Network-Constrained Unit Commitment with Flexible Temporal Resolution

TL;DR

The paper tackles the computational burden of network-constrained unit commitment by introducing a congestion-aware flexible temporal-resolution method. It determines adaptive time periods by optimizing a congestion-aware impact metric and integrates this with ramping constraint transformations and adaptive-period parameter updates. Through IEEE 118-bus and Polish 2736-bus case studies, the method achieves large speedups (up to tens of times) with negligible cost variation, while preserving feasibility with correction when needed. The approach offers near-optimal day-ahead scheduling with significantly reduced computation, and can be combined with other acceleration techniques.

Abstract

Modern network-constrained unit commitment (NCUC) bears a heavy computational burden due to the ever-growing model scale. This situation becomes more challenging when detailed operational characteristics, complicated constraints, and multiple objectives are considered. We propose a novel simplification method to determine the flexible temporal resolution for acceleration and near-optimal solutions. The flexible temporal resolution is determined by analyzing the impact on generators in each adaptive time period with awareness of congestion effects. Additionally, multiple improvements are employed on the existing NCUC model compatible with flexible temporal resolution to reduce the number of integer variables while preserving the original features. A case study using the IEEE 118-bus and the Polish 2736-bus systems verifies that the proposed method achieves substantial acceleration with low cost variation and high accuracy.

Figures (10)

• Figure 1: Framework of the proposed NCUC with flexible temporal resolution.
• Figure 2: Illustration of the power output variation model. To ensure the feasibility of the model, the square green area \ref{['eq:theor-nrange']}, the blue line \ref{['eq:theor-nbal']} and the yellow half-plane \ref{['eq:theor-ndpf']} must intersect.
• Figure 3: Illustration of the extreme circumstances concerning the original time periods for the ramping constraints. For $RU_i^{(t)}$, the power output of the unit increases with the maximum ramping capability. For $SU_i^{(t)}$, the power output of the unit increases with the maximum ramping capability from $\underline{P}_i$ to $\overline{P}_i$.
• Figure 4: Power flow on Line 31 and Line 51 ignoring and considering congestion. These two lines become congested at approximately 18:00, when the demand is at its highest.
• Figure 5: The flexible temporal resolution determined by M1 and M2.