A transformer operating loss condition optimization monitoring method

Abstract

The present application belongs to the technical field of transformer operation loss monitoring, and in particular relates to a transformer operation loss condition optimization monitoring method. The present application calculates abnormality degree based on oil temperature and vibration characteristics, generates real-time working conditions in combination with load intensity, analyzes real-time operation loss abnormality index in association with cumulative operation duration, realizes dynamic association of loss and actual state of equipment, extracts fluctuation parameters through load characteristic change curve analysis, calculates comprehensive load fluctuation degree, corrects operation loss abnormality index to eliminate load fluctuation interference, finally determines dynamic reference threshold value according to historical loss curve, determines sustained over-limiting through double rules and triggers early warning, simultaneously accurately identifies causes such as load overload and internal heating abnormality, and outputs corresponding adjustment instructions, which improves loss monitoring accuracy and scene adaptability, realizes operation and maintenance closed loop of early warning, cause and disposal, and is suitable for transformer full life cycle loss monitoring.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of transformer operation loss monitoring, and in particular, relates to a transformer operation loss condition optimization monitoring method. BACKGROUND

[0002] As the core equipment of power conversion and transmission in the power system, the operation loss condition of the transformer directly determines the economy of the power system, affects the energy consumption cost and resource utilization rate, and is related to the insulation life of the transformer and the voltage stability of the power system. In order to ensure the safe life of the transformer at the lowest loss cost and support the stable operation of the power grid, the operation loss of the transformer must be optimized and monitored.

[0003] The prior art such as the transformer loss monitoring method and the transformer loss monitoring system disclosed in the Chinese patent application with the application number 202110384183.1 realizes the basis of the internal loss condition of the transformer in the operation process by monitoring the real-time loss of the transformer and giving a warning when the loss of the transformer is greater than the preset range, thereby solving the problem that the existing transformer loss measurement method cannot provide the basis for judging the internal condition of the transformer in the operation process.

[0004] According to the above prior art, the current fault prediction method only calculates the real-time loss value through the voltage and current signals, compares it with the standard loss value, and does not associate the load strength, oil temperature and vibration in the actual operation of the transformer, nor does it consider the dynamic influence of load fluctuation on loss. There are still the following problems: 1. The current method only calculates a single loss value by relying on voltage and current, and compares it with the standard value, ignoring the limitations of load fluctuation and equipment state on loss abnormality.

[0005] 2. The current method only detects the transformer loss by comparing the transformer loss value obtained by real-time calculation with the standard loss value, without considering the influence of equipment state changes and dynamic fluctuations in the actual operation on the standard loss value, resulting in significant deviation in loss determination.

[0006] 3. The current method does not further analyze the alarm causes and does not provide corresponding disposal schemes, resulting in the need for maintenance personnel to spend a lot of time to troubleshoot the root cause and delay fault handling. SUMMARY

[0007] In view of this, in order to solve the above problems, a transformer operation loss condition optimization monitoring method is proposed.

[0008] The purpose of the application can be achieved by the following technical scheme: the application provides a transformer operation loss condition optimization monitoring method, which comprises the following steps: S1, real-time acquisition of the load strength, oil temperature and vibration characteristics of the transformer, and introduction of the transformer input and output power characteristic time sequence.

[0009] S2, based on the oil temperature and vibration characteristics, respectively, the oil temperature abnormality degree and vibration abnormality degree are counted, and the real-time working condition is generated combined with the load intensity, the real-time running loss abnormality index is analyzed based on the real-time working condition and the cumulative running time.

[0010] S3, based on the input and output power characteristics time sequence, the comprehensive load fluctuation degree is obtained through the load characteristic change curve analysis, and the running loss abnormality index is corrected, and the corrected real-time running loss abnormality index is output.

[0011] S4, based on the historical loss curve, the running loss abnormality index reference threshold is determined, the running loss abnormality is judged according to the preset abnormality judgment rule, when it is judged that it is continuously over limit, the early warning is triggered, and the early warning cause and the corresponding adjustment instruction are identified.

[0012] Compared with the prior art, the beneficial effects of the present application are as follows: (1) the present application corrects the preset rated power loss by the loss factor, so that the final running loss abnormality index not only covers the influence of equipment state abnormality, but also adapts the difference of aging degree, avoids the one-sidedness of single loss comparison, makes the final corrected loss abnormality index more truly reflect the actual loss abnormality of the transformer, and provides accurate data support for subsequent judgment.

