Phase imbalance correction device and correction method for distribution system based on energy storage collaboration

Abstract

The present invention discloses a phase-to-phase imbalance correction device and correction method for a power distribution system based on energy storage collaboration, which relates to the field of power conversion technology and includes: collecting historical phase-to-phase operating data of the power distribution system, and predicting the phase-to-phase operating curve based on the historical operating data; collecting the actual phase-to-phase operating curve of the power distribution system in real time, calculating the error coefficient based on the predicted phase-to-phase operating curve and the actual phase-to-phase operating curve, and generating a corrected phase-to-phase operating curve based on the error coefficient; obtaining the phase-to-phase imbalance time point and correction value at a future moment based on the corrected phase-to-phase operating curve, and using the imbalance time point and correction value as correction parameters for the energy storage unit. By analyzing the error coefficient between the predicted phase-to-phase operating curve and the actual phase-to-phase operating curve, and generating a corrected phase-to-phase operating curve based on the error coefficient, the phase-to-phase imbalance time point and correction value at a future moment are extracted from the corrected phase-to-phase operating curve, thereby performing early correction on the phases of the power distribution system and reducing the failure rate of phase imbalance.

Description

Technical Field

[0001] The present invention belongs to the technical field of power conversion, and in particular relates to a phase imbalance correction device and correction method for a power distribution system based on energy storage collaboration. Background Art

[0002] Three-phase imbalance in a power system is caused by unbalanced three-phase loads and asymmetric three-phase parameters of system components. The balance of three-phase voltage in a power system is a key indicator of power quality, and phase imbalances are currently corrected.

[0003] For example, the Chinese invention patent with the authorization announcement number CN105870944B discloses a method for controlling interphase power balance of a power electronic transformer, which includes the following steps: calculating the d-axis component and q-axis component of the zero-sequence voltage using the difference in active power under the zero-sequence voltage to be output by any two phases of the transformer and the active power under the zero-sequence voltage to be output by the third phase of the three phases; calculating the zero-sequence voltage based on the d-axis component of the zero-sequence voltage, the q-axis component of the zero-sequence voltage, and the grid angular frequency; superimposing the three-phase modulation wave output by the high-voltage stage of the transformer with the zero-sequence voltage to generate a new three-phase modulation wave; generating corresponding trigger pulses based on the new three-phase modulation wave to control the high-voltage stage of the transformer and realize interphase power balance control. This patent intervenes in the correction after the grid voltage has an unbalanced fault, and cannot perform correction in advance before the fault occurs to avoid the occurrence of phase imbalance.

[0004] To this end, the present invention proposes a phase imbalance correction device and correction method for a power distribution system based on energy storage collaboration. Summary of the Invention

[0005] The present invention aims to address at least one of the technical problems existing in the prior art. To this end, the present invention proposes a device and method for correcting interphase imbalance in a power distribution system based on energy storage collaboration. This device and method predict future interphase imbalances based on historical power usage data, and uses energy storage units to perform preemptive corrections, thereby reducing the failure rate of interphase imbalances.

[0006] To achieve the above objectives, a distribution system interphase imbalance correction device and correction method based on energy storage collaboration are proposed, including:

[0007] Collect historical inter-phase operating data of the distribution system and predict inter-phase operating curves based on the historical operating data;

[0008] Collect the actual phase-to-phase operating curve of the distribution system in real time, calculate the error coefficient based on the predicted phase-to-phase operating curve and the actual phase-to-phase operating curve, and generate the corrected phase-to-phase operating curve based on the error coefficient;

[0009] Based on the corrected inter-phase operation curve, the inter-phase imbalance time point and the correction value at a future moment are obtained, and the imbalance time point and the correction value are used as correction parameters of the energy storage unit.

