A power equipment monitoring analysis and early warning system

Abstract

The application relates to the technical field of power transformers, in particular to a power equipment monitoring, analyzing and early warning system, which comprises a data acquisition module, which is used for acquiring the running state of power equipment, wherein the running state comprises the working period and running data of the power equipment; the running data comprises real-time load change information, running temperature, overvoltage shock times, water content of insulating oil and dielectric loss tangent of insulating oil; a trend identification module, which is used for acquiring the fault characteristics of the power equipment under the corresponding running state based on the running state of the power equipment, identifying the fault trend of the power equipment, and calculating the fault coefficient of the power equipment according to the fault trend of the power equipment; and a fault analysis module, which is used for generating the corresponding fault analysis tree of the power equipment based on the fault coefficient, and sequentially determining the fault relationship coefficient, data fault proportion and fault period of the power equipment; the reliability and precision of power equipment detection are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power transformers, in particular to a power equipment monitoring, analyzing and early warning system. BACKGROUND

[0002] The stable operation of power equipment is crucial to the reliability of the power system. However, power equipment is affected by various factors during long-term operation, such as environmental conditions, load changes, aging, etc., which may cause equipment performance degradation or even failure. Traditional power equipment maintenance methods mainly rely on periodic inspection and experience-based judgment, which is not only inefficient, but also difficult to identify potential failure risks in a timely manner.

[0003] For example, Chinese Patent Publication No. CN115936535A discloses a method, device, electronic device and storage medium for improving the service life of a power transformer. The method includes constructing a health status indicator for the power transformer, calculating the indicator weight of the health status indicator, calculating the indicator membership degree of the health status indicator, performing health status analysis on the power transformer to obtain a health analysis result of the power transformer, performing fault analysis on the power transformer to obtain a fault analysis index, collecting risk history data of the power transformer, calculating the cycle loss degree and service life cost of the power transformer, performing preventive risk analysis on the power transformer to obtain a risk analysis result, and performing residual cycle analysis on the power transformer to obtain a residual analysis result.

[0004] For example, Chinese Patent Publication No. CN117172555A discloses an intelligent generation system for power grid risk early warning notification, which includes a transformer information acquisition module, an environmental information acquisition module, a transformer overload risk analysis module, a transformer operating temperature risk analysis module, a transformer oil explosion risk analysis module, a transformer risk level confirmation module, a cloud database and a power grid risk early warning notification generation terminal. The present application calculates the transformer operation risk evaluation index through the overload risk level, the operating temperature risk level and the oil explosion risk level, confirms the risk level and the risk reason of the transformer, realizes the multi-dimensional analysis of the transformer early warning notification, improves the coverage of the transformer early warning notification analysis, and reduces the error of the risk level and the risk reason analysis.

[0005] The prior art mainly describes the risk problems of the cycle and the transformer in the oil explosion aspect, but when evaluating the overall transformer, the control of the current operating condition of the transformer is insufficient, the current, voltage, temperature and corresponding insulating oil of the transformer cannot be specifically identified in time, which leads to the inability to predict the change trend of the current transformer under the condition of overload high temperature, resulting in the inability to identify the state of the transformer subsequently, and causing the reduction of identification accuracy. SUMMARY

[0006] To solve the above technical problems, the technical scheme adopted by the present application is: a power equipment monitoring analysis and early warning system, comprising: a data acquisition module, configured to acquire a running state of power equipment, the running state comprising a working cycle and running data of the power equipment; the running data comprising real-time load change information, running temperature, overvoltage shock times, water content of insulating oil and dielectric loss tangent of insulating oil.

[0007] A trend identification module is configured to acquire a fault feature of the power equipment under a corresponding running state based on the running state of the power equipment, and identify a fault trend of the power equipment, and calculate a fault coefficient of the power equipment according to the fault trend of the power equipment.

[0008] A fault analysis module is configured to generate a fault analysis tree corresponding to the power equipment based on the fault coefficient, and sequentially determine a fault relationship coefficient, a data fault proportion and a fault cycle of the power equipment.

[0009] A fault influence module is configured to evaluate a fault loss of the power equipment according to the fault cycle, and calculate a fault influence coefficient of the power equipment in a corresponding fault cycle.

[0010] An influence range determination module is configured to analyze an influence range coefficient of running data that appears to be faulty according to the fault relationship coefficient and the data fault proportion.

[0011] An early warning evaluation module is configured to calculate a fault promotion coefficient of the power equipment under different cycle lengths and values of running data according to the fault influence coefficient and the influence range coefficient, and obtain an analysis and early warning result of the current power equipment.

[0012] The present application has the beneficial effects that: the present application acquires the running state of the power equipment in real time through the data acquisition module, including the working cycle and the running data, ensuring the timeliness and accuracy of the data; the trend identification module identifies the fault trend of the equipment based on the historical data and the current running state, calculates the fault coefficient, and improves the accuracy of fault diagnosis; the fault analysis module generates the fault analysis tree, determines the fault relationship coefficient, the data fault proportion and the fault cycle, and provides detailed fault cause analysis; the fault influence module evaluates the specific loss of the fault to the equipment and the system, calculates the fault influence coefficient, and provides a basis for decision-making; the influence range determination module analyzes the influence range coefficient of the fault on other parts of the system, ensuring the comprehensiveness of the maintenance measures; the early warning evaluation module calculates the fault promotion coefficient, generates the analysis and early warning result of the current power equipment, and provides scientific maintenance suggestions. BRIEF DESCRIPTION OF DRAWINGS

[0013] The present application will be further described below in combination with the drawings and examples.

