Distributed scenario-based optimization for asset management in a hierarchical decision making environment

Abstract

Abstract-Asset management attempts to keep the power system in working conditions. It requires much coordination between multiple entities and long term planning often months in advance. In this work we introduce a mid-term asset management formulation as a stochastic optimization problem, that includes three hierarchical layers of decision making, namely the midterm, short-term and real-time. We devise a tractable scenario approximation technique for efficiently assessing the complex implications a maintenance schedule inflicts on a power system. This is done using efficient Monte-Carlo simulations that tradeoff between accuracy and tractability. We then present our implementation of a distributed scenario-based optimization algorithm for solving our formulation, and use an updated PJM 5-bus system to show a solution that is cheaper than other maintenance heuristics that are likely to be considered by TSOs.

Unit Commitment Using Nearest Neighbor as a Short-Term Proxy

“…We begin with defining the accurate UC solution; notations are adopted from [4]. The optimal UC decision u * p (y s , u m ) is the solution of the following optimization problem:…”

A computational model of mid-term outage scheduling for long-term system studies

“…Another direction is taken in [18]- [20], where the authors model the impact of maintenance on asset failure rate and the associated failure costs. Among these works, only [20] explicitly models the RT context as an additional instance beyond the DA, and considers uncertainty in the future operating conditions by means of Monte Carlo simulations.…”

“…Among these works, only [20] explicitly models the RT context as an additional instance beyond the DA, and considers uncertainty in the future operating conditions by means of Monte Carlo simulations. However, this leads to tractability issues that the authors handle by using a reduced number of time periods when generating Monte Carlo samples.…”

Probabilistic Reliability Management Approach and Criteria for power system real-time operation

“…3) From short-term to long-term decision making: As soon as the usefulness and practicality of the new RMAC is established in the context of real-time operation, we need to see what are the implications of its use in real-time, when it comes to decision making in other -longer-term -decision making contexts, such as operation planning (a few hours, days, weeks ahead in time), asset management (maintenance, inspection, and replacement, over months or many years) [18] and grid development (to be justified over decades). In principle these later decision making contexts aim at choosing their strategies while targeting to ensure the reliable operation of the system in real-time.…”