A power supply facility diagnosis and early warning method and system for a smart grid

Abstract

The application belongs to the technical field of power supply facility monitoring, and discloses a power supply facility diagnosis and early warning method and system for a smart power grid; the method comprises the following steps: a platform end receives long-term operation data collected by a facility end; the platform end analyzes the long-term operation data, obtains analysis data, and determines whether to generate an abnormal instruction; if the abnormal instruction is generated, the platform end receives real-time operation data collected by the facility end; the platform end analyzes the analysis data and the real-time operation data, and determines whether to generate a decline instruction; if the decline instruction is generated, the platform end diagnoses the residual life of the power grid facility according to the analysis data and the real-time operation data; the platform end generates an early warning instruction, and sends the early warning instruction, the analysis data, the real-time operation data and the residual life to the corresponding facility end; the application realizes in-depth learning and dynamic decision of the life management of the power grid facility, and improves the management efficiency and operation safety of the power grid facility.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power supply facility monitoring, and more particularly, to a power supply facility diagnosis and early warning method and system for a smart grid. BACKGROUND

[0002] The life monitoring and diagnosis of power grid facilities are crucial to ensuring the safe and reliable operation of power grid facilities, reducing maintenance costs, achieving personalized management, and making early arrangements for plans.

[0003] With the development of the Internet, the smart grid has become one of the main directions of the development of the power system, and therefore the intelligent life monitoring and diagnosis method of power grid facilities has emerged. For example, the Chinese patent application with the publication number CN108181580A discloses a solid insulation switch device intelligent diagnosis device and life evaluation method. The device includes a solid insulation switch device, a solid insulation switch device monitoring terminal, and a life evaluation system. Based on online real-time monitoring of the operating state of the switch device, the physical data (including electrical parameters and non-electrical parameters) monitored by each sensor are compared and analyzed, and a reference model established based on historical data is relied on to timely detect faults and remove them. At the same time, the electrical parameters at the time of the fault are recorded and uploaded to evaluate the reliability and life of the device. The device can effectively monitor the operating condition of the high-voltage switch device and realize intelligent fault diagnosis of the intelligent switch device, thereby improving the reliability of the power grid operation. Through the evaluation of the life of the device, the system maintenance is reasonably arranged, and the cost of human resources is effectively reduced.

[0004] However, the above-mentioned technologies directly compare the preset threshold value with the collected data when making an abnormal decision. The performance of the power grid facility changes over time, and the preset threshold value is difficult to dynamically optimize and cannot adapt to the individual differences of different power grid facilities. At the same time, the life evaluation method only mentions the mechanical life evaluation method and the electrical life evaluation method, and does not detail the specific steps in the evaluation method, i.e., the specific life monitoring and diagnosis logic cannot be known.

[0005] In view of the above problems, the present application provides a power supply facility diagnosis and early warning method and system for a smart grid. SUMMARY

[0006] In order to overcome the above-mentioned defects of the prior art and achieve the above-mentioned purposes, the present application provides the following technical solution: a power supply facility diagnosis and early warning method for a smart grid, comprising:

[0007] The platform end receives long-term operation data collected by the facility end;

[0008] The platform end analyzes long-term running data, obtains analysis data, and determines whether to generate an abnormal instruction;

[0009] If the abnormal instruction is generated, the platform end receives real-time running data collected by the facility end;

[0010] The platform end analyzes the analysis data and the real-time running data, and determines whether to generate a decline instruction;

[0011] If the decline instruction is generated, the platform end diagnoses the residual life of the power grid facility according to the analysis data and the real-time running data;

[0012] The platform end generates a warning instruction, and sends the warning instruction, the analysis data, the real-time running data, and the residual life to the corresponding facility end.

