An outgrowth of the systems dynamics approach is what we called systems diagrams tools. The software discussed here has many more icons than the stocks, flows, and parameters that the systems dynamics tool operates with. Whole submodels or solvers for mathematical equations, say, partial differential equations (PDEs), may be embedded into specially designed icons that later on become part of the toolbox for future applications. Once again, we get more functionality and flexibility, but certainly at the expense of a much steeper learning curve.
• Extend - Imagine That Inc., http://www.imaginetha-tinc.com/ - free demo, Win/Mac
As it follows from the name of the product, the system is extendable. It encourages modularity, providing the functionality to encapsulate certain processes and subsystems into blocks that can be further reused. Extend models are constructed with library-based iconic blocks. Each block describes a calculation or a step in a process. Interprocess communication allows two applications communicating and sharing data with one another. This feature allows the integration of external data and applications into and out of Extend models. Information is automatically updated between Extend and Excel, can be connected with databases (Open DataBase Connectivity), has embedded ActiveX or OLE (object linking and embedding), and works with DLL. Block building is based on ModL - a language that provides high-level functions and features while having a familiar look and feel for users with experience of programming in C. Also allows scripting to develop 'wizards' or self-modifying models. Evolutionary optimizer employs powerful enhanced evolutionary algorithms to determine the best model configuration.
• GoldSim - GoldSim Technology Group, http:// www.goldsim.com/ - free evaluation and student version, Win
Same approach based on an extendable library of icons ('hierarchical containers') for a variety of processes. The user controls the sequence of events and can superimpose the occurrence and consequences of discrete events onto continuously varying systems. Other features include particularly strong stochastic, Monte Carlo simulation component to treat uncertainty and risks inherent in all complex systems; embedded optimization, sensitivity analyses (e.g., tornado charts, statistical measures); external dynamic links to programs and spreadsheets, direct exchange of data with ODBC-compliant databases. Models can be saved as player files. There are several extension modules (e.g., for contaminant transport using solvers for PDE, financial analysis, etc.).
• Simulink - The MathWorks, http://www. mathworks.com/ - free trial and web demo, Win/Mac/ UNIX
Built on top of MATLAB (see below). Provides an interactive graphical environment and a customizable set of block libraries, which can be extended for specialized applications. More power, but harder to master. Can generate C code for your models, which can be further embedded into other applications. Based on the same concept of expandable libraries of predefined blocks, with an interactive graphical editor for assembling and managing block diagrams, with functionality to interface with other simulation programs and incorporate handwritten code, including MATLAB algorithms. Has full access to MATLAB for analyzing and visualizing data, developing GUIs, and creating model data and parameters.
Pros. Power, versatility, flexibility, expandability.
Cons. In a way, the pros become their cons, since after investing much time to fully master these systems it is most likely that they will become your 'hammer' for the future. Besides getting wedded to proprietary software, there is always a risk of running into limitations that will be hard to overcome.
Was this article helpful?