电网可用输电能力的计算方法研究

发布时间：2019-05-15 15:41

【摘要】：可用输电能力(Available Transfer Capability, ATC)不仅是指导电力市场交易顺利进行的重要指标,也是评估系统安全可靠性的重要测度,对电力系统调度运行与规划建设具有重要的指导作用。近年来,与ATC相关的输电断面辨识、确定性ATC计算、概率性ATC计算等问题在分析计算领域受到了广泛关注,但在大型区域电网互联、新能源发电系统广泛接入的条件下,传统方法在全面考虑各方面的影响因素、计算速度与计算精度的平衡、不确定性因素的处理等方面遇到了新的挑战。一方面,弱互联的大型电网存在更多的稳定性问题,为了考虑暂态稳定性等动态约束条件,传统确定性方法的计算时间将大幅度增加,难以满足实用需求；另一方面,系统运行方式的复杂多变导致不确定性因素大幅度增加,传统概率分析方法在计算速度上遇到了瓶颈,而且针对风电场、光伏电站等新增不确定性因素也需要建立更加适用的概率分析模型。本文以输电断面辨识、确定性ATC计算和概率性ATC计算为研究重点,从计算方法的角度开展研究。提出了一种基于复杂网络理论的输电断面快速辨识方法。该方法基于电网的小世界特性提出了输电介数指标用于从输电线路中辨识关键输电环节,然后采用GN算法对电网分区得到与关键输电环节相关的分区断面,最后通过分区断面的优化组合快速辨识输电断面同时给出重要性排序。对CEPRI-36系统与某省级电网的算例表明,新算法计算量小,辨识结果能够完整地反映出电网拓扑结构与运行方式的特点并计及不确定性因素的影响,对于复杂多变的大电网具有很好的适应性。提出了一种考虑方向性和风险性的大型互联电网ATC快速计算方法。该方法在以重复潮流为基本框架的ATC计算过程中,引入了3种具有实际指导意义的功率增长方向以考虑不同功率增长过程对ATC计算结果的影响,同时考虑暂态失稳概率,求取系统在承担不同失稳风险时的ATC。此外,该方法还通过优化校验过程与自适应步长控制提高了计算速度。某省级电网的算例表明,所提算法能够满足大型互联电网在线应用需求,不同功率增长方向和失稳风险下的ATC结果将为运行人员提供更加全面的输电能力信息。提出了一种重复潮流与连续潮流相结合的主从迭代ATC改进算法。该方法以重复潮流为主迭代过程,自适应地计算迭代步长和功率增长方向,当潮流不收敛时转入连续潮流的从迭代过程,并将重复潮流的步长与方向作为连续潮流的功率调整目标,从而既发挥了重复潮流灵活性的特点,也确保了潮流轨迹的收敛性和稳定性。此外,还将连续潮流应用到N-1约束校验过程和多断面功率控制。CEPRI-36系统和华中-华北联网系统的算例结果表明,改进算法能够灵活地考虑多方面影响因素,计算速度快,而且对潮流收敛性较差的大型电网具有很好的适应性。提出了一种基于拉丁超立方采样和聚类思想的概率ATC快速计算方法。该方法在蒙特卡罗仿真的采样过程中,采用拉丁超立方采样提高采样效率,并利用Cholesky分解处理输入随机变量的相关性；对于得到的大量样本,采用聚类分析技术得到相似度较高的分类场景,推导出最优潮流灵敏度快速计算这些分类场景的ATC。IEEE-RTS系统的算例结果表明,所提算法计算速度快,计算结果精度高、鲁棒性好,并且针对不同的应用需求可以实现计算精度与计算时间的协调。此外,算例结果还表明风速相关性增强了ATC的波动性,需要在计算模型中予以考虑,以确保分析结果的准确性。提出了一种基于随机响应面法的概率ATC快速计算方法。该方法将随机响应面法应用到概率ATC计算中,将输入随机变量和输出响应表达为标准正态随机变量的函数关系,利用小规模的采样快速得到ATC的概率分布。为了处理分布函数未知且具有相关性的连续随机变量,引入了多项式正态变换结合正交变换技术；在考虑离散随机变量时,将ATC概率分布分解为连续部分与离散部分分别求取。对含风电场的IEEE-RTS系统的算例表明随机响应面法计算速度快,计算精度满足实用需求。针对考虑时序特性的概率ATC计算,提出了一种基于马尔科夫链蒙特卡罗仿真的风光发电功率时间序列模拟方法。该模型直接对风电场、光伏电站的发电功率进行时序状态转移过程建模,采用Gibbs抽样生成单个或多个间歇性电源的发电功率时间序列,将高阶马尔科夫链的思想应用到多间歇性电源模拟以考虑彼此之间的相关性。对德国2家电力公司所管辖的风电场、光伏电站进行了发电功率时间序列模拟,结果表明所提方法具有较高的模拟精度,而且能很好地考虑多个间歇性电源间的相关性。
[Abstract]:Available Transfer Capability (ATC) is not only an important index to guide the smooth progress of the power market, but also an important measure to evaluate the system safety and reliability. In recent years, the problems of transmission cross-section identification, deterministic ATC calculation and probabilistic ATC calculation related to ATC are widely concerned in the field of analysis and calculation, but under the condition that the large-scale regional power grid is connected and the new energy power generation system is widely accessed, The traditional method has encountered a new challenge in the aspects of the comprehensive consideration of the factors of the influence, the balance between the calculation speed and the calculation accuracy, the processing of the uncertainty factors, and so on. On the one hand, there are more stability problems in the weak-connected large-scale power grid. In order to consider the dynamic constraint conditions such as transient stability, the calculation time of the traditional deterministic method will be greatly increased, and it is difficult to meet the practical demand; on the other hand, The complex and changeable system operation mode results in a large increase in the uncertainty factor, and the traditional probability analysis method has a bottleneck in the calculation speed, and a more suitable probability analysis model is also required for the newly added uncertainty factors such as the wind farm and the photovoltaic power station. In this paper, based on the identification of power transmission section, the calculation of deterministic ATC and the calculation of probability ATC, the research is carried out from the angle of the calculation method. A rapid identification method of power transmission section based on complex network theory is proposed The method is based on the small world characteristics of the power grid, which is used for identifying the key transmission link from the power transmission line, and then using the GN algorithm to obtain the partition off related to the key transmission link by the GN algorithm. and finally, a transmission section is quickly identified by the combination of the section of the section and the importance bank is given at the same time An example of the CEPRI-36 system and a provincial power grid shows that the calculation of the new algorithm is small, and the identification result can completely reflect the characteristics of the power grid topology and the operation mode and take into account the influence of the uncertainty factors, and has good adaptability to the complex and changeable large power grid. In this paper, a kind of large-scale interconnected power grid