An oscillation instability propagation analysis system for a high-proportion new energy grid-connected system

Abstract

The application discloses an oscillation instability propagation analysis system for a high-proportion new energy grid-connected system, and relates to the technical field of power system safety and stability control. The system adopts a hierarchical distributed architecture and is divided into five levels, namely, a perception and collection layer, a network transmission layer, a data storage layer, a core analysis layer and an application display layer. The levels are cooperatively matched to realize whole-process closed-loop management and control from data collection to analysis application. The overall architecture has the characteristics of high compatibility, strong expansibility and high real-time performance and can be adapted to high-proportion new energy grid-connected systems of different scales and topologies. In combination with the generation mechanism and propagation characteristics of wideband oscillation of double-high power systems, the system integrates advanced technologies such as adaptive signal processing, Copula transfer entropy nonlinear causal analysis, multi-source data fusion and real-time online monitoring, breaks through the limitations of traditional models, realizes wideband oscillation full-dimension identification and accurate tracing of a propagation path, and fills the technical gap in the field of wideband oscillation analysis of high-proportion new energy grid-connected systems.

Description

Technical Field

[0001] This invention relates to the field of power system safety and stability control technology, and in particular to an oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems. Background Technology

[0002] For new power system technologies, such as virtual synchronous machine control, flexible DC transmission, grid-connected converters, and electrochemical energy storage, the existing systems have not been adapted to address the new oscillation characteristics brought about by these technologies. The following shortcomings exist: they cannot identify the oscillation excitation sources and propagation paths caused by the new equipment; they lack a multi-dimensional, full-scenario verification system, with existing technologies relying primarily on simulation analysis, lacking dual verification from real-time digital dynamic model experiments and actual grid measurement data; experimental data and waveform graphs are insufficient, and the reliability and engineering applicability of the analysis results are inadequate.

[0003] In addition, most existing systems perform offline analysis, which has poor real-time performance and cannot achieve online monitoring of oscillations, real-time path tracking and early warning, making it difficult to meet the needs of real-time power grid dispatch and emergency control. Summary of the Invention

[0004] This invention provides an oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems, which overcomes the limitations of traditional models and achieves full-dimensional identification of broadband oscillations and accurate tracing of propagation paths.

[0005] On one hand, this invention provides an oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems. The system adopts a layered distributed architecture, comprising: a sensing and acquisition module, a network transmission module, a data storage module, a core analysis module, and an application display module. The sensing and acquisition module synchronously collects multi-dimensional electrical quantities from the grid-connected system. The network transmission module transmits the data collected by the sensing and acquisition module to the data storage module. The data storage module uses a hybrid storage mode combining real-time and historical databases to classify and store the received data. The core analysis module, based on the data in the data storage module, performs modal parameter identification, propagation path analysis, and oscillation source localization for oscillation instability by integrating adaptive signal processing and Copula transfer entropy nonlinear causal analysis. The application display module provides multi-dimensional visualization and early warning of the analysis results from the core analysis module.

[0006] In one implementation of the present invention, the sensing and acquisition module integrates multiple synchronous acquisition devices, including a power grid synchronous phasor measurement device, a fault recording device, a new energy power station grid connection point monitoring and control terminal, a converter operation status monitoring device, a flexible DC transmission system monitoring terminal, and an electrochemical energy storage grid connection point monitoring device; the multi-dimensional electrical quantities include bus voltage, line current, active power, reactive power, frequency, converter control signal, and energy storage charging and discharging power; the sensing and acquisition module adopts BeiDou / GPS dual-mode timing.

[0007] In one implementation of the present invention, the network transmission module is compatible with IEC61850, IEC60870-5-104 and MQTT communication protocols, constructs a transmission channel combining fiber optic private network and wireless private network, controls the transmission delay within a preset time threshold, and has data encryption and packet loss retransmission functions.

[0008] In one implementation of the present invention, the data storage module includes a real-time database for storing millisecond-level real-time monitoring data to meet online analysis requirements; a historical database employing a distributed storage architecture for storing full-volume time-domain waveforms, frequency-domain spectra, oscillation mode parameters, propagation path data, experimental data, and field-measured data; and a data quality verification mechanism for eliminating bad and redundant data.

[0009] In one implementation of the present invention, the core analysis module specifically includes: The signal preprocessing submodule is used to perform noise reduction processing on the original signal using the AVMD-SVD joint noise reduction algorithm; The wideband oscillation mechanism analysis submodule is used to analyze the oscillation generation mechanism; The modal parameter identification submodule is used to extract oscillation mode parameters from the denoised signal; The oscillation propagation path analysis submodule is used to construct a directed weighted electrical network based on the Copula transfer entropy algorithm to trace the oscillation propagation path; The oscillation source positioning submodule is used to locate the oscillation excitation source; The propagation vulnerability assessment submodule is used to assess critical nodes and vulnerable branches in oscillation propagation networks.

