A high-supply and low-metering special transformer metering method and device with three-phase half-wave rectifier load

Abstract

The present invention discloses a metering method and device for a high-supply, low-metering special transformer with a three-phase half-wave rectifier load, which belongs to the field of electric power metering, wherein the method includes: obtaining a target three-phase AC power supply, and performing load simulation of three-phase half-wave rectification to obtain a first simulated load characteristic; when the constraint of the first target load characteristic is met, issuing a first metering instruction; training to obtain an integrated fusion model of power consumption; activating a metering cabinet based on the first metering instruction to obtain first metering information, as input information of the integrated fusion model of power consumption, to obtain a first simulated power consumption; using the first simulated power consumption as the metered power of three-phase half-wave rectification under the first target load characteristic. The present invention solves the technical problem of inaccurate metering of high-supply, low-metering special transformers with three-phase half-wave rectification loads in the prior art, and achieves the technical effect of accurately simulating the load characteristics of three-phase half-wave rectification and improving the metering accuracy of high-supply, low-metering special transformers.

Description

Technical Field

[0001] The invention relates to the field of electric power metering, in particular to a metering method and device for a high-supply and low-metering special transformer with a three-phase half-wave rectifier load. Background Art

[0002] High-supply, low-metering transformer metering is a commonly used metering method, especially suitable for high-voltage power supply and low-voltage metering scenarios. However, in actual applications, many electrical devices will generate nonlinear loads, among which three-phase half-wave rectifier load is a typical case. This load will cause harmonics and power factor reduction in the power grid, and also bring challenges to accurate metering.

[0003] The existing high-supply and low-metering special transformer metering method is difficult to accurately capture and process the non-sinusoidal waveform generated by half-wave rectification when processing three-phase half-wave rectifier loads, resulting in deviations in metering results. In addition, due to the complexity and variability of load characteristics, the existing simulation method is difficult to fully reflect the actual power consumption, which in turn affects the accuracy of metering. Summary of the invention

[0004] The present invention provides a metering method and device for a high-supply, low-metering special transformer with a three-phase half-wave rectifier load, aiming to solve the technical problem of inaccurate metering of a high-supply, low-metering special transformer with a three-phase half-wave rectifier load in the prior art.

[0005] In view of the above problems, the present invention adopts the following technical solutions.

[0006] The first aspect of the present invention provides a metering method for a high-supply, low-metering special transformer with a three-phase half-wave rectifier load, the method comprising: obtaining a target three-phase AC power supply, and performing a load simulation of three-phase half-wave rectification under the target three-phase AC power supply to obtain a first simulated load characteristic; when the first simulated load characteristic meets the constraints of the first target load characteristic, issuing a first metering instruction; obtaining an integrated fusion model of power consumption based on data training in historical metering records of high-supply, low-metering special transformers; activating a metering cabinet based on the first metering instruction to obtain first metering information, and using the first metering information as input information of the integrated fusion model of power consumption to obtain a first simulated power consumption; using the first simulated power consumption as the metered power of the three-phase half-wave rectifier under the first target load characteristic.

[0007] Another aspect of the present invention provides a metering device for a high-supply, low-meter dedicated transformer with a three-phase half-wave rectifier load, which is used to implement the high-supply, low-meter dedicated transformer metering method, and the device includes: a power supply simulator, which is used to provide a target three-phase AC power supply and is connected to a metering cabinet; the metering cabinet includes: a three-phase thyristor, which is used to set and regulate the conduction angle; a trigger board, which is connected to the three-phase thyristor and is used to trigger the three-phase thyristor to set and regulate the conduction angle; an input control board, which is used to receive an external control signal and activate the trigger board based on the external control signal; in cooperation with the power supply simulator, the high-supply, low-meter dedicated transformer metering of the three-phase half-wave rectifier load is performed through the metering cabinet.

[0008] One or more technical solutions provided by the present invention have at least the following technical effects or advantages:

[0009] By obtaining the target three-phase AC power supply and performing load simulation of three-phase half-wave rectification under this power supply, the first simulated load characteristics are obtained, and a simulation environment is created to reproduce the electrical characteristics of the actual three-phase half-wave rectifier load; when the first simulated load characteristics meet the constraints of the first target load characteristics, a first metering instruction is issued to ensure that the simulated load characteristics are consistent with the behavior of the actual half-wave rectifier load, and subsequent metering is performed only when the simulation achieves the expected effect; an integrated fusion model of electricity consumption is obtained by training based on data in the historical high-supply and low-metering special transformer metering records to improve the accuracy and adaptability of metering; based on the first metering instruction, the metering cabinet is activated to obtain the first metering signal The first metering information is used as the input information of the electricity consumption integrated fusion model to obtain the first simulated electricity consumption, and the real-time metering data is combined with the pre-trained model to improve the accuracy of the metering result; the first simulated electricity consumption is used as the metered electricity of the three-phase half-wave rectifier under the first target load characteristic to determine the final metering result, and the technical solution with comprehensive consideration of the simulated load characteristics and model prediction to improve the metering accuracy solves the technical problem of inaccurate metering of high-supply and low-metering special transformers with three-phase half-wave rectifier loads in the prior art, and achieves the technical effect of accurately simulating the load characteristics of the three-phase half-wave rectifier and improving the metering accuracy of the high-supply and low-metering special transformers.

[0010] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A schematic flow chart of a high-supply and low-metering special transformer metering method with a three-phase half-wave rectifier load is provided for an embodiment of the present invention;

[0012] Figure 2 A structural schematic diagram of a high-supply and low-meter special transformer metering device with a three-phase half-wave rectifier load is provided for an embodiment of the present invention.