ATC rapid meter considering the direction and risk is put forward. In the process of ATC calculation with repeated power flow as the basic frame, three kinds of power growth direction with practical guiding significance are introduced to take into account the influence of different power growth process on the calculation result of ATC, and the transient loss is considered. When the system is at the risk of different instability, the stability probability is obtained. ATC. In addition, the method is also improved by the optimization of the verification process and the adaptive step control An example of a provincial power grid shows that the proposed algorithm can meet the demand of on-line application of the large-scale interconnected power grid, and the ATC results in different power growth direction and the instability risk will provide more comprehensive power transmission for the operators. Force information. A master-slave iterative AT with a combination of a repeated flow and a continuous flow is presented. C. The method uses the repeated power flow as the main iterative process to adaptively calculate the iteration step size and the power growth direction, and when the power flow does not converge, the iterative process of the continuous power flow is transferred, and the step length and the direction of the repeated power flow are taken as the work of the continuous power flow. The goal of rate adjustment is not only the characteristic of the flexibility of the repeated power flow, but also the convergence of the power flow path. Performance and stability. In addition, the continuous flow is applied to the N-1 constraint checking process and the multi-break The results of the CEPRI-36 system and the Central China-North China networking system show that the improved algorithm can take into account many factors, the calculation speed is fast, and the large power grid with poor convergence is very important. Good adaptability. A probabilistic AT based on Latin hypercube sampling and clustering is presented. C. In the sampling process of the Monte-Carlo simulation, the sampling efficiency is improved by using the Latin hypercube sampling, and the correlation of the input random variables is processed by using the Cholesky decomposition process, and the similarity degree is obtained by the clustering analysis technique for a large number of samples obtained. The experimental results of the IEEE-RTS system show that the proposed algorithm is fast, the results are high, the robustness is good, and the calculation accuracy can be achieved for different application requirements. The calculation time is coordinated. In addition, the results of the example show that the correlation of wind speed increases the fluctuation of ATC and needs to be taken into consideration in the calculation model to ensure the distribution. The accuracy of the results is presented in this paper. The probability of the random response surface method is presented. In the method, the random response surface method is applied to the probability ATC calculation, and the input random variable and the output response are expressed as a function relation of the normal normal random variable, and the small-scale sampling is used to quickly In order to deal with a continuous random variable with an unknown distribution function and having a correlation, a polynomial normal transformation combined orthogonal transform technique is introduced, and when a discrete random variable is considered, the ATC probability distribution is decomposed into a continuous part The calculation results of the IEEE-RTS system with the wind farm show that the calculation speed of the random response surface method is fast, This paper presents a kind of wind and light emission based on the Markov chain Monte-Carlo simulation for the calculation of the probability ATC considering the time-sequence characteristics. The electric power time series simulation method is used for modeling the power generation power of the wind farm and the photovoltaic power station directly, Power generation power time series and application of high-order Markov chain to multi-intermittent power supply simulation In order to consider the correlation of each other, the power generation power time series of the wind farm and the photovoltaic power station under the jurisdiction of 2 electric power companies in Germany are simulated. The results show that the proposed method has high simulation precision and can be well considered.
【学位授予单位】：华中科技大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TM744

【参考文献】