Cointegration Modelling for Health and Condition Monitoring of Wind Turbines - An Overview

The cointegration method has recently attracted a growing interest from scientists and engineers as a promising tool for the development of wind turbine condition monitoring systems. This paper presents a short review of cointegration-based techniques developed for condition monitoring and fault detection of wind turbines. In all reported applications, cointegration residuals are used in control charts for condition monitoring and early failure detection. This is known as the residual-based control chart approach. Vibration signals and SCADA data are typically used with cointegration in these applications. This is due to the fact that vibration-based condition monitoring is one of the most common and effective techniques (used for wind turbines); and the use of SCADA data for condition monitoring and fault detection of wind turbines has become more and more popular in recent years.

Procedural Guide for System-Level Impact Evaluation of Industrial AI-Driven Technologies: Application to Risk-Based Investment Analysis for Condition Monitoring Systems in Manufacturing

Industrial artificial intelligence (IAI) and other analysis tools with obfuscated internal processes are growing in capability and ubiquity within industrial settings. Decision makers share concern regarding the objective evaluation of such tools and their impacts at the system level, facility level, and beyond. One application where this style of tool is making a significant impact is in Condition Monitoring Systems (CMSs). This paper addresses the need to evaluate CMSs, a collection of software and devices that alert users to changing conditions within assets or systems of a facility. The presented evaluation procedure uses CMSs as a case study for a broader philosophy evaluating the impacts of IAI tools. CMSs can provide value to a system by forewarning of faults, defects, or other unwanted events. However, evaluating CMS value through scenarios that did not occur is rarely easy or intuitive. Further complicating this evaluation are the ongoing investment costs and risks posed by the CMS from imperfect monitoring. To overcome this, an industrial facility needs to regularly and objectively review CMS impacts to justify investments and maintain competitive advantage. This paper's procedure assesses the suitability of a CMS for a system in terms of risk and investment analysis. This risk-based approach uses the changes in the likelihood of good and bad events to quantify CMS value without making any one-time pointwise estimates. Fictional case studies presented in this paper illustrate the procedure and demonstrate its usefulness and validity.

The Deterministic Nature of Sensor-Based Information for Condition Monitoring of the Cutting Process

Condition monitoring of the cutting process is a core function of autonomous machining and its success strongly relies on sensed data. Despite the enormous amount of research conducted so far into condition monitoring of the cutting process, there are still limitations given the complexity underlining tool wear; hence, a clearer understanding of sensed data and its dynamical behavior is fundamental to sustain the development of more robust condition monitoring systems. The dependence of these systems on acquired data is critical and determines the success of such systems. In this study, data is acquired from an experimental setup using some of the commonly used sensors for condition monitoring, reproducing realistic cutting operations, and then analyzed upon their deterministic nature using different techniques, such as the Lyapunov exponent, mutual information, attractor dimension, and recurrence plots. The overall results demonstrate the existence of low dimensional chaos in both new and worn tools, defining a deterministic nature of cutting dynamics and, hence, broadening the available approaches to tool wear monitoring based on the theory of chaos. In addition, recurrence plots depict a clear relationship to tool condition and may be quantified considering a two-dimensional structural measure, such as the semivariance. This exploratory study unveils the potential of non-linear dynamics indicators in validating information strength potentiating other uses and applications.

Advanced fluid models for resonant online oil condition monitoring systems

Two main topics are presented in this work which enable more efficient use of oil condition monitoring systems based on resonant fluid sensing. A new fluid model for a recently introduced compact measurement unit for oil condition monitoring based on simultaneous measurement of viscosity and density is discussed. It is shown that a new fluid model allows achieving higher accuracies, which is demonstrated by comparison to earlier models. The second topic deals with measuring fluid parameters over varying temperatures and thus providing additional monitoring parameters and enhanced data consistency. We propose an alternative representation of the Vogel model using transformed parameters having a clear physical meaning and which are more stable in presence of measurement noise.

Brake Health Prediction Using LogitBoost Classifier Through Vibration Signals

Brake is one of the crucial elements in automobiles. If there is any malfunction in the brake system, it will adversely affect the entire system. This leads to tribulation on vehicle and passenger safety. Therefore the brake system has a major role to do in automobiles and hence it is necessary to monitor its functioning. In recent trends, vibration-based condition monitoring techniques are preferred for most condition monitoring systems. In the present study, the performance of various fault diagnosis models is tested for observing brake health. A real vehicle brake system was used for the experiments. A piezoelectric accelerometer is used to obtain the signals of vibration under various faulty cases of the brake system as well as good condition. Statistical parameters were extracted from the vibration signals and the suitable features are identified using the effect of the study of the combined features. Various versions of machine learning models are used for the feature classification study. The classification accuracy of such algorithms has been reported and discussed.

Convolutional Neural Network for Wind Turbine Failure Classification Based on SCADA Data

As a renewable energy source and an alternative to fossil fuels, the wind power industry is growing rapidly. However, due to harsh weather conditions, wind turbines (WT) still face many failures that raise the price of energy produced and reduce the reliability of wind energy. Hence, the use of reliable monitoring and diagnostic systems of WTs is of great importance. Operation and maintenance expenses represent 30% of the total cost of large wind farms. The installation of offshore and remote wind farms has increased the need for efficient fault detection and condition monitoring systems. In this work, without using specific custom devices for monitoring conditions, but only increasing the sampling frequency in the sensors already available (in all commercial WT) of the supervisory control and data acquisition system (SCADA), datadriven multiple fault detection is performed, and a classification strategy is developed. The data is processed, and subsequently, using a convolutional neural network (CNN), six faults are classified and evaluated with different metrics. Finally, it should be noted that the classification speed allows the implementation of this strategy to monitor conditions online in real under-production WTs.

The Fault Detection of Gears of Electromechanical Power Transmission System using Frequency Domain Approach

Electric motors are widely used in the industry and convert electrical power to mechanical power. Gears are connected with the motor shaft for frictionless power transmission. The power transmission efficiency is dropped when the gear teeth is damaged. It causes huge financial loss. There are various commercially available condition monitoring tools for the preventive maintenance and condition monitoring of the gearing systems. However, such tools and techniques are intrusive and costly. This paper presents a non-intrusive approach for timely detection of gear damages and can be integrated to existing condition monitoring systems to save time and cost. The proposed method is based on the analysis of power spectra of the frequency amplitude plots. The amplitude difference at gear harmonics gives indication of the presence of the fault. The proposed technique has been tested and validated at 1400 rpm and 1370 rpm of the motor which is operating at 75 % load and full load.

Assessing Wear Evolutions in Railway Wheelsets Using a Survival Modeling Approach

The present paper provides a method to predict maintenance needs for the railway wheelsets by modeling the wear out affecting the wheelsets during its life cycle using survival analysis. Wear variations of wheel profiles are discretized and modelled through a censored survival approach, which is appropriate for modeling wheel profile degradation using real operation data from the condition monitoring systems that currently exist in railway companies. Several parametric distributions for the wear variations are modeled and the behavior of the selected ones is analyzed and compared with wear trajectories computed by a Monte Carlo simulation procedure. This procedure aims to test the independence of events by adding small fractions of wear to reach larger wear values. The results show that the independence of wear events is not true for all the established events, but it is confirmed for small wear values. Overall, the proposed framework is developed in such a way that the outputs can be used to support predictions in condition-based maintenance models and to optimize the maintenance of wheelsets.