Fast Scenario Reduction for Power Systems by Deep Learning

Abstract

Scenario reduction is an important topic in stochastic programming problems. Due to the random behavior of load and renewable energy, stochastic programming becomes a useful technique to optimize power systems. Thus, scenario reduction gets more attentions in recent years. Many scenario reduction methods have been proposed to reduce the scenario set in a fast speed. However, the speed of scenario reduction is still very slow, in which it takes at least several seconds to several minutes to finish the reduction. This limitation of speed prevents stochastic programming to be implemented in real-time optimal control problems. In this paper, a fast scenario reduction method based on deep learning is proposed to solve this problem. Inspired by the deep learning based image process, recognition and generation methods, the scenario data are transformed into a 2D image-like data and then to be fed into a deep convolutional neural network (DCNN). The output of the DCNN will be an "image" of the reduced scenario set. Since images can be processed in a very high speed by neural networks, the scenario reduction by neural network can also be very fast. The results of the simulation show that the scenario reduction with the proposed DCNN method can be completed in very high speed.

The energy distance for ensemble and scenario reduction

“…Sophisticated selection methods that work for the Wasserstein metrics can applied to the energy distance as well, see e.g. Li and Gao (2019).…”

“…Growe-Kuska et al (2003); Römisch (2009); Feng and Ryan (2013). However, other alternatives like particle swarm optimization, neural network base deep learning methods and other heuristics are applied too, see Li and Gao (2019).…”

A Stochastic Decision-Making Tool Suite for Distributed Energy Resources Integration in Energy Markets

“…The heuristic approach employed in this stage identifies and eliminates scenarios that are unlikely or insignificant, and that have minimal or no impact on the final result. In this instance, the Kantorovich distance has been employed as the criterion for this reduction, as outlined in [90].…”

“…Such consideration reflects a proficient handling of the uncertainty associated with this variable, thereby influencing the decision-making process facilitated by this tool. It is important to note that it is assumed that all scenarios, including 50 scenarios (obtained using scenario reduction techniques such as FastForward [90]) to balance the computational burden and model formulation, have the same probability of occurrence for both variables with uncertainty.…”