Search

Fault detection and classification are considered as one of the most mandatory techniques in nowadays industrial monitoring. The necessity of fault monitoring is due to the fact that early detection can restrain high-cost maintenance. Due to the complexity of the wind turbines and the considerable amount of data available via SCADA systems, machine learning methods and specifically deep learning approaches seem to be powerful means to solve the problem of fault detection in wind turbines. In this article, a novel deep learning fault detection and classification method is presented based on the time-series analysis technique and convolutional neural networks (CNN) in order to deal with some classes of faults in wind turbine machines. To validate this approach, challenging scenarios, which consists of less than 5% performance reduction (which is hard to identify) in the two actuators or four sensors of the wind turbine along with sensors noise are investigated, and the appropriate structures of CNN are suggested. Finally, these algorithms are evaluated in simulation based on the data of a 4.8 MW wind turbine benchmark and their accuracy approves the convincing performance of the proposed methods. The proposed algorithm are applicable to both on-shore and off-shore wind turbine machines.

Machine learning, as a significant component of the Industrial Internet of Things (IIoT), has been widely applied in many fields. The continuously generated data from the various sensors are collected and stored, this is also known as a data stream. However, the non-stationary phenomenon in data stream, concept drift, is important to be detected immediately for the operation of the IoT system. Therefore, the detection method for concept drift is needed to alert the requirement to maintain or replace some components in advance, so as to avoid or mitigate the risk of malfunction of the IoT system. The majority of existing literature focuses on concept drift detection on binary classification. To fill this gap, here we propose an algorithm to detect multi-class. Moreover to improve the performance of detection, we also introduce an algorithm which integrates the existing error rate which is widely used with our newly proposed False Positive rate. The new method is called Drift Detection Method with False Positive rate for multi-label classification (DDM-FP-M). The DDM-FP-M firstly defines the false positive rate calculation method in multi-label classification, then integrates it with Drift Detection method with False positive rate (DDM-FP). The performance of the proposed method is evaluated through Intel Lab data and is found to outperform the Drift Detection method(DDM) over 50% cases.

Microscopic blood cell analysis is an important methodology for medical diagnosis, and complete blood cell counts (CBCs) are one of the routine tests operated in hospitals. Results of the CBCs include amounts of red blood cells, white blood cells and platelets in a unit blood sample. It is possible to diagnose diseases such as anemia when the numbers or shapes of red blood cells become abnormal. The percentage of white blood cells is one of the important indicators of many severe illnesses such as infection and cancer. The amounts of platelets are decreased when the patient suffers hemophilia. Doctors often use these as criteria to monitor the general health conditions and recovery stages of the patients in the hospital. However, many hospitals are relying on expensive hematology analyzers to perform these tests, and these procedures are often time consuming. There is a huge demand for an automated, fast and easily used CBCs method in order to avoid redundant procedures and minimize patients’ burden on costs of healthcare. In this research, we investigate a new CBC detection method by using deep neural networks, and discuss state of the art machine learning methods in order to meet the medical usage requirements. The approach we applied in this work is based on YOLOv3 algorithm, and our experimental results show the applied deep learning algorithms have a great potential for CBCs tests, promising for deployment of deep learning methods into microfluidic point-of-care medical devices. As a case of study, we applied our blood cell detector to the blood samples of COVID-19 patients, where blood cell clots are a typical symptom of COVID-19.

As the need for early fault detection increases day by day in large industries, the importance of a reliable fault diagnosis becomes more obvious. Moreover, sensors in industrial systems are prone to faults or malfunctions due to aging or accidents. Motivated by the above, in this study, a neural network sensor fault diagnosis approach is proposed and the stability and convergence of the algorithm are proven by using the robust estimation theorem and input-to-state stability (ISS). The proposed algorithm is applied to a wind turbine benchmark with 4.8 MW rated power. 10% to 30% of the sensor performance reduction is considered to illustrate the effective performance of the addressed algorithm.

