ISO 20414:2020 pdf free.Fire safety engineering一Verification and validation protocol for building fire evacuation models.
5 Verification
5.1 General
Verification is the process of determining that the implementation of a calculation method accurately represents the developer’s conceptual description of the calculation method and of the solution to the calculation method. This does not imply that the governing equations or rules are appropriate, only that they are being implemented and solved correctly and that the implementation accurately represents the developer’s conceptual description of the calculation method and of the solution to the calculation method.
This section presents a list of components for the verification of evacuation models for the study of human behaviour in building fires. The components subjected to verification are divided into different categories. The first category relates to the basic components that shall be verified in order to have a model able to represent basic evacuation scenarios. Some of these basic components have been defined in line with the work conducted by the International Maritime Organization in the context of the verification of evacuation models for passenger ships[2]. If one of these components is not present in the model, the associated verification test is not conducted, but the model tester shall clearly state that this component is missing. The second category relates to the behavioural components that a model may include in relation to its predictive behavioural capabilities. The third category refers to fire-people interaction components. These components deal with the representation of the interactions between people and the effects of fire (e.g. reduced visibility, toxicity). The last category refers to the components that a model may include in relation to the specific building/application in consideration. Evacuation models may include further components in addition to the list presented in this section. If additional components are available in an evacuation model, this shall be included in the verification testing procedure.
The basic components of all evacuation models shall be verified. Evacuation model testers may further verify the behavioural components, the fire/people interaction components and the building-specific components in relation to the stated field of applicability of the model under consideration.
Components can be categorized in different mannersWl. Core components relate to different functionalities, namely Pre-evacuation (P), Movement (M), Navigation/Route selection (N&R) and Flow conditions/constraints (F). The levels which are addressed in the components can be 1) Individual, 2) Aggregate or 3) Scenario level.
This document presents the components, as well as a list of possible tests that can be used for the verification of them. Different tests can be used by model testers if they are able to verify the components under consideration in a satisfactory manner (i.e. the tester shall provide an equivalent proof that the component has been verified). In some instances, evacuation models can require different test methods for the analysis of their results regarding the input variables required for their calibration. In these instances, a justification of the different test method shall be provided by the model tester.
Models may adopt two approaches when representing an emergent behaviour. This can be the result of a deterministic user-defined assumption for the outcome or the result of a predictive sub-model. On one hand, models employing a deterministic approach can be tested for verification only, i.e. no validation studies can be made since the simulated outcome is based on an a priori assumption. On the other hand, models including predictive sub-models can be tested using both verification and validation testing. For this reason, the model tester shall document what type of sub-model is being tested (deterministic user-defined or predictive).ISO 20414 pdf download.