A hybrid measurement based power grid state estimation analysis method and system

Abstract

The application discloses a kind of power grid state estimation analysis method and system based on mixed measurement, belong to power system dispatching automation technical field.Method includes: to the time-space synchronous calibration of multi-source measurement data, generates mixed measurement dataset;According to data precision characteristics, calculate adaptive weight matrix;Based on dataset and weight matrix, construct improved robust mixed measurement state estimation model;Through improved fast convergence algorithm iteration solution, obtain preliminary estimation result;Preliminary result is compared and verified with local high-precision reference data, if error is over threshold value, then modify weight matrix and reiterate, until the final result that meets accuracy requirement is obtained.The application solves the problem of insufficient coverage of single measurement system, weak robustness of traditional algorithm and insufficient dynamic tracking capability, improves the accuracy, real-time performance and robustness of complex power grid operation state estimation.

Description

Technical Field

[0001] This invention relates to a power grid state estimation and analysis method and system based on hybrid measurements, belonging to the field of power system dispatch automation technology. Background Technology

[0002] Power grid state estimation is a core function of energy management systems. By analyzing and processing measurement data, it accurately obtains key state information such as node voltage amplitude, phase angle, and power flow distribution, providing data support for power grid dispatch optimization, fault diagnosis, and power quality control. With the large-scale integration of new energy sources and the advancement of smart grid construction, the operating characteristics of the power grid are becoming increasingly complex, and a single measurement system can no longer meet the requirements for high-precision, real-time state estimation.

[0003] Currently, the measurement equipment widely used in power grids mainly includes synchronous phasor measurement units (PMUs), Supervisory Control and Data Acquisition (SCADA) systems, and advanced measurement systems (AMIs). Among them, PMUs have high precision, high synchronization, and high-frequency refresh characteristics, and can accurately capture dynamic information such as voltage phase angle. However, due to economic cost limitations, their configuration coverage is relatively low. SCADA systems have a wide measurement range and high data redundancy, and can provide measurement data such as power and voltage amplitude. However, they have drawbacks such as long sampling periods, low synchronization accuracy, and large measurement errors. AMI systems can acquire detailed load data on the user side with high accuracy, but their long sampling periods make them difficult to adapt to the dynamic state monitoring needs of the power grid.

[0004] Existing hybrid measurement state estimation techniques mostly focus on the fusion of two types of measurement data, neglecting the synergistic and complementary value of the three types of measurement data, and have several technical bottlenecks: First, the spatiotemporal synchronization of multi-source measurement data is poor, and the sampling time and delay characteristics of different measurement devices vary significantly, which can easily lead to data pollution and affect estimation accuracy. Second, measurement data has uncertainties, including equipment noise, outliers, and data missing issues. Existing algorithms lack robustness and are prone to estimation results deviating from the actual state. Third, the hybrid measurement dimension is high under complex power grids, and traditional estimation algorithms suffer from slow convergence speed and high computational complexity, making it difficult to meet real-time scheduling requirements. Fourth, existing models do not fully consider the bidirectional and fluctuating impact of power flow brought about by distributed energy access, resulting in insufficient estimation adaptability.

[0005] Therefore, there is an urgent need to design a power grid state quantity estimation method and system that can achieve accurate fusion of multi-source measurement data, and has high estimation accuracy, fast convergence speed and strong robustness, so as to meet the operation monitoring needs of complex power grids. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and provide a power grid state estimation and analysis method and system based on hybrid measurement. This solves the problems of insufficient coverage of a single measurement system, weak robustness of traditional algorithms, and insufficient dynamic tracking capability, thereby improving the accuracy, real-time performance, and robustness of complex power grid operating state estimation.

[0007] To achieve the above objectives, the present invention is implemented using the following technical solution:

[0008] In a first aspect, the present invention provides a power grid state estimation and analysis method based on hybrid measurements, comprising:

[0009] The acquired multi-source measurement data are subjected to time synchronization calibration and spatial synchronization calibration to generate a spatiotemporally unified hybrid measurement dataset;

[0010] Based on the accuracy characteristics and real-time status of each data point in the spatiotemporally unified hybrid measurement dataset, an adaptive weight matrix is ​​calculated.

[0011] Based on the spatiotemporally unified hybrid measurement dataset and the adaptive weight matrix, an improved robust hybrid measurement state estimation model is constructed.

