An amorphous alloy transformer energy efficiency optimization control system for smart grid

Abstract

The application discloses an amorphous alloy transformer energy efficiency optimization control system for smart grids, which collects transformer electrical quantities, leakage magnetic flux, vibration, temperature and switch state, applies controlled perturbation of positive and negative integral to offset each other in a stable operation time window, separates core loss, winding loss and harmonic additional loss, and constructs a dynamic loss surface; according to the gate voltage integral, current attenuation and gate phase, the residual magnetism is estimated, and the application relates to the technical field of transformer operation optimization. The amorphous alloy transformer energy efficiency optimization control system for smart grids can reduce the influence of load natural fluctuation and magnetic flux bias on the identification result, realize online separation of core loss, winding loss and harmonic additional loss, and obtain incremental active power, excitation current, core leakage magnetic flux and magnetostriction vibration response under the condition that the target amorphous alloy transformer maintains normal power supply.

Description

Technical Field

[0001] This invention relates to the field of transformer operation optimization, and more specifically, to an energy efficiency optimization control system for amorphous alloy transformers used in smart grids. Background Technology

[0002] Amorphous alloy transformers, characterized by low no-load losses, have been gradually applied in smart distribution networks with significant load fluctuations and a high proportion of renewable energy integration. In existing technologies, CN110008645A discloses a transformer loss calculation method that calculates core and winding losses through no-load tests, JA dynamic models, and harmonic winding models. However, its results mainly rely on pre-built models, making it difficult to continuously correct actual losses using controlled responses under uninterrupted power supply conditions. CN102244384B discloses an optimized operation method for main transformers based on economic equivalence analysis, which can select operating schemes and adjust switching states based on loss analysis results. However, it does not include the inrush current energy, magnetization recovery energy, and thermal recovery losses caused by switching on and off in the energy-saving benefits. CN103238197A and CN112564039B disclose inrush current suppression schemes based on voltage integral or residual magnetism state to determine the closing phase, but they do not form a unified closed loop by integrating residual magnetism state with dynamic loss modeling and candidate combination evaluation. Therefore, existing technologies cannot simultaneously achieve online identification of dynamic losses, quantification of switching transient energy, and full-cycle net energy saving control, which can easily lead to problems such as predicted energy saving but actual increased energy consumption or frequent switching. Summary of the Invention

[0003] The purpose of this invention is to provide an energy efficiency optimization control system for amorphous alloy transformers used in smart grids, which solves the problems of existing control methods that make it difficult to continuously identify actual losses under uninterrupted power supply conditions, fail to include switching transient energy in energy-saving benefits, and are disconnected from steady-state loss optimization and residual magnetism closing control.

[0004] This invention achieves the above objectives through the following technical solution: an energy efficiency optimization control system for amorphous alloy transformers used in smart grids, the system comprising: The system includes a data acquisition module, a perturbation identification module, a loss modeling module, a transient energy management module, a candidate combination generation module, an energy efficiency decision-making module, a switching control module, and a feedback correction module. The data acquisition module is used to collect electrical quantities, core leakage flux, magnetostrictive vibration, core temperature, winding temperature, tap position and circuit breaker status of amorphous alloy transformers, and to perform timing synchronization. The perturbation identification module is used to determine the identification time window when the target amorphous alloy transformer meets the load, voltage and harmonic stability conditions, and to apply a controlled perturbation whose difference between the positive volt-second integral and the negative volt-second integral does not exceed the volt-second balance threshold. The loss modeling module is used to establish a dynamic loss surface between the operating variables and the losses based on the incremental active power, excitation current, core leakage flux and magnetostrictive vibration response before, during and after the perturbation, and the winding resistance is corrected according to the winding temperature. It separates the core loss, winding loss and harmonic additional loss. The transient energy management module is used to determine the residual magnetism state based on the voltage integral, current decay characteristics and opening phase of each phase before and after the circuit breaker is opened, and to associate the residual magnetism state with the outage duration, closing phase, excitation inrush current energy, magnetization recovery energy and thermal recovery loss to form a transient energy ledger. The candidate combination generation module is used to generate candidate operation combinations based on the load forecast results and new energy output forecast results within the scheduling cycle. The energy efficiency decision module is used to determine the steady-state cumulative loss of candidate operating combinations based on the dynamic loss surface, determine the transient additional energy for combination switching based on the transient energy ledger, subtract the transient additional energy from the steady-state cumulative loss difference between the current operating combination and the candidate operating combination to obtain the net energy saving over the entire cycle, and determine the target operating combination. The switching control module is used to determine the closing phase and closing sequence based on the residual magnetism state of the amorphous alloy transformer to be switched on. The feedback correction module is used to correct the dynamic loss surface, transient energy ledger, switching threshold, and minimum hold time based on the actual steady-state loss, actual inrush current energy, and actual bus voltage disturbance after switching.

[0005] Furthermore, the data acquisition module includes: Electrical quantity acquisition unit, status quantity acquisition unit, time synchronization unit, and data validity processing unit; The electrical quantity acquisition unit is used to acquire the phase voltage, phase current, active power, reactive power, voltage harmonic distortion rate, and current harmonic distortion rate of the primary and secondary sides. The status acquisition unit is used to acquire core leakage flux, magnetostrictive vibration, core temperature, winding temperature, tap position, and circuit breaker status. The time synchronization unit is used to add a unified timestamp to the collected data and perform resampling and time alignment according to the identified sampling period; The data validity processing unit is used to interpolate data whose continuous missing duration does not exceed the missing tolerance, and to mark the time period in which the continuous missing duration exceeds the missing tolerance as an invalid identification time period.

[0006] Furthermore, the perturbation identification module includes: The system includes a steady-state determination unit, a time window determination unit, a perturbation generation unit, and a perturbation execution unit. The steady state judgment unit is used to calculate the load rate range, the effective value range of the primary side voltage, and the voltage harmonic distortion rate range during the observation period. The time window determination unit is used to determine the time period that continuously meets the stability condition as the identification time window when the three ranges do not exceed the corresponding stability thresholds. The perturbation generation unit is used to generate perturbation commands that sequentially include positive perturbation segments and negative perturbation segments, such that the ratio of the absolute value of the sum of the voltage time integrals of the two segments to the sum of the absolute values ​​of the voltage time integrals of the two segments does not exceed the volt-second balance threshold. The perturbation execution unit is used to execute the perturbation command through the energy storage converter or active filter, and when the core leakage magnetic baseline deviation exceeds the magnetic flux recovery threshold after the perturbation ends, the current identification data is marked as invalid.

[0007] Furthermore, the loss modeling module includes: Response extraction unit, total loss calculation unit, winding loss calculation unit, harmonic additional loss calculation unit, and core loss calculation unit; The response extraction unit is used to obtain the active power, excitation current, core leakage flux, and magnetostrictive vibration characteristics of the reference section, positive perturbation section, negative perturbation section, and recovery section. The total loss calculation unit is used to determine the total loss by the difference between the active power on the primary side and the active power on the secondary side. The winding loss calculation unit is used to determine the winding loss based on the winding material temperature coefficient, the winding resistance at the reference temperature, the current winding temperature, and the effective value of the current in each phase. The harmonic additional loss calculation unit is used to determine the harmonic additional loss based on the harmonic current and the winding AC resistance at the corresponding frequency. The core loss calculation unit is used to deduct winding losses and harmonic additional losses from the total loss, and the symmetric deviation of the excitation current, core leakage flux and magnetostrictive vibration response under positive and negative perturbations does not exceed the response deviation threshold as a valid condition for the identification result.

[0008] Furthermore, the dynamic loss surface takes load rate, core temperature, winding temperature, voltage harmonic distortion rate, current harmonic distortion rate, three-phase imbalance, tap position and reactive power compensation as input variables, and core loss, winding loss, harmonic additional loss and total loss as output variables. The loss modeling module is also used to write the input variables and output variables obtained from each effective identification into the loss sample library. When the input variable to be evaluated is within the coverage area of ​​the loss sample, it performs partition interpolation based on the adjacent loss samples. When the input variable to be evaluated is outside the coverage area, or when the number of times the deviation between the actual total loss and the predicted total loss exceeds the loss deviation threshold reaches the update number, a new controlled perturbation is triggered and the dynamic loss surface is updated using the newly added loss samples.

[0009] Furthermore, the transient energy management module includes: The system includes a tripping recording unit, a residual magnetism estimation unit, a transient energy calculation unit, and a ledger writing unit. The tripping recording unit is used to record the voltage of each phase terminal, the current of each phase, the tripping phase, and the current decay time before and after tripping. The residual magnetism estimation unit is used to integrate the voltage of each phase terminal within the integration interval formed by the opening determination time and the convergence time of each phase terminal voltage, determine the direction of residual magnetism based on the sign of the integration result, and determine the residual magnetism amplitude based on the absolute value of the integration result and the current decay time. The transient energy calculation unit is used to integrate the product of the instantaneous voltage and instantaneous current after closing within the inrush current duration interval to obtain the excitation inrush current energy, integrate the difference between the actual loss and the corresponding steady-state loss within the magnetization recovery interval to obtain the magnetization recovery energy, and integrate the additional copper loss generated during the winding temperature recovery process to obtain the thermal state recovery loss. The ledger writing unit is used to write the excitation inrush current energy, magnetization recovery energy and thermal recovery loss using the transformer identifier, residual magnetization direction, residual magnetization amplitude, outage duration, winding temperature and closing phase as indexes.

