Power system source network load storage panoramic operation state accurate evaluation and monitoring method, system, device and medium

Abstract

The invention discloses a power system source network load storage panoramic operation state accurate evaluation and monitoring method, system and device and a medium, and belongs to the technical field of data evaluation, and the method comprises the steps: collecting the real-time operation data of a power system, and carrying out the preprocessing; based on the preprocessed real-time operation data, establishing a first mathematical model, and performing dynamic refined modeling on system power balance; performing comprehensive analysis on the real-time operation data based on the first mathematical model, the first algorithm and the second algorithm, and outputting a panoramic operation state result; and utilizing an ARIMA time sequence prediction model and combining a panoramic operation state result to predict a future operation trend of the power system, identifying potential risk points and generating corresponding early warning information. On the basis of a traditional power balance model, a line loss dynamic correction parameter and an energy storage charging and discharging efficiency coefficient are introduced, and the calculation deviation caused by neglecting the power transmission loss time-varying characteristic and the energy storage energy conversion loss in the traditional model is effectively overcome.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data evaluation, and in particular to a power system source-grid-load-storage panoramic operation state accurate evaluation and monitoring method, system, device and medium. BACKGROUND

[0002] With the large-scale access of high-proportion renewable energy represented by wind power and photovoltaic power to the power grid, and the wide participation of flexible resources such as electric vehicles, distributed energy storage and temperature-controlled loads in power system operation, the power source structure, load characteristics and network morphology of the power system have undergone profound changes, and the coupling relationship between the "source-grid-load-storage" links is increasingly close and dynamically interactive and complex.

[0003] Traditional power system monitoring and evaluation methods are mostly focused on a single link (such as power generation output or power grid flow), lack unified modeling and panoramic perception ability for multi-element collaborative operation state, and are difficult to fully reflect the overall operation situation of the system, with the problem of "local observability and global distortion".

[0004] At the same time, the existing evaluation models generally have the defect of insufficient accuracy: on the one hand, the time-varying characteristics of transmission line losses are often ignored, and fixed values or idealized assumptions are used, resulting in deviation of power balance calculation from reality; on the other hand, the modeling of energy storage units often does not consider the energy conversion efficiency in the charging and discharging process, causing the regulation capacity evaluation to be overly optimistic and affecting the reliability of dispatching decisions.

[0005] In addition, multi-source heterogeneous data from SCADA, PMU, EMS and other sources have problems such as time asynchronization, different accuracy, and local conflicts, and traditional methods lack effective fusion mechanisms, which can easily introduce noise and errors. More importantly, current monitoring systems mostly rely on threshold alarm mechanisms, which belong to passive response mode, and are difficult to identify potential risks such as energy storage overrun, voltage instability, supply-demand imbalance and other gradual characteristics in a forward-looking manner.

[0006] Due to the lack of a unified mathematical model and analysis framework that can integrate the dynamic behavior of power sources, power grids, loads and energy storage, and their physical constraints, existing technologies face serious challenges in achieving high-precision, real-time and predictable panoramic operation state evaluation. Therefore, a power system source-grid-load-storage panoramic operation state accurate evaluation and monitoring method and system are provided. SUMMARY

[0007] In view of the above problems, the present application is proposed.

[0008] Therefore, the technical problem solved by the present application is: how to overcome the problems of insufficient model accuracy, difficulty in multi-source data fusion and lag of risk early warning in the prior art, and provide a unified technical solution capable of realizing high-precision, strong real-time and predictable power system source-grid-load-storage panoramic operation state perception and forward-looking risk early warning.

[0009] To solve the above technical problems, the present application provides the following technical solution: a power system source-grid-load-storage panoramic operation state accurate evaluation and monitoring method, which comprises, collecting real-time operation data of the power system for preprocessing; based on the preprocessed real-time operation data, establishing a first mathematical model to dynamically and finely model the system power balance; based on the first mathematical model, and the first algorithm and the second algorithm, comprehensively analyzing the real-time operation data, and outputting panoramic operation state results; using an ARIMA time series prediction model, combining the panoramic operation state results, predicting the future operation trend of the power system, identifying potential risk points and generating corresponding early warning information.

[0010] As a preferred scheme of the power system source-grid-load-storage panoramic operation state accurate evaluation and monitoring method, wherein: the real-time operation data of the power source end, the power grid end, the load end and the energy storage end of the power system are collected, and the real-time operation data is preprocessed, comprising, collecting real-time operation data of the power system; performing outlier identification and elimination operations on the real-time operation data; unifying the data time base of real-time operation data of different sources.

[0011] As a preferred scheme of the power system source-grid-load-storage panoramic operation state accurate evaluation and monitoring method, wherein: based on the preprocessed real-time operation data, a first mathematical model is established to dynamically and finely model the system power balance, comprising, based on the physical connection relationship of the power grid, defining the network topology of the first mathematical model; by integrating the core physical constraints of each link of the power system, a constraint system of the first mathematical model is constructed; and when constructing the global power balance of the first mathematical model, introducing line loss dynamic correction parameters and energy storage charging and discharging efficiency coefficients to complete the dynamic modeling of the system power balance.

[0012] The application overcomes the precision defects of traditional evaluation models caused by ignoring the time-varying characteristics of loss and energy conversion loss by constructing a first mathematical model and introducing a line loss dynamic correction parameter and a storage charging and discharging efficiency coefficient in the global power balance relationship. This improvement enables the power balance calculation between power generation, load, storage and power grid to dynamically fit the actual operating conditions, improving the authenticity and accuracy of panoramic operation state assessment.

