Power distribution network state estimation method based on multi-time scale hybrid measurement

Abstract

The application discloses a power distribution network state estimation method based on multi-time scale mixed measurement, and the method comprises the following steps: a multi-objective optimization function containing measurement point configuration and partition scale is constructed, D-PMU and SCADA measurement points are taken as boundary nodes, a power distribution network is divided into multiple sub-areas, the multiple sub-areas exchange topological information through the boundary nodes, the topologies of adjacent sub-areas are equivalent and condensed to the boundary nodes, the equivalent injection power of the boundary nodes is calculated, an interval dynamic state estimation model is independently constructed for each sub-area, the dynamic state estimation model of each sub-area is solved, the estimation results of each sub-area are aggregated, dynamic state estimation of the whole network is carried out, a multi-source measurement data fusion strategy considering non-synchronous measurement deviation is adopted, and the influence of non-synchronization of multi-source measurement sampling time on dynamic state estimation precision is reduced; when the topology changes, only the dynamic state estimation model of the sub-area with the changed topology needs to be reconstructed, and the calculation efficiency of dynamic state estimation is improved.

Description

Technical Field

[0001] This invention relates to the field of power distribution network technology, and more specifically to a power distribution network state estimation method based on multi-timescale hybrid measurements. Background Technology

[0002] The widespread integration of distributed generation has transformed distribution networks from traditional unidirectional power supply networks into active distribution networks with bidirectional power flow. This transformation places higher demands on the state estimation, operation control, and regional division of distribution networks.

[0003] Existing distribution network estimation methods can guarantee accuracy to a certain extent in the first few state estimations between two AMI measurements. However, the accuracy of the pseudo-measurement data of AMI calculated by this method gradually decreases thereafter, leading to a reduction in the accuracy of state estimation. When the operation mode of the distribution network changes, resulting in changes in the local topology, the dynamic state estimation model of the entire network needs to be reconstructed, which will greatly affect the computational efficiency of state estimation. Since there are nonlinear power measurements in the measurement data, the iterative solution of the Jacobian matrix in the recursive process will introduce a large number of nonlinear interval operations, exacerbating the expansion of the solution interval, and thus making the state estimation results unusable. Summary of the Invention

[0004] The proposed method for power distribution network state estimation based on multi-timescale hybrid measurement addresses the technical problems of existing channel estimation methods, which suffer from rapidly increasing pilot overhead and computational complexity, hindering practical system deployment, resulting in low channel estimation accuracy and poor convergence, numerical instability caused by manual parameter tuning and the use of excessively large penalty parameters, and low adaptability to different channel environments.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: The distribution network state estimation method based on multi-time-scale hybrid measurement of the present invention includes the following steps: S1. Construct a multi-source measurement dataset including D-PMU measurement points, SCADA measurement points, and AMI measurement points; S2. Construct a multi-objective optimization function that includes measurement point configuration and partition size, and divide the power grid into multiple sub-regions with the D-PMU and the SCADA measurement points as boundary nodes. S3. The multiple sub-regions interact with topological information through boundary nodes, condense the topological equivalents of adjacent sub-regions to the boundary nodes, calculate the equivalent injection power of the boundary nodes, and each sub-region independently constructs an interval dynamic state estimation model. S4. Solve the dynamic state estimation model of each sub-region, and perform dynamic state estimation of the entire network by aggregating the estimation results of each sub-region. S5. Calculate the evaluation index of the dynamic state estimation of the whole network, verify the interval estimation results, obtain the evaluation results, and optimize the parameters of the dynamic state estimation model based on the evaluation results.