[0013] (2) the present application constructs the load change curve based on the input and output power time sequence, calculates the input and output load fluctuation degree respectively, judges the correlation of the two by the Pearson correlation coefficient to obtain the comprehensive load fluctuation degree, and then corrects the running loss abnormality index, which not only quantifies the interference of load fluctuation on loss, but also avoids the problem of repeated calculation of input and output fluctuation, so that the corrected loss abnormality index can more truly reflect the comprehensive state of loss value and fluctuation interference, and avoids the loss abnormality misjudgment or omission caused by ignoring fluctuation.

[0014] (3) the present application accurately locates four kinds of early warning causes based on load intensity, oil temperature abnormality degree, vibration abnormality degree, comprehensive load fluctuation degree and other indexes, and formulates customized adjustment instruction for each kind of cause, so as to avoid blind operation and maintenance, make the transformer loss monitoring from problem discovery to problem solving closed-loop management, improve the operation and maintenance response efficiency, and avoid the risk of equipment damage caused by untimely abnormality disposal. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical scheme of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description, and obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creating any creative labor.

[0016] Figure 1 This is a schematic diagram of the monitoring process for transformer operating losses according to the present invention.

[0017] Figure 2 This is a schematic diagram of the analysis process for the transformer operating loss anomaly index of the present invention.

[0018] Figure 3 This is a schematic diagram of the process for extracting the fluctuation amplitude of the present invention.

[0019] Figure 4 This is a schematic diagram of the extraction process for the duration of a single fluctuation in this invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see Figure 1 As shown, the present invention provides an optimized monitoring method for transformer operating losses, characterized in that the method includes: S1, real-time acquisition of transformer load intensity, oil temperature and vibration characteristics, and simultaneous import of transformer input and output power characteristic timing.

[0022] It should be noted that vibration characteristics include, but are not limited to, vibration frequency, vibration amplitude, and duration of vibration exceeding limits. Among them, the duration of vibration exceeding limits is the maximum value between the duration of vibration frequency exceeding limits and the duration of vibration amplitude exceeding limits.

[0023] It should be further noted that the threshold values ​​for judging excessive vibration frequency and excessive vibration amplitude are determined by taking the normal vibration range specified in the transformer's technical manual as the initial threshold, and then calibrating the values ​​according to relevant industry standards. The calibrated values ​​are then used as the threshold values ​​for judgment.

[0024] S2. Based on the oil temperature and vibration characteristics, respectively, the oil temperature anomaly and vibration anomaly are statistically analyzed, and combined with the load intensity, real-time operating conditions are generated. Based on the real-time operating conditions and the cumulative running time, the real-time operating loss anomaly index is analyzed.

[0025] The statistical process for the oil temperature anomaly and vibration anomaly includes: retrieving the preset normal oil temperature range according to the transformer model, and retrieving the median temperature of the normal oil temperature range.

[0026] If the oil temperature collected at a certain time point is not located in the preset normal range of oil temperature, a relative deviation value of the oil temperature collected at the certain time point and the temperature median is calculated, and the maximum relative deviation value is taken as the oil temperature abnormality degree.

[0027] The preset vibration feature normal range is called, and the vibration abnormality degree is obtained in the same way as the calculation of the oil temperature abnormality degree.

[0028] Exemplarily, it is assumed that the vibration features involved in the present application include three vibration indexes of vibration frequency, vibration amplitude and vibration overrun duration, if multiple vibration indexes collected at a certain time point are not located in the corresponding index normal range, a vibration abnormality value is obtained by calculating the relative deviation value of each vibration index collected at the certain time point and the corresponding vibration index and performing weighted summation, and the maximum vibration abnormality value is taken as the vibration abnormality degree, if a single vibration index collected at a certain time point is not located in the corresponding index normal range, the vibration abnormality degree is obtained in the same way as the calculation of the oil temperature abnormality degree.

[0029] It should be noted that the weights of the vibration frequency, the vibration amplitude and the vibration overrun duration can be set by referring to the recommended values in the manufacturer's technical manual of the transformer.

[0030] Please refer to Figure 2 The analysis process of the operation loss abnormality index includes: calculating the difference between the input power and the output power of the transformer, denoted as the actual total loss, and calling the preset rated power loss of the transformer.

[0031] The usage degree is obtained according to the proportion of the cumulative running time and the effective use time of the transformer.

[0032] The load intensity, the oil temperature abnormality degree and the vibration abnormality degree of the transformer allowed to carry under the corresponding usage degree are matched based on the usage degree, and the allowed carrying working condition is obtained by combination.

[0033] By comparing each index in the real-time working condition with the allowed carrying working condition, if there is an index exceeding the corresponding allowed carrying index, the following steps are performed: the index proportion exceeding the corresponding allowed carrying index and the exceeding degree of each exceeding index are counted, and the exceeding degree is quantified by the relative deviation value.

