Double-Layer Optimization Mechanism for Multi-Area OPF Considering Valve-Point Loading Effect

In terms of the multi-area optimal power flow (OPF) problem, the optimized objectives are always a fuel cost function expressed by a second-order polynomial. However, the valve-point loading effect, whose cost curve is a transcendental function formed by the superposition of the sine and polynomial...

Full description

Saved in:
Bibliographic Details
Published in:CSEE Journal of Power and Energy Systems Vol. 11; no. 2; pp. 683 - 691
Main Authors: Jizhong Zhu, Cong Zeng, Yun Liu, Xuancong Xu
Format: Journal Article
Language:English
Published: China electric power research institute 2025
Subjects:
Distributed computation platform
distributed meta-heuristic optimization algorithm
double-layer optimization mechanism
multi-area optimal power flow
ISSN:2096-0042
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In terms of the multi-area optimal power flow (OPF) problem, the optimized objectives are always a fuel cost function expressed by a second-order polynomial. However, the valve-point loading effect, whose cost curve is a transcendental function formed by the superposition of the sine and polynomial function, will make the objective function non-convex and non-differentiable. Conventional distributed optimization technologies can hardly make a solution directly. Therefore, it is necessary to realize a distributed solution for multi-area OPF from another point of view. In this paper, we constitute a new double-layer optimization mechanism. The proposed distributed meta-heuristic optimization (DMHO) algorithm is put on the top layer to optimize the dispatching of each area, and in each iteration a distributed power flow calculation method is embedded as the bottom layer to minimize the mismatch of power balance. Numerical experiments demonstrate that the proposed approach not only implements a multi-area OPF distributed solution but also accelerates the convergence rate, improves the solution accuracy and enhances the robustness. In addition, a fully decentralized computation experiment is performed in an actual distributed environment to test its practicability and computation efficiency.
ISSN:2096-0042
DOI:10.17775/CSEEJPES.2022.08660