[0010] Furthermore, the candidate combination generation module is used to generate candidate operating combinations based on the active load prediction sequence, reactive load prediction sequence, new energy output prediction sequence, three-phase load distribution prediction sequence, and load duration prediction results. These combinations include maintaining the current combination, putting in standby transformers, taking out operating transformers, and adjusting the load distribution ratio of parallel transformers. The module also deletes candidate operating combinations that do not meet the operating constraints based on the rated capacity, upper limit of winding temperature rise, bus voltage range, upper limit of three-phase imbalance, reactive power demand, and lower limit of standby capacity. The energy efficiency decision module is used to read the predicted losses of the reserved candidate operation combinations from the dynamic loss surface according to the scheduling time interval, and to accumulate the product of the predicted losses and the duration to obtain the steady-state cumulative losses.

[0011] Furthermore, the energy efficiency decision module is used to select ledger records from the transient energy ledger that have the same residual magnetization direction and whose residual magnetization amplitude, shutdown duration, winding temperature and proposed closing phase are respectively located within a preset grouping interval. The sum of the average values ​​of excitation inrush current energy, magnetization recovery energy and thermal state recovery loss in the selected ledger records is determined as the transient additional energy. The net energy saving over the entire cycle is obtained by subtracting the steady-state cumulative loss of the candidate operating combination and the transient additional energy from the steady-state cumulative loss of the current operating combination. The candidate operating combination with the largest net energy saving over the entire cycle is determined as the target operating combination only when the net energy saving over the entire cycle exceeds the switching threshold and the load duration is not less than the minimum hold time; otherwise, the current operating combination is maintained.

[0012] Furthermore, the switching control module includes: The closing phase calculation unit, the closing sequence determination unit, and the action delay compensation unit are all included. The closing phase calculation unit is used to calculate the expected magnetic flux peak value at each candidate closing time based on the residual magnetism direction and residual magnetism amplitude of each phase of the amorphous alloy transformer to be put into operation, as well as the magnetic flux increment generated by each phase voltage of the bus at the candidate closing time, and to determine the candidate closing time with the smallest expected magnetic flux peak value as the target closing time. The closing sequence determination unit is used to determine the first closing phase and subsequent closing phases based on the three-phase magnetic circuit coupling relationship, the residual magnetic amplitude of each phase and the phase-by-phase operation conditions of the circuit breaker. The action delay compensation unit is used to compensate for the sending time of the closing command based on the historical closing delay of each phase of the circuit breaker, so that the contacts close at the target closing time.

[0013] Furthermore, the feedback correction module is used to determine the loss deviation between the actual steady-state loss and the predicted steady-state loss, the transient deviation between the actual excitation inrush current energy and the predicted excitation inrush current energy, and the disturbance deviation between the actual bus voltage disturbance and the predicted bus voltage disturbance after the target operating combination has completed the switching. When the loss deviation exceeds the loss deviation threshold, the actual input variables and the actual steady-state loss are written into the loss sample library and the dynamic loss surface is updated. When the transient deviation exceeds the transient deviation threshold, the residual magnetism amplitude is corrected and the transient energy ledger is updated based on the actual closing phase and closing current waveform. The switching threshold is adjusted based on the 95th percentile of the loss deviation, transient deviation and disturbance deviation within the continuous statistical period, and the minimum holding time is adjusted based on the circuit breaker's allowable operating frequency and the predicted load duration.

[0014] The beneficial effects of this invention are as follows: 1. Under the condition that the target amorphous alloy transformer is kept under normal power supply, the present invention obtains incremental active power, excitation current, core leakage flux and magnetostrictive vibration response by controlling perturbation through the cancellation of positive volt-second integral and negative volt-second integral. This can reduce the influence of natural load fluctuation and magnetic flux bias on the identification results and realize the online separation of core loss, winding loss and harmonic additional loss.

[0015] 2. This invention uses load rate, core temperature, winding temperature, harmonic distortion rate, three-phase imbalance, tap position and reactive power compensation as operating variables to construct a dynamic loss surface. When the deviation between the measured loss and the predicted loss exceeds a threshold, it triggers the update of new loss samples, which can reflect the actual loss drift of amorphous alloy transformers under different operating conditions.

[0016] 3. This invention establishes a residual magnetism-transient energy ledger, which incorporates the residual magnetism direction, residual magnetism amplitude, outage duration, closing phase, excitation inrush current energy, magnetization recovery energy, and thermal state recovery loss of each phase into the evaluation of candidate operation combinations, so that the operation combination decision can cover the complete energy process of transformer decommissioning, shutdown, and re-energization.

[0017] 4. This invention obtains the net energy saving for the entire cycle by subtracting the transient additional energy from the steady-state cumulative loss difference between the current operating combination and the candidate operating combination, and determines the target operating combination by combining the switching threshold and the minimum holding time. This can reduce the ineffective switching caused by short-term load fluctuations and the actual increase in energy consumption due to predicted energy saving.

[0018] 5. This invention determines the closing phase and closing sequence based on the residual magnetism of the amorphous alloy transformer to be put into operation, and performs feedback correction based on the actual inrush current energy, actual steady-state loss and bus voltage disturbance, forming a closed-loop control process of loss identification, combination evaluation, switching execution and model correction. Attached Figure Description

[0019] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a system block diagram of the present invention; Figure 2 This is a flowchart of the energy efficiency optimization control of the present invention; Figure 3 This is a flowchart of the controlled perturbation loss identification and dynamic loss surface update of the present invention; Figure 4 This is a schematic diagram of the structure of the residual magnetism-transient energy ledger and the full-cycle net energy saving assessment of the present invention; Figure 5 This is a flowchart of the residual magnetism matching phase selection closing and feedback correction of the present invention. Detailed Implementation

[0020] The present application will now be described in further detail with reference to the accompanying drawings. It should be noted that the following specific embodiments are only used to further illustrate the present application and should not be construed as limiting the scope of protection of the present application. Those skilled in the art can make some non-essential improvements and adjustments to the present application based on the above application content.

[0021] Example 1: Please see Figure 1-5 This invention provides an energy efficiency optimization control system for amorphous alloy transformers used in smart grids, comprising a data acquisition module, a disturbance identification module, a loss modeling module, a transient energy management module, a candidate combination generation module, an energy efficiency decision-making module, a switching control module, and a feedback correction module.

[0022] The data acquisition module is used to collect electrical quantities, core leakage flux, magnetostrictive vibration, and core temperature of amorphous alloy transformers. Winding temperature The system tracks tap position and circuit breaker status, and performs unified timing, resampling, and timing alignment. The perturbation identification module determines the identification time window when the target amorphous alloy transformer meets load, voltage, and harmonic stability conditions, and applies a positive signal through an energy storage converter or active filter. Integral and Negative Controlled perturbations that cancel each other out in the integral. The loss modeling module is used to extract incremental active power, excitation current, core leakage flux, and magnetostrictive vibration response before, during, and after the controlled perturbation, and to combine this with winding temperature. The core loss, winding loss, and harmonic-induced additional loss are separated to establish a dynamic loss surface.

[0023] The transient energy management module is used to determine the residual magnetism of each phase based on the voltage integral, current decay characteristics, and opening phase of each phase before and after the amorphous alloy transformer is tripped, and to correlate the residual magnetism with the downtime and winding temperature. The closing phase, inrush current energy, magnetization recovery energy, and thermal recovery loss are correlated to form a transient energy ledger. The candidate combination generation module is used to generate a candidate combination based on future adjustments. The system generates candidate operating combinations based on the predicted results of active load, reactive load, renewable energy output, three-phase load distribution, and load duration within a given cycle. The energy efficiency decision module determines the net energy saving for the entire cycle by comprehensively considering the steady-state cumulative losses of the candidate operating combinations and the transient additional energy generated during combination switching, and selects the target operating combination. The switching control module determines the closing phase and sequence based on the residual magnetism of each phase of the amorphous alloy transformer to be switched on. The feedback correction module corrects the dynamic loss surface, transient energy ledger, switching threshold, and minimum hold time based on the actual steady-state losses after switching, the actual inrush current energy, and the actual bus voltage disturbance.

[0024] This embodiment applies an energy efficiency optimization control system to a smart substation equipped with two or more amorphous alloy transformers. The data acquisition module, perturbation identification module, and switching control module are deployed at the substation edge controller, while the candidate combination generation module and energy efficiency decision module are deployed at the substation dispatching center. The server, loss modeling module, transient energy management module, and feedback correction module are respectively located at the station-end edge controller and the dispatching... A copy of the synchronized data is stored on the server.

