An analysis method and system for energy saving and loss reduction of a transformer

Abstract

The application discloses an analysis method and system for energy saving and loss reduction of a transformer, comprising: collecting standby state data of a transformer group, identifying a high-loss unit and predicting a parameter drift trend thereof; acquiring an electromagnetic shock intensity signal in a working state; fusing the parameter drift trend and the electromagnetic shock intensity signal, evaluating a risk of synergistic disorder; and adaptively performing dynamic voltage regulation and load balancing optimization according to a risk evaluation result and a risk cause. Through cross-condition correlation analysis and closed-loop control on the transformer group as a power supply core, the application not only can effectively reduce standby no-load loss and delay equipment aging, reduce operation cost, but also can ensure long-term stability and reliability of a power supply system, and can be widely applied to industrial fields such as pulse electric fields and precision manufacturing.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power transmission and transformation automatic control, in particular to an analysis method and system for transformer energy saving and loss reduction. BACKGROUND

[0002] Pulsed electric field (PEF) technology is a non-thermal food processing technology that can achieve sterilization or improve product quality without significantly increasing food temperature, thus having significant advantages in preserving the original flavor, color and nutritional ingredients of food, and is gradually being applied in the food industry. When PEF technology is applied on a large scale in industrial production, the long-term operation cost and stability of the high-voltage pulse power supply system, which is the core equipment of PEF technology, become a technical bottleneck for its widespread application. Industrial production lines usually have the characteristics of intermittent work, and during the production interval of changing food batches, pipe cleaning or equipment maintenance, the PEF power supply system is often in a long standby state. When processing food, the power supply system needs to output high-power electric pulses of microseconds with high frequency and high stability.

[0003] This coexistence of "long standby" and "instant impact" working conditions raises two major technical problems. First, the energy efficiency and cost problem. During standby, there is a continuous no-load loss in the transformer group inside the power supply system, which directly translates into operating electricity expenses. At the same time, during production, the equipment ages due to repeated high-power electromagnetic shocks, which can lead to a decrease in energy conversion efficiency, further increasing the processing cost per product. Second, the stability and quality problem. The food processing industry has very high requirements for the uniformity of product quality, and the electrical parameters of each transformer unit in the power supply system will gradually drift under repeated shocks, causing the coordinated work of each unit to be out of sync. This imbalance can cause the pulse electric field strength output by the PEF device to be unstable, directly affecting the consistency of food processing results, which can lead to incomplete sterilization or over-processing, and thus cause food safety risks or product scrap.

[0004] The energy efficiency and stability problems caused by the coexistence of "long standby" and "instant impact" working conditions are common in many industries that use transformer groups for power supply, and are a common technical bottleneck that restricts the operation cost and reliability of such systems. SUMMARY

[0005] The technical problem to be solved by the present application is the energy efficiency decay and stability degradation of the transformer group as the power supply core under complex working conditions with intermittent and impact loads.

[0006] To achieve the above object, the application provides an analysis method for energy saving and loss reduction of a transformer, which is applied to a power supply system comprising a plurality of transformer units, and comprises the following steps:

[0007] Step one, collecting and processing standby state data to obtain the no-load loss distribution characteristics of the power supply system. Specifically, the current, voltage data and temperature indicators of each transformer unit in the power supply system in the standby state during production intervals are collected in real time, and the Fourier transform method is used to process the collected raw data, so as to obtain the no-load loss distribution characteristics capable of representing the energy consumption level of each transformer unit.

[0008] Step two, identifying high-loss units and fitting their parameter drift trends. According to the no-load loss distribution characteristics, a high-loss unit set is selected from the transformer group constituting the power supply system, and the electrical parameter history records of each unit in the set are extracted. If the extracted records show that the parameter drift trend exceeds the preset degradation threshold, the least square method is used to fit the history records to obtain the parameter drift trend curve representing the performance degradation law. The application uses the no-load loss data generated by the power supply system in the standby state as the basis for diagnosing the health status and predicting future performance degradation, thereby performing cross-condition correlation diagnosis on the state of the power supply system.

[0009] Step three, obtaining the electromagnetic shock intensity signal in the working state. For the transformer units that have been identified to have significant parameter drift trends, the electromagnetic shock intensity signals acting on these units when the power supply system is in high-power output are further obtained, which reflect the electrical stress level of the equipment in the working state.

