High-Level Processing
During high-level processing for simulation, the cumulative number of system successes, downtime of the system, and spares cost for the system are all considered. The basic steps in the high-level processing follow.
1. Simulate the system for a specified number of times (simulation runs).
2. Find the total (cumulative) number of system successes (uninterrupted services), the total downtime of the system, and the total spares cost.
3. Find the average by dividing the total number of system successes by the number of simulation runs. This provides the expected reliability value:
◦ Reliability = (Total number of system successes) / (Simulation runs)
◦ Expected downtime of the system = (Total downtime of the system) / (Simulation runs)
◦ Expected spares cost = (Total spares cost) / (Simulation runs)
If reliability is 0 or 1 and theoretical reliability is less than 1, when you force simulation, you should increase the number of simulations to get a minimum number of failures or successes. If the desired accuracy is not obtained, increasing the number of simulations even further will provide more precise results.
4. Find other related results. Examples follow:
◦ Confidence intervals
◦ Availability (which is calculated directly)
◦ Hazard rate