基于网络化Euler-Lagrange系统的电力系统暂态稳定控制方法研究

It is critical to secure power system’ operation for social stability and economic prosperity. Conventionally, centralized control is employed to enhance power systems’ transient stability. However, with scale expanding and operation environment complicating, centralized control has not already adapted to the need of power system’s stability, even is infeasible. Actually, enhancing power systems’ transient stability is to force all generators’ power angles to theirs steady-state’s vales and rotor angular speeds to the rated speed, which indeed is a second-order consensus problem. A model for enhancing power systems’ transient stability based on Networked Euler-Lagrange Systems is established. Transient instability is caused by the asynchrony between generator clusters. Based on the instability mechanism and considering communication cost for control, a practical and effective consensus algorithm based on second-order neighbors' information is proposed, in which the first-order neighbors and second-order neighbors are decided by coherence generators identification and clustering. Parameters and communication network topology vary have significant influence on the performance of consensus algorithm. Through Lyapunov stability theory, the effects of parameters vary and communication network topology vary on the convergence performance and convergence rate are investigated. According to the investigation, the optimization of parameters and communication network topology is studied.

保障电力系统稳定运行是关乎社会稳定和经济发展的重大问题。传统暂态稳定控制大都采用集中式方式。随着电力系统规模扩大和运行环境复杂化，集中式控制已经不能完全适应电力系统运行的需要，甚至是不可行的。提高电力系统的暂态稳定性实质上是要控制发电机趋向相同的转子角速度和稳态运行点的功角，是一个二阶系统一致性问题。据此建立了适合于研究多机电力系统暂态稳定控制问题的网络化Euler-Lagrange系统模型。暂态失稳大多表现为机群与机群之间失去同步，根据该机理并综合考虑通信成本，提出了较实用的二层邻居信息的一致性算法，其中第一层邻居和第二层邻居通过发电机机群同调识别和分群后选取得到。控制算法参数大小和通信网络拓扑等对一致性算法的性能有重大影响。基于Lyapunov稳定性定理等，研究了他们对一致性算法的收敛性和收敛速度的影响，并提出算法参数和通信拓扑的优化方法。

在项目执行期间围绕“基于NELS一致性的功角稳定性控制研究”展开研究。从以电力系统功角稳定控制为研究对象，以提高系统的功角稳定性为目标，通过对比Euler-Lagrange方程与发电机运动方程之间的同异，建立发电机Euler-Lagrange方程。设计基于一致性理论的分布式控制方法，提出采用基于二层邻居信息的通信拓扑。主要研究内容有：.（1）对比分析发电机动态运动方程与Euler-Lagrange方程之间的同异，建立基于网络化Euler-Lagrange 系统的适合于多机协调控制的电力系统模型。.（2）基于电力网络划分，研究发电机机群同调群分组方法。基于发电机同调分群，设计一致性算法；.（3）分析控制算法参数和通信网络拓扑等对系统性能和稳定性的影响，并扩展分布式控制到基于安全域的安全分析方法。仿真验证算法参数和通信拓扑对控制效果的影响。