[0010] Step four, fusion of multi-source information and collaborative disorder risk level evaluation. The Kalman filtering algorithm is used to fuse and process the parameter drift trend curve and the electromagnetic shock intensity signal. This step combines the inherent aging trend of the equipment with the impact of the actual working condition to dynamically judge and quantitatively evaluate the collaborative disorder risk level that may be caused by the inconsistent performance of each unit in the power supply system.

[0011] Step five, adaptively assigning weight to risk source and determining adjustment strategy. After judging the level of synergy disorder risk, further dynamically assess the main factors leading to parameter drift. Specifically, analyze the correlation between the recent fluctuation characteristics (e.g. frequency, amplitude change rate) of the electromagnetic impact intensity signal and the slope change of the parameter drift trend curve. If the correlation is high, it indicates that the recent high-intensity processing task is the main cause of the accelerated parameter drift, at which time a higher adjustment weight will be given to the "suppress aging" goal; otherwise, if the correlation is low, it indicates that the parameter drift is more due to device inherent aging or environmental factors, at which time a higher adjustment weight will be given to the "standby energy saving" goal. The present application can further trace the causes of the risk on the basis of predicting the risk, and dynamically adjust the focus of the subsequent control strategy according to different causes, so as to achieve more targeted adaptive adjustment.

[0012] Step six, performing dynamic voltage adjustment and optimizing energy distribution. If the level of synergy disorder risk judged in step four is higher than the preset risk threshold, then according to the adjustment strategy weight determined in step five, the input voltage of the low-loss unit set with relatively stable performance in the power supply system is adjusted through the dynamic voltage adjustment protocol, so as to obtain an optimized energy distribution matrix, which is used to balance the energy saving and anti-aging needs of the entire system.

[0013] Step seven, updating load balancing configuration and determining device aging suppression threshold. According to the optimized energy distribution matrix, when the power supply system is instantaneously switched from standby state to high-load working state, the load balancing configuration of the transformer group in the power supply system is updated using the matrix. This configuration transfers more load and impact stress to units in better health, while determining a device aging suppression threshold for units whose performance has already drifted, to slow down their further performance degradation.

[0014] Step eight, calibrating energy efficiency evaluation value and outputting final running parameters. From the device aging suppression threshold, a correction coefficient is derived to represent the difference in system stability. The correction coefficient is used to calibrate the energy efficiency evaluation value of the entire production line, so that the energy efficiency evaluation not only considers the immediate power consumption, but also takes into account the long-term impact due to the decline in stability, and finally obtains a set of final running parameters for guiding the operation of the power supply system.

[0015] Another aspect of the present application also provides an analysis system for transformer energy saving and loss reduction, which is configured to perform the above method, comprising:

[0016] A data acquisition module for acquiring current-voltage data and temperature indicators of each transformer unit in the power supply system in standby state during production intermittence in real time;

[0017] a loss analysis module configured to process the data using a Fourier transform method to obtain no-load loss distribution characteristics of the transformer bank, and identify a high-loss unit according to the characteristics;

[0018] a trend prediction module configured to extract an electrical parameter history record of the high-loss unit, and fit the record using a least squares method to obtain a parameter drift trend curve;

[0019] a risk assessment module configured to obtain an electromagnetic impact strength signal, and fuse the parameter drift trend curve and the electromagnetic impact strength signal using a Kalman filter algorithm to determine a level of risk of synergistic disorder inside the power supply system;

[0020] an adaptive adjustment module configured to adaptively assign adjustment weights according to an attribution analysis of the risk sources;

[0021] an optimal control module configured to, if the determined level of risk of synergistic disorder is higher than a preset threshold, adjust an input voltage of the low-loss unit set through a dynamic voltage adjustment protocol to obtain an optimized energy distribution matrix, and update a load balancing configuration according to the optimized energy distribution matrix;

[0022] a parameter calibration module configured to derive a correction coefficient of the poor stability index from a device aging suppression threshold, and calibrate an overall production line energy efficiency evaluation value using the correction coefficient to obtain a final set of power supply system operating parameters.