[0013] Description of reference numerals: power simulator 10 , metering cabinet 20 , three-phase thyristor 21 , trigger board 22 , input control board 23 . DETAILED DESCRIPTION

[0014] The overall idea of ​​the technical solution provided by the present invention is as follows:

[0015] The embodiment of the present invention provides a high-supply, low-metering special transformer metering method and device with a three-phase half-wave rectifier load. First, a target three-phase AC power supply is obtained, and on this basis, a load simulation of three-phase half-wave rectification is performed to obtain a simulation result consistent with the actual load characteristics, and a first simulated load characteristic is obtained. Secondly, a verification mechanism is introduced. Only when the first simulated load characteristic meets the constraints of the first target load characteristic, the subsequent metering process is triggered to ensure the effectiveness of the simulation. Thirdly, the data in the historical high-supply, low-metering special transformer metering records are used to train an integrated fusion model of power consumption to provide data support for subsequent metering. In the actual metering process, the first metering information obtained in real time is input into the pre-trained integrated fusion model of power consumption to obtain a more accurate first simulated power consumption. Finally, the first simulated power consumption processed by the model is used as the final metering result, that is, the metered power under the first target load characteristic, which effectively improves the accuracy of the metering, thereby solving the problem of insufficient accuracy of traditional metering technology when facing a three-phase half-wave rectifier load.

[0016] After introducing the basic principles of the present invention, various non-limiting implementation methods of the present invention will be specifically described below in conjunction with the accompanying drawings.

[0017] Embodiment 1, as Figure 1 As shown, this embodiment provides a high-supply and low-metering special transformer metering method with a three-phase half-wave rectifier load, the method comprising:

[0018] S1: Obtain a target three-phase AC power supply, and perform a load simulation of three-phase half-wave rectification under the target three-phase AC power supply to obtain a first simulated load characteristic.

[0019] Specifically, first, obtain the target three-phase AC power supply for simulating the actual power supply situation on site. The target three-phase AC power supply is provided by a power simulator. The power simulator provides three-phase AC power with different parameters according to demand, and provides the required power supply environment for subsequent load simulation. After obtaining the target three-phase AC power supply, the three-phase half-wave rectifier load will be simulated under the power supply of the power supply through the relevant components in the metering cabinet, such as three-phase thyristors, trigger boards, etc. Among them, the three-phase thyristors, under the control of the trigger board, adjust their conduction angles to simulate voltage, current and other characteristics similar to the actual three-phase half-wave rectifier load.

[0020] After load simulation, the load characteristics of the simulated three-phase half-wave rectifier load when running under the target three-phase AC power supply can be obtained, which is recorded as the first simulated load characteristics. The first simulated load characteristics characterize the waveform characteristics of the voltage, current, etc. of the simulated load circuit, providing a basis for subsequent judgment of whether the simulated load meets the actual load constraints.

[0021] S2: When the first simulated load characteristic satisfies the constraint of the first target load characteristic, a first metering instruction is issued.

[0022] Specifically, after obtaining the first simulated load characteristic of the simulated three-phase half-wave rectifier load, it is determined whether the first simulated load characteristic meets the characteristic constraint of the actual three-phase half-wave rectifier load, that is, the constraint of the first target load characteristic, thereby ensuring that the three-phase half-wave rectifier load can accurately reflect the operating status of the actual three-phase half-wave rectifier load, providing a reliable basis for high supply and low metering.

[0023] The first target load characteristic is obtained from field measurements or historical data, and characterizes the voltage, current and other characteristic parameters of the actual three-phase half-wave rectifier load during operation. In order to determine whether the first simulated load characteristic meets the constraints of the first target load characteristic, it is achieved by comparing the characteristic parameters of the two. The specific comparison method will be elaborated in detail later. When the first simulated load characteristic meets the constraints of the first target load characteristic, it means that the current simulated three-phase half-wave rectifier load has similar electrical characteristics to the actual three-phase half-wave rectifier load, which can be used for subsequent metering processes. At this time, the first metering instruction is issued to notify the metering cabinet to start the high supply and low metering operation of the simulated three-phase half-wave rectifier load.

[0024] If the first simulated load characteristic fails to meet the constraint of the first target load characteristic, the parameters of the simulated three-phase half-wave rectifier load need to be adjusted until the constraint is met. This adjustment process can be achieved through the trigger board, which will be described in detail in other steps.

[0025] S3: The electricity consumption integration fusion model is obtained by training the data in the historical high-supply and low-metering special transformer metering records.

[0026] Specifically, in order to improve the accuracy of the metering of high-supply and low-metering dedicated transformers, the data in the historical high-supply and low-metering dedicated transformer metering records are used to construct an integrated fusion model of electricity consumption, fully explore the electricity consumption prediction rules contained in the historical data, and accurately estimate the actual electricity consumption of the current simulated three-phase half-wave rectifier load.

[0027] First, obtain the historical high-supply and low-metering special transformer metering records, which contain various metering information (such as active power, reactive power, apparent power, etc.) and the corresponding actual power consumption. Preprocess and extract features of the historical high-supply and low-metering special transformer metering records to form a structured training data set. Then, use the machine learning algorithm to model and train the training data set. Considering the complexity of the power consumption prediction problem, the idea of ​​ensemble learning is adopted to improve the prediction performance by combining multiple basic models. For example, using common ensemble learning algorithms, such as Bagging and Boosting, multiple power consumption prediction sub-models are obtained through parallel training or serial training. After obtaining multiple power consumption prediction sub-models, the prediction results of each sub-model are integrated through a fusion strategy to obtain the final power consumption prediction value. Among them, the fusion strategy includes weighted average, voting, etc. By fusing multiple sub-models, the advantages of each sub-model are fully utilized, the limitations of a single model are reduced, and the generalization ability of the overall prediction is improved. After the training and fusion process, the power consumption integrated fusion model can be obtained, which comprehensively utilizes the information of historical data to estimate power consumption from multiple angles, and has strong robustness and adaptability.