[0025] I. Operational Data Acquisition and Synchronization Processing The data acquisition module is connected to the voltage transformer, current transformer, core leakage flux sensor, magnetostrictive vibration sensor, and core temperature sensor, respectively. Sensors, winding temperature The sensors, tap changer controllers, and circuit breaker controllers are connected for communication.

[0026] The collected electrical quantities include primary three-phase voltage, primary three-phase current, secondary three-phase voltage, secondary three-phase current, primary active power, secondary active power, reactive power, voltage total harmonic distortion (THD), and current total harmonic distortion (THD). The collected equipment status quantities include core leakage flux signal, magnetostrictive vibration signal, and core temperature. Winding temperature Tap position and circuit breaker open / closed status.

[0027] For a power grid frequency of 50 In application scenarios, voltage, current, core leakage flux, and magnetostrictive vibration signals are measured using a 12.8... The sampling frequency, corresponding to each 256 sampling points. This sampling frequency can meet the needs of harmonic extraction up to the 25th order, excitation inrush current waveform identification, and magnetostrictive vibration characteristic analysis.

[0028] The output periods for active power, reactive power, total harmonic distortion of voltage, and total harmonic distortion of current are set to 1. Core temperature and winding temperature The sampling period is set to 5. The tap position and circuit breaker status are recorded using an event-triggered method.

[0029] All data acquisition terminals use the same time source, and the time synchronization error does not exceed 100. The data acquisition module uses 1 As a baseline time interval, power, temperature The harmonic and state data are resampled and time-aligned, while the original high-frequency waveforms of voltage, current, leakage flux and vibration signals are preserved for subsequent controlled perturbation response analysis and transient energy calculation.

[0030] When the consecutive missing time does not exceed 3 When the missing time period is 3, linear interpolation is performed using valid data before and after the missing time period; when the consecutive missing time exceeds 3... When this occurs, the corresponding time period will be marked as an invalid identification period, and the triggering of controlled perturbations will be prohibited during that time period. 3 Missing tolerance is based on power calculation cycle and temperature. The sampling period and the duration of the perturbation response are determined to ensure that the data entering the loss identification process is continuous in time.

[0031] Data exceeding the sensor's measurement range, or data whose change per unit time exceeds the equipment's physical change limit, is marked as abnormal data. The equipment's physical change limit is determined based on the rated parameters of the amorphous alloy transformer and the sensor's response speed. And determined by continuous operation tests before commissioning.

[0032] II. Determining the Time Window for Judging and Identifying Stable Operating Status The perturbation identification module uses 60 As a steady-state observation period, the load rate range, primary voltage effective value range, and total harmonic distortion rate range of the target amorphous alloy transformer are calculated respectively.

[0033] In this embodiment, when the load rate range does not exceed 2 The effective value range of the primary voltage does not exceed 1% of the rated voltage, and the range of the total harmonic distortion (THD) does not exceed 0.5. At that time, it was determined that the target amorphous alloy transformer met the conditions for stable operation.

[0034] The aforementioned stability thresholds were determined based on the calibration results of on-site data. During the initial commissioning of the transformer, at least 30 sets of natural stable operating data were collected near load rates of 20%, 40%, 60%, and 80%, assuming no circuit breaker operation, tap changer adjustment, or external fault disturbances. The 95th percentile values ​​of the load rate range, the effective voltage range, and the total harmonic distortion rate range were calculated. The 95th percentile value was then added to the maximum measurement error of the corresponding measurement link to obtain the initial stability threshold value.

[0035] When the on-site statistical value is lower than the aforementioned set value, the aforementioned set value shall be used to avoid frequent interruptions in the identification process due to measurement noise; when the load rate range exceeds 3... The effective voltage range exceeds 1.5% of the rated voltage, or the total harmonic distortion range exceeds 0.8%. At that time, the corresponding time period is not used as the loss identification period.

[0036] When the target amorphous alloy transformer is continuously 60 When stable operating conditions are met, the perturbation identification module will subsequently perform 2 Up to 4 The time interval is defined as the identification time window. If a circuit breaker trips, tap change, load rate changes abruptly, bus voltage exceeds limits, or a protection alarm occurs within the identification time window, the identification process is terminated, and the intermediate data generated during the identification process is deleted.

[0037] III. Controlled Perturbation Generation and Volt-Sec Balance Control like Figure 3 As shown, the perturbation identification module applies positive and negative perturbations to the target amorphous alloy transformer sequentially through an energy storage converter or an active filter.

[0038] The initial amplitude of both positive and negative perturbations is set to 0.8% of the rated phase voltage, and each perturbation segment lasts for 6 seconds. Two intervals are set between the positive perturbation section and the negative perturbation section. The transition range.

[0039] The perturbation amplitude starts at 0.5% of the rated phase voltage and increases gradually by 0.1 units each time. When the peak values ​​of the excitation current, core leakage flux, and magnetostrictive vibration response all reach three times the corresponding root mean square value of the background noise, the current perturbation amplitude is maintained; when any response fails to meet the aforementioned signal-to-noise ratio requirement, the perturbation amplitude is further increased. The perturbation amplitude must not exceed 1.2% of the rated phase voltage and must not trigger overvoltage protection, undervoltage protection, or tap adjustment.

[0040] Positive perturbations and negative perturbations according to Controlled by the principle of balance, The imbalance rate satisfies formula (1): In the formula, for The imbalance rate is a dimensionless parameter. This is the positive perturbation voltage, in units of ; The negative perturbation voltage is expressed in units of 1. ; and These are the start and end times of the positive perturbation, respectively, in units of 1 / 2. ; and These are the start and end times of the negative perturbation, respectively, in units of 1 / 2. ; for Balance threshold.

[0041] In this embodiment, Balance threshold The threshold is set to 0.02. This threshold was determined through a baseline recovery test of the core leakage flux after perturbation. When When the imbalance rate does not exceed 0.02, the perturbation ends 2 The relative deviation between the core leakage magnetic baseline inside the perturbation and the core leakage magnetic baseline before the perturbation should not exceed 2%.

[0042] If the core leakage magnetic baseline deviation exceeds 2%, the identification data will be marked as invalid and processed according to the positive perturbation method. Integral and Negative Perturbation The actual difference in the integral is used to adjust the duration or amplitude of the negative perturbation, and the controlled perturbation is reimplemented.

[0043] The 2% baseline recovery threshold for magnetic flux leakage should be higher than the sum of the measurement error of the magnetic flux leakage sensor and the natural fluctuation of magnetic flux leakage under stable operating conditions. In this embodiment, the maximum measurement error of the magnetic flux leakage sensor is 0.6%, and the 95th percentile of the natural fluctuation of magnetic flux leakage under stable operating conditions is 0.8%. The sum of these two values ​​is multiplied by 1.4. The full coefficient is rounded down to 2%.

[0044] IV. Separation of core loss, winding loss and harmonic-related loss The loss modeling module divides the identification time window into a reference segment, a positive perturbation segment, a negative perturbation segment, and a recovery segment, and extracts the primary active power, secondary active power, fundamental current, harmonic currents, core leakage flux amplitude, and magnetostrictive vibration amplitude for each segment.

[0045] The difference between the active power on the primary side and the active power on the secondary side is used to determine the total loss of the target amorphous alloy transformer. Winding losses, harmonic losses, and core losses are determined according to formula (2): In the formula, The fundamental winding loss is expressed in units of 1000 ppm. ; Harmonic additional loss, unit is ; Core loss, unit: ; The active power on the primary side is expressed in units of 1. ; This refers to the active power on the secondary side, in units of... ; For the first The effective value of the fundamental phase current, in units of ; For the first Xiangdi The effective value of the second harmonic current, in units of ; For reference temperature Next Phase winding equivalent resistance, unit: ; For reference temperature Next Phase winding in the first AC resistance at the second harmonic frequency, in units of ; For the first Phase winding temperature The unit is ; Reference temperature for winding resistance The unit is ; The resistance temperature of the winding conductor coefficient, unit is per The formula uses the power of ten to the power of negative cube to convert watts to... .

[0046] This embodiment uses copper windings, with a reference temperature. Take 75 Winding conductor resistance temperature coefficient Take 0.00393 per Reference temperature The equivalent resistance of the windings should be based on the DC resistance test data from the transformer's factory; if the factory data is missing, the DC resistance test of the windings should be performed again while the transformer is out of service.

[0047] The AC resistance corresponding to each harmonic was determined through a frequency sweep impedance test before the transformer was put into operation. The upper limit for the harmonic order was set to 25th because at 12.8... The sampling frequency can stably extract harmonic components up to the 25th order, and field tests show that the energy of the first 25 harmonics accounts for no less than 99% of the total harmonic energy. If the energy of harmonics above the 25th order accounts for more than 1%, the sampling frequency should be increased accordingly and the range of harmonic loss calculation should be expanded.

[0048] The loss identification results must also satisfy the perturbation response symmetry condition. The core loss identification values ​​for the positive and negative perturbation sections are calculated separately, and the changing directions of excitation current, core leakage flux, and magnetostrictive vibration response are compared.

[0049] When the relative deviation between the core loss identification values ​​corresponding to the positive and negative perturbations does not exceed 5%, and the changing directions of the excitation current, core leakage flux, and magnetostrictive vibration response correspond to the perturbation direction, the identification results are determined as valid loss samples.