[0010] Preferably, the method for predicting the inter-phase operation curve based on historical operation data includes:

[0011] A voltage identification model that predicts phase-to-phase voltage based on historical operating data training;

[0012] Continuously collect n pieces of distribution system operation data at a preset frequency, where n is an integer greater than 1, input the n pieces of operation data into a voltage identification model in sequence, and output n phase-to-phase voltages;

[0013] The n phase-to-phase voltages are mapped in a coordinate system with time as the horizontal coordinate and the phase-to-phase voltage as the vertical coordinate according to the order of acquisition time, and the n phase-to-phase voltages are connected in sequence to obtain the predicted phase-to-phase operation curve.

[0014] Preferably, the historical operation data includes H group condition data and phase-to-phase voltage corresponding to the H group condition data;

[0015] Where H is the preset amount of training data, H=1, 2, 3…H.

[0016] Preferably, the voltage recognition model is trained in the following manner:

[0017] A set of conditional data and the phase-to-phase voltages corresponding to the conditional data are converted into a set of first eigenvectors, and the first eigenvectors are used as inputs of a machine learning model A. The machine learning model A uses the predicted phase-to-phase voltage for each set of conditional data as output, and minimizes the sum of the prediction accuracies of all phase-to-phase voltages as a training objective. The machine learning model A is trained until the sum of the prediction accuracies reaches convergence, and the training is stopped. The machine learning model A obtained through training is used as a voltage recognition model, and the machine learning model A is a polynomial regression model or a support vector machine model.

[0018] Preferably, the predicted interphase operation curve is a curve of the interphase voltage change over time predicted by the power distribution system based on current operation data and combined with historical operation data;

[0019] The actual phase-to-phase operation curve is a curve drawn by collecting the phase-to-phase voltage changes over time in real time by the power distribution system.

[0020] Preferably, the error coefficient is calculated based on the predicted inter-phase operation curve and the actual inter-phase operation curve as follows:

[0021] Calculate the sub-error coefficients of the predicted phase-to-phase operating curve and the actual phase-to-phase operating curve. The calculation expression is:

[0022]

[0023] in, It represents the sub-error coefficient at the time t of the nth running data; It is the voltage value of the predicted phase-to-phase operating curve at the nth operating data and the tth moment; is the voltage value collected at the tth moment of the actual phase-to-phase operation curve;

[0024] The error coefficient is calculated based on n sub-error coefficients. The error coefficient calculation expression is:

[0025] ,in is the error coefficient.

[0026] Preferably, the method for generating the correction phase-to-phase operation curve based on the error coefficient is:

[0027] Step 1: Calculate the correction voltage at time t+L based on the error coefficient, where L is an integer greater than 1. The calculation expression is:

[0028] , is the compensation coefficient, D is the correction voltage;

[0029] Step 2: Repeat step 1 and continue to update the value of L, and calculate the updated correction voltage value of t+L in sequence to obtain A correction voltage value, , and By connecting the corrected voltage values in chronological order, the corrected phase-to-phase operation curve can be drawn.

[0030] Preferably, the method for obtaining the interphase imbalance time point and correction value at a future moment based on the corrected interphase operation curve is:

[0031] Compare the corrected voltage at a future moment with a preset voltage threshold. If the corrected voltage is greater than or less than the preset voltage threshold, it is determined to be phase unbalanced, and the corresponding moment of the corrected voltage is marked. If the corrected voltage is equal to the preset voltage threshold, it is determined to be phase balanced.

[0032] Collect multiple sets of historical correction data, where the historical correction data includes imbalance data and correction values corresponding to the imbalance data;

[0033] A correction recognition model that predicts correction values is trained based on historical correction data;

[0034] Obtain real-time operating data of the power distribution system, input the real-time operating data into the correction identification model to output the correction value.

[0035] Preferably, the correction recognition model is trained in the following manner:

[0036] A set of unbalanced data and the correction values corresponding to the unbalanced data are converted into a set of second eigenvectors, and the second eigenvectors are used as inputs of a machine learning model B. The machine learning model B outputs the predicted correction value for each set of unbalanced data, and minimizes the sum of the prediction accuracies of all correction values as a training goal; the machine learning model B is trained until the sum of the prediction accuracies reaches convergence, and the training is stopped. The trained machine learning model B is used as a correction recognition model, and the machine learning model B is a polynomial regression model or a support vector machine model.