[0023] Compared with the prior art, the present application has the following beneficial effects:

[0024] The present application performs correlation analysis and predictive closed-loop control on data in the standby and pulse working states of the power supply system, which helps to ensure the long-term stability and accuracy of the output pulse of the PEF and similar devices with similar working characteristics, thereby ensuring the consistency of the working effect and reducing the risk of product quality fluctuations caused by uneven processing. At the same time, through real-time analysis of the no-load loss in the standby state of the power supply system and identification of the high-loss unit, energy distribution can be optimized to reduce energy waste during non-production time; by actively suppressing device aging, the system can maintain high energy conversion efficiency during the production process, thereby reducing operating costs. In addition, the closed-loop adjustment and load balancing strategy used in the present application can actively slow down the aging process of key electrical components, which helps to improve the overall reliability and uptime of the PEF production line, and prolong the service life of the power supply system. BRIEF DESCRIPTION OF DRAWINGS

[0025] Embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0026] Figure 1 is a flowchart of an energy-saving loss-reducing analysis method for a pulse electric field power supply system provided by an embodiment of the present application.

[0027] Figure 2 is a structural block diagram of an energy-saving loss-reducing analysis system for a pulsed electric field power supply system provided by an embodiment of the present application.

[0028] Figure 3 is a schematic diagram for illustrating parameter drift trend and electromagnetic shock strength signal in an embodiment of the present application.

[0029] Figure 4 is a schematic diagram of multi-source information fusion using Kalman filtering algorithm in an embodiment of the present application.

[0030] Figure 5 is a schematic diagram of adaptive allocation of adjustment strategy weight in an embodiment of the present application.

[0031] Figure 6 is a schematic diagram of transformer group no-load loss distribution and high-loss unit identification in a specific application scenario in an embodiment of the present application.

[0032] Figure 7 is a comparative schematic diagram of the impact of load balancing strategy before and after optimization on unit shock current in an embodiment of the present application.

[0033] Figure 8 is a comparative curve diagram of traditional energy efficiency evaluation and calibrated energy efficiency evaluation in an embodiment of the present application. DETAILED DESCRIPTION

[0034] To make the purpose, technical solutions and advantages of the present application clearer, the specific embodiments of the present application will be described in detail below with reference to the drawings. It should be understood that the specific embodiments herein are only used to explain the present application and not to limit the present application.

[0035] To make those skilled in the art better understand the technical solutions of the present application, the present application will be described in detail below with pulsed electric field (PEF) power supply system as a typical application scenario. It should be understood that the inventive concept of the present application is not limited to this specific scenario, and any system that uses transformer group power supply and has similar "long standby" and "instantaneous shock" working conditions can apply the technical solutions of the present application to achieve the purpose of energy saving and loss reduction and stability improvement.

[0036] Embodiment 1

[0037] The present embodiment provides an energy-saving loss-reducing analysis method for a pulsed electric field (PEF) power supply system, which is applied to scenarios such as food processing production lines. The power supply system is usually composed of a transformer group composed of multiple transformer units in parallel or series, which provides high-voltage power supply for PEF equipment. As shown in the figure, the method comprises the following steps: Figure 1 ​

[0038] Step S1, collect and process standby state data to obtain the no-load loss distribution characteristics of the power supply system.

[0039] When the PEF production line enters the production interval, such as replacing batch materials or cleaning the equipment, the power supply system is in standby state. This step uses this standby window to make a preliminary diagnosis of the health status of the system. Specifically, by deploying high-precision current sensors (such as Hall effect sensors) and voltage sensors (such as voltage dividers) at the input of each transformer unit, real-time acquisition of current and voltage waveform data in standby state is carried out at a sampling frequency of not less than 10 kHz. At the same time, temperature data is collected by temperature sensors (such as PT100 thermal resistance) attached to the core or winding surface of each transformer unit. Since the load current in standby state is very small, the input power at this time mainly reflects the no-load loss, including the hysteresis loss and eddy current loss in the core. Due to device aging or insulation performance degradation, the standby current waveform may be distorted. Therefore, this step preferably uses the Fast Fourier Transform (FFT) algorithm to process the collected non-sinusoidal current and voltage waveform data, and decomposes it into fundamental and harmonic components. The no-load loss (P_loss) can be calculated according to the following formula:

[0040] where n is the harmonic number, V_n and I_n are the effective values of the n-th harmonic voltage and current, and φ_n is the phase difference between the n-th harmonic voltage and current. By calculating all units in the transformer group, a no-load loss distribution characteristic that can quantitatively represent the energy consumption level of each unit is obtained.

