Multi-dimensional instability dominant factor detection method and system based on cyclic dft method

Abstract

The application discloses a multi-dimensional instability dominant factor detection method and system based on a cyclic DFT method, obtains a detection data set; determines a static voltage stability margin, a static limit current and a static limit power angle of a power distribution network; based on a single-machine control mode, takes voltage control state parameters of the power distribution network as optimization variables, takes the static limit current and the static limit power angle as control variables, sequentially determines a coupling model of multi-dimensional instability dominant factors of the power distribution network, a distribution function of instability factors on stability of the power distribution network, an adaptive optimization model of power distribution network partition scheduling, a historical data distribution function of voltage of the power distribution network, corrects wideband measurement data in the detection data set, and obtains steady-state characteristic quantities under different wideband oscillation modes; when a change amount of the steady-state characteristic quantity under any wideband oscillation mode is greater than the static voltage stability margin of the power distribution network, it is determined that there is instability of the wideband oscillation mode, and the steady-state characteristic quantity under the wideband oscillation mode is taken as a detection result, thereby providing decision support for power grid dispatching.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of power system control, and in particular, relates to a multi-dimensional instability dominant factor detection method and system based on a cyclic DFT method for a high-proportion renewable energy power system. BACKGROUND

[0002] In recent years, the application of renewable energy and power electronic devices in the power grid has made great progress, and the power system is developing towards "high proportion of renewable energy" and "high proportion of power electronic devices", which has become an important technical feature of the new power system. Since the stability analysis of the traditional power system is derived from the characteristics of synchronous machines, it has been difficult to adapt to the rapid development of "double high", and the "double high" system often occurs wide frequency oscillation, which leads to the decrease of system stability and the decline of weak node voltage quality. To solve the above problems, the network type new energy + energy storage construction scheme is proposed to improve the strength and stability of the power grid, which can change the virtual inertia provision mode from passive control to autonomous support. To realize this scheme, multi-dimensional instability dominant factor analysis technology and wide frequency oscillation mode real-time extraction technology are proposed. From the operation principle, the former can achieve power balance function to improve system stability by comprehensively considering multiple main factors that can make the power system unstable to dispatch power, and the latter can accurately analyze the modal frequency to realize wide frequency monitoring function.

[0003] In the prior art, the resolution of traditional frequency analysis and the selection length of data window are closely related. The longer the selection length of data window is, the higher the resolution is. However, a long data window needs a large amount of data calculation, which is not suitable for implementation on a device, and the analysis of frequency and amplitude has great hysteresis, which is not suitable for real-time online dynamic calculation. The power system is a complex whole, and the instability dominant factors are power angle and voltage. It is generally difficult to determine the order of occurrence of voltage stability and power angle stability problems, which may be only one instability state or voltage stability and power angle stability may appear instability state at the same time. For the discrimination of power angle instability, the power system is relatively mature in research, and many discrimination methods are proposed. However, there is less research on the discrimination of voltage instability at the present stage, especially the comprehensive discrimination problem of voltage instability and power angle instability. The results of single research on voltage stability and power angle stability problem also have many deviations. For the discrimination of the two, there is little exploration at home and abroad, and there is no standard that can comprehensively analyze and discriminate voltage stability and power angle stability. There is also no unified statement on the interaction between voltage stability and power angle stability. SUMMARY

[0004] In order to solve the problems in the prior art, the application provides a multi-dimensional instability dominant factor detection method and system based on a cyclic DFT method, a dispatch set for active distribution network voltage control is constructed by analyzing weak nodes, analyzing the influence of renewable resources on static voltage stability of the distribution network and analyzing characteristics of the renewable resources, voltage stability is realized through multi-party decision-making, so that electric energy can be balanced faster, and stability and strength of the power grid are improved.In addition, the system adopts a cyclic discrete Fourier transform (DFT) correction algorithm based on a sliding data window, which can effectively avoid the influence of frequency leakage under the premise of accurate analysis of modal frequency, avoids the hysteresis problem caused by a large amount of data calculation, improves the power grid sensing function, prevents oscillation risk and gives accurate early warning, and provides decision support for power grid dispatching.

[0005] The application adopts the following technical scheme.

[0006] The application provides a multi-dimensional instability dominant factor detection method based on a cyclic DFT method, which comprises the following steps.

[0007] Wideband measurement data is obtained by using a sliding data window, and a wideband oscillation modal phasor in the sliding data window at the current time is determined by using a cyclic discrete Fourier transform method; the wideband oscillation modal phasor in the sliding data window at the next time is predicted based on the obtained wideband oscillation modal phasor in the sliding data window at the current time; if it is determined that wideband oscillation occurs according to the predicted value of the wideband oscillation modal phasor in the sliding data window at the next time, the wideband measurement data in the sliding data window at the current time is recorded as a detection data set;

[0008] According to a power flow calculation result of the distribution network system, static voltage stability indexes of weak nodes and new energy grid-connected points in the distribution network are determined, and the minimum value in the static voltage stability indexes is taken as a static voltage stability margin of the distribution network; curves of active power and reactive current components output by an AC side of the grid-connected converter and curves of the converter power angle are drawn respectively; extreme points on the curves of the active power and the reactive current components are extracted as static limit currents, and extreme points on the curves of the active power and the converter power angle are extracted as static limit power angles;

[0009] The coupling model of the multi-dimensional instability dominant factor of the power distribution network is established based on the single-machine control mode, taking the voltage control state parameters of the power distribution network as optimization variables, and taking the static limit current and the static limit power angle as control variables; the distribution function of the influence of the instability factor on the stability of the power distribution network is obtained based on the coupling model of the multi-dimensional instability dominant factor of the power distribution network; the adaptive optimization model of the power distribution network partition scheduling is established by using the distribution function of the influence of the instability factor on the stability of the power distribution network and the closed-loop control characteristic set of the output current of the grid-connected converter; and the historical data distribution function of the voltage of the power distribution network is obtained based on the adaptive optimization model of the power distribution network partition scheduling, taking the closed-loop control characteristic set of the output current of the grid-connected converter in each scheduling partition and the voltage control state parameters of the power distribution network in each scheduling partition.

[0010] The wideband measurement data in the detection data set is corrected by using the historical data distribution function of the voltage of the power distribution network and the voltage control state parameters of the power distribution network in each scheduling partition, and the steady-state characteristic quantity of the power distribution network under different wideband oscillation modes is obtained; when the change of the steady-state characteristic quantity under any wideband oscillation mode is greater than the static voltage stability margin of the power distribution network, it is determined that the power distribution network is unstable under the wideband oscillation mode, and the steady-state characteristic quantity under the wideband oscillation mode is taken as the detection result of the multi-dimensional instability dominant factor.

[0011] Preferably, the model of the cyclic discrete Fourier transform method satisfies the following relationship:

[0012]

[0013] In the formula, x(m) is the mth frequency domain signal, x(n) is the nth time domain signal, and N is the number of time domain discrete signals. Therefore, the input of the discrete Fourier transform is N time domain discrete signals, and the output is N frequency domain discrete point signals.

[0014] Preferably, if it is determined that wideband oscillation occurs according to the predicted value of the wideband oscillation mode phasor in the next time point sliding data window, the wideband measurement data in the current time point sliding data window is recorded as a detection data set, and the sliding data window is continuously moved to collect sample data, and the wideband measurement data in the current time point sliding data window is recorded as a detection data set.

[0015] If it is determined that wideband oscillation will not occur according to the predicted value of the wideband oscillation mode phasor in the next time point sliding data window, the sliding data window is continuously moved to collect sample data.

[0016] Preferably, the new energy grid-connected point includes a PQ type node and a PV type node; wherein the active power and the reactive power of the PQ type node are opposite to the active power and the reactive power of the load, and the active power and the voltage amplitude of the PV type node are constant values.