[0050] The 5% response deviation threshold is based on the same load rate range and the same temperature. The interval and the same tap position are determined by at least 20 repeated perturbation tests. If the 95th percentile of the relative deviation of the repeated tests is below 5%, 5% is used as the response deviation threshold; if the 95th percentile is between 5% and 8%, the actual 95th percentile is used; if the 95th percentile exceeds 8%, the current sampling accuracy is determined. If the perturbation amplitude or operational stability is insufficient, the identification data will not be received by further relaxing the threshold.

[0051] V. Construction and Update of Dynamic Loss Surface The loss modeling module uses load rate and core temperature as parameters. Winding temperature Total harmonic distortion of voltage, total harmonic distortion of current, and three-phase current imbalance. Using tap position and reactive power compensation as input variables, and core loss, winding loss, harmonic additional loss and total loss as output variables, a dynamic loss surface corresponding to each amorphous alloy transformer is established.

[0052] The dynamic loss surface is configured with separate data layers for each tap position. Within the same tap position, the load factor is calculated based on 10... Divide the interval, core temperature and winding temperature According to 5 The voltage total harmonic distortion (THD) and current total harmonic distortion (THD) are divided into intervals according to one... Dividing the intervals, three-phase current imbalance According to 0.5 The reactive power compensation is divided into intervals based on 5% of the transformer's rated capacity.

[0053] The above interval width Sensitive to changes in loss and loss sample density Determined. When there are fewer than 12 effective loss samples in a single interval, a loss unit is not formed separately, but is merged with the adjacent interval; when there are 12 effective loss samples, partitioned multilinear interpolation is used to determine the predicted loss in that interval.

[0054] Setting 12 groups as the minimum number of samples is to ensure that each loss unit covers at least three operating levels of the input variable: low, medium, and high values, and that each operating level retains no fewer than 4 sets of repeated data, thereby reducing the impact of single measurement errors on the interpolation results.

[0055] When the operating condition to be evaluated is within the coverage of existing loss samples, the loss samples that are closest to the operating condition to be evaluated and distributed in different directions are selected for partition interpolation; when the operating condition to be evaluated is outside the coverage of existing loss samples, no extrapolation calculation is performed, but the operating condition is marked as an operating condition to be supplemented for identification.

[0056] The actual total loss is determined by the difference between the active power on the primary side and the active power on the secondary side. When the relative deviation between the actual total loss and the predicted total loss from the dynamic loss surface exceeds 5% for three consecutive times, the loss modeling module sends an update command to the perturbation identification module and implements controlled perturbation when the target amorphous alloy transformer meets the stable operating conditions again.

[0057] The 5% loss deviation threshold is determined based on power measurement error, time synchronization error, and the natural fluctuation of total loss under stable operating conditions. In this embodiment, the overall error of the power measurement link does not exceed 1%, and the 95th percentile of the natural fluctuation of total loss under stable operating conditions is 2.1%. The sum of these two values ​​is multiplied by 1.5. The full coefficient is 4.65%, which is rounded down to 5%.

[0058] VI. Estimation of Remanence like Figure 4 As shown, when the amorphous alloy transformer performs a tripping operation, the transient energy management module records the phase voltage, phase current, tripping phase, actual disconnection time of the circuit breaker contacts, and current decay time before and after the tripping.

[0059] The system starts integrating the phase terminal voltage from the moment the corresponding phase magnetic field passes through zero most recently before the circuit breaker trips, until the phase terminal voltage is continuously integrated for three consecutive times. The voltage is 0.5% lower than the rated phase voltage, and the first phase voltage is obtained according to formula (3). Phase remanence estimate: In the formula, For the first Remanence estimate, in Weber; The magnetic flux value at the start of integration is expressed in Weber. For the first Number of turns in phase winding; For the first Phase terminal voltage, unit is ; The starting time of integration, in units of ; For the first Phase terminal voltage convergence time, in units of .

[0060] When the integration begins Choosing the moment when the magnet passes through zero, Set the value to zero. Determine the direction of remanence based on the sign of the estimated remanence value, and determine the amplitude of remanence based on the absolute value of the estimated remanence value.

[0061] The residual magnetism is not corrected by weighted summation of current decay time, but by a segmented correction rule. When the relative deviation between the actual current decay time and the median value of the historical current decay time under the same opening phase does not exceed 20%, the residual magnetism amplitude is not corrected; when the actual current decay time is more than 20% higher than the historical median value, the corresponding positive correction coefficient is selected according to the pre-established no-load opening test table; when the actual current decay time is more than 20% lower than the historical median value, the corresponding reverse correction coefficient is selected according to the no-load opening test table.

[0062] The 20% cutoff threshold is determined through no fewer than 30 no-load tripping tests and should be more than three times the measurement error of the current decay time. The correction factor is determined by the ratio of the residual magnetism value measured in the no-load tripping test to the estimated voltage integral value, without using manually set weights.

[0063] VII. Construction of Transient Energy Ledger After the amorphous alloy transformer is re-energized, the transient energy management module records the actual re-energizing phase, outage duration, and winding temperature before re-energizing. Peak value of inrush current, duration of inrush current and bus voltage disturbance.

[0064] The inrush current duration period begins from the actual closing moment of the circuit breaker contacts and extends to five consecutive peak currents in each phase. The process ends when the current does not exceed 1.2 times the peak steady-state current. The 1.2 times threshold is determined based on the upper limit of the natural fluctuation of the peak current under stable operating conditions and the current measurement error, and is used to distinguish between the inrush current section and the steady-state current section.

[0065] The magnetization recovery interval begins at the end of the inrush current duration interval and continues until 60 consecutive days after the actual total loss. The process ends when the value is within ±5% of the predicted dynamic loss surface value.

[0066] The inrush current energy is obtained by integrating the additional input power after closing over the inrush current duration; the magnetization recovery energy is obtained by integrating the difference between the actual total loss and the corresponding steady-state loss of the dynamic loss over the magnetization recovery interval; the thermal recovery loss is obtained by integrating the winding temperature... The difference between the actual winding loss during the recovery period and the target steady-state winding loss is obtained by integration.

[0067] In the above energy integration, instantaneous power is uniformly expressed in watts, and integration time is uniformly expressed in watts. The integral result, after being divided by 3.6 million, is converted to... This ensures that transient energy and steady-state cumulative loss use the same dimensions.

[0068] The transient energy ledger includes transformer identification, residual magnetism direction of each phase, normalized residual magnetism amplitude range, outage duration range, and winding temperature. The interval and closing phase interval are used as indexes to store excitation inrush current energy, magnetization recovery energy, and thermal recovery loss.

[0069] Wide range of normalized remanence amplitude Set to 0.1 Wide range of downtime Set to 30 Winding temperature Interval width Set to 5 Wide closing phase range Set to 5 The direction of remanence is divided into positive, negative, and near-zero states, with the normalized remanence amplitude not exceeding 0.05. At that time, it was determined to be near-zero remanence.

[0070] 0.05 The error is determined based on the maximum error of the residual magnetism measurement link. In this embodiment, the maximum error of the residual magnetism measurement link is 0.018. Multiplying this by 2.5 times the error margin gives 0.045. Rounded down, it is determined to be 0.05. This is to avoid misjudging the direction of residual magnetism due to measurement noise.

[0071] When the number of transient energy records within the same index interval reaches 10 sets, the maximum and minimum values ​​are removed, and the arithmetic mean of the remaining 8 sets of records is calculated and used as the predicted transient energy value for that index interval. This processing method can reduce the impact of occasional load disturbances and circuit breaker malfunctions on the prediction results, and does not require setting subjective weights.

[0072] VIII. Generation of Candidate Combinations and Constraint Screening Candidate combination generation module obtains future tuning The prediction results of active load, reactive load, renewable energy output, three-phase load distribution, and load duration within the cycle.

[0073] In this embodiment, future adjustment The cycle is set to 6. , adjust The time interval is set to 15. The system generates candidate operating combinations based on the currently operating amorphous alloy transformers, standby amorphous alloy transformers, and the rated capacity of each amorphous alloy transformer. These combinations include maintaining the current operating combination, putting one standby amorphous alloy transformer into operation, taking one currently operating amorphous alloy transformer out of operation, and adjusting the load distribution ratio of parallel-operating amorphous alloy transformers.

[0074] Candidate operating combinations must simultaneously meet the following operating constraints: the predicted load rate of any amorphous alloy transformer does not exceed 90%; the predicted winding hot spot temperature... No more than 95 Bus voltage is within 95% to 105% of rated voltage; three-phase current imbalance. No more than 10%; after disconnecting any single amorphous alloy transformer, the remaining usable capacity shall not be less than 110% of the predicted maximum load.

[0075] The above operating constraints are determined based on the nameplate parameters of the amorphous alloy transformer, insulation heat resistance class, substation protection settings, and substation operating procedures. Regarding allowable load factor and hot spot temperature... For application scenarios with different limit or reserve capacity requirements, the values ​​specified in the corresponding equipment manufacturing documents and station operation procedures should be adopted.

