Modeling is a method of solving problems in which the system under study is replaced by a simpler object, which in turn describes a real model. There are cases when it is unacceptable or pointless to conduct an experiment on real situations due to the fragility, or the high cost of creating a prototype, or the long time of the experiment. It is in such situations that modeling is applied (Wainer & Mosterman, 2018). A simulation model is a computer program that describes the design and recreates the behavior of a real system over time (Wainer & Mosterman, 2018). It makes possible to get a detailed simulation model which provides statistics about different aspects of the system, which is due to different input data.
The use of simulation models provides many advantages:
- Price. For example, a reduction in the number of jobs in an organization can lead to a decrease in the quality of service, and then to the loss of customers (Wainer & Mosterman, 2018). To make the right decision in such a situation, simulation modeling can be applied, what would make it possible to predict the results of any actions in the company.
- Time. In real time, the efficiency of using any equipment or opening, for example, some new subsidiaries, can take a very long time (Wainer & Mosterman, 2018). The simulation model, on the other hand, is able to deduce the most probable outcome of such actions in few minutes.
- Repeatability. At present, organizations of various types must respond very quickly to all sorts of, even minor changes in the market. Their further development, and maybe even their existence in principle, may depend on this (Wainer & Mosterman, 2018). The simulation model runs a huge number of experiments with different parameters to find out what is best to do in order to avoid adverse moments and make the right decision.
It should be noted that simulation modeling has a number of disadvantages. Although it may take a lot of time and effort to create a simulation model, there is no guarantee that the resulting model will provide answers to all questions (Wainer & Mosterman, 2018). There is no method for proving that a model performs exactly the same as the real model. The simulation is based on multiple repetitions of sequences, which in turn are based on the generation of random numbers that recreate the occurrence of different situations.
Modeling cannot recreate the system with such accuracy as mathematical analysis, since it is based on the generation of random numbers. If it is possible to represent the system using a mathematical model, then it is better to do so (Wainer & Mosterman, 2018). The disadvantage of simulation modeling so far is that there are no defined standards (Wainer & Mosterman, 2018). Therefore, it may turn out that if different analysts recreate the same real model, the result may be completely different models.
However, the disadvantages of simulation models are leveled using Markov analysis. The fact is that such an approach is designed for mathematical calculations and assumptions, but is limited by the framework of the model. This means that Markov analysis is a quantitative method that has the characteristics of continuity (Yoe, 2019). This tool analyzes the state of the model, predicting and detecting failures (Yoe, 2019). Therefore, Markov analysis is able to minimize the above described risks and disadvantages of simulation. Markov analysis is supplemented with the help of statistical quality tools, which are aimed at controlling quantitative characteristics (Yoe, 2019). With a controlled process, the assessment of each subsequent probability will give a similar curve, if the process is out of control, the curves will be different. Accordingly, in order for simulation models to give maximum performance, they must be controlled by additional tools, such as Markov analysis and static quality.
References
Muzy, A. & Hill, D. R. C. (2018). Stochastic modeling strategies for the simulation of large (spatial) distributed systems: Application to fire spread. In Wainer, G. A. & Mosterman, P. J. (Eds.), Discrete-event modeling and simulation (pp. 331-357). CRC Press.
Yoe, C. (2019). Principles of risk analysis. Decision making under uncertainty. 2nd edition. CRC Press.