[0034] The loss factor is obtained from the working condition loss reference table according to the index proportion exceeding and the exceeding degree of each exceeding index, and the reference rated power loss is obtained by modifying the preset rated power loss.

[0035] It should be noted that the modification refers to multiplying the loss factor by the preset rated power loss.

[0036] Further, the exceeding degree of each exceeding index and the index proportion exceeding in the above-mentioned working condition loss reference table constitute an associated mapping relationship together with the loss factor.

[0037] The data in the working condition loss reference table can be obtained from the technical manual of the transformer manufacturer and combined with the influence degree of the cumulative running time of the transformer on the loss.

[0038] If the actual total loss is less than or equal to the reference rated power loss, the operation loss abnormality index is 0.

[0039] On the contrary, the relative deviation value of the actual total loss and the reference rated power loss is calculated as the operation loss abnormality index.

[0040] It should be noted that the index ratio is used to quantify the proportion of the number of dimensions of the equipment deviating from the normal permitted working condition, which can directly reflect the deviation range of the overall operation state of the equipment. The exceeding degree is used to accurately measure the severity of the single item exceeding the limit, clearly the influence weight of a single indicator on equipment loss, and the matching loss factor of the two is used for rated power loss correction, mainly to avoid the correction deviation caused by single parameter judgment, avoid ignoring the overall risk superposition caused by multiple indicators slightly exceeding the limit, and at the same time avoid missing the local risk amplification caused by single indicator serious exceeding, finally make the corrected reference rated power loss accurately fit the actual operation risk of multiple dimensions and multiple severity of the equipment, and ensure the accuracy of the subsequent operation loss abnormality index calculation.

[0041] S3, based on the input and output power characteristic time sequence, the comprehensive load fluctuation degree is obtained through load characteristic change curve analysis, and the running loss abnormality index is corrected based on the comprehensive load fluctuation degree. Output the corrected real-time running loss abnormality index.

[0042] Understandably, the above correction means multiplying the comprehensive load fluctuation degree and the running loss abnormality index.

[0043] Wherein, the process of load characteristic change curve analysis includes: taking time as abscissa and input power as ordinate to construct input load change curve, counting fluctuation frequency of the curve, and extracting fluctuation amplitude and single fluctuation interval length.

[0044] The single fluctuation interval length is compared with the single fluctuation interval length under the normal running state of the transformer. If the single fluctuation interval length is greater than the single fluctuation interval length under the normal running state of the transformer, the input load fluctuation degree is obtained by normalizing and weighting the sum of the comprehensive fluctuation frequency and the fluctuation amplitude.

[0045] It should be noted that the weight of the above comprehensive fluctuation frequency and fluctuation amplitude can be obtained based on industry experience, such as taking 0.6 as the weight of the fluctuation amplitude and taking 0.4 as the weight of the fluctuation frequency.

[0046] Otherwise, the complement of the ratio of the adjacent fluctuation interval length and the reference fluctuation interval length is taken as the fluctuation intensity, and the input load fluctuation degree is obtained by normalizing and weighted sum of the comprehensive fluctuation intensity, fluctuation frequency and fluctuation amplitude.

[0047] The output load fluctuation degree is obtained in the same way as the input load fluctuation degree.

[0048] The Pearson correlation coefficient of the input load fluctuation degree and the output load fluctuation degree is calculated, and if the Pearson correlation coefficient is positive and less than a preset correlation threshold, the comprehensive fluctuation degree is calculated by weighted average method.

[0049] If the Pearson correlation coefficient is equal to the preset correlation threshold, the comprehensive load fluctuation degree is calculated by variance fusion.

[0050] It should be noted that the above-mentioned preset correlation threshold can take the industry experience value 0.7 as the early warning threshold, and after 2-4 weeks of trial operation, according to the sensitivity and false positive rate, the principle of insufficient adjustment is adjusted by 0.02-0.05, and finally the preset correlation threshold is obtained.

[0051] It should be pointed out that the weights of the above-mentioned comprehensive fluctuation intensity, fluctuation frequency and fluctuation amplitude, and the weights of the input load fluctuation degree and the output load fluctuation degree can be based on the transformer model and call the corresponding industry influence priority weight corresponding to the corresponding transformer model as the weight.

[0052] For example: if it is an industrial transformer, the fluctuation amplitude weight is 0.4, the fluctuation intensity weight is 0.35, and the fluctuation frequency weight is 0.25. If it is a civil distribution transformer, the fluctuation intensity weight is 0.4, the fluctuation frequency weight is 0.35, and the fluctuation amplitude weight is 0.25.