[0006] Preferably, step S1 includes the following steps: S11 and D-PMU use D-PMU measurement data at the state estimation time to perform state estimation, and select SCADA measurement data closest to the state estimation time for state estimation. S12. Set a time window to determine whether the AMI measurement data is updated at the state estimation time; S13. If the state estimation time has been updated, then the updated AMI measurement data shall be used for state estimation. S14. If the state is not updated at the state estimation time, use the predicted value of the state quantity to calculate the pseudo-measurement data of the AMI node for state estimation, and construct a robust factor to adjust the weight of the AMI pseudo-measurement in the state estimation. S15. Calculate the asynchronous measurement deviation caused by the asynchronous sampling time.

[0007] Preferably, step S3 includes the following steps: S31. Each sub-region exchanges topological information through boundary nodes, and the topological condensation of adjacent sub-regions is achieved through the equivalent admittance matrix. S32. Calculate the equivalent injected power of the boundary node by measuring the power of the boundary node and the node voltage in the sub-region. S33. Interval numbers are used to characterize the asynchronous measurement bias of multi-source measurement data, and an interval dynamic state estimation model for multi-source measurement fusion is constructed.

[0008] Preferably, step S4 includes the following steps: S41. State Prediction: Calculate the state prediction quantity and the prediction error covariance matrix, and transform the prediction error covariance matrix into an upper bound matrix by combining the upper bound optimization idea. S42. Measurement Transformation: Converts power measurement into equivalent current measurement; S43. The filtering calculation is completed through the upper bound matrix to obtain the state estimation results of each sub-region; S44. Aggregate the interval state estimation results of each sub-region to form the dynamic state estimation output of the entire network.

[0009] Preferably, step S5 includes the following steps: S51. Calculate the evaluation index for the dynamic state estimation of the entire network using the evaluation index calculation formula; S52. Based on the evaluation results, optimize the multi-source measurement fusion strategy and dynamic state estimation model parameters in reverse.

[0010] Preferably, the calculation formula for the AMI measurement data is as follows:

[0011] in, For AMI in Measurement data at time, () represents the time series of AMI measurement data. (·) represents the measurement function of AMI. In order to be in The most recently updated sampling time in AMI before the current time. For time window, for State prediction at any given time; The robustness factor is defined as:

[0012] in, The robustness factor This is the robustness factor correction function; The formula for calculating the asynchronous measurement deviation is as follows:

[0013] in, This is due to asynchronous measurement deviation; The average rate of change of the measured quantity over the interval between the sampling time and the state estimation time is given by the metric. This represents the time deviation between the sampling time and the state estimation time.

[0014] Preferably, the equivalent admittance matrix is ​​as follows:

[0015] In this context, subscript A represents an internal node, and subscript B represents a boundary node; The formula for calculating the equivalent injected power of the boundary node is as follows:

[0016] in, as well as Boundary nodes Equivalent injection power, Boundary nodes The set of line measurements connected to nodes within the sub-region. After equivalence to Ward and boundary nodes A set of connected equivalent lines, , For boundary nodes Equivalent admittance of connected branches, , For sub-regions and boundary nodes Power measurements and predicted status values ​​of connected lines. , They are respectively Time Node The real and imaginary parts of the state prediction quantity , They are respectively Time Node The real and imaginary parts of the state prediction quantity; The expression for the interval dynamic state estimation model is as follows:

[0017] in, For interval state variables, For interval measurement, for n System process noise, for m Dimensional measurement noise, This is the state transition function. This is the measurement function.

[0018] Preferably, the formulas for calculating the state prediction quantity and the prediction error covariance matrix are as follows:

[0019] in, for State prediction at time t, for The state estimate at time t, where F is the state transition matrix. for The covariance matrix of the prediction error at any given time. for Time-estimation error covariance matrix for The covariance matrix;

[0020] in, for The Jacobian matrix of the time measurement function, for The gain matrix at time step I, where I is the identity matrix. for The measurement error matrix.