[0041] Step S2, identify high-loss units and fit their parameter drift trend.

[0042] Based on the no-load loss distribution characteristics obtained in step S1, a high-loss screening threshold is set to identify abnormal performance units from the transformer group. Preferably, the threshold can be set to 1.5 times the average no-load loss of all units, or 120% higher than the factory calibration value. For the high-loss units screened out, the historical records of their key electrical parameters, such as no-load current, winding resistance, insulation resistance, etc., are retrieved from the system's historical database. If the records show that the parameters present a monotonic or fluctuating drift trend over time, and the cumulative drift exceeds the preset degradation threshold (for example, the winding resistance changes by more than 5%), it is considered that the unit has obvious performance degradation. At this time, the least squares method is used to fit the historical data points of the parameter to obtain a parameter drift trend curve that can represent the performance degradation law. For example, it can be fitted as a quadratic polynomial where P(t) is the parameter value, t is the running time, a, b, c are coefficients determined by fitting the historical data points by least squares method. This curve can be used to predict the performance state of the unit in the future period of time. As shown in Figure 3 the parameter drift trend 301 is the drift trend curve of the electrical parameter (such as no-load current) of a certain high-loss unit over time. This step associates the standby state energy consumption data with the historical electrical parameters, realizing the cross-condition, predictive diagnosis of the device health condition.

[0043] Step S3, acquire electromagnetic shock intensity signals in the working state.

[0044] For the transformer unit identified in step S2 as having a significant parameter drift trend, it is necessary to further evaluate the stress level it bears in the actual working state. When the PEF device is processing food, i.e. the power supply system outputs high-voltage pulses at high frequency, through high-bandwidth Rogowski coils and other sensors, the pulse current waveform acting on these units is collected. The electromagnetic shock intensity signal can be quantified by calculating the rise rate (di / dt) or peak amplitude of the pulse current. For example, the di / dt peak value of each pulse is taken as a data point to form a time series signal. As shown in Figure 3 302 is a series of electromagnetic shock intensity pulses that the unit bears in the working state. This signal directly reflects the electrical stress that the device bears when working, and is a key external factor for analyzing performance accelerated degradation.

[0045] Step S4, fuse multi-source information and perform collaborative misadjustment risk assessment.

[0046] To achieve accurate evaluation of the overall stability of the power supply system, Kalman filtering algorithm is adopted in this step to fuse the parameter drift trend curve (from step S2) representing the intrinsic aging law of the equipment and the electromagnetic shock intensity signal (from step S3) reflecting the external working condition stress. The state prediction model of the Kalman filter is based on the parameter drift trend curve 401 to predict the electrical parameter state at the next moment. The state update link of the Kalman filter takes the real-time acquired electromagnetic shock intensity signal 402 as the external input to correct the prediction result. Specifically, the size of the shock intensity signal can be associated with the process noise covariance matrix Q, that is, when the shock intensity increases, the value of Q is increased accordingly, indicating that the uncertainty of the system state is increased. Through the iterative operation of the filter, a more accurate parameter state estimation value 403 that fuses the long-term trend and the instantaneous shock influence is obtained. Based on the estimation value, the parameter difference between the unit and other healthy units is evaluated. When the difference exceeds the allowable range of cooperative work, it is judged that there is a risk of cooperative disorder. The risk level can be quantified, for example, defined as (estimated parameter value - average parameter value of healthy unit) / average parameter value of healthy unit. This fusion processing method can dynamically and quantitatively judge the risk of instability of the entire power supply system caused by the performance degradation of individual units.

[0047] Step S5, adaptively assigning weights to risk sources and determining adjustment strategies.

[0048] After judging the level of cooperative disorder risk, this step further traces the main cause of the risk in order to take more targeted adjustment strategies. Specifically, the Pearson correlation coefficient between the fluctuation characteristics (such as standard deviation or change rate) of the electromagnetic shock intensity signal in the recent time window (for example, the past week) and the change of the slope of the parameter drift trend curve (i.e. the parameter degradation acceleration) is calculated. If the correlation is high (for example, the correlation coefficient is greater than 0.7), it indicates that the recent heavy load and high intensity processing task is the main reason for the accelerated drift of the parameter, and the focus of the system adjustment strategy should be placed on “suppressing aging”, giving this goal a higher adjustment weight, such as the weight W_aging in Figure 5 . On the contrary, if the correlation is low (for example, the correlation coefficient is less than 0.3), it indicates that the parameter drift is more due to the inherent natural aging of the equipment or environmental temperature and other factors, and the focus of the system adjustment should be placed on “standby energy saving”, that is, the high loss problem in standby state is given priority to, and a higher adjustment weight is given to this goal, such as the weight W_standby in Figure 5 . This step realizes the optimal allocation of control resources by intelligently attributing the problem to the root cause, so that the subsequent control adjustment can be adaptively adjusted according to the root cause.