[0028] S4: activating a metering cabinet based on the first metering instruction to obtain first metering information, and using the first metering information as input information of the power consumption integrated fusion model to obtain first simulated power consumption.

[0029] Specifically, when the first metering instruction is issued, the metering cabinet is instructed to perform high supply and low metering on the current simulated three-phase half-wave rectifier load. After receiving the first metering instruction, the metering cabinet begins to collect and record various electrical parameters of the simulated load circuit, and obtains a series of metering information, which is recorded as the first metering information. The first metering information includes but is not limited to the following: active power, reactive power, apparent power, current, voltage, power factor, etc. of the simulated load circuit. The first metering information reflects the power consumption of the simulated three-phase half-wave rectifier load from different angles, and is an important basis for estimating the actual power consumption.

[0030] After obtaining the first metering information, it is input into the trained integrated fusion model of electricity consumption. The integrated fusion model of electricity consumption receives the first metering information as an input feature, and obtains the corresponding estimated electricity consumption value, i.e., the first simulated electricity consumption, through calculation and reasoning. The first simulated electricity consumption reflects the estimated electricity consumption of the high-supply and low-metering transformer in one metering cycle under the current simulated three-phase half-wave rectifier load state. Since the integrated fusion model of electricity consumption comprehensively considers historical data and multi-model fusion, its estimation result has a high degree of credibility, providing strong support for subsequent electricity metering and analysis.

[0031] In practical applications, the collection frequency and metering cycle of the metering cabinet are set according to demand. A higher collection frequency and a shorter metering cycle can provide a more precise estimate of power consumption, but it will also increase the system's computing and storage overhead. Therefore, it is necessary to balance accuracy and efficiency and select appropriate parameter configurations.

[0032] S5: Using the first simulated power consumption as the measured power consumption of three-phase half-wave rectification under the first target load characteristics.

[0033] Specifically, after obtaining the first simulated power consumption, it is used as the metered power consumption of the three-phase half-wave rectifier under the first target load characteristic. The first simulated power consumption is obtained through the power consumption integration fusion model, which reflects the estimated power consumption of the simulated three-phase half-wave rectifier load in the current state. Since it has been ensured that the first simulated load characteristic meets the constraints of the first target load characteristic, the first simulated power consumption accurately reflects the power consumption of the actual three-phase half-wave rectifier load under the same conditions.

[0034] The first simulated power consumption is used as the metered power consumption. In the subsequent power metering and statistical process, this value can be directly used to represent the actual load power consumption without other measurements or estimates. This avoids the difficulty of directly measuring the three-phase half-wave rectifier load in actual scenarios, and also reduces the complexity and cost of the metering device.

[0035] By taking the first simulated power consumption as the metered power consumption of three-phase half-wave rectification under the first target load characteristic, a simple and efficient high-supply and low-metering special transformer metering method is provided, which overcomes the limitations of traditional metering methods and improves metering accuracy.

[0036] Furthermore, the high supply and low metering special transformer metering method of the present invention also includes:

[0037] Start the trigger board according to the first external control signal obtained by the input control board, and start the three-phase thyristor through the trigger board;

[0038] The three-phase thyristor obtains the first simulated load characteristic under the first conduction angle control strategy.

[0039] In a feasible implementation, in order to realize the simulation of the three-phase half-wave rectifier load, the voltage and current of the simulated load circuit are adjusted by controlling the conduction state of the three-phase thyristor. The control of the three-phase thyristor is realized by a trigger board, which receives a control signal from the input control board and triggers the three-phase thyristor according to the instruction of the signal.

[0040] Specifically, first, the input control board obtains the first external control signal. The signal can come from the user's manual input or from other control modules or systems. The first external control signal carries control instructions for the simulated three-phase half-wave rectifier load, such as the setting of the conduction angle, the trigger timing, etc. When the input control board receives the first external control signal, it will start the trigger board according to the content of the signal. After receiving the start instruction, the trigger board starts working and generates a corresponding trigger pulse sequence according to the requirements of the first external control signal. These trigger pulse sequences will be transmitted to the control end of the three-phase thyristor to control its conduction state. Under the control of the trigger board, the three-phase thyristor starts to conduct, so that the simulated load circuit is connected to the power supply. By adjusting the conduction time and angle of the three-phase thyristor, different load characteristics can be simulated.

[0041] The conduction state of the started three-phase thyristor is set according to the first conduction angle control strategy. The first conduction angle control strategy refers to a strategy for regulating the conduction angle of the three-phase thyristor. By reasonably designing and optimizing this strategy, the characteristics of the simulated three-phase half-wave rectifier load can be made closer to the actual three-phase half-wave rectifier load. Among them, the first conduction angle control strategy is implemented based on a pre-established mathematical model or lookup table, which describes the correspondence between the conduction angle and the load characteristics. Given a target load characteristic, the corresponding conduction angle setting value can be obtained through the first conduction angle control strategy. According to the first conduction angle control strategy, the trigger board generates a corresponding trigger pulse sequence to control the conduction of the three-phase thyristor. Under this control, the simulated load circuit exhibits specific voltage and current waveforms, as well as corresponding characteristic parameters such as power and power factor, which constitute the first simulated load characteristics. The first simulated load characteristics reflect the electrical behavior of the simulated load under the current conduction angle setting. If the characteristic is similar to the characteristic of the actual three-phase half-wave rectifier load, it can be considered that the simulated three-phase half-wave rectifier load has performed a good simulation of the actual three-phase half-wave rectifier load under the current state.