[0021] Preferably, the expression for the upper bound matrix is ​​as follows:

[0022] in, This is the upper bound matrix of the interval prediction error covariance matrix. Parameters can be derived from a matrix The Frobenius norm is obtained from the matrix. elements The definition is as follows:

[0023] in, To determine the center value of the prediction error covariance matrix, This is an augmented matrix of the prediction error covariance matrix; The filtering calculation using the upper bound matrix is ​​as follows:

[0024] in, is the Jacobian matrix after measurement transformation, and is a constant matrix; This is the measurement error matrix after measurement transformation; =1 / a k .

[0025] Preferably, the evaluation index is calculated using the following formula:

[0026]

[0027]

[0028] in, This is a true value coverage metric for the interval. To estimate the number of times, For nodes The number of times the true value falls within the estimated interval. The total number of nodes. It is a conservative indicator for the range. , These are the upper and lower bounds of the interval state estimation results, respectively. As an indicator of the deviation within the interval, , These are the upper and lower bounds of the true interval, respectively.

[0029] As can be seen from the above technical solution, this invention provides a distribution network state estimation method based on multi-timescale hybrid measurements. Compared with the prior art, this invention has the following advantages: by taking into account the multi-source measurement data fusion strategy of asynchronous measurement deviation, the impact of asynchronous sampling time of multi-source measurements on the accuracy of dynamic state estimation is reduced; the distribution network is divided into multiple sub-regions, and the topology information between sub-regions is exchanged through network equivalence, and the equivalent injected power is updated based on the power balance equation. When the topology changes, only the dynamic state estimation model of the sub-region with topology change needs to be reconstructed, which improves the computational efficiency of dynamic state estimation; the uncertainty of asynchronous measurement deviation is represented by interval numbers, an interval dynamic state estimation model of the sub-region is constructed, and an improved interval Kalman filter method based on upper bound optimization and measurement function linearization is proposed to suppress the interval expansion caused by a large number of interval operations during the solution process. Attached Figure Description

[0030] Figure 1 This is a flowchart illustrating the distribution network state estimation method based on multi-timescale hybrid measurement of the present invention. Figure 2 This is a schematic diagram of multi-source measurement data fusion in this invention; Figure 3 This is a schematic diagram of topological information interaction in this invention; Figure 4 This is a flowchart of the solution process for the interval dynamic state estimation model in this invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0032] like Figure 1 As shown, the distribution network state estimation method based on multi-time-scale hybrid measurement in this embodiment includes the following steps: S1. Construct a multi-source measurement dataset including D-PMU measurement points, SCADA measurement points, and AMI measurement points; S2. Construct a multi-objective optimization function that includes measurement point configuration and partition scale, and divide the power grid into multiple sub-regions with D-PMU and SCADA measurement points as boundary nodes. S3. Multiple sub-regions exchange topological information through boundary nodes, condense the topological equivalents of adjacent sub-regions to the boundary nodes, calculate the equivalent injection power of the boundary nodes, and each sub-region independently constructs an interval dynamic state estimation model. S4. Solve the dynamic state estimation model for each sub-region, and perform dynamic state estimation for the entire network by aggregating the estimation results of each sub-region. S5. Calculate the evaluation index of the dynamic state estimation of the whole network, verify the interval estimation results, obtain the evaluation results, and optimize the parameters of the dynamic state estimation model based on the evaluation results.

[0033] S1 includes the following steps: S11 and D-PMU use D-PMU measurement data at the state estimation time to perform state estimation, and select SCADA measurement data closest to the state estimation time for state estimation. S12. Set a time window to determine whether the AMI measurement data is updated at the state estimation time; S13. If the state estimation time has been updated, then the updated AMI measurement data shall be used for state estimation. S14. If the state is not updated at the state estimation time, use the predicted value of the state quantity to calculate the pseudo-measurement data of the AMI node for state estimation, and construct a robust factor to adjust the weight of the AMI pseudo-measurement in the state estimation. S15. Calculate the asynchronous measurement deviation caused by the asynchronous sampling time.