[0049] Step S6, dynamic voltage regulation and optimized energy distribution.

[0050] If the level of risk of synergistic imbalance determined in step S4 is higher than the preset threshold (for example, the risk quantification value is greater than 0.1), the optimization control is started according to the weight of the adjustment strategy determined in step S5. In this step, a dynamic voltage regulation protocol is used to fine-tune the input voltage of the low-loss unit set with relatively stable performance in the power supply system. For example, if the weight of "aging suppression" is high, the protocol will instruct the solid-state voltage regulator or thyristor voltage regulator connected to the front end of the power supply system to slightly reduce the input voltage of the high-loss unit and slightly increase the input voltage of the healthy unit. The goal of this operation is to generate an optimized energy distribution matrix. This matrix is a diagonal matrix, and the diagonal elements represent the adjustment coefficients of the voltage (or power) of each transformer unit. For high-loss units that need to be protected, the coefficient is less than 1; for healthy units that need to bear more load, the coefficient is greater than 1, and the adjustment of all coefficients needs to ensure that the total output power of the system remains unchanged.

[0051] Step S7, update load balancing configuration and determine aging suppression threshold.

[0052] According to the optimized energy distribution matrix generated in step S6, when the production line is instantaneously switched from standby state to PEF pulse processing state, the central controller of the power supply system will use the matrix to update the load balancing configuration of the transformer group. Specifically, when distributing pulse power tasks, more load (i.e., higher pulse current) is directed to healthy units with adjustment coefficients greater than 1, while the load allocated to high-loss units is correspondingly reduced. At the same time, a dynamic device aging suppression threshold is determined for these high-loss units, which is a specific upper limit of an operating parameter, for example, limiting the peak operating current to within 80% of its rated value, or controlling the maximum operating temperature to be 5 degrees Celsius lower than the normal value. This actively slows down the further degradation of its performance and extends its effective service life through "de-rating use".

[0053] Step S8, calibrate energy efficiency evaluation value and output final operating parameters.

[0054] The traditional energy efficiency evaluation value only focuses on the instantaneous input-output power ratio, ignoring the long-term stability and reliability costs caused by inconsistent device performance. This step is used to make up for this deficiency. From the device aging suppression threshold determined for the high-loss units in step S7, a correction coefficient is derived to represent the difference in system stability. For example, the correction coefficient C = 1- (unit rated peak current - device aging suppression threshold) / unit rated peak current. For healthy units that are not limited, their correction coefficient is 1. Then, the product of the correction coefficients of all units is used to calibrate the energy efficiency evaluation value of the entire production line. The calibrated energy efficiency value wherein η_original is the original energy efficiency evaluation value, Π(C_i) is the product of all unit correction coefficients. This calibrated energy efficiency index not only measures the instantaneous power consumption, but also contains the consideration of the long-term stable operation capability of the system. Finally, the system outputs a set of final operation parameters based on this comprehensive evaluation, including the updated load balancing strategy, the aging suppression threshold of each unit, and the optimized standby management scheme, forming a complete closed-loop optimization control.

[0055] Embodiment 2

[0056] The embodiments of the present application also provide an energy-saving loss-reducing analysis system for a pulse electric field power supply system, as shown in the figure, which is configured to perform the method described in Embodiment 1. The system can be embedded in the central controller of the PEF power supply system or exist as an independent monitoring and analysis unit. The system comprises: Figure 2

[0057] The data acquisition module 201 is used to collect the current, voltage data and temperature indicators in the standby state of production intermittence through the sensors deployed on each transformer unit of the power supply system, and transmit the data to the subsequent modules.

[0058] The loss analysis module 202 internally integrates a Fourier transform processing unit, which is used to process the raw data sent by the data acquisition module 201, calculate the no-load loss of each transformer unit, form the no-load loss distribution characteristics of the power supply system, and identify the high-loss units according to the preset rules.