[0042] Through the input control board, trigger board and three-phase thyristor, the simulation of three-phase half-wave rectifier load characteristics is realized, which has the advantages of high control accuracy and strong flexibility, and provides a reliable foundation for the subsequent high-supply and low-meter special transformer measurement.

[0043] Furthermore, the high supply and low metering special transformer metering method of the present invention also includes:

[0044] When the first simulated load characteristic does not satisfy the constraint of the first target load characteristic, a first reset instruction is issued, and the first conduction angle control strategy is reset and adjusted based on the first reset instruction.

[0045] In a feasible implementation, when the first simulated load characteristic does not meet the constraints of the first target load characteristic, that is, the simulated three-phase half-wave rectifier load characteristic does not match the first target load characteristic, dynamic optimization of the simulated load characteristic is achieved by resetting and adjusting the first conduction angle control strategy.

[0046] The three-phase thyristor is controlled by the first conduction angle control strategy to obtain the first simulated load characteristic. However, since the working conditions of the actual load may fluctuate, or the first conduction angle control strategy itself is insufficient, it is possible that there is a large deviation between the first simulated load characteristic and the first target load characteristic. In order to solve this deviation, it is first determined whether the first simulated load characteristic meets the constraints of the first target load characteristic, and by comparing the key characteristic parameters of the two, such as the waveform, amplitude, phase, etc. of the voltage and current, it is calculated whether the deviation is within the allowable range. If the first simulated load characteristic does not meet the constraints of the first target load characteristic, that is, the deviation between the two exceeds the preset threshold, it is considered that the current simulation effect is not ideal, and the first conduction angle control strategy needs to be adjusted. At this time. A first reset instruction is issued to trigger the reset operation of the conduction angle control strategy. The first reset instruction may contain a variety of control information, such as the direction, amplitude, step size, etc. of the adjustment, which is dynamically generated according to factors such as the size and sign of the deviation, the effect of the historical adjustment, etc. The purpose of the first reset instruction is to guide the targeted modification of the first conduction angle control strategy to reduce the difference between the simulated load characteristic and the target load characteristic.

[0047] After receiving the first reset instruction, the trigger board will adjust the parameters of the first conduction angle control strategy according to the content of the instruction. For example, modify the conduction angle, conduction time, trigger timing, etc. of the three-phase thyristor. The amplitude and direction of the adjustment are determined by the first reset instruction to ensure that the effect of the adjustment is in the direction of reducing the deviation. The first conduction angle control strategy after reset adjustment will be reapplied to the control of the three-phase thyristor. Under the new strategy, the electrical characteristics of the simulated load circuit will change, and an updated simulated load characteristic will be obtained. After one or more adjustments, the new simulated load characteristics will be closer to the target load characteristics and meet its constraints. Among them, reset adjustment is an iterative process. If the simulated load characteristics still do not meet the requirements after one adjustment, a new reset instruction is generated according to the new deviation situation, and the next round of adjustment is performed until the simulated load characteristics meet the constraints of the target load characteristics.

[0048] Through the adaptive adjustment mechanism, the matching accuracy of the simulated load characteristics is improved, the fluctuation of the actual load conditions is effectively responded to, and it is ensured that the simulated load can better simulate the actual load under different conditions, thereby providing a more reliable foundation for the subsequent high-supply and low-metering special transformer metering.

[0049] Furthermore, the high supply and low metering special transformer metering method of the present invention also includes:

[0050] Acquire a first simulation record of the first simulation load characteristic, the first simulation record comprising a first simulation voltage waveform and a first simulation current waveform;

[0051] Reading a predetermined waveform feature, and respectively acquiring a first waveform feature value set of the first analog voltage waveform and a second waveform feature value set of the first analog current waveform based on the predetermined waveform feature;

[0052] Obtaining a first actual record of a first actual load characteristic, the first actual record comprising a first actual voltage waveform and a first actual current waveform;

[0053] Respectively acquiring a third waveform characteristic value set of the first actual voltage waveform and a fourth waveform characteristic value set of the first actual current waveform;

[0054] Comparing the first waveform feature value set with the third waveform feature value set, and the second waveform feature value set with the fourth waveform feature value set in sequence to obtain a comparison result;

[0055] A determination is made based on the comparison result whether the first simulated load characteristic satisfies a constraint of the first target load characteristic.

[0056] In a preferred embodiment, by extracting and comparing the voltage and current waveforms of the simulated three-phase half-wave rectifier load and the actual three-phase half-wave rectifier load, a quantitative evaluation of the simulation effect is achieved. The three-phase thyristor is controlled by the first conduction angle control strategy to obtain the first simulated load characteristic. In order to analyze and evaluate the characteristic, the corresponding voltage and current waveform data, i.e., the first simulation record, are obtained. Voltage and current sensors are set in the simulated load circuit to collect the voltage and current signals of the simulated three-phase half-wave rectifier load in real time during operation. After amplification, filtering, sampling and other processing, these signals are processed to obtain the first digital simulated voltage waveform and the first simulated current waveform. The data points of these two waveforms are arranged in chronological order, reflecting the voltage and current changes of the simulated load in an operation cycle, as the first simulation record. Then, the predetermined waveform characteristics are read from the memory. The predetermined waveform characteristics refer to a set of quantitative indicators reflecting the waveform morphology that are predefined according to the electrical characteristics of the three-phase half-wave rectifier load, such as the peak value, valley value, peak-to-valley difference, average value, effective value, fundamental amplitude, total harmonic distortion rate, etc. of the waveform, which describe the key attributes of the waveform from different angles. Subsequently, the first analog voltage waveform and the first analog current waveform are analyzed and calculated based on the predetermined waveform characteristics to obtain a first waveform characteristic value set and a second waveform characteristic value set. The two sets respectively contain specific values ​​of multiple characteristic indicators, which are used to characterize the morphological characteristics of the first analog voltage waveform and the first analog current waveform.

