基于数据驱动的一次调频能力建模与优化决策

Data-Driven Modeling and Optimization of Primary Frequency Modulation Ability

文章设计了一套基于数据驱动的一次调频优化运行评价决策系统。通过在线获取发电机主变高压侧频率信号、负荷信号,抽选具代表性的一次调频事件,可深度还原《并网发电厂辅助服务管理实施细则》和《发电厂并网运行管理实施细则》,并建立不同形式下的边界考核条件,模拟电网考核过程,对火电机组一次调频能力进行建模。同时,通过专家库比对、智能算法在线计算,可诊断出一次调频性能劣化原因及考核原因,推送诊断结果,实现控制逻辑、控制参数的闭环优化,提升一次调频整体性能。

This paper designs a decision-making system for optimization of primary frequency modulation.By obtaining the frequency of the high voltage side of the main transformer as well as the load,the system selects representative events and can thus restore the″two rules″in depth.It also sets the boundary conditions of different forms,simulates the power grid assessment process and models the primary frequency modulation ability of the thermal power unit.Moreover,through expert library comparison and intelligent algorithm calculation,the system diagnoses the reasons for degradation in primary frequency modulation performance and assessment,delivers the results and realizes the closed-loop optimization of control logic and control parameters.The overall performance of primary frequency modulation is thus improved.