[0017] According to the power flow calculation results of the power distribution network system, the static voltage stability indexes of the weak nodes, PQ-type nodes and PV-type nodes are obtained by the following relationship:

[0018]

[0019] In the formula, L j is the static voltage stability index of node j, node j includes weak nodes, PQ-type nodes and PV-type nodes, Z ij , R ij and X ij are the impedance, resistance and reactance of the line between node i and node j, respectively, and satisfy S j , P j and Q j are the apparent power, active power and reactive power of node j, respectively, and satisfy U i is the voltage amplitude of node i.

[0020] Preferably, the minimum value in the static voltage stability indexes of each node is taken as the static voltage stability margin of the power distribution network, and the following relationship is satisfied:

[0021]

[0022] In the formula, is the static voltage stability margin of the power distribution network.

[0023] Preferably, the curves of the active power and the reactive current component of the AC side output of the grid-connected converter and the curves of the power angle of the converter are drawn respectively, including:

[0024] Based on the fundamental frequency equivalent circuit of the AC side of the grid-connected converter in the dq axis rotating coordinate, the active power of the AC side output of the grid-connected converter satisfies the following relationship:

[0025] P out =1.5(V od I Q +V oq I P )

[0026] In the formula, P out is the active power of the AC side output of the grid-connected converter, V od and V oq are the d-axis component and q-axis component of the grid point voltage respectively, I Q is the reactive current component, and I P is the active current component.

[0027] When the output power of the grid-connected converter is maximum, the active power of the AC side output of the grid-connected converter and the power angle δ of the converter satisfy the following relationship:

[0028]

[0029] wherein V g is the grid voltage amplitude, x g is the equivalent reactance between the grid point and the grid, and δ is the converter power angle;

[0030] The curves of active power and reactive current components of the AC side output of the grid-connected converter are plotted based on the above formula respectively, and the curves are P out -I Q , active power and converter power angle curves P out -δ.

[0031] Preferably, the coupling model of the multi-dimensional instability dominant factor of the power distribution network satisfies the following relationship:

[0032]

[0033] wherein M is the coupling function of the multi-dimensional instability dominant factor of the power distribution network; b i is the static limit current on the ith transmission line; n is the number of transmission lines; s(e) is the static limit power angle, and C is the voltage control state parameter of the power distribution network.

[0034] Preferably, the distribution function of the influence of the instability factor on the stability of the power distribution network satisfies the following relationship:

[0035] k l =M+(b i +s(e))

[0036] wherein k l is the distribution function of the influence of the instability factor on the stability of the power distribution network.

[0037] Preferably, the adaptive optimization model of the sub-area dispatching of the power distribution network satisfies the following relationship:

[0038]

[0039] wherein r(w) is the adaptive optimization function of the sub-area dispatching, and Φ is the closed-loop control characteristic set of the output current of the grid-connected converter.

[0040] Preferably, the historical data distribution function of the voltage of the power distribution network satisfies the following relationship:

[0041]

[0042] wherein F is the historical data distribution function of the voltage of the power distribution network, is the optimal solution of the sub-area dispatching model, and corresponds to the dispatching instruction of the dispatching sub-area w, and Φ wa set of closed-loop control characteristics of the grid-connected converter output current in the dispatching partition w, a power distribution network voltage control state parameter of the dispatching partition w, w being the total number of dispatching partitions.

[0043] Preferably, the steady-state characteristics of the power distribution network under different wide-frequency oscillation modes satisfy the following relationship:

[0044]

[0045] In the formula, m is the wide-frequency measurement data in the detection data set, and t is the steady-state characteristics of the power distribution network under different wide-frequency oscillation modes.

[0046] Preferably, the steady-state characteristics of the power distribution network under different wide-frequency oscillation modes include, but are not limited to, the voltage amplitude and phase under each wide-frequency oscillation mode obtained by wide-frequency measurement, the closed-loop control characteristics of the grid-connected converter output current, and the historical data of the power distribution network voltage.

[0047] The application also provides a multi-dimensional instability dominant factor detection system based on a cyclic DFT method, comprising:

[0048] The data processing module is configured to obtain wide-frequency measurement data by using a sliding data window, determine the wide-frequency oscillation mode phasor of the wide-frequency measurement data in the sliding data window at the current time by using a cyclic discrete Fourier transform method, predict the wide-frequency oscillation mode phasor in the sliding data window at the next time based on the obtained wide-frequency oscillation mode phasor in the sliding data window at the current time, and record the wide-frequency measurement data in the sliding data window at the current time as a detection data set if wide-frequency oscillation is determined to occur according to the predicted value of the wide-frequency oscillation mode phasor in the sliding data window at the next time.

[0049] The index establishing module is configured to determine the static voltage stability indexes of the weak nodes and the new energy grid-connected points in the power distribution network according to the power flow calculation results of the power distribution network system, take the minimum value in each static voltage stability index as the static voltage stability margin of the power distribution network, draw the curves of the active power and the reactive current component output by the grid-connected converter on the alternating current side and the curves of the converter power angle, extract the extreme points on the curves of the active power and the reactive current component as the static limit current, and extract the extreme points on the curves of the active power and the converter power angle as the static limit power angle.

[0050] The instability dominant factor detection module is configured to, based on the single-machine control mode, take the power distribution network voltage control state parameters as optimization variables, take the static limit current and the static limit power angle as control variables, and establish a coupling model of the multi-dimensional instability dominant factor of the power distribution network; based on the coupling model of the multi-dimensional instability dominant factor of the power distribution network, obtain a distribution function of the influence of the instability factor on the stability of the power distribution network; based on the distribution function of the influence of the instability factor on the stability of the power distribution network and the closed-loop control characteristic set of the output current of the grid-connected converter, establish an adaptive optimization model of the power distribution network partition scheduling; based on the adaptive optimization model of the power distribution network partition scheduling, take the closed-loop control characteristic set of the output current of the grid-connected converter in each scheduling partition and the power distribution network voltage control state parameters of each scheduling partition, and obtain a historical data distribution function of the power distribution network voltage; based on the historical data distribution function of the power distribution network voltage and the power distribution network voltage control state parameters of each scheduling partition, correct the wide-frequency measurement data in the detection data set, and obtain the steady-state characteristic quantity of the power distribution network under different wide-frequency oscillation modes; when the change quantity of the steady-state characteristic quantity under any wide-frequency oscillation mode is greater than the static voltage stability margin of the power distribution network, it is determined that the power distribution network has instability of the wide-frequency oscillation mode, and the steady-state characteristic quantity under the wide-frequency oscillation mode is taken as the detection result of the multi-dimensional instability dominant factor.

[0051] The model of the cyclic discrete Fourier transform method satisfies the following relationship:

[0052]

[0053] In the formula, x(m) is the mth frequency domain signal, x(n) is the nth time domain signal, and N is the number of time domain discrete signals. Therefore, the input of the discrete Fourier transform is N time domain discrete signals, and the output is N frequency domain discrete point signals.

[0054] If it is determined that wide-frequency oscillation occurs according to the predicted value of the wide-frequency oscillation mode phasor in the next time point in the sliding data window, the wide-frequency measurement data in the sliding data window at the current time point is recorded as a detection data set, and the sliding data window is continuously moved to collect sample data, and the wide-frequency measurement data in the sliding data window at the current time point is recorded as a detection data set;

[0055] If it is determined that wide-frequency oscillation does not occur according to the predicted value of the wide-frequency oscillation mode phasor in the next time point in the sliding data window, the sliding data window is continuously moved to collect sample data.