[0010] In one implementation of the present invention, the modal parameter identification submodule employs a joint identification algorithm combining the Prony algorithm, the ESPRIT algorithm, and wavelet transform to achieve accurate extraction of oscillation modes across the entire frequency band. Specifically, the Prony algorithm is used for preliminary identification of the frequency, amplitude, damping ratio, and phase of stationary oscillation signals; the ESPRIT algorithm is used for identification of harmonic and oscillatory coupled signals; and the wavelet transform is used for joint time-frequency domain analysis to locate the oscillation occurrence time and frequency distribution.

[0011] In one implementation of the present invention, the oscillation propagation path analysis submodule quantifies the nonlinear, non-Gaussian oscillation propagation causal relationship through the Copula transfer entropy algorithm, outputs the oscillation propagation path, propagation delay and causal strength, and dynamically identifies the main propagation branch and key nodes.

[0012] In one implementation of the present invention, the oscillation source location submodule adopts a multi-dimensional oscillation source location algorithm, which, based on the principle of energy function or impedance identification and combined with the time-frequency domain characteristics of electrical quantities of each node, accurately locates the equipment or site where the oscillation excitation source is located.

[0013] In one implementation of the present invention, the system further comprises a three-level verification system, which includes: network topology structure demonstration mechanism analysis, MATLAB / Simulink simulation verification, and RT-LAB real-time digital dynamic model experimental verification, for comprehensively verifying the accuracy and practicality of the system.

[0014] In one implementation of the present invention, the application display module automatically generates early warnings and decision suggestions based on the analysis results, providing support for broadband oscillation suppression and system stability control.

[0015] The present invention provides an oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems, which has the following advantages: 1. This invention adopts a combination of data-driven and mechanism analysis, and integrates the joint identification algorithm of Prony, ESPRIT and wavelet transform, covering the entire frequency band from 0.1Hz to 500Hz. It achieves accurate identification of subsynchronous, supersynchronous and mid-to-high frequency multimodal broadband oscillations of high-proportion new energy grid-connected systems, breaking through the limitations of single-mode analysis in traditional technology.

[0016] 2. This invention introduces a transfer entropy algorithm optimized by the Copula function to quantify the causal relationship of nonlinear and non-Gaussian oscillation propagation, constructs a directed weighted electrical network, and realizes accurate tracking of oscillation propagation path, propagation delay and causal strength. It can dynamically identify the main propagation branch and key nodes, with an analysis accuracy of over 95%.

[0017] 3. This invention constructs a three-level verification system: theoretical proof of network topology machines, MATLAB / Simulink simulation verification, and RT-LAB real-time digital dynamic model experiment. It is equipped with complete experimental data and waveform graphs, which realizes multi-dimensional and full-scenario cross-verification of analysis results, significantly improving the system's credibility and engineering applicability.

[0018] 4. By adopting a hierarchical distributed architecture and a full-process online operation mechanism, this invention integrates high-precision synchronous acquisition, low-latency transmission, and real-time analysis and early warning functions, realizing closed-loop management of the entire process from data acquisition to decision-making recommendations, and meeting the needs of real-time power grid dispatching and emergency control. Attached Figure Description

[0019] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the overall system architecture provided in an embodiment of the present invention; Figure 2 This is a flowchart of the oscillation instability propagation analysis provided in an embodiment of the present invention; Figure 3 A schematic diagram of a new energy system control model provided in an embodiment of the present invention; Figure 4 A schematic diagram of a four-station, two-area new energy power generation grid connection system and oscillation propagation provided in an embodiment of the present invention; Figure 5 The amplitude-time characteristic diagrams of each oscillation mode component provided in the embodiments of the present invention; Figure 6 Frequency-time characteristic diagrams of each oscillation mode component provided in the embodiments of the present invention; Figure 7 A schematic diagram of the entropy transfer of a four-station, two-region new energy power generation grid connection system provided in an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0021] This invention, combining the generation mechanism and propagation characteristics of broadband oscillations in high-voltage power systems, integrates cutting-edge technologies such as adaptive signal processing, Copula transfer entropy nonlinear causal analysis, multi-source data fusion, and real-time online monitoring. It proposes an oscillation propagation path analysis system that combines data-driven approaches with mechanism analysis, overcoming the limitations of traditional model-dependent systems. This system achieves full-dimensional identification of broadband oscillations and accurate tracing of their propagation paths, filling a technological gap in the analysis of broadband oscillations in high-proportion renewable energy grid-connected systems. The technical solution proposed in this invention will be described in detail below with reference to the accompanying drawings.