[0013] Further, the long-term running data includes a current value and an oil product index;

[0014] A preset collection interval is set, and the current value is collected according to the collection interval;

[0015] The oil product index is calculated according to the collection interval, and the calculation method of the oil product index is as follows: oil product influence data including total acid value, mineral content, and lead content is collected according to the collection interval; a lead content threshold value is preset, and the lead content is compared with the lead content threshold value; if the lead content is less than or equal to the lead content threshold value, the lead content in the oil product influence data is adjusted to 0; if the lead content is greater than the lead content threshold value, the lead content is subtracted from the lead content threshold value to obtain a lead content difference value, and the lead content in the oil product influence data is adjusted to the lead content difference value; and the oil product index is calculated according to the oil product influence data.

[0016] Further, the method for obtaining analysis data includes:

[0017] A current change curve is constructed according to the n current values obtained, and an oil product index change curve is constructed according to the n oil product indexes obtained; the current change curve is a curve of continuous change of the current during the operation of the power grid facility, and the oil product index change curve is a curve of continuous change of the oil product index during the operation of the power grid facility; an instant corresponding to the n th current value in the current change curve is marked as a prediction instant; the first current values in the current change curve are input into a trained current prediction model, a current value corresponding to the prediction instant is predicted and marked as a predicted current value; and the first oil product indexes in the oil product index change curve are input into a trained index prediction model, an oil product index corresponding to the prediction instant is predicted and marked as a predicted oil product index;

[0018] The n th current value in the current change curve is divided by the predicted current value to obtain a current error rate, and the n th oil product index in the oil product index change curve is divided by the predicted oil product index to obtain an index error rate;

[0019] The n-th current value in the current change curve is subtracted from the first current value to obtain a current change amplitude, and the n-th oil product index in the oil product index change curve is subtracted from the first oil product index to obtain an index change amplitude.

[0020] The current change amplitude, the index change amplitude, the current error rate and the index error rate are taken as analysis data.

[0021] Further, the method of judging whether to generate an abnormal instruction comprises:

[0022] The analysis data is input into a trained probability analysis model to predict an abnormal probability of the power grid facility;

[0023] A probability threshold is preset ; the abnormal probability is compared with the probability threshold ;

[0024] If , no abnormal instruction is generated.

[0025] If , an abnormal instruction is generated.

[0026] Further, the real-time operation data comprises a polarization rate, a hot spot number and a flaw number.

[0027] The polarization rate is an imbalance degree of vibration of the power grid facility in different directions.

[0028] The hot spot number is an abnormal point number of temperature values on the surface of the power grid facility.

[0029] The flaw number is a number of physical structure defects on the surface of the power grid facility.

[0030] Further, the method of obtaining the polarization rate comprises: obtaining three-direction vibration values, the three-direction vibration values comprising an X-direction vibration value, a Y-direction vibration value and a Z-direction vibration value; marking the vibration value with the largest value as a maximum vibration value and the vibration value with the smallest value as a minimum vibration value in the three-direction vibration values; obtaining a vibration deviation value according to the difference between the maximum vibration value and the minimum vibration value; obtaining a vibration average value by adding each vibration value in the three-direction vibration values in turn and taking an average; and obtaining the polarization rate according to the ratio of the vibration deviation value to the vibration average value.

[0031] ​​The method for obtaining the number of hot spots comprises: obtaining a temperature distribution image; presetting a temperature threshold, comparing the temperature value of each pixel point in the temperature distribution image with the temperature threshold, marking the pixel point with a temperature value greater than the temperature threshold as a hot spot, and not marking the pixel point with a temperature value less than or equal to the temperature threshold; and counting the number of hot spots to obtain the number of hot spots.

[0032] The method for obtaining the number of defects comprises:

[0033] A three-dimensional coordinate system is established for the power grid facility, and m coordinate points corresponding to the surface of the power grid facility are obtained; the surface of the power grid facility is scanned by using a laser sensor at multiple coordinate points, that is, the m coordinate points are all scanned; the laser sensor comprises a laser emitter and a laser receiver, the laser emitter emits a laser beam, the laser beam is a short pulse laser beam, and the laser receiver receives the laser beam reflected by the surface of the power grid facility; the reflection time of each coordinate point is obtained, the reflection time being the duration from laser emission to laser reception; the emission angle and the receiving angle of each coordinate point are obtained, the emission angle being the angle between the laser emitter and the horizontal plane when the laser is emitted, and the receiving angle being the angle between the laser receiver and the horizontal plane when the laser is received;

