The Process of Verification and Validation of Building Fire Evacuation Models

Abstract

To date, there is no International standard on procedures and tests to assess the verification and validation (V&V) of building fire evacuation models. Often it is the case that model testers adopt inconsistent procedures, or tests designed for other model uses or they do not test them for all features embedded in their model. For instance, the tests presented within the MSC/Circ.1238 (Guidelines for evacuation analysis for new and existing passenger ships) provided by the International Maritime Organization are often employed for the V&V of models outside their original context of use (e.g. building fires instead of maritime applications). This document is intended to open a discussion on the main issues associated with the definition of a standard procedure for the V&V of building fire evacuation models. A review of the current procedures, tests and methods available in the literature to assess the V&V of building evacuation models is provided. The capabilities of building evacuation models are evaluated by studying their five main core components, namely 1) pre-evacuation time, 2) movement and navigation, 3) exit usage, 4) route availability and 5) flow constraints. A set of tests and recommendations about the verification and validation of building evacuation models is proposed. These tests include suggestions on using simple tests of emergent behaviours together with examples of experimental data-sets suitable for the analysis of different core components. The uncertainties associated with evacuation modelling are discussed. In particular, a method for the analysis of behavioural uncertainty (uncertainty due to the use of distributions or stochastic variables to simulate human behaviour in evacuation modelling) is presented. The method consists of a set of convergence criteria based on functional analysis. The last part of this document presents a discussion on the issues associated with the definition of the acceptance criteria of a standard V&V protocol.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 137 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.