[0059] The trend prediction module 203 is used to connect the historical database, extract the electrical parameter history records of the high-loss units identified by the loss analysis module 202, and through the logic unit of the built-in least square method or other fitting algorithms, fit the historical records to obtain the parameter drift trend curve representing the performance degradation.

[0060] The risk assessment module 204 is used to obtain the electromagnetic shock intensity signal in the working state from the sensor network, and internally integrate a Kalman filter algorithm processor. The core function of this module is to fuse the parameter drift trend curve output by the trend prediction module 203 and the real-time electromagnetic shock intensity signal, dynamically judge and quantitatively evaluate the risk level of the coordination disorder inside the power supply system.

[0061] The adaptive adjustment module 205 is used to perform attribution analysis on the risk output by the risk assessment module 204. Through analyzing the correlation between the electromagnetic shock intensity signal and the parameter drift change, this module adaptively allocates adjustment weights for the two goals of “aging suppression” and “standby energy saving” to determine the subsequent control strategy. ​

[0062] The optimization control module 206 is activated when it receives a signal that the risk level is above the threshold. It generates and executes dynamic voltage regulation instructions to adjust the input voltage of the low-loss unit set according to the weights determined by the adaptive adjustment module 205, so as to obtain an optimized energy distribution matrix, and updates the load balancing configuration of the system according to the matrix.

[0063] The parameter calibration module 207 derives a stability difference correction coefficient from the device aging suppression threshold determined by the optimization control module 206, and calibrates the energy efficiency evaluation value of the overall production line using the coefficient, and finally generates and outputs a set of final operating parameters for guiding the operation of the PEF power supply system.

[0064] The present application performs correlation analysis and closed-loop control on the power supply system through the above-mentioned method and system, thereby saving energy and reducing consumption, and maintaining the stability and high efficiency of the system during long-term operation.

[0065] Embodiment 3

[0066] This embodiment will further illustrate the actual operation process and effect of the method described in the present application in a specific application scenario. In a specific application scenario, a pulse electric field (PEF) system for juice sterilization is deployed on a certain juice processing production line, and the high-voltage pulse power supply system thereof is composed of 12 parallel transformers (Unit 1 to Unit 12). After two years of system operation, the energy-saving and loss-reducing analysis system described in the present application is introduced.

[0067] During the 30-minute standby window period of a production batch replacement in the early stage of system application, the method of the present application starts to execute. First, the data acquisition module acquires the standby current, voltage and temperature data of all 12 transformer units. After Fourier transform processing by the loss analysis module, it is found that the no-load loss of most units is stable at about 80 watts, while the no-load loss of Unit 4 and Unit 7 is as high as 110 watts and 135 watts respectively, which are identified by the system as high-loss units. Figure 6

[0068] Subsequently, the trend prediction module automatically retrieves the running history database of Unit 4 and Unit 7 units in the past two years. The data shows that the winding resistance of Unit 7 unit has increased by 6.2% compared with the factory value, exceeding the preset degradation threshold of 5%. The system uses the least squares method to fit the historical data points of the resistance of Unit 7 unit, and obtains a quadratic polynomial parameter drift trend curve:

[0069] ​where t is the running month, and R(t) is the predicted resistance value (Ohm). This indicates that the aging process of Unit 7 has shown an accelerating trend.

[0070] In the following production cycle, when the PEF system is pulsed to process fruit juice, the risk assessment module begins to focus on Unit 7. By deploying a Rogowski coil on the output side of Unit 7, it obtains the real-time electromagnetic shock intensity signal it is subjected to and finds that when processing high-viscosity fruit pulp, the peak value of the rising rate (di / dt) of the pulse current frequently reaches 1.5 kA / μs. The Kalman filter algorithm in the risk assessment module takes this high shock intensity signal as an external input to real-time correct the resistance state prediction value based on the aforementioned drift curve. After fusion processing, the system obtains a more accurate current equivalent resistance estimate of Unit 7 and calculates the parameter difference with the healthy unit based on this, and the quantified level of synergistic disorder risk is 0.13, which exceeds the system-set risk threshold of 0.1.

