The large-scale automated collection of energy consumption data raises new optimization problems. The issue addressed in this thesis is to propose solutions for modeling and solving optimization models for multi-stream energy consumption management. The consumption data are represented as a time series.
We first develop deterministic and robust models for the optimization of electricity contracting. This problem is formulated as the minimization of a convex objective function under special structure constraints (order constraints); it is solved by a new efficient algorithm that considerably reduces the computational effort, in the presence of a large amount of data, needed to evaluate the objective function.
We then present deterministic and robust models around the optimal management of water networks using an automated strategy based on the use of trigger levels. We also show that the problem can be reduced to smaller daily problems without losing too much accuracy.
Numerical results are presented and reveal the possibility of strong savings in the different energy systems studied.