[0012] The improved robust hybrid measurement state estimation model is iteratively solved by an improved fast convergence algorithm to obtain preliminary estimation results of power grid state variables.

[0013] The preliminary estimation results of the power grid state variables are compared and verified with the preset local high-precision reference data. If the error exceeds the preset threshold, the adaptive weight matrix is ​​corrected and the iteration is repeated until the final power grid state estimation result that meets the accuracy requirements is obtained.

[0014] Furthermore, the method for acquiring the multi-source measurement data includes:

[0015] Acquire PMU measurement data from synchronous phasor measurement devices, SCADA measurement data from data acquisition and monitoring systems, and AMI measurement data from advanced measurement systems;

[0016] The PMU measurement data includes node voltage phasors and branch current phasors; the SCADA measurement data includes node injected power, branch power flow, and node voltage amplitude; and the AMI measurement data includes user-side load power.

[0017] Furthermore, before performing spatiotemporal synchronization calibration on the multi-source measurement data, preprocessing of the acquired multi-source measurement data is also included, specifically:

[0018] The 3σ criterion is used to identify and remove outliers in various measurement data.

[0019] To fill the data gaps after outliers were removed, the historical average interpolation method was used.

[0020] Furthermore, the step of performing time synchronization calibration and spatial synchronization calibration on the acquired multi-source measurement data to generate a spatiotemporally unified hybrid measurement dataset includes:

[0021] The timescale of the PMU measurement data Using time as a reference, a delay compensation model is constructed to calculate the SCADA data calibration time. And extract from the preprocessed SCADA dataset by interpolation SCADA data at any given time, to obtain a SCADA time synchronization dataset synchronized with the PMU time profile;

[0022] An improved linear extrapolation interpolation model is constructed to extract interpolation values ​​from the preprocessed AMI dataset. The improved linear extrapolation interpolation model is as follows: AMI time synchronization dataset is obtained by matching AMI data at specific times. ;

[0023] In the formula, For AMI raw data at time The value, For PMU data at time The value, For interpolation Real-time AMI data;

[0024] Based on the power grid topology, a spatial synchronization correction model is constructed to correct power loss in the branch power flow data in the SCADA time synchronization dataset and the AMI time synchronization dataset.

[0025] The power loss-corrected SCADA time synchronization dataset and AMI time synchronization dataset are integrated with the PMU measurement data to generate a spatiotemporally unified hybrid measurement dataset.

[0026] Furthermore, based on the accuracy characteristics and real-time status of each data point in the spatiotemporally unified hybrid measurement dataset, an adaptive weight matrix is ​​calculated, including:

[0027] Based on the variance of various measurement data Calculate the base weights The basic weights are inversely proportional to the variance; a dynamic correction factor is constructed. ,in This refers to the trend deviation of data predicted based on historical trends. This refers to the equipment operating status coefficient;

[0028] Multiplying the base weights by the dynamic correction factor yields the dynamic weights. And suppress the weight of suspected abnormal data;

[0029] Construct a diagonal weight matrix based on the dynamic weights of all measurement data. .

[0030] Furthermore, the construction of the improved robust hybrid measurement state estimation model includes:

[0031] Based on node voltage amplitude and phase angle State variables Construct measurement equations that include three types of measurement data: PMU, SCADA, and AMI. ,in, For measurement functions, This is due to measurement error;

[0032] Introducing a robustness factor, a weighted least squares objective function combining robustness penalty and robustness penalty is constructed. The objective function is as follows: ;

[0033] in, This is the robustness adjustment coefficient. This is a robust penalty function using the Huber function;

[0034] Introduce probabilistic constraints to modify the measurement function. The node injection power term is defined in the following probabilistic constraints: ,in, For the output fluctuation value of distributed energy resources, To allow the maximum fluctuation value, Confidence level;

[0035] The output is an improved robust hybrid measurement state estimation model that includes the objective function, measurement equations, and probabilistic constraints.

[0036] Furthermore, the step of iteratively solving the improved robust hybrid measurement state estimation model using an improved fast convergence algorithm to obtain preliminary estimation results of the power grid state variables includes:

[0037] In the iterative solution of the improved robust hybrid measurement state estimation model, the measurement function... Perform linearization and calculate the Jacobian matrix. ;

[0038] The Jacobian matrix is ​​divided into blocks for decoupling, weakly coupled blocks are ignored, and a compensation decoupling matrix is ​​introduced. After correcting the decoupling error, the simplified Jacobian matrix is ​​obtained. ;

[0039] Introducing an adaptive step size factor The iteration step size is dynamically adjusted according to the residual change trend, and the state variable correction amount is updated.