[0076] If any candidate operating combination does not meet any of the aforementioned operating constraints, it will be removed from the candidate set and will not be included in the full-cycle net energy saving assessment process.

[0077] IX. Calculation of Net Energy Savings Throughout the Cycle and Determination of Target Operating Combinations The energy efficiency decision module is based on 15 The tone The time interval is used to read the predicted loss corresponding to each candidate running combination from the dynamic loss surface, and the predicted loss for each time period is multiplied by 0.25. After multiplying, the results are summed to obtain the possible combinations of running conditions for the next 6 months. Steady-state cumulative loss within.

[0078] For candidate operating combinations that include the operation of a standby amorphous alloy transformer, selection is made from the transient energy ledger based on the consistent residual magnetism direction and the normalized residual magnetism amplitude, outage duration, and winding temperature. The proposed method is to use transient energy records where the closing phase is located within the corresponding index interval.

[0079] If the number of transient energy records in the corresponding index interval is no less than 3, the transient energy prediction value of the index interval is used; if the number of records is less than 3, the record closest to the current residual magnetism state and operating state is selected from the adjacent index interval; if no less than 3 records can still be obtained in the adjacent index interval, the upper limit value of transient energy obtained from the transformer commissioning test is used to avoid underestimating the commissioning cost of candidate operating combinations.

[0080] No. The transient additional energy and the net energy saved over the entire cycle for each candidate operating combination satisfy formula (4): In the formula, For the first The transient additional energy of each candidate running combination, in units of ; The inrush current energy is expressed in units of 1000 kJ / m². ; The energy for magnetization recovery is expressed in units of 1. ; Thermal recovery loss, unit: ; For the first The total net energy savings over the entire cycle for each candidate operating combination, in units of ; To maintain the steady-state cumulative loss during the current operating combination, the unit is... ; For the first The steady-state cumulative loss of each candidate operating combination, in units of .

[0081] Excitation inrush current energy, magnetization recovery energy, and thermal state recovery loss are all energy values ​​with the same physical dimensions. The three are directly added together to determine the transient additional energy, without any artificial weights.

[0082] In this embodiment, the initial cutoff threshold is set to 5.5. The switching threshold consists of the prediction error margin, the energy equivalent to the circuit breaker operation, and the minimum effective energy saving benefit.

[0083] The prediction error margin is calculated using the most recent 30 adjustments. The 95th percentile of the absolute value of the steady-state cumulative loss prediction error during the process is 3.2 in this embodiment. The energy consumption for circuit breaker operation is calculated based on the remaining mechanical life of the circuit breaker, the cost of a single maintenance operation, and the allowable number of operations. In this embodiment, it is 0.8. The minimum effective energy saving benefit is determined based on the energy efficiency management targets of the substation; in this example, it is 1.5. All three have dimensions. Adding them directly gives 5.5 The cut-off threshold.

[0084] The prediction error margin is recalculated every 10 effective operation combination switches; the energy converted from circuit breaker operation is updated monthly based on the remaining mechanical life of the circuit breaker; the minimum effective energy saving benefit is set by the operation and management personnel based on the electricity cost and energy saving target.

[0085] The minimum hold time is initially set to 3. The minimum holding time is determined based on the thermal time constant of the amorphous alloy transformer, the continuous reliable time of load prediction, and the daily allowable number of circuit breaker operations. When the net energy saving of a candidate operating combination over the entire cycle exceeds the switching threshold, and the predicted duration of the corresponding load state is not less than the minimum holding time, the candidate operating combination with the largest net energy saving over the entire cycle is determined as the target operating combination; otherwise, the current operating combination is maintained. 10. Residual magnetism matching phase selection closing like Figure 5 As shown, for the amorphous alloy transformers that need to be put into the target operating combination, the switching control module uses 1 As the phase search interval, traverse 0 Up to 359 The candidate closing phase.

[0086] For each candidate closing phase, the power flux increment after closing is determined based on the integral of the voltage of each phase of the bus. The power flux increment is then added to the estimated residual magnetism of the corresponding phase to obtain the expected peak flux of each phase. The maximum value among the three expected peak fluxes is taken as the evaluation result of the corresponding candidate closing phase, and the candidate closing phase with the smallest maximum value is determined as the target closing phase.

[0087] The closing phase determination process directly uses the minimum expected peak value of the three-phase magnetic flux as the control objective, without using the weighted sum of multiple evaluation indicators.

[0088] The closing sequence prioritizes the phase whose power flux increment is opposite to the direction of residual magnetism as the first closing phase, and determines the closing sequence of the other two phases based on the flux change generated by the three-phase magnetic circuit coupling after the first closing phase is closed.

[0089] The operation delay compensation unit statistically analyzes the closing delay corresponding to the most recent 30 effective closing operations of each phase of the circuit breaker and calculates the median value of the closing delay for each phase. Based on the target closing time and the median value of the corresponding phase closing delay, the system sends phase-specific closing commands in advance.

[0090] When the standard deviation of the corresponding phase closure delay does not exceed 2 When the median value of the closing delay is used for motion compensation; when the standard deviation of the closing delay exceeds 2... If the circuit breaker's mechanical action is deemed to have excessive dispersion, the phase selection and closing function of the corresponding phase will be suspended, and a circuit breaker maintenance prompt will be generated.

[0091] 2 Discreteness threshold based on 50 The conversion relationship between voltage phase and time in the power grid is determined. 2 Corresponding to approximately 36 electric angle If the time deviation exceeds this limit, the actual closing phase of the circuit breaker may deviate significantly from the target closing phase, and will no longer meet the requirements of residual magnetism matching control.

[0092] During the phase selection and closing process, the peak value of the inrush current, bus voltage disturbance, and three-phase current imbalance are monitored in real time. In this embodiment, the peak inrush current control threshold is set to 6 times the rated current, the maximum bus voltage drop control threshold is set to 5% of the rated voltage, and the three-phase current imbalance is... The control threshold is set to 10%.

[0093] The above The full threshold is determined as 80% of the corresponding protection action value. When the reference value for inrush current protection action is 7.5 times the rated current, the control threshold is set to 6 times the rated current to retain a 20% control margin before the protection device operates.

[0094] When any monitoring value exceeds the corresponding When the threshold is full, the closing operation of subsequent phases is stopped, and the current operating state is maintained or restored to the operating state before the combined switching is based on the capacity of the transformers currently in operation and the bus voltage status.

[0095] XI. Feedback and Corrections After the target operating combination completes the switch, the feedback correction module calculates the loss deviation between the actual steady-state loss and the predicted steady-state loss, the transient energy deviation between the actual excitation inrush current energy and the predicted excitation inrush current energy, and the disturbance deviation between the actual bus voltage disturbance and the predicted bus voltage disturbance.

[0096] The loss deviation threshold is initially set to 5%. When the relative deviation between the actual steady-state loss and the predicted steady-state loss exceeds 5% for three consecutive times, the corresponding actual operating variables and actual steady-state loss are written into the loss sample library, and the dynamic loss surface is updated.

[0097] The transient energy deviation threshold is initially set to 10%. Transient energy calculation is affected by the actual closing time of the circuit breaker contacts, the peak instantaneous current, and the transient integral interval identification error. Therefore, the transient energy deviation threshold is higher than the steady-state loss deviation threshold.

[0098] Every 30 valid operational combination switches, the system calculates the 95th percentile values ​​of the absolute values ​​of loss deviation and transient energy deviation. When the corresponding 95th percentile value is lower than the initial threshold, the initial threshold is used; when the 95th percentile value is higher than the initial threshold but does not exceed 15%, the actual 95th percentile value is used; when the 95th percentile value exceeds 15%, it is determined that there is an anomaly in the measurement link, residual magnetism estimation process, or transient energy model, and data is not received by further relaxing the threshold.

[0099] When the transient energy deviation exceeds the corresponding threshold, the residual magnetization amplitude of each phase before closing is reversed according to the actual closing phase, the actual excitation inrush current peak value and the excitation inrush current decay process, and the corrected residual magnetization state, actual excitation inrush current energy, actual magnetization recovery energy and actual thermal state recovery loss are written into the transient energy ledger.

[0100] When the actual transient additional energy within a continuous statistical period is higher than the predicted transient additional energy, the switching threshold is increased; when the actual transient additional energy is lower than the predicted transient additional energy, and the bus voltage disturbance does not exceed [a certain threshold], the switching threshold is increased. When the threshold is full, decrease the switching threshold. The single adjustment range of the switching threshold. The threshold should not exceed 10% of the original threshold value to prevent the control parameters from changing abruptly due to a single abnormal result.

[0101] When the actual operating frequency of the circuit breaker within a preset statistical period exceeds 80% of the allowable operating frequency, the minimum holding time will be extended by 30 seconds. When the actual frequency of actions is less than 40% of the allowable frequency of actions, and the accuracy of load duration prediction exceeds 90%, the minimum hold time will be reduced by 30%. The minimum holding time must not be less than the main thermal time constant of the amorphous alloy transformer.

