Research On Detection Method And Application For Regional Power Grid Voltage Stability

The operating point of power system draws nearer to its stability limit under severalconditions, such as the rapid development of the power industry in our country, the widerdomain large-scale grid interconnection, the deepening of the electricity market-orientedreforms, and the input of the intermittent renewable energy sources. Moreover, naturaldisasters have repeatedly caused accidents of regional power grid voltage collapse, whichmake voltage stability problem of power system more prominent than ever before. Generally,if the bus voltage angle, power system frequency and the amplitude of voltage are relativelystable, the system protection and warning devices cannot predict or act in advance when thevoltage instability of the power grid happens. As a result, how to accurately online identifyand predict the regional power grid voltage stability on the complex operation conditions hasbecame one of the great challenges that the power dispatching departments are faced with.PMU technology provides conditions for the analysis and prediction of voltage stabilitydynamics, which are based on wide area simultaneous measurement data of the regional powergrid. On the basis of the study of synchronous phasor measurement algorithm, the model ofthe analysis and prediction of voltage stability dynamic characteristics are established in thispaper. Thus, online identification and prediction system of the regional power grid voltagestability is developed. It provides guidance for the voltage stability control to the powersystem operation management department.The main contents are as follows?(1)The DFT algorithm causes the measurement errors due to the frequency leakage andfrequency mismatch when the frequency of the sampling signals changes. In this study, anonline model is developed which is using PMU to synchronously collect the noise content anddynamic characteristics of each phase voltage and current signal. It uses dynamiccharacteristics variance and noise variance to quantify the sampling system real-time state.Therefore, frequency adaptive model and solve algorithm are established. In this way, thetracking algorithm has both fast tracking and good anti-noise characteristics under dynamicconditions. With the analysis of the interaction between the dynamic characteristics ofsampling signal and decaying DC component, the dynamic fundamental wave filter algorithm is established. It can estimate the dynamic characteristics of the current and voltage phasesdecaying DC component. Based on the improved frequency tracking algorithm of DFT, thecombination of the adaptive frequency tracking model and DC component tracking model cancalculate accurate PMU measurements.(2) In order to establish the optimization location method of a regional grid PMUmonitoring which is based on node voltage weak state analysis index and its algorithm, thenonlinear programming model and assessment method for the voltage stability critical pointand the weakness indicator of the node are studied. With electric power system grid structureparameters in a particular area, a research for the nonlinear function relations is accomplishedwhich include the modules in grid and the dynamic process of load and voltages of nodes,considering that the voltage instability starts from local diffusion to the collapse of the whole.The model of the relationships between the process of system approach voltage instabilitycritical state and the voltage weakness status indicators is studied according to the regionalgrid numerical simulation data. A detection model is built up to accurately identify theregional grid weak nodes in accordance with the degree of node voltage weakness. The datacomes from the sizes of sensitivity indicators and sensitivity change rate indicators, whichcharacterize the change speed of all nodes sensitivities. Therefore, the optimizationconfiguration method of the not completely observable PMU monitoring points is establishedon the basis of the monitor at voltage weak areas in order to insure the grid observable andstability of the voltage.(3) Based on synchronized pharos measurements and the modules and algorithms of nodevoltage stability rapid judgment, the nonlinear time series prediction modules and algorithmsof voltage stability trajectory are studied. The parameters of real time grid equivalent model,aimed at the data of power grid local node synchronous pharos measurement, are calculated byinvertible decoupling and impedance modulus algorithms with synchronous voltage andcurrent pharos of nodes and related loads, and by which the whole system could be equivalentto a two nodes grid. The Thevenin equivalent model which includes load nodes andimpedance model limit algorithms is used to deduce a synchronous grid voltage stability indexfor rapid judgment of grid voltage stability. In view of the grid voltage and current real-timemonitoring data, the grid dynamic voltage nonlinear time series analysis method is studied toreconstruct phase point in phase space trajectory prediction model and to determine the phase space and the parameters of the voltage stability of nonlinear system. Thus, the voltagestability global forecast model can be set up by using the nonlinear mapping of support vectormachine model for dynamic voltage in the global nonlinear time series prediction. A nonlinearregression method of nonlinear time sequence of the phase trajectory in phase space isestablished to predict the dynamic voltage stability of power grid trajectory.(4) In accordance of the requirements of Wide Area Measure System (WAMS) and theactual power grid operating conditions, establishment of regional power grid voltage stabilitycriterion system is studied, and software for regional power grid voltage stability onlinedetection and prediction is developed. On the basis of an online detection model of regionalpower grid single node voltage stability, multiple node voltage stability online identificationmodels and nonlinear time series prediction model of grid voltage stability, a mathematicalmodel and an algorithm for voltage stability criterion are established. Based on the DSP,Visual C++and Oracle database technologies, a regional grid voltage stability detectionapplication system platform is developed with the improved frequency tracking dataacquisition algorithm and mathematical model of voltage stability criterion system as the core.It can accomplish the online detection and short-term prediction of regional grid voltagestability and its relevant parameters with real-time measurements of the grid nodes. It can alsobe used for providing voltage stability criterion and forecast data in real time to the gridoperation management departments as reference and guidance to make adjustments to theelectric power system voltage stability control strategies.(5) According to the technological standard in “Automatic Voltage Control TechnicalSpecification of The Grid (Q/GDW747-2012)” of State Grid Corporation of China, theregional grid voltage stability criterion system is applied to the grid operation management,combined with the actual situation of Liaoning grid voltage stability controlling system. Theregional grid voltage stability detection application system is successfully put into practice at a500kV substation in June2012. The hardware and software of system are stable and the resultsof online voltage stability criterion and warning in actual operation are reasonable andacceptable. As a conclusion, the regional grid voltage stability judging system and its onlinevoltage stability identification parameters and prediction results can meet the practicalengineering requirements of regional power grid voltage stability control.