Data-driven anomaly detection in cyber-physical production systems

2015 ◽  
Vol 63 (10) ◽  
Author(s):  
Oliver Niggemann ◽  
Christian Frey
Keyword(s):  
Anomaly Detection ◽  
Production Systems ◽  
Data Driven ◽  
Plant Analysis ◽  
Product Complexity ◽  
Automation Systems ◽  
Data Driven Approach ◽  
Assistance Systems ◽  
Cyber Physical Production Systems ◽  
Intelligent Assistance

AbstractDue to global competition and increasing product complexity, the complexity of production systems has grown significantly in recent years. This places an increasing burden on automation developers, systems engineers and plant constructors. Intelligent assistance systems and smart automation systems are a possible solution to face this complexity: The machines, i.e. the software and assistance systems, take over tasks that were previously carried out manually by experts. At the heart of this concept are intelligent anomaly detection approaches based on models of the system behaviors. Intelligent assistance systems learn these models automatically: Based on data, these systems extract most necessary knowledge about the diagnosis task. This paper outlines this data-driven approach to plant analysis using several use cases from industry.

A comparative study of data-driven and physics-based gas turbine fault recognition approaches

2021 ◽  
pp. 1748006X2198964
Author(s):  
Juan Luis Pérez-Ruiz ◽  
Igor Loboda ◽  
Iván González-Castillo ◽  
Víctor Manuel Pineda-Molina ◽  
Karen Anaid Rendón-Cortés ◽  
...  
Keyword(s):  
Anomaly Detection ◽  
Gas Turbine ◽  
Recognition Accuracy ◽  
Data Driven ◽  
Fault Identification ◽  
Fault Recognition ◽  
Data Driven Approach ◽  
Diagnostic Approaches ◽  
Artificial Neural Network Ann ◽  
Two Stages

The present paper compares the fault recognition capabilities of two gas turbine diagnostic approaches: data-driven and physics-based (a.k.a. gas path analysis, GPA). The comparison takes into consideration two differences between the approaches, the type of diagnostic space and diagnostic decision rule. To that end, two stages are proposed. In the first one, a data-driven approach with an artificial neural network (ANN) that recognizes faults in the space of measurement deviations is compared with a hybrid GPA approach that employs the same type of ANN to recognize faults in the space of estimated fault parameter. Different case studies for both anomaly detection and fault identification are proposed to evaluate the diagnostic spaces. They are formed by varying the classification, type of diagnostic analysis, and deviation noise scheme. In the second stage, the original GPA is reconstructed replacing the ANN with a tolerance-based rule to make diagnostic decisions. Here, two aspects are under analysis: the comparison of GPA classification rules and whole approaches. The results reveal that for simple classifications both spaces are equally accurate for anomaly detection and fault identification. However, for complex scenarios, the data-driven approach provides on average slightly better results for fault identification. The use of a hybrid GPA with ANN for a full classification instead of an original GPA with tolerance-based rule causes an increase of 12.49% in recognition accuracy for fault identification and up to 54.39% for anomaly detection. As for the whole approach comparison, the application of a data-driven approach instead of the original GPA can lead to an improvement of 12.14% and 53.26% in recognition accuracy for fault identification and anomaly detection, respectively.

Designing And Optimizing Production In A High Variety / Low Volume Environment Through Data-Driven Simulation

2021 ◽  
Author(s):  
Virginia Fani ◽  
Bianca Bindi ◽  
Romeo Bandinelli
Keyword(s):  
Simulation Model ◽  
Production Systems ◽  
Discrete Event ◽  
Simulation Models ◽  
Product Variety ◽  
Data Driven ◽  
Business Decisions ◽  
Footwear Industry ◽  
Data Driven Approach ◽  
High Level

HVLV environments are characterized by high product variety and small lot production, pushing companies to recursively design and optimize their production systems in a very short time to reach high-level performance. To increase their competitiveness, companies belonging to these industries, often SMEs working as third parties, ask for decision-making tools to support them in a quick and reactive reconfiguration of their production lines. Traditional discrete event simulation models, widely studied in the literature to solve production-related issues, do not allow real-time support to business decisions in dynamic contexts, due to the time-consuming activities needed to re-align parameters to changing environments. Data-driven approach overcomes these limitations, giving the possibility to easily update input and quickly rebuild the model itself without any changes in the modeling code. The proposed data-driven simulation model has also been interfaced with a commonly-used BI tool to support companies in the iterative comparison of different scenarios to define the optimal resource allocation for the requested production plan. The simulation model has been implemented into a SME operating in the footwear industry, showing how this approach can be used by companies to increase their performance even without a specific knowledge in building and validating simulation models.

scholarly journals Early anomaly detection and localisation in distribution network: a data-driven approach

2020 ◽  
Vol 14 (18) ◽  
pp. 3814-3825
Author(s):  
Xin Shi ◽  
Robert Qiu ◽  
Xing He ◽  
Zenan Ling ◽  
Haosen Yang ◽  
...  
Keyword(s):  
Anomaly Detection ◽  
Distribution Network ◽  
Data Driven ◽  
Data Driven Approach
Sign in / Sign up

Export Citation Format

Share Document