[0040] Based on the objective function value of the current iteration Compared with the objective function value of the previous iteration Based on the comparison results, the adaptive step size factor is dynamically adjusted. :

[0041] like Then take ;

[0042] like Then take ;

[0043] in, and They are respectively Time and The state variable at any given time;

[0044] Based on the simplified Jacobian matrix The diagonal weight matrix and the adaptive step size factor Calculate the state variable correction amount The formula is as follows:

[0045] ;

[0046] The improved robust mixed measurement state estimation model is solved iteratively until the objective function value of the current iteration is reached. Compared with the objective function value of the previous iteration If the absolute value of the difference is less than the preset accuracy threshold, the iteration stops and the current state variable is output as the preliminary estimation result of the power grid state variables.

[0047] Furthermore, the preliminary estimation results of the power grid state variables are compared and verified with preset local high-precision reference data, including:

[0048] Using PMU measurement data as local high-precision reference data, the node voltage amplitude error, node voltage phase angle error, and branch power flow error are calculated.

[0049] If any error index exceeds the preset threshold, the dynamic correction factor and robustness adjustment coefficient in the adaptive weight matrix are adjusted, and the iterative solution is performed again.

[0050] If all error indicators are less than the preset threshold, the current result is determined to be the final power grid state estimation result and output.

[0051] Secondly, the present invention provides a power grid state estimation and analysis system based on hybrid measurements, used to implement the power grid state estimation and analysis method based on hybrid measurements described in any one of the preceding claims, comprising:

[0052] The calibration module is used to perform time synchronization calibration and spatial synchronization calibration on the acquired multi-source measurement data to generate a spatiotemporally unified hybrid measurement dataset.

[0053] The calculation module is used to calculate the adaptive weight matrix based on the accuracy characteristics and real-time status of each data in the spatiotemporally unified hybrid measurement dataset.

[0054] The model building module is used to build an improved robust hybrid measurement state estimation model based on the spatiotemporally unified hybrid measurement dataset and the adaptive weight matrix.

[0055] The iterative solution module is used to iteratively solve the improved robust hybrid measurement state estimation model through an improved fast convergence algorithm to obtain preliminary estimation results of power grid state variables.

[0056] The comparison and verification module is used to compare and verify the preliminary estimation results of the power grid state variables with the preset local high-precision reference data. If the error exceeds the preset threshold, the module is returned to correct the adaptive weight matrix and iteratively solve the problem again until the final power grid state estimation result that meets the accuracy requirements is obtained.

[0057] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0058] Fourthly, the present invention provides an electronic device, comprising:

[0059] Memory, used to store computer programs / instructions;

[0060] A processor for executing the computer program / instructions to implement the steps of any of the methods described above.

[0061] Fifthly, the present invention provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of any of the methods described above.

[0062] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0063] This invention provides a power grid state estimation and analysis method and system based on hybrid measurements. By fusing local high-precision reference data from the synchronous phasor measurement unit (PMU) with broad-spectrum measurement data from the Supervisory Control and Data Acquisition (SCADA) system, and combining spatiotemporal correlation pseudo-measurement construction technology with an improved adaptive robust estimation algorithm, the invention achieves accurate estimation of power grid state quantities. This invention solves the problems of insufficient coverage of a single measurement system, weak robustness of traditional algorithms, and insufficient dynamic tracking capability, and improves the accuracy, real-time performance, and robustness of complex power grid operating state estimation. Attached Figure Description

[0064] Figure 1 This is a flowchart of a power grid state estimation and analysis method based on hybrid measurements provided in an embodiment of the present invention. Detailed Implementation

[0065] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0066] Example 1: This example introduces a power grid state estimation and analysis method based on hybrid measurements, including:

[0067] The acquired multi-source measurement data are subjected to time synchronization calibration and spatial synchronization calibration to generate a spatiotemporally unified hybrid measurement dataset;

[0068] Based on the accuracy characteristics and real-time status of each data point in the spatiotemporally unified hybrid measurement dataset, an adaptive weight matrix is ​​calculated.