[0034] According to the reflection time and the speed of light of each coordinate point, the laser path of each coordinate point is calculated, the laser path being the path passed by the laser when the laser is emitted to the laser receiver; according to the laser path, the emission angle and the receiving angle of each coordinate point, the straight-line distance of each coordinate point is calculated, the straight-line distance being the vertical distance between each coordinate point and the laser sensor;

[0035] The point cloud data is formed according to the straight-line distances corresponding to the m coordinate points, the point cloud data is reproduced by using a polygon mesh method, the three-dimensional surface structure of the power grid facility is obtained, and is marked as a real-time structure; the real-time structure is compared with the preset power grid facility structure, the number of regions with different structures is counted, and the number of defects is obtained.

[0036] Further, the method for judging whether to generate a decline instruction comprises:

[0037] The analysis data and the real-time operation data are taken as test data, the test data is input into the trained life decline analysis model, and it is judged whether the life of the power grid facility has declined;

[0038] According to the predicted judgment label, a corresponding judgment result is obtained to judge whether the life of the power grid facility has declined under the condition of the test data;

[0039] If the judgment result is that the life has declined, a decline instruction is generated;

[0040] If the judgment result is that the life is constant, no decline instruction is generated.

[0041] Further, the step of diagnosing the residual life of the power grid facility comprises:

[0042] Step 1: encode the residual life into chromosomes, randomly generate N chromosomes to form an initial population;

[0043] Step 2: the fitness is the accuracy, the accuracy is determined by the reciprocal of the life error value, the life error value is the absolute value of the difference between the expected life and the residual life;

[0044] Step 3: keep the top K chromosomes with the highest fitness and directly copy them to the next generation, randomly select U chromosomes for crossover according to the fitness probability, generate new chromosomes, and replace the individuals in the original population with the best ones. Select the remaining chromosomes to make up the population capacity N;

[0045] Step 4: randomly select U chromosomes for crossover, generate new chromosomes, and replace the individuals in the original population with the best ones, randomly exchange the gene values in the chromosomes according to the preset mutation probability;

[0046] Step 5: when the preset maximum number of iterations is reached or a chromosome with a fitness value exceeding the threshold value appears in the population, terminate the loop, output the optimal solution, and the optimal solution is the residual life corresponding to the chromosome with the maximum fitness.

[0047] Further, the platform end analyzes the residual life of the power grid facility and matches the corresponding warning instruction;

[0048] Obtain the initial operating life and importance of the power grid facility; the initial operating life is the expected life of the power grid facility under normal operating conditions; the importance is the importance of the power grid facility in the power grid, and the importance range is , J is an integer greater than 1;

[0049] Divide the residual life by the initial operating life to obtain the life change rate; according to the importance of the power grid facility, divide the corresponding warning instruction into V different levels of warning instruction, , that is, the corresponding warning instruction includes the first warning instruction, …, the Vth warning instruction; then divide 1 by the importance to obtain the matching division degree, and the life change rate range corresponding to each level of warning instruction is , where t is the warning instruction level, , is the matching division degree; then compare the life change rate of the power grid facility with the life change rate range corresponding to each level of warning instruction, and the warning instruction corresponding to the life change rate within the life change rate range and the highest level of warning instruction are matched as the warning instruction of the power grid facility.

[0050] The utility model provides a power supply facility diagnosis early warning system for smart grid, implements the utility model discloses a power supply facility diagnosis early warning method for smart grid, comprising:

[0051] Long -term data receiving module, for platform end receives the long -term operation data of facility end collection;

[0052] Long -term data analysis module, for platform end carries out analysis to long -term operation data, obtains analysis data and judges whether generates exception instruction;

[0053] Real -time data receiving module, if generates exception instruction, then platform end receives the real -time operation data of facility end collection;

[0054] Life decline analysis module, for platform end carries out analysis to analysis data and real -time operation data, judges whether generates decline instruction;

[0055] Life diagnosis module, if generates decline instruction, then platform end diagnoses the residual life of power grid facility according to analysis data and real -time operation data;

[0056] Early warning module, for platform end generates early warning instruction, and sends early warning instruction, analysis data, real -time operation data and residual life to corresponding facility end.