[0013] As a preferred scheme of the power system source network load storage panoramic operation state precise evaluation and monitoring method, wherein: the first mathematical model is based on the first algorithm and the second algorithm to comprehensively analyze the real-time operation data, and output the panoramic operation state result, including, mapping each end real-time operation data to the corresponding node and state parameter of the first mathematical model, establishing the association relationship between data and model; based on the topology and constraints of the first mathematical model, the multi-source heterogeneous data are processed by the first algorithm to generate a joint state data matrix; using the second algorithm, the dynamic operation state of the nodes of the whole system is calculated and updated according to the joint state data matrix and the constraint system of the first mathematical model; output the panoramic operation state result of the power system whole-link operation index.

[0014] As a preferred scheme of the power system source network load storage panoramic operation state precise evaluation and monitoring method, wherein: the ARIMA time series prediction model is used to predict the future operation trend of the power system, identify potential risk points and generate corresponding warning information, including, collecting historical time series data and panoramic operation state results in the first mathematical model as input of the ARIMA time series prediction model; running the ARIMA time series prediction model to output the prediction trend of the key operation parameters of the power system in the future specified time window; comparing and analyzing the panoramic operation state result and the prediction trend to identify potential security out-of-limit risk points in the system; the identified potential security out-of-limit risk points are converted into structured warning information containing risk position, type and time window.

[0015] The application realizes quantitative prediction of future operation trend of the power system and early identification of potential risks by combining the ARIMA time series prediction model with the panoramic operation state result, breaks through the passive response mode of the traditional threshold alarm mechanism, and can issue a forward-looking early warning for risk points with gradual characteristics such as line overload, voltage instability, and energy storage out-of-limit.

[0016] As a preferred scheme of the power system source network load storage panoramic operation state accurate evaluation and monitoring method, the line loss dynamic correction parameter is used in the power balance equation of the first mathematical model, and the total active loss of the system is dynamically corrected according to the square relationship between the line current and the resistance. The energy storage charging and discharging efficiency coefficient is a dynamic variable depending on the real-time operation state of the energy storage unit, which is used in the energy storage state of charge dynamic equation of the first mathematical model, and represents the energy conversion loss in the charging and discharging process.

[0017] As a preferred scheme of the power system source network load storage panoramic operation state accurate evaluation and monitoring method, the first algorithm is a data fusion algorithm based on weighted average, and the real-time operation data of different sources is assigned a weight to form a fused joint state data matrix. The second algorithm is a state estimation algorithm, which calculates and updates the dynamic state parameters of each node based on the joint state data matrix and the constraint conditions of the first mathematical model.

[0018] The application provides a power system source network load storage panoramic operation state accurate evaluation and monitoring system.

[0019] To solve the above technical problems, the application provides the following technical scheme: a power system source network load storage panoramic operation state accurate evaluation and monitoring system, comprising: a data collection module, a mathematical model construction module, a comprehensive analysis module, and an output module. The data collection module collects real-time operation data of the power system for preprocessing. The mathematical model construction module establishes a first mathematical model based on the preprocessed real-time operation data to dynamically and finely model the system power balance. The comprehensive analysis module performs comprehensive analysis on the real-time operation data based on the first mathematical model, a first algorithm, and a second algorithm, and outputs panoramic operation state results. The output module uses an ARIMA time series prediction model and combines the panoramic operation state results to predict future operation trends of the power system, identify potential risk points, and generate corresponding early warning information.

[0020] The application provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and wherein the processor implements steps of the power system source network load storage panoramic operation state accurate evaluation and monitoring method when executing the computer program.

[0021] The application provides a computer readable storage medium, which stores a computer program, wherein the computer program implements steps of the power system source network load storage panoramic operation state accurate evaluation and monitoring method when executed by a processor.

[0022] The application has the beneficial effects that: the application introduces a line loss dynamic correction parameter and an energy storage charging and discharging efficiency coefficient on the basis of a traditional power balance model, effectively overcomes calculation deviations caused by ignoring time-varying characteristics of power transmission loss and energy storage energy conversion loss in the traditional model, and makes the energy flow relationship between power generation, load, energy storage and the power grid closer to the real working condition through real-time correction of line loss and actual energy conversion efficiency of energy storage, especially in complex scenes such as new energy fluctuation and load mutation, and still maintains high modeling accuracy. Not only the sensitivity to system power imbalance is enhanced, but also reliable data basis is provided for subsequent state estimation, trend prediction and risk warning, and the authenticity, credibility and stability of panoramic operation state evaluation are improved.

[0023] The application realizes quantitative prediction of future operation trend and early identification of potential risks of the power system by combining an ARIMA time series prediction model with panoramic operation state results. The passive response mode of the traditional threshold alarm mechanism is broken through, and a forward-looking early warning can be given to risk points with gradual characteristics such as line overload, voltage instability and energy storage overrun. This provides a valuable decision window for dispatchers and realizes a leap from passive response to fault to active prevention of risks. BRIEF DESCRIPTION OF DRAWINGS

[0024] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0025] Figure 1 A power system source network load storage panoramic operation state accurate evaluation and monitoring method provided by an embodiment of the application is shown in the general flowchart. DETAILED DESCRIPTION

[0026] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the protection scope of the present application.

