Simulation–optimization is often used in enterprise decision-making processes, both operational and tactical. This paper shows how an intuitive mapping from descriptive problem to optimization model can be realized with Constraint Programming (CP). It shows how a CP model can be constructed given a simulation model and a set of business goals. The approach is to train a neural network (NN) on simulation model inputs and outputs, and embed the NN into the CP model together with a set of soft constraints that represent business goals. We study this novel simulation–optimization approach through a set of experiments, finding that it is flexible to changing multiple objectives simultaneously, allows an intuitive mapping from business goals expressed in natural language to a formal model suitable for state-of-the-art optimization solvers, and is realizable for diverse managerial problems.

Wind farms suffer from so-called wake effects: when turbines are located in the wind shadows of other turbines, their power output is substantially reduced. These losses can be partially mitigated via actively changing the yaw from the individually optimal direction. Most existing wake control techniques have two major limitations: they use simplified wake models to optimize the control strategy, and they assume that the atmospheric conditions remain stable. In this paper, we address these limitations by applying reinforcement learning (RL). RL forgoes the wake model entirely and learns an optimal control strategy based on the observed atmospheric conditions and a reward signal, in this case the power output of the farm. It also accounts for random transitions in the observations, such as turbulent fluctuations in the wind. To evaluate RL for active wake control, we provide a simulator based on the state-of-the-art FLORIS model in the OpenAI gym format. Next, we propose three different state-action representations of the active wake control problem and investigate their effect on the performance of RL-based wake control. Finally, we compare RL to a state-of-the-art wake control strategy based on FLORIS and show that RL is less sensitive to changes in unobservable data.