The Pointer-2 strategy controls the following optimization methods:
For more information about these methods, see About the Optimization Techniques. This complementary set of algorithms was selected because each succeeds or fails for different topography features. It has been found that a hybrid combination of these methods solves a broad range of design optimization problems. The Pointer-2 strategy can control one algorithm at a time or all algorithms at once. As the optimization proceeds, the strategy determines which algorithms are most successful and the optimal internal control parameters (e.g., step sizes, numbers of iterations, number of restarts, etc.). The goal of the Pointer-2 strategy is to enable the non-optimization expert to successfully utilize these optimization techniques. In addition to running optimization techniques, the Pointer-2 strategy periodically generates internal approximations using data points taken from the optimization history. The Pointer-2 strategy performs an optimization on the approximation and verifies the optimum with an exact execution. This approach can accelerate the convergence of the Pointer-2 strategy for design spaces that can be approximated accurately. The Pointer-2 strategy also allows you to select a single technique that is used instead of the hybrid combination. This gives you the advantage of working with a technique with which you have experience to try and achieve improved performance. | |||||||