[0069] Based on the spatiotemporally unified hybrid measurement dataset and the adaptive weight matrix, an improved robust hybrid measurement state estimation model is constructed.

[0070] The improved robust hybrid measurement state estimation model is iteratively solved by an improved fast convergence algorithm to obtain preliminary estimation results of power grid state variables.

[0071] The preliminary estimation results of the power grid state variables are compared and verified with the preset local high-precision reference data. If the error exceeds the preset threshold, the adaptive weight matrix is ​​corrected and the iteration is repeated until the final power grid state estimation result that meets the accuracy requirements is obtained.

[0072] like Figure 1 As shown in the figure, the power grid state estimation and analysis method based on hybrid measurements provided in this embodiment involves the following steps in its application process:

[0073] Step 1 (Multi-source measurement data acquisition and preprocessing) Specific implementation techniques:

[0074] Power grid measurement data were collected using three types of measurement devices: PMU, SCADA, and AMI. PMU data included node voltage phasors and branch current phasors; SCADA data included node active / reactive power injection, branch active / reactive power flow, and node voltage amplitude; and AMI data included user-side load active / reactive power data. The collected data underwent preprocessing to remove obvious outliers and fill in missing data, laying the foundation for subsequent data fusion.

[0075] (1) Collect multi-source measurement data by category. The process is as follows:

[0076] PMU (Synchronous Phasor Measurement Unit): Acquires node voltage phasors. Branch current phasors sampling frequency Output the original dataset Where k is the sampling time, This represents the number of sampling points.

[0077] SCADA (Supervisory Control and Data Acquisition) system: Active / Reactive power injected into the acquisition node. Branch road contributing / not contributing to the current Node voltage amplitude sampling frequency Output the original dataset Where m is the sampling time, This represents the number of sampling points.

[0078] AMI (Advanced Metering System): Collects active / reactive power data from the user side load. sampling frequency Output the original dataset Where n is the sampling time, This represents the number of sampling points.

[0079] (2) Data preprocessing, the process is as follows:

[0080] Outlier removal: The 3σ criterion is used to identify outlier data. For each data category, the mean μ and standard deviation σ are calculated for each indicator. If a data point x satisfies |x−μ|>3σ, it is considered an outlier and removed. Missing data imputation.

[0081] Missing data imputation: To fill data gaps after outlier removal, historical mean imputation is used. For the missing data x(t) at time t, k sets of data from the same time period over the past 7 days are selected. Calculate the mean As filler values.

[0082] Output: Preprocessed qualified dataset .

[0083] Step Two (Spatiotemporal Synchronization Calibration of Hybrid Measurement Data): Specific Implementation Techniques

[0084] The purpose of this step is to eliminate the differences in time scale and spatial location among the three types of data, and to generate a spatiotemporally unified measurement dataset.

[0085] (1) Time synchronization calibration, the process is as follows:

[0086] 1) Selection of reference time scale: based on High-precision time scale in The time reference is provided by the PMU's GPS synchronization module, with a time error ≤1 μs;

[0087] 2) SCADA data time calibration:

[0088] Model Construction: Let the expected SCADA measurement delay be... Then the first The SCADA calibration time corresponding to each PMU time is:

[0089] ;

[0090] Data matching: From Interpolation extraction SCADA data at specific times is used to generate a SCADA dataset aligned with PMU time. .

[0091] 3) AMI data time interpolation completion:

[0092] Model Construction: An improved linear extrapolation method was used, combined with the dynamic trends of PMU high-frequency data, to establish an AMI data interpolation model.

[0093] ;

[0094] In the formula, For AMI raw data at time The value, For PMU data at time The value, For interpolation AMI data at any given time.

[0095] Output: An AMI dataset that perfectly matches the PMU timeline. .

[0096] (2) Spatial synchronization calibration, the process is as follows:

[0097] Model Construction: Based on the power grid topology, a spatially synchronized correction model is established to eliminate deviations caused by differences in the installation locations of measurement equipment. For branch power flow measurement, the correction formula is:

[0098] ;

[0099] ;

[0100] In the formula, , This refers to the branch power flow data after time synchronization. , This is the correction value for branch power loss, derived from the branch resistance. Reactance With current The calculation yielded: .

