Incremental Learning Method for Wind Turbine Fault Detection Models Considering False Negatives

Condition monitoring of wind turbine based on incremental learning and multivariate state estimation technique

The real-time fault detection and diagnosis algorithm of a liquid rocket engine is the basis of online reconfiguration of guidance and the control system of a launch vehicle, which is directly related to the success or failure of space mission. Based on previous related works, this paper carries out comparative experimental studies of relevant intelligent algorithm models for real-time fault detection engineering application requirements of a liquid hydrogen–oxygen rocket engine. Firstly, the working state and detection parameters’ selection of a hydrogen–oxygen engine are analyzed, and the proposed three real-time intelligent fault detection algorithm model design methods are elaborated again. Fault detection calculation and analysis are carried out through normal test data and fault test data. The comparative analysis results of real-time intelligent fault detection algorithm models is presented from three dimensions: detection time, fault detection, and stability and consistency. Finally, based on a correlation analysis, a comprehensive intelligent fault diagnosis model design framework is given to further solve the requirements of real-time fault detection and diagnosis engineering development of a liquid rocket engine, a complex piece of equipment.

<strong class="journal-contentHeaderColor">Abstract.</strong> Wind turbine Supervisory Control and Data Acquisition (SCADA) datasets available for research usually contain a limited number of failure events. This limitation hinders the successful application of Deep Learning (DL) methods for fault detection and prognosis, as they require large datasets for robust training and generalisation. This work proposes a method using Conditional Generative Adversarial Networks (cGANs) to generate synthetic SCADA time series that replicate wind turbine behaviour under controllable operational, environmental, and degradation conditions. Given a set of SCADA time series representing these conditions, the cGAN generates temperature and pressure time series simulating gearbox operation. Results show that augmenting the training set of an Artificial Neural Network (ANN) fault detection model with synthetic time series reduces false positives in the detected gearbox faults by 84 % on average, enabling the model to blindly detect a fault in a test wind turbine without prior knowledge of the event. Furthermore, training a Convolutional Autoencoder-based unsupervised health indicator (HI) model with both real and synthetic SCADA time series leads to an HI that more accurately captures the expected degradation trend. Using this HI, the gearbox's remaining useful life (RUL) can be predicted within the defined error bounds from around 4.5 months before the detection of the fault, while the HI obtained without the synthetic data fails to produce reliable RUL estimates.

Fault detection and diagnosis algorithms for an open-cycle liquid propellant rocket engine using the Kalman filter and fault factor methods

This study presents an advanced data augmentation framework to enhance fault diagnostics in industrial centrifugal pumps using vibration data. The proposed framework addresses the challenge of insufficient defect data in industrial settings by integrating traditional augmentation techniques, such as Gaussian noise (GN) and signal stretching (SS), with advanced models, including Long Short-Term Memory (LSTM) networks, Autoencoders (AE), and Generative Adversarial Networks (GANs). Our approach significantly improves the robustness and accuracy of machine learning (ML) models for fault detection and classification. Key findings demonstrate a marked reduction in false positives and a substantial increase in fault detection rates, particularly in complex operational scenarios where traditional statistical methods may fall short. The experimental results underscore the effectiveness of combining these augmentation techniques, achieving up to a 30% improvement in fault detection accuracy and a 25% reduction in false positives compared to baseline models. These improvements highlight the practical value of the proposed framework in ensuring reliable operation and the predictive maintenance of centrifugal pumps in diverse industrial environments.

This paper proposes a new method for estimating the topology of radial distribution networks and a novel data-driven fault detection technique using smart meter data. An improved method to estimate voltage sensitivity coefficients from smart meter data is proposed that takes into account the variability of the transformer voltage. These coefficients are used in both the topology estimation and the fault detection algorithms. In the topology estimation algorithm, an improved graph learning algorithm, which iteratively constructs the network graph, is devised. In the fault detection algorithm, the objective is to look for sudden changes in the estimated voltage sensitivity coefficients. The performance of the proposed methods is first assessed on randomly generated topologies under various noise conditions. Then, the performance of the algorithms is validated using real smart meter measurements obtained from Australian distribution networks. The results show a significant improvement in the accuracy of estimated topologies compared to the state-of-the-art methods while the fault detection technique is successful in detecting faults using real smart meter data.

