【作者】 张坤；

【导师】 毛承雄； 陆继明；

【作者基本信息】 华中科技大学 ， 电力系统及其自动化， 2013， 博士

【摘要】 随着风电穿透率的不断增加,风电系统与电网之间的相互影响也变得越来越大。其主要表现在：一方面,由于风能资源具有波动性,导致风电出力也具有波动性,这在风电穿透率较高的情况下会对电网的电能质量及其稳定性产生较大的影响。另一方面,随着风电在电网中所占比例不断增大,若风电机组在电网发生故障时仍采取被动保护式的解列方式,则会增加整个系统的恢复难度,甚至可能加剧故障,严重影响到电网的安全运行。为此,在风电系统中配置一定容量的储能装置将起到平抑功率波动、提高低电压穿越能力、维持发电／负荷动态平衡、保持电压／频率稳定的作用,从而实现风力发电系统安全、经济、高效、优质地运行。本文将针对基于储能的直驱风力发电系统优化控制策略开展研究。1、根据直驱风力发电系统的运行特点,针对目前风电系统存在的功率平抑以及低电压穿越这两大核心问题,提出了三种综合控制策略,即：储能系统的双模功率控制、储能系统的直接功率控制以及储能系统的直接电压控制。在这三种控制策略下,直驱风电系统在电网正常运行时可实现平滑的有功输出,而当电网发生故障时,且同时具有较强的低电压穿越能力。并且,在储能系统直接电压控制的基础上,结合超级电容器和蓄电池在技术性能上的互补优势,进一步提出了一种混合储能系统的综合控制策略,从而可以大幅度地提高系统的经济性能和技术性能。2、根据风电低电压穿越运行的相关技术要求,当电网的故障持续时间超出一定范围后,风电系统将被允许脱离电网。为了最大利用风能资源以及提高当地负荷的供电可靠性,针对基于超导储能的直驱风力发电系统提出了一种协调控制策略,使其能够实现并网／离网这两种运行模式之间的快速平滑切换。3、结合风电功率的短期预测以及储能系统荷电状态的实时检测,提出了一种变平滑时间常数的双重模糊控制策略,从而能够有效地控制储能系统的充放电功率大小,使其在平滑风电功率波动的同时,还能够避免出现过度充电或深度放电的状况,并向着合适的荷电状态转变。4、根据风电系统接入电网的相关技术规定,提出了一种能够定量反映功率曲线平滑度的判据标准,并在此基础上建立了储能系统特性参数-风电功率平滑度的短期神经网络模型,然后综合考虑了储能系统的技术性能和经济成本,建立了储能系统特性参数-风电功率平滑度、储能系统成本特性的长期数学模型,进而通过遗传算法得到储能系统最佳的特性参数组合。更多还原

【Abstract】 The interplay between wind power system (WPS) and grid is becoming increasinglysignificant with the increased penetration of wind power, as is shown in two aspects: Onone hand, since the wind energy resources are quite fluctuant and unstable, thecorresponding wind-generated electricity is also full of fluctuations, which has greatinfluence on the quality and stability of the grid power on condition that the penetration ofwind power is relatively high; on the other hand, as the proportion of wind power in thewhole electronic grid is gradually increasing, if the passively protective disconnectionmethod is still adopted when grid faults occur, the difficulty of recovering the whole systemwill be increased, and to make things worse, the faults may become more serious, whichwill have significantly negative effects on the safety of grid operation. Therefore,integrating the energy storage system (ESS) with certain capacity will help to smooth windpower fluctuations, enhance the low voltage ride-through (LVRT) capability, maintain thedynamic balance of power generation/load, and keep the voltage and frequency stable, andhence realize the safe, economic, effective and high-quality operation of WPS. In this paper,studies are carried out on the optimizing and controlling strategies toward energy storagebased direct drive WPS.1. According to the features of direct drive WPS operation, based on the two essentialissues (smoothing wind power fluctuation and low voltage riding-through) existing incurrent WPS, three comprehensive control strategies are proposed: Dual-mode control,direct power control and direct voltage control of energy storage system (ESS). Under thesethree controlling strategies, the direct drive WPS can output smoother active power undernormal conditions and maintain relatively high low voltage ride-through capability whengrid faults occur. More importantly, based on the direct voltage control of ESS andcombining the complementary advantages of supercapacitors and batteries to each other interms of technical performance, a comprehensive strategy for controlling hybrid ESS isfurther brought up, which can hence greatly enhance the economic and technical performance of the system.2. According to the related technical requirements of low voltage ride-through forwind power, the WPS will be permitted to be cut off from the grid when the duration of gridfault goes beyond a certain range. In order to make the most of wind resources and improvethe reliability of the local loads power supply, a coordinated controlling strategy forsuperconducting magnetic energy storage-based direct drive WPS is introduced, so as torealize the fast and smooth switch between grid-connected and off-grid modes.3. According to the short-period prediction of wind power and real-time detection ofthe state-of-charge for ESS, a double fuzzy logic control strategy optimizing themanagement of the ESS is proposed to effectively control the charging and dischargingpower of ESS. Therefore, the wind power fluctuations can be smoothed, overcharge anddeep discharge can be effectively avoided, and the transformation towards suitablestate-of-charge can be realized.4. Considering the relevant technical regulations of grid-connected WPS, this paperfurther comes up with a judgmental standard to quantitatively reflect the level-of-smoothing of power curves, based on which the short-period neural network model of ESScharacteristic parameters–level-of-smoothing of wind output power is established. Thentaken the technical and economic demand into the comprehensive consideration, along-term mathematical model of ESS characteristic parameters-lever-of-smooth of windoutput power and ESS cost is then established so that the optimized characteristicparameters of the ESS can be obtained through genetic algorithm.更多还原