[0037] A distribution system interphase imbalance correction device based on energy storage collaboration is used to implement the above-mentioned distribution system interphase imbalance correction method based on energy storage collaboration, including a historical data collection module, an acquisition module, an operation curve prediction module, a correction parameter generation module, a correction module and an energy storage unit; wherein each module is connected to each other via wired and / or wireless means;

[0038] Historical data collection module, used to collect historical operation data between phases;

[0039] Acquisition module, used to collect actual phase-to-phase operation curves;

[0040] An operation curve prediction module is used to predict the phase operation curve based on historical operation data;

[0041] A correction parameter generation module calculates an error coefficient based on the predicted interphase operation curve and the actual interphase operation curve, generates a correction interphase operation curve based on the error coefficient, and obtains the interphase imbalance time point and correction value at a future moment based on the correction interphase operation curve;

[0042] The correction module controls the coordination of the energy storage units based on the imbalance time point and the correction value, and adjusts the phases by injecting or absorbing energy.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] The present invention collects historical operating data of the distribution system, predicts the phase-to-phase operating curve based on the historical operating data, analyzes the error coefficient between the predicted phase-to-phase operating curve and the actual phase-to-phase operating curve, and generates a corrected phase-to-phase operating curve based on the error coefficient. The phase-to-phase imbalance time point and correction value at a future moment are extracted from the corrected phase-to-phase operating curve, thereby performing early correction of the phases of the distribution system and reducing the failure rate of phase imbalance. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flow chart of a method for correcting inter-phase imbalance in a power distribution system based on energy storage collaboration in Example 1 of the present invention;

[0046] Figure 21 is an example diagram of the predicted inter-phase operation curve and the actual inter-phase operation curve in Example 1 of the present invention;

[0047] Figure 3 This is a module connection diagram of a power distribution system inter-phase imbalance correction device based on energy storage collaboration in Example 2 of the present invention;

[0048] Figure 4 This is a relationship diagram of the application of the phase imbalance correction device of the power distribution system based on energy storage cooperation in embodiment 2 of the present invention to the power distribution system. DETAILED DESCRIPTION

[0049] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0050] Example 1

[0051] like Figure 1 and Figure 2 As shown, the method for correcting inter-phase imbalance in a distribution system based on energy storage collaboration includes:

[0052] Collect historical interphase operating data of the distribution system, the historical operating data including H groups of conditional data and the phase-to-phase voltages corresponding to the H groups of conditional data; where H is a preset amount of training data, H=1, 2, 3, ..., H. The conditional data in this embodiment covers different time periods, different load conditions, and various factors that may affect the phase-to-phase voltage, including but not limited to information such as phase-to-phase voltage values, current values, load sizes, and power factors. Preprocess the collected historical operating data, including data cleaning, outlier removal, and data normalization, to ensure data accuracy and reliability. Predict the phase-to-phase operating curve based on the historical operating data. A specific method for predicting the phase-to-phase operating curve based on the historical operating data is to first train a voltage recognition model for predicting the phase-to-phase voltage based on the historical operating data.

[0053] Continuously collecting n pieces of distribution system operating data at a preset frequency. The distribution system operating data includes the same features as historical operating data so that it can be input into a trained voltage identification model for prediction. The preset frequency in this embodiment is determined by those skilled in the art based on actual needs. For example, the AC frequency of each phase of three-phase power is used as the preset frequency. n is an integer greater than 1. The n pieces of operating data are sequentially input into the voltage identification model, and n phase-to-phase voltage prediction values are output.