[0057] Next, the voltage and current waveform data of the actual three-phase half-wave rectifier load during operation are obtained through field measurement or historical archives, including the first actual voltage waveform and the first actual current waveform, to form a first actual record. The first actual voltage waveform and the first actual current waveform reflect the law of change of the electrical quantity of the actual three-phase half-wave rectifier load during operation. Then, the first actual voltage waveform and the first actual current waveform are feature extracted using predetermined waveform features to obtain a third waveform feature value set and a fourth waveform feature value set. These two sets reflect the morphological characteristics of the actual load waveform, and are compared with the feature value set of the simulated load waveform to evaluate the quality of the simulation effect.

[0058] Subsequently, the characteristic value set of the simulated load waveform is compared with the characteristic value set of the actual load waveform one by one to obtain a comparison result. Specifically, the difference between each characteristic value in the first waveform characteristic value set and the corresponding characteristic value in the third waveform characteristic value set is calculated to obtain a set of characteristic value deviations of the voltage waveform; similarly, the difference between the second waveform characteristic value set and the fourth waveform characteristic value set is calculated to obtain a set of characteristic value deviations of the current waveform. The characteristic value deviation of the voltage waveform and the characteristic value deviation of the current waveform are used as the comparison result, wherein the magnitude of the deviation reflects the degree of coincidence between the simulated waveform and the actual waveform in different characteristics. Thereafter, the characteristic value deviations in the obtained comparison result are comprehensively analyzed to determine whether the first simulated load characteristic meets the constraints of the first target load characteristic. Specifically, for each characteristic value deviation, a corresponding deviation threshold is set. When a certain deviation is less than or equal to its threshold, it is considered that the simulation effect on the feature meets the requirements; otherwise, it is considered that the requirements are not met. If all characteristic value deviations meet their corresponding deviation thresholds, it is considered that the first simulated load characteristic meets the constraints of the first target load characteristic as a whole, and the simulation effect is ideal. If the deviation of any characteristic value exceeds the threshold, it is considered that there is a large difference between the simulated load characteristics and the target load characteristics, and optimization adjustment is required.

[0059] By extracting and comparing waveform features, the degree of fit between the simulated load characteristics and the actual load characteristics is quantitatively evaluated, providing an objective basis for subsequent simulation optimization and helping to improve simulation accuracy and reliability. At the same time, by introducing multiple waveform features, the characteristics of voltage and current in different aspects are comprehensively considered, which can fully reflect the pros and cons of the simulation effect and provide reliable guarantee for high-supply and low-meter special transformer measurement.

[0060] Furthermore, the high supply and low metering special transformer metering method of the present invention also includes:

[0061] The predetermined waveform characteristics include waveform width characteristics and waveform height characteristics. The waveform width characteristics include maximum waveform width, minimum waveform width and waveform width difference. The waveform height characteristics include maximum waveform height, minimum waveform height and waveform height difference.

[0062] In a preferred embodiment, the predetermined waveform features are divided into two categories: waveform width features and waveform height features, and three specific quantitative indicators are defined in each category.

[0063] The waveform width feature is used to describe the span and changes of the waveform in the time dimension, including the maximum waveform width, the minimum waveform width and the waveform width difference. Among them, the maximum waveform width refers to the maximum value of the duration of the waveform in one cycle, reflecting the maximum duration of the waveform; the minimum waveform width refers to the minimum value of the duration of the waveform in one cycle, reflecting the minimum duration of the waveform; the waveform width difference refers to the difference between the maximum waveform width and the minimum waveform width, reflecting the range of change of the waveform duration. Through these three indicators, the expansion and contraction characteristics of the waveform in the time dimension are characterized, and the extreme values ​​and variation range of the waveform width are captured.

[0064] The waveform height feature is used to describe the size and change of the waveform in the amplitude dimension, including the maximum waveform height, minimum waveform height and waveform height difference. Among them, the maximum waveform height refers to the maximum value of the waveform amplitude in a cycle, reflecting the peak level of the waveform; the minimum waveform height refers to the minimum value of the waveform amplitude in a cycle, reflecting the valley level of the waveform; the waveform height difference refers to the difference between the maximum waveform height and the minimum waveform height, reflecting the range of change of the waveform amplitude. Through these three indicators, the fluctuation characteristics of the waveform in the amplitude dimension are characterized, and the extreme values ​​and change range of the waveform height are captured.

[0065] The waveform width feature and the waveform height feature are combined to comprehensively describe the morphological features of the waveform. In practical applications, some of the indicators can be selected or other indicators can be added according to specific needs to meet the waveform feature extraction requirements in different scenarios. By introducing waveform width features and waveform height features, the content of the predetermined waveform features is enriched, providing a more comprehensive basis for subsequent waveform comparison and simulation effect evaluation. Through the multi-dimensional and multi-indicator waveform feature extraction method, the morphological characteristics of the waveform can be more finely portrayed, the matching degree between the simulated load characteristics and the actual load characteristics can be improved, and a strong guarantee can be provided for the accuracy of high-supply and low-metering special transformer metering.