[0056] The new energy grid-connected point includes a PQ type node and a PV type node; wherein the active power and the reactive power of the PQ type node are opposite to the active power and the reactive power of the load, and the active power and the voltage amplitude of the PV type node are constant values;

[0057] According to the power flow calculation results of the power distribution network system, the static voltage stability indexes of the weak nodes, PQ-type nodes and PV-type nodes are obtained by the following relationship:

[0058]

[0059] In the formula, L j is the static voltage stability index of node j, node j includes weak nodes, PQ-type nodes and PV-type nodes, Z ij , R ij and X ij are the impedance, resistance and reactance of the line between node i and node j, respectively, and satisfy S j , P j and Q j are the apparent power, active power and reactive power of node j, respectively, and satisfy U i is the voltage amplitude of node i.

[0060] The minimum value in the static voltage stability indexes of each node is taken as the static voltage stability margin of the power distribution network, and the following relationship is satisfied:

[0061]

[0062] In the formula, is the static voltage stability margin of the power distribution network.

[0063] The curves of the active power and the reactive current component of the AC side output of the grid-connected converter, and the curves of the converter power angle are drawn respectively, including:

[0064] Based on the fundamental frequency equivalent circuit of the AC side of the grid-connected converter in the dq axis rotating coordinate, the active power of the AC side output of the grid-connected converter satisfies the following relationship:

[0065] P out =1.5(V od I Q +V oq I P )

[0066] In the formula, P out is the active power of the AC side output of the grid-connected converter, V od and V oq are the d-axis component and q-axis component of the grid point voltage respectively, I Q is the reactive current component, and I P is the active current component.

[0067] When the output power of the grid-connected converter is maximum, the active power of the AC side output of the grid-connected converter and the converter power angle δ satisfy the following relationship:

[0068]

[0069] In the formula, V g is the grid voltage amplitude, x g is the equivalent reactance between the grid connection point and the grid, and δ is the converter power angle.

[0070] Based on the above formula, the curves of the active power and the reactive current component of the AC side output of the grid-connected converter are drawn respectively P out -I Q , the active power and the converter power angle curve P out -δ.

[0071] The coupling model of the multi-dimensional instability dominant factor of the distribution network satisfies the following relationship:

[0072]

[0073] In the formula, M is the coupling function of the multi-dimensional instability dominant factor of the distribution network; b i is the static limit current on the i-th transmission line; n is the number of transmission lines; s(e) is the static limit power angle, and C is the voltage control state parameter of the distribution network.

[0074] The distribution function of the influence of the instability factor on the stability of the distribution network satisfies the following relationship:

[0075] k l =M+(b i +s(e))

[0076] In the formula, k l is the distribution function of the influence of the instability factor on the stability of the distribution network.

[0077] The adaptive optimization model of the distribution network partition scheduling satisfies the following relationship:

[0078]

[0079] In the formula, r(w) is the adaptive optimization function of the partition scheduling, and Φ is the closed-loop control characteristic set of the output current of the grid-connected converter.

[0080] The historical data distribution function of the voltage of the distribution network satisfies the following relationship:

[0081]

[0082] In the formula, F is the historical data distribution function of the voltage of the distribution network, is the optimal solution of the partition scheduling model, corresponding to the scheduling instruction of the scheduling partition w, and Φ w is the closed-loop control characteristic set of the output current of the grid-connected converter in the scheduling partition w, The power distribution network voltage control state parameters of the scheduling partition w are scheduled, and w is the total number of scheduling partitions.

[0083] The steady-state characteristic quantities of the power distribution network under different wide-frequency oscillation modes satisfy the following relationship:

[0084]

[0085] In the formula, m is the wide-frequency measurement data in the detection data set, and t is the steady-state characteristic quantity of the power distribution network under different wide-frequency oscillation modes.

[0086] The steady-state characteristic quantities of the power distribution network under different wide-frequency oscillation modes include but are not limited to the voltage amplitude and phase under each wide-frequency oscillation mode obtained by wide-frequency measurement, the closed-loop control characteristics of the grid-connected converter output current, and the historical data of the power distribution network voltage.

[0087] A terminal includes a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method.

[0088] A computer-readable storage medium has a computer program stored thereon, which is executed by a processor to implement the steps of the method.

[0089] The beneficial effects of the present application at least include that, compared with the prior art, the method proposed by the present application is suitable for a multi-dimensional support decision system of a high-proportion renewable energy power system, a scheduling set of active power distribution network voltage control is constructed through weak node analysis, renewable resource influence analysis on static voltage stability of the power distribution network, and renewable resource characteristic analysis, voltage stability is realized through multi-party decision, so that electric energy can be balanced faster, and the stability and strength of the power grid are improved. In addition, the system adopts a cyclic DFT correction algorithm based on a sliding data window, which can effectively avoid the influence of frequency leakage under the premise of accurate analysis of modal frequency, avoids the hysteresis problem caused by a large amount of data calculation, improves the power grid sensing function, prevents oscillation risk and gives accurate early warning, and provides decision support for power grid dispatching. BRIEF DESCRIPTION OF DRAWINGS

[0090] Figure 1 The present application proposes a flowchart of a multi-dimensional instability dominant factor analysis method of a high-proportion renewable energy power system. DETAILED DESCRIPTION

[0091] To make the purpose, technical scheme and advantages of the present application clearer, the technical scheme of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. The embodiments described in the present application are only a part of the embodiments of the present application, not all the embodiments. All other embodiments obtained by those skilled in the art without creative labor on the basis of the spirit of the present application belong to the protection scope of the present application.

[0092] The application proposes a multi-dimensional instability dominant factor detection method based on a cyclic DFT method, which can adapt to multi-dimensional instability dominant factor support decision of a high-proportion renewable energy power system, such as Figure 1 as shown, comprising:

[0093] Step 1, wideband measurement data is obtained by using a sliding data window, wideband oscillation modal phasors in the sliding data window are determined by using a cyclic discrete Fourier transform method, wideband oscillation modal phasors in the sliding data window at the next moment are predicted based on the obtained wideband oscillation modal phasors in the sliding data window at the current moment, and if it is determined that wideband oscillation occurs according to the predicted value of the wideband oscillation modal phasors in the sliding data window at the next moment, the wideband measurement data in the sliding data window at the current moment is recorded as a detection data set.

[0094] The full name of wideband measurement is wideband multi-frequency signal measurement, and its connotation is to realize unified measurement of signals in a wide frequency domain or wide frequency band. At present, wideband measurement is preliminarily defined in related standards: based on high-speed sampling of no less than 12.8 kHz, unified measurement of fundamental wave, harmonic and inter-harmonic signals in the range of 0-2500 Hz of the power grid is realized.

[0095] In a non-limiting preferred embodiment, a cyclic DFT method based on a sliding data window is used to monitor the system for wideband to obtain wideband measurement data. The traditional extraction method has a large amount of data to process, is not suitable for implementation on a device, and has a large lag in frequency and amplitude analysis, which cannot meet the requirements of real-time online dynamic operation. The cyclic DFT method based on the sliding data window can analyze very accurate modal frequencies while retaining the transient change trend information of the amplitude. The cyclic DFT method is used to calculate the modal frequency and modal phasor (including amplitude and phase) for fixed-interval sampling data, which can effectively avoid the influence of frequency leakage.

[0096] In the embodiment, the wideband measurement data is stored in an array, a sliding data window is used to slide on the array, the wideband measurement data in the sliding data window is obtained during the window sliding process, and the wideband measurement data is transformed from the time domain to the frequency domain by using a cyclic DFT method to obtain a wideband oscillation modal; wherein the model of the cyclic DFT method satisfies the following relationship:

[0097]

[0098] In the formula, x(m) is the mth frequency domain signal, x(n) is the nth time domain signal, and N is the number of time domain discrete signals. Therefore, the input of the discrete Fourier transform is N time domain discrete signals, and the output is N frequency domain discrete point signals.

[0099] Based on the model of the cyclic DFT method, each point in x(n) is multiplied by cos(2pi*i*n / N) and sin(2pi*i*n / N) respectively when m=i to obtain the real part and the imaginary part of x(i), wherein i is the sample point ordinal number. Thus, the wide-frequency oscillation mode phasor (including amplitude and phase) is obtained.