[0101] Output: A unified measurement dataset after spatiotemporal synchronization ,in , This is the spatially synchronized SCADA and AMI dataset.

[0102] Step 3 (Adaptive Weight Allocation of Measurement Data) Specific Implementation Techniques:

[0103] Based on data accuracy, real-time performance, and equipment status, dynamic weights are assigned to the three types of measurement data, and a weight matrix is ​​constructed.

[0104] (1) Calculation of basic weights, the process is as follows:

[0105] Model construction: For the spatiotemporally synchronized measurement data, calculate the variance of each type of measurement. (m represents PMU / SCADA / AMI), the basic weights are inversely proportional to the variance, as shown in the formula:

[0106] ;

[0107] In the formula, ,therefore .

[0108] (2) Dynamic weight adjustment, the process is as follows:

[0109] 1) Correction factor design;

[0110] Introducing dynamic correction factors Taking into account the deviation of data trends With equipment operating status coefficient The formula is as follows:

[0111] ;

[0112] in:

[0113] , This is a historical trend prediction value; the greater the deviation, The smaller;

[0114] Equipment condition coefficient : Assign values ​​based on the device's operating status, such as when the PMU synchronization status is normal. Lost time 4 hours When SCADA communication quality is good When the packet loss rate is higher than 10% ;

[0115] 2) Dynamic weight calculation;

[0116] ;

[0117] Outlier weight suppression: If a data point is identified as a suspected outlier (residual pre-judgment), its weight is adjusted to... .

[0118] (3) Construction of the weight matrix, the process is as follows:

[0119] Construct a diagonal weight matrix Matrix dimension and mixed measurement vector Consistent, the diagonal elements are the dynamic weights of each measurement data. Off-diagonal elements are 0:

[0120] ;

[0121] in: The dimension of the hybrid measurement vector. The dynamic weights correspond to the PMU, SCADA, and AMI measurement data, respectively.

[0122] Output: Adaptive weight matrix , which serves as the core parameter input for the robust state estimation model.

[0123] Step 4 (Constructing an improved robust hybrid measurement state estimation model): Specific implementation techniques and methods.

[0124] A state estimation model incorporating three types of measurement data is established, and robustness factors and probability constraints are introduced to improve the model's robustness.

[0125] (1) The state variables and measurement vectors are defined as follows:

[0126] 1) State variables ;

[0127] Based on the voltage amplitude of the power grid nodes and phase angle For state variables, , where n is the number of power grid nodes.

[0128] 2) Hybrid Measurement Vector ;

[0129] Integrating the three types of measurement data after spatiotemporal synchronization, ,in:

[0130] PMU measurement vector, which includes the magnitude and phase angle of node voltage phasors and branch current phasors;

[0131] SCADA measurement vectors, including node injected power, branch power flow, and voltage amplitude;

[0132] : AMI measurement vector, which includes user-side load power.

[0133] (2) The measurement equation is constructed as follows:

[0134] The basic form of the measurement equation is:

[0135] ;

[0136] In the formula:

[0137] : Measurement function vector, which represents the state variables The nonlinear function is represented in blocks as follows: Each block function corresponds to the physical mapping relationship of the three types of measurements;

[0138] The measurement error vector follows a constant with a mean of zero and a covariance matrix of... The Gaussian distribution, i.e. W is the adaptive weight matrix output in step 3.

[0139] (3) Improve the construction of the robust objective function, the process is as follows:

[0140] Introducing a robustness factor, we construct an objective function that combines weighted least squares with robustness penalty:

[0141] ;

[0142] In the formula: This is the robustness adjustment coefficient, ranging from [0.1, 1], adaptively adjusted by the measurement anomaly rate (the higher the anomaly rate, the better). The larger);

[0143] The robustness penalty function, using the Huber function, is in the form of... ,in To measure the residual, The residual threshold is determined by the measurement accuracy level.

[0144] (4) Introduction of probabilistic constraints for distributed energy resources, the process is as follows:

[0145] To address the uncertainty in output from distributed energy sources such as wind and solar power, probabilistic constraints are introduced:

[0146] ;

[0147] in, For the output fluctuation value of distributed energy resources, To allow the maximum fluctuation value, The confidence level is set to 0.05.

[0148] This constraint quantifies the impact of power output fluctuations on state estimation by correcting the nodal injection power term in the measurement function H(X).