[0014] Figure 1It is a system framework diagram of a power equipment monitoring analysis and early warning system.

[0015] Figure 2 It is a system schematic diagram of a power equipment monitoring analysis and early warning system.

[0016] Figure 3 It is a flowchart of a trend identification module of a power equipment monitoring analysis and early warning system.

[0017] Figure 4 It is a flowchart of a fault impact module of a power equipment monitoring analysis and early warning system. DETAILED DESCRIPTION

[0018] Embodiments of the present application are described in detail below. The embodiments described below are exemplary only, and are not to be construed as limiting the present application. Unless otherwise noted, technical or conditions not specified in the embodiments are performed according to the techniques or conditions described in the literature in the art or according to the product instructions.

[0019] Reference Figure 1 A power equipment monitoring analysis and early warning system, comprising: a data acquisition module, a trend identification module, a fault analysis module, a fault impact module, an impact range determination module, and a warning evaluation module.

[0020] The data acquisition module acquires the operating state of the power equipment, including the working cycle and the operating data, and outputs the data to the trend identification module; the trend identification module acquires the fault characteristics of the equipment based on the operating state of the power equipment, and outputs to the fault analysis module; the fault analysis module identifies the relevant data of the fault characteristics, and outputs to the fault impact module; the fault impact module evaluates the fault cycle, and outputs the corresponding data to the impact range determination module; the impact range determination module analyzes the influence degree of the fault on other parts of the system, and outputs to the warning evaluation module; the warning evaluation module comprehensively considers the influence of the current fault, and completes the analysis of the fault.

[0021] As Figure 2 shown, the data acquisition module is used to acquire the operating state of the power equipment, and the operating state includes the working cycle and the operating data of the power equipment; the operating data includes real-time load change information, operating temperature, overvoltage shock times, water content of insulating oil, and dielectric loss tangent of insulating oil.

[0022] The trend identification module is used to acquire the fault characteristics of the power equipment under the corresponding operating state based on the operating state of the power equipment, and identify the fault trend of the power equipment, and calculate the fault coefficient of the power equipment according to the fault trend of the power equipment.

[0023] The fault analysis module is configured to generate a fault analysis tree corresponding to the power equipment based on the fault coefficient, and sequentially determine a fault relationship coefficient, a data fault proportion, and a fault period of the power equipment.

[0024] The fault influence module is configured to evaluate a fault loss of the power equipment according to the fault period, and calculate a fault influence coefficient of the power equipment in the corresponding fault period.

[0025] The influence range determination module is configured to analyze an influence range coefficient of the running data that appears to be faulty according to the fault relationship coefficient and the data fault proportion.

[0026] The early warning evaluation module is configured to calculate a fault promotion coefficient of the power equipment under different period lengths and values of the running data according to the fault influence coefficient and the influence range coefficient, and obtain an early warning result of the current power equipment.

[0027] At this time, the power equipment used is a transformer, and all power equipment in the following text represents a transformer. The calculation of the fault is to analyze whether the performance of the transformer degrades during the overall use, and what causes the performance degradation, so as to perform real-time early warning on the transformer. When the performance of the transformer degrades too fast or the performance of the transformer is significantly low, the transformer is replaced in time to improve the effect of the analysis of the power equipment.

[0028] The transformer may have overloading and overheating, and oil quality degradation. When these two situations occur, the running state can be identified to determine whether these abnormal running conditions occur at this time. In addition, the use time of the transformer can also cause the transformer to have a certain degree of performance degradation, which has a smaller impact than overloading and overheating and oil quality degradation. Therefore, at this time, the overloading and overheating and oil quality degradation are mainly identified, and these situations are processed according to the identified period to verify the form of influence on the performance of the transformer when these problems occur.

[0029] Overloading and overheating: The transformer may overheat if it is subjected to a load that exceeds its rated current for a long time. Overheating can accelerate the wear and tear of the winding and insulation materials, thereby accelerating the aging or damage of the transformer. Poor contact of the conductor, failure of the cooling system, or poor heat dissipation of the transformer may also cause overheating.

[0030] Oil quality degradation: For oil-immersed transformers, the quality and performance of the oil are crucial for the normal operation. Improper operation or maintenance may cause the oil to be contaminated, with excessive gas, moisture, and impurities in the oil, thereby causing degradation. Degraded oil can reduce insulation performance, increase loss, and thus affect the service life of the transformer.

[0031] In an embodiment of the present application, the data acquisition module is configured to acquire the operating state of the power equipment, wherein the operating state comprises the operating cycle and the operating data of the power equipment.