[0100] If it is determined that the wide-frequency oscillation occurs according to the predicted value of the wide-frequency oscillation mode phasor in the sliding data window at the next moment, the wide-frequency measurement data in the sliding data window at the current moment is recorded as a detection data set.

[0101] In the present application, the wide-frequency measurement data collected by the sliding data window is used to predict the amplitude and phase of the wide-frequency oscillation mode phasor. If the predicted value meets the system wide-frequency oscillation condition, the wide-frequency measurement data in the sliding data window at the current moment is recorded as a detection data set, and the sliding data window is continuously moved to collect sample data, while the wide-frequency measurement data in the sliding data window at the current moment is recorded as a detection data set. At the next moment, the previous wide-frequency measurement data needs to be detected to find the multi-dimensional instability dominant factor when the wide-frequency oscillation occurs, so as to give an early warning of system instability.

[0102] If it is determined that the wide-frequency oscillation will not occur according to the predicted value of the wide-frequency oscillation mode phasor in the sliding data window at the next moment, the sliding data window is continuously moved to collect sample data.

[0103] In the embodiment, a waveform graph of the equivalent impedance of the grid-connected inverter changing with the current is output through a large number of simulations and experiments. It is found through a large number of simulations and experiments that when the wide-frequency oscillation occurs in the "double high" system, the equivalent impedance of the inverter is negative resistance and capacitance, the phase of the equivalent impedance does not change but the amplitude increases sharply with the increase of the current, and at this time the power system shows subsynchronous oscillation. Under the wide-frequency oscillation characteristics of the "double high" system, the traditional extraction method has a large amount of data processing, is not suitable for implementation on a device, and has a large lag in frequency and amplitude analysis, which is not suitable for real-time online dynamic calculation. The cyclic DFT correction algorithm based on the sliding data window can analyze very accurate modal frequencies while not losing the transient change trend information of the amplitude. The DFT correction algorithm is used to calculate the modal frequency and modal phasor (i.e. amplitude and phase) for the fixed-interval sampling data, which can effectively avoid the influence of frequency leakage. Through the cyclic DFT correction algorithm of the sliding data window, it is determined whether the system is in a wide-frequency oscillation state. If it is in the state, the multi-dimensional support decision system is started.

[0104] Through the cyclic DFT correction algorithm of the sliding data window, the accuracy of the modal frequency sampling can be ensured on the basis of reducing the amount of data processing, and the problem of large lag in frequency and amplitude analysis of the traditional modal sampling can be solved, which is more suitable for device implementation and real-time online dynamic calculation.

[0105] Step 2, according to the power distribution network system flow calculation results, determine the weak nodes in the power distribution network and the static voltage stability index of the new energy grid-connected point, and take the minimum value in each static voltage stability index as the static voltage stability margin of the power distribution network.

[0106] When the node voltage exceeds the limit, voltage collapse phenomenon often occurs at the weakest node, leading to overall collapse of the power system, so more attention should be paid to the weak nodes during power dispatch. The judgment method of weak nodes includes:

[0107] The P-V curve of the node under the power grid operation scenario is generated by using the continuous power flow method, and the inflection point on the P-V curve is taken as the weak node; the voltage gradually decreases with the increase of active power, and the slope is small before the voltage inflection point comes, and the curvature becomes larger after the voltage crosses the inflection point, the node that is more likely to reach the inflection point is weaker, and accordingly the weak node is extracted; the voltage amplitude and active power of the weak node constitute the feature data of the weak node, which satisfies the following relationship:

[0108] Z k =(U k ,P k )

[0109] In the formula, Z k is the feature data of the kth weak node, U k is the voltage amplitude of the kth weak node, and P k is the active power of the kth weak node.

[0110] Generally, there are only PQ nodes and balance nodes in the power grid. When renewable energy is connected to the power grid, the number and type of nodes in the power grid will change. Since the connection of renewable energy to the power grid does not affect the frequency regulation of the system, the output active power can be approximately considered as a constant value, and the operation mode of reactive power and voltage needs to be analyzed according to the actual situation. The new energy grid-connected point includes PQ-type node and PV-type node; among them, the active power and reactive power of the PQ-type node are opposite to the active power and reactive power of the load, and the active power and voltage amplitude of the PV-type node are constant values.

[0111] According to the power distribution network system flow calculation results, the static voltage stability index of the weak node, PQ-type node and PV-type node is obtained by the following relationship:

[0112]

[0113] In the formula, L j is the static voltage stability index of node j, node j includes weak nodes, PQ-type nodes and PV-type nodes, Z ij , R ij and X ijrespectively, impedance, resistance and reactance of the line between node i and node j, satisfying S j , P j and Q j respectively, apparent power, active power and reactive power of node j, satisfying U i is the voltage amplitude of node i.

[0114] The minimum value in the static voltage stability index of each node is taken as the static voltage stability margin of the power distribution network, satisfying the following relationship:

[0115]

[0116] In the formula, is the static voltage stability margin of the power distribution network.

[0117] The static voltage stability index L j of node j is smaller, the static voltage stability of node j is poorer, and the static voltage stability margin of the power distribution network is smaller, which indicates that the voltage of the power distribution network is less stable.

[0118] The most developed renewable energy sources at present include solar energy, wind energy, geothermal energy, biomass energy and ocean energy. These renewable resources have volatility and intermittency, and the distribution of renewable energy in different regions is also different. In order to consider the characteristics of renewable energy in power distribution, the generation uncertainty should be predicted under different time scales and multi-scenario operation requirements. According to the time scale division, it includes medium and long-term prediction and short-term and ultra-short-term prediction. According to the method division, it includes reserve setting method, stochastic optimization method, chance constraint programming method and robust optimization method. The new energy power uncertainty prediction unit is used for clustering historical data to obtain multiple type clusters to describe the power uncertainty.

[0119] In order to cope with the power uncertainty brought by new energy generation, the present application adopts a method of discretization and clustering based on historical data to generate multiple type clusters to describe the power uncertainty. In essence, this method is a random optimization method of multiple discrete scenes. The random optimization method generates random scenes from the probability distribution of uncertain variables (actual and predicted power error of renewable energy) and extracts them, and then discretizes the continuous probability distribution into the form of multiple scene sets.

[0120] In the embodiment, the random optimization method is used to obtain active power and reactive power, and apparent power is calculated; the static voltage stability index of the new energy grid connection point is calculated by using the optimized apparent power, active power and reactive power, so as to realize the elimination of the influence of new energy generation uncertainty on the calculation result of the static voltage stability index.

[0121] Step 3, draw the curves of active power and reactive current component of the grid-connected converter AC side output, and the curves of converter power angle respectively; extract the extreme points on the curves of active power and reactive current component as static limit current, and extract the extreme points on the curves of active power and converter power angle as static limit power angle.

[0122] The grid-connected converter is connected to the grid at one end and to new energy or energy storage equipment at the other end. The distribution network is regarded as a single-machine infinite system, and the equivalent impedance between the grid-connected point and the infinite grid is assumed to be R g , x g , and the grid-connected point phase is 0°.

[0123] The grid-connected point voltage vector V o ∠0° is decomposed to dq axis, and when the q-axis component of the grid-connected point voltage is zero, the output power of the grid-connected converter is maximum, and at this time the d-axis is oriented on the grid-connected point voltage vector V o ∠0°.

[0124] Based on the fundamental frequency equivalent circuit of the renewable resource grid-connected converter AC side in the dq rotating coordinate system, the renewable resource grid-connected converter is equivalent to a single-machine infinite system, and the grid-connected point voltage amplitude satisfies the following relationship:

[0125]

[0126] In the formula, U o is the grid-connected point voltage amplitude, V g is the grid voltage amplitude, I Q is the reactive current component, and I P is the active current component.

