Abstract
Unit commitment, one of the significant tasks in power system operations, faces new challenges as the system uncertainty increases dramatically due to the integration of time-varying resources, such as wind. To address these challenges, we propose the formulation and solution of a generalized unit commitment problem for a system with integrated wind resources. Given the prespecified interval information acquired from real central wind forecasting system for uncertainty representation of nodal wind injections with their correlation information, the proposed unit commitment problem solution is computationally tractable and robust against all uncertain wind power injection realizations. We provide a solution approach to tackle this problem with complex mathematical basics and illustrate the capabilities of the proposed mixed integer solution approach on the large-scale power system of the Northwest China Grid. The numerical results demonstrate that the approach is realistic and not overly conservative in terms of the resulting dispatch cost outcomes.
Highlights
Unit commitment (UC) is a critically important function in operation scheduling, as it provides the linkage from the maintenance and production scheduling to economic dispatch [1]
We provide a solution approach to tackle this problem with complex mathematical basics and illustrate the capabilities of the proposed mixed integer solution approach on the large-scale power system of the Northwest China Grid
In today’s competitive electricity market environment, UC is a basic tool used by an independent system operator (ISO) or regional transmission organization (RTO) to clear the day-ahead markets and by a resource manager to optimize its offering strategy [2]
Summary
Unit commitment (UC) is a critically important function in operation scheduling, as it provides the linkage from the maintenance and production scheduling to economic dispatch [1]. To consider the grid line power flow constraints, we explicitly represent the correlations between the uncertain wind power outputs based on the principal component analysis (PCA) techniques. The PCA techniques can capture effectively the correlation relationship between the interrelated random variables and its based transformation allows us to convert a large number of interrelated variables into uncorrelated principal components (PCs) [15, 16] In this way, the multivariate statistical wind power interdependency can be represented by a series of uncorrelated variables, PCs; this will make it easier to devise a solution approach for the extended UC problem. After transforming the correlated wind power outputs by their corresponding uncorrelated principal components, we can use the proposed solution approach to obtain the final scheduling results.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