This paper proposes a fault analysis and detection algorithm for a product industrial assembly line workflow. The images that describe the orientation and composition of the assembly line products are acquired using a video camera and analyzed using an algorithm for fault analysis and detection developed in Matlab. The small amount of time for image analysis along with a higher recognition rate for fault detection, recommend the proposed methodology for this type of industrial scenarios.

Fault detection algorithm for grid-connected photovoltaic plants

A fault-tolerant technique of a grid-side converter (GSC) is a very important task because the unbalanced grid power endangers the overall system. Since the GSC is very sensitive to grid disturbance, the complete system needs to be stopped suddenly once an open-switch fault occurs. To improve the reliability of system, the continuous operation should be guaranteed. In this paper, a redundant topology based fault-tolerant algorithm is proposed for a GSC in a wind power generation system. The proposed scheme consists of the fault detection and fault-tolerant algorithms. The fault detection algorithm employs the durations of positive and negaitive cycles of three-phase grid currents as well as normalized root mean square (RMS) currents. Once a fault is detected, the corresponding faulty phase is identified and isolated to enable the fault-tolerant operation. The faulty phase is replaced by redundant one rapidly to recover the original shape of the grid currents, which ensures the continuity in operation. In contrast with the conventional methods, the proposed fault detection and fault-tolerant algorithms work effectively even in the presence of the open faults in multiple switches in the GSC. Simulation results verify the effectiveness of the proposed fault diagnosis and fault-tolerant control algorithms.

This paper proposes a novel fault detection algorithm in a brushless DC (BLDC) motor drive system. This proposed method is configured without the additional sensor for fault detection and identification. The fault detection and identification are achieved by a simple algorithm using the operating characteristic of the BLDC motor. The drive system after the fault identification is reconfigured by four-switch topology connecting a faulty leg to the middle point of DC-link using bidirectional switches. This proposed method can also be embedded into existing BLDC motor drive system as a subroutine without an excessive computational effort. The feasibility of a novel fault detection algorithm is validated in simulation and experiment.

This report describes models of building faults created for OpenStudio to support the ongoing development of fault detection and diagnostic (FDD) algorithms at the National Renewable Energy Laboratory. Building faults are operating abnormalities that degrade building performance, such as using more energy than normal operation, failing to maintain building temperatures according to the thermostat set points, etc. Models of building faults in OpenStudio can be used to estimate fault impacts on building performance and to develop and evaluate FDD algorithms. The aim of the project is to develop fault models of typical heating, ventilating and air conditioning (HVAC) equipment in the United States, and the fault models in this report are grouped as control faults, sensor faults, packaged and split air conditioner faults, water-cooled chiller faults, and other uncategorized faults. The control fault models simulate impacts of inappropriate thermostat control schemes such as an incorrect thermostat set point in unoccupied hours and manual changes of thermostat set point due to extreme outside temperature. Sensor fault models focus on the modeling of sensor biases including economizer relative humidity sensor bias, supply air temperature sensor bias, and water circuit temperature sensor bias. Packaged and split air conditioner fault models simulate refrigerant undercharging, condenser fouling, condenser fan motor efficiency degradation, non-condensable entrainment in refrigerant, and liquid line restriction. Other fault models that are uncategorized include duct fouling, excessive infiltration into the building, and blower and pump motor degradation.