[0057] The utility model discloses a power supply facility diagnosis early warning method and system for smart grid's technical effect and advantage:

[0058] 1. through the long -term operation data of facility end and analysis data of analysis of gathering, can real -time monitoring the running state of power grid facility, and discovers abnormal situation in time;Real -time operation data is gathered again in -depth analysis in combination with analysis data, can in -depth research and judge whether the life of power grid facility declines, thereby early warning and generating corresponding instruction in advance;In addition, through genetic algorithm realizes power grid facility individualization residual life diagnosis, greatly improves detection precision and diagnosis effect, can provide scientific basis for the maintenance and update of power grid facility, improves the reliability and stability of power grid operation;Realize the in -depth learning and dynamic decision of power grid facility life management, improve the management efficiency and operation safety of power grid facility.

[0059] 2. according to the importance of power grid facility, divides early warning instruction into multiple grades, and calculates life change rate to match corresponding grade early warning instruction;Can improve the accuracy and practicality of early warning, provides personalized and target-oriented intelligent decision support for operation and maintenance personnel, helps relevant operation and maintenance personnel accurately master the life change of power grid facility, optimizes resource allocation and management, provides more effective support for the management and operation of power grid facility. BRIEF DESCRIPTION OF DRAWINGS

[0060] Figure 1The figure is a schematic diagram of the power supply facility diagnosis and early warning system for the smart power grid according to Embodiment 1 of the present application;

[0061] Figure 2 The figure is a schematic diagram of the principle of obtaining the number of defects according to Embodiment 1 of the present application;

[0062] Figure 3 The figure is a schematic diagram of the power supply facility diagnosis and early warning system for the smart power grid according to Embodiment 2 of the present application;

[0063] Figure 4 The figure is a flow chart of the power supply facility diagnosis and early warning method for the smart power grid according to Embodiment 3 of the present application. DETAILED DESCRIPTION

[0064] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0065] Embodiment 1

[0066] Please refer to Figure 1 The power supply facility diagnosis and early warning system for the smart power grid according to the present embodiment includes a long-term data receiving module, a long-term data analysis module, a real-time data receiving module, a life decline analysis module, a life diagnosis module and an early warning module. The modules are connected by wired and / or wireless means to realize data transmission between the modules.

[0067] The long-term data receiving module is used for receiving long-term operation data collected by the facility end at the platform end.

[0068] The platform end is a smart network connection cloud control platform, and one power grid facility is one facility end.

[0069] The long-term operation data includes current values and oil indicators.

[0070] The preset collection interval is used to collect the current values, which are obtained by the current sensor built in the power grid facility. The collection interval is preset by those of ordinary skill in the art according to the current change rate when the power grid facility is running. The greater the current change range is, the shorter the collection interval is, and the smaller the current change range is, the longer the collection interval is.

[0071] The oil product indicators are calculated based on the data collection interval. The calculation method is as follows: Oil product impact data, including total acid value, mineral content, and lead content, is collected at the collection interval. A lead content threshold is preset, and the lead content is compared with this threshold. If the lead content is less than or equal to the threshold, the lead content in the oil product impact data is adjusted to 0. If the lead content is greater than the threshold, the lead content is subtracted from the threshold to obtain the lead content difference, which is then used to adjust the lead content in the oil product impact data. The oil product indicators are calculated based on the oil product impact data, and the expression for the oil product indicators is:

[0072] In the formula, For oil quality indicators, The total acid value after standardization. For standardized mineral content, To standardize the lead content, add 1 to the denominator to avoid the risk of dividing by zero and ensure... Within the [0,1] range, it better meets the actual evaluation needs. , , All are preset weighting coefficients; the specific values ​​of the weighting coefficients in the formula can be set according to the actual situation. The weighting coefficients reflect the degree of influence of each type of oil product impact data on the oil product index. Those skilled in the art can preset the corresponding weighting coefficients according to the actual degree of influence of each type of oil product impact data on the oil product index, so as to accurately assess the oil product index during the operation of power grid facilities.