[0027] Embodiment 1, refer to Figure 1 For an embodiment of the present application, the embodiment provides a power system source network load storage panoramic operation state accurate evaluation and monitoring method, comprising: In the prior art, there are three outstanding problems in power system monitoring: first, the traditional model ignores the time-varying characteristics of line loss and the energy conversion efficiency of energy storage, which leads to power balance calculation deviation and causes "evaluation distortion". Second, the multi-source data such as SCADA and PMU lack effective fusion mechanism, and it is difficult to form accurate global state perception, and "panoramic difficulty" appears. Third, the passive mode relying on threshold alarm cannot prospectively identify gradual risks.

[0028] Therefore, the present application realizes accurate modeling by constructing a source-network-load-storage joint mathematical model introducing dynamic correction parameters, completes multi-source data comprehensive analysis by using fusion algorithm and state estimation, realizes the leap from real-time evaluation to prospective warning combined with ARIMA model, and finally forms the accurate evaluation and active defense system of panoramic operation state.

[0029] S1, collecting real-time operation data of the power system for preprocessing; S2, based on the preprocessed real-time operation data, a first mathematical model is established to dynamically and finely model the system power balance; S3, based on the first mathematical model, and the first algorithm and the second algorithm are used to comprehensively analyze the real-time operation data, and output the panoramic operation state result; S4, using an ARIMA time series prediction model, and combining the panoramic operation state result, predicting the future operation trend of the power system, identifying potential risk points and generating corresponding warning information.

[0030] Embodiment 2, for an embodiment of the present application, based on the previous embodiment, a power system source network load storage panoramic operation state accurate evaluation and monitoring method is provided, comprising: Step S1, collecting real-time operation data of the power system for preprocessing, comprising steps S11-S13: S11, collecting real-time operation data of the power system.

[0031] Collecting real-time operation data of the power source end, the power grid end, the load end and the energy storage end of the power system.

[0032] Real-time operating data includes: generator output power, node voltage, current, load curve, and state of charge.

[0033] S12. Perform outlier identification and removal operations on real-time running data.

[0034] Furthermore, a statistical detection method based on the 3σ criterion is used to identify and eliminate abnormal data that exceeds the normal fluctuation range for real-time operating data; data that clearly does not conform to the operating rules of the power system is identified and eliminated through data rationality verification; and data consistency verification is applied to identify and eliminate contradictory multi-source measurement data.

[0035] S13. Establish a unified data time benchmark for real-time operational data from different sources.

[0036] A unified timestamp system is established for data from different acquisition systems; an interpolation algorithm is used to unify data from different sampling frequencies to a common time base; and a data buffering mechanism is used to handle timing inconsistencies caused by communication delays.

[0037] In the implementation of this application, the preprocessing of S1 involves collecting real-time operating data from the power source, grid, load, and energy storage, including multi-dimensional parameters such as generator output power, node voltage, load curve, and state of charge. Subsequently, abnormal and contradictory data are identified and eliminated through anomaly detection, data rationality verification, and consistency check based on the 3σ criterion. Finally, a unified timestamp system is established, and interpolation algorithms and data buffering mechanisms are used to solve the time synchronization problem of multi-source data.

[0038] In one alternative embodiment, the preprocessing of S1 adopts a simplified process based on a data quality scoring mechanism. By establishing a quality scoring system for each data source, high-quality data sources are automatically selected, thereby reducing the complexity of data processing.

[0039] In another alternative embodiment, the preprocessing of S1 adopts a simplified process based on a data quality scoring mechanism. By establishing a quality scoring system for each data source, high-quality data sources are automatically selected, reducing the complexity of data processing.

[0040] This invention fundamentally eliminates the state assessment bias caused by data quality defects by performing systematic anomaly identification, rationality verification and time synchronization processing on multi-source heterogeneous data, and effectively avoids the error transmission and amplification problems caused by data defects in traditional methods.

[0041] Step S2: Based on the preprocessed real-time operating data, establish the first mathematical model to perform dynamic and refined modeling of the system power balance, including the following steps S21~S23: S21. Based on the physical connection relationship of the power grid, define the network topology of the first mathematical model.

[0042] Based on the power system topology, power supply units, load units, energy storage units, and key substation nodes are generated and used as nodes of the power system. Each node contains corresponding real-time operating parameters: power supply unit: output power, unit status; load unit: instantaneous load, current demand; energy storage unit: state of charge (SOC), charging and discharging power.

[0043] The transmission lines, distribution lines, and grid connections are defined as edges. Power transfer relationships and line impedance relationships between nodes are established to form a grid topology matrix. Specifically, based on the power system topology, the electrical connections of power source, load, energy storage, and key substation nodes are identified and each node is uniquely numbered. Then, based on the physical connections of transmission and distribution lines, parameter information for each line is extracted, including line length, cross-sectional area, conductor material, and rated voltage level. Based on this, the impedance parameters (resistance, reactance) and admittance parameters of the line are calculated. On this basis, power transfer equations between nodes are constructed to describe the active and reactive power exchange between nodes and adjacent nodes, forming node power balance equations. Finally, all node and line parameters are organized in matrix form to generate the grid topology matrix and impedance matrix.

[0044] The real-time operating parameters of each node are combined in time series order within a sliding time window w to form a multi-dimensional feature vector, which is used to reflect the dynamic operating characteristics of each link of source-grid-load-storage. At the same time, historical data is integrated to increase the time continuity of the model.

