Efficient Coordination of Top-Down and Bottom-Up Models for Energy System Design : An Algorithmic Approach

We consider linked top-down and bottom-up models in which the top-down model is a mixed-complementarity program describing a macroeconomic system, while the bottom-up model is a linear program describing the energy system. We introduce a coordination algorithm that speeds up convergence for the linkage of the two models. The coordination algorithm uses optimal bases from previous iterations to predict the solution of the bottom-up model. If such a prediction is successful, the bottom-up problem need not be solved in that iteration, resulting in a time gain. In numerical experiments on an energy system design problem, we show that our coordination algorithm correctly predicts the bottom-up solution in most iterations, resulting in a significant reduction in overall computation time