[0054] Map n phase-to-phase voltages in the order of acquisition time in a coordinate system with time as the horizontal axis and phase-to-phase voltage as the vertical axis, and connect n phase-to-phase voltages in sequence to obtain the predicted phase operation curve. In this way, the trend of phase-to-phase voltage change over time can be observed intuitively. Figure 2 The curve marked as Y in the figure represents a prediction of future phase-to-phase voltage changes based on current operating data and historical experience. By way of example, this embodiment analyzes and predicts the phase-to-phase operating curve to understand the changing trend and potential imbalance problems of the distribution system's phase-to-phase voltage. For example, by observing characteristics such as the slope and fluctuation amplitude of the curve, the stability and imbalance level of the system can be judged. Based on the predicted phase-to-phase operating curve, measures can be taken in advance to prevent the occurrence of phase-to-phase imbalance problems. For example, load distribution can be adjusted, and the charging and discharging strategies of the energy storage system can be optimized to maintain phase-to-phase voltage balance. Using the predicted phase-to-phase operating curve as input and combining it with other system parameters, more in-depth analysis and decision-making can be carried out. For example, the system's power loss can be calculated, and the reliability of the equipment can be evaluated, providing a basis for system optimization and maintenance.

[0055] Specifically, the training method of the voltage recognition model is:

[0056] Historical operating data is collected under the experimental environment. The historical operating data includes H-group condition data and the phase-to-phase voltage corresponding to the H-group condition data. That is, in the experimental environment, multiple sets of condition data are simulated, the distribution system is operated under these condition data, and the phase-to-phase voltage at this time is recorded as the phase-to-phase voltage corresponding to the condition data. By simulating different H-group condition data in this way, the phase-to-phase voltage corresponding to the H-group condition data can be obtained;

[0057] A set of conditional data and the phase-to-phase voltages corresponding to the conditional data are converted into a set of first eigenvectors. The first eigenvectors are used as inputs to a machine learning model A. The machine learning model A uses the predicted phase-to-phase voltage for each set of conditional data as output, and minimizes the sum of the prediction accuracies of all phase-to-phase voltages as a training objective. The prediction accuracy is calculated as follows: ,in, is the number of the first eigenvector, For prediction accuracy, For the The predicted phase-to-phase voltage corresponding to the first eigenvector of the group, For the The actual phase-to-phase voltage corresponding to the first eigenvector of the group is obtained; the machine learning model A is trained until the sum of the prediction accuracies reaches convergence and the training is stopped.

[0058] It should be noted that the predicted phase-to-phase operation curve is a curve showing the change of phase-to-phase voltage over time, predicted based on the current operating data and historical operating data of the distribution system. Specifically, when generating the predicted phase-to-phase operation curve, the current distribution system operating data is first analyzed. These operating data may include the current load conditions, power factor, system topology, etc. At the same time, combined with a large amount of historical operating data, advanced data processing and analysis technologies, such as machine learning algorithms and statistical models, are used to predict future trends in phase-to-phase voltage changes. The predicted phase-to-phase operation curve generated in this way can provide forward-looking guidance for system operation and management, helping to discover potential problems in advance and take corresponding measures.

[0059] The actual phase-to-phase operation curve is a curve drawn by collecting the phase-to-phase voltage changes over time in real time from the distribution system. During the operation of the distribution system, sensors and monitoring equipment installed at different locations collect phase-to-phase voltage data in real time at a certain frequency. These data are recorded in chronological order, and the actual phase-to-phase operation curve is drawn with the phase-to-phase voltage as the vertical axis and time as the horizontal axis. The actual phase-to-phase operation curve reflects the true state of the distribution system in actual operation and is an important basis for evaluating the system performance and stability. By comparing the predicted phase-to-phase operation curve with the actual phase-to-phase operation curve, the difference between the two can be discovered in a timely manner to determine whether the system operation meets expectations. If a large difference is found, it may mean that there is an abnormality in the system, and further cause analysis and corresponding adjustment measures are needed to ensure the safe, stable and efficient operation of the distribution system.