[0053] For example: the input load fluctuation degree weight of the industrial transformer is 0.6, and the output load fluctuation degree weight is 0.4. The input load fluctuation degree weight of the civil distribution transformer is 0.4, and the output load fluctuation degree weight is 0.6. The input load fluctuation degree weight of the new energy supporting transformer is 0.75, and the output load fluctuation degree weight is 0.25.

[0054] It should be noted that, in the above calculation of the comprehensive load fluctuation degree, the corresponding industry priority weight is retrieved according to the difference in load characteristics of different types of transformers, to ensure that the fluctuation degree calculation is consistent with the actual operation scene of the transformer, rather than a unified standard. According to the Pearson correlation coefficient of the input and output load fluctuation degree, different calculation methods are matched, which highlights the key side fluctuation influence and balances the uncertainty of the double-side fluctuation, so that the comprehensive fluctuation degree can better reflect the real load change trend. The accurate quantitative comprehensive load fluctuation degree can effectively distinguish between normal load fluctuation and abnormal load fluctuation, avoid misjudgment of slight fluctuation of loss caused by normal load change as abnormal, and finally improve the accuracy of operation loss abnormality judgment and reduce false alarm or missed alarm.

[0055] It should also be noted that the load characteristic change curve presents the continuous change trend of the load intuitively with time as the axis, which can clearly reflect the continuity, suddenness and periodicity of the fluctuation and other dynamic characteristics. Pure data analysis cannot completely restore these time sequence characteristics through discrete numerical values, avoiding the bias of equating different change patterns with the same numerical values, facilitating targeted extraction of key fluctuation parameters, and the curve as a visual carrier can convert abstract parameters such as high fluctuation frequency and large amplitude into intuitive morphological features, making the subsequent weight allocation, fusion calculation and other logic more easily correspond to the actual operation scene.

[0056] Please refer to Figure 3 The fluctuation amplitude extraction process includes: extracting a starting time point and an ending time point to form an analysis time period, and counting the analysis duration.

[0057] The maximum amplitude is extracted from the curve, and the average amplitude is calculated, and the proportion of the duration of the curve segment to which the maximum amplitude belongs to the analysis duration is obtained as the fluctuation duration proportion.

[0058] If the fluctuation duration proportion is greater than a preset duration proportion threshold, the maximum amplitude is taken as the final extracted fluctuation amplitude.

[0059] If the fluctuation duration proportion is less than the preset duration proportion threshold, the difference between the maximum amplitude and the average amplitude is calculated, and when the difference is less than a preset fluctuation difference threshold, the average amplitude is taken as the final extracted fluctuation amplitude.

[0060] It should be noted that the above-mentioned preset duration proportion threshold is a critical value for determining whether the value of the fluctuation amplitude is the maximum amplitude or the average amplitude. The preset duration proportion threshold can be determined comprehensively based on the factory design parameters and technical specifications of the transformer and the actual application load characteristics of the transformer, and the comprehensive determination result is taken as the preset duration proportion threshold.

[0061] When the difference is greater than the preset fluctuation difference threshold, the average value of the maximum amplitude and the average amplitude is taken as the final extracted fluctuation amplitude.

[0062] It should be noted that the above preset fluctuation difference threshold is to determine whether the value of the fluctuation amplitude is the maximum amplitude or the critical value of the average of the maximum amplitude and the average amplitude under the premise that the fluctuation duration ratio is less than the preset duration ratio threshold. The preset fluctuation difference threshold can be determined based on the factory design parameters and core component tolerance of the device, combined with the statistical data of the difference between the maximum amplitude and the average amplitude in the short-term fluctuation scene in the actual application scene characteristics and historical operation of the device, and also can be calibrated and verified by industry standards, and dynamically iterated and optimized according to the state change of the device. The final optimized result is used as the fluctuation difference threshold.

[0063] Please refer to Figure 4 The extraction process of the single fluctuation interval duration includes extracting the interval duration of each adjacent peak point from the curve, extracting the minimum interval duration and the average interval duration.

[0064] The curve is divided into curve segments with the peak points as the division points, and the duration of each curve segment is extracted.

[0065] The proportion of the number of curve segments with a duration less than the average interval duration to the total number of curve segments is calculated and recorded as the dense curve segment ratio.

[0066] If the dense curve segment ratio is less than the first preset ratio threshold, the average interval duration is taken as the single fluctuation interval duration.