[0102] Example 2: This example uses a substation equipped with three units, each with a rated capacity of 1000. Taking an amorphous alloy transformer as an example, the first and second amorphous alloy transformers are in operation, while the third amorphous alloy transformer is in standby mode.

[0103] The system follows 12.8 The sampling frequency collects phase voltage, phase current, core leakage flux, and magnetostrictive vibration signals, according to 5 Sampling periodically collects core temperature and winding temperature .

[0104] In 60 During the steady-state observation period, the load rate range of the first amorphous alloy transformer was 1.3. The effective range of the primary voltage is 0.62% of the rated voltage, and the range of the total harmonic distortion (THD) is 0.27. None of them exceeded the corresponding stability threshold, so they entered the perturbation identification state.

[0105] The energy storage converter applies positive and negative perturbations with amplitudes of 0.8% of the rated phase voltage to the first amorphous alloy transformer. The positive and negative perturbations last for six seconds each. Calculations show that... The imbalance rate is 0.014, which is less than 0.02. Balance threshold: After the perturbation, the core leakage magnetic baseline deviation was 1.2%, which is less than the 2% leakage magnetic baseline recovery threshold. Therefore, the perturbation data was determined as valid identification data.

[0106] After loss separation, the fundamental winding loss of the first amorphous alloy transformer under current operating conditions is 3.76%. The harmonic-induced additional loss is 0.35. The core loss is 0.93. The total loss is 5.04. .

[0107] The fundamental winding loss of the second amorphous alloy transformer under current operating conditions is 3.98. The harmonic-induced additional loss is 0.39. The core loss is 0.96. The total loss was 5.33. The system will use the load rate and core temperature of the two transformers as parameters. Winding temperature Harmonic distortion rate, three-phase current imbalance The tap position and reactive power compensation amount are written into the corresponding loss sample library.

[0108] Future 6 The load forecast results indicate that, while maintaining the operation of both the first and second amorphous alloy transformers, the steady-state cumulative loss is 128.6%. After the third amorphous alloy transformer was put into operation and the load was redistributed, the steady-state cumulative loss was 108.4. Steady-state cumulative losses decreased by 20.2%. .

[0109] The transient energy ledger of the third amorphous alloy transformer shows that, under the current residual magnetism state, outage duration, and winding temperature... Under the proposed closing phase, the predicted inrush current energy is 1.4. The predicted magnetization recovery energy is 0.8. The predicted thermal recovery loss is 1.0. Therefore, the transient additional energy corresponding to the input of the third amorphous alloy transformer is 3.2. The net energy saving over the entire cycle is 17.0. .

[0110] The threshold for this cut is 5.5. The minimum holding time is 3 The predicted duration of the high-load condition is 4.2 hours. Since the net energy saving over the entire cycle is greater than the switching threshold and the predicted load duration is greater than the minimum hold time, the energy efficiency decision module will activate the third amorphous alloy transformer and redistribute the load to determine the target operating combination.

[0111] Before phase selection and closing, the normalized residual magnetism of the first phase of the third amorphous alloy transformer is positive 0.31. The normalized remanence of the second phase is negative 0.27. The normalized remanence of the third phase is positive 0.04. Since the normalized remanence amplitude of the third phase does not exceed 0.05... The third phase is determined to be in a near-zero remanence state.

[0112] The switching control module traverses 0 Up to 359 The candidate closing phases were determined, and the second phase was selected as the first closing phase. The target closing phase was the second phase voltage of the busbar, 167. Position. The first and third phases close at predetermined subsequent phases.

[0113] The median delay of the second phase closing in the last 30 circuit breakers is 41.6. The standard deviation is 1.3. No more than 2 The action discreteness threshold, therefore the system is 41.6 seconds before the target closing time. Issue the second phase closing command.

[0114] After the switchover was completed, the measured peak inrush current was 4.7 times the rated current, the maximum bus voltage drop was 2.3% of the rated voltage, and the three-phase current was unbalanced. The maximum value was 6.1%, none of which exceeded the corresponding... Full threshold.

[0115] The actual inrush current energy is 1.62. Compared to 1.4 The predicted value had a deviation of 15.7%, exceeding the 10% transient energy deviation threshold. The feedback correction module corrected the residual magnetism amplitude of the third amorphous alloy transformer based on the actual closing phase, peak excitation inrush current, and excitation inrush current decay curve, and wrote the actual excitation inrush current energy, actual magnetization recovery energy, and actual thermal recovery loss into the transient energy ledger.

[0116] The relative deviation between the actual steady-state loss after the switch and the predicted value of the dynamic loss surface is 3.9%, which does not exceed the 5% loss deviation threshold. Therefore, the dynamic loss surface update will not be triggered this time.

[0117] Therefore, this invention achieves online separation of core loss, winding loss, and harmonic-induced loss of amorphous alloy transformers through controlled micro-perturbation, reflects the actual loss characteristics under different operating conditions through dynamic loss surface, quantifies the impact of residual magnetism on excitation inrush current energy, magnetization recovery energy, and thermal recovery loss through transient energy ledger, and implements operation combination decisions based on net energy savings throughout the entire cycle, thereby avoiding the problem of actual increased consumption caused by predictive energy saving based solely on steady-state loss or fixed load rate thresholds.

[0118] Example 3 This embodiment illustrates how the present invention determines the decommissioning time, re-energizing time, and phase selection closing parameters of an amorphous alloy transformer when the net load of a power distribution station continuously decreases due to increased output from new energy sources, followed by a rebound in load at night.

[0119] In this embodiment, the substation is equipped with four units, each with a rated capacity of 1250. The amorphous alloy transformers are designated as the first, second, third, and fourth amorphous alloy transformers, respectively. The first, second, and third amorphous alloy transformers are in operation, while the fourth amorphous alloy transformer is in standby status. The rated voltage of the substation bus is 10 kV. The power grid frequency is 50. Distributed optical network access within the station Power generation system.

[0120] The data acquisition module is designed according to 12.8. The sampling frequency was used to collect primary side voltage, primary side current, secondary side voltage, secondary side current, core leakage flux, and magnetostrictive vibration signals of each amorphous alloy transformer, according to 1 The output cycle generates active power, reactive power, voltage total harmonic distortion (THD), and current total harmonic distortion (THD), according to 5 The sampling period collects the core temperature and winding temperature .

[0121] During the operating period from 9:45 AM to 10:45 AM that day, distributed optical... Output gradually increased, and the net apparent load of the substation increased from 2180 Dropped to 1030 Load forecasting results indicate that 1030 Up to 1120 The low load will continue until 16:45, after which it will decrease as light... With decreased power output and increased electricity load, the net apparent load of the substation will rise to 2600 before 18:00. .

[0122] Between 10:15 and 10:16, 60 During the observation period, the load factor range of the first amorphous alloy transformer was 1.1. The effective range of the primary voltage is 0.54% of the rated voltage, and the range of the total harmonic distortion (THD) is 0.31. The load factor range corresponding to the second amorphous alloy transformer is 1.4. The effective range of the primary voltage is 0.61%, and the range of the total harmonic distortion (THD) is 0.34. The load factor range corresponding to the third amorphous alloy transformer is 1.2. The effective range of the primary voltage is 0.58%, and the range of the total harmonic distortion (THD) is 0.29. Since none of the above operating parameters exceeded the corresponding stability threshold, the perturbation identification module performed loss identification on the three amorphous alloy transformers in operation.

[0123] The disturbance identification module applies positive and negative disturbances of 0.7% of the rated phase voltage to the in-station energy storage converter. Both the positive and negative disturbance sections last for 6 seconds. The first amorphous alloy transformer, the second amorphous alloy transformer, and the third amorphous alloy transformer correspond to... The imbalance rates were 0.012, 0.016, and 0.013, respectively, all less than 0.02. The baseline deviations of the core leakage flux after the perturbation were 0.9%, 1.4%, and 1.1%, respectively, all less than the 2% baseline recovery threshold. Therefore, the data obtained from the three controlled perturbations were all determined to be valid identification data.

[0124] After processing by the loss modeling module, the core loss of the first amorphous alloy transformer under the current operating conditions is 0.88. The fundamental winding loss is 1.74. The harmonic-induced additional loss is 0.28. The total loss was 2.90. The core loss of the second amorphous alloy transformer is 0.91. The fundamental winding loss is 1.81. The harmonic-induced additional loss is 0.31. The total loss was 3.03. The core loss of the third amorphous alloy transformer is 1.04. The fundamental winding loss is 1.93. The harmonic-induced additional loss is 0.39. The total loss was 3.36. .

[0125] The total loss of the third amorphous alloy transformer is higher than that of the first and second amorphous alloy transformers. According to the corresponding dynamic loss surface, the core loss of the third amorphous alloy transformer is consistently higher under the same load rate, and its current winding temperature... 71 65 higher than the first amorphous alloy transformer 67 of the second amorphous alloy transformer Therefore, the candidate combination generation module prioritizes exiting the third amorphous alloy transformer as the candidate solution, and simultaneously generates candidate solutions for exiting the first amorphous alloy transformer, exiting the second amorphous alloy transformer, and maintaining the current operating combination.