[0060] See Figure 2 As shown, the actual phase-to-phase operation curve of the distribution system is collected in real time. Figure 2 The curve marked S is the actual interphase operation curve. The error coefficient is calculated based on the predicted interphase operation curve and the actual interphase operation curve, and the corrected interphase operation curve is generated based on the error coefficient. Specifically, the error coefficient is calculated based on the predicted interphase operation curve and the actual interphase operation curve as follows:

[0061] Calculate the sub-error coefficients of the predicted phase-to-phase operating curve and the actual phase-to-phase operating curve. The calculation expression is:

[0062]

[0063] in, It represents the sub-error coefficient at the time t of the nth running data; It is the voltage value of the predicted phase-to-phase operating curve at the nth operating data and the tth moment; is the voltage value collected at time t by the actual phase-to-phase operating curve. The purpose is to use this formula to calculate the sub-error coefficient corresponding to each operating data at each moment, reflecting the degree of deviation between the predicted value and the actual value.

[0064] The error coefficient is calculated based on n sub-error coefficients. The error coefficient calculation expression is:

[0065] ,The purpose is to obtain a comprehensive error coefficient by averaging all the sub-error coefficients, which represents the ,prediction error of the entire distribution system at a specific moment.

[0066] Furthermore, the method for generating the correction phase operation curve based on the error coefficient is as follows:

[0067] Step 1: Calculate the corrected voltage at time t+L based on the error coefficient, that is, predict the corrected voltage at the future time at the current time. The corrected voltage is the voltage between the operating phases of the distribution system at the future time, that is, at time t+L, after calculating the error between the predicted phase-to-phase operating curve and the actual phase-to-phase operating curve. L is an integer greater than 1, and the calculation expression is:

[0068] ,in is the error coefficient, is a compensation coefficient, which is obtained by those skilled in the art based on a large amount of experimental data and is not specifically limited here. D is a correction voltage. The purpose is to adjust the voltage value at the current moment by considering the error coefficient and the compensation coefficient to obtain the correction voltage at time t+L.

[0069] Step 2: Repeat step 1 and continue to update the value of L, and calculate the updated correction voltage value of t+L in sequence to obtain A correction voltage value, , and By connecting the corrected voltage values in chronological order, a corrected phase-to-phase operating curve can be drawn. This curve reflects the predicted change in the distribution system's phase-to-phase voltage over time after correction. By continuously predicting voltage values at future times, the operating status of the distribution system can be more accurately understood, providing a strong basis for subsequent phase-to-phase imbalance correction. Based on the corrected phase-to-phase operating curve, the time points and correction values of phase imbalances at future times are obtained. These imbalance time points and correction values serve as correction parameters for the energy storage units.

[0070] Specifically, the method for obtaining the interphase imbalance time point and correction value at a future moment based on the corrected interphase operation curve is as follows:

[0071] Compare the corrected voltage at a future moment with a preset voltage threshold. If the corrected voltage is greater than or less than the preset voltage threshold, it is determined to be phase unbalanced, and the corresponding moment of the corrected voltage is marked. If the corrected voltage is equal to the preset voltage threshold, it is determined to be phase balanced.

[0072] Collect multiple sets of historical correction data, which include imbalance data and correction values corresponding to the imbalance data. That is, in an experimental environment, simulate multiple sets of imbalance data. The imbalance data includes but is not limited to information such as phase-to-phase voltage, current, load size, and power factor monitored from the power distribution system when phase power imbalance occurs. By simulating imbalance data in different time periods and different load conditions, the phases are corrected using different correction values. Personnel skilled in the art test the corrected phases and select the correction value with the best correction effect to correspond to the imbalance data. By simulating multiple sets of different imbalance data in this way, multiple sets of historical correction data can be obtained.

[0073] A correction recognition model that predicts correction values is trained based on historical correction data;

[0074] Obtain real-time operating data of the power distribution system, input the real-time operating data into the correction identification model to output the correction value.