[0022] Figure 1 This is a system composition diagram for oscillation instability propagation analysis of a high-proportion renewable energy grid-connected system, provided as an embodiment of the present invention. (See diagram below.) Figure 1 As shown, the system mainly consists of five levels: a sensing and acquisition module, a network transmission module, a data storage module, a core analysis module, and an application display module. Each level works together to achieve closed-loop control of the entire process from data acquisition to analysis and application. The overall architecture has the characteristics of high compatibility, strong scalability, and high real-time performance, and can be adapted to high-proportion new energy grid-connected systems of different scales and topologies.

[0023] Sensing and Acquisition Module: As the data source of the system, it integrates multiple types of synchronous acquisition devices, including power grid synchronous phasor measurement devices (PMU / WAMS), fault recording devices, new energy power plant grid connection point monitoring and control terminals, converter operation status monitoring devices, flexible DC transmission system monitoring terminals, and electrochemical energy storage grid connection point monitoring equipment, to realize the synchronous acquisition of multiple dimensions of electrical quantities such as bus voltage, line current, active / reactive power, frequency, converter control signals, and energy storage charging and discharging power.

[0024] Network transmission module: Adopting a dedicated communication protocol for power systems, compatible with multiple communication protocols, and constructing a transmission channel combining fiber optic private network and wireless private network, it achieves low-latency and high-reliability transmission of sensing module data, with transmission latency controlled within 50ms, ensuring the security and integrity of data transmission.

[0025] Data storage module: It adopts a hybrid storage mode that combines real-time database and historical database. The real-time database stores millisecond-level real-time monitoring data to meet the needs of online analysis; the historical database stores full-volume time-domain waveforms, frequency-domain spectra, oscillation mode parameters, propagation path data, experimental data, and field measurement data. It adopts a distributed storage architecture to support long-term storage and fast query of massive data, and at the same time establishes a data quality verification mechanism to remove bad data and redundant data.

[0026] Core Analysis Module: As the core functional module of the system, it integrates signal preprocessing submodule, broadband oscillation mechanism analysis submodule, modal parameter identification submodule, oscillation propagation path analysis submodule, oscillation source localization submodule, and propagation vulnerability assessment submodule. It deeply integrates core algorithms such as AVMD-SVD joint noise reduction, Prony-ESPRIT-Wavelet joint modal identification, Copula transfer entropy nonlinear causal analysis, and multi-dimensional oscillation source localization to achieve intelligent analysis throughout the entire process from signal processing to path analysis.

[0027] To address the wideband multimodal coupled oscillation characteristics of high-voltage power systems, this invention employs a joint identification algorithm combining Prony, ESPRIT, and wavelet transform to accurately extract oscillation modes across the entire frequency band from 0.1Hz to 500Hz, overcoming the limitations of low accuracy and poor adaptability of single algorithms. Specifically, the Prony algorithm provides preliminary identification of oscillation frequency, amplitude, damping ratio, and phase, adapting to stationary oscillation signals; the ESPRIT algorithm improves the identification accuracy of harmonic and oscillation coupling signals, suitable for non-stationary signals; and wavelet transform enables joint time-frequency domain analysis, accurately locating the oscillation occurrence time and frequency distribution. Through the fusion and complementarity of these three algorithms, it can simultaneously identify subsynchronous oscillations (5Hz-45Hz), supersynchronous oscillations (55Hz-100Hz), and mid-frequency / high-frequency oscillations (100Hz-500Hz), outputting the frequency, amplitude, damping ratio, time-domain waveform, and frequency-domain spectrum of each mode, and determining the convergence, constant amplitude, or divergence characteristics of the oscillations. Traditional path analysis methods cannot quantify the causal relationships of nonlinear, non-Gaussian oscillation propagation. This invention introduces a transfer entropy algorithm optimized by the Copula function to construct a directed weighted electrical network, thereby achieving accurate tracking of oscillation propagation paths and quantification of causal strength.

