An integrated architecture for simulation

‘The explosion in computer performance allows average engineers to use simulation as part of the design process’ and enables them ‘to search all possible design alternatives, and find the one that best meets the overall design criteria’ by performing hundreds of simulations rather than just one or two. Yet, ‘the advanced analysts resist the effort, concerned that it would be hard to get the correct answers.’ An integrated architecture for simulation is proposed in this paper, which reduces the effort required for finding of the correct answers through integration of the simulation package with data-management (DM) software. The integrated architecture is based on three elements: input/output (I/O) handling, autonomous operation and control. I/O handling is provided by an add-on interface constructed in the DM tool, taking advantage of the fact that simulation software tools are increasingly able to exchange information with other software tools in an integrated way. All user interaction with the simulation can be done via this interface—input data entry, triggering the execution of simulation runs, and analyses of the results, which are automatically transferred by the simulation and recorded in the DM tool. Autonomous operation allows the execution of batch of simulation runs in the background leaving the user free to engage in other activities. Control, beyond providing the means for autonomous operation, enables the inclusion of simulation within larger context such as simulation–optimization schemes. Consequently, the integrated architecture allows simulation to be ‘routinely applied as a powerful, flexible and general purpose engine for optimization from the initial concept across the full product life cycle’.