[0067] If the dense curve segment ratio is greater than the first preset ratio threshold and less than the second preset ratio threshold, the average of the minimum interval duration and the average interval duration is taken as the single fluctuation interval duration, and the second preset ratio threshold is greater than the first preset ratio threshold.

[0068] If the dense curve segment ratio is greater than the second preset ratio threshold, the minimum interval duration is taken as the single fluctuation interval duration.

[0069] It should be noted that the first preset ratio threshold and the second preset ratio threshold can be located and distinguished by the distribution characteristics of the proportion of the curve segments with a duration less than the average interval duration, which are the natural dividing points of low density and high density, and are used as the initial thresholds of the first preset ratio threshold and the second preset ratio threshold. The first preset ratio threshold and the second preset ratio threshold are finally output after adjusting the initial thresholds according to the application scene of the device.

[0070] Further, by constructing an input load change curve, extracting three core parameters of fluctuation frequency, fluctuation amplitude and single fluctuation interval duration, and combining with the type of transformer, the input or output load fluctuation degree is calculated through normalization weighting, which converts the abstract load change into specific quantitative indicators, avoids subjective judgment of load fluctuation, and makes the subsequent correction of the abnormal loss index have a clear reference to the load state. At the same time, through multiple threshold judgments, such as the first preset proportion threshold and the second preset proportion threshold, the normal load fluctuation and the abnormal load fluctuation can be accurately distinguished, and by taking the normal influence of load fluctuation on loss into account, the modified real-time operation loss abnormal index can be more in line with the real loss state of the equipment, providing more reliable basic data for subsequent continuous over-limit judgment and early warning trigger, and reducing the probability of false alarm or missed alarm.

[0071] S4, determining an operation loss abnormal index benchmark threshold based on a historical loss curve, performing operation loss abnormality judgment according to a preset abnormality judgment rule, triggering an early warning when the judgment is continuous over-limit, identifying a warning cause, and triggering a corresponding adjustment instruction.

[0072] The acquisition process of the loss abnormality index benchmark threshold comprises: selecting a loss curve segment located in a preset effective analysis time window from the historical loss curve as a target loss curve segment, extracting the slope of the target loss curve segment, and extracting the operation loss abnormality index of each historical operation from the target loss curve segment.

[0073] Further, the setting of the above-mentioned preset effective analysis time window needs to be combined with the abnormal loss law to exclude invalid or interference periods, determine the basic time range, determine the window length, and verify the rationality of the preset window with historical normal operation data. Illustratively, the preset effective analysis time window can be set to 24 hours.

[0074] It should be noted that the slope of the above-mentioned target loss curve segment can be extracted by using linear fitting, wherein the extraction of the slope by linear fitting is a prior art and will not be described and illustrated in detail in the present application.

[0075] If the slope is less than a preset abnormality threshold, or the slope is greater than the preset abnormality threshold and the difference between the slope and the preset abnormality threshold is less than a preset difference threshold, the average of the loss abnormality index is taken as the loss abnormality index benchmark threshold.

[0076] It should be noted that the above-mentioned preset abnormality threshold provides a reasonable basis for the subsequent judgment of whether the slope is abnormal. The preset abnormality threshold can be based on a historical normal slope sample set, and an initial value is obtained by statistical calculation, and then a critical value is obtained through actual verification and long-term data dynamic optimization.

[0077] It should be noted that the above historical normal slope sample set is obtained by screening samples consistent with the current target loss curve segment scene from long period historical loss data, and then cutting the effective curve segment of the samples according to the preset effective analysis time window rule, calculating the slope of each effective curve segment by linear fitting, and finally forming the historical normal slope sample by integration.

[0078] Otherwise, the relative deviation value of the difference value and the preset abnormal threshold value is calculated, the relative deviation value is input as an input variable into an exponential decay function to obtain a decay factor, and the running loss abnormal index is corrected as a loss abnormal index benchmark threshold.

[0079] It should be noted that the above correction means multiplying the decay factor and the running loss abnormal index.

[0080] Further, the above exponential decay function is: , wherein is a natural constant. is a decay coefficient determining the decay speed, The larger the decay coefficient is, the faster the decay is, and the decay coefficient The initial value of the decay coefficient can be determined according to the type of the device and the industry experience, and the initial value is verified through historical abnormal data to finally obtain the decay coefficient . is a relative deviation value. is a decay factor.

[0081] The deviation rate determination rule determines the content of the deviation rate, which is obtained by calculating the deviation rate of the corrected real-time running loss abnormal index and the running loss abnormal index benchmark threshold.

[0082] It should be noted that the above deviation rate is a relative deviation value.