[0028] This invention establishes a three-tiered verification system: network topology structure demonstration mechanism analysis, MATLAB / Simulink simulation verification, and RT-LAB real-time digital dynamic model experiment, comprehensively verifying the system's accuracy and practicality. The network topology structure demonstration mechanism analysis, based on the proportion of new energy power plants in the system, the types of new energy power plants, and the mathematical derivation of the LC oscillation induction mechanism, theoretically demonstrates the oscillation induction process of a new high-proportion new energy power system. Simulation verification constructs models of dual-high power systems with different new energy penetration rates, different short-circuit ratios, and different control strategies to simulate various broadband oscillation scenarios and acquire simulation waveforms and data. The real-time dynamic model experiment uses the RT-LAB hardware-in-the-loop simulation platform to build an experimental system combining physical dynamic model and digital simulation, simulating real power grid operating conditions, collecting experimental data and waveforms, and collecting oscillation event data to verify the system's application effect in practical engineering.

[0029] The figures listed in the system of this invention, Figure 2 This document details the flowchart of the system's method for analyzing the propagation of oscillation instability. The system method describes a complete, data-driven, broadband oscillation end-to-end analysis framework. For complex oscillation scenarios in high-proportion renewable energy power systems, it achieves closed-loop analysis from data acquisition to oscillation source location, vulnerability assessment, and early warning. The technical principles and core value of each step are as follows.

[0030] 1. Multi-source synchronous data acquisition This step is the foundation of the entire analysis system described in this invention. Its core objective is to acquire comprehensive measurement data of the power system. The data sources primarily include synchronously collected PMU (phasor measurement unit) and SCADA / WAMS data from renewable energy power plants (wind / solar), conventional generating units, transmission lines, and busbars, covering key electrical / mechanical quantities such as voltage, current, power, frequency, and rotational speed. A crucial requirement for this step is the time synchronization of multi-node, multi-type data (typically based on GPS / BeiDou timing) to ensure the temporal consistency of subsequent modal identification and transfer entropy calculations, avoiding analysis errors caused by time offsets. This step is adapted to the wide-band oscillation requirements (from subsynchronous and supersynchronous to high-frequency bands) of high-proportion renewable energy systems, covering measurement data from different frequency bands to support subsequent mechanism analysis.

[0031] 2. Bad data removal combined with AVMD-SVD noise reduction This step is the core of data preprocessing, addressing the issues of "dirty data" and noise interference in on-site measurement data. Bad data removal primarily relies on methods such as data consistency verification, Gaussian threshold filtering, and the 3σ principle to eliminate outliers and missing values ​​caused by equipment failures or communication interruptions, ensuring data reliability. Secondly, the AVMD-SVD joint noise reduction method mainly consists of two sub-modules: First, AVMD (Adaptive Variational Mode Decomposition) adaptively decomposes the noisy signal into multiple modal components, avoiding the limitations of traditional VMD which requires manually preset parameters, and accurately separating oscillation components and noise components in different frequency bands. The SVD (Singular Value Decomposition) module then performs singular value decomposition on the modal components decomposed by AVMD, filtering out singular values ​​corresponding to noise by setting thresholds, further suppressing white noise, power frequency interference, and other noise components. The advantage of designing AVMD-SVD is that it combines the multi-scale decomposition capability of AVMD with the strong noise suppression capability of SVD, significantly improving the signal-to-noise ratio while preserving oscillation characteristic information, providing high-quality input data for subsequent modal identification.

[0032] 3. Wideband Oscillation Mechanism Analysis Module Based on the denoised data, a qualitative analysis of the oscillation mechanism is conducted to clarify the type and cause of the oscillation: First, we need to classify the types of oscillations: distinguish between oscillation modes with different mechanisms, such as subsynchronous / supersynchronous oscillations caused by grid connection of new energy sources, high-frequency oscillations caused by inverter control parameter mismatch, and electromechanical coupling oscillations.

[0033] Then comes the mechanism analysis logic: combining the control characteristics of new energy equipment (such as the control loop parameters of wind turbine converters and photovoltaic inverters), grid topology and impedance characteristics, we can initially determine the "source type" of oscillation (equipment level / network level) to provide directional guidance for subsequent mode identification and source tracing.

[0034] 4. Prony-ESPRIT-Wavelet Joint Modality Identification This module is the core technology of the entire system, enabling accurate parameter identification of wideband oscillation modes: Wavelet transform performs multi-scale time-frequency analysis on signals, locating the time-frequency distribution of oscillations and initially separating oscillation components in different frequency bands, thus solving the problem of multi-mode aliasing in broadband signals. The Prony algorithm, based on a complex exponential model, fits signals and extracts key parameters such as frequency, amplitude, damping ratio, and phase of oscillation modes. It is suitable for processing decaying oscillation signals but is sensitive to noise, thus relying on prior noise reduction processing. ESPRIT (rotation-invariant signal parameter estimation), a parameter identification algorithm based on subspace methods, is more robust to noise and can complement the Prony algorithm, improving the stability and accuracy of modal parameter estimation. This module effectively combines the three methods, offering significant advantages. The combination of these three methods balances time-frequency localization, parameter accuracy, and noise robustness, and can simultaneously identify broadband oscillation characteristics across multiple frequency bands and modes, overcoming the limitations of a single algorithm.

