Data-driven decision making in power systems with probabilistic guarantees: Theory and applications of chance-constrained optimization

Geng, Xinbo; Xie, Le

doi:10.1016/j.arcontrol.2019.05.005

Cited by 95 publications

(58 citation statements)

References 88 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In power systems literature, CCO has been previously used to address security constrained economic dispatch and unit commitment problems [27]. This class of problems is difficult to solve, due to two main reasons:…”

Section: B Chance-constrained Optimizationmentioning

confidence: 99%

Chance-Constrained Optimal Distribution Network Partitioning to Enhance Power Grid Resilience

2021

View full text Add to dashboard Cite

This paper proposes a method for identifying potential self-adequate sub-networks in the existing power distribution grid. These sub-networks can be equipped with control and protection schemes to form microgrids capable of sustaining local loads during power systems contingencies, thereby mitigating disasters. Towards identifying the best microgrid candidates, this work formulates a chance-constrained optimal distribution network partitioning (ODNP) problem addressing uncertainties in load and distributed energy resources; and presents a solution methodology using the sample average approximation (SAA) technique. Practical constraints like ensuring network radiality and availability of grid-forming generators are considered. Quality of the obtained solution is evaluated by comparison with-a) an upper bound on the probability that the identified microgrids are supply-deficient, and b) a lower bound on the objective value for the true optimization problem. Performance of the ODNP formulation is illustrated through case-studies on a modified IEEE 37-bus feeder. It is shown that the network flexibility is well utilized; the partitioning changes with risk budget; and that the SAA method is able to yield good quality solutions with modest computation cost.

show abstract

Section: B Chance-constrained Optimizationmentioning

confidence: 99%

Chance-Constrained Optimal Distribution Network Partitioning to Enhance Power Grid Resilience

2021

View full text Add to dashboard Cite

show abstract

“…In Reference 19, Xia et al improve the complexity of the model by computing the likelihood of a ground truth permutation, and maximize it to learn the model parameter in their solution named ListMLE. Note that here robust, 20,21 stochastic, 22 and chance-constrained 23 uncertainty handling methods can be good alternatives for optimization.…”

Section: Related Studiesmentioning

confidence: 99%

Learning to rank by using multivariate adaptive regression splines and conic multivariate adaptive regression splines

Altinok

Karagöz

Batmaz

2020

Computational Intelligence

View full text Add to dashboard Cite

Learning to rank is a supervised learning problem that aims to construct a ranking model for the given data. The most common application of learning to rank is to rank a set of documents against a query. In this work, we focus on point-wise learning to rank, where the model learns the ranking values. Multivariate adaptive regression splines (MARS) and conic multivariate adaptive regression splines (CMARS) are supervised learning techniques that have been proven to provide successful results on various prediction problems. In this article, we investigate the effectiveness of MARS and CMARS for point-wise learning to rank problem. The prediction performance is analyzed in comparison to three well-known supervised learning methods, artificial neural network (ANN), support vector machine, and random forest for two datasets under a variety of metrics including accuracy, stability, and robustness. The experimental results show that MARS and ANN are effective methods for learning to rank problem and provide promising results.

show abstract

“…Matpower and YALMIP were used to formulate the c-UC problem in MatlabR2018a. The c-UC problem was converted to s-UC via ConvertChanceConstraint in [5], then solved using Gurobi 8.10 till the MIP gap is smaller than 0.01%.…”

Section: Sample Complexity For S-ucmentioning

confidence: 99%

Chance-constrained Unit Commitment via the Scenario Approach

Geng

Xie

2019

2019 North American Power Symposium (NAPS)

Self Cite

View full text Add to dashboard Cite

Keeping the balance between supply and demand is a fundamental task in power system operational planning practices. This task becomes particularly challenging due to the deepening penetration of renewable energy resources, which induces a significant amount of uncertainties. In this paper, we propose a chance-constrained Unit Commitment (c-UC) framework to tackle challenges from uncertainties of renewables. The proposed c-UC framework seeks cost-efficient scheduling of generators while ensuring operation constraints with guaranteed probability. We show that the scenario approach can be used to solve c-UC despite of the non-convexity from binary decision variables. We reveal the salient structural properties of c-UC, which could significantly reduce the sample complexity required by the scenario approach and speed up computation. Case studies are performed on a modified 118-bus system. The authors are with the Recently, the growing amount of uncertainties from renewables pose new challenges on the operations of power systems. UC, as a critical part of day-ahead scheduling, needs to be improved to consider the impacts of uncertainties. Broadly speaking, there are two approaches for decision making under uncertainties: stochastic optimization (SO) and robust optimization (RO). SO relies on probabilistic models to explain uncertainties and often optimizes the objective function in the presence of randomness. SO has found many successful applications in power systems. References [1]-[3] formulate and solve the stochastic unit commitment problem, which typically minimizes expected commitment and dispatch costs in the presence of uncertainties. RO takes an alternative approach, in which the uncertainty model is set-based and deterministic. Recently, researchers in [4] formulated and solved the robust unit commitment problem, which minimizes the commitment and dispatch costs for the worst case in a predefined uncertainty set. Both approaches attract a lot of attention and are relatively successful in addressing the challenges related with uncertainties. This paper looks at the UC problem through the lens of chance-constrained optimization (CCO), which is closely related with both stochastic and robust optimization [5]. The main difference of CCO from SO or RO is the chance constraint (i.e. (1b) and (2b)), which explicitly considers the feasibility of solutions under uncertainties.Various formulations of chance-constrained (security-constrained) unit commitment have been proposed, e.g.[6]- [14]. As mentioned in [5], chance-constrained optimization problems can be solved via different methods. We take chance-constrained unit commitment problem as an example. It can be solved using sample average approximation [8]-[10], [12]-[14] or robust optimization based techniques [15]. The scenario approach, which might be the most wellknown method to solve chance-constrained optimization, was not directly applied on the unit commitment. The only related references we found are [16], [17], which are built upon a variation of the scen...

show abstract

Data-driven decision making in power systems with probabilistic guarantees: Theory and applications of chance-constrained optimization

Cited by 95 publications

References 88 publications

Chance-Constrained Optimal Distribution Network Partitioning to Enhance Power Grid Resilience

Chance-Constrained Optimal Distribution Network Partitioning to Enhance Power Grid Resilience

Learning to rank by using multivariate adaptive regression splines and conic multivariate adaptive regression splines

Chance-constrained Unit Commitment via the Scenario Approach

Contact Info

Product

Resources

About