energy production is retained as close as possible to the desired value. As fatigue damage is strongly influenced by the inflow parameters, a suitable control strategy is adopted to reduce the bending moments in the blades as one typical example. The contribution discusses the case when the damage accumulation is suddenly increased due to an unexpected situation (for instance high crack propagation rate) targeting to show that even if it happens, it is possible to retain the planned service lifetime through an suitably adopted control strategy. Influencing factors on the fatigue damage progression are pointed out. Flap-wise, edge-wise blade bending moments, fore-aft and side-to-side tower bending moments time series data simulated using FAST model developed by NREL are used for these purposes. Furthermore, the system state determination based on accumulated fatigue damage is done using the diagnosis-based data-filtering algorithm, and is represented in form of traffic-light-like coding. Here each color describes a specific system state.

Improvement and Comparison of Wear-oriented State-of-health Classification Methods Using Optimization Techniques

Mechatronic systems are monitored while its operation parameters are subjected to several influences affecting the efficiency, functionality, and safety. It is of major interest to infer the actual State-of-Health of critical components from acquired data. The degree of wear and the quality of mechatronic systems or cost-sensitive machine components are significant for the system reliability. The overall reliability of technical systems must be ensured in order to reduce the risks and costs of a system failure. Automated monitoring of wear and the classification of the machine state are necessary for State-of-Health evaluation. These techniques were used in the currently developed approach to judge both fault probability and system reliability. Optimization techniques are used to improve developed approaches concerning the reliability of State-of-Health evaluation. Core of this contribution is the comparison of two algorithms which can be easily used, applied, and handled. These algorithms were specially developed facing industrial data or measurements from technical systems. As example in this contribution a hydraulically driven machine part sliding over another one is used. A connection between measured hydraulic data to the degree of wear of the lubricated surface is used to calculate information about the state of a sliding surface between the two machine parts. The pressure time behavior is taken and filtered for better evaluation using arithmetic mean value and sliding window technique [1]. For further generation of suitably defined characteristics, the data has to be post processed and analyzed. Using the filtered data, the wear state is classified using two different methods. The first method uses thresholds to distinguish the surface condition into three states of wear. For the second method, further filtering and calculation of the trend lead to a classification of the wear state. In extension to previous work of the authors [1], here the thresholds of method one and the window size of method two are optimized to minimize the difference of the classification results for the same experiment. The main idea of this contribution is to make the results similar by means of optimization in order to assure the plausibility of the employed approaches. doi: 10.12783/SHM2015/80

Development of a State-related Evaluation for Diagnostic-oriented Data Filtering Approach

Ensuring the total reliability and availability of complex systems such as safe mechatronic systems or cost-sensitive machine components is of increasing importance. To monitor such systems during operation, parameters subjected to several influences affecting the efficiency, functionality, and safety are measured. It is of major interest to infer the actual State-of-Health of critical components from acquired data. The supervision of critical components of mechatronic systems becomes an important task and will be discussed here. The degree of wear and the functionality of complex systems are significant for the system reliability. The overall reliability of technical systems must be ensured in order to reduce the risks and costs of a system failure. To classify the machine state using easy-to-measure signals, two aspects are important: the filtering and the interpretation of the measured data. Core of this contribution is the development and application of a state-related evaluation for diagnostic-oriented data filtering approach to be used directly with industrial data or measurements from technical systems in operation. The data used in this contribution are taken from experiments generated using a wear test rig. The test rig designed for evaluation of tribological effects consists of two wear plates which interact with each other. A connection between measured macroscopic data (sliding force equivalent data resulting from test rig’s operation) and the degree of wear of the lubricated surface is established to calculate information about the state of a sliding surface between the two machine parts. The time behavior of the pressure is taken using arithmetic mean value and sliding window technique. The filtering method is explained in previous work of the authors [1]. The actual level of wear, calculated using the integrated damage increments, and the actual changes of the level of wear are utilized to define three states (healthy/good condition; small changes in condition; non healthy/not in good condition) heuristically. This concept is briefly introduced in [1]. A comparison of different measurements with the same operating conditions shows similar results. doi: 10.12783/SHM2015/76

Comparison of Three — Easy to Apply and Innovative — Signal-Based Approaches for Diagnosis of a Technical System With Wear

Ensuring the reliability and availability of complex systems such as safe mechatronic systems or cost-sensitive machine components is of increasing importance. Besides the availability of problem-specific sensors and filtering techniques three major issues are of interest:i) Preparation of the measured data (filtering),ii) Interpretation of the data with respect to the machines state as well as to the machines’ remaining lifetime or guaranteed functionality, andiii) establishing the required knowledge behind from available measurements and data.Core of this contribution is the development and application of easy to apply and easy to interpret algorithms to be used directly with industrial data or measurements from technical systems in operation.The three approaches applied areAI: Acoustic Emission (AE) characterized by measurements in combination using a suitable filter to be designed,AII: data analysis using operating system data feature capturing technique, andAIII: Adaptive fuzzy-based filtering [1] with training and classification modules.The approaches are developed in detail due to former research work [2], here they are applied and compared using the same experiment, the shown results are based on experiments. The three approaches use different algorithms but partially different signal sources (from the same experiment). Each of the approaches allows a different and distinguished problem-oriented insight to the complex wear process of the considered system, typical for mechanical engineering-related machines. The comparison between three different approaches for wear diagnosis can be considered as the main idea of this paper which allows insights into the advantages and disadvantages of each of these approaches.© 2014 ASME