ProcedureActionTimeBranches.txt
Purpose
The ProcedureActionTimeBranches.txt input file allows multiple accident sequence branches to be generated when a procedure action is executed. The ADS-IDAC procedure following model allows the analyst to specify the time taken to execute each step action using a three-parameter Weibull probability density distribution. This file controls when and how many action execution time branches are generated to capture crew-to-crew variability.
- When only one action time branch is generated, the required action execution time equals the mean value of the Weibull distribution.
- When more than one branch is generated, the probability distribution is partitioned into segments and the time for each sequence branch is determined by the mean value over the associated partition.
Weibull Partition Method (Figure 19, Equation 11)
The partition boundaries are determined by dividing the probability range of the Weibull cumulative distribution function (CDF) into a number of segments equal to the number of desired branches. Each branch receives an equal probability mass from the CDF. For example, five branches produce five equal-probability segments with partition boundaries at CDF increments of 0.2.
The execution time for each branch is the mean time over its associated partition. For the first and second partitions (Equation 11):
Where
INFO
Branching rules are tied to a specific combination of procedure, step, and action. Executing the same action from a different procedure step will not trigger branching.
Input File Format
Each entry is one line — procedure_name step_name action_name number_of_branches (matching the shipped BV4 file; the parser is token-based, so line breaks and other whitespace are interchangeable):
Procedure_Action_Time_Branches number_of_items
procedure_name_1 step_name_1 action_name_1 number_of_branches_1
.
.
procedure_name_n step_name_n action_name_n number_of_branches_nParameter Descriptions
| Parameter | Type | Description | Example Value |
|---|---|---|---|
number_of_items | Integer, ≥ 0 | Number of action time branching events in the file. Setting to 0 disables all branching without requiring deletion of entries. | 2 |
procedure_name | String | Procedure associated with the branching rule. Must match a valid procedure_name in Procedures.txt. | MPBG_SG_A_Increase_FWRV |
step_name | String | Procedure step associated with the branching rule. Must match a valid step_name in Procedures.txt. | Step_1 |
action_name | String | Control panel action subject to time branching. Should match a control_name in ControlPanel.txt. The name is not validated when the file is read; if the control is never executed by the specified step, no branches are generated and no error is reported. Time delay steps use the reserved word Time. | X_SG_A_FWRV |
number_of_branches | Integer, ≥ 1 | Number of event sequence branches to generate. With 1, no branching occurs — the action time is the mean of the Weibull distribution. | 2 |
TIP
Setting number_of_items to 0 disables all branching events without needing to delete entries. To activate only one branching event, move it to the top of the list and set number_of_items to 1. Only the first number_of_items entries are read — the BV4 reference file ships with a count of 0, so the example rows listed below its header are inert.
Execution
Each alternative defined here is opened as a child sequence that restarts from the branch point with that alternative applied. See How a Branch Alternative Executes.
Sample Input
Procedure_Action_Time_Branches 2
MPBG_SG_A_Increase_FWRV Step_1 X_SG_A_FWRV 2
E_3 Briefing_Hold_1 Time 3Sample Explanation
When action
X_SG_A_FWRVis executed by procedureMPBG_SG_A_Increase_FWRV, stepStep_1, 2 time branches are generated to model variability in the specified action time. Execution ofX_SG_A_FWRVby any other procedure step will not generate branching events.When the time delay action
Timeis executed by procedureE_3, stepBriefing_Hold_1, 3 action time branches are generated. This branching rule produces three delay times — a short time, a nominal time, and a long time — based on the time probability distribution specified for the action.