[0083] If the deviation rate is greater than the preset warning deviation rate for a duration greater than the preset duration threshold or the number of times that the deviation rate is greater than the preset warning deviation rate within the preset time window is greater than the preset occurrence threshold, it is determined that the deviation rate is continuously out of limit, otherwise it is determined that the deviation rate is not continuously out of limit.

[0084] Further, the above preset duration threshold is a duration boundary value for defining transient interference and persistent risk. The preset duration threshold can be obtained by retrieving the design parameters of the device from the manufacturer's manual of the transformer, determining the initial threshold by combining the historical damage data of the same type of transformer, and then correcting the initial threshold according to the running scene of the transformer to output the final threshold as the preset duration threshold.

[0085] It should be noted that the above design parameters include the heat aging tolerance time of the insulation material and the overload tolerance duration of the winding wire.

[0086] It should be noted that the above preset time window is a fixed statistical cycle for determining whether the cumulative frequency of the abnormality reaches the risk standard, and the preset time window can be determined by analyzing historical data to filter out invalid transient fluctuations and samples of failure precursor abnormalities, determining an initial basic time length based on the principle of covering effective early warning precursor period, and then adjusting the initial basic time length according to the device type, load characteristics and abnormal type to obtain the final preset time window. Exemplarily, the preset time window can be set to 50 minutes.

[0087] It should be noted that the above preset warning deviation rate is a critical value for determining the real-time abnormality deviation rate that needs to trigger attention. The preset warning deviation rate can be calculated by calculating the deviation rate of the real-time loss abnormality index in the historical normal operation period data and the reference threshold value, forming a historical normal deviation rate sample set, and then calculating the historical normal deviation rate sample set by quantile method to obtain the corresponding preset warning deviation rate.

[0088] It should be noted that the above preset occurrence threshold is a critical number for determining that the deviation rate in the preset time window exceeds the warning value and needs to trigger continuous over-limit determination. The preset occurrence threshold can be obtained by referring to the experience value of similar devices in the industry.

[0089] The identification content of the warning cause includes: if the load intensity is greater than the preset intensity threshold, the warning cause is load overload.

[0090] If the oil temperature abnormality degree is greater than the preset oil temperature abnormality degree threshold, the warning cause is internal heating abnormality.

[0091] If the vibration abnormality degree is greater than the preset vibration abnormality degree threshold, the warning cause is structural failure.

[0092] If the comprehensive load fluctuation degree is greater than the preset load fluctuation degree threshold, the warning cause is load instability.

[0093] It should be noted that the load intensity quantifies the actual working load of the device, and the preset intensity threshold is referenced based on the rated bearing capacity of the device. If the load intensity is greater than the preset intensity threshold, it means that the actual load exceeds the design safety range, so it directly corresponds to load overload.

[0094] Further, the above oil temperature abnormality degree threshold reflects the deviation of oil temperature from the normal interval. The normal oil temperature is determined by the normal loss heat of the device and is stable. If the oil temperature abnormality degree is greater than the preset oil temperature abnormality degree threshold, it means that the oil temperature deviates from the reasonable range, and the root cause is the generation of additional abnormal heat inside, i.e. internal heating abnormality.

[0095] It should be noted that the vibration abnormality threshold, i.e., the deviation of vibration from the normal reference, is the inherent slight vibration of the mechanical movement of the equipment, and if the vibration abnormality is greater than the preset vibration abnormality threshold, it means that the vibration balance is broken, which only occurs when the structural component fails, so it corresponds to the structural failure.

[0096] Further, the comprehensive load fluctuation quantifies the load fluctuation amplitude and frequency, and the preset load fluctuation threshold corresponds to the stable fluctuation range under normal working conditions, and if the comprehensive load fluctuation is greater than the preset load fluctuation threshold, it means that the load suddenly rises or falls or fluctuates at a high frequency, which is beyond the normal working condition category, and the essence is that the load is unstable.

[0097] The adjustment content of the adjustment instruction includes: when the early warning guide is due to load overload, triggering a load adjustment instruction.

[0098] When the early warning guide is due to internal heating anomaly, triggering an oil tank oil detection temperature adjustment instruction.

[0099] When the early warning guide is due to structural failure, triggering a structure maintenance adjustment instruction.

[0100] When the early warning guide is due to load instability, triggering a load distribution adjustment instruction.