[0066] Furthermore, the high supply and low metering special transformer metering method of the present invention also includes:

[0067] Extracting a first historical record from the historical high-supply and low-metering special transformer metering record, the first historical record including first historical metering information and first historical power consumption;

[0068] Dividing the training data group formed based on the first historical metering information and the first historical power consumption to obtain a division result, wherein the division result includes a first data group and a second data group;

[0069] The first data group is trained based on the recurrent neural network principle to obtain a first model, and the second data group is trained based on the gradient boosting decision tree principle to obtain a second model;

[0070] Performing supervised training on the first output result of the first model, the second output result of the second model, and the first historical power consumption to obtain a meta-learner;

[0071] The first model and the second model are used as initial learners and constructed with the meta-learner to obtain the electricity consumption integrated fusion model.

[0072] In a preferred embodiment, an accurate and robust electricity consumption prediction model is achieved by extracting, dividing and training historical data, combining recurrent neural networks, gradient boosting decision trees and meta-learning techniques.

[0073] First, extract valid training samples from the historical high-supply and low-metering special transformer metering records as the first historical records. The first historical records contain two types of information: the first historical metering information and the first historical electricity consumption. Among them, the first historical metering information refers to the power supply parameters at a certain moment in history, such as voltage, current, power, etc.; the first historical electricity consumption refers to the actual electricity consumption corresponding to that moment. The first historical metering information and the first historical electricity consumption constitute the input and output of the model training. After the first historical records are organized into a training data group, they are divided to obtain two sub-data sets. Specifically, a division strategy, such as random division, stratified sampling, etc., is used to divide the training data group into a first data group and a second data group. These two data groups will be used to train different sub-models respectively to achieve the effect of integrated learning.

[0074] After obtaining the first data group and the second data group, two algorithms, recurrent neural network and gradient boosting decision tree, are used to train them respectively to obtain two sub-models. Recurrent neural network is a deep learning model suitable for processing sequence data. By introducing a loop structure and a gating mechanism, it can effectively capture the time dependency of data. The first data group is trained using a recurrent neural network, and the first model is obtained by learning the mapping relationship between historical metering information and power consumption. The first model can predict the corresponding power consumption based on the input metering information. The gradient boosting decision tree is a machine learning algorithm based on ensemble learning. It iteratively constructs multiple decision trees and uses the gradient descent method to optimize the parameters of each tree to continuously improve the prediction accuracy. The second data group is trained using a gradient boosting decision tree, and the second model is obtained by learning the nonlinear relationship between historical metering information and power consumption. The second model can predict the corresponding power consumption based on the input metering information.

[0075] After obtaining the first model and the second model, a meta-learner is constructed to fuse and correct the outputs of the two sub-models. Specifically, the metering information in the first historical record is used as input, and the first model and the second model are used for prediction respectively to obtain the first output result and the second output result. Then, the two output results are used together with the corresponding first historical power consumption as training data for the meta-learner for supervised learning. By learning the relationship between the output of the sub-model and the actual power consumption, the meta-learner can adaptively adjust the weight of each sub-model to obtain the optimal fusion strategy. After obtaining the first model, the second model and the meta-learner, they are combined to form a complete integrated fusion model of power consumption. Specifically, the first model and the second model are used as initial learners to predict the input metering information respectively to obtain two preliminary power consumption prediction values. Then, the two prediction values ​​are input into the meta-learner, and the meta-learner performs weighted fusion and correction to obtain the final power consumption prediction result. Through the integrated fusion model structure, the advantages of different algorithms can be comprehensively utilized to improve the accuracy and generalization ability of prediction.

[0076] By making full use of historical data and combining technologies such as recurrent neural networks, gradient boosting decision trees and meta-learning, a multi-level, adaptive electricity consumption prediction model is constructed. Compared with the traditional single model, this integrated fusion model can more comprehensively mine the laws in the data, accurately model complex electricity consumption behaviors, and provide reliable electricity consumption estimates for high-supply and low-metering special transformer metering.

[0077] Furthermore, the high supply and low metering special transformer metering method of the present invention also includes:

[0078] The first historical metering information includes first historical active power, first historical reactive power, first historical apparent power and first historical power factor.

[0079] In a feasible implementation, the first historical metering information includes at least the first historical active power, the first historical reactive power, the first historical apparent power and the first historical power factor. The first historical active power refers to the active power value actually consumed at a certain moment in the historical high-supply and low-metering special transformer metering record, reflecting the actual electric energy consumed by the load. The first historical reactive power refers to the reactive power value actually consumed at a certain moment in the historical high-supply and low-metering special transformer metering record, reflecting the reactive power demand of the load on the power grid during operation. Although it does not directly consume electric energy, it will affect the quality of electric energy and transmission efficiency. The first historical apparent power refers to the apparent power value actually shown at a certain moment in the historical high-supply and low-metering special transformer metering record. It is the vector sum of active power and reactive power, reflecting the overall power demand of the load on the power grid. The first historical power factor refers to the power factor value actually shown at a certain moment in the historical high-supply and low-metering special transformer metering record, reflecting the proportional relationship between active power and apparent power, and is an important indicator for measuring the efficiency of electric energy utilization.

[0080] The first historical active power, the first historical reactive power, the first historical apparent power and the first historical power factor are included in the first historical metering information to reflect the energy consumption characteristics of the three-phase half-wave rectifier load in historical operation from different angles. When constructing the integrated fusion model of power consumption, the first historical active power, the first historical reactive power, the first historical apparent power and the first historical power factor are used as input features to provide the model with richer and more comprehensive information, which helps to improve the accuracy and reliability of power consumption prediction and make it applicable to the three-phase half-wave rectifier load under different operating conditions.

[0081] Furthermore, the high supply and low metering special transformer metering method of the present invention also includes:

[0082] Obtaining a target power supply of the target three-phase AC power source;

[0083] Obtaining a deviation between the first simulated power consumption and the target power consumption, recorded as a first simulated deviation;

[0084] The first simulation deviation is used to characterize the degree of metering error of the high-supply and low-metering special transformer under the first target load characteristic.