[0071] At this time, the adaptive adjustment module starts risk attribution analysis. By calculating the Pearson correlation coefficient between the average value of daily electromagnetic shock intensity and the change in the slope of the Unit 7 resistance drift curve in the past month, a high correlation result of 0.85 is obtained. This indicates that the recent high-intensity production task is the main cause of the accelerated degradation of Unit 7 performance. Therefore, the system adaptively tilts the adjustment strategy weight to “suppress aging” and gives it a weight of 0.8, while the weight of “standby energy saving” is 0.2.

[0072] Since the level of synergistic disorder risk has exceeded the threshold, the optimization control module generates an optimized energy distribution matrix according to the “suppress aging” priority strategy through the dynamic voltage adjustment protocol. The matrix instructs the solid-state voltage regulator connected to the front end of the power system to adjust the input voltage coefficient of Unit 4 and Unit 7 to 0.9, while setting the voltage coefficient of the four units with the best performance state to 1.05 to compensate for the total power output. When the production line switches from standby to working state, the matrix is used to update the load balancing configuration to transfer more pulse load tasks to units with better performance. At the same time, the system determines an aging suppression threshold for Unit 7, i.e., limits its peak working current to within 85% of its rated value. As Figure 7 shown, the optimized load balancing strategy significantly reduces the impact current that Unit 4 (Unit 4) and Unit 7 (Unit 7) with high loss are subjected to when working, while letting the healthy Unit 2 (Unit 2) and Unit 5 (Unit 5) bear more load.

[0073] Finally, the parameter calibration module derives a stability difference correction coefficient C_T6=0.85 from the 85% current limit, and uses it to calibrate the energy efficiency evaluation value of the overall production line, so that the energy efficiency evaluation result can comprehensively reflect the long-term health status of the system. Finally, the system outputs a final set of operation parameters including the latest load balancing scheme, Unit 4 and Unit 7 unit operation limits, and the optimized standby strategy, and issues them to the central controller of the PEF power system for execution. After three months of continuous optimization operation, the monitoring data shows that the resistance drift rate of Unit 7 unit is slowed down by about 40%, the unit energy consumption of the production line per ton of juice is reduced by 3.5%, and the uniformity of the product sterilization effect is improved. As shown in FIG. 8, the calibrated energy efficiency evaluation curve can reflect the long-term operation benefit of the system compared with the traditional energy efficiency evaluation. Figure 8

[0074] The above merely illustrates the preferred embodiments of the present application, but is not intended to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.​

Claims

1. An analysis method for energy saving and loss reduction of transformers, which is applied to a power supply system including a plurality of transformer units, characterized by, The method comprises: Collecting current data and voltage data of each transformer unit in the standby state of the power supply system, and processing the collected current data and voltage data by using the Fourier transform method to determine the no-load loss distribution characteristics capable of representing the energy consumption level of each transformer unit; According to the no-load loss distribution characteristics, identifying high-loss units and low-loss units from the plurality of transformer units; Extracting the electrical parameter history record of the high-loss unit, and fitting the electrical parameter history record by using the least squares method to obtain a parameter drift trend curve representing the performance degradation rule of the high-loss unit; For the high-loss unit, obtain the electromagnetic shock intensity signal acting on the high-loss unit in the high-power output working state of the power supply system, and the electromagnetic shock intensity signal reflects the electrical stress level borne by the high-loss unit in the working state; Using Kalman filtering algorithm, the parameter drift trend curve and the electromagnetic shock intensity signal are fused and processed to dynamically quantify and evaluate the risk level of the collaborative disorder caused by the inconsistent performance of the high-loss unit and other transformer units; If the collaborative disorder risk level is higher than the preset risk threshold, adjust the input voltage of the low-loss unit with stable performance in the plurality of transformer units by the dynamic voltage regulation protocol to obtain an optimized energy distribution matrix; and according to the optimized energy distribution matrix, update the load balancing configuration of the plurality of transformer units when the power supply system switches from the standby state to the high-load working state; the step of updating the load balancing configuration of the plurality of transformer units comprises: transferring the load and shock stress to the low-loss unit, and determining a device aging suppression threshold for the high-loss unit, wherein the device aging suppression threshold is the upper limit of the peak working current or the upper limit of the highest working temperature of the high-loss unit; derive a correction coefficient from the device aging suppression threshold for representing the system stability difference; calibrate an original energy efficiency evaluation value by using the correction coefficient to obtain a calibrated energy efficiency value taking into account the long-term impact of the stability decline; and output a final operating parameter set for controlling the operation of the power supply system based on the calibrated energy efficiency value; the calculation method of the correction coefficient is: correction coefficient C = 1-(unit rated peak current-device aging suppression threshold) / unit rated peak current; and the calculation method of the calibrated energy efficiency value η_calibrated is: where η_original is the original energy efficiency assessment value, Π(C_i) is the product of all transformer unit correction factors, and i is the index of the transformer unit.