Microscopic blood cell analysis is an important methodology for medical diagnosis, and complete blood cell counts (CBCs) are one of the routine tests operated in hospitals. Results of the CBCs include amounts of red blood cells, white blood cells and platelets in a unit blood sample. It is possible to diagnose diseases such as anemia when the numbers or shapes of red blood cells become abnormal. The percentage of white blood cells is one of the important indicators of many severe illnesses such as infection and cancer. The amounts of platelets are decreased when the patient suffers hemophilia. Doctors often use these as criteria to monitor the general health conditions and recovery stages of the patients in the hospital. However, many hospitals are relying on expensive hematology analyzers to perform these tests, and these procedures are often time consuming. There is a huge demand for an automated, fast and easily used CBCs method in order to avoid redundant procedures and minimize patients’ burden on costs of healthcare. In this research, we investigate a new CBC detection method by using deep neural networks, and discuss state of the art machine learning methods in order to meet the medical usage requirements. The approach we applied in this work is based on YOLOv3 algorithm, and our experimental results show the applied deep learning algorithms have a great potential for CBCs tests, promising for deployment of deep learning methods into microfluidic point-of-care medical devices. As a case of study, we applied our blood cell detector to the blood samples of COVID-19 patients, where blood cell clots are a typical symptom of COVID-19.

In this paper, a deep learning fault detection approach is proposed based on the convolutional neural network in order to cope with one class of faults in wind turbine systems. Fault detection is very vital in nowadays industries due to the fact that instantly detection can prevent waste of cost and time. Deep learning as one of the powerful approaches in machine learning is a promising method to identify and classify the intrigued problems, which are hard to solve by classical methods. In this case, less than 5% performance reduction in generator torque along with sensor noise, which is challenging to identify by an operator or classical diagnosis methods is studied. The proposed algorithm, which is evolved from convolutional neural network idea, is evaluated in simulation based on a 4.8 MW wind turbine benchmark and the accuracy of the results confirms the persuasive performance of the suggested approach.

It is well-recognised that to achieve long-term sustainable and resilient land management we need to understand the coupled dynamics of social and ecological systems. Land use change scenarios will often aim to understand (i) the behaviours of land management, influenced by direct and indirect drivers, (ii) the resulting changes in land use and (iii) the environmental implications of these changes. While the literature in this field is extensive, approaches to parameterise coupled systems through integration of empirical social science based models and ecology based models still need further development. We propose an approach to land use dynamics modelling based on the integration of behavioural models derived from choice experiments and spatially explicit systems dynamics modelling. This involves the specification of a choice model to parameterise land use behaviour and the integration with a spatial habitat succession model. We test this approach in an upland socio-ecological system in the United Kingdom. We conduct a choice experiment with land managers in the Peak District National Park. The elicited preferences form the basis for a behavioural model, which is integrated with a habitat succession model to predict the landscape level vegetation impacts. The integrated model allows us to create projections of how land use may change in the future under different environmental and policy scenarios, and the impact this may have on landscape vegetation patterns. We illustrate this by showing future projection of landscape changes related to hypothetical changes to EU level agricultural management incentives. The advantages of this approach are (i) the approach takes into account potential environmental and management feedbacks, an aspect often ignored in choice modelling, (ii) the behavioural rules are revealed from actual and hypothetical choice data, which allow the research to test the empirical evidence for various determinants of choice, (iii) the behavioural choice models generate probabilities of alternative behaviours which make them ideally suited for integration with simulation models. The paper concludes that the modelling approach offers a promising route for linking socio-economic and ecological features of socio-ecological systems. Furthermore, our proposed approach allows testing of the underlying socio-economic and environmental drivers and their interaction in real environmental systems.

There are wide range of cloud services commercially available. However there is limited research that investigates the strengths and weaknesses of their cost models in relation to different types of usage requirements. We propose a new costing model that systematically evaluates cloud services, and which combines compute, disk storage, and memory requirements. This paper demonstrates the proposed costing model on a data set that was derived from a real-world industrial data centre workload by calculating the precise cost of service provision from two leading cloud providers.