應用電荷搜尋演算法進行含風力發電之系統最佳化無效功率調度研究

Study of the Optimal Reactive Power Dispatch for Wind Power Systems by using Charged System Search Algorithm

郭子祥(Tzu-Hsiang Kuo)

指導教授 ： 陳昭榮

共同指導教授 ： 李清吟

國立臺北科技大學/電資學院/電機工程系所/碩士(2014年)

摘要

無效功率與電壓控制問題對於電力系統而言是非常重要的，適當的調整不僅可以減少有效功率損失，且能改善電壓穩定問題，使得系統操作於穩定的運作範圍內。傳統無效功率的調度方式一般採用調整發電機端電壓、配電變電所內主變壓器有載分接頭位置、以及加裝無效功率補償裝置，從而控制電壓以供給用戶穩定的電壓，提昇系統的供電品質。傳統系統中因僅考慮負載變動對於系統電壓與無效功率的影響，所以相關的技術與理論已經相當成熟，然而在新世代的電力系統中，由於智慧電網以及再生能源的導入，系統無效功率與電壓控制問題將產生新的挑戰以及可能的契機。再生能源由於具有高度的不確定性，因此其產生的電力輸出變化極高，系統電壓控制的問題將顯得更為複雜。此外，再生能源電力轉換器的控制技術亦會影響再生能源輸出的特性。本研究將以隨機變數來描述再生能源的不確定性，並以機率負載潮流以及最佳化演算法來分析含再生能源系統之電壓控制的問題。
本論文採用的最佳化演算法為電荷搜尋演算法，它是一種多重代理人的方法，將代理人視為帶電粒子(Charged Particle, CP)，每個CP基於其適應值與距離從而影響其他CP，藉由計算吸引力以更新位置尋找最佳解，並在一些基準函數與工程問題上驗證其效率。本論文的目標函數是在滿足所有系統運轉的限制式情況下，使輸電損失率最小化，並將發電機的端電壓、變壓器分接頭設定與並聯電容器組當作控制變數，調整電力系統之電壓大小。本論文透過分析模擬IEEE 30-bus証實所提方法之可行性。後續將研究澎湖高占比風力系統，當考量再生能源不確定因素後，探討在電力系統為機率性潮流下的最佳化無效功率之調度。

關鍵字

無效功率 ； 電壓控制 ； 再生能源 ； 電荷搜尋演算法

並列摘要

Problems on reactive power and voltage control are significant for power systems. Suitable adjustment can not only reduce active power loss, but improve voltage stability so as to maintain a stable operation. Traditional reactive power dispatch methods use the adjustment of generator terminal voltage, load taps of transformers in substations, and reactive power compensation devices to control voltage and enhance the supply quality. Traditional systems only considered the effect of load variations on the system voltage and reactive power, and associated technologies and theories are mature. However, in modern power systems, reactive power and voltage control problems would suffer from new challenge owing to the introduction of smart grids and renewable energy. Renewable energy posses high uncertainty, so the variation of its power output is high, leading to more complicated solutions for the voltage control. Additionally, the control of power converters in renewable energy would have an effect on its output characteristics. This study will describe the uncertainty of renewable energy by random variables, and analyze the voltage control problems in renewable energy systems by using probability load flows and optimal algorithms.
This project uses the Charged System Search (CSS), one of the optimal algorithms, to solve the reactive power dispatch. It is a multi-agent-based method, which regards the agent as the charged particle (CP). Each CP affects other CPs based on the fitness values and distance, search for the optimal solutions by calculating the attractive force to adjust locations, and verifying its efficiency on basic functions and industry problems. The objective function of this study is to minimize the transmission loss with satisfying all of the system constraints, and sets the terminal voltage of generators, tap setting of transformers, and capacitors as control variables to adjust the system voltage. The initial study has proved the feasibility of the proposed method by using the IEEE 30- bus simulation. Next, this study will consider the high wind power penetration system in Penghu. This study will discuss the optimal reactive power dispatch taking renewable energy uncertainty and probability load flow into account.

並列關鍵字

Reactive Power ； Voltage Control ； Renewable Energy ； Charged System Search

參考文獻

[1] 陳彥豪、陳彥宏、胡桓祥，「各國微電網示範計畫介紹」，台灣經濟研究月刊，第三十三卷，第七期，2010，第103-112頁。

[29] 羅一峰，應用累積元與基因演算法求解配電系統無效電力控制，碩士論文，中原大學電機工程系，桃園，2008。

[32] 張銘仁，應用類神經網路於高占比風能電力系統中發電機跳機之最低頻率預測，碩士論文，國立臺北科技大學電機工程系，臺北，2013。

[2] R. H. Liang and C. K. Cheng, "Dispatch of main transformer ULTC and capacitors in a distribution system," IEEE Transactions on Power Delivery, Vol. 16, Issue 4, 2001, pp. 625-630.

[4] C. T. Su and C. T. Lin, "Fuzzy-based voltage/reactive power scheduling for voltage security improvement and loss reduction," IEEE Transactions on Power Delivery, Vol. 16, Issue 2, 2001, pp. 319-323.