Identification of representative operating conditions of HVAC systems in passenger rail vehicles based on sampling virtual train trips. Issue 2 (April 2016)

• Record Type: Journal Article
• Authors: Luger, Christian; Kallinovsky, Johann; Rieberer, René
• Is Part Of: Advanced engineering informatics. Volume 30:Issue 2(2016:Apr.)
• Page Start: 157
• Page End: 167
• Publication Date: 2016-04
• Subjects: Rail vehicle thermal loads -- Rail HVAC load cycles -- Representative operating conditions -- Monte-Carlo-based data sampling -- General transit feed specification; Computer-aided engineering -- Periodicals; Engineering -- Data processing -- Periodicals; 620.00285
• DOI: 10.1016/j.aei.2016.02.006
• Languages: English
• ISSNs: 1474-0346

Abstract

Graphical abstract: Highlights: Monte-Carlo-simulation for sampling virtual train trips. Use of GTFS feeds on train schedules and trips as simulation input. Application of an acausal thermal rail vehicle model. Tool to identify HVAC operating conditions for stationary and dynamic simulation. Abstract: Simulation-driven development and optimization of heating, ventilation and air-conditioning (HVAC) systems in passenger rail vehicles is of growing relevance to further increase product quality and energy efficiency. However, today required knowledge of realistic operating conditions is mostly unavailable. This work introduces methodologies and tools to identify representative operating conditions of HVAC systems in passenger rail vehicles. First a Monte-Carlo-simulation approach was employed to acquire a large set of close-to-reality HVAC operating conditions based on simulated train trips. Sampling simulated train trips bypassed the issue of unavailability of appropriate real-world data. Furthermore the approach allowed high flexibility in considering HVAC-relevant factors associated to different categories of trains, rail networks, operation profiles and meteorological conditions. Second, algorithms and methodologies such as k -means clustering and an adapted Finkelstein–Schafer statistical method were implemented to identify representative HVAC operating conditions from the sampled dataset. Final results comprise a set of time-independent HVAC operating points with associated frequencies of occurrence (ROC-points) as well as a set of time-domain signals for representative days (ROC-signals). These results are input for stationary or dynamic system-level simulations, which are used to support design decisions. The developed methodology was exemplarily applied to urban/suburban trains in Switzerland.