[0032] The operating cycle of the transformer generally refers to the time period experienced by the transformer from the start of operation to the current time, and the load change experienced by the transformer in the time period.

[0033] The operating data of the transformer is an important basis for evaluating the state of the transformer. In the present application, in order to verify whether the transformer has performance degradation and failure problems at this time, the operating data will select the real-time operating parameters of the power equipment and the temperature data collected on the power equipment, wherein the temperature data includes but is not limited to winding hot spot temperature rise and oil top layer temperature rise, which can represent the operating temperature of the power equipment, to assist in verifying the change of the current operating temperature.

[0034] The winding hot spot temperature rise reflects the temperature change of the transformer winding during operation, and is an important indicator for evaluating the thermal performance of the transformer.

[0035] The oil top layer temperature rise reflects the temperature change of the transformer oil during operation, and is closely related to the heat dissipation performance of the winding.

[0036] After obtaining the operating temperature, it is also necessary to verify whether the equipment has generated overload, such as the number of overvoltage shocks, to evaluate whether there are obvious problems at this time. The number of overvoltage shocks can evaluate the voltage shock resistance of the power equipment at this time.

[0037] The operating data obtained at the same time can also include the water content of the insulating oil and the dielectric loss tangent of the insulating oil, so as to reflect the current operating condition of the power equipment to be identified.

[0038] The water content of the insulating oil: Water will reduce the resistance value of the insulating oil and increase the dielectric loss, which poses a threat to the safe operation of the transformer.

[0039] The dielectric loss tangent of the insulating oil: reflects the loss characteristics of the insulating oil under alternating voltage, and is an important indicator for evaluating the performance of the insulating oil.

[0040] The operating data at this time are represented as real-time load change information, operating temperature, number of overvoltage shocks, water content of insulating oil, and dielectric loss tangent of insulating oil; in order to verify the operating condition of the current power equipment under the conditions of overload and overheating, oil quality deterioration, so as to timely monitor the power equipment and give an early warning when an abnormal condition is monitored.

[0041] Real-time load change information is usually obtained through load testing tools or system monitoring tools. These tools can simulate or monitor the load of the system and record the load change information in real time. For example, monitoring the current and voltage values of the system, recording the current and voltage value change ratio of the current equipment, as the load change information recorded at this time, so as to understand the real-time load of the system; The change ratio of current and voltage value is the ratio of the change ratio of current and voltage value to the average value of current and voltage value in the historical data.

[0042] The running temperature is usually measured by a thermometer or an infrared thermometer. The infrared thermometer can directly measure the instantaneous temperature of the motor winding by irradiating the running motor.

[0043] The number of overvoltage shocks is usually measured by special overvoltage monitoring equipment or system. These devices can monitor the overvoltage of the power system in real time and record the number and amplitude of overvoltage shocks. During the closing and opening process of the disconnecting switch, multiple breakdowns may occur. Counting the number of breakdown pulses during a single closing or opening process is equivalent to counting the number of breakdowns. This helps to evaluate the resistance of the power system to overvoltage shocks.

[0044] The measurement of water content in insulating oil usually adopts the following methods: Karl Fischer method: by reacting the insulating oil sample with Karl Fischer reagent, water will react with iodine in the reagent to generate hydrogen iodide, and then the water content is determined by titration. This method has high sensitivity and accuracy for trace water.

[0045] Electrochemical method: use electrodes to react with water in insulating oil to detect water content. This method usually uses special electrochemical sensors or electrodes to measure the change of current or voltage to determine the water content.

[0046] Infrared spectroscopy: determine the water content by detecting the absorption characteristics of water molecules in insulating oil. Water in insulating oil will absorb specific wavelengths of infrared spectrum, and the water content is calculated by measuring the intensity or proportion of absorption peaks.

[0047] The measurement of dielectric loss tangent of insulating oil usually uses special dielectric loss tester or Westinghouse bridge and other equipment. These devices can measure the dielectric constant and dielectric loss factor of insulating oil, and calculate the dielectric loss tangent. The measurement methods include balance measurement method and angle difference measurement method. Among them, the balance measurement method is to adjust the bridge parameters to make the bridge reach the balance state, so as to measure the dielectric loss tangent; The angle difference measurement method is to calculate the dielectric loss tangent by directly measuring the angle difference between voltage and current.

[0048] At this time, after analyzing the five values from the running data, the running condition of the current power equipment can be effectively evaluated.

[0049] In one embodiment of the present application, the trend identification module is configured to obtain fault characteristics of the power equipment under corresponding operating states based on the operating states of the power equipment, and identify a fault trend of the power equipment, and calculate a fault coefficient of the power equipment according to the fault trend of the power equipment.

[0050] In this module, the purpose of identifying the fault trend of the power equipment is to identify the operating data in abnormal states, and verify how many abnormal conditions exist in the current working period, or the frequency of occurrence of abnormal conditions, to quantify the fault trend of the power equipment.