[0034] like Figure 2 As shown, specifically, in order to reduce the impact of asynchronous sampling times and sampling frequency differences of multi-source measurements on state estimation, it is necessary to fuse the multi-source measurement data at the state estimation time to construct the multi-source measurement dataset at the state estimation time. Figure 2 Chinese: T sE To estimate the period, the D-PMU uses D-PMU measurement data at the state estimation time for state estimation. Simultaneously, SCADA measurement data closest to the state estimation time is selected for state estimation, and the asynchronous measurement bias caused by the asynchronous sampling time is calculated.

[0035] S3 includes the following steps: S31. Each sub-region exchanges topological information through boundary nodes, and the topological condensation of adjacent sub-regions is achieved through the equivalent admittance matrix. S32. Calculate the equivalent injected power of the boundary node by measuring the power of the boundary node and the node voltage in the sub-region. S33. Interval numbers are used to characterize the asynchronous measurement bias of multi-source measurement data, and an interval dynamic state estimation model for multi-source measurement fusion is constructed.

[0036] S4 includes the following steps: S41. State Prediction: Calculate the state prediction quantity and the prediction error covariance matrix, and transform the prediction error covariance matrix into an upper bound matrix by combining the upper bound optimization idea. S42. Measurement Transformation: Converts power measurement into equivalent current measurement; S43. The filtering calculation is completed through the upper bound matrix to obtain the state estimation results of each sub-region; S44. Aggregate the interval state estimation results of each sub-region to form the dynamic state estimation output of the entire network.

[0037] S5 includes the following steps: S51. Calculate the evaluation index for the dynamic state estimation of the entire network using the evaluation index calculation formula; S52. Based on the evaluation results, optimize the multi-source measurement fusion strategy and dynamic state estimation model parameters in reverse.

[0038] The formula for calculating AMI measurement data is as follows:

[0039] in, For AMI in Measurement data at time, () represents the time series of AMI measurement data. (·) represents the measurement function of AMI. In order to be in The most recently updated sampling time in AMI before the current time. For time window, for State prediction at any given time; Specifically, when the AMI measurement is not updated, this paper predicts the pseudo-measurement of the AMI at the current time from the estimated value of the state quantity at the previous time step. In the first few state estimations between two AMI measurement times, this method can guarantee its accuracy to a certain extent. However, the accuracy of the AMI pseudo-measurement data calculated by this method will gradually decrease thereafter, leading to a reduction in the accuracy of the state estimation. Therefore, this section uses the physical equation between the measurement and the state quantity to quantify the error of the AMI pseudo-measurement, and then constructs a robustness factor. Adjust the weight of AMI pseudo-measurements in state estimation to reduce the impact of decreased pseudo-measurement accuracy.

[0040] The robustness factor is defined as:

[0041] in, The robustness factor This is the robustness factor correction function; The formula for calculating asynchronous measurement deviation is as follows:

[0042] in, This is due to asynchronous measurement deviation; The average rate of change of the measured quantity over the interval between the sampling time and the state estimation time is given by the metric. This represents the time deviation between the sampling time and the state estimation time.

[0043] Measurement error taking into account asynchronous measurement deviation for:

[0044] in, The measurement equipment error refers to the error generated during the data acquisition process of the measurement equipment. The measurement equipment error and the asynchronous measurement deviation are independent of each other, and the variance of the measurement error can be expressed as:

[0045] in, Let V be the expected variance of the measurement equipment error; The expected variance of the asynchronous measurement bias; The equivalent admittance matrix is ​​as follows:

[0046] In this context, subscript A represents an internal node, and subscript B represents a boundary node; Specifically, assuming the power grid is divided into sub-regions 1 and 2, the corresponding admittance matrix Y can be expressed as:

[0047] like Figure 3 As shown, Y1 and Y2 are the admittance matrices of subregions 1 and 2, respectively; Taking the topology information exchange process of sub-region 1 as an example, firstly, the topology information of sub-region 2 is condensed and equivalently represented to the boundary nodes, and the equivalent boundary node admittance matrix is ​​calculated. The calculation formula is as follows:

[0048] Subsequently, sub-regions 1 and 2 exchange topological information to obtain the boundary node admittance matrix of sub-region 2 after it is equivalent. Then, combining its own topology information, the equivalent admittance matrix containing the entire network topology information is obtained through Ward's equivalent values. ;

[0049]

[0050] in, The boundary node admittance matrix after subregion 1 is equivalent represents the influence of adjacent subregions on subregion 1; According to the power balance constraint theory, the equivalent injected power of the boundary node can be calculated through power measurements at the boundary node and the node voltages within the sub-region. Since obtaining accurate node voltage information is difficult during state estimation, predicted values ​​of state variables are used for calculation. The formula for calculating the equivalent injected power of the boundary node is as follows:

[0051] in, as well as Boundary nodes Equivalent injection power, Boundary nodes The set of line measurements connected to nodes within the sub-region. After equivalence to Ward and boundary nodes A set of connected equivalent lines, , For boundary nodes Equivalent admittance of connected branches, , For sub-regions and boundary nodes Power measurements and predicted status values ​​of connected lines. , They are respectively Time Node The real and imaginary parts of the state prediction quantity , They are respectively Time Node The real and imaginary parts of the state prediction quantity; As shown above, when the network topology remains unchanged, each sub-region completes its independent dynamic state estimation after updating the equivalent injection power of the corresponding boundary nodes. When the topology changes, it is only necessary to reconstruct the dynamic state estimation model of the sub-region with the changed topology. By exchanging the topology information of each sub-region and updating the equivalent injection power of the boundary nodes, the estimation model can be solved to obtain the dynamic state estimation results of each region.

[0052] The expression for the interval dynamic state estimation model is as follows:

[0053] in, For interval state variables, For interval measurement, for n System process noise, for m Dimensional measurement noise, This is the state transition function. This is the measurement function.

[0054] The formulas for calculating the state prediction and the prediction error covariance matrix are as follows:

[0055] in, for State prediction at time t, for The state estimate at time t, where F is the state transition matrix. for The covariance matrix of the prediction error at any given time. for Time-estimation error covariance matrix for The covariance matrix;

[0056] in, for The Jacobian matrix of the time measurement function, for The gain matrix at time step I, where I is the identity matrix. for The measurement error matrix.

[0057] Because the measurement dataset contains nonlinear power measurements, the iterative solution of the Jacobian matrix during the recursive process introduces a large number of nonlinear interval operations, which exacerbates the expansion of the solution interval and makes the state estimation results unusable. To address this, an improved interval Kalman filter algorithm is proposed. This algorithm combines the idea of ​​upper bound optimization to transform the prediction error covariance matrix into an upper bound matrix, and transforms the power measurement into an equivalent current measurement through measurement transformation. At the same time, it uses error propagation theory to correct the measurement error of the equivalent current measurement, thus suppressing the interval expansion problem caused by nonlinear interval operations while ensuring the accuracy of the solution algorithm.

[0058] The expression for the upper bound matrix is ​​as follows:

[0059] in, This is the upper bound matrix of the interval prediction error covariance matrix. Parameters can be derived from a matrix The Frobenius norm is obtained from the matrix. elements The definition is as follows:

[0060] in, To determine the center value of the prediction error covariance matrix, This is an augmented matrix of the prediction error covariance matrix; The filtering calculation is performed using the upper bound matrix as follows:

[0061] in, is the Jacobian matrix after measurement transformation, and is a constant matrix; This is the measurement error matrix after measurement transformation; =1 / a k .