[0045] S22. By integrating the core physical constraints of each link in the power system, a constraint system for the first mathematical model is constructed.

[0046] By establishing generator output power constraint and power balance models for the power source, power flow equations and node voltage constraint models for the grid, and load forecasting models for the load (specifically: collecting historical power data, electricity consumption characteristics, environmental factors (such as temperature, holidays, etc.) and real-time load measurements for each load node, and cleaning, normalizing, and removing outliers; then, selecting the ARIMA model based on time series, constructing input feature vectors, and training the model to learn load change patterns; next, evaluating the model accuracy through cross-validation and adjusting model parameters according to load node characteristics; finally, applying the trained load forecasting model to real-time data and outputting the predicted power demand values ​​for each node within the future time window), and establishing dynamic changes in state of charge and charging / discharging constraint models for the energy storage end, the nodes, lines, and dynamic state parameters of each end are integrated according to the system topology to form a source-grid-load-storage joint mathematical model. This joint mathematical model can be used for joint evaluation of system power balance, voltage stability, and energy storage regulation capabilities, realizing a panoramic description and analysis of the overall operating status of all aspects of the entire power system.

[0047] Establishing a generator output power constraint and power balance model for the power source includes the following steps: Identify each generator node at the power supply end, including parameters such as the rated capacity, operating status, maximum power generation, and minimum power generation of each unit, and assign a unique identifier to each generator node; Based on the rated capacity and operating range of the unit, establish a generator output power constraint model: , ; in, For the first The generator at time 'output power' and For the minimum and maximum allowable power of the unit, This represents the total number of generators at the power supply end, i.e., the number of generators participating in power output.

[0048] Identify all nodes on the power grid side, including substation nodes, load aggregation nodes, and power access nodes; determine the transmission lines and distribution lines between each node and their electrical parameters (impedance, rated capacity, conductor length, etc.); uniquely identify nodes and lines in the model to establish a topology matrix; Based on the nodal injected power (generating power, load power, and energy storage power) and line impedance, the AC power flow equations are established: ; ; In the formula, Inject active power into the nodes, Inject reactive power into the node. For nodes voltage amplitude, The phase angle difference of the node voltage. , These are elements of the line admittance matrix; For nodes The active power injected into the load or the active power injected into the generator (positive indicates generation, negative indicates load). For nodes The load injects reactive power or the generator injects reactive power. For nodes The set of connected neighboring nodes. For nodes voltage amplitude, This is the real part of the admittance matrix of a line or node, representing conductivity. This is the imaginary part of the line or node admittance, representing the susceptance.

[0049] Voltage constraint equations are established for each node, and power flow constraints are established for each line to ensure that the lines are not overloaded. When the node voltage deviates from the allowable range, the reactive power output of the generator or the power factor of the energy storage terminal is adjusted through the model feedback mechanism to ensure voltage stability. The voltage constraint equation is as follows: In the formula, For a moment node voltage, and The upper and lower limits of the allowable voltage are set to ensure the stable operation of the power grid; Current constraints are: In the formula, For a moment Line complex power, This is the rated capacity.

[0050] Furthermore, a dynamic state of charge (SOC) change and charge / discharge constraint model is established for the energy storage device, including the following steps: Identify all energy storage units in the power system and record the rated capacity of each energy storage unit. Initial state of charge Maximum charge and discharge power and minimum charge / discharge power ; Based on energy storage charging and discharging power and energy storage charge and discharge efficiency coefficient Establish the dynamic equations of the charged state: ; In the formula: For the first Each energy storage unit at time The state of charge; The charging power is greater than 0. The discharge power is greater than 0. For time step; Charge state constraints are established based on the state of charge, and upper and lower limit constraints are established based on the energy storage charging and discharging power to ensure the safe operation of the equipment. The state of charge constraint is as follows: ; The upper and lower limit constraints are: ; In the formula, This represents the minimum permissible state of charge for the energy storage unit; below this value, the safe operation of the energy storage will be affected. This represents the maximum permissible state of charge for the energy storage unit; exceeding this value will lead to overcharging risks. This is the minimum allowable power (discharge capacity or lower limit of safe power) for the energy storage unit. This refers to the maximum power allowed for the energy storage unit (charging capacity or safe power limit).

[0051] The nodes, lines, and dynamic status parameters at each end are integrated according to the system topology to form a joint mathematical model of source-grid-load-storage, including the following steps: Generator node at the power supply end Load end load node Energy storage nodes and key substation nodes on the power grid side Summarize and form a joint node set. Each node is assigned a unique identifier, and its type, capacity, operating status, and dynamic parameters (power, state of charge, voltage, current, etc.) are recorded. The connection relationships between transmission lines, distribution lines, and power sources, loads, and energy storage are defined as edge sets. For each edge, electrical parameters (impedance, rated capacity), connection nodes, and direction information are recorded to ensure a clear power flow relationship; Collect dynamic state parameters (such as power generation) from nodes at the power source, load, energy storage, and grid ends. Load power Energy storage state of charge Voltage amplitude The dynamic state parameters are integrated into a joint state matrix according to the time series. .

