基于故障穿越演化特性的SVG电磁暂态模型测辨方法

Identification method of an electromagnetic transient model for an SVG based on fault ride-through evolution characteristics

建立准确的静止无功发生器(static var generators,SVG)白盒电磁暂态仿真模型是分析电网电压稳定特性的前提。然而,由于SVG的控制器结构和参数保密,其建模大都基于典型控制结构和参数,模型的暂态输出特性与实际差异较大。针对上述问题,提出了基于SVG厂家封装黑盒模型故障穿越(fault ride-through,FRT)演化特性的电磁暂态模型测辨方法。首先,分析了厂家黑盒模型的拓扑特征,通过多工况故障穿越响应测试,厘清了其故障穿越演化特性。然后,通过分析不同控制环节暂态切换过程对SVG故障穿越响应特性的影响和作用途径,提出了基于SVG故障穿越响应演化形态的控制器结构辨识方法。通过分析SVG不同控制环节参数对其故障穿越响应特性的分阶段作用原理,提出了基于故障穿越响应幅值的控制器参数分步辨识方法,形成了SVG的白盒化电磁暂态模型测辨方法体系。最后,将建立的不同型号白盒仿真模型与对应厂家黑盒模型进行了故障穿越响应特性对比分析,发现其误差远小于现行标准的允许误差,证明了提出方法的有效性和通用性。

Establishing an accurate white box electromagnetic transient simulation model for static var generators(SVGs)is a prerequisite for analyzing the voltage stability characteristics of the power grid.However,existing models often use generic control structures and parameters because of confidentiality issues of the SVG controller,resulting in significant discrepancies between transient output characteristics and actual performance.To solve these issues,this paper proposes an electromagnetic transient model identification method based on the fault ride-through(FRT)evolution characteristics of the SVG manufacturer packaged black box model.First,this paper analyzes topological characteristics of the manufacturer’s black box model,and clarifies its FRT evolution characteristics through multi-condition FRT response testing.Then,by analyzing the impacts and pathways of transient switching processes of different control parts of the SVG,a controller structure identification method based on the evolution form of the FRT response is proposed.By analyzing the influencing mechanism of different SVG control parameters on FRT response characteristics by stages,a step-by-step identification method for controller parameters based on FRT response amplitude is proposed,forming a white box electromagnetic transient model identification method system for the SVG.Finally,a comparison is made between the established white box simulation models of different models and the corresponding manufacturer’s black box models in terms of FRT response characteristics.This finds that the error is much smaller than the allowable error of the current standard,proving the effectiveness and universality of the proposed method.