[0126] The candidate combination generation module is based on 15 The tone At time intervals, operational constraints are verified for each candidate operating combination. After the third amorphous alloy transformer is decommissioned, the predicted average load rates of the first and second amorphous alloy transformers during low-load periods are 41.6% and 40.8%, respectively, with peak load rates of 45.3% and 44.7%, respectively; the predicted winding hot spot temperature... 74 respectively and 75 The predicted bus voltage is within 98.1% to 101.4% of the rated voltage; three-phase current imbalance. Not exceeding 5.8%.

[0127] After either the first or second amorphous alloy transformer is taken out of service, the rated capacity of the remaining single amorphous alloy transformer is 1250. It is still higher than the predicted maximum apparent load of 1120 during low-load periods. 1.1 times. Therefore, the candidate operating combinations for exiting the third amorphous alloy transformer meet the requirements of capacity, winding temperature, etc. Bus voltage, three-phase imbalance And standby capacity constraints.

[0128] The energy efficiency decision module determines the adjustment period from 10:30 to 18:00 based on the dynamic loss surface. During the period, if the first, second, and third amorphous alloy transformers are operated simultaneously, the predicted steady-state cumulative loss is 96.4%. .

[0129] If the third amorphous alloy transformer is disconnected at 10:30 and reconnected at 16:45, the predicted steady-state cumulative loss is 79.1%. If the second amorphous alloy transformer is deactivated and reactivated at 16:45, the predicted steady-state cumulative loss is 80.0%. If the first amorphous alloy transformer is deactivated and reactivated at 16:45, the predicted steady-state cumulative loss is 80.8%. .

[0130] The energy efficiency decision module further reads the transient energy ledger of the third amorphous alloy transformer. Based on the projected outage of the third amorphous alloy transformer for 6... 15 Expected winding temperature before closing 42 Based on the residual magnetism of each phase and the proposed closing phase, the inrush current energy upon reactivation is determined to be 1.2. The magnetization recovery energy is 0.7. The thermal recovery loss is 0.7. The transient additional energy is 2.6. .

[0131] According to formula (4), the net energy saving over the entire cycle when the third amorphous alloy transformer is deactivated and reactivated before the evening load rises is 14.7. The net energy saving over the entire lifecycle corresponding to the removal of the second amorphous alloy transformer is 13.9. The net energy saving over the entire lifecycle corresponding to the removal of the first amorphous alloy transformer is 12.8. .

[0132] This adjustment The cut threshold is 6.0. Among them, the most recent 30 adjustments The 95th percentile of the absolute value of the steady-state cumulative loss prediction error is 3.5. The equivalent energy of the circuit breaker operation is 0.9. The minimum effective energy saving benefit is 1.6. All three items are based on The threshold is determined by directly adding the values ​​in units of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 ...

[0133] This adjustment The minimum holding time is 2 30 This time is determined based on the thermal time constant of the third amorphous alloy transformer, the continuous reliable time of load forecasting, and the remaining allowable number of circuit breaker operations for the day. The predicted duration of the low load condition is 6. 15 This is greater than the minimum holding time; the net energy saving over the entire cycle corresponding to the withdrawal of the third amorphous alloy transformer is 14.7. Greater than 6.0 The switching threshold is determined. Therefore, the energy efficiency decision module identifies the third amorphous alloy transformer being shut down at 10:30 and reconnected at 16:45 as the target operating combination.

[0134] At 10:30, when the tripping operation was performed, the transient energy management module recorded the phase voltage, phase current, tripping phase, and actual opening time of the circuit breaker contacts of the third amorphous alloy transformer, and calculated the residual magnetism of each phase according to formula (3). The normalized residual magnetism of the first phase was calculated to be positive 0.28. The normalized remanence of the second phase is negative 0.33. The normalized remanence of the third phase is positive 0.02. .

[0135] Because the normalized remanence amplitude of the third phase is less than 0.05 The third phase is determined to have near-zero residual magnetism. The deviations between the current decay time of the first and second phases and the historical median value of the same tripping phase are 8.6% and 11.2%, respectively, neither of which exceeds the 20% segment correction limit. Therefore, no additional correction is made to the residual magnetism amplitude of the first and second phases.

[0136] During the outage of the third amorphous alloy transformer, the system continued to acquire load and light. The rolling forecast results for output. When the reassessment was performed at 16:15, the predicted maximum apparent load from 16:45 to 18:00 was revised from the initial forecast of 2600. Adjusted to 2550 It is still necessary to operate three amorphous alloy transformers together, and the original target operating combination remains unchanged.

[0137] Before reconnecting the third amorphous alloy transformer, the switching control module uses 1 For phase interval traversal 0 Up to 359 The candidate closing phase. Based on the residual magnetism of each phase and the magnetic flux increment generated by the bus voltage, the second phase is determined as the first closing phase, and the bus voltage of the second phase is set to 171. The position is determined to be the target's first conjunction phase.

[0138] The median value of the effective closing delay of the second phase of the circuit breaker in the most recent 30 transactions is 42.3 seconds. The standard deviation is 1.5. No more than 2 The action discreteness threshold. Therefore, the switching control module is 42.3 seconds before the target closing time. Send the second phase closing command, and control the first and third phases to close sequentially based on the three-phase magnetic circuit coupling results.

[0139] During the reactivation process, the peak inrush current of the third amorphous alloy transformer was measured to be 4.3 times the rated current, the maximum bus voltage drop was 2.0% of the rated voltage, and the three-phase current was unbalanced. The maximum value was 5.6%, none of which exceeded the corresponding... Full threshold, the switching process was completed normally.

[0140] After being reactivated, the actual inrush current energy was 1.28. The actual magnetization recovery energy is 0.74. The actual thermal recovery loss is 0.78. The actual transient additional energy is 2.80. The actual transient additional energy is relative to 2.60. The relative deviation of the predicted value is 7.69%, which does not exceed the 10% transient energy deviation threshold. Therefore, the forced correction of the residual magnetic amplitude is not triggered, and the actual transient energy is simply recorded in the transient energy ledger.

[0141] Adjust At the end of the cycle, the corrected reference loss corresponding to the original operating combination is maintained at 96.9. The actual steady-state cumulative loss after executing the target operating combination is 79.9. In addition, the actual transient additional energy is 2.8. Afterwards, the actual total lifecycle loss was 82.7. The actual net energy saving is 14.2. .

[0142] Actual net energy savings vs. predicted net energy savings: 14.7 The deviation between the two is 3.4%, which does not exceed the 5% loss deviation threshold. Therefore, this operation combination switch does not trigger dynamic loss surface reconstruction, but only writes the actual steady-state loss sample as a valid sample into the loss sample library of the third amorphous alloy transformer.

[0143] This embodiment demonstrates that, in an operating scenario where changes in new energy output cause the net load to first decrease and then rebound, the present invention can not only determine whether it is necessary to temporarily shut down some amorphous alloy transformers, but also pre-calculate the excitation inrush current energy, magnetization recovery energy, and thermal recovery loss generated by their subsequent reactivation into the adjustment parameters. Evaluation can avoid making exit decisions based solely on steady-state loss reduction during low-load periods, thereby improving the accuracy of full-cycle energy efficiency optimization across multiple operating states.

[0144] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A smart grid-based energy efficiency optimization control system for amorphous alloy transformers, characterized in that, The system includes: The system includes a data acquisition module, a perturbation identification module, a loss modeling module, a transient energy management module, a candidate combination generation module, an energy efficiency decision-making module, a switching control module, and a feedback correction module. The data acquisition module is used to collect electrical quantities, core leakage flux, magnetostrictive vibration, core temperature, winding temperature, tap position and circuit breaker status of amorphous alloy transformers, and to perform timing synchronization. The perturbation identification module is used to determine the identification time window when the target amorphous alloy transformer meets the load, voltage and harmonic stability conditions, and to apply a controlled perturbation whose difference between the positive volt-second integral and the negative volt-second integral does not exceed the volt-second balance threshold. The loss modeling module is used to establish a dynamic loss surface between the operating variables and the losses based on the incremental active power, excitation current, core leakage flux and magnetostrictive vibration response before, during and after the perturbation, and the winding resistance is corrected according to the winding temperature. It separates the core loss, winding loss and harmonic additional loss. The transient energy management module is used to determine the residual magnetism state based on the voltage integral, current decay characteristics and opening phase of each phase before and after the circuit breaker is opened, and to associate the residual magnetism state with the outage duration, closing phase, excitation inrush current energy, magnetization recovery energy and thermal recovery loss to form a transient energy ledger. The candidate combination generation module is used to generate candidate operation combinations based on the load forecast results and new energy output forecast results within the scheduling cycle. The energy efficiency decision module is used to determine the steady-state cumulative loss of candidate operating combinations based on the dynamic loss surface, determine the transient additional energy for combination switching based on the transient energy ledger, subtract the transient additional energy from the steady-state cumulative loss difference between the current operating combination and the candidate operating combination to obtain the net energy saving over the entire cycle, and determine the target operating combination. The switching control module is used to determine the closing phase and closing sequence based on the residual magnetism state of the amorphous alloy transformer to be switched on. The feedback correction module is used to correct the dynamic loss surface, transient energy ledger, switching threshold, and minimum hold time based on the actual steady-state loss, actual inrush current energy, and actual bus voltage disturbance after switching.