[0052] Based on node set Edge set and dynamic state matrix Constructing a joint mathematical model of source-grid-load-storage : This model can fully describe the real-time operating status of all aspects of the entire power system, providing a foundation for panoramic operating status assessment, short-term forecasting, and risk warning. Constraints from the power source, load, energy storage, and grid ends are integrated into a joint mathematical model of source-grid-load-storage. The constraints include: upper and lower limits of power and power balance constraints at the power source end; power flow equations, node voltage constraints, and line power constraints at the grid end; power prediction and regulation constraints at the load end; and state of charge and charging / discharging power constraints at the energy storage end.

[0053] S23. When constructing the global power balance of the first mathematical model, dynamic correction parameters for line loss and energy storage charging and discharging efficiency coefficients are introduced to complete the dynamic modeling of the system power balance.

[0054] Combined with the total load of the power system Line loss and energy storage charging and discharging power A power balance model is established, and line loss correction parameters are introduced into the power balance model. and energy storage charge and discharge efficiency coefficient This is used to improve the accuracy of the power balance model. Furthermore, in the comprehensive operational status assessment of the power system's power generation, grid, load, and storage, traditional power balance models typically assume that transmission line losses are fixed or negligible, making it difficult to reflect dynamic loss changes caused by different operating conditions (such as line temperature variations, load fluctuations, and voltage deviations). To address this issue, a line loss correction parameter is introduced. It is used to dynamically correct the theoretically calculated line loss value, so that the model can adaptively adjust the loss estimate according to the real-time operating status.

[0055] Its advantage lies in significantly improving the accuracy of power balance calculation, avoiding generation dispatch errors and node power imbalances caused by loss estimation deviations, thereby achieving accurate assessment of the actual energy flow status of the power grid. In the source-grid-load-storage system, energy storage units experience energy losses during charging and discharging. If an idealized energy conversion relationship is directly adopted, it will lead to a deviation between the energy balance at the energy storage end and the total power balance of the system, affecting the authenticity of the assessment results.

[0056] To address this issue, an energy storage charge / discharge efficiency coefficient is introduced. This is used to describe the energy conversion efficiency of an energy storage system during the charging and discharging process.

[0057] Its advantage lies in its ability to accurately characterize the actual energy conversion characteristics of energy storage units, making the model more consistent with real working conditions when calculating supply and demand balance and energy storage regulation capacity, thereby improving the credibility and stability of panoramic operation status monitoring and risk warning. The power balance model after introducing line loss correction parameters and energy storage charge / discharge efficiency coefficients is as follows: In the formula, This refers to the total number of energy storage devices in the power system. For the index of the energy storage unit, For the first Each energy storage unit at time The charging and discharging power.

[0058] Ensure that the total power generation can meet the real-time load demand and energy storage regulation requirements of the system.

[0059] In the implementation of this application, the first mathematical model of S2 is: to establish a joint mathematical model of source-grid-load-storage.

[0060] The network topology framework is formed by generating a set of nodes and a set of connecting edges based on the power system topology. The power balance constraints at the source, the power flow equations and voltage security constraints at the grid, the power characteristic constraints at the load, and the dynamic constraints of the state of charge at the energy storage end are integrated to form a unified constraint system. Furthermore, dynamic correction parameters for line losses and energy storage charging and discharging efficiency coefficients are introduced into the global power balance relationship to complete the dynamic and refined modeling of the system power balance.

[0061] In one alternative embodiment, the first mathematical model of S2 is a simplified modeling method based on equivalent circuits.

[0062] The generator is equivalent to a controlled current source, the transmission and distribution line is equivalent to a π-type equivalent circuit, the load is equivalent to a parallel connection of a variable impedance and a current source, and the energy storage system is equivalent to a controlled voltage source series internal resistance model. A unified electrical equation for the entire network is established by the nodal voltage method, and a loss correction term based on the square of the line current and an energy storage efficiency compensation term based on the SOC-efficiency curve are introduced into the total power balance equation, which significantly reduces the computational complexity while ensuring the accuracy of the model.

[0063] In another alternative embodiment, the first mathematical model of S2 is a generalized modeling method based on energy flow.

[0064] A multi-timescale energy flow equation, encompassing second, minute, and hourly scales, is established to describe the entire dynamic process from electromagnetic transients to electromechanical transients. In the energy transfer path, a power-to-energy conversion model is established on the power source side, an energy flow transmission model considering distributed parameters is established on the grid side, an energy consumption characteristic model is established on the load side, and an energy throughput model considering aging conditions is established on the energy storage side. By embedding time-varying efficiency correction factors at the energy conversion interfaces of each stage, an accurate characterization of the energy flow process of the entire system is achieved.

[0065] This invention addresses the power balance calculation errors caused by traditional models that neglect real-time changes in line losses and the charging and discharging efficiency of energy storage by establishing a joint mathematical model and introducing two key parameters. This makes the energy flow relationship between power generation, load, energy storage, and the power grid more consistent with reality, avoiding scheduling decision errors caused by inaccurate models.

[0066] Step S3: Based on the first mathematical model, and using the first and second algorithms to comprehensively analyze the real-time running data, the panoramic running status result is output, including the following steps S31~S34: S31. Map the real-time running data of each terminal to the corresponding nodes and state parameters of the first mathematical model to establish the relationship between data and model.

[0067] Map the real-time operational data collected from each terminal to the source-grid-load-storage joint mathematical model. The corresponding nodes and state parameters.

[0068] S32. Based on the topology and constraints of the first mathematical model, the first algorithm is used to collaboratively process multi-source heterogeneous data to generate a joint state data matrix. S33. Using the second algorithm, based on the constraint system of the joint state data matrix and the first mathematical model, calculate and update the dynamic operating state of all system nodes; S34. Output the panoramic operation status results of all aspects of the power system's operation indicators.

