高渗透率分布式新能源发电接入下的电网谐波不确定性交互影响机理研究

The grid is gradually dominated by power electronic inverter-based devices such as renewable energy generation, AC-DC hybrid transmission and custom power technology. Its operational characteristics has changed significantly. The harmonic interference, previously generated locally by customers in the distribution network, is propagating obviously in the whole grid. With the higher levels of variable renewable energy, it presents a complex spatial and temporal distribution characteristics. Moreover, it perhaps results in resonance and unstability in the grid with distributed new energy power generation..In the grid with multi harmonic sources, the harmonic power flow is the most valid method to study power system harmonic. However, it fails to process the variable and uncertainty of harmonic distribution which cannot be ignored. The grid operators have difficulty in choosing the best way to integrate the distributed generation. In this case, they also need to design the market products providing economic incentives to the harmonic pollution customers in the market regulation. At present, the Monte Carlo method is used to analyze the randomness of harmonic power flow by sampling parameters and calculating a large number of deterministic operating scenarios. Owing to the difficulty of acquiring the parameter probability distribution and the strength of calculation, the Monte Carlo method generally used in small scale grid..The project will use Affine Arithmetic to build the uncertain model of harmonic source and calculate the interval harmonic flow. It extracts the distinct feature based on interval principal component analysis and accordingly distinguish the typical mode from the harmonic monitoring data by clustering procedure. Based on interval correlation analysis the coupled harmonic volt-ampere characteristic will be decoupled into harmonic constant-current source, voltage controlled harmonic current source and complex harmonic source. According to the attributes of different typical operating modes, it establishes harmonic interval model and affine model, and the next, the parameters of the model are identified. .It also builds the complex affine model of wind turbine considering the output uncertainty. It deduces the harmonic power flow equations in the form of complex affine by Ybus-Gauss method. Modified affine arithmetic will be used to solve the equations iteratively. The similar noise elements are merged to improve the calculation speed while maintaining high calculation precision. And the centrosymmetric convex polygon according to the complex affine will be drew to compute the amplitude and phase angle interval of node voltage precisely. .The project will build a test system to verify the uncertain model and complex affine harmonic power flow.

当前电网呈现出明显的电力电子化趋势，加上分布式新能源发电的波动性、随机性与谐波干扰负荷的不确定性交互影响，使得谐波产生机理、传播特性更加复杂，更易引发谐振以及稳定性问题，严重制约了新能源发电的接入规模。.项目针对分布式新能源发电和谐波源交互影响的不确定性，提出并解决谐波源不确定性发射机理和电网谐波不确定性传播机理两个科学问题。拟采用区间主成分分析提取谐波源运行特征量并实现典型模态聚类，基于区间相关分析解耦各模态的谐波伏安特性，建立区间型谐波源、单一受控型谐波源及混合型谐波源的仿射模型并辨识参数，实现基于数据驱动的谐波源统计建模；根据风电、光伏运行机理建立其基波功率和谐波特性仿射模型，提出复仿射谐波潮流理论和算法，以仿射修正算法、仿射多边形理论为基础减小计算结果的保守性。项目将基于仿射算术提出、研究和验证谐波源和新能源发电交互影响的不确定性理论，为含分布式新能源发电的优化提供理论和技术支撑。

项目围绕谐波源不确定性发射机理和电网谐波不确定性传播机理两个科学问题开展研究工作，具体研究内容主要包括：（1）研究谐波源仿射计算模型的建模理论和一般方法；（2）研究复仿射谐波潮流理论和求解方法；（3）分布式新能源发电与谐波源之间的谐波交互影响机理；（4）理论研究成果的完备性和保守性仿真验证方法。.项目主要完成了以下几方面工作：（1）提出基于主要特征量聚类的典型运行模态提取方法，分别建立了谐波源发射的仿射模型和风电机组出力的仿射模型，给出了典型模态参数的辨识方法；（2）提出复仿射谐波潮流算法，利用仿射修正算法、同类噪声元合并等手段，在保证结果完备性的同时减小保守性，并在电网谐波不确定性传播、新能源规划、多能流系统不确定性分析、不确定性优化等方面实现了应用；（3）以仿射谐波谐波潮流算法为手段，研究了谐波源运行不确定性与新能源发电功率波动性的交互影响，提出了计及出力不确定性的新能源谐波模型，分析交互影响对电网谐波的影响规律，提出多场景下的分布式新能源发电优化接入模型；（4）建设了数据驱动的不确定性建模分析系统、复仿射谐波潮流软件包，以及Etap、PSCAD平台上的验证模型。.项目在理论和方法上突破了传统的确定性谐波交互影响机理分析构架，能够计及谐波源运行的不确定性、分布式发电的功率波动性及其与谐波输出特性之间的相关性，为分析分布式新能源发电和谐波源的不确定性交互影响机理建立理论基础。以仿射多边形理论为基础，提出仿射修正算法、同类噪声元合并等方法，保证仿射方程组迭代求解的收敛性及保守性，解决区间潮流算法的保守性问题，为高渗透率分布式新能源发电接入下的电网不确定性优化提供技术手段。.依托该项目培养硕士研究生18名，目前在读博士研究生2名、硕士研究生5名，授权发明专利7项、受理10项，发表SCI检索期刊论文5篇、EI检索期刊论文19篇、EI检索会议论文4篇，录用SCI检索期刊论文1篇、EI检索期刊论文3篇。