[0073] The total acid value is the hydrogen ion content in the oil used in power grid facilities. This total acid value is obtained from a hydrogen ion selective electrode installed in the oil tank of the power grid facilities. A higher total acid value indicates faster corrosion of metal components within the power grid facilities, lower oil quality indicators, and a shorter lifespan for the power grid facilities, and vice versa. The mineral content is the content of minerals (such as calcium, aluminum, and magnesium) in the oil used in power grid facilities. This mineral content is obtained from a sequential spectrometer installed at the oil outlet in the oil tank of the power grid facilities. Higher mineral content indicates higher conductivity in the oil, reducing insulation performance and bubble point; therefore, lower oil quality indicators and a shorter lifespan for the power grid facilities, and vice versa. The lead content is the lead ion content in the oil used in power grid facilities. This lead content is obtained from a lead ion selective electrode installed in the oil tank of the power grid facilities. The greater the lead content exceeds the lead content threshold, the more severe the arc corrosion problem in the power grid facilities, resulting in lower oil quality indicators and a shorter lifespan for the power grid facilities, and vice versa. The lead content threshold is preset by a person skilled in the art based on industry standards for power grid facilities.

[0074] The long-term data analysis module is used by the platform to analyze long-term running data, obtain analysis data, and determine whether abnormal instructions are generated.

[0075] The method for obtaining analysis data comprises:

[0076] According to the obtained n current values, a current change curve is constructed, and according to the obtained n oil product indicators, an oil product indicator change curve is constructed; the current change curve is a curve of continuous change of current during operation of the power grid facility, and the oil product indicator change curve is a curve of continuous change of oil product indicator during operation of the power grid facility; an instant corresponding to the n current value in the current change curve is marked as a prediction instant; the first n current values in the current change curve are input into a trained current prediction model, a current value corresponding to the prediction instant is predicted, and is marked as a predicted current value; the first n oil product indicators in the oil product indicator change curve are input into a trained indicator prediction model, an oil product indicator corresponding to the prediction instant is predicted, and is marked as a predicted oil product indicator; wherein, since the oil product indicators and the current values are obtained according to the collection interval, the instant corresponding to the n oil product indicator in the oil product indicator change curve is also the prediction instant. The specific training process of the current prediction model comprises:

[0077] The specific training process of the current prediction model comprises:

[0078] a continuous current values are collected in advance according to the collection interval, ; a sliding step L and a sliding window length are preset according to the actual experience of a person skilled in the art; the a current values are converted into corresponding training samples by using the sliding window method, the training samples are used as the input of the current prediction model, the current values after the sliding step L are used as the output, the subsequent current values of each training sample are used as the prediction target, the mean absolute percentage error MAPE is used to evaluate the precision of the prediction result, and when the calculated MAPE is less than the preset MAPE, the training of the current prediction model is completed; wherein, the calculation formula of the MAPE is , wherein, is the prediction target corresponding to the d predicted current value, is the d predicted current value, and R is the number of predicted current values; a current prediction model for predicting the current value at a future instant according to the current value is generated; wherein, the current prediction model is an RNN neural network model; it should be noted that the preset MAPE is set in advance by the staff according to the required precision of the model.

[0079] For example, the collected current value data set A contains 10 current values, , is the h current value, , a plurality of training samples are constructed by using the sliding window, the length of the sliding window is defined as 3, the sliding step L is 1, each training sample contains 3 continuous current values, and the next current value of the 3 continuous current values is used as the prediction target; for example:

[0080] ​ As the training data, The corresponding prediction target is ;

[0081] As the training data, The corresponding prediction target is ; and so on, for training the current prediction model;

[0082] The specific training process of the index prediction model is consistent with that of the current prediction model, and is also an RNN neural network model.

[0083] Divide the nth current value in the current change curve by the predicted current value to obtain the current error rate, and divide the nth oil product index in the oil product index change curve by the predicted oil product index to obtain the index error rate;

[0084] Subtract the first current value from the nth current value in the current change curve to obtain the current change amplitude, and subtract the first oil product index from the nth oil product index in the oil product index change curve to obtain the index change amplitude.

