Data tab of the Fault Tree Node Definition dialog
Entering probability and cost data for values and binary events is done on the Data tab of the Fault Tree Node Definition dialog. For fault trees that will not be embedded as modules, each basic event must have a probability specified for it which represents the probability of the event being in its true state (see below for fault trees designed to be embedded). A node may also have true cost and/or false cost data.
Entering Probability and Cost Data
- Double-click the value/binary event whose data you wish to edit. The Node Definition dialog appears with the Data tab active.
- For binary events, select the column corresponding to the data you wish to edit: Probability, True Cost, or False Cost.
- For binary events, enter a number.
- For values, you may enter a number or an expression. The expression can contain variable names, mathematical operators and/or numbers. The nodes corresponding to the variables used in the expression must be connected to the node. (You can build connections from formulas via the ribbon after you enter the expression.)
- Click OK when done.
Specified versus Calculated Probabilities
DPL must be able to evaluate the probability of each binary event in the fault tree. In a fault tree that is not going to be embedded as a module, each basic event must either have a number specified for the probability or be connected to another node. If an expression is specified for a binary event, it can only contain functions and numbers. A value may also use other values or events in an expression but they must be connected to the value. If you are developing a fault tree that will be embedded as a module, you may leave the probability data blank for basic events and supply the input data from the fault tree or decision model in which you will embed the module.
For derived events that are connected to a gate, module or event, the probability expression must be left blank. DPL calculates the event's probability using the logic of the subtree beneath it.
Using Time Series and Multiple Experts
Versions: DPL Fault Tree