2. The method of claim 1, wherein, According to the no-load loss distribution characteristics, identifying high-loss units from the plurality of transformer units, specifically comprising: Setting a high-loss screening threshold; and Comparing the no-load loss values of each transformer unit with the high-loss screening threshold, and identifying the transformer unit with the no-load loss value higher than the high-loss screening threshold as a high-loss unit.

3. The method of claim 1, wherein, The method further comprises: After dynamically quantitatively evaluating the synergistic disorder risk level, a Pearson correlation coefficient between fluctuation characteristics of the electromagnetic shock intensity signal and slope change of the parameter drift trend curve in a preset time window is calculated to obtain a correlation between the two; and Based on the Pearson correlation coefficient, adaptive adjustment weights are assigned to the aging inhibition target and the standby energy-saving target, wherein the adjustment weight assignment content includes: If the Pearson correlation coefficient is higher than a preset first correlation threshold, the aging inhibition target is given a higher adjustment weight than the standby energy-saving target; and If the Pearson correlation coefficient is lower than a preset second correlation threshold, the standby energy-saving target is given a higher adjustment weight than the aging inhibition target.

4. The method of claim 1, wherein, The collected data in the standby state further includes temperature indicators of each transformer unit.

5. The method of claim 1, wherein, by calculating the rise rate di / dt or the peak amplitude of the pulsed current acting on the high-loss element c where d represents the differential operator, i represents the current, and t represents the time. c by calculating the rise rate di / dt or the peak amplitude of the pulsed current acting on the high-loss element 6. An analysis system for energy saving and loss reduction of transformers, which is applied to a power supply system including a plurality of transformer units, characterized by, The system includes: A data collection module for collecting current data and voltage data of each transformer unit in the standby state of the power supply system; A loss analysis module for processing the current data and voltage data collected by the data collection module using a Fourier transform method to determine no-load loss distribution characteristics capable of representing energy consumption levels of each transformer unit, and identifying high-loss units and low-loss units according to the no-load loss distribution characteristics; A trend prediction module for extracting electrical parameter historical records of the high-loss units and fitting the electrical parameter historical records by least squares to obtain a parameter drift trend curve representing performance degradation rules of the high-loss units; A risk assessment module for obtaining electromagnetic shock intensity signals acting on the high-loss units in a high-power output state of the power supply system, and fusing the parameter drift trend curve and the electromagnetic shock intensity signals using a Kalman filter algorithm to dynamically quantitatively evaluate synergistic disorder risk levels that may be caused by inconsistent performance of the high-loss units and other transformer units; An optimization control module for adjusting input voltages of low-loss units with stable performance in the multiple transformer units by a dynamic voltage regulation protocol to obtain an optimized energy distribution matrix if the synergistic disorder risk level is higher than a preset risk threshold; and According to the optimized energy distribution matrix, updating load balancing configurations of the multiple transformer units when the power supply system switches from the standby state to the high-load working state; The step of updating the load balancing configurations of the multiple transformer units includes: Transferring loads and shock stresses to the low-loss units, and determining a device aging inhibition threshold for the high-loss units, wherein the device aging inhibition threshold is an upper limit of a peak working current or an upper limit of a maximum working temperature of the high-loss units; A parameter calibration module for deriving a correction coefficient representing system stability differences from the device aging inhibition threshold; Calibrating an original energy efficiency evaluation value using the correction coefficient to obtain a calibrated energy efficiency value taking into account long-term effects due to stability decline; and Based on the calibrated energy efficiency value, output a final set of operation parameters for controlling operation of the power supply system; The correction coefficient is calculated in the following manner: correction coefficient C = 1 - (unit rated peak current - equipment aging suppression threshold) / unit rated peak current; and the calibrated energy efficiency value η_calibrated is calculated in the following manner: where η_original is the original energy efficiency assessment value, Π(C_i) is the product of all transformer unit correction factors, and i is the index of the transformer unit.

Images