[0035] 5. Extraction of dominant oscillation components This module, based on joint modal identification results, filters out the oscillation components that have the greatest impact on system safety. The selection of oscillation components is primarily based on a combination of the mode's damping ratio (the lower the damping ratio, the higher the oscillation risk), amplitude, and participation factor (reflecting the degree of participation of different devices / nodes in that mode), thereby determining the dominant oscillation component. Its core value lies in focusing on key risk modes from complex multimodal oscillations, avoiding interference from secondary modes in subsequent analysis, and significantly improving analysis efficiency and relevance.

[0036] 6. Copula function optimization for transit entropy calculation This step is crucial for oscillation propagation path analysis, addressing the problem that traditional correlation analysis cannot capture nonlinear causal relationships. Transfer entropy is an information theory-based causality measurement method that quantifies the information flow and intensity between two node signals, thereby determining the direction and strength of oscillation propagation. However, traditional transfer entropy calculations suffer from sensitivity to data distribution and insufficient capture of nonlinear relationships.

[0037] Copula functions describe the nonlinear dependency structure between variables. By estimating propagation entropy through Copula entropy, it eliminates the need to assume that the data follows a specific distribution, significantly improving the accuracy and robustness of propagation entropy calculation. The core function of this module is to calculate the "causal strength" between different nodes, providing a basis for weighting subsequent construction of directed weighted networks and quantifying the propagation path and impact of oscillations in the system.

[0038] 7. Construction of the Directed Weighted Network Main / Secondary Propagation Path Extraction Module Based on the transfer entropy calculation results, the power system is abstracted into a directed weighted complex network. From the directed weighted network, based on network flow analysis, the shortest path / maximum weight path algorithm, and the participation factor of oscillation modes, the main channels for oscillation to spread from the source to the whole system are located, and the key paths of oscillation propagation are identified: the main path: the path with the highest weight and the greatest transfer entropy intensity, that is, the main propagation channel of oscillation energy, which is the core propagation link of system oscillation risk; the secondary path: the path with the next highest weight, reflecting the secondary propagation channel of oscillation, which plays a supplementary or amplifying role to the main path.

[0039] 8. Location of oscillation source and vulnerability assessment of propagation branches This is the core objective of the methodology, enabling precise traceability from the network level to the site level and even the device level: Based on the centrality analysis of directed weighted networks, the bus / regional node where the oscillation source is located is determined, narrowing down the location range. Then, combined with equipment measurement data and participation factor analysis from renewable energy power plants, the power plant or unit group where the oscillation source is located is pinpointed. Through the identification of control parameters of key equipment and the matching analysis of impedance characteristics and oscillation modes, the specific equipment causing the oscillation (such as wind turbine converters, SVG controllers, etc.) is finally located. Then, by combining the branch's transfer entropy weight, power flow characteristics, and impedance characteristics, the "amplification / attenuation" effect of the branch on oscillation propagation and the impact of branch faults on oscillation propagation are evaluated, identifying the weak link in oscillation propagation, providing a basis for system hardening, and providing targeted targets for subsequent prevention and control measures. Finally, the accuracy of the analysis results is verified by visualization and early warning closed-loop, forming a traceable analysis record.

[0040] To comprehensively verify the accuracy and practicality of the method, a three-level verification system is constructed: "theoretical proof of topology machine - offline simulation verification - real-time dynamic model experiment". 1. Level 1: Theoretical Proof of Network Topology Based on the topological characteristics of the power system and combined with the electrical coupling mechanism of oscillation propagation, the rationality of the method is verified from a theoretical perspective. Based on the matching of system impedance topology, node participation factor and oscillation mode, the intrinsic relationship between oscillation propagation path and network structure is demonstrated. Combined with the grid-connected control characteristics of new energy equipment, the consistency between the oscillation source location results and the key nodes of the topology is verified, and the qualitative verification at the mechanism level is completed.

[0041] 2. Second level: MATLAB / Simulink offline simulation verification A power system simulation model with a high proportion of new energy sources was built, and the quantitative accuracy of the method was verified through a controllable scenario: typical broadband oscillation scenarios such as subsynchronous / supersynchronous / high frequency were reproduced, and standard simulation data were output; the method was applied to perform modal identification, propagation path analysis and oscillation source tracing, and the results were compared with the simulation preset scenario to verify the algorithm.

