Define system, model, simulation

Definitions

• System: a collection of entities that act and interact together to achieve some logical end. Example: a hospital outpatient department with patients, doctors, nurses and rooms.
• Model: a representation of a system, usually simpler than the system, used to study it. It keeps the aspects that matter for the purpose of the study and leaves out the rest. Example: a queue with one arrival stream and several servers that represents the outpatient department.
• Simulation: the imitation of the operation of a real system over time by experimenting with its model. Example: running the queue model on a computer for a simulated month, with random patient arrivals and service times, and recording the average waiting time.

Example

A bank has two tellers. System: customers, tellers, queue. Model: arrivals are random with a mean gap of 2 minutes, service takes about 3 minutes, and one queue feeds two tellers. Simulation: generate arrival and service times, step the clock through events, and measure average waiting time and teller utilization. The bank can then see the effect of a third teller before hiring one.

Advantages of simulation

• Tests changes without disturbing the real system.
• Time can be compressed or expanded.
• Gives insight into interactions and bottlenecks.
• Works for complex stochastic systems with no analytical solution.
• Cheaper and safer than real experiments; good for training.

Disadvantages of simulation

• Needs special training to build a good model.
• Output is random and can be difficult to interpret.
• Can be costly and slow to develop and run.
• May be used when a simple analytical solution is available.
• Results are only as good as the model and its data.