[0127] From the fundamental frequency equivalent circuit, the active power output by the renewable resource grid-connected converter AC side satisfies the following relationship:

[0128] P out =1.5(V od I Q +V oq I P )

[0129] In the formula, P out is the active power output by the grid-connected converter AC side, V od and V oq are the d-axis component and q-axis component of the grid-connected point voltage respectively.

[0130] When the phase-locked loop dynamics is ignored, the d-axis component is directly oriented on the grid-connected point voltage vector V o ∠0°, at this time V oq is zero, and the active power output by the grid-connected converter AC side satisfies the following relationship:

[0131]

[0132] reactive current component I Q The following relationship is satisfied between the transformer power angle δ and the active power output on the AC side of the grid-connected converter:

[0133] I Q x g = V g sin δ

[0134] The following relationship is satisfied between the active power output on the AC side of the grid-connected converter and the transformer power angle δ:

[0135]

[0136] In the above formula, the active power P output on the AC side of the grid-connected converter is a function of the reactive current component I and the transformer power angle δ, and based on the above formula, the curves P out -I Q of the active power and the reactive current component and the curves P out - δ of the active power and the transformer power angle are respectively drawn, the extreme point on the P Q -I out curve is extracted as the static limit current, and the extreme point on the P out - δ curve is extracted as the static limit power angle. Q out

[0137] In the embodiment, the static voltage stability margin of the power distribution network is first determined, and then the static limit current and the static limit power angle are taken as the steady-state characteristic quantities of the power distribution network.

[0138] Step 4, based on the single-machine control mode, taking the voltage control state parameters of the power distribution network as optimization variables and taking the static limit current and the static limit power angle as control variables, a coupling model of multi-dimensional instability dominant factors of the power distribution network is established; based on the coupling model of multi-dimensional instability dominant factors of the power distribution network, a distribution function of the influence of instability factors on the stability of the power distribution network is obtained; using the distribution function of the influence of instability factors on the stability of the power distribution network and the closed-loop control characteristic set of the output current of the grid-connected converter, an adaptive optimization model of power distribution network partition scheduling is established; based on the adaptive optimization model of power distribution network partition scheduling, taking the closed-loop control characteristic set of the output current of the grid-connected converter in each scheduling partition and the voltage control state parameters of the power distribution network in each scheduling partition, a historical data space distribution function of the voltage of the power distribution network is obtained.

[0139] Specifically, the coupling model of multi-dimensional instability dominant factors of the power distribution network satisfies the following relationship:

[0140] ​​

[0141] wherein M is a coupling function of the multi-dimensional instability dominant factor of the power distribution network; b i is the static limit current on the ith transmission line; n is the number of transmission lines; s(e) is the static limit power angle, and C is the voltage control state parameter of the power distribution network.

[0142] In the embodiment, the static limit current and the static limit power angle are taken as control independent variables, a single-machine control method of the grid-connected converter is adopted to identify the multi-dimensional instability dominant factor of the power distribution network, and the coupling effect of various factors such as the multi-dimensional instability dominant factor, the load form parameter distribution, and the static voltage stability margin is comprehensively considered.

[0143] When the system is connected to a weak AC system, the influence of the comprehensive weak node on the system stability, the renewable energy uncertainty prediction, and the influence of each factor leading to system instability in the set of factors affecting the system stability are analyzed; therefore, based on the coupling model of the multi-dimensional instability dominant factor of the power distribution network, the distribution function of the influence of the instability factor on the stability of the power distribution network is obtained, and the following relationship is satisfied:

[0144] k l =M+(b i +s(e))

[0145] wherein k l is the distribution function of the influence of the instability factor on the stability of the power distribution network.

[0146] Based on the generalized characteristic analysis method defined by the static characteristic, the closed-loop control characteristics of the grid-connected converter output current are obtained according to the voltage control state parameter of the power distribution network, the power distribution network is divided according to the load form parameter distribution, the adaptive optimization model of the power distribution network partition scheduling is established, and the following relationship is satisfied:

[0147]

[0148] wherein r(w) is the adaptive optimization function of the partition scheduling, and Φ is the set of closed-loop control characteristics of the grid-connected converter output current.

[0149] In the embodiment, the solution of the adaptive optimization model of the power distribution network partition scheduling is the scheduling instruction of each scheduling partition.

[0150] Each scheduling partition has an optimal solution when responding to the multi-dimensional support decision instruction of the power distribution network, and therefore the historical data distribution function of the entire power distribution network voltage satisfies the following relationship:

[0151]

[0152] wherein F is the historical data distribution function of the power distribution network voltage, Φ is the scheduling instruction corresponding to the scheduling partition w for the optimal solution of the partition scheduling model, w is a closed-loop control feature set of the grid-connected converter output current in the scheduling partition w, W is the total number of scheduling partitions.

[0153] Step 5, correcting the wideband measurement data in the detection data set by the historical data distribution function of the power grid voltage and the power grid voltage control state parameters of each scheduling partition, to obtain the steady-state characteristic quantity of the power grid under different wideband oscillation modes; when the change of the steady-state characteristic quantity under any wideband oscillation mode is greater than the static voltage stability margin of the power grid, it is determined that the power grid has the wideband oscillation mode instability, and the steady-state characteristic quantity under the wideband oscillation mode is used as the multi-dimensional instability leading factor detection result.

[0154] The steady-state characteristic quantity of the power grid under different wideband oscillation modes satisfies the following relationship:

[0155]

[0156] In the formula, m is the wideband measurement data in the detection data set, and t is the steady-state characteristic quantity of the power grid under different wideband oscillation modes.

[0157] In the embodiment, the steady-state characteristic quantity of the power grid under different wideband oscillation modes includes but is not limited to the voltage amplitude and phase under each wideband oscillation mode obtained by wideband measurement, the closed-loop control features of the grid-connected converter output current, and the historical data of the power grid voltage; when the change of the characteristic quantity under any wideband oscillation mode is greater than the static voltage stability margin of the power grid, it is determined that the power grid has the wideband oscillation mode instability, and the characteristic quantity under the wideband oscillation mode is used as the multi-dimensional instability leading factor detection result.

[0158] The method provided by the application can adapt to a high-proportion new energy power system, can more accurately and quickly balance electric energy, and the multi-dimensional support control decision technology considers more comprehensively when electric energy is dispatched, so that the voltage of each node can be more stable, and the stability of the system is effectively improved.

[0159] The application further provides a multi-dimensional instability leading factor detection system based on a cyclic DFT method, which comprises:

[0160] The data processing module is configured to obtain wideband measurement data by using a sliding data window, determine wideband oscillation mode phasors of the wideband measurement data in the sliding data window at a current time by using a cyclic discrete Fourier transform method, predict wideband oscillation mode phasors in the sliding data window at a next time based on the obtained wideband oscillation mode phasors in the sliding data window at the current time, and record the wideband measurement data in the sliding data window at the current time as a detection data set if it is determined that wideband oscillation occurs according to the predicted value of the wideband oscillation mode phasors in the sliding data window at the next time.

[0161] The index establishing module is configured to determine static voltage stability indexes of weak nodes and new energy grid-connected points in the power distribution network according to a power flow calculation result of the power distribution network system, take a minimum value in the static voltage stability indexes as a static voltage stability margin of the power distribution network, draw curves of active power and reactive current components output by an AC side of the grid-connected converter and curves of the active power and a converter power angle, and extract extreme points on the curves of the active power and the reactive current components as static limit currents and extract an extreme point on the curve of the active power and the converter power angle as a static limit power angle.