[0042] 3. Level Three: RT-LAB Real-Time Digital Dynamic Model Experiment Verification Based on the RT-LAB real-time simulation platform, a semi-physical dynamic model experiment was conducted to verify the applicability of the method in actual engineering scenarios: a closed-loop real-time simulation system was constructed by connecting real new energy controllers, protection devices and other physical equipment to simulate the dynamic response characteristics of the actual power grid; the method was integrated into a real-time monitoring system to verify the real-time performance and reliability of oscillation propagation analysis, source location and early warning functions under experimental conditions, and to complete the final verification before engineering application.

[0043] The three-level verification system realizes full-dimensional verification from theoretical mechanism to offline simulation and then to real-time dynamic model. It not only ensures the theoretical rigor of the method, but also verifies its accuracy and practicality in actual engineering through simulation and experiment, providing reliable technical support for the prevention and control of oscillation instability in high-proportion new energy power systems.

[0044] according to Figure 2 The aforementioned analysis process and the three-level analysis system constructed in this invention are intended to build a system with a high proportion of new energy sources, with new energy power stations as... Figure 3 The control system shown controls the photovoltaic system.

[0045] The core of this control system is to describe the photovoltaic array using equivalent circuits, thereby obtaining the characteristics of new energy power plants represented by photovoltaics. It enables calculations of output current, power, etc., and can be used for subsequent MPPT control, grid-connected simulation, and other scenarios. The control system's functional modules are divided as follows: First, the standard parameter and environmental correction module. Its control logic corrects the output characteristics of the photovoltaic cells based on illumination and temperature, calculating current and voltage parameters under actual operating conditions. The corrected parameters will be used for subsequent calculations of the mathematical model equations for the new energy power plant. Second, the equation calculation module, the core of the model, is used to solve the coefficients of complex terms such as exponential terms in the equations. It calculates the nonlinear current-voltage characteristics describing the new energy power plant through dividers, adders, and exponent modules. For the photovoltaic array, the generation of exponential terms is mainly achieved through the Math Function module, simulating nonlinear conduction characteristics and reflecting the current variation law of the photovoltaic PN junction. Then, the output characteristic calculation module calculates and combines the results of complex terms, using multipliers and adders to achieve the complete calculation of the equations and obtain the output characteristics. Additionally, the Rs module introduces voltage drop correction to simulate the impact of internal losses on the output characteristics. The output results are used in grid-connected simulations.

[0046] Figure 4 This presents the full topology of the example system provided by this invention, and performs broadband oscillation analysis after targeted upgrades to the two-region power system topology combined with a high-proportion renewable energy power plant broadband application scenario. This dual-region interconnected power system mainly comprises two regions, left and right, interconnected by two parallel tie lines, forming a typical two-region system. Each region is equipped with dual power sources: G1 and G2 are region 1 power sources, and G3 and G4 are region 2 power sources, connected to the regional bus via step-up transformers Tr1~Tr4. G2 and G4 are wind farms / photovoltaic power plants, while G1 and G3 are retained as synchronous generators, simulating a high renewable energy proportion scenario. Generators within the region are connected to the regional bus via transformers, and local loads can be configured. The two tie lines between regions enable power exchange and are key channels for inter-regional oscillations and the cross-regional propagation of broadband oscillations.

[0047] Synchronous generating units G1 and G3 employ a subtransient model, configuring speed governors and excitation systems to simulate electromechanical transient characteristics. New energy power stations G2 and G4 are photovoltaic power stations; RLC parameters are configured to simulate transmission line impedance characteristics. By adjusting control parameters and RLC parameters, and setting different impedances for the tie lines, weak grid conditions in inter-regional power transmission are simulated, amplifying the risk of broadband oscillations and simulating actual operating conditions to induce oscillations. Broadband oscillation scenarios are constructed by adjusting the new energy penetration rate, the output ratio of G2 and G4 to 50%, the tie line power level, and control parameter deviations. Measurement points are set at the generator outlet, the high-voltage side of the transformer, the regional busbar, and both ends of the tie lines to collect signals such as three-phase voltage, current, and active / reactive power.