[0051] The fault characteristics of the power equipment are divided according to real-time load change information, operating temperature, overvoltage shock times, water content in insulating oil, and dielectric loss tangent of insulating oil, for example, the occurrence of insulation loss is found according to the change of the dielectric loss tangent of the insulating oil, the load capacity and overload are found according to the real-time load change information, whether there is a high-temperature operation causing failure is found according to the operating temperature, the insulation and power supply resistance are found according to the overvoltage shock times, and the change of the insulating oil used in the transformer is found according to the water content in the insulating oil; after these operating data are extracted as fault characteristics, the specific conditions of the operating data can be identified, and the fault trend of the power equipment at this time is used to divide these identified conditions according to corresponding specific descriptions, to identify whether the power equipment has a corresponding fault trend.

[0052] As shown in Figure 3 The processing mode for identifying the fault trend of the power equipment includes the following contents: based on the current working period, a first index corresponding to the operating data is extracted, the first index is used to extract data exceeding the average value of historical data from the operating data; a second index in a time period adjacent to the first index is obtained; the second index is used to identify data in the adjacent time period when the first index is identified, to identify abnormal conditions of the current operating data in the adjacent time period.

[0053] Based on the operating data corresponding to the first index and the second index, at least one fault characteristic corresponding to the operating data is obtained; the fault characteristic represents the performance decline or failure of the system or the equipment, and the fault characteristic is obtained according to the corresponding data in the real-time load change information, the operating temperature, the overvoltage shock times, the water content in the insulating oil, and the dielectric loss tangent of the insulating oil, to identify the specific conditions of these data when the failure occurs.

[0054] The first probability of the fault feature in the operation data corresponding to the first index and the second probability of the fault feature in the operation data corresponding to the second index are calculated; the first probability represents the probability of the fault feature in the operation data corresponding to the first index, and the second probability represents the probability of the fault feature in the operation data corresponding to the second index, that is, the possibility or frequency of the fault feature in the adjacent time period of the first index.

[0055] The first probability and the second probability are compared to obtain the fault trend of the power equipment; at this time, the fault trend is identified by comparing the significance of the fault feature in the two time periods, for example, in the case where the first probability corresponding to the first index takes the maximum value, whether the fault feature in the adjacent time period of the first index is also the problem of high frequency or low frequency of occurrence, and the fault trend represented by the fault feature, the change trend of the fault feature is rising, falling or remaining stable in the time period corresponding to the first index and the adjacent time period of the first index, so that the current fault feature can be determined, and the fault coefficient related to the fault trend is calculated; at this time, the first probability and the second probability record the values changing with the time period, and are output as the change trend of the power equipment.

[0056] The fault coefficient of the power equipment is obtained based on the currently identified change trend, so the implementation manner of the fault coefficient of the power equipment can be that the change amount of the first probability and the second probability is calculated, and the fault coefficient of the power equipment is calculated based on the change amount of the first probability and the second probability and by using the quantitative analysis manner of absolute change and relative change.

[0057] For example, the fault coefficient of the power equipment is represented as ; wherein, represents the fault coefficient, represents the first probability, represents the second probability, represents the change amount of the first probability, represents the change amount of the second probability, wherein the change amount ratio of the first probability and the second probability represents comparing the change amount of the first probability and the second probability to identify which probability is more obvious, and the fault coefficient set at this time mainly identifies the difference between the fault condition identified in the case where the operation data deviates obviously from the normal operation value and the fault condition identified in the adjacent time period, to describe the fault condition of the power equipment in the corresponding scene; for the change amount of the first probability and the change amount of the second probability at this time, the change amount of the first probability and the change amount of the second probability are obtained by comparing the first probability and the second probability with the average value of the first probability and the second probability in the historical data.

[0058] In an embodiment of the present application, the fault analysis module is configured to generate a fault analysis tree corresponding to the power equipment based on the fault coefficient, and sequentially determine the fault relationship coefficient, the data fault proportion and the fault period of the power equipment.

[0059] In the present module, the fault analysis tree takes the fault feature as an intermediate node and takes the power equipment as a starting point. At this time, the fault feature represents the main aspect or stage of equipment failure, such as "insulation performance failure", "mechanical strength decline", "thermal performance degradation", etc. The value of the leaf node of the intermediate node is represented by using the fault coefficient, and the edge of the leaf node is set with a weight according to the fault feature, thereby constructing the fault analysis tree corresponding to the current power equipment.

[0060] The fault analysis tree can be represented as: power equipment (root node): insulation performance failure (intermediate node): insulation material aging (leaf node, weight: 0.6), high humidity of operating environment (leaf node, weight: 0.3), insulation layer damage (leaf node, weight: 0.1); mechanical strength decline (intermediate node): vibration and impact (leaf node, weight: 0.5), component wear (leaf node, weight: 0.3), loose fastener (leaf node, weight: 0.2); thermal performance degradation (intermediate node): poor heat dissipation (leaf node, weight: 0.7), overload operation (leaf node, weight: 0.2), cooling system failure (leaf node, weight: 0.1); at this time, the current fault analysis tree can be analyzed according to the weight of the fault feature and the value of the leaf node to obtain the values of the fault relationship coefficient, the data fault proportion and the fault period.