[0162] The instability dominant factor detection module is configured to establish a coupling model of multi-dimensional instability dominant factors of the power distribution network based on a single-machine control mode, take voltage control state parameters of the power distribution network as optimization variables, and take the static limit currents and the static limit power angle as control variables, obtain a distribution function of an influence of instability factors on stability of the power distribution network based on the coupling model of the multi-dimensional instability dominant factors of the power distribution network, establish an adaptive optimization model of power distribution network partition scheduling by using the distribution function of the influence of the instability factors on the stability of the power distribution network and a closed-loop control characteristic set of output currents of the grid-connected converters, obtain a historical data distribution function of the voltage of the power distribution network based on the adaptive optimization model of the power distribution network partition scheduling, take the historical data distribution function of the voltage of the power distribution network and the voltage control state parameters of the power distribution network in each scheduling partition, correct the wideband measurement data in the detection data set, and obtain steady-state characteristic quantities of the power distribution network under different wideband oscillation modes, and determine that the power distribution network is unstable under a wideband oscillation mode if a variation of a steady-state characteristic quantity under the wideband oscillation mode is greater than the static voltage stability margin of the power distribution network, and take the steady-state characteristic quantity under the wideband oscillation mode as a detection result of the multi-dimensional instability dominant factor.

[0163] A model of the cyclic discrete Fourier transform method satisfies the following relationship:

[0164]

[0165] In the formula, x(m) is the mth frequency domain signal, x(n) is the nth time domain signal, and N is the number of time domain discrete signals. Therefore, the input of the discrete Fourier transform is N time domain discrete signals, and the output is N frequency domain discrete point signals.

[0166] If it is determined that wide frequency oscillation occurs according to the predicted value of the wide frequency oscillation mode phasor in the sliding data window at the next moment, the wide frequency measurement data in the sliding data window at the current moment is recorded as a detection data set, and the sliding data window is continuously moved to collect sample data, and the wide frequency measurement data in the sliding data window at the current moment is recorded as a detection data set;

[0167] If it is determined that wide frequency oscillation does not occur according to the predicted value of the wide frequency oscillation mode phasor in the sliding data window at the next moment, the sliding data window is continuously moved to collect sample data.

[0168] The new energy grid connection point includes a PQ type node and a PV type node; wherein the active power and the reactive power of the PQ type node are opposite to the active power and the reactive power of the load, and the active power and the voltage amplitude of the PV type node are constant values;

[0169] According to the power flow calculation result of the distribution network system, the static voltage stability index of the weak node, the PQ type node and the PV type node is obtained by the following relationship:

[0170]

[0171] In the formula, L j is the static voltage stability index of node j, node j includes a weak node, a PQ type node and a PV type node, Z ij , R ij and X ij are the impedance, resistance and reactance of the line between node i and node j, respectively, and satisfy S j , P j and Q j are the apparent power, active power and reactive power of node j, respectively, and satisfy U i is the voltage amplitude of node i.

[0172] The minimum value in the static voltage stability index of each node is taken as the static voltage stability margin of the distribution network, and the following relationship is satisfied:

[0173]

[0174] In the formula, is the static voltage stability margin of the distribution network.

[0175] Draw the curves of active power and reactive current component of the grid-connected converter AC side output, and the curves of the converter power angle, including:

[0176] Based on the fundamental frequency equivalent circuit of the grid-connected converter AC side in the dq axis rotating coordinate, the active power of the grid-connected converter AC side output satisfies the following relationship:

[0177] P out =1.5(V od I Q +V oq I P )

[0178] In the formula, P out is the active power of the grid-connected converter AC side output, V od and V oq are the d-axis component and q-axis component of the grid point voltage respectively, I Q is the reactive current component, and I P is the active current component;

[0179] When the output power of the grid-connected converter is maximum, the active power of the grid-connected converter AC side output and the converter power angle δ satisfy the following relationship:

[0180]

[0181] In the formula, V g is the grid voltage amplitude, x g is the equivalent reactance between the grid point and the grid, and δ is the converter power angle;

[0182] Based on the above formula, draw the curves of active power and reactive current component of the grid-connected converter AC side output P out -I Q , and the curves of active power and converter power angle P out -δ.

[0183] The coupling model of the multi-dimensional instability dominant factor of the distribution network satisfies the following relationship:

[0184]

[0185] In the formula, M is the coupling function of the multi-dimensional instability dominant factor of the distribution network; b i is the static limit current on the i-th transmission line; n is the number of transmission lines; s(e) is the static limit power angle, and C is the voltage control state parameter of the distribution network.

[0186] The distribution function of the influence of the instability factor on the stability of the distribution network satisfies the following relationship:

[0187] k l =M+(bi + s (e)

[0188] In the formula, k l is a distribution function of the influence of the instability factor on the stability of the power distribution network.

[0189] The adaptive optimization model of the power distribution network partition scheduling satisfies the following relationship:

[0190]

[0191] In the formula, r(w) is an adaptive optimization function of the partition scheduling, and Φ is a closed-loop control feature set of the output current of the grid-connected converter.

[0192] The historical data distribution function of the power distribution network voltage satisfies the following relationship:

[0193]

[0194] In the formula, F is a historical data distribution function of the power distribution network voltage, is an optimal solution of the partition scheduling model, corresponding to the scheduling instruction of the scheduling partition w, and Φ w is a closed-loop control feature set of the output current of the grid-connected converter in the scheduling partition w, is a power distribution network voltage control state parameter of the scheduling partition w, and W is the total number of scheduling partitions.

[0195] The steady-state characteristic quantity of the power distribution network under different wide-frequency oscillation modes satisfies the following relationship:

[0196]

[0197] In the formula, m is wide-frequency measurement data in the detection data set, and t is a steady-state characteristic quantity of the power distribution network under different wide-frequency oscillation modes.

[0198] The steady-state characteristic quantity of the power distribution network under different wide-frequency oscillation modes includes but is not limited to the voltage amplitude and phase under each wide-frequency oscillation mode obtained by wide-frequency measurement, the closed-loop control features of the output current of the grid-connected converter, and the historical data of the power distribution network voltage.

[0199] The present disclosure can be a system, a method, and / or a computer program product. The computer program product can include a computer readable storage medium having computer readable program instructions loaded thereon, which instructions, when executed by a processor, cause the processor to implement various aspects of the present disclosure.

[0200] Computer readable storage media can be tangible storage media which can retain and store instructions for use by an instruction execution device. Computer readable storage media can be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer readable storage media include the following: a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanically encoded device such as punch-cards or raised structures in a groove having instructions recorded thereon, and any suitable combination of the foregoing. A computer readable storage medium, as used herein, is not to be construed as being transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission media (e.g., light pulses passing through a fiber-optic cable), or electrical signals transmitted through a wire.

[0201] Computer readable program instructions described herein can be downloaded to respective computing / processing devices from a computer readable storage medium or to an external computer or external storage device via a network, for example, the Internet, a local area network, a wide area network and / or a wireless network. The network can comprise copper transmission cables, optical transmission fibers, wireless transmission, routers, firewalls, switches, gateway computers and / or edge servers. A network adapter card or network interface in each computing / processing device receives computer readable program instructions from the network and forwards the computer readable program instructions for storage in a computer readable storage medium within the respective computing / processing device.

[0202] Computer readable program instructions for carrying out operations of the present disclosure can be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state-setting data, or any combination of source code or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C++ or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The computer readable program instructions can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate arrays (FPGA), or programmable logic arrays (PLA) can execute the computer readable program instructions by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present disclosure.

[0203] Finally, it should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, but not to limit it. Although the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or replaced, and any modification or replacement without departing from the spirit and scope of the present application should be covered in the protection scope of the claims of the present application.