In data-based method for fault detection, principal component analysis (PCA) has been used successfully for fault detection in system with highly correlated variables. The aim of this paper is to combine the exponentially weighted moving average (EWMA) control scheme with PCA model in order to improve fault detection performance. In fact, PCA is used to provide a modeling framework for the develop fault detection algorithm. Because of the ability of EWMA control scheme for detecting small changes, this technique is appropriate to improve the detection of a small fault in PCA model. The performance of the PCA-based EWMA fault detection algorithm is illustrated and compared to conventional fault detection methods using simulated continuously stirred tank reactor (CSTR) data. The results show the effectiveness of the developed algorithm.

This thesis presents a critical appraisal of several differing approaches to the design and testing of fault detection (FD) algorithms monitoring the instrumentation used in the alternated aerobic/anoxic cycles (AC) process for nitrogen removal. Several features are selected as the basis of the FD, involving the slope and the timing of the process measurements of the various nitrogen compounds. Two separate FD algorithms have been developed for the anoxic and the aerobic phases, requiring a separate tuning but sharing the same principles: first some low-level checks are performed on the raw signals, discriminating gross malfunctions like missing data and spikes, then more sophisticated methods are used to investigate the presence of more subtle anomalies that were not detected by the previous screening. The FD problem is treated either in terms of classification problem, testing different algorithms such as binary trees, support vector machines (SVM) and principal component analysis (PCA), or as a forecasting one, using the Bayesian theory to predict the faulty or normal state of the process based on the previous records. An operational data set obtained from a municipal plant was used to first train the algorithm. However, due to the fairly limited information which could be extracted, a more comprehensive data set was created building an AC model based on the standard Benchmark simulation model with improved nitrogen kinetics and seasonal temperature variations. Detailed sensor models were also included, so that the occurrence of faults could be totally controlled, both in kind and timing. The performances of the various methods on either the operational and the synthetic datasets have been assessed comparing the anomalies detected by the methods with those actually observed. While the great majority of the gross faults is successfully detected by the preliminary screening, differing performances of the subsequent finer detection are obtained, depending on both the quality of data set and the detection method used: poorer results are observed using the plant data, in part due to an insufficient characterization of the fault events and in part due to the limited number of signals monitored. The higher availability of measurements provided by the numerical model, instead, enhances the discrimination capabilities of the tested methods, especially the nonlinear SVM, while the PCA-based approach and the Bayesian predictor results less affected by a change in the combination of diagnostic parameters used.

Transmission lines are protected by distance relays as main protection and overcurrent relays as backup protection. Both of distance relays and overcurrent relays have mal-operation during high Impedance faults, faults in the week in-feed systems and faults occurred during power swing. Fluctuations in voltage and current during power swing cause fluctuations in measured impedance by distance relay. In some measured impedance moves to protected zone and causes mal-operation. So, distance relay is blocked duration of a power swing. In cases that fault occurs during a power swing, the relay must detect fault inception. In this paper, a new algorithm for fault detection in cases with low fault current and fault detection during power swing is presented. Diagnostic criteria in this algorithm are the calculation of the spectral energy of current signal by the Adaptive Short-Time Fourier transform (ASTFT). Fault currents consist of a considerable amount of high-frequency components which they do not exist the in-line current during overload or power swing. The presented method is based on cooperation between existing fault detection methods and a fault detection using ASTST. Different conditions are Simulated in this paper in the software / EMTDC PSCAD and MATLAB to verify of correct performance of the proposed algorithm.

The major concern in a Controller Area Network based safety critical system are node failures, its early detection and failure tolerance. Aiming at the improvement of node fault detection timing response and reliability of a CAN based distributed system, this paper mainly focuses on a new fault detection and tolerance algorithm named Distributed Node Fault Detection and Tolerance [DNFDT] Algorithm for Controller Area Network. The purpose of the algorithm is to have an effective fault detection method for (n-1) node faults for n node system by reducing the fault detection cycle timings and by limited checking on fault free nodes. A four node experimental hardware platform was implemented and performance evaluation of the time elapsed for one complete fault detection cycle under varying CAN bus loads have been obtained.