[0048] The parameter identification principle and implementation logic used in this invention for broadband oscillations in a high-proportion renewable energy system with four generators and two regions, encompassing theoretical, simulation, and real-time experimental scenarios, is the system described in the aforementioned flowchart. Broadband oscillations in high-proportion renewable energy power systems are characterized by multi-mode, nonlinear, high-noise, and time-varying features. The core objective of parameter identification is to extract key parameters of the oscillation modes from voltage / current / power time-series signals, such as frequency, amplitude, phase, damping ratio, and participation factor, to distinguish between dominant and secondary oscillation modes, providing quantitative basis for subsequent propagation path analysis, oscillation source localization, and damping control.

[0049] In the core component, MATLAB / Simulink simulation combined with RT-LAB real-time digital dynamic model experimental analysis is used. The MATLAB / Simulink simulation achieves noise-free, controllable, and reproducible wideband oscillation parameter identification data, which is mainly used for algorithm verification and parameter calibration.

[0050] 1. Data Acquisition: In the Simulink four-machine two-region model, voltage, current, and power measurement modules are set up to export the three-phase signals of each key node (new energy power station outlet, both ends of the tie line, and regional bus) and convert them into per-unit values ​​or named value sequences.

[0051] 2. Preprocessing process. Defect removal removes singular points generated during simulation due to step size and initial value issues; noise reduction employs AVMD-SVD combined noise reduction to separate noise and oscillation components, improving identification accuracy.

[0052] 3. The core identification algorithm performs time-frequency localization of broadband oscillations by multi-scale decomposition, projecting the signal into subspaces of different frequencies to achieve joint time-frequency domain analysis.

[0053] The core principle of parameter identification in RT-LAB real-time digital dynamic model experiments is real-time simulation and semi-physical closed-loop verification. Data comes from the real-time simulation platform or connected physical devices, more closely resembling real-world power grid conditions, thus placing higher demands on the real-time performance and robustness of the identification algorithm. RT-LAB experimental data includes real-time simulation noise, AD / DA conversion errors, and interference from semi-physical devices. The signal is a continuous data stream, and online identification meets hard real-time requirements, with algorithm computation latency typically in the microsecond range. RT-LAB divides the four-machine, two-region model into two parts: the main simulation core (SM core) runs the main power system model, outputting voltage and current signals from key nodes; the auxiliary core (SS core) runs the real-time parameter identification algorithm, receives the data stream from the main core, and calculates modal parameters online. Low-latency signal transmission is achieved through shared memory or a real-time communication bus, enabling real-time output of the oscillation mode parameters obtained from data interaction identification, which can be displayed on the RT-LAB monitoring interface.

[0054] Through simulation and semi-physical real-time analysis on the RT-LAB platform, the amplitude-time characteristics and amplitude-frequency characteristics of the system can be obtained, such as... Figure 5 , Figure 6 As shown. Based on these two characteristics, the propagation path of the system oscillation and the location of the oscillation source can be determined. Figure 4 And the diagram illustrating the transfer entropy is as follows: Figure 7 As shown.

[0055] The modal identification results show that the system has four typical oscillation modes: Mode 1 is a 50Hz power frequency synchronous component with a damping ratio that is stable at around 0.45 and shows no divergence trend; Mode 2 is an 80Hz supersynchronous oscillation with a damping ratio close to 0 and persistent, which is the dominant instability mode of the system; Modes 3 and 4 are subsynchronous / high-frequency secondary coupling modes with a damping ratio that gradually increases over time, posing a risk of multimodal coupling.

[0056] Based on modal identification and transfer entropy analysis, the oscillation propagation characteristics of a high-proportion renewable energy power system can be summarized as follows: Oscillation source location: Node 2 of the renewable energy power station is the excitation source of this supersynchronous oscillation. Its interaction with the weak power grid triggers a low-damping dominant oscillation mode. Propagation path: The oscillation energy propagates across regions along the main channel of nodes 2-7-11-4. The transfer entropy intensity reaches a peak of 1.254 at node 4, which is the most critical oscillation amplification node in the system. Simultaneously, the oscillation energy couples with the generator shaft oscillation through the inter-regional tie line, forming a multi-mode superposition effect.

[0057] Risk Assessment: The interaction oscillation between the photovoltaic power plant node 2 and the weak power grid is classified as a Level 1 risk mode, posing a risk of continuous divergence and instability. The inter-regional tie lines 7-11 and node 4 are critical vulnerable links in the system, requiring targeted additional damping control measures to suppress the cross-regional propagation and amplification of oscillation energy. This result indicates that the interaction between the photovoltaic power plant and the weak power grid is the core cause of this broadband oscillation, and the tie line branches are the key amplification links for oscillation energy, necessitating targeted damping control and reinforcement of vulnerable branches.