[0069] It outputs complete real-time operating status results, including: real-time generation power, output margin, and power deviation at the power source; node voltage amplitude, phase angle distribution, power flow direction, and line load rate at the grid; actual power consumption, demand response level, and power deviation at the load; state of charge (SOC), charging and discharging power, and adjustable capacity at the energy storage end; and comprehensive evaluation results such as overall system power balance deviation, energy loss rate, steady-state margin, and operational risk level, which reflect the overall real-time operational health and safety of the power system.

[0070] In the embodiments of this application, the first algorithm of S32 is a weighted average data fusion algorithm, which assigns fusion weights based on data reliability and measurement accuracy to real-time running data from different data sources, performs collaborative processing on multi-source heterogeneous data, and finally generates a unified joint state data matrix.

[0071] By utilizing the topological relationships and node and line constraints in the source-grid-load-storage joint model, the operational data from different ends are fused based on the weighted average method to form a fused joint state data matrix.

[0072] Specifically, based on the topology in the joint model, the connection relationships and power transfer paths between power sources, loads, energy storage, and various nodes of the power grid are determined; then, the real-time operating data (power generation, load power, voltage, current, state of charge, etc.) collected from each end are mapped to the state variables of the corresponding nodes.

[0073] Next, based on node type, data source reliability, and correlation with system power balance, fusion weighting coefficients are set, such as the weight of the generator end. Load side weight Energy storage weight Grid weight .

[0074] At every moment Weighted average fusion is performed on multi-source data from the same node or line to calculate the fusion status value. (In the formula, For the first Data sources at time Input value, These are the weighting coefficients. The number of data sources participating in the fusion is used to verify and correct the fusion results through line constraints (such as power flow equations, voltage range, and capacity limits).

[0075] Finally, the fusion results of each node and line are integrated in chronological order to form a unified joint state data matrix. It is used to reflect the comprehensive operational characteristics of the entire source-grid-load-storage system.

[0076] In one alternative embodiment, the first algorithm of S32 is a dynamic data fusion algorithm based on Kalman filtering.

[0077] A system state-space model is established, using multi-source measurement data as observed variables. The optimal estimate of the state variables is dynamically updated through a prediction-correction loop. The fusion weights of each data source are adaptively adjusted using the covariance matrix to achieve adaptive suppression of noise and outliers, thereby improving the robustness and dynamic tracking capability of data fusion.

[0078] In another alternative embodiment, the first algorithm of S32 is a dynamic data fusion algorithm based on Kalman filtering.

[0079] A system state-space model is established, using multi-source measurement data as observed variables. The optimal estimate of the state variables is dynamically updated through a prediction-correction loop. The fusion weights of each data source are adaptively adjusted using the covariance matrix to achieve adaptive suppression of noise and outliers, thereby improving the robustness and dynamic tracking capability of data fusion.

[0080] In the embodiment of this application, the second algorithm of S33 is a state estimation algorithm. Based on the joint state data matrix, and using the topological framework and physical constraint system of the first mathematical model (source-grid-load-storage joint mathematical model), the algorithm calculates and updates the optimal estimated values ​​of the dynamic operating states (such as voltage, phase angle, power, etc.) of all system nodes through mathematical solutions.

[0081] Based on the fused joint state data matrix and the constraints of the source-grid-load-storage joint model, the dynamic state parameters of each node are calculated and updated, including: Power source side: calculating the output power deviation and available power of each generator (specifically: extracting the real-time output power of each generator from the model based on the fused joint state data matrix). And its planned or predicted output Compare and calculate the output power deviation. .

[0082] At the same time, combined with the upper and lower limits of generator power constraints Calculate the available adjustable power margin This is used to evaluate its ability to participate in system power regulation.

[0083] This approach reflects both the operational deviations at the power source and dynamically quantifies its regulation potential. At the grid end: Node voltage magnitudes, phase angles, and line power flow are calculated (specifically, based on the fused node voltages and line parameters, admittance information between adjacent nodes is extracted from the topology matrix, and the node voltage magnitudes are iteratively solved using AC power flow equations). With phase angle Furthermore, the active and reactive power flow of each line is calculated. .

[0084] If the power flow exceeds the rated capacity or the node voltage exceeds the allowable range, it is marked as a potentially risky node or an overloaded line, realizing dynamic updating and safety verification of the grid's operating status; the load side calculates the actual power consumption and deviation of the load (extracting the real-time power consumption of each load node from the fusion matrix). Combined with historical load forecasts or model forecasts Calculate load deviation It is used to reflect the dynamic characteristics of load changes.

[0085] By further combining voltage and current measurement data, the actual power factor and load efficiency can be calculated to identify abnormal fluctuations or sudden load increases, providing a basis for load-side scheduling and demand response.

[0086] The energy storage unit calculates its state of charge (SOC) and adjustable power (based on the SOC dynamic equations in the joint model, utilizing the fused charge / discharge power and efficiency coefficients to update the SOC of the energy storage unit in real time). .

[0087] Simultaneously, based on the state of charge constraints and upper and lower power limits, the current adjustable power range is calculated. This allows for the determination of the rechargeable or dischargeable capacity of energy storage units, thereby providing a dynamic adjustment basis for system frequency regulation, peak shaving, or emergency support.