[0062] To verify the effectiveness and accuracy of the proposed method, a truth coverage index for the solution interval of state estimation is defined. Conservative indicators Deviation indicators To evaluate the state estimation results, Characterize the completeness of the state estimation interval. The higher the value, the more complete the interval estimation result. Under the premise of satisfying completeness, and The smaller the value, the higher the reference value of the interval estimation result; The formulas for calculating the evaluation indicators are as follows:

[0063]

[0064]

[0065] in, This is a true value coverage metric for the interval. To estimate the number of times, For nodes The number of times the true value falls within the estimated interval. The total number of nodes. It is a conservative indicator for the range. , These are the upper and lower bounds of the interval state estimation results, respectively. As an indicator of the deviation within the interval, , These are the upper and lower bounds of the true interval, respectively.

[0066] In summary, firstly, a multi-source measurement data fusion strategy considering asynchronous measurement bias is proposed to reduce the impact of asynchronous sampling times on the accuracy of dynamic state estimation. Secondly, the distribution network is divided into multiple sub-regions, and the topology information between sub-regions is exchanged through network equivalence. The equivalent injected power is updated based on the power balance equation. When the topology changes, only the dynamic state estimation model of the sub-region with the topology change needs to be reconstructed, which improves the computational efficiency of dynamic state estimation. Finally, the uncertainty of asynchronous measurement bias is represented by interval numbers, and an interval dynamic state estimation model for the sub-region is constructed. An improved interval Kalman filter method based on upper bound optimization and measurement function linearization is proposed to suppress interval expansion caused by a large number of interval operations during the solution process.

[0067] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state disk (SSD)).

[0068] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0069] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0070] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. The distribution network state estimation method based on multi-timescale hybrid measurement of the present invention is characterized in that, Includes the following steps: S1. Construct a multi-source measurement dataset including D-PMU measurement points, SCADA measurement points, and AMI measurement points; S2. Construct a multi-objective optimization function that includes the configuration of measurement points, and divide the power grid into multiple sub-regions with the D-PMU measurement points and the SCADA measurement points as boundary nodes. S3. The multiple sub-regions interact with topological information through boundary nodes, condense the topological equivalents of adjacent sub-regions to the boundary nodes, calculate the equivalent injection power of the boundary nodes, and each sub-region independently constructs an interval dynamic state estimation model. S4. Solve the dynamic state estimation model of each sub-region, and perform dynamic state estimation of the entire network by aggregating the estimation results of each sub-region. S5. Calculate the evaluation index of the dynamic state estimation of the whole network, verify the interval estimation results, obtain the evaluation results, and optimize the parameters of the dynamic state estimation model based on the evaluation results.

2. The distribution network state estimation method based on multi-time-scale hybrid measurement according to claim 1, characterized in that: S1 includes the following steps: S11 and D-PMU use D-PMU measurement data at the state estimation time to perform state estimation, and select SCADA measurement data closest to the state estimation time for state estimation. S12. Set a time window to determine whether the AMI measurement data is updated at the state estimation time; S13. If the state estimation time has been updated, then the updated AMI measurement data shall be used for state estimation. S14. If the state is not updated at the state estimation time, use the predicted value of the state quantity to calculate the pseudo-measurement data of the AMI node for state estimation, and construct a robust factor to adjust the weight of the AMI pseudo-measurement in the state estimation. S15. Calculate the asynchronous measurement deviation caused by the asynchronous sampling time.

3. The distribution network state estimation method based on multi-timescale hybrid measurement according to claim 2, characterized in that: S3 includes the following steps: S31. Each sub-region exchanges topological information through boundary nodes, and the topological condensation of adjacent sub-regions is achieved through the equivalent admittance matrix. S32. Calculate the equivalent injected power of the boundary node by measuring the power of the boundary node and the node voltage in the sub-region. S33. Interval numbers are used to characterize the asynchronous measurement bias of multi-source measurement data, and an interval dynamic state estimation model for multi-source measurement fusion is constructed.