[0085] The current change amplitude, the index change amplitude, the current error rate, and the index error rate are used as analysis data.

[0086] The method for determining whether to generate an abnormal instruction comprises:

[0087] The analysis data is input into the trained probability analysis model to predict the abnormal probability of the power grid facility;

[0088] The specific training process of the probability analysis model comprises:

[0089] A group of analysis data corresponding to the abnormal probability is collected in advance, the analysis data and the corresponding abnormal probability are converted into a corresponding group of feature vectors, and b is an integer greater than 1; wherein the abnormal probability corresponding to the analysis data is analyzed by a person skilled in the art in the historical power grid facility life monitoring process, a group of different analysis data is collected, and the corresponding abnormal probability is analyzed under the condition of each group of analysis data combined with actual experience.

[0090] Each group of feature vectors is used as the input of the probability analysis model, the probability analysis model takes a group of predicted abnormal probabilities corresponding to each group of analysis data as the output, takes the actual abnormal probability corresponding to each group of analysis data as the prediction target, and the actual abnormal probability is the abnormal probability collected in advance corresponding to the analysis data; the sum of the prediction errors of all analysis data is minimized as the training target; the probability analysis model is trained until the sum of the prediction errors converges, and the training is stopped.

[0091] The probability analysis model is specifically a deep neural network model, which includes an input layer, a hidden layer and an output layer; each hidden layer includes a plurality of neurons, each neuron is connected with the next layer of neurons, and the connection includes a weight which determines the importance and influence of data transmission in the neural network; an activation function is applied to each neuron between the hidden layer and the output layer, the activation function is mapped to a nonlinearity, allowing the network to learn more complex patterns and features.

[0092] The preset probability threshold , the probability threshold It is set in advance by a person skilled in the art according to the required detection accuracy.

[0093] The abnormal probability is compared with the probability threshold ;

[0094] If , no abnormal instruction is generated; it indicates that the probability of the power grid facility being abnormal is small, and the remaining diagnostic analysis of the power grid facility is not needed;

[0095] If , an abnormal instruction is generated; it indicates that the probability of the power grid facility being abnormal is large, which may affect the service life of the power grid facility, and the power grid facility needs to be further diagnosed and analyzed.

[0096] The real-time data receiving module receives the real-time running data collected by the facility end if an abnormal instruction is generated.

[0097] The real-time running data includes a polarization rate, a number of hot spots and a number of defects.

[0098] The polarization rate is an imbalance degree of the power grid facility vibrating in different directions; the greater the polarization rate, the more serious the imbalance of the power grid facility vibrating in different directions, and the imbalance of vibration is usually caused by mechanical problems such as aging of internal materials and damage of parts of the power grid facility, and also means that the mechanical stress is concentrated in the power grid facility, which is more prone to failure such as fracture, so the service life of the power grid facility is shorter, and vice versa;

[0099] The method for obtaining the polarization rate is: a three-axis vibration sensor installed on the surface of the power grid facility is used to obtain three-direction vibration values, including X-direction vibration value, Y-direction vibration value and Z-direction vibration value; among the three-direction vibration values, the vibration value with the largest value is marked as the maximum vibration value, and the vibration value with the smallest value is marked as the minimum vibration value; the vibration deviation value is obtained according to the difference between the maximum vibration value and the minimum vibration value, and the expression of the vibration deviation value is: ; in the formula, is the vibration deviation value, is the maximum vibration value, The minimum vibration value; each vibration value in the three-dimensional vibration value is sequentially added and averaged to obtain a vibration average value, and the expression of the vibration average value is: ; wherein, is the vibration average value, is the X-direction vibration value, is the Y-direction vibration value, is the Z-direction vibration value; according to the ratio of the vibration deviation value and the vibration average value, a polarization rate is obtained, and the expression of the polarization rate is: ; wherein, is the polarization rate.

