Prev | Next |
Model Simulation
Model Simulation brings your behavioral models to life with instant, real-time behavioral model execution. Coupled with tools to manage triggers, events, guards, effects, breakpoints and simulation variables, plus the ability to visually track execution at run-time, the Simulator is a powerful means of 'watching the wheels turn' and verifying the correctness of your behavioral models. With Simulation you can explore and test the dynamic behavior of models. In the Corporate and above editions, you can also use JavaScript as a run-time execution language for evaluating guards, effects and other script-able pieces of behavior.
Extensive support for triggers, trigger sets, nested states, concurrency, dynamic effects and other advanced simulation capabilities, provides a remarkable environment in which to build interactive and working models that help explore, test and visually trace complex business, software and system behavior. With JavaScript enabled, it is also possible to create embedded COM objects that will do the work of evaluating guards and executing effects - allowing the simulation to be tied into a much larger set of dependent processes. For example, a COM object evaluating a guard condition on a State Transition might query a locally running process, read and use a set of test data, or even connect to an SOA web service to obtain some current information.
As Enterprise Architect uses a dynamic, script driven Simulation mechanism there is no need to generate code or compile your model before running a simulation. It is even possible to update simulation variables in real time using the Simulation console window. This is useful for testing alternate branches and conditions 'on the fly', either at a set Simulation break point or when the Simulation reaches a point of stability (for example, when the Simulation is 'blocked').
In the Professional version of Enterprise Architect, you can manually walk through simulations - although no JavaScript will execute - so all choices are manual decisions. This is useful for testing the flow of a behavioral model and highlighting possible choices and processing paths.
Simulation Overview
Notes