[0088] In one alternative embodiment, the second algorithm of S33 is a state estimation algorithm based on weighted least squares.

[0089] An optimization model is established with the objective function of minimizing the weighted sum of squares of the differences between measured and estimated values. The power flow equations and power balance constraints of the first mathematical model are used as equality constraints, and the voltage safety limit and line transmission capacity are used as inequality constraints. The optimal estimates of the node voltage, phase angle and power distribution of the entire system are obtained through iterative solution, and bad data are identified and eliminated.

[0090] In another alternative embodiment, the second algorithm of S33 employs an artificial intelligence-based deep learning state estimation method.

[0091] A deep neural network model is trained using historical operating data to establish a nonlinear mapping relationship from measurement data to system state. The joint state data matrix is ​​used as input features, and the system state estimation result is directly output through network forward propagation. At the same time, the network output is corrected by combining the physical constraints of the first mathematical model to ensure that the estimation result meets the physical laws of the power system.

[0092] This invention effectively solves the old problems of messy data and unclear status by first fusing data and then analyzing the status.

[0093] Data from different monitoring devices are processed uniformly to eliminate errors and inconsistencies, forming a complete and reliable data foundation. Using the previously established precise model, these data are analyzed in depth. This not only yields directly observable operating parameters but also allows for the calculation of state variables that are crucial to system safety but cannot be directly measured, such as the voltage phase angle of each node and the actual power transmission of the lines.

[0094] Through this processing, the previously scattered monitoring data is transformed into a comprehensive and accurate system operation status report. This report includes real-time operational data from various components such as power sources, the power grid, loads, and energy storage, and clearly demonstrates the system's safe operating range and regulation capabilities. This allows operators to fully grasp the actual operating status of the power grid, providing a solid basis for subsequent load forecasting, risk warning, and dispatching decisions, thereby improving the safety of power grid operation.

[0095] Step S4: Using the ARIMA time series forecasting model and combining it with the panoramic operation status results, predict the future operation trend of the power system, identify potential risk points, and generate corresponding early warning information, including the following steps S41~S44: S41. Collect historical time-series data and panoramic operation status results from the first mathematical model as input to the ARIMA time series prediction model.

[0096] Collect historical operational data (power generation, load power, energy storage state of charge, voltage amplitude, current, etc.) of nodes, lines, and dynamic state parameters in the source-grid-load-storage joint model.

[0097] S42. Run the ARIMA time series forecasting model and output the predicted trends of key operating parameters of the power system within a specified future time window.

[0098] The ARIMA model is used to predict the state of the power system within a future time window f, including the changing trends of power generation at the source, power demand at the load end, node voltage and line power flow at the grid end, and state of charge and available power at the energy storage end.

[0099] Historical time-series data of each node and line, such as power generation, load power, node voltage, line power flow and energy storage state of charge, are extracted from the fused joint state data matrix and normalized and outlier removed.

[0100] These time-series data are input into the ARIMA model, which learns the temporal dependencies between different nodes and the dynamic characteristics of the system. During the prediction process, joint model constraints such as power balance constraints, energy storage efficiency, line capacity, and voltage upper and lower limits are considered to iteratively generate the predicted state sequence for future time steps.

[0101] The forecast results are integrated into dynamic indicators such as power generation, load power, node voltage, line power flow and energy storage status of each node within the future time window, providing a quantitative basis for potential risk identification and early warning generation.

[0102] Furthermore, the architecture of the ARIMA model comprises three core parts: The input historical time series data is differentially processed (I layer) to eliminate trends and non-stationarity, thereby making the series stationary.

[0103] In the autoregressive (AR) layer, the current value is linearly combined with the previous p lag values ​​to extract the autocorrelation features of the sequence. Simultaneously, in the moving average (MA) layer, the prediction error from the previous q periods is incorporated for correction, eliminating the impact of random fluctuations on the prediction. The final output is the predicted value for the next f steps, and the corresponding confidence interval can be generated to quantify uncertainty and assist in risk assessment.

[0104] S43. Based on the comparison and analysis of the panoramic operation status results and the predicted trends, identify potential safety over-limit risk points in the system.

[0105] Based on the prediction results and combined with the S3 comprehensive analysis results, potential risk points were identified: Line overload: The predicted power flow exceeds the line's rated capacity.

[0106] Energy storage anomaly: State of charge exceeds the allowable range or charging / discharging power is abnormal.

[0107] Supply and demand imbalance: power generation and load, as well as energy storage charging and discharging, are not in balance.

[0108] S44. Transform the identified potential safety violation risk points into structured early warning information that includes the risk location, type and time window.

[0109] Potential risk points are transformed into visualized or signaled early warning information, including: risk node number, risk type, and prediction time window; and suggested control or optimization measures (such as energy storage scheduling, power generation adjustment, and load management).

[0110] Example 3 is an embodiment of the present invention. This embodiment provides a precise assessment and monitoring system for the panoramic operation status of power system sources, grids, loads and storage, including a data collection module, a mathematical model construction module, a comprehensive analysis module and an output module.

[0111] The data collection module is responsible for collecting and preprocessing real-time operating data of the power system.

[0112] The mathematical model building module is based on preprocessed real-time operating data to establish the first mathematical model and perform dynamic and refined modeling of the system power balance.

[0113] The comprehensive analysis module is based on the first mathematical model and uses the first and second algorithms to comprehensively analyze real-time operating data and output panoramic operating status results.