[0149] Output: An improved robust hybrid measurement state estimation model (including objective function, measurement equation, and probability constraints) is used as the model input for the algorithm in step five.

[0150] Step 5 (Improved Fast Convergence Algorithm Solution): Specific Implementation Techniques

[0151] Efficient solutions are provided for nonlinear state estimation models to quickly obtain power grid state variable estimation results.

[0152] (1) Model linearization process, as follows:

[0153] Using the Gauss-Newton method, the nonlinear measurement function H(X) is iterated with initial values... Performing a Taylor expansion and ignoring higher-order terms, we obtain the linearized measurement equation:

[0154] ;

[0155] In the formula:

[0156] : No. The state variable values ​​for the next iteration;

[0157] : State variable correction amount;

[0158] : No. The Jacobian matrix of the next iteration, where the matrix elements are the partial derivatives of the measurement function with respect to the state variables;

[0159] (2) Jacobian matrix block decoupling, the process is as follows:

[0160] Decoupling strategy: Based on the characteristics that active power P is mainly related to the voltage phase angle θ and reactive power Q is mainly related to the voltage amplitude V, the Jacobian matrix is... The blocks are as follows:

[0161] ;

[0162] Ignore weakly coupled blocks and Construct a simplified Jacobian matrix At the same time, a compensation decoupling matrix is ​​introduced. After correcting the decoupling error, the final result is: Its function is to decompose the solution of a high-dimensional matrix into the solution of a low-dimensional submatrix, thereby reducing computational complexity.

[0163] (3) Adaptive iterative step size adjustment, the process is as follows:

[0164] Introducing an adaptive step size factor Dynamically adjust the step size based on the residual change trend:

[0165] If residuals This indicates that the iteration direction is correct. ;

[0166] If residuals This indicates that the step size is too large; take... ;

[0167] in, and They are respectively Time and The state variable at any given time;

[0168] The formula for updating the correction amount is:

[0169] ;

[0170] (4) Iterative convergence determination, the process is as follows:

[0171] The iteration terminates when the residual meets the set precision threshold. :

[0172] ;

[0173] In the formula, Usually taken .

[0174] Output: Estimation results of power grid state variables that meet convergence accuracy. Input data used for result verification and correction.

[0175] Step Six (Estimation Result Verification and Correction): Specific Implementation Techniques

[0176] (1) Error calculation and accuracy verification, the process is as follows:

[0177] 1) Selecting a validation benchmark: Using PMU high-precision measurement data as the benchmark (error ≤ 0.1%), and comparing it with historical state estimation results from the same period.

[0178] 2) Calculate the critical error:

[0179] Indicator node voltage amplitude error: ;

[0180] Node voltage phase angle error: ;

[0181] Branch flow error: ;

[0182] 3) Accuracy judgment: If all error indicators are less than the preset threshold, the judgment result is qualified; otherwise, it is judged as unqualified.

[0183] (2) Closed-loop verification, the process is as follows:

[0184] 1) If the result is unsatisfactory, return to step three and adjust the dynamic correction factor of the weight matrix W. With robustness adjustment factor ;

[0185] 2) Repeat steps 3 through 4, 5, and 6 until the error index meets the threshold requirement.

[0186] Final output: High-precision power grid state estimation results .

[0187] Example 2: This example provides a power grid state estimation and analysis system based on hybrid measurements, including:

[0188] The calibration module is used to perform time synchronization calibration and spatial synchronization calibration on the acquired multi-source measurement data to generate a spatiotemporally unified hybrid measurement dataset.

[0189] The calculation module is used to calculate the adaptive weight matrix based on the accuracy characteristics and real-time status of each data in the spatiotemporally unified hybrid measurement dataset.

[0190] The model building module is used to build an improved robust hybrid measurement state estimation model based on the spatiotemporally unified hybrid measurement dataset and the adaptive weight matrix.

[0191] The iterative solution module is used to iteratively solve the improved robust hybrid measurement state estimation model through an improved fast convergence algorithm to obtain preliminary estimation results of power grid state variables.

[0192] The comparison and verification module is used to compare and verify the preliminary estimation results of the power grid state variables with the preset local high-precision reference data. If the error exceeds the preset threshold, the module is returned to correct the adaptive weight matrix and iteratively solve the problem again until the final power grid state estimation result that meets the accuracy requirements is obtained.

