Guide to Advanced Empirical
Download 1.5 Mb. View original pdf
|
3299771.3299772, BF01324126
| 5. Simulation Tools Today, many software tools are available to support the various simulation techniques described above. Compared to the first tools available in the s, sands, most of today’s more popular tools have a user-friendly interface and are inexpensive, making them practical to use fora large variety of decision making situations. Today, most tools. Allow for rapid model development through using, for example (a) Drag and drop of iconic building blocks (b) Graphical element linking (c) Syntactic constraints on how elements are linked. Are very reliable. Require little training. Are easy to understand 126 MM ller and D. Pfahl Because of these features, simulation tools allow modellers to develop large detailed models rapidly. Modern tools have followed the evolution of software languages and software development environments. Now they focus on model design and a proper visualization rather than on programming the simulation logic. The simulation tools in today’s marketplace are robust and reasonably inexpensive. Most tools cost in the range of $1,000–10,000, and free versions are available for experimentation and evaluation. They run on standard PC hardware, and are therefore affordable even for small organizations with tight budgets. The number of simulation tools is large, in particular if one counts the ever-growing number of simulation environment research prototypes developed at universities allover the world. In principle, a simulation model based on any of the above mentioned simulation techniques can also be implemented in an ordinary programming languages (e.g., Java, or by using general purpose simulation languages (e.g., MATLAB ® ). However, several commercial simulation tools use the most important simulation techniques and are suited to support software engineering problems. Table 2 characterizes three popular examples of simulation tools supporting SD, DE, and Monte Carlo simulation, respectively. The choice of a simulation tool environment depends on several factors. Since the prices are comparatively low, the most important factor is the appropriateness of the simulation technique that is supported. Ina professional simulation environment, in conjunction to the simulation modelling tool, other tools are often used. Professional simulation studies typically involve information systems or databases which store the input, calibration, and output data, a statistical distribution fitter to analyze the calibration data, and an optimizer. High-end tools such as the more expensive versions of VENSIM ® and EXTEND ® already include the distribution fitters and optimizers. Table 2 Examples of commercial simulation tools used in software engineering Tool name Main focus Characterization Interesting features VENSIM ® (Vensim, Support of SD simulation Dynamic, continuous, deterministic and stochastic, quantitative Optimization function, calibration support, graphical modelling language (using standard SD symbols, animation, can emulate event-driven simulation to some extend by introducing if-then-else-conditions EXTEND ® (Extend, Support of DE and SD simulation Dynamic, event-driven and continuous, deterministic and stochastic, quantitative Optimization support, graphical modelling language, strong modularization capability statistical fitting (StatFit ® ), library source code available @RISK ® (@Risk, Monte Carlo simulation Static, deterministic, stochastic, quantitative Can easily be integrated with standard spreadsheet tools (i.e., Microsoft’s EXCEL, provides functionality for distribution fitting (BestFit ® ) 5. Simulation Methods Next follows a brief introduction into the SD modelling tool VENSIM ® , which will be used in the presentation of a process simulation example in Sect. 6 below. Download 1.5 Mb. Share with your friends:
|