[0101] It should be noted that when the early warning guide is due to load overload, the adjustment instruction is to take the non-core load on the load side as the object, and adjust by cutting off the non-core load. When the early warning guide is due to internal heating anomaly, the adjustment instruction is to take the internal heating detection equipment as the object, and adjust by stopping the operation of the winding. When the early warning guide is due to structural failure, the adjustment instruction is to take the transformer core fixing structure as the object, and adjust by tightening the core clamping bolt. When the early warning guide is due to load instability, the adjustment instruction is to take the load-side electrical equipment as the object, and adjust by coordinating the load peak-shaving start-stop to optimize the load distribution.

[0102] The above content is only an example and description of the concept of the present application, and those skilled in the art can make various modifications or supplements or use similar ways to replace the described specific embodiments, as long as they do not deviate from the concept of the present application or exceed the scope defined by the present application, which shall be within the protection scope of the present application.

Claims

1. A method for optimizing and monitoring transformer operating losses, characterized in that, The method includes: S1. Real-time acquisition of transformer load intensity, oil temperature and vibration characteristics, and import of transformer input and output power characteristic timing. S2. Based on the oil temperature and vibration characteristics, the oil temperature anomaly and vibration anomaly are statistically analyzed respectively, and real-time operating conditions are generated in combination with the load intensity. Based on the real-time operating conditions and the cumulative running time, the real-time operating loss anomaly index is analyzed. The difference between the transformer's input power and output power is calculated and recorded as the actual total loss. The preset rated power loss of the transformer is then retrieved. The usage level is obtained based on the ratio of the cumulative operating time to the effective usage time of the transformer. Based on the usage level, the permissible load intensity, oil temperature anomaly, and vibration anomaly of the transformer under the corresponding usage level are matched and combined to obtain the permissible load condition. The indicators in the real-time operating condition and the permissible load condition are compared. If any indicator exceeds the corresponding permissible load indicator, the following steps are performed: the proportion of indicators exceeding the corresponding permissible load indicator and the degree of exceeding each exceeding indicator are statistically analyzed, and the degree of exceeding is quantified by a relative deviation value. The loss factor is obtained from the operating condition loss reference table based on the proportion of exceeding indicators and the degree of exceeding each exceeding indicator, and the preset rated power loss is corrected to obtain the reference rated power loss. If the actual total loss is less than or equal to the reference rated power loss, the operating loss anomaly index is 0; otherwise, the relative deviation value between the actual total loss and the reference rated power loss is calculated as the operating loss anomaly index. S3. Based on the input and output power characteristic time series, the comprehensive load fluctuation is obtained by analyzing the load characteristic change curve. Based on the comprehensive load fluctuation, the operating loss anomaly index is corrected and the corrected real-time operating loss anomaly index is output. S4. Determine the benchmark threshold for the abnormal operating loss index based on the historical loss curve, and determine the abnormal operating loss according to the deviation rate judgment rule. When it is determined to be continuously exceeding the limit, an early warning is triggered, and the cause of the early warning is identified and the corresponding adjustment instruction is triggered.

2. The method for optimizing and monitoring transformer operating losses as described in claim 1, characterized in that: The statistical process for the oil temperature anomaly and vibration anomaly includes: Based on the transformer model, retrieve the preset normal oil temperature range and then retrieve the median temperature within that normal range. If the oil temperature collected at a certain time point is not within the preset normal oil temperature range, calculate the relative deviation between the oil temperature collected at that time point and the median temperature, and take the maximum relative deviation as the oil temperature anomaly degree; The normal range of the preset vibration characteristics is retrieved, and the vibration anomaly degree is obtained in the same way as the calculation of the oil temperature anomaly degree.

3. The method for optimizing and monitoring transformer operating losses as described in claim 1, characterized in that: The process of analyzing the load characteristic change curve includes: Using time as the horizontal axis and input power as the vertical axis, an input load variation curve is constructed. The fluctuation frequency of the curve is statistically analyzed, and the fluctuation amplitude and the duration of a single fluctuation interval are extracted. The duration of the single fluctuation interval is compared with the duration of the single fluctuation interval under normal transformer operation. If the duration of the single fluctuation interval is longer than the duration of the single fluctuation interval under normal operation, the input load fluctuation degree is obtained by normalizing and weighting the comprehensive fluctuation frequency and fluctuation amplitude. Otherwise, the complement of the ratio of the duration of the adjacent fluctuation interval to the duration of the reference fluctuation interval will be taken as the fluctuation intensity. The input load volatility will be obtained by normalizing and weighting the comprehensive fluctuation intensity, fluctuation frequency, and fluctuation amplitude. The output load fluctuation is obtained in the same way as the input load fluctuation. Calculate the Pearson correlation coefficient between the input load volatility and the output load volatility. If the Pearson correlation coefficient is positive and less than the preset correlation threshold, calculate the overall volatility using the weighted average method. If the Pearson correlation coefficient is equal to the preset correlation threshold, the overall load volatility is calculated by variance fusion.

