Many design algorithms (e.g., optimization techniques, DOE techniques, etc.) use random numbers to generate parameter values (or combinations of parameter values) to use when searching the design space. If you execute a model with one of these design algorithms and execute it a second time, you will not get the same results (though you will often get near the same solution, but through a different path). If you want the model execution to follow a consistently repeatable path, you must set a fixed seed for the random number generator to use when it provides random numbers (otherwise, it often uses something like the current time in milliseconds). With a fixed seed, the random number generator always produces the same sequence of random numbers. For more information about setting the seed, see Setting the Execution Options. | |||||||