An a priori Kriging method based on genetic algorithms for process fluid loads

Unlike steady fluid loads, process fluid loads are usually constrained by initial conditions and other constraints. Traditional Kriging methods do not consider these constraints, resulting in significant deviations between model predictions and actual results. In response to the above issues, this article improves the traditional Kriging method by incorporating relevant constraints into the modeling process, and uses genetic algorithms to solve the corresponding problem. The simulation shows that the proposed a priori Kriging model has significantly better predictions than the traditional Kriging model.