[0061] The fault relationship coefficient is obtained by analyzing the leaf nodes of the fault analysis tree and calculating the Pearson correlation coefficient; the data fault proportion is the ratio of the number of leaf nodes corresponding to the fault relationship coefficient to the total number of leaf nodes under the same value; and the fault period represents the average interval time of each fault identified when the fault analysis tree is generated by the fault coefficient.

[0062] The analysis process of the fault relationship coefficient is as follows: based on all the leaf nodes in the fault analysis tree, the fault coefficient and the weight coefficient corresponding to each leaf node are obtained; the standardized covariance between the fault coefficient set and the weight coefficient set is calculated by statistical analysis method, so as to quantify the linear correlation degree between the two, and the result is the fault relationship coefficient.

[0063] Specifically The calculation method of the fault relationship coefficient can be shown as follows: ; wherein, represents the fault relationship coefficient, represents the fault coefficient corresponding to the i-th leaf node in the fault analysis tree, a standard value of the failure coefficient, a number of leaf nodes in the failure analysis tree, i is in a range of 1 to n, a weight coefficient corresponding to the i th leaf node in the failure analysis tree, a standard value of the weight coefficient; at this time, the standard values of the failure coefficient and the weight coefficient are represented by the average values of the corresponding failure coefficient and the weight coefficient in the historical data.

[0064] In an embodiment of the present application, the failure impact module is used to evaluate the failure loss of the power equipment according to the failure period, and calculate the failure impact coefficient of the power equipment in the corresponding failure period.

[0065] In this module, the failure is evaluated according to the average time of the failure, for example, the failure characteristics represented by each parameter in the operation data are calculated respectively, and the relative comprehensive value of the failure characteristics is taken as the failure impact coefficient output at this time; the purpose of calculating the failure impact coefficient is to identify the change of the current power equipment in different periods, and at this time, the failure impact coefficient can be quantified by calculating the failure uniform distribution degree and the damage difference index.

[0066] As shown in Figure 4 , the implementation of the failure impact coefficient is to divide the failure characteristics into first failure characteristics and second failure characteristics; at this time, each failure characteristic contains the content of the corresponding five numerical values in the operation data, and the division of the first failure characteristics and the second failure characteristics is to separate the original failure characteristics according to the median data, at this time, the number of the first failure characteristics and the second failure characteristics is the same, and the serial numbers are also the same.

[0067] According to the real-time load change information, the operating temperature and the overvoltage impact times corresponding to the first failure characteristics, the failure uniform distribution degree related to the failure distribution is calculated.

[0068] According to the water content of the insulating oil and the dielectric loss tangent value of the insulating oil corresponding to the second failure characteristics, the damage difference index related to the damage of the equipment is calculated.

[0069] Based on the failure uniform distribution degree and the damage difference index, the failure impact coefficient is obtained.

[0070] Assuming that all the failure characteristics are one-to-one at this time, the number of the first failure characteristics and the second failure characteristics is consistent, and the arrangement order of these failure characteristics is consistent, at this time, there are J failure characteristics, and the serial number of each failure characteristic is j, at this time, the number of the first failure characteristics and the second failure characteristics is J, and the serial numbers are also the same; at this time, the failure uniform distribution degree can be realized in the following way.

[0071] The calculation process of the fault uniform distribution degree comprises the following steps: A. Normalizing the real-time load change information, operating temperature and overvoltage impact number corresponding to the first fault feature respectively.

[0072] B. Based on the normalization processing result, calculating the probability distribution of each parameter in the first fault feature.

[0073] C. For the probability distribution of each parameter, calculating the information entropy value thereof as the uniform distribution degree of the parameter, and the calculation of the information entropy value reflects the sum of the product of the probability of all possible states of the parameter and the logarithm value of the probability.

[0074] D. Combining the uniform distribution degrees of the three parameters obtained in step C: firstly, calculating each uniform distribution degree value as the index of a natural exponential function to obtain the corresponding exponential result; then adding the three exponential results as the numerator; at the same time, calculating the sum of the three uniform distribution degrees as the index of a natural exponential function to obtain the denominator; finally, dividing the numerator by the denominator, and the obtained result is the fault uniform distribution degree.

[0075] Among them, the first fault feature comprises three parameters of real-time load change information, operating temperature and overvoltage impact number.

[0076] The uniform distribution degree (information entropy value) of the parameter reflects the uniform degree of the probability distribution of the parameter on all possible states thereof, and the greater the value, the more uniform the distribution.

[0077] The information entropy is expressed as: ; wherein, represents the value of the information entropy, represents the probability value of the jth fault feature, and the information entropy value in the current state is calculated based on the probability value of each fault feature at this time.

[0078] The combination operation refers to the specific index processing and proportional calculation of the uniform distribution degree values of the three parameters, and the calculation essentially reflects the weighted geometric mean relationship based on the exponential transformation.

[0079] For example, the fault uniform distribution degree is expressed as: the real-time load change information, operating temperature and overvoltage impact number corresponding to the first fault feature are normalized to calculate the probability of the first fault feature; the probability of the first fault feature is calculated according to the serial number of the fault feature to obtain the uniform distribution degree of the first fault feature; and the uniform distribution degree of the first fault feature is combined according to the real-time load change information, operating temperature and overvoltage impact number to calculate the fault uniform distribution degree.