[0075] More specifically, the training method of the correction recognition model is:

[0076] Converting a set of unbalanced data and correction values corresponding to the unbalanced data into a set of second eigenvectors;

[0077] The second eigenvector is used as the input of the machine learning model B. The machine learning model B outputs the predicted correction value for each set of imbalanced data, and minimizes the sum of the prediction accuracy of all correction values as the training goal. The calculation formula for prediction accuracy is: , where g is the number of the second eigenvector, For prediction accuracy, is the predicted correction value corresponding to the second eigenvector of the g-th group, is the actual correction value corresponding to the second eigenvector of the g-th group; the machine learning model B is trained until the sum of the prediction accuracies reaches convergence, and the training is stopped. The trained machine learning model B is used as the correction recognition model;

[0078] It should be noted that machine learning model A and machine learning model B are polynomial regression models or support vector machine models.

[0079] Specifically, this embodiment obtains the imbalance time point and correction value that will occur in the future, and then starts the compensation operation of the energy storage unit in advance for the imbalance time point based on the response capability and load forecast of the selected energy storage unit. The purpose is that the energy storage unit can inject or absorb energy in advance based on the predicted voltage fluctuation trend to avoid the impact when the imbalance actually occurs.

[0080] Example 2

[0081] Based on the above embodiments, see Figure 3 and Figure 4 As shown, a phase imbalance correction device for a power distribution system based on energy storage collaboration includes a historical data collection module, an acquisition module, an operation curve prediction module, a correction parameter generation module, a correction module and an energy storage unit; wherein each module is connected to each other by wired and / or wireless means;

[0082] A historical data collection module is used to collect historical interphase operating data. In this embodiment, the historical operating data includes but is not limited to information such as interphase voltage, current, load, and power factor. The historical data collection module includes multiple units and uses sensors that can collect information such as interphase voltage, current, load, and power factor. The sensors are deployed on the power distribution system. The specific deployment location and sensor model are not specifically limited here.

[0083] The acquisition module is used to collect the actual phase-to-phase operation curve. The acquisition module includes a sensor for collecting phase-to-phase voltage values. The sensor is deployed on the power distribution system. The specific deployment location and sensor model are not limited here. Figure 2 As shown, the collected phase-to-phase voltage values and the timestamp of the collection are plotted as shown in Figure 2 The curve shown can be used to obtain the actual phase-to-phase operation curve;

[0084] The operation curve prediction module is used to predict the phase-to-phase operation curve based on historical operation data. The specific method is to calculate the sub-error coefficient between the predicted phase-to-phase operation curve and the actual phase-to-phase operation curve. The calculation expression is:

[0085]

[0086] in, It represents the sub-error coefficient at the time t of the nth running data; It is the voltage value of the predicted phase-to-phase operating curve at the nth operating data and the tth moment; is the voltage value collected at time t by the actual phase-to-phase operating curve. The purpose is to use this formula to calculate the sub-error coefficient corresponding to each operating data at each moment, reflecting the degree of deviation between the predicted value and the actual value.

[0087] The error coefficient is calculated based on n sub-error coefficients. The error coefficient calculation expression is:

[0088] ,The purpose is to obtain a comprehensive error coefficient by averaging all the sub-error coefficients, which represents the prediction error of the entire distribution system at a specific moment;

[0089] The correction parameter generation module calculates the error coefficient based on the predicted interphase operation curve and the actual interphase operation curve, generates the corrected interphase operation curve based on the error coefficient, and obtains the interphase imbalance time point and correction value at the future moment based on the corrected interphase operation curve. The specific method is as follows: Step 1, calculates the corrected voltage at time t+L based on the error coefficient, that is, predicts the corrected voltage at the future moment at the current moment. The corrected voltage is the voltage between the operating phases of the distribution system at the future moment, that is, at time t+L, predicted after calculating the error between the predicted interphase operation curve and the actual interphase operation curve. L is an integer greater than 1, and the calculation expression is:

[0090] ,in is the error coefficient, is a compensation coefficient, which is obtained by those skilled in the art based on a large amount of experimental data and is not specifically limited here. D is a correction voltage. The purpose is to adjust the voltage value at the current moment by considering the error coefficient and the compensation coefficient to obtain the correction voltage at time t+L.