[0114] The output module uses the ARIMA time series forecasting model and combines it with the panoramic operation status results to predict the future operation trend of the power system, identify potential risk points, and generate corresponding early warning information.

[0115] This embodiment also provides an electronic device applicable to a method for accurate assessment and monitoring of the panoramic operation status of a power system's power generation, grid, load, and storage, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the method for accurate assessment and monitoring of the panoramic operation status of a power system's power generation, grid, load, and storage as proposed in the above embodiment.

[0116] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements a method for accurate assessment and monitoring of the panoramic operation status of power system sources, grids, loads, and storage as proposed in the above embodiments.

[0117] The storage medium proposed in this embodiment belongs to the same inventive concept as the method for accurate assessment and monitoring of the panoramic operation status of power system source, grid, load and storage proposed in the above embodiments. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0118] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0119] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for accurate assessment and monitoring of the panoramic operation status of power system sources, grid, load, and storage, characterized in that: The application relates to a power system risk early warning method and device. Real-time operation data of a power system is collected and preprocessed; Based on the preprocessed real-time operation data, a first mathematical model is established to dynamically and finely model system power balance; Based on the first mathematical model, a first algorithm and a second algorithm are used to comprehensively analyze the real-time operation data, and panoramic operation state results are outputted; An ARIMA time series prediction model is used to predict the future operation trend of the power system, identify potential risk points and generate corresponding early warning information.

2. The method of claim 1, wherein the method is characterized by: The real-time operation data of the power source end, the grid end, the load end and the energy storage end of the power system is collected and preprocessed, including Real-time operation data of a power system is collected and preprocessed; Anomaly value identification and elimination operation is performed on the real-time operation data; The real-time operation data of different sources is unified in terms of data time reference.

3. The method of claim 2, wherein the method is characterized by: Based on the preprocessed real-time operation data, a first mathematical model is established to dynamically and finely model system power balance, including Based on the physical connection relationship of the power grid, the network topology of the first mathematical model is defined; By integrating the core physical constraints of each link of the power system, the constraint system of the first mathematical model is constructed; When constructing the global power balance of the first mathematical model, the line loss dynamic correction parameter and the energy storage charging and discharging efficiency coefficient are introduced to complete the dynamic modeling of the system power balance.

4. The method of claim 3, wherein the method is characterized by: Based on the first mathematical model, a first algorithm and a second algorithm are used to comprehensively analyze the real-time operation data, and panoramic operation state results are outputted, including Map each end real-time operation data to the corresponding node and state parameter of the first mathematical model to establish the association between data and model; Based on the topology and constraints of the first mathematical model, the first algorithm is used to cooperatively process the multi-source heterogeneous data to generate a joint state data matrix; Using the second algorithm, the dynamic operation state of the nodes of the whole system is calculated and updated according to the joint state data matrix and the constraint system of the first mathematical model; The panoramic operation state results of the power system full-link operation index are outputted.

5. The method of claim 4, wherein the method is characterized by: The ARIMA time series prediction model is used to predict the future operation trend of the power system, identify potential risk points and generate corresponding early warning information, including Collect the historical time series data in the first mathematical model and the panoramic operation state results as the input of the ARIMA time series prediction model; Run the ARIMA time series prediction model to output the prediction trend of the key operation parameters of the power system in the future specified time window; Based on the panoramic operation state results and the prediction trend, the potential security risk points in the system are identified; The identified potential security risk points are converted into structured early warning information containing risk position, type and time window.

6. The method of claim 5, wherein the method is characterized by: The line loss dynamic correction parameter is used in the power balance equation of the first mathematical model, and the total active loss of the system is dynamically corrected according to the square relationship between the line current and the resistance; The energy storage charging and discharging efficiency coefficient is a dynamic variable dependent on the real-time operation state of the energy storage unit, which is used in the energy storage state of charge dynamic equation of the first mathematical model, and represents the energy conversion loss in the charging and discharging process.

7. The method of claim 6, wherein the method is characterized by: The first algorithm is a data fusion algorithm based on weighted average, which forms a fused joint state data matrix by assigning weights to real-time operation data from different sources; The second algorithm is a state estimation algorithm, which calculates and updates the dynamic state parameters of each node based on the joint state data matrix and the constraint conditions of the first mathematical model.

8. A power system source network load storage panoramic operation state accurate evaluation and monitoring system, applying a power system source network load storage panoramic operation state accurate evaluation and monitoring method according to any one of claims 1-7, characterized in that, It comprises: a data collection module, a mathematical model construction module, a comprehensive analysis module, and an output module; The data collection module collects real-time operation data of the power system for preprocessing; The mathematical model construction module establishes a first mathematical model based on the preprocessed real-time operation data to dynamically and finely model the system power balance; The comprehensive analysis module performs comprehensive analysis on the real-time operation data based on the first mathematical model, the first algorithm, and the second algorithm, and outputs panoramic operation state results; The output module uses an ARIMA time series prediction model and combines the panoramic operation state results to predict the future operation trend of the power system, identify potential risk points, and generate corresponding warning information. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8 when the computer program is executed by the processor. The processor executes the computer program to realize the steps of the power system source-grid-load-storage panoramic operation state precise evaluation and monitoring method according to any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the steps of the power system source-grid-load-storage panoramic operation state precise evaluation and monitoring method according to any one of claims 1 to 7.

Images