[0080] The uniform distribution degree of the first fault feature is expressed as: ; wherein, a uniform distribution degree of the first fault feature, a probability of the first fault feature corresponding to the jth serial number, a serial number of the fault feature, and j is in a range of 1 to J; the probability of the first fault feature is represented as a ratio of a number of values corresponding to real-time load change information, operating temperature and overvoltage shock times in the current first fault feature to a total number of historical data.

[0081] The fault uniform distribution degree is represented as, and the uniform distribution degree of the first fault feature is divided into uniform distribution degrees corresponding to real-time load change information, operating temperature and overvoltage shock times; the uniform distribution degrees corresponding to real-time load change information, operating temperature and overvoltage shock times are obtained, and the fault uniform distribution degree is calculated.

[0082] ; wherein, a fault uniform distribution degree, a uniform distribution degree corresponding to real-time load change information, a uniform distribution degree corresponding to operating temperature, a uniform distribution degree corresponding to overvoltage shock times, an exponential constant.

[0083] The calculation process of the damage difference index comprises the following steps: E1. The insulation oil water content index value and the insulation oil dielectric loss tangent index value contained in the second fault feature are normalized respectively.

[0084] E2. For each detection sample point, the following is performed: the deviation of the insulation oil water content index value of the sample point from its preset standard value is calculated, denoted as deviation value one; the deviation of the insulation oil dielectric loss tangent index value of the sample point from its preset standard value is calculated, denoted as deviation value two; the deviation value one and the deviation value two are multiplied by the respective preset weight coefficients, and the weighted deviations are added to obtain the weighted comprehensive deviation of the sample point.

[0085] E3. For the weighted comprehensive deviations of all sample points, the following is performed: the weighted comprehensive deviation value of each sample point is squared; the sum of all squared values is calculated.

[0086] E4. The square root of the result obtained in step E3 is taken, and the ratio of the weighted comprehensive deviation of the sample point is the damage difference index.

[0087] The second fault feature contains the insulation oil water content index value and the insulation oil dielectric loss tangent index value.

[0088] The preset standard value is a reference baseline value of the insulation oil water content and the insulation oil dielectric loss tangent value preset in advance.

[0089] The preset weight coefficient refers to a proportion factor that is preset to reflect the importance of the water content of the insulating oil and the dielectric loss tangent of the insulating oil.

[0090] The sample points refer to data points at different detection times or different detection positions distinguished by the serial number j, and the total number of j is the total number of sample points.

[0091] The damage difference index is essentially a statistical dispersion measure based on a weighted comprehensive deviation calculation and adjusted by the sample degree of freedom, and reflects the overall deviation of the index value from the standard value.

[0092] For example, the damage difference index is calculated by normalizing the water content of the insulating oil and the dielectric loss tangent of the insulating oil corresponding to the second fault feature.

[0093] ; wherein, represents the damage difference index, represents the index value of the water content of the insulating oil in the second fault feature corresponding to the jth serial number, represents the standard value of the index value of the water content of the insulating oil, represents the index value of the dielectric loss tangent of the insulating oil in the second fault feature corresponding to the jth serial number, represents the standard value of the index value of the dielectric loss tangent of the insulating oil, represents the weight coefficient of the water content of the insulating oil, represents the weight coefficient of the dielectric loss tangent of the insulating oil; at this time, the index value used is obtained from the specific value corresponding to the water content of the insulating oil and the dielectric loss tangent of the insulating oil extracted from the second fault feature, and the extracted index value is normalized to eliminate the dimension, and the standard value of the index value used is obtained by extracting the average value from the historical data.

[0094] Preferably, the weight coefficients corresponding to the water content of the insulating oil and the dielectric loss tangent of the insulating oil are set to 0.5 and 0.5, respectively, in the present application, for measuring the difference between the second fault feature and the historical data.

[0095] The calculation process of the fault influence coefficient is as follows: the sum of the ratio of the fault uniform distribution degree to its preset standard value and the ratio of the damage difference index to its preset standard value is obtained, and the sum is multiplied by the preset weight coefficient of the fault uniform distribution degree and the preset weight coefficient of the damage difference index in turn, and the obtained calculation result is taken as the fault influence coefficient.

[0096] The two weight coefficients set here are different from those set in the foregoing, and the weight coefficients of the fault uniform distribution degree and the damage difference index are set by the ratio of the data quantity corresponding to the value of the current fault uniform distribution degree and the damage difference index to the total quantity of historical data, and the standard value of the fault uniform distribution degree and the damage difference index is set by the standard deviation of the fault uniform distribution degree and the damage difference index in the historical data.

[0097] In an embodiment of the application, the influence range determination module is configured to analyze the influence range coefficient of the running data with faults according to the fault relationship coefficient and the data fault proportion.