2. The energy efficiency optimization control system for amorphous alloy transformers used in smart grids according to claim 1, characterized in that, The data acquisition module includes: Electrical quantity acquisition unit, status quantity acquisition unit, time synchronization unit, and data validity processing unit; The electrical quantity acquisition unit is used to acquire the phase voltage, phase current, active power, reactive power, voltage harmonic distortion rate, and current harmonic distortion rate of the primary and secondary sides. The status acquisition unit is used to acquire core leakage flux, magnetostrictive vibration, core temperature, winding temperature, tap position, and circuit breaker status. The time synchronization unit is used to add a unified timestamp to the collected data and perform resampling and time alignment according to the identified sampling period; The data validity processing unit is used to interpolate data whose continuous missing duration does not exceed the missing tolerance, and to mark the time period in which the continuous missing duration exceeds the missing tolerance as an invalid identification time period.

3. The energy efficiency optimization control system for amorphous alloy transformers used in smart grids according to claim 2, characterized in that, The perturbation identification module includes: The system includes a steady-state determination unit, a time window determination unit, a perturbation generation unit, and a perturbation execution unit. The steady state judgment unit is used to calculate the load rate range, the effective value range of the primary side voltage, and the voltage harmonic distortion rate range during the observation period. The time window determination unit is used to determine the time period that continuously meets the stability condition as the identification time window when the three ranges do not exceed the corresponding stability thresholds. The perturbation generation unit is used to generate a perturbation command that sequentially includes a positive perturbation segment and a negative perturbation segment, such that the ratio of the absolute value of the sum of the voltage time integrals of the two segments to the sum of the absolute values ​​of the voltage time integrals of the two segments does not exceed the volt-second balance threshold. The perturbation execution unit is used to execute the perturbation command through the energy storage converter or active filter, and when the core leakage magnetic baseline deviation exceeds the magnetic flux recovery threshold after the perturbation ends, the current identification data is marked as invalid.

4. The energy efficiency optimization control system for amorphous alloy transformers used in smart grids according to claim 3, characterized in that, The loss modeling module includes: Response extraction unit, total loss calculation unit, winding loss calculation unit, harmonic additional loss calculation unit, and core loss calculation unit; The response extraction unit is used to obtain the active power, excitation current, core leakage flux, and magnetostrictive vibration characteristics of the reference section, positive perturbation section, negative perturbation section, and recovery section. The total loss calculation unit is used to determine the difference between the active power on the primary side and the active power on the secondary side as the total loss; The winding loss calculation unit is used to determine the winding loss based on the winding material temperature coefficient, the winding resistance at the reference temperature, the current winding temperature, and the effective value of the current in each phase. The harmonic additional loss calculation unit is used to determine the harmonic additional loss based on the harmonic current and the winding AC resistance at the corresponding frequency. The core loss calculation unit is used to deduct winding losses and harmonic additional losses from the total loss, and the symmetric deviation of the excitation current, core leakage flux and magnetostrictive vibration response under positive and negative perturbations does not exceed the response deviation threshold as a valid condition for the identification result.

5. The energy efficiency optimization control system for amorphous alloy transformers used in smart grids according to claim 4, characterized in that: The dynamic loss surface takes load rate, core temperature, winding temperature, voltage harmonic distortion rate, current harmonic distortion rate, three-phase imbalance, tap position and reactive power compensation as input variables, and core loss, winding loss, harmonic additional loss and total loss as output variables. The loss modeling module is also used to write the input variables and output variables obtained from each effective identification into the loss sample library. When the input variable to be evaluated is within the coverage area of ​​the loss sample, it performs partition interpolation based on the adjacent loss samples. When the input variable to be evaluated is outside the coverage area, or when the number of times the deviation between the actual total loss and the predicted total loss exceeds the loss deviation threshold reaches the update number, a new controlled perturbation is triggered and the dynamic loss surface is updated using the newly added loss samples.

6. The energy efficiency optimization control system for amorphous alloy transformers used in smart grids according to claim 1, characterized in that, The transient energy management module includes: The system includes a tripping recording unit, a residual magnetism estimation unit, a transient energy calculation unit, and a ledger writing unit. The tripping recording unit is used to record the voltage of each phase terminal, the current of each phase, the tripping phase, and the current decay time before and after tripping. The residual magnetism estimation unit is used to integrate the voltage of each phase terminal within the integration interval formed by the opening determination time and the convergence time of each phase terminal voltage, determine the direction of residual magnetism based on the sign of the integration result, and determine the residual magnetism amplitude based on the absolute value of the integration result and the current decay time. The transient energy calculation unit is used to integrate the product of the instantaneous voltage and instantaneous current after closing within the inrush current duration interval to obtain the excitation inrush current energy, integrate the difference between the actual loss and the corresponding steady-state loss within the magnetization recovery interval to obtain the magnetization recovery energy, and integrate the additional copper loss generated during the winding temperature recovery process to obtain the thermal state recovery loss. The ledger writing unit is used to write the excitation inrush current energy, magnetization recovery energy and thermal recovery loss using the transformer identifier, residual magnetization direction, residual magnetization amplitude, outage duration, winding temperature and closing phase as indexes.

7. The energy efficiency optimization control system for amorphous alloy transformers used in smart grids according to claim 5, characterized in that: The candidate combination generation module is used to generate candidate operating combinations based on the active load prediction sequence, reactive load prediction sequence, new energy output prediction sequence, three-phase load distribution prediction sequence, and load duration prediction results. These combinations include maintaining the current combination, putting in standby transformers, taking out operating transformers, and adjusting the load distribution ratio of parallel transformers. The module also deletes candidate operating combinations that do not meet the operating constraints based on the rated capacity, upper limit of winding temperature rise, bus voltage range, upper limit of three-phase imbalance, reactive power demand, and lower limit of standby capacity. The energy efficiency decision module is used to read the predicted losses of the reserved candidate operation combinations from the dynamic loss surface according to the scheduling time interval, and to accumulate the product of the predicted losses and the duration to obtain the steady-state cumulative losses.

8. The energy efficiency optimization control system for amorphous alloy transformers used in smart grids according to claim 7, characterized in that: The energy efficiency decision module is used to select ledger records from the transient energy ledger that have the same residual magnetization direction and whose residual magnetization amplitude, shutdown duration, winding temperature and proposed closing phase are respectively located within a preset grouping interval. The sum of the average values ​​of excitation inrush current energy, magnetization recovery energy and thermal state recovery loss in the selected ledger records is determined as the transient additional energy. The net energy saving over the entire cycle is obtained by subtracting the steady-state cumulative loss of the candidate operating combination and the transient additional energy from the steady-state cumulative loss of the current operating combination. The candidate operating combination with the largest net energy saving over the entire cycle is determined as the target operating combination only when the net energy saving over the entire cycle exceeds the switching threshold and the load duration is not less than the minimum hold time; otherwise, the current operating combination is maintained.

9. The energy efficiency optimization control system for amorphous alloy transformers used in smart grids according to claim 6, characterized in that, The switching control module includes: The closing phase calculation unit, the closing sequence determination unit, and the action delay compensation unit are all included. The closing phase calculation unit is used to calculate the expected magnetic flux peak value at each candidate closing time based on the residual magnetism direction and residual magnetism amplitude of each phase of the amorphous alloy transformer to be put into operation, as well as the magnetic flux increment generated by each phase voltage of the bus at the candidate closing time, and to determine the candidate closing time with the smallest expected magnetic flux peak value as the target closing time. The closing sequence determination unit is used to determine the first closing phase and subsequent closing phases based on the three-phase magnetic circuit coupling relationship, the residual magnetic amplitude of each phase and the phase-by-phase operation conditions of the circuit breaker. The action delay compensation unit is used to compensate for the sending time of the closing command based on the historical closing delay of each phase of the circuit breaker, so that the contacts close at the target closing time.

10. The energy efficiency optimization control system for amorphous alloy transformers for smart grids according to claim 9, characterized in that: The feedback correction module is used to determine the loss deviation between the actual steady-state loss and the predicted steady-state loss, the transient deviation between the actual excitation inrush current energy and the predicted excitation inrush current energy, and the disturbance deviation between the actual bus voltage disturbance and the predicted bus voltage disturbance after the target operating combination has completed the switching. When the loss deviation exceeds the loss deviation threshold, the actual input variables and the actual steady-state loss are written into the loss sample library and the dynamic loss surface is updated. When the transient deviation exceeds the transient deviation threshold, the residual magnetism amplitude is corrected and the transient energy ledger is updated based on the actual closing phase and closing current waveform. The switching threshold is adjusted based on the 95th percentile of the loss deviation, transient deviation and disturbance deviation within the continuous statistical period, and the minimum holding time is adjusted based on the circuit breaker's allowable operating frequency and the predicted load duration.

Images