Construction of multiband uncertainty set for building energy management with uncertain loads and solar power

For the optimal operation of building energy systems with uncertain energy consumption and renewable solar energy supply, the construction of uncertainty set is becoming especially important due to the conservatism of the solution in robust optimization. The general two-stage robust optimization with single-band uncertainty model considers the entire uncertain set for any uncertainty, which would result in too conservative solutions. In order to flexibly adapt the decision solutions for the robust optimal operation of building energy systems, the multiband uncertainty concept is introduced in adaptive robust optimization. An efficient method is defined to partitioning the single uncertainty band into multiple sub-bands on the basis of stochastic programming. And each individual uncertainty set is associated with a certain probability. The numerical study shows that the concept of multiband uncertainty set introduces great flexibility into the adaptive robust optimization model for the optimal operation of building energy systems.