[0091] Step 2: Repeat step 1 and continuously update the value of L. Calculate the updated corrected voltage value of t+L in sequence to obtain a series of corrected voltage data points. Connect these data points in chronological order to draw the corrected phase-to-phase operation curve.

[0092] The correction module controls the coordination of the energy storage units based on the imbalance time point and correction value, adjusting the phases by injecting or absorbing energy. Specifically, the correction module obtains the imbalance time point and correction value that will occur in the future, and then, based on the response capability and load forecast of the selected energy storage unit, initiates compensation operations of the energy storage unit in advance for the imbalance time point. The purpose is to enable the energy storage unit to inject or absorb energy in advance based on the predicted voltage fluctuation trend to avoid the impact of the actual imbalance. In this embodiment, the energy storage unit adopts a battery energy storage system or a supercapacitor, such as a device with the function of storing and releasing electrical energy, which is not specifically limited here.

[0093] The parts of the above technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive redundancy.

[0094] The above-described specific embodiments further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above description is merely a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

[0095] The above preset parameters or preset thresholds are all set by those skilled in the art according to actual conditions or obtained through large amounts of data simulation.

[0096] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A method for correcting interphase imbalance in a power distribution system based on energy storage collaboration, characterized in that: include: Collect historical inter-phase operating data of the distribution system and predict inter-phase operating curves based on the historical operating data; Collect the actual phase-to-phase operating curve of the distribution system in real time, calculate the error coefficient based on the predicted phase-to-phase operating curve and the actual phase-to-phase operating curve, and generate the corrected phase-to-phase operating curve based on the error coefficient; Based on the corrected inter-phase operation curve, the inter-phase imbalance time point and correction value are obtained at a future moment, and the imbalance time point and correction value are used as correction parameters of the energy storage unit; The method for obtaining the future interphase imbalance time point and correction value based on the corrected interphase operation curve is as follows: Compare the corrected voltage at a future moment with a preset voltage threshold. If the corrected voltage is greater than or less than the preset voltage threshold, it is determined to be phase unbalanced, and the corresponding moment of the corrected voltage is marked. If the corrected voltage is equal to the preset voltage threshold, it is determined to be phase balanced. Collect multiple sets of historical correction data, where the historical correction data includes imbalance data and correction values corresponding to the imbalance data; A correction recognition model that predicts correction values is trained based on historical correction data; Obtain real-time operating data of the power distribution system, input the real-time operating data into the correction identification model to output the correction value.

2. The method for correcting inter-phase imbalance in a power distribution system based on energy storage collaboration according to claim 1, characterized in that: Methods for predicting interphase operating curves based on historical operating data include: A voltage identification model that predicts phase-to-phase voltage based on historical operating data training; Continuously collect n pieces of distribution system operation data at a preset frequency, where n is an integer greater than 1, input the n pieces of operation data into a voltage identification model in sequence, and output n phase-to-phase voltages; The n phase-to-phase voltages are mapped in a coordinate system with time as the horizontal coordinate and the phase-to-phase voltage as the vertical coordinate according to the order of acquisition time, and the n phase-to-phase voltages are connected in sequence to obtain the predicted phase-to-phase operation curve.

3. The method for correcting inter-phase imbalance in a power distribution system based on energy storage collaboration according to claim 2, characterized in that: The historical operation data includes H groups of condition data and phase-to-phase voltages corresponding to the H groups of condition data; wherein H is a preset amount of training data, H=1, 2, 3...H.