[0193] The specific functions of each module described above are explained in the relevant content of the method in Embodiment 1, and will not be repeated here.

[0194] Example 3: This example provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described in Example 1.

[0195] Example 4: This example provides an electronic device, including:

[0196] Memory, used to store computer programs / instructions;

[0197] A processor for executing the computer program / instructions to implement the steps of any of the methods described in Embodiment 1.

[0198] Example 5: This example provides a computer program product, including a computer program / instructions, which, when executed by a processor, implement the steps of the method described in any one of Examples 1.

[0199] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

[0200] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, systems, or computer program products. Therefore, this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this disclosure can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0201] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0202] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0203] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0204] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this disclosure and not to limit its protection scope. Although this disclosure has been described in detail with reference to the above embodiments, those skilled in the art should understand that after reading this disclosure, they can still make various changes, modifications or equivalent substitutions to the specific implementation of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the pending claims.

Claims

1. A power grid state estimation and analysis method based on hybrid measurements, characterized in that, include: The acquired multi-source measurement data are subjected to time synchronization calibration and spatial synchronization calibration to generate a spatiotemporally unified hybrid measurement dataset; Based on the accuracy characteristics and real-time status of each data point in the spatiotemporally unified hybrid measurement dataset, an adaptive weight matrix is ​​calculated. Based on the spatiotemporally unified hybrid measurement dataset and the adaptive weight matrix, an improved robust hybrid measurement state estimation model is constructed. The improved robust hybrid measurement state estimation model is iteratively solved by an improved fast convergence algorithm to obtain preliminary estimation results of power grid state variables. The preliminary estimation results of the power grid state variables are compared and verified with the preset local high-precision reference data. If the error exceeds the preset threshold, the adaptive weight matrix is ​​corrected and the iteration is repeated until the final power grid state estimation result that meets the accuracy requirements is obtained.

2. The power grid state estimation and analysis method based on hybrid measurements according to claim 1, characterized in that, The method for acquiring multi-source measurement data includes: Acquire PMU measurement data from synchronous phasor measurement devices, SCADA measurement data from data acquisition and monitoring systems, and AMI measurement data from advanced measurement systems; The PMU measurement data includes node voltage phasors and branch current phasors; the SCADA measurement data includes node injected power, branch power flow, and node voltage amplitude; and the AMI measurement data includes user-side load power.

3. The power grid state estimation and analysis method based on hybrid measurements according to claim 1, characterized in that, Before performing spatiotemporal synchronization calibration on the multi-source measurement data, preprocessing of the acquired multi-source measurement data is also included, specifically: The 3σ criterion is used to identify and remove outliers in various measurement data. To fill the data gaps after outliers were removed, the historical average interpolation method was used.

4. The power grid state estimation and analysis method based on hybrid measurements according to claim 2, characterized in that, The process of performing time synchronization calibration and spatial synchronization calibration on the acquired multi-source measurement data to generate a spatiotemporally unified hybrid measurement dataset includes: The timescale of the PMU measurement data Using time as a reference, a delay compensation model is constructed to calculate the SCADA data calibration time. And extract from the preprocessed SCADA dataset by interpolation SCADA data at any given time, to obtain a SCADA time synchronization dataset synchronized with the PMU time profile; An improved linear extrapolation interpolation model is constructed to extract interpolation values ​​from the preprocessed AMI dataset. The improved linear extrapolation interpolation model is as follows: AMI time synchronization dataset is obtained by matching AMI data at specific times. ; In the formula, For AMI raw data at time The value, For PMU data at time The value, For interpolation AMI data at any given time; Based on the power grid topology, a spatial synchronization correction model is constructed to correct power loss in the branch power flow data in the SCADA time synchronization dataset and the AMI time synchronization dataset. The power loss-corrected SCADA time synchronization dataset and AMI time synchronization dataset are integrated with the PMU measurement data to generate a spatiotemporally unified hybrid measurement dataset.

5. The power grid state estimation and analysis method based on hybrid measurement according to claim 1, characterized in that, Based on the accuracy characteristics and real-time status of each data point in the spatiotemporally unified hybrid measurement dataset, an adaptive weight matrix is ​​calculated, including: Based on the variance of various measurement data Calculate the base weights The basic weights are inversely proportional to the variance; a dynamic correction factor is constructed. ,in This refers to the trend deviation of data predicted based on historical trends. This refers to the equipment operating status coefficient; Multiplying the base weights by the dynamic correction factor yields the dynamic weights. And suppress the weight of suspected abnormal data; Construct a diagonal weight matrix based on the dynamic weights of all measurement data. .