[0098] The influence range coefficient is used to evaluate the influence degree of the fault on the equipment at this time, and the data fault proportion is used to represent the ratio of the number of leaf nodes corresponding to the fault relationship coefficient to the total number of leaf nodes under the same value of the fault relationship coefficient. In the scenario of calculating the influence range coefficient at this time, the data fault proportion can also represent the ratio of the number of corresponding fault features to the total number of fault features under the same value of the fault relationship coefficient.

[0099] At this time, the influence range coefficient corresponding to the current running data can be identified according to the data corresponding to the fault relationship coefficient and the data fault proportion, for example, by adding the working cycle of the current power equipment and combining the fault relationship coefficient and the data fault proportion to quantify the influence range coefficient.

[0100] The finally calculated influence range coefficient represents the relative distribution of the fault features corresponding to the fault relationship coefficient and the data fault proportion, and represents the correlation between the fault relationship coefficient and the actual distribution of faults.

[0101] The influence range coefficient can be represented by a normal distribution form of the current fault relationship coefficient and the data fault proportion to obtain the range influenced at this time.

[0102] For example, the average value of the fault relationship coefficient and the standard deviation of the data fault proportion are calculated in sequence according to the fault relationship coefficient and the data fault proportion, and the influence range coefficient is calculated by a normal distribution analysis method.

[0103] The influence range coefficient is represented as ; wherein represents the influence range coefficient, represents the fault relationship coefficient, represents the average value of the fault relationship coefficient, represents the standard deviation of the data fault proportion, represents the circular constant, represents an exponential constant. At this time, the influence range coefficient can represent a specific value to evaluate the distribution of the current fault relationship coefficient under the corresponding data fault proportion. At this time, the fault relationship coefficient and the data fault proportion do not have dimensions in calculation, and in order to obtain more accurate values, the fault relationship coefficient and the data fault proportion are standardized to make the value range of the two closer. When the influence range coefficient needs to output the corresponding range, the average value of the current fault relationship coefficient and the standard deviation of the data fault proportion can be used to quantify the range of the fault relationship coefficient and the data fault proportion that need to be identified at this time. The range value can be, ; wherein represents an adjustment coefficient, and 1.96 can be selected as the value used at this time. At this time, the value represented by the data fault proportion is small, and when identifying the fault relationship coefficient and the data fault proportion that need to be paid attention to through the range, the corresponding value can be accurately identified.

[0104] In an embodiment of the present application, the early warning evaluation module is configured to calculate the fault promotion coefficient of the power equipment under different cycle lengths and the value of the running data according to the fault influence coefficient and the influence range coefficient, and obtain the analysis and early warning result of the current power equipment.

[0105] At this time, in order to quantify the fault promotion coefficient, the fault cycle of the power equipment is converted into a proportional value relative to the average value of the fault cycle in the historical data, which is denoted as a cycle proportional value. The number of running data under the corresponding fault cycle relative to the total data is denoted as a data proportional value. Therefore, the fault promotion coefficient at this time is represented as: the cycle proportional value and the data proportional value of the fault cycle under different cycle lengths are obtained, and the fault promotion coefficient is calculated based on the data proportional value, the cycle proportional value, the fault influence coefficient and the influence range coefficient.

[0106] ; wherein, represents the fault promotion coefficient, represents the fault influence coefficient, represents the influence range coefficient, represents the cycle proportional value, represents the data proportional value, , , , represents a pending coefficient; the pending coefficient is set to -0.05, -0.1, 1 and 0.2 in sequence.

[0107] After obtaining the fault promotion coefficient, the state of the current power equipment is divided into multiple levels according to the value of the fault promotion coefficient at this time, for example, into five levels. Low risk: > 0.9; medium-low risk: 0.8 <0.9; medium risk: 0.6 <0.8; high risk: 0.5 <0.6; extremely high risk: <0.5; the identified grades and failure promotion coefficients are output, so that the current power equipment analysis warning result can be obtained.

[0108] By matching the current failure promotion coefficient with the specific state of the current power equipment, the specific situation of the current power equipment in use can be accurately identified, the unstable operation of the power equipment can be prevented in time, and the monitoring result of the operation state of the power equipment can be improved in combination with the current situation and timely analysis.

[0109] Although the embodiments of the present application have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application, which are still covered by the protection scope of the present application.

Claims

1. A power equipment monitoring analysis and early warning system, characterized in that, Comprise: Data acquisition module, obtain the running state of power equipment, the running state includes the working cycle and running data of power equipment; Trend identification module, based on the running state of power equipment, obtain the fault characteristics of power equipment under the corresponding running state, and identify the fault trend of power equipment, and calculate the fault coefficient of power equipment according to the fault trend of power equipment; Fault analysis module, based on the fault coefficient, generate the fault analysis tree corresponding to the power equipment, and determine the fault relationship coefficient, data fault proportion and fault period of the power equipment in turn; Fault influence module, according to the fault period, the fault loss of power equipment is evaluated, and the fault influence coefficient of power equipment in the corresponding fault period is calculated; Influence range determination module, according to the fault relationship coefficient and data fault proportion, the influence range coefficient of the running data with fault is analyzed; Early warning evaluation module, according to the fault influence coefficient and influence range coefficient, the fault promotion coefficient of power equipment under different cycle length and the value of running data is calculated, and the current power equipment analysis early warning result is obtained.