4. The method for correcting inter-phase imbalance in a power distribution system based on energy storage collaboration according to claim 3, characterized in that: The training method of the voltage recognition model is: Converting a set of conditional data and the phase-to-phase voltages corresponding to the conditional data into a set of first eigenvectors, using the first eigenvectors as input to a machine learning model A. The machine learning model A outputs the predicted phase-to-phase voltage for each set of conditional data, and minimizes the sum of the prediction accuracies of all phase-to-phase voltages as a training objective; The machine learning model A is trained until the sum of the prediction accuracies reaches convergence, and the training is stopped. The trained machine learning model A is used as a voltage recognition model. The machine learning model A is a polynomial regression model or a support vector machine model.

5. The method for correcting inter-phase imbalance in a power distribution system based on energy storage collaboration according to claim 1, characterized in that: The predicted interphase operation curve is a curve of the interphase voltage change over time predicted by the power distribution system based on current operation data and combined with historical operation data; The actual phase-to-phase operation curve is a curve drawn by collecting the phase-to-phase voltage changes over time in real time by the power distribution system.

6. The method for correcting inter-phase imbalance in a power distribution system based on energy storage collaboration according to claim 2, characterized in that: The method for calculating the error coefficient based on the predicted phase-to-phase operating curve and the actual phase-to-phase operating curve is: Calculate the sub-error coefficients of the predicted phase-to-phase operating curve and the actual phase-to-phase operating curve. The calculation expression is: ； in, It represents the sub-error coefficient at the time t of the nth running data; It is the voltage value of the predicted phase-to-phase operating curve at the nth operating data and the tth moment; is the voltage value collected at the tth moment of the actual phase-to-phase operation curve; The error coefficient is calculated based on n sub-error coefficients. The error coefficient calculation expression is: ,in is the error coefficient.

7. The method for correcting inter-phase imbalance in a power distribution system based on energy storage coordination according to claim 6, characterized in that: The method for generating the corrected phase-to-phase operating curve based on the error coefficient is: Step 1: Calculate the correction voltage at time t+L based on the error coefficient, where L is an integer greater than 1. The calculation expression is: , is the compensation coefficient, D is the correction voltage; Step 2: Repeat step 1 and continue to update the value of L, and calculate the updated correction voltage value of t+L in sequence to obtain A correction voltage value, , and By connecting the corrected voltage values in chronological order, the corrected phase-to-phase operation curve can be drawn.

8. The method for correcting inter-phase imbalance in a power distribution system based on energy storage coordination according to claim 7, characterized in that: The training method of the correction recognition model is: Converting a set of unbalanced data and correction values corresponding to the unbalanced data into a set of second eigenvectors; The second eigenvector is used as input to a machine learning model B, wherein the machine learning model B outputs a predicted correction value for each set of imbalanced data, and minimizes the sum of the prediction accuracies of all correction values as a training objective; The machine learning model B is trained until the sum of the prediction accuracies reaches convergence, and the training is stopped. The trained machine learning model B is used as the correction recognition model. The machine learning model B is a polynomial regression model or a support vector machine model.

9. A device for correcting interphase imbalance in a power distribution system based on energy storage collaboration, which is used to implement the method for correcting interphase imbalance in a power distribution system based on energy storage collaboration as described in any one of claims 1 to 8, characterized in that: It includes a historical data collection module, an acquisition module, an operation curve prediction module, a correction parameter generation module, a correction module and an energy storage unit; wherein each module is connected by wired and / or wireless means; Historical data collection module, used to collect historical operation data between phases; Acquisition module, used to collect actual phase-to-phase operation curves; An operation curve prediction module is used to predict the phase operation curve based on historical operation data; A correction parameter generation module calculates an error coefficient based on the predicted interphase operation curve and the actual interphase operation curve, generates a correction interphase operation curve based on the error coefficient, and obtains the interphase imbalance time point and correction value at a future moment based on the correction interphase operation curve; The correction module controls the coordination of the energy storage units based on the imbalance time point and the correction value, and adjusts the phases by injecting or absorbing energy.

Images