4. The method for optimizing and monitoring transformer operating losses as described in claim 3, characterized in that: The process of extracting the fluctuation amplitude includes: Extract the start and end times to form the analysis period, and then calculate the analysis duration. Extract the maximum amplitude from the curve and calculate the average amplitude. At the same time, locate the ratio of the duration of the curve segment to which the maximum amplitude belongs to the analysis duration to obtain the fluctuation duration ratio. If the duration of fluctuation exceeds the preset duration threshold, the maximum amplitude will be used as the final extracted fluctuation amplitude. If the duration of fluctuation is less than the preset duration threshold, calculate the difference between the maximum amplitude and the average amplitude. When the difference is less than the preset fluctuation difference threshold, use the average amplitude as the final extracted fluctuation amplitude. When the difference is greater than the preset fluctuation difference threshold, the average of the maximum amplitude and the average amplitude will be used as the final extracted fluctuation amplitude.

5. The method for optimizing and monitoring transformer operating losses as described in claim 3, characterized in that: The process for extracting the duration of a single fluctuation includes: Extract the interval duration of each adjacent peak point from the curve, and extract the minimum interval duration and the average interval duration; Using the peak point as the dividing point, the curve is divided into curve segments, and the duration of each curve segment is extracted; The ratio of the number of curve segments whose duration is shorter than the average interval to the total number of curve segments is denoted as the dense curve segment ratio. If the ratio of dense curve segments is less than the first preset ratio threshold, the average interval duration is taken as the single fluctuation interval duration. If the ratio of dense curve segments is greater than the first preset ratio threshold but less than and less than or equal to the preset second preset ratio threshold, the average of the minimum interval duration and the average interval duration is taken as the single fluctuation interval duration, and the second preset ratio threshold is greater than the first preset ratio threshold. If the ratio of dense curve segments is greater than the second preset ratio threshold, the minimum interval duration is taken as the single fluctuation interval duration.

6. The method for optimizing and monitoring transformer operating losses as described in claim 1, characterized in that: The process of obtaining the benchmark threshold for the anomaly index includes: The loss curve segment within the preset effective analysis time window is selected from the historical loss curve and recorded as the target loss curve segment. The slope of the target loss curve segment is extracted, and the abnormal operation loss index of each historical run is extracted from the target loss curve segment. If the slope is less than the preset abnormal threshold, or the slope is greater than the preset abnormal threshold and the difference between the slope and the preset abnormal threshold is less than the preset difference threshold, then the mean of the loss abnormality index will be used as the benchmark threshold for the loss abnormality index. Otherwise, the relative deviation between the difference and the preset abnormal threshold is calculated, and the relative deviation is used as an input variable to input the exponential decay function to obtain the decay factor. The abnormal operating loss index is then corrected as the benchmark threshold for the abnormal loss index.

7. The method for optimizing and monitoring transformer operating losses as described in claim 1, characterized in that: The determination criteria for the deviation rate include: The real-time anomaly deviation rate is obtained by calculating the deviation rate between the corrected real-time operating loss anomaly index and the benchmark threshold of the operating loss anomaly index. If the duration of the deviation rate exceeding the preset warning deviation rate is greater than the preset duration threshold, or if the cumulative number of times the deviation rate exceeds the preset warning deviation rate within the preset time window is greater than the preset occurrence threshold, it is determined to be a continuous over-limit; otherwise, it is determined to be a non-continuous over-limit.

8. The method for optimizing and monitoring transformer operating losses as described in claim 1, characterized in that: The identification content of the warning cause includes: If the load intensity exceeds the preset intensity threshold, an alarm will be triggered due to overload. If the oil temperature abnormality exceeds the preset oil temperature abnormality threshold, an alarm will be triggered due to abnormal internal heating. If the vibration anomaly exceeds the preset vibration anomaly threshold, an early warning will be issued indicating a structural fault. If the overall load fluctuation exceeds the preset load fluctuation threshold, an early warning will be issued because the load is unstable.

9. The method for optimizing and monitoring transformer operating losses as described in claim 8, characterized in that: The adjustment content of the adjustment instruction includes: When an early warning is triggered due to overload, a load adjustment command is executed. When the warning system detects abnormal internal heating, it triggers a fuel tank oil temperature adjustment command. When a structural fault is detected by the early warning system, a structural maintenance and adjustment command is triggered. When an early warning is triggered due to unstable load, a load distribution adjustment command is initiated.

Images