[0058] The various embodiments in this invention are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0059] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0060] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A system for analyzing the propagation of oscillation instability in grid-connected systems with a high proportion of renewable energy, characterized in that, The system adopts a layered distributed architecture, comprising: a sensing and acquisition module, a network transmission module, a data storage module, a core analysis module, and an application display module. The sensing and acquisition module synchronously collects multi-dimensional electrical quantities in the grid-connected system. The network transmission module transmits the data collected by the sensing and acquisition module to the data storage module. The data storage module uses a hybrid storage mode combining real-time and historical databases to classify and store the received data. The core analysis module, based on the data in the data storage module, performs modal parameter identification, propagation path analysis, and oscillation source localization for oscillation instability by integrating adaptive signal processing and Copula transfer entropy nonlinear causal analysis. The application display module provides multi-dimensional visualization and early warning of the analysis results from the core analysis module.

2. The oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems according to claim 1, characterized in that, The sensing and acquisition module integrates multiple synchronous acquisition devices, including a power grid synchronous phasor measurement device, a fault recording device, a new energy power station grid connection point monitoring and control terminal, a converter operation status monitoring device, a flexible DC transmission system monitoring terminal, and an electrochemical energy storage grid connection point monitoring device; the multi-dimensional electrical quantities include bus voltage, line current, active power, reactive power, frequency, converter control signals, and energy storage charging and discharging power; the sensing and acquisition module adopts BeiDou / GPS dual-mode time synchronization.

3. The oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems according to claim 1, characterized in that, The network transmission module is compatible with IEC61850, IEC60870-5-104 and MQTT communication protocols, constructs a transmission channel combining fiber optic private network and wireless private network, controls the transmission delay within a preset time threshold, and has data encryption and packet loss retransmission functions.

4. The oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems according to claim 1, characterized in that, In the data storage module, the real-time database is used to store millisecond-level real-time monitoring data to meet online analysis needs; the historical database adopts a distributed storage architecture to store full-volume time-domain waveforms, frequency-domain spectra, oscillation mode parameters, propagation path data, experimental data, and field-measured data; the data storage module also establishes a data quality verification mechanism to remove bad and redundant data.

5. The oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems according to claim 1, characterized in that, The core analysis module specifically includes: The signal preprocessing submodule is used to perform noise reduction processing on the original signal using the AVMD-SVD joint noise reduction algorithm; The wideband oscillation mechanism analysis submodule is used to analyze the oscillation generation mechanism; The modal parameter identification submodule is used to extract oscillation mode parameters from the denoised signal; The oscillation propagation path analysis submodule is used to construct a directed weighted electrical network based on the Copula transfer entropy algorithm to trace the oscillation propagation path; The oscillation source positioning submodule is used to locate the oscillation excitation source; The propagation vulnerability assessment submodule is used to assess critical nodes and vulnerable branches in oscillation propagation networks.

6. The oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems according to claim 5, characterized in that, The modal parameter identification submodule employs a joint identification algorithm combining the Prony algorithm, the ESPRIT algorithm, and wavelet transform to achieve accurate extraction of oscillation modes across the entire frequency band. Specifically, the Prony algorithm is used for the preliminary identification of the frequency, amplitude, damping ratio, and phase of stationary oscillation signals; the ESPRIT algorithm is used for the identification of harmonic and oscillatory coupled signals; and the wavelet transform is used for joint time-frequency domain analysis to locate the oscillation occurrence time and frequency distribution.

7. The oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems according to claim 5, characterized in that, The oscillation propagation path analysis submodule quantifies the nonlinear and non-Gaussian oscillation propagation causal relationship through the Copula transfer entropy algorithm, outputs the oscillation propagation path, propagation delay and causal strength, and dynamically identifies the main propagation branch and key nodes.

8. The oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems according to claim 5, characterized in that, The oscillation source location submodule adopts a multi-dimensional oscillation source location algorithm, which is based on the principle of energy function or impedance identification and combined with the time-frequency domain characteristics of electrical quantities of each node to accurately locate the equipment or site where the oscillation excitation source is located.

9. The oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems according to claim 1, characterized in that, The system also has a three-level verification system, which includes: network topology structure demonstration mechanism analysis, MATLAB / Simulink simulation verification, and RT-LAB real-time digital dynamic model experimental verification, to comprehensively verify the accuracy and practicality of the system.

10. The oscillation instability propagation analysis system for high-proportion renewable energy grid-connected systems according to claim 1, characterized in that, The application demonstration module automatically generates early warnings and decision suggestions based on the analysis results, providing support for broadband oscillation suppression and system stability control.

Images