Pramod Bangalore

An Artificial Neural Network Approach for Early Fault Detection of Gearbox Bearings

Gearbox has proven to be a major contributor toward downtime in wind turbines. The majority of failures in the gearbox originate from the gearbox bearings. An early indication of possible wear and tear in the gearbox bearings may be used for effective predictive maintenance, thereby reducing the overall cost of maintenance. This paper introduces a self-evolving maintenance scheduler framework for maintenance management of wind turbines. Furthermore, an artificial neural network (ANN)-based condition monitoring approach using data from supervisory control and data acquisition system is proposed. The ANN-based condition monitoring approach is applied to gearbox bearings with real data from onshore wind turbines, rated 2 MW, and located in the south of Sweden. The results demonstrate that the proposed ANN-based condition monitoring approach is capable of indicating severe damage in the components being monitored in advance.

An approach for self evolving neural network based algorithm for fault prognosis in wind turbine

In recent years Supervisory Control and Data Acquisition (SCADA) system has been used to monitor the condition of wind turbine components. SCADA being an integral part of wind turbines comes at no extra cost and measures an array of signals. This paper proposes to use artificial neural networks (ANN) algorithm for analysis of SCADA data for condition monitoring of components. The first step to build an ANN model is to create the training data set. Here an automated process to decide the training data set has been presented. The approach reduces the number of samples in the training data set compared to the conventional method of hand picking the data set. Further the approach describes how the ANN model could be kept in tune with the changes in the operating conditions of the wind turbine by updating the ANN model. The fault prognosis obtained from the model can be used to optimize the maintenance scheduling activity.

Self Evolving Neural Network Based Algorithm for Fault Prognosis in Wind Turbines : A Case Study

Asset management of wind turbines has gained increased importance in recent years. High maintenance cost and longer downtimes of wind turbines have led to research in methods to optimize maintenance activities. Condition monitoring systems have proven to be a useful tool towards aiding maintenance management of wind turbines. Methods using Supervisory Control and Data Acquisition (SCADA) system along with artificial intelligence (AI) methods have been developed to monitor the condition of wind turbine components. Various researchers have presented different artificial neural network (ANN) based models for condition monitoring of components in a wind turbine. This paper presents an application of the approach to decide and update the training data set needed to create an accurate ANN model. A case study with SCADA data from a real wind turbine has been presented. The results show that due to a major maintenance activity, like replacement of component, the ANN model has to be re-trained. The results show that application of the proposed approach makes it possible to update and re-train the ANN model.

Cost efficient maintenance strategies for wind power systems using LCC

Maintenance costs for wind power plants are a significant part of the total life cycle cost, especially for offshore wind power plants, situated at remote sites. In order to decrease the cost of maintenance, monitoring systems have been used to estimate the condition of critical components in wind turbines. This paper proposes Life Cycle Cost analysis (LCC) approach for maintenance management of wind turbines. The LCC approach for maintenance management presented in this paper is an extension on previous work by J. Nilsson and L. Bertling, where a comparison has been made with this previous work and the same is extended with new data and models. Case studies are presented based on data from three different wind turbines rated 3 and 6MW. Three different scenarios have been studied and the effect of condition monitoring system (CMS) has been analysed. For any chosen value the CMS proves to be a profitable option.

Extension of test system for distribution system reliability analysis with integration of Electric Vehicles in distribution system

Over the years the electric power system has seen an exponential growth in terms of size and technology. A similar growth and development has taken place in the terms of probabilistic applications used to analyze the power systems. A literature survey was done to analyze the existing widely used test systems; IEEE RTS and RBTS [5]. Conclusions were drawn towards how these test systems could be updated or modified to be sufficient with respect to the modern power systems “Smart Grid”. Extensions to Bus-2 distribution system of RBTS have been proposed to include the integration of Electric Vehicles into the distribution system. A sample study investigating the effect of Vehicle to Grid supply on the Energy Not Supplied in the system has been carried out; the results for the same have been presented. The aim of the work is to be able to provide a test system with electric vehicles for probabilistic reliability applications.

On the use of reliability test systems: A literature survey

Over the years the reliability test systems such as the IEEE Reliability Test System (RTS) and Roy Billinton Test System (RBTS) have been used extensively by researchers, as a bench mark system, for reliability assessment and other developments in the field of probabilistic applications in power systems. This paper presents an extensive literature survey of previous publications in which the RTS and RBTS or other test systems have been used. From the survey, several observations can be made, such as on the purpose of the use of the test systems, or where the studies were made. With the development of the electric power system, both in size and technology, this survey serves as a basis for assessing the appropriateness of the existing IEEE RTS system to address these developments. Such developments would mainly include wind energy, increased use of HVDC transmission and the state-of-the-art communication systems applied to power systems. The survey shows that it is necessary to extend the existing test systems to better address the demands of future electric power system.

Data-driven algorithm for throughput bottleneck analysis of production systems

ABSTRACT The digital transformation of manufacturing industries is expected to yield increased productivity. Companies collect large volumes of real-time machine data and are seeking new ways to use it in furthering data-driven decision making. A challenge for these companies is identifying throughput bottlenecks using the real-time machine data they collect. This paper proposes a data-driven algorithm to better identify bottleneck groups and provide diagnostic insights. The algorithm is based on the active period theory of throughput bottleneck analysis. It integrates available manufacturing execution systems (MES) data from the machines and tests the statistical significance of any bottlenecks detected. The algorithm can be automated to allow data-driven decision making on the shop floor, thus improving throughput. Real-world MES datasets were used to develop and test the algorithm, producing research outcomes useful to manufacturing industries. This research pushes standards in throughput bottleneck analysis, using an interdisciplinary approach based on production and data sciences. GRAPHICAL ABSTRACT

A data-driven algorithm to predict throughput bottlenecks in a production system based on active periods of the machines

Abstract Smart manufacturing is reshaping the manufacturing industry by boosting the integration of information and communication technologies and manufacturing process. As a result, manufacturing companies generate large volumes of machine data which can be potentially used to make data-driven operational decisions using informative computerized algorithms. In the manufacturing domain, it is well-known that the productivity of a production line is constrained by throughput bottlenecks. The operational dynamics of the production system causes the bottlenecks to shift among the production resources between the production runs. Therefore, prediction of the throughput bottlenecks of future production runs allows the production and maintenance engineers to proactively plan for resources to effectively manage the bottlenecks and achieve higher throughput. This paper proposes an active period based data-driven algorithm to predict throughput bottlenecks in the production system for the future production run from the large sets of machine data. To facilitate the prediction, we employ an auto-regressive integrated moving average (ARIMA) method to predict the active periods of the machine. The novelty of the work is the integration of ARIMA methodology with the data-driven active period technique to develop a bottleneck prediction algorithm. The proposed prediction algorithm is tested on real-world production data from an automotive production line. The bottleneck prediction algorithm is evaluated by treating it as a binary classifier problem and adapted the appropriate evaluation metrics. Furthermore, an attempt is made to determine the amount of past data needed for better forecasting the active periods.