Safety passenger transportation is more important than efficiency or reliability. Therefore, it is vital to maintain the proper condition of the equipment related to the passengers’ comfort and safety. This manuscript presents the methodology of complete development and implementation of both hybrid model and digital twin 3.0 for an HVAC in railways. The objective of this is to monitor the condition of the HVAC where it matters to the comfort and safety of the passengers in the trains. The level 3.0 of digital twin will be developed for the diagnosis and prognosis of HVAC by using hybrid modeling. The description illustrated in this paper is focused on the methodology used to implement a hybrid model-based approach, and both the need and advantages of using hybrid model approaches instead of data-based approaches. The development considers the importance of safety and environmental risks, which are included in the risk quantification of failure modes. Railway’s maintainers replace critical components in early stages of degradation; thus, the use of a data-driven model loses essential information related to advanced stages of degradation which might decrease the accuracy of the maintenance instructions provided. Physics-based model can be used to generate synthetic data to overcome the lack of data in advanced stages of degradation, and then, the synthetic data can be combined with the real data, which is collected by sensor located in the real system, to build the data-driven model. The combination leads to form hybrid-model based approach with a large number of failure modes that were unpredictable. Finally, the outcome is beneficial for the proper functioning of systems; hence, safety of the passengers. Introduction The diagnostics and prognostics are the main processes in condition-based maintenance (CBM). These processes are broadly used in industrial assets, mainly, at levels of part, components, sub-systems, and systems. Fault detection and diagnostics (FDD) is the identification of a faulty component through the detection and isolation of a fault. Once a fault appears in a system, diagnostics process might detect that fault and identify the faulty part. Prognostics process is performed for estimating the remaining useful life (RUL) of a system. It is estimated using the behavior data while the system works, allowing maintainers to avoid the corrective maintenance [1]. Thus, appropriate implementation of CBM leads to reduce costs and increase the reliability and safety of systems. As it is shown in Figure 1, there are highlighted four techniques for estimating the RUL and, on a consequence, for implementing FDD and prognostics processes [1]: experience-based approaches, model-based approaches, data-driven approaches, and hybrid modelbased approaches (HyMAs). Figure 1: Remaining useful life (RUL) estimation models. SNE 31(3), 2021, 121-126, DOI: 10.11128/sne.31.tn.10572 Received: March 10, 2021 (Selected EUROSIM 2019 Postconf. Publ.), Revised: August 30, 2021; Accepted: September 2, 2021 SNE Simulation Notes Europe, ARGESIM Publisher Vienna, ISSN Print 2305-9974, Online 2306-0271, www.sne-journal.org Gálvez et al. Hybrid Models and Digital Twins for Condition Monitoring: HVAC System for Railway 122 SNE 31(3) – 9/2021 T N The model-based approach are mathematical models of the physical system [2]. These models incorporate such characteristics as material properties, thermodynamic, and mechanical responses. Furthermore, different operational conditions, components defect, and degradation must be modelled for fault detection. Thus, these models accurately describe the behaviour of systems [3] [4]. Data-driven approaches use operational data collected by sensors embedded in the system for obtaining information related to the degradation state of components. Thus, these approaches can trace the relationship between the data acquired from the system with its degradation [5] [6]. As Figure 1 shows, HyMAs is the combination of data-driven and model-based approaches. The main advantage of this combination is the reduction in both the historical information required to train a data-driven model and the information needed for a robust physicsbased model. This data fusion aims to improve FDD and prognostics processes. The model-based approaches generates synthetic data to overcome the lack of historical data, thus improving the ability of a data-driven apprach to detect failure modes (FMs) and reducing the appearance of hidden FMs, metaphorically known as “black swan losses” [7]. 1 Problem Discussion The case study used in this research work is a heating, ventilation and air conditioning (HVAC) system which is installed in a high-speed passenger train. Currently, many types of research are based on datadriven approaches but missing important information and failures. This occurs because maintainers replace critical elements in early stages of degradation for safety, reliability, economic, and environmental reasons. Therefore, it is challenging to acquire data in faulty stages of the system and advanced stages of degradation [3]. The solution proposed in this manuscript is to transfer the methodology to railway subsystems to scale, develop, and validate an HyMAs for railway equipment through the combination of modeling based on physical laws and modeling based on data. As a primary purpose, it will be developed to improve the diagnostics and prognostics processes of components of railway systems. The outcome will be to provide railway companies with the necessary tools to reach a level 3 of digital twin based on HyMA. The Digital twin levels are listed in Table 1 and shown in Figure 2, Figure 3, and Figure 4. Digital Twin Data Layer Level 1.0 SCADA Data Online Data OT (SCADA) Operational Alarms based on anomalies 2.0 Adding: Offline data, GMAO, ERP, taxonomy, ontology management data OT (SCADA) and IT (ERP, management data...) Tactical Planning of corrective maintenance instructions. 2.1 Context data Operative context 3.0 Physics model Hybrid digital twin OT + IT + Engineering technology Strategic Renew of assets, performance improvements Table 1: Digital twins levels description. Figure 2: Digital twin 2.0, extracted from [8]. Figure 3 shows the Digital Twin 2.1, who combines the Digital Twin 2.0 (see Figure 2) and context data. Context-awareness is considered as an application’s ability to adapt to changing circumstances and respond according to the context of use. Figure 3: Digital twin 2.1, extracted from [8]). Figure 4 shows the Digital Twin 3.0, who combines the Digital Twin 2.0 with a HyMA. It could be used from component to system level. Gálvez et al. Hybrid Models and Digital Twins for Condition Monitoring: HVAC System for Railway SNE 31(3) – 9/2021 123 T N Figure 4: Digital twin 3.0, extracted from [8].
[1]
Joseph Mathew,et al.
Rotating machinery prognostics. State of the art, challenges and opportunities
,
2009
.
[2]
Marcantonio Catelani,et al.
Architecture for hybrid modelling and its application to diagnosis and prognosis with missing data
,
2017
.
[3]
Marcantonio Catelani,et al.
Context awareness for maintenance decision making: A diagnosis and prognosis approach
,
2015
.
[4]
Biswajit Basu,et al.
Residential HVAC fault detection using a system identification approach
,
2017
.
[5]
Prashant J. Shenoy,et al.
WattHome: A Data-driven Approach for Energy Efficiency Analytics at City-scale
,
2018,
KDD.
[6]
Dawn An,et al.
Practical options for selecting data-driven or physics-based prognostics algorithms with reviews
,
2015,
Reliab. Eng. Syst. Saf..
[7]
Terje Aven,et al.
On the meaning of a black swan in a risk context
,
2013
.