Best Practice Data Sharing Guidelines for Wind Turbine Fault Detection Model Evaluation

Abstract

In this paper, a set of best practice data sharing guidelines for wind turbine fault detection model evaluation is developed, which can help practitioners overcome the main challenges of digitalisation. Digitalisation is one of the key drivers for reducing costs and risks over the whole wind energy project life cycle. One of the largest challenges in successfully implementing digitalisation is the lack of data sharing and collaboration between organisations in the sector. In order to overcome this challenge, a new collaboration framework called WeDoWind was developed in recent work. The main innovation of this framework is the way it creates tangible incentives to motivate and empower different types of people from all over the world to share data and knowledge in practice. In this present paper, the challenges related to comparing and evaluating different SCADA-data-based wind turbine fault detection models are investigated by carrying out a new case study, the “WinJi Gearbox Fault Detection Challenge”, based on the WeDoWind framework. A total of six new solutions were submitted to the challenge, and a comparison and evaluation of the results show that, in general, some of the approaches (Particle Swarm Optimisation algorithm for constructing health indicators, performance monitoring using Deep Neural Networks, Combined Ward Hierarchical Clustering and Novelty Detection with Local Outlier Factor and Time-to-failure prediction using Random Forest Regression) appear to exhibit high potential to reach the goals of the Challenge. However, there are a number of concrete things that would have to have been done by the Challenge providers and the Challenge moderators in order to ensure success. This includes enabling access to more details of the different failure types, access to multiple data sets from more wind turbines experiencing gearbox failure, provision of a model or rule relating fault detection times or a remaining useful lifetime to the estimated costs for repairs, replacements and inspections, provision of a clear strategy for training and test periods in advance, as well as provision of a pre-defined template or requirements for the results. These learning outcomes are used directly to define a set of best practice data sharing guidelines for wind turbine fault detection model evaluation. The guidelines can be used by researchers in the sector in order to improve model evaluation and data sharing in the future.

On the Issue of the Creation and Functioning of Energy Efficiency Management Systems for Technological Processes of Mining Enterprises

“…Nevertheless, due to the nonlinearity of processes and the stochastic nature of possible events-accidents and repairs of ECO, weather and climatic phenomena-the construction of a forecast and a plan for energy consumption should proceed not only from statistically observed values, but also from a dynamic assessment of the current situation, taking into account a complex set of loosely coupled factors. In general, such solutions are possible using data mining and machine learning methods, which has been confirmed in a large volume of scientific and engineering works [5][6][7]16,[29][30][31][32]. The formation of the initial feature space of the training sample, although it may cause difficulties, should not be a significant problem in the way of applying such methods.…”

“…In our earlier works, we gave both a justification for the need for such requirements and an example of their implementation in the form of a reference meta-model of the architecture of an industrial enterprise, DEA 1.0 (Digital Enterprise Architecture) [33]. The proposed model of data-centric microservice architecture formalizes the ways of building functional components of the system and the processes of information exchange between such components through specialized services, conventionally combined by the general term "Digital Platform", which is also consistent with the results of scientific and engineering research in international and domestic practice [24][25][26]29,31,34].…”

“…However, the latter thesis can be more applicable when it comes to creating some high-precision models or a Digital Twin. In cases of building dispatching automated systems with a sufficient degree of accuracy (meaning permissible error) and reliability, the computational models included in such systems can be formed with a high generalizing ability, i.e., in such a way that they will be applicable to the same type of energy-consuming objects of different enterprises, regardless of their initial specific technical-technological and cumulative operational properties [5][6][7]22,31]. This means that for this type of energyconsuming object (ECO), such as excavators of the same series and brand, it is possible to build a unified model for predicting energy consumption with a unified list of input parameters and invariant to the specific initial characteristics of each individual excavator at each individual enterprise.…”

“…However, for this purpose, the optimization criteria and the corresponding control parameters must be formulated considering the heterogeneity of the types of energy resources consumed and the energy consumption objects themselves, which is a very difficult task that can be considered an independent scientific study. Undoubtedly, in one way or another, the tasks of dynamic power management are widely known, and there are quite a few high-class solutions [5,6,11,12,17,23,31,35,36]. However, for mining industries, this type of energy is not the only one, and in some cases, it may not even be the main one the type of resource.…”