Claims

1. A method for detecting multidimensional instability dominant factors based on cyclic DFT, characterized in that, include: Wideband measurement data is acquired using a sliding data window, and the wideband oscillation mode phasor of the wideband measurement data within the sliding data window at the current moment is determined using the cyclic discrete Fourier transform method. Based on the broadband oscillation mode phasors obtained in the current sliding data window, predict the broadband oscillation mode phasors in the next sliding data window; If broadband oscillation is determined to have occurred based on the predicted phasor values ​​of the broadband oscillation modes in the sliding data window at the next time step, then the broadband measurement data in the sliding data window at the current time step is recorded as the detection dataset. Based on the power flow calculation results of the distribution network system, the static voltage stability index of weak nodes and new energy grid connection points in the distribution network is determined, and the minimum value among the static voltage stability indices is taken as the static voltage stability margin of the distribution network. Plot the curves of active power and reactive current components output from the AC side of the grid-connected converter, and the curve of the converter's power angle, respectively; extract the extreme points on the curves of active power and reactive current components as static limiting currents, and extract the extreme points on the curves of active power and converter's power angle as static limiting power angles. Based on the single-machine control mode, a coupled model of the dominant factors of multidimensional instability in the distribution network is established, using the distribution network voltage control state parameters as optimization variables and the static limit current and static limit power angle as control variables. Based on this coupled model, the distribution function of the impact of instability factors on distribution network stability is obtained. Using the distribution function of the impact of instability factors on distribution network stability and the closed-loop control feature set of the grid-connected converter output current, an adaptive optimization model for distribution network zone scheduling is established. Based on this adaptive optimization model, using the closed-loop control feature set of the grid-connected converter output current in each scheduling zone and the distribution network voltage control state parameters of each scheduling zone, the historical data distribution function of the distribution network voltage is obtained. By using the historical data distribution function of the distribution network voltage and the distribution network voltage control state parameters of each dispatching zone, the broadband measurement data in the detection dataset is corrected to obtain the steady-state characteristic quantities of the distribution network under different broadband oscillation modes. When the change of the steady-state characteristic quantity under any broadband oscillation mode is greater than the static voltage stability margin of the distribution network, it is determined that the distribution network has instability in that broadband oscillation mode. The steady-state characteristic quantity under that broadband oscillation mode is used as the detection result of the dominant factor of multidimensional instability.

2. The method for detecting multidimensional instability dominant factors based on the cyclic DFT method according to claim 1, characterized in that, The model of the cyclic discrete Fourier transform method satisfies the following relationship: In the formula, x(m) is the m-th frequency domain signal, x(n) is the n-th time domain signal, and N is the number of discrete time domain signals; therefore, the input of the discrete Fourier transform is N discrete time domain signals, and the output is N discrete frequency domain signals.

3. The method for detecting multidimensional instability dominant factors based on the cyclic DFT method according to claim 2, characterized in that, If broadband oscillation is determined to occur based on the predicted value of the broadband oscillation mode phasor in the sliding data window at the next time step, the broadband measurement data in the sliding data window at the current time step is recorded as the detection dataset, and the sliding data window is moved to collect sample data. At the same time, the broadband measurement data in the sliding data window at the current time step is recorded as the detection dataset. If it is determined that broadband oscillation will not occur based on the predicted phasor value of the broadband oscillation mode within the sliding data window at the next time step, then the sliding data window continues to move to collect sample data.

4. The method for detecting multidimensional instability dominant factors based on the cyclic DFT method according to claim 1, characterized in that, New energy grid connection points include PQ type nodes and PV type nodes; among them, the active power and reactive power of PQ type nodes are opposite to the active power and reactive power of the load, while the active power and voltage amplitude of PV type nodes are constant values. Based on the power flow calculation results of the distribution network system, the static voltage stability indices of weak nodes, PQ-type nodes, and PV-type nodes are obtained using the following formulas: In the formula, L j Z represents the static voltage stability index of node j, which includes weak nodes, PQ-type nodes, and PV-type nodes. ij R ij and X ij Let be the impedance, resistance, and reactance of the line between node i and node j, respectively, satisfying... S j P j and Q j Let be the apparent power, active power, and reactive power of node j, respectively, satisfying . U i Let be the voltage amplitude at node i.

5. The method for detecting multidimensional instability dominant factors based on the cyclic DFT method according to claim 4, characterized in that, The minimum value among the static voltage stability indices of each node is taken as the static voltage stability margin of the distribution network, satisfying the following relationship: In the formula, This refers to the static voltage stability margin of the distribution network.

6. The method for detecting multidimensional instability dominant factors based on the cyclic DFT method according to claim 1, characterized in that, Plot the curves of active power and reactive current components output from the AC side of the grid-connected converter, and the curve of the converter's power angle, including: Based on the fundamental frequency equivalent circuit of the AC side of the grid-connected converter in the dq-axis rotating coordinate system, the active power output of the AC side of the grid-connected converter satisfies the following relationship: P out =1.5(V od I Q +V oq I P ) In the formula, P out V represents the active power output from the AC side of the grid-connected converter. od V oq These are the d-axis and q-axis components of the grid connection point voltage, respectively. Q For the reactive current component, I P This refers to the active current component; When the grid-connected converter has maximum output power, the active power output on the AC side of the grid-connected converter and the converter power angle δ satisfy the following relationship: In the formula, V g x represents the grid voltage amplitude. g δ is the equivalent reactance between the grid connection point and the power grid, and δ is the converter power angle. Based on the above formula, plot the curves P of the active power and reactive current components of the AC side output of the grid-connected converter. out -I Q The curve P of active power versus converter power angle out -δ.

7. The method for detecting multidimensional instability dominant factors based on the cyclic DFT method according to claim 1, characterized in that, The coupling model of the dominant factors of multidimensional instability in distribution networks satisfies the following relationship: In the formula, M is the coupling function of the dominant factors of multidimensional instability in the distribution network; b i Let be the static limiting current on the i-th transmission line; n be the number of transmission lines; s(e) be the static limiting power angle; and C be the voltage control state parameter of the distribution network.

8. The method for detecting multidimensional instability dominant factors based on the cyclic DFT method according to claim 7, characterized in that, The distribution function of the impact of instability factors on the stability of the distribution network satisfies the following relationship: k l =M+(b i +s(e)) In the formula, k l This is the distribution function of the impact of instability factors on the stability of the distribution network.

9. The method for detecting multidimensional instability dominant factors based on the cyclic DFT method according to claim 8, characterized in that, The adaptive optimization model for distribution network zone scheduling satisfies the following relationship: In the formula, r(w) is the adaptive optimization function of partitioned scheduling, and Φ is the closed-loop control characteristic set of the grid-connected converter output current.

10. The method for detecting multidimensional instability dominant factors based on the cyclic DFT method according to claim 9, characterized in that, The historical voltage distribution function of the distribution network satisfies the following relationship: In the formula, F is the historical data distribution function of the distribution network voltage. For the optimal solution of the partitioned scheduling model, the scheduling instruction corresponding to partition w is Φ w This represents the set of closed-loop control characteristics of the output current of the grid-connected converter within the scheduling partition w. Here are the distribution network voltage control status parameters for dispatch zone w, where W is the total number of dispatch zones.

11. The method for detecting multidimensional instability dominant factors based on the cyclic DFT method according to claim 10, characterized in that, The steady-state characteristic quantities of a power distribution network under different broadband oscillation modes satisfy the following relationship: In the formula, m represents the broadband measurement data in the detection dataset, and t represents the steady-state characteristic quantity of the distribution network under different broadband oscillation modes.

12. The method for detecting multidimensional instability dominant factors based on the cyclic DFT method according to claim 11, characterized in that, The steady-state characteristics of the distribution network under different broadband oscillation modes include, but are not limited to, the voltage amplitude and phase under each broadband oscillation mode obtained by broadband measurement, the closed-loop control characteristics of the output current of the grid-connected converter, and historical data of the distribution network voltage.