4. The distribution network state estimation method based on multi-timescale hybrid measurement according to claim 3, characterized in that: S4 includes the following steps: S41. State Prediction: Calculate the state prediction quantity and the prediction error covariance matrix, and transform the prediction error covariance matrix into an upper bound matrix by combining the upper bound optimization idea. S42. Measurement Transformation: Converts power measurement into equivalent current measurement; S43. The filtering calculation is completed through the upper bound matrix to obtain the state estimation results of each sub-region; S44. Aggregate the interval state estimation results of each sub-region to form the dynamic state estimation output of the entire network.

5. The distribution network state estimation method based on multi-timescale hybrid measurement according to claim 4, characterized in that: S5 includes the following steps: S51. Calculate the evaluation index for the dynamic state estimation of the entire network using the evaluation index calculation formula; S52. Based on the evaluation results, optimize the multi-source measurement fusion strategy and dynamic state estimation model parameters in reverse.

6. The distribution network state estimation method based on multi-timescale hybrid measurement according to claim 5, characterized in that: The formula for calculating the AMI measurement data is as follows: in, For AMI in Measurement data at time, () represents the time series of AMI measurement data. (·) represents the measurement function of AMI. In order to be in The most recently updated sampling time in AMI before the current time. For time window, for State prediction at any given time; The robustness factor is defined as: in, The robustness factor This is the robustness factor correction function; The formula for calculating the asynchronous measurement deviation is as follows: in, This is due to asynchronous measurement deviation; The average rate of change of the measured quantity over the interval between the sampling time and the state estimation time is given by the metric. This represents the time deviation between the sampling time and the state estimation time.

7. The distribution network state estimation method based on multi-timescale hybrid measurement according to claim 6, characterized in that: The equivalent admittance matrix is ​​as follows: In this context, subscript A represents an internal node, and subscript B represents a boundary node; The formula for calculating the equivalent injected power of the boundary node is as follows: in, as well as Boundary nodes Equivalent injection power, Boundary nodes The set of line measurements connected to nodes within the sub-region. After equivalence to Ward and boundary nodes A set of connected equivalent lines, , For boundary nodes Equivalent admittance of connected branches, , For sub-regions and boundary nodes Power measurements and predicted status values ​​of connected lines. , They are respectively Time Node The real and imaginary parts of the state prediction quantity , They are respectively for Time Node The real and imaginary parts of the state prediction quantity; The expression for the interval dynamic state estimation model is as follows: in, For interval state variables, For interval measurement, for n System process noise, for m Dimensional measurement noise, This is the state transition function. This is the measurement function.

8. The distribution network state estimation method based on multi-timescale hybrid measurement according to claim 7, characterized in that: The formulas for calculating the state prediction and the prediction error covariance matrix are as follows: in, for State prediction at time t, for The state estimate at time t, where F is the state transition matrix. for The covariance matrix of the prediction error at any given time. for Time-estimation error covariance matrix for The covariance matrix; in, for The Jacobian matrix of the time measurement function, for The gain matrix at time step I, where I is the identity matrix. for The measurement error matrix.

9. The distribution network state estimation method based on multi-timescale hybrid measurement according to claim 8, characterized in that: The expression for the upper bound matrix is ​​as follows: in, This is the upper bound matrix of the interval prediction error covariance matrix. Parameters can be derived from a matrix The Frobenius norm is obtained from the matrix. elements The definition is as follows: in, To determine the center value of the prediction error covariance matrix, This is an augmented matrix of the prediction error covariance matrix; The filtering calculation using the upper bound matrix is ​​as follows: in, is the Jacobian matrix after measurement transformation, and is a constant matrix; This is the measurement error matrix after measurement transformation; =1 / a k .

10. The distribution network state estimation method based on multi-time-scale hybrid measurement according to claim 9, characterized in that: The calculation formula for the evaluation indicators is as follows: in, This is a true value coverage metric for the interval. To estimate the number of times, For nodes The number of times the true value falls within the estimated interval. The total number of nodes. It is a conservative indicator for the range. , These are the upper and lower bounds of the interval state estimation results, respectively. As an indicator of the deviation within the interval, , These are the upper and lower bounds of the true interval, respectively.

Images