以多目標演化演算法結合資源分配機制求解動態電力調度之成本與污染最佳化問題

Abstract

在當前能源需求快速增長的背景下，動態電力調度之成本與污染最佳化問題成為電力系統最佳化的重要挑戰。該問題涉及在多時段內，滿足用電需求的同時，最小化燃料成本與污染物排放，並考量各種系統限制（如發電機輸出範圍、負載平衡及電力升降限制）。由於動態電力調度之成本與污染最佳化問題的多目標性、高維度與非線性特性，傳統方法難以有效解決該問題。本研究提出一種動態資源分配機制的多目標演化演算法，用以求解動態電力調度之成本與污染最佳化問題。該方法改進了傳統演算法在資源分配效率與解的多樣性上的不足，透過外部解集合引導搜尋方向，並運用比例動態調整決策機制處理不可行解。實驗中，透過多組公開測試數據進行比較，結果顯示本論文之方法和二十六個既有方法相比，展現優秀的求解能力。With the rapid increase in power demand, solving the dynamic economic emission dispatch (DEED) problem has become a crucial challenge in power system optimization. The DEED problem aims to minimize fuel costs and pollutant emissions while satisfying power demand over multiple time periods, considering various system constraints such as generator output limits, load balance, and ramp rate constraints. Due to the multiple objectives, high dimensionality, and nonlinearity of the DEED problem, traditional methods struggle to solve it effectively.This study develops an algorithm based on the external archive-guided multi-objective evolutionary algorithm based on decomposition (EAG-MOEA/D). EAG-MOEA/D decomposes a multi-objective problem into subproblems using an aggregation function, stores non-dominated solutions in an external archive, and allocates computing resources to subproblems based on their contribution to the archive. In this thesis, we propose three key improvements: first, we increase the update frequency of environmental selection to allow high-quality solutions to exert greater influence on the evolutionary process; second, we enhance the resource allocation mechanism by incorporating diversity contribution; and third, we hybridize the weighted sum and Tchebycheff functions to balance convergence and diversity. Experimental results on multiple publicly available benchmark datasets demonstrate that the proposed method outperforms twenty-six existing algorithms, exhibiting superior problem-solving capabilities.