[0100] The number of hot spots is the number of temperature value abnormal points on the surface of the power grid facility; the more the number of hot spots, the larger the temperature value abnormal area on the surface of the power grid facility, and there are problems such as insulation damage or poor contact, so the service life of the power grid facility is shorter, and vice versa.

[0101] The method for obtaining the number of hot spots is: using a thermal imager aligned with the power grid facility to obtain a temperature distribution image; presetting a temperature threshold, comparing the temperature value of each pixel point in the temperature distribution image with the temperature threshold, marking the pixel point with a temperature value greater than the temperature threshold as a hot spot, and not marking the pixel point with a temperature value less than or equal to the temperature threshold; counting the number of hot spots to obtain the number of hot spots; wherein, the temperature threshold is obtained by a person skilled in the art according to the technical manual or parameter table of the power grid facility to obtain the normal working temperature range, and the temperature threshold is preset according to the maximum value of the normal working temperature range.

[0102] The number of defects is the number of physical structure defects on the surface of the power grid facility; the more the number of defects, the more defects on the surface of the power grid facility, the more failure points of the power grid facility damage, which will destroy the structural continuity, reduce the material stiffness and durability, and weaken the insulation protection ability, thereby resulting in shorter service life of the power grid facility, and vice versa.

[0103] Referring to FIG. 1, Figure 2 , the method for obtaining the number of defects includes:

[0104] A three-dimensional coordinate system is established for the power grid facility, and m coordinate points corresponding to the surface of the power grid facility are obtained; a laser sensor aligned with the power grid facility is used to scan the surface of the power grid facility at multiple coordinate points, i.e., scanning each of the m coordinate points; the laser sensor includes a laser emitter and a laser receiver, the laser emitter emits a laser beam, the laser beam is a short pulse laser beam, and the laser receiver receives the laser beam reflected back by the surface of the power grid facility; the reflection time of each coordinate point is obtained, the reflection time is the duration from laser emission to laser reception, the reflection time is obtained by a built-in timer of the laser emitter, when the laser emitter emits the laser beam, the timer starts timing, when the laser receiver receives the laser beam, the timer stops timing, and the data collected by the timer is the reflection time; the emission angle and the receiving angle of each coordinate point are further obtained, the emission angle is the angle between the laser emitter and the horizontal plane when the laser is emitted, and the receiving angle is the angle between the laser receiver and the horizontal plane when the laser is received, the emission angle is obtained by an absolute encoder built in the laser emitter, and the receiving angle is obtained by an absolute encoder built in the laser receiver.

[0105] According to the reflection time and the speed of light of each coordinate point, the laser path of each coordinate point is calculated, the laser path is the path passed by the laser when the laser is emitted to the laser receiver; the expression of the laser path is: ; in the formula, is the laser path, is the reflection time, is the speed of light, and the speed of light is defined as ; according to the laser path, the emission angle and the receiving angle of each coordinate point, the straight-line distance of each coordinate point is calculated, the straight-line distance is the vertical distance between each coordinate point and the laser sensor; the expression of the straight-line distance is: ; in the formula, is the straight-line distance, is the emission angle, is the receiving angle;

[0106] Point cloud data is formed according to the straight-line distances corresponding to the m coordinate points, polygon meshing is used to grid the point cloud data, the three-dimensional surface structure of the power grid facility is obtained, and is marked as a real-time structure; the real-time structure is compared with a preset power grid facility structure, the number of regions with different structures is counted, and the number of defects is obtained; the polygon meshing is a prior art and will not be described in detail here; the power grid facility structure is modeled and set by a person skilled in the art according to the design drawings of the power grid facility.

[0107] The life decline analysis module is used to analyze the analysis data and the real-time running data on the platform side to determine whether to generate a decline instruction;

[0108] The method for determining whether to generate a decline instruction includes:

[0109] The analysis data and the real-time operation data are taken as test data, the test data is input into the trained life decline analysis model, and whether the life of the power grid facility declines is determined.

