【摘要】：進(jìn)入21世紀(jì)以來(lái),具有能量自治管理與控制能力的光伏微網(wǎng)作為可再生能源利用的主要形式之一,得到了迅速發(fā)展。但由于光伏微網(wǎng)的電源單一,出力波動(dòng)大,且配置的儲(chǔ)能容量一般也比較有限,較小的電源或負(fù)荷波動(dòng)都會(huì)對(duì)其優(yōu)化運(yùn)行調(diào)度帶來(lái)較大影響。因此,必須針對(duì)光伏微網(wǎng)自身的特點(diǎn),建立適用于光伏微網(wǎng)安全穩(wěn)定運(yùn)行的能量?jī)?yōu)化調(diào)度模型,并研究相應(yīng)的求解方法進(jìn)行仿真分析,具有重要的理論和工程意義。本文首先介紹了包含分布式電源的微網(wǎng)結(jié)構(gòu),并從微網(wǎng)優(yōu)化調(diào)度策略、優(yōu)化調(diào)度模型及其模型求解算法這三個(gè)方面闡述關(guān)于微網(wǎng)能量?jī)?yōu)化調(diào)度的研究歷史與研究現(xiàn)狀,在此基礎(chǔ)上,重點(diǎn)研究了考慮需求需求側(cè)管理(demand side managment,DSM)的多目標(biāo)優(yōu)化調(diào)度問(wèn)題、考慮預(yù)測(cè)誤差的隨機(jī)優(yōu)化調(diào)度問(wèn)題和多時(shí)間尺度的光伏微網(wǎng)隨機(jī)優(yōu)化調(diào)度問(wèn)題。主要的研究工作為以下3個(gè)部分:(1)在考慮需求側(cè)管理的光伏微網(wǎng)多目標(biāo)優(yōu)化調(diào)度問(wèn)題中。首先對(duì)典型光伏微網(wǎng)的運(yùn)行狀態(tài)進(jìn)行了簡(jiǎn)要分析,并針對(duì)考慮電動(dòng)汽車(chē)參與光伏微網(wǎng)能量?jī)?yōu)化調(diào)度與不考慮電動(dòng)汽車(chē)參與時(shí)對(duì)系統(tǒng)運(yùn)行經(jīng)濟(jì)性的影響,分別建立了不包含電動(dòng)汽車(chē)充電的、及包含電動(dòng)汽車(chē)充電的優(yōu)化調(diào)度模型,并綜合考慮功率平衡、儲(chǔ)能系統(tǒng)荷電狀態(tài)、負(fù)荷可轉(zhuǎn)移的時(shí)間范圍、電動(dòng)汽車(chē)的充電時(shí)間等約束條件,提出了基于非支配排序遺傳算法(NSGA-Ⅱ)的優(yōu)化求解方法。仿真結(jié)果分析表明了該方法的可行性和有效性,且考慮需求側(cè)管理和電動(dòng)汽車(chē)充電對(duì)于提高系統(tǒng)運(yùn)行經(jīng)濟(jì)性效果顯著。(2)在考慮預(yù)測(cè)誤差的隨機(jī)優(yōu)化調(diào)度問(wèn)題中,針對(duì)光伏發(fā)電功率及負(fù)荷功率的預(yù)測(cè)精準(zhǔn)度對(duì)系統(tǒng)安全經(jīng)濟(jì)運(yùn)行的影響,采用概率密度分布函數(shù)對(duì)系統(tǒng)運(yùn)行中的不確定因素進(jìn)行模擬,綜合考慮實(shí)時(shí)電價(jià)、實(shí)時(shí)旋轉(zhuǎn)備用電價(jià)、系統(tǒng)旋轉(zhuǎn)備用約束等約束條件,然后提出了一種基于機(jī)會(huì)約束規(guī)劃理論的并網(wǎng)型光伏微電網(wǎng)優(yōu)化調(diào)度模型。其中引入了隨機(jī)變量和機(jī)會(huì)約束,不易直接求解,但應(yīng)用蒙特卡羅模擬與遺傳算法相結(jié)合的混合算法可將機(jī)會(huì)約束規(guī)劃模型轉(zhuǎn)化為確定性問(wèn)題進(jìn)行求解,為含不確定性因素的光伏微網(wǎng)系統(tǒng)經(jīng)濟(jì)調(diào)度問(wèn)題提供了一個(gè)有效方法。(3)在多時(shí)間尺度的光伏微網(wǎng)隨機(jī)優(yōu)化調(diào)度問(wèn)題中,考慮到由于可再生能源及負(fù)荷的預(yù)測(cè)誤差而造成的日前調(diào)度和實(shí)時(shí)調(diào)度的偏差,提出了一種考慮需求側(cè)管理的多時(shí)間尺度的光伏微網(wǎng)隨機(jī)優(yōu)化調(diào)度模型,以減輕由光伏和負(fù)荷側(cè)的預(yù)測(cè)誤差引起的功率波動(dòng)。在超短期日前光伏和負(fù)荷功率預(yù)測(cè)的基礎(chǔ)上,基于機(jī)會(huì)約束規(guī)劃理論建立日前的經(jīng)濟(jì)調(diào)度模型,并在實(shí)時(shí)調(diào)度時(shí)采用實(shí)時(shí)的功率調(diào)整策略對(duì)日前的調(diào)度方案進(jìn)行修改。仿真結(jié)果分析表明通過(guò)此種多時(shí)間尺度的調(diào)度方法可以有效補(bǔ)償系統(tǒng)實(shí)時(shí)運(yùn)行的功率波動(dòng),保證系統(tǒng)運(yùn)行的安全可靠性,提高了方案的可行性。
[Abstract]:Since the beginning of the 21st century, photovoltaic microgrids, which have the ability to manage and control energy autonomously, have been developed rapidly as one of the main forms of renewable energy utilization. However, because of the single power supply of the photovoltaic microgrid, the output force fluctuates greatly, and the energy storage capacity of the configuration is generally limited, the smaller power supply or load fluctuation will have a great impact on the optimal operation and scheduling of the photovoltaic microgrid. Therefore, according to the characteristics of photovoltaic microgrid, it is necessary to establish an energy optimal scheduling model suitable for the safe and stable operation of photovoltaic microgrid, and to study the corresponding solution method for simulation and analysis, which has important theoretical and engineering significance. This paper first introduces the microgrid structure including distributed power supply, and expatiates the history and research status of microgrid energy optimal scheduling from three aspects: microgrid optimal scheduling strategy, optimal scheduling model and its model solving algorithm. On this basis, the multi-objective optimal scheduling problem with demand-side management (demand side), stochastic optimal scheduling problem with prediction error and stochastic scheduling problem with multiple time scales for photovoltaic microgrid are studied. The main research work is as follows: (1) in the multi-objective scheduling problem of photovoltaic microgrid considering demand-side management. Firstly, the operation state of typical photovoltaic microgrid is briefly analyzed, and the effect of considering the energy optimization of photovoltaic microgrid and not considering the participation of electric vehicle on the operation economy of the system is discussed. The optimal scheduling models without electric vehicle charging and with electric vehicle charging are established, and the power balance, the charging state of energy storage system and the time range of load transfer are considered synthetically. Based on the charging time constraints of electric vehicles, an optimization method based on the non-dominated sorting genetic algorithm (NSGA- 鈪，

本文編號(hào)：2133447