6. The power grid state estimation and analysis method based on hybrid measurements according to claim 5, characterized in that, The construction of the improved robust hybrid measurement state estimation model includes: Based on node voltage amplitude and phase angle State variables Construct measurement equations that include three types of measurement data: PMU, SCADA, and AMI. ,in, For measurement functions, This is due to measurement error; Introducing a robustness factor, a weighted least squares objective function combining robustness penalty and robustness penalty is constructed. The objective function is as follows: ; in, This is the robustness adjustment coefficient. This is a robust penalty function using the Huber function; Introduce probabilistic constraints to modify the measurement function. The node injection power term is defined in the following probabilistic constraints: ,in, For the output fluctuation value of distributed energy resources, To allow the maximum fluctuation value, Confidence level; The output is an improved robust hybrid measurement state estimation model that includes the objective function, measurement equations, and probabilistic constraints.

7. The power grid state estimation and analysis method based on hybrid measurements according to claim 6, characterized in that, The step of iteratively solving the improved robust hybrid measurement state estimation model using an improved fast convergence algorithm to obtain preliminary estimation results of power grid state variables includes: In the iterative solution of the improved robust hybrid measurement state estimation model, the measurement function... Perform linearization and calculate the Jacobian matrix. , Sampling time; The Jacobian matrix is ​​divided into blocks for decoupling, weakly coupled blocks are ignored, and a compensation decoupling matrix is ​​introduced. After correcting the decoupling error, the simplified Jacobian matrix is ​​obtained. ; Introducing an adaptive step size factor The iteration step size is dynamically adjusted according to the residual change trend, and the state variable correction amount is updated. Based on the objective function value of the current iteration Compared with the objective function value of the previous iteration Based on the comparison results, the adaptive step size factor is dynamically adjusted. : like Then take ; like Then take ; in, and They are respectively Time and The state variable at any given time; Based on the simplified Jacobian matrix The diagonal weight matrix and the adaptive step size factor Calculate the state variable correction amount The formula is as follows: ； The improved robust mixed measurement state estimation model is solved iteratively until the objective function value of the current iteration is reached. Compared with the objective function value of the previous iteration If the absolute value of the difference is less than the preset accuracy threshold, the iteration stops and the current state variable is output as the preliminary estimation result of the power grid state variables.

8. The power grid state estimation and analysis method based on hybrid measurements according to claim 1, characterized in that, The preliminary estimation results of the power grid state variables are compared and verified with preset local high-precision reference data, including: Using PMU measurement data as local high-precision reference data, the node voltage amplitude error, node voltage phase angle error, and branch power flow error are calculated. If any error index exceeds the preset threshold, the dynamic correction factor and robustness adjustment coefficient in the adaptive weight matrix are adjusted, and the iterative solution is performed again. If all error indicators are less than the preset threshold, the current result is determined to be the final power grid state estimation result and output.

9. A power grid state estimation and analysis system based on hybrid measurements, used to implement the power grid state estimation and analysis method based on hybrid measurements as described in any one of claims 1-8, characterized in that, include: The calibration module is used to perform time synchronization calibration and spatial synchronization calibration on the acquired multi-source measurement data to generate a spatiotemporally unified hybrid measurement dataset. The calculation module is used to calculate the adaptive weight matrix based on the accuracy characteristics and real-time status of each data in the spatiotemporally unified hybrid measurement dataset. The model building module is used to build an improved robust hybrid measurement state estimation model based on the spatiotemporally unified hybrid measurement dataset and the adaptive weight matrix. The iterative solution module is used to iteratively solve the improved robust hybrid measurement state estimation model through an improved fast convergence algorithm to obtain preliminary estimation results of power grid state variables. The comparison and verification module is used to compare and verify the preliminary estimation results of the power grid state variables with the preset local high-precision reference data. If the error exceeds the preset threshold, the module is returned to correct the adaptive weight matrix and iteratively solve the problem again until the final power grid state estimation result that meets the accuracy requirements is obtained.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When executed by a processor, the computer program implements the steps of the method described in any one of claims 1-6.

Images