Robust Conservation Voltage Reduction Evaluation Using Soft Constrained Gradient Analysis

Zixiao Ma

doi:10.1109/TPWRS.2022.3146842

2022-11-01

Zixiao Ma

Yingmeng Xiang

Zhaoyu Wang *

DOI：10.1109/TPWRS.2022.3146842

原文链接

原文求助

摘要

En 中文

Evaluation of conservation voltage reduction (CVR) factor is critical to peak load reduction, 556 energy conservation, and many CVR-related power system control/optimization strategies. This paper proposes a novel robust CVR factor evaluation method using soft constrained gradient analysis based on time-varying load modeling and real utility field measurements. Firstly, the time-varying ZIP parameter identification is formulated as an over-determinant problem composed of first-order gradient with respect to each coefficient and soft constraints representing temporal correlation of loads in each time step, in order to improve the robustness and smoothness of CVR factor evaluation. Then, problems in time series are coordinated with a sliding window approach. A necessary condition for selecting the smallest window size is proposed and strictly proved to guarantee the existence and uniqueness of the solution of time-varying load modeling problem. Finally, time-varying CVR factors are accurately and robustly calculated with field measurements and the identified load model. Case studies are performed using sufficient field measurements obtained from two real utilities. Unlike existing methods that require a large number of measurements to obtain precise estimation of CVR factor, the proposed method is sufficiently accurate even when the measurements are limited or of low time resolution. Further, the accuracy and robustness of the proposed approach are validated under different types of uncertainties and compared with other existing data processing and CVR factor evaluation methods.

Keyword:

Load modeling

Voltage measurement

Mathematical models

Analytical models

Voltage control

Power measurement

Power system dynamics

Conservation voltage reduction

time-varying load modeling

robust evaluation

limited measurements