[0085] In a feasible implementation, by introducing the target power supply, the deviation between the first simulated power consumption and the target power consumption is calculated to obtain a first simulated deviation, which is used to characterize the metering error degree of the high-supply and low-metering special transformer under the first target load characteristics.

[0086] When performing high-supply and low-metering special transformer metering, in addition to obtaining the power consumption of the simulated load, the target power supply of the target three-phase AC power supply is obtained as a reference for evaluating the metering error. The target power supply refers to the amount of electric energy actually provided by the target three-phase AC power supply in a metering cycle, which is obtained by installing a special power metering device on the power supply side, such as a high-precision power meter, power quality analyzer, etc. The target power supply reflects the power output level of the actual power supply under the current operating conditions. Ideally, the first simulated power consumption should be basically consistent with the target power supply. However, due to the influence of factors such as metering errors and line losses, there is often a certain deviation between the two.

[0087] After obtaining the target power supply, compare it with the obtained first simulated power consumption, calculate the difference between the two, and obtain the first simulated deviation. The first simulated deviation quantifies the error degree of the high-supply and low-metering special transformer measurement under the first target load characteristics. If the first simulated deviation is a positive value, it means that the first simulated power consumption is greater than the target power supply, that is, the power consumption of the simulated load is estimated to be too high; conversely, if the first simulated deviation is a negative value, it means that the first simulated power consumption is less than the target power supply, that is, the power consumption of the simulated load is estimated to be too low. The larger the absolute value of the first simulated deviation, the greater the measurement error.

[0088] By introducing the first simulation deviation, the accuracy of the high-supply and low-metering transformer can be intuitively evaluated. In practical applications, an error threshold can be set. When the absolute value of the first simulation deviation is less than the threshold, the metering error is considered to be within an acceptable range; conversely, if the absolute value of the first simulation deviation exceeds the threshold, the metering method needs to be optimized and improved. By making full use of the metering information on the power supply side, comparing the deviation between the simulated power consumption and the actual power supply, and quantitatively evaluating the accuracy of the metering, it provides an important basis for optimizing the metering method and improving the metering reliability.

[0089] Embodiment 2, as Figure 2 As shown, this embodiment provides a high-power-low-meter dedicated transformer metering device with three-phase half-wave rectifier load, which is used to implement the high-power-low-meter dedicated transformer metering method described in Example 1. The device is composed of a power simulator 10 and a metering cabinet 20.

[0090] A power simulator 10 is used to provide a target three-phase AC power supply and is connected to a metering cabinet.

[0091] Metering cabinet 20, the metering cabinet comprises:

[0092] Three-phase thyristor 21, the three-phase thyristor is used to set and control the conduction angle;

[0093] A trigger board 22, which is connected to the three-phase thyristor and is used to trigger the three-phase thyristor to set and adjust the conduction angle;

[0094] The input control board 23 is used to receive an external control signal and activate the trigger board based on the external control signal.

[0095] In cooperation with the power simulator, the metering cabinet is used to measure the high supply and low metering of the three-phase half-wave rectified load.

[0096] In this embodiment, the high-supply low-meter dedicated transformer metering device includes a power simulator 10 and a meter cabinet 20. The power simulator 10 is used to provide a three-phase AC power supply and connect to the meter cabinet to simulate the actual power supply situation on site. The power simulator 10 provides three-phase AC power with different parameters according to demand, such as voltage amplitude, frequency, etc., to meet the metering requirements under different working conditions. The meter cabinet 20 is the core component of the entire device, which contains multiple functional modules for simulating and metering three-phase half-wave rectifier loads.

[0097] Specifically, the metering cabinet 20 includes a three-phase thyristor 21, a trigger board 22 and an input control board 23. Among them, the three-phase thyristor 21 is a controllable rectifier device. By adjusting its conduction angle, the waveform of the output voltage and current is changed, thereby simulating the characteristics of the three-phase half-wave rectifier load. The trigger board 22 is connected to the three-phase thyristor 21, and is used to provide a trigger pulse to the three-phase thyristor 21 to control its conduction and shutdown. The trigger board 22 uses a 32-bit ARM chip microcomputer as the control core, and has strong computing and control capabilities. By adjusting the trigger pulse output by the trigger board 22, the conduction angle of the three-phase thyristor 21 can be accurately controlled, thereby realizing the adjustment of the load voltage, current, and power, and simulating the characteristics of the on-site half-wave rectifier load. The input control board 23 is used to receive an external control signal and activate the trigger board 22 according to the instruction of the control signal. The control input of the input control board 23 is a DC voltage signal of DC1~5V, and different voltage amplitudes correspond to different conduction angle settings. The conduction angle can be adjusted from 0 to 180 degrees, which can cover the load characteristics under various actual working conditions.

[0098] In actual operation, the power simulator 10 provides a stable three-phase AC power supply and is connected to the meter cabinet 20. The input control board 23 receives the external DC control voltage, converts it into the corresponding conduction angle control instruction, and sends it to the trigger board 22. According to the conduction angle control instruction, the trigger board 22 uses the ARM single-chip microcomputer to accurately calculate the phase and width of the trigger pulse, and outputs it to the control end of the three-phase thyristor 21 to achieve accurate control of its conduction angle and simulate the voltage and current characteristics of the three-phase half-wave rectifier load. The meter cabinet 20 is also provided with a voltage and current sensor for collecting the voltage and current signals of the simulated load circuit. Considering that the actual load may have a DC component, a transformer with anti-DC component is selected as a current sensor to ensure the accuracy of the measurement result. After amplification, filtering, A / D conversion and other processing, the voltage and current signals are processed to obtain digital voltage and current waveform data for subsequent metering analysis and power consumption estimation.