13. A multidimensional instability dominance factor detection system based on the cyclic DFT method, used to implement the method described in any one of claims 1 to 12, characterized in that, include: The data processing module is used to acquire broadband measurement data using a sliding data window, determine the broadband oscillation mode phasor of the broadband measurement data within the sliding data window at the current time using the cyclic discrete Fourier transform method, and predict the broadband oscillation mode phasor of the sliding data window at the next time based on the obtained broadband oscillation mode phasor within the sliding data window at the current time. If broadband oscillation is determined to have occurred based on the predicted phasor values ​​of the broadband oscillation modes in the sliding data window at the next time step, then the broadband measurement data in the sliding data window at the current time step is recorded as the detection dataset. The index establishment module is used to determine the static voltage stability index of weak nodes and new energy grid connection points in the distribution network based on the power flow calculation results of the distribution network system, and to use the minimum value among the static voltage stability indices as the static voltage stability margin of the distribution network. Plot the curves of active power and reactive current components output from the AC side of the grid-connected converter, and the curve of the converter's power angle, respectively; extract the extreme points on the curves of active power and reactive current components as static limiting currents, and extract the extreme points on the curves of active power and converter's power angle as static limiting power angles. The instability-dominant factor detection module is used to establish a coupled model of multidimensional instability-dominant factors in the distribution network based on a single-machine control mode, using the distribution network voltage control state parameters as optimization variables and the static limiting current and static limiting power angle as control variables. Based on this coupled model, the distribution function of the impact of instability factors on distribution network stability is obtained. Using the distribution function of the impact of instability factors on distribution network stability and the closed-loop control characteristic set of the grid-connected converter output current, an adaptive optimization model for distribution network zone scheduling is established. Based on this adaptive optimization model, the grid-connected converters within each scheduling zone... The historical data distribution function of the distribution network voltage is obtained by combining the closed-loop control feature set of the output current and the distribution network voltage control state parameters of each dispatching zone. The broadband measurement data in the detection dataset is corrected using the historical data distribution function of the distribution network voltage and the distribution network voltage control state parameters of each dispatching zone to obtain the steady-state characteristic quantity of the distribution network under different broadband oscillation modes. When the change of the steady-state characteristic quantity under any broadband oscillation mode is greater than the static voltage stability margin of the distribution network, it is determined that the distribution network has instability in that broadband oscillation mode. The steady-state characteristic quantity under that broadband oscillation mode is used as the detection result of the dominant factor of multidimensional instability.

14. The multidimensional instability dominance factor detection system based on the cyclic DFT method according to claim 13, characterized in that, The model of the cyclic discrete Fourier transform method satisfies the following relationship: In the formula, x(m) is the m-th frequency domain signal, x(n) is the n-th time domain signal, and N is the number of discrete time domain signals; therefore, the input of the discrete Fourier transform is N discrete time domain signals, and the output is N discrete frequency domain signals.

15. The multidimensional instability dominance factor detection system based on the cyclic DFT method according to claim 14, characterized in that, If broadband oscillation is determined to occur based on the predicted value of the broadband oscillation mode phasor in the sliding data window at the next time step, the broadband measurement data in the sliding data window at the current time step is recorded as the detection dataset, and the sliding data window is moved to collect sample data. At the same time, the broadband measurement data in the sliding data window at the current time step is recorded as the detection dataset. If it is determined that broadband oscillation will not occur based on the predicted phasor value of the broadband oscillation mode within the sliding data window at the next time step, then the sliding data window continues to move to collect sample data.

16. The multidimensional instability dominance factor detection system based on the cyclic DFT method according to claim 13, characterized in that, New energy grid connection points include PQ type nodes and PV type nodes; among them, the active power and reactive power of PQ type nodes are opposite to the active power and reactive power of the load, while the active power and voltage amplitude of PV type nodes are constant values. Based on the power flow calculation results of the distribution network system, the static voltage stability indices of weak nodes, PQ-type nodes, and PV-type nodes are obtained using the following formulas: In the formula, L j Z represents the static voltage stability index of node j, which includes weak nodes, PQ-type nodes, and PV-type nodes. ij R ij and X ij Let be the impedance, resistance, and reactance of the line between node i and node j, respectively, satisfying... S j P j and Q j Let be the apparent power, active power, and reactive power of node j, respectively, satisfying . U i Let be the voltage amplitude at node i.

17. The multidimensional instability dominance factor detection system based on the cyclic DFT method according to claim 16, characterized in that, The minimum value among the static voltage stability indices of each node is taken as the static voltage stability margin of the distribution network, satisfying the following relationship: In the formula, This refers to the static voltage stability margin of the distribution network.

18. The multidimensional instability dominance factor detection system based on the cyclic DFT method according to claim 13, characterized in that, Plot the curves of active power and reactive current components output from the AC side of the grid-connected converter, and the curve of the converter's power angle, including: Based on the fundamental frequency equivalent circuit of the AC side of the grid-connected converter in the dq-axis rotating coordinate system, the active power output of the AC side of the grid-connected converter satisfies the following relationship: P out =1.5(V od I Q +V oq I P ) In the formula, P out V represents the active power output from the AC side of the grid-connected converter. od V oq These are the d-axis and q-axis components of the grid connection point voltage, respectively. Q For the reactive current component, I P This refers to the active current component; When the grid-connected converter has maximum output power, the active power output on the AC side of the grid-connected converter and the converter power angle δ satisfy the following relationship: In the formula, V g x represents the grid voltage amplitude. g δ is the equivalent reactance between the grid connection point and the power grid, and δ is the converter power angle. Based on the above formula, plot the curves P of the active power and reactive current components of the AC side output of the grid-connected converter. out -I Q The curve P of active power versus converter power angle out -δ.

19. The multidimensional instability dominance factor detection system based on the cyclic DFT method according to claim 13, characterized in that, The coupling model of the dominant factors of multidimensional instability in distribution networks satisfies the following relationship: In the formula, M is the coupling function of the dominant factors of multidimensional instability in the distribution network; b i Let be the static limiting current on the i-th transmission line; n be the number of transmission lines; s(e) be the static limiting power angle; and C be the voltage control state parameter of the distribution network.

20. The multidimensional instability dominance factor detection system based on the cyclic DFT method according to claim 19, characterized in that, The distribution function of the impact of instability factors on the stability of the distribution network satisfies the following relationship: k l =M+(b i +s(e)) In the formula, k l This is the distribution function of the impact of instability factors on the stability of the distribution network.

21. The multidimensional instability dominance factor detection system based on the cyclic DFT method according to claim 20, characterized in that, The adaptive optimization model for distribution network zone scheduling satisfies the following relationship: In the formula, r(w) is the adaptive optimization function of partitioned scheduling, and Φ is the closed-loop control characteristic set of the grid-connected converter output current.

22. The multidimensional instability dominance factor detection system based on the cyclic DFT method according to claim 21, characterized in that, The historical voltage distribution function of the distribution network satisfies the following relationship: In the formula, F is the historical data distribution function of the distribution network voltage. For the optimal solution of the partitioned scheduling model, the scheduling instruction corresponding to partition w is Φ w This represents the set of closed-loop control characteristics of the output current of the grid-connected converter within the scheduling partition w. Here are the distribution network voltage control status parameters for dispatch zone w, where W is the total number of dispatch zones.

23. The multidimensional instability dominance factor detection system based on the cyclic DFT method according to claim 22, characterized in that, The steady-state characteristic quantities of a power distribution network under different broadband oscillation modes satisfy the following relationship: In the formula, m represents the broadband measurement data in the detection dataset, and t represents the steady-state characteristic quantity of the distribution network under different broadband oscillation modes.

24. The multidimensional instability dominance factor detection system based on the cyclic DFT method according to claim 23, characterized in that, The steady-state characteristics of the distribution network under different broadband oscillation modes include, but are not limited to, the voltage amplitude and phase under each broadband oscillation mode obtained by broadband measurement, the closed-loop control characteristics of the output current of the grid-connected converter, and historical data of the distribution network voltage.

25. A terminal, comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1-12.

26. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1-12.

Images