2. The power equipment monitoring, analyzing and early warning system according to claim 1, characterized in that, The processing mode for identifying the fault trend of power equipment includes the following contents: Based on the current working cycle, the first index corresponding to the running data is extracted, and the second index in the adjacent time period of the first index is obtained; Based on the running data corresponding to the first index and the second index, at least one fault feature corresponding to the running data is obtained; The first probability of the fault feature in the running data corresponding to the first index and the second probability of the fault feature in the running data corresponding to the second index are calculated; The first probability and the second probability are compared, and the fault trend of the power equipment is obtained.

3. The power equipment monitoring, analyzing and early warning system according to claim 2, characterized in that, The implementation mode of the fault coefficient of the power equipment is that the change amount of the first probability and the second probability is calculated, and the fault coefficient of the power equipment is calculated based on the change amount of the first probability and the second probability and by using the quantitative analysis mode of absolute change and relative change.

4. The power equipment monitoring, analyzing and early warning system according to claim 1, characterized in that, The analysis process of the fault relationship coefficient is that, based on all leaf nodes in the fault analysis tree, the fault coefficient and weight coefficient corresponding to each leaf node are obtained; the standardized covariance between the fault coefficient set and the weight coefficient set is calculated by statistical analysis method, so as to quantify the linear correlation degree between the two, and the result is the fault relationship coefficient.

5. The power equipment monitoring, analyzing and early warning system according to claim 1, characterized in that, The implementation mode of the fault influence coefficient is that the fault feature is divided into first fault feature and second fault feature; According to the real-time load change information, operating temperature and overvoltage impact number corresponding to the first fault feature, the fault uniform distribution degree related to fault distribution is calculated; According to the water content of insulating oil and dielectric loss tangent value corresponding to the second fault feature, the damage difference index related to equipment damage is calculated; Based on the fault uniform distribution degree and the damage difference index, the fault influence coefficient is obtained.

6. The power equipment monitoring and analysis early warning system according to claim 5, wherein The calculation process of the fault uniform distribution degree comprises the following steps: A. The real-time load change information, operating temperature and overvoltage impact number corresponding to the first fault feature are normalized respectively; B. based on the normalization processing result, calculate the probability distribution of each parameter in the first fault feature; C. For each parameter, the information entropy value of its probability distribution is calculated as the uniform distribution degree of the parameter, and the information entropy ; wherein, represents the value of the information entropy, represents the probability value of the jth fault feature, and the information entropy value in the current state is calculated based on the probability value of each fault feature at this time. D. combine the three uniform distribution degrees obtained in step C: first, take each uniform distribution degree value as the index of a natural exponential function to obtain the corresponding exponential result; then add the three exponential results as the numerator; at the same time, take the sum of the three uniform distribution degrees as the index of a natural exponential function to obtain the denominator; finally, divide the numerator by the denominator, and the result obtained is the fault uniform distribution degree.

7. The power equipment monitoring analysis and early warning system according to claim 6, characterized in that, the calculation process of the damage difference index comprises the following steps: E1. normalizing the insulating oil water content index value and the insulating oil dielectric loss tangent value included in the second fault feature; E2. for each detection sample point: calculate the deviation of the insulating oil water content index value of the sample point from its preset standard value, denoted as deviation value one; calculate the deviation of the insulating oil dielectric loss tangent value of the sample point from its preset standard value, denoted as deviation value two; multiply deviation value one and deviation value two by their respective preset weight coefficients, and add the weighted deviations to obtain the weighted comprehensive deviation of the sample point; E3. for the weighted comprehensive deviations of all sample points: square the weighted comprehensive deviation value of each sample point; calculate the sum of all squared values; E4. take the square root of the result obtained in step E3, and the ratio of the weighted comprehensive deviation of the sample point is the damage difference index.

8. The power equipment monitoring, analyzing and early warning system according to claim 7, characterized in that, The calculation process of the fault influence coefficient is: sum the ratio of the fault uniform distribution degree to its preset standard value and the ratio of the damage difference index to its preset standard value to obtain a sum value, and multiply the sum value by the preset weight coefficient of the fault uniform distribution degree and the preset weight coefficient of the damage difference index in turn, and the calculation result is taken as the fault influence coefficient.

9. The power equipment monitoring, analyzing and early warning system according to claim 1, characterized in that, The influence range coefficient is represented as: according to the fault relationship coefficient and the data fault proportion, the average value of the fault relationship coefficient and the standard deviation of the data fault proportion are calculated in turn, and the influence range coefficient is calculated by normal distribution analysis.

10. The power equipment monitoring, analyzing and early warning system according to claim 1, characterized in that, The fault promotion coefficient is represented as: the cycle proportion value of the fault cycle and the data proportion value of the operation data under different cycle lengths are obtained, and the fault promotion coefficient is calculated by multi-parameter fusion weighting based on the data proportion value, the cycle proportion value, the fault influence coefficient and the influence range coefficient.

Images