[0099] Through the cooperation of the power simulator 10 and the metering cabinet 20, the three-phase half-wave rectifier load can be accurately simulated and measured. Among them, the trigger board uses a high-performance ARM microcontroller for precise control, and the input control board supports a wide range of conduction angle adjustment, which can flexibly simulate various load characteristics. At the same time, the high-supply low-meter special transformer metering also uses a transformer with DC component resistance to improve the accuracy of current measurement, thereby effectively improving the accuracy and reliability of the high-supply low-meter special transformer metering.

[0100] Any step of the method described above may be stored as a computer instruction or program in an unlimited computer memory, and may be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present invention, without any unnecessary restrictions.

[0101] Furthermore, the first or second mentioned above may not only represent an order relationship, but may also represent a specific concept, and / or refer to the selection of multiple elements individually or in whole. Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalents, the present invention is intended to include these changes and variations.

Claims

1. A high supply and low metering special transformer metering method with three-phase half-wave rectifier load, characterized in that: include: Obtain a target three-phase AC power supply, and perform a load simulation of three-phase half-wave rectification under the target three-phase AC power supply to obtain a first simulated load characteristic; When the first simulated load characteristic satisfies the constraint of the first target load characteristic, issuing a first metering instruction; The integrated fusion model of electricity consumption is obtained by training the data in the historical high-supply and low-metering special transformer metering records; activating the metering cabinet based on the first metering instruction to obtain first metering information, and using the first metering information as input information of the power consumption integrated fusion model to obtain first simulated power consumption; Using the first simulated power consumption as the measured power consumption of three-phase half-wave rectification under the first target load characteristic; Wherein, the high supply and low metering special transformer metering method with three-phase half-wave rectifier load also includes: Acquire a first simulation record of the first simulation load characteristic, the first simulation record comprising a first simulation voltage waveform and a first simulation current waveform; Reading a predetermined waveform feature, and respectively acquiring a first waveform feature value set of the first analog voltage waveform and a second waveform feature value set of the first analog current waveform based on the predetermined waveform feature; Obtaining a first actual record of a first actual load characteristic, the first actual record comprising a first actual voltage waveform and a first actual current waveform; Respectively acquiring a third waveform characteristic value set of the first actual voltage waveform and a fourth waveform characteristic value set of the first actual current waveform; Comparing the first waveform feature value set with the third waveform feature value set, and the second waveform feature value set with the fourth waveform feature value set in sequence to obtain a comparison result; Determining whether the first simulated load characteristic satisfies the constraint of the first target load characteristic according to the comparison result; The predetermined waveform characteristics include waveform width characteristics and waveform height characteristics, wherein the waveform width characteristics include the maximum waveform width, the minimum waveform width and the waveform width difference, and the waveform height characteristics include the maximum waveform height, the minimum waveform height and the waveform height difference; Wherein, the high supply and low metering special transformer metering method with three-phase half-wave rectifier load also includes: Start the trigger board according to the first external control signal obtained by the input control board, and start the three-phase thyristor through the trigger board; The three-phase thyristor obtains the first simulated load characteristic under the first conduction angle control strategy; When the first simulated load characteristic does not satisfy the constraint of the first target load characteristic, a first reset instruction is issued, and the first conduction angle control strategy is reset and adjusted based on the first reset instruction.

2. According to claim 1, a high supply and low metering special transformer metering method with three-phase half-wave rectifier load is characterized in that: Also includes: Extracting a first historical record from the historical high-supply and low-metering special transformer metering record, the first historical record including first historical metering information and first historical power consumption; Dividing the training data group formed based on the first historical metering information and the first historical power consumption to obtain a division result, wherein the division result includes a first data group and a second data group; The first data group is trained based on the recurrent neural network principle to obtain a first model, and the second data group is trained based on the gradient boosting decision tree principle to obtain a second model; Performing supervised training on the first output result of the first model, the second output result of the second model, and the first historical power consumption to obtain a meta-learner; The first model and the second model are used as initial learners and constructed with the meta-learner to obtain the electricity consumption integrated fusion model.

3. A high supply and low metering special transformer metering method with three-phase half-wave rectifier load according to claim 2, characterized in that: The first historical metering information includes first historical active power, first historical reactive power, first historical apparent power and first historical power factor.

4. The high supply and low metering special transformer metering method with three-phase half-wave rectifier load according to claim 1 is characterized in that: Also includes: Obtaining a target power supply of the target three-phase AC power source; Obtaining a deviation between the first simulated power consumption and the target power consumption, recorded as a first simulated deviation; The first simulation deviation is used to characterize the degree of metering error of the high-supply and low-metering special transformer under the first target load characteristic.

5. A high-supply low-meter dedicated transformer metering device with three-phase half-wave rectifier load, used to implement the high-supply low-meter dedicated transformer metering method according to any one of claims 1 to 4, characterized in that: include: A power simulator, which is used to provide a target three-phase AC power supply and is connected to a metering cabinet; Metering cabinet, including three-phase thyristor, trigger board and input control board: Three-phase thyristor, the three-phase thyristor is used to set and control the conduction angle; A trigger board, the trigger board is connected to the three-phase thyristor and is used to trigger the three-phase thyristor to set and regulate the conduction angle; an input control board, the input control board being used to receive an external control signal and activate the trigger board based on the external control signal; In cooperation with the power simulator, the metering cabinet is used to measure the high supply and low metering of the three-phase half-wave rectified load.

6. A high supply and low metering transformer metering device with three-phase half-wave rectifier load according to claim 5, characterized in that: The adjustment range of the conduction angle is 0~180 degrees.

Images