Researhc

Prompt

This algorithm is useful while working with expensive, time-consuming objective functions that entail a great deal of computational complexity and effort. It uses a surrogate function that approximates the actual objective function but takes less time to evaluate. The algorithm chooses quasirandom points within the bounds on input variables and then constructs the surrogate by interpolating upon the true objective using a radial basis function interpolator [47]. Thus, the best point minimizing the surrogate can be taken as a reasonable approximation for the minimizer of the objective. The method tries to balance two competing goals: (1) Complete exploration of search space for a global minimum and (2) Increase speed by arriving at a good solution using as few objective function evaluations as possible. This works best when the objective is smooth and continuous, and has been proven to arrive at globally optimal solutions on bounded domains even though the convergence is not fast [47]. However, smoothness is not strictly required to use this technique. Such an approach can greatly speed up computational times for complex component sizing and economic dispatch problems with non-linear cost functions and a large number of variables and constraints. There is usually no stopping criterion that can accurately recognize when the solver is near a global optimum. Instead the program terminates upon reaching a prescribed number of iterations, function evaluations or time duration and outputs the ’best’ solution found within that preset computational budget11