Condition monitoring plays a significant role in the safety and reliability
of modern industrial systems. Artificial intelligence (AI) approaches are
gaining attention from academia and industry as a growing subject in industrial
applications and as a powerful way of identifying faults. This paper provides
an overview of intelligent condition monitoring and fault detection and
diagnosis methods for industrial plants with a focus on the open-source
benchmark Tennessee Eastman Process (TEP). In this survey, the most popular and
state-of-the-art deep learning (DL) and machine learning (ML) algorithms for
industrial plant condition monitoring, fault detection, and diagnosis are
summarized and the advantages and disadvantages of each algorithm are studied.
Challenges like imbalanced data, unlabelled samples and how deep learning
models can handle them are also covered. Finally, a comparison of the
accuracies and specifications of different algorithms utilizing the Tennessee
Eastman Process (TEP) is conducted. This research will be beneficial for both
researchers who are new to the field and experts, as it covers the literature
on condition monitoring and state-of-the-art methods alongside the challenges
and possible solutions to them.
Condition monitoring is a crucial aspect of maintaining safety and reliability in industrial systems. With the rise of artificial intelligence (AI) approaches, there is growing interest in using these techniques for fault detection and diagnosis in industrial plants. This paper focuses on the Tennessee Eastman Process (TEP) as a benchmark for intelligent condition monitoring.
The paper provides an overview of both deep learning (DL) and machine learning (ML) algorithms that are popular and state-of-the-art in the field. DL algorithms, such as neural networks, have shown great potential in handling complex and unstructured data, while ML algorithms, such as support vector machines, have proven to be effective in classification tasks. By summarizing these algorithms, the paper offers a comprehensive understanding of the tools available for condition monitoring.
One of the key challenges in condition monitoring is dealing with imbalanced data and unlabelled samples. Imbalanced data occurs when certain classes or categories are underrepresented in the dataset, which can lead to biased models. Unlabelled samples, on the other hand, pose a challenge when traditional supervised learning methods require labeled data for training. The paper addresses these challenges and explores how deep learning models can handle them.
A noteworthy aspect of this research is its multidisciplinary nature. It combines concepts from AI, industrial engineering, and data science to provide a comprehensive understanding of condition monitoring. By examining the accuracies and specifications of different algorithms using the TEP benchmark, the paper also allows for a comparison of performance.
Overall, this research contributes to the field of condition monitoring by providing a literature review of state-of-the-art methods and addressing the challenges faced in implementing these techniques. It is valuable not only for researchers new to the field but also for experts looking to stay updated with the latest advancements. The use of open-source benchmarks like TEP further enhances the reproducibility and comparability of research in this area.