[0110] The specific training process of the life decline analysis model includes:

[0111] The corresponding determination result is set for the c groups of test data in advance, c is an integer greater than 1, the determination result includes life decline and constant life, different digital labels are set for the life decline and the constant life, for example, the digital label is set to 0 for the life decline, and the digital label is set to 1 for the constant life; the determination result corresponding to the test data is collected by the person skilled in the art in the historical power grid facility life monitoring process, c groups of different test data are collected, and the person skilled in the art sets the corresponding determination result for the c groups of different test data according to actual experience under the condition of the c groups of different test data, whether the life of the power grid facility declines;

[0112] The digital label of the determination result is marked as a determination label, and the test data and the corresponding determination label are converted into a corresponding group of feature vectors;

[0113] Each group of feature vectors is taken as the input of the life decline analysis model, the life decline analysis model takes a group of predicted determination labels corresponding to each group of test data as the output, takes the actual determination label corresponding to each group of test data as the prediction target, and the actual determination label is the digital label of the determination result corresponding to the test data set in advance; the sum of all prediction errors is minimized as the training target; wherein the calculation formula of the prediction error is , wherein is the prediction error, is the group number of the feature vector corresponding to the test data, is the predicted determination label corresponding to the first group of test data, is the actual determination label corresponding to the first group of test data; the life decline analysis model is trained until the sum of the prediction errors converges and the training is stopped;

[0114] The above life decline analysis model is specifically a deep neural network model;

[0115] The corresponding determination result is obtained according to the predicted determination label, to determine whether the life of the power grid facility declines under the condition of the test data;

[0116] If the determination result is life decline, a decline instruction is generated;

[0117] If the determination result is constant life, no decline instruction is generated;

[0118] The life diagnosis module diagnoses the residual life of the power grid facility according to the analysis data and the real-time operation data if the decrease instruction is generated.

[0119] The step of diagnosing the residual life of the power grid facility comprises:

[0120] Step 1: encode the residual life into a chromosome, randomly generate N chromosomes to form an initial population, and determine the residual life range based on the technical specifications of the facility;

[0121] Step 2: the fitness is the accuracy, and the accuracy is determined by the reciprocal of the life error value, which is the absolute value of the difference between the expected life and the residual life;

[0122] Step 3: retain the top K chromosomes with the highest fitness and directly copy them to the next generation, randomly select U chromosomes for crossover according to the fitness probability, generate new chromosomes, and replace the individuals in the original population with the best ones. Select the remaining chromosomes to supplement the population capacity N;

[0123] Step 4: randomly select U chromosomes for crossover, generate new chromosomes, and replace the individuals in the original population with the best ones, randomly exchange the gene values in the chromosomes according to the preset mutation probability (such as 2%), and enhance the diversity of the population.

[0124] Step 5: when the preset maximum number of iterations is reached or a chromosome with a fitness value exceeding the threshold Q appears in the population, terminate the loop, output the optimal solution, and the optimal solution is the residual life corresponding to the chromosome with the maximum fitness.

[0125] The expected life is calculated by inversely integrating the negative factors such as current change amplitude and current error rate, index change amplitude and index error rate, hot spot number, flaw number and polarization rate, etc. The larger the sum of the independent weight items in the denominator, the shorter the expected life.

[0126] The larger the life error value, the smaller the corresponding accuracy, and vice versa. The expected life is expressed as follows:

[0127] ; in the formula, is the expected life, is the current change amplitude, is the current error rate, is the index change amplitude, is the index error rate, is the hot spot number, is the flaw number, is the polarization rate, the square root of the current change amplitude is used to suppress the nonlinear influence of extreme values; represents the coupling effect of the current change amplitude and the error rate (the unit is unified as the square of the current); , respectively represents the independent influence of the index change amplitude and its error, avoiding unit conflict; , , Still linear superposition, consistent with the independent defect assumption; , , , , , are weight factors; the specific values of the weight factors can be set according to actual conditions, and the weight factors reflect the influence degree of each test data on the expected life of the power grid facility. Through historical data regression analysis or machine learning optimization, the actual correlation degree of each parameter and life is quantified, and the optimal weight is automatically fitted, or the subjective score of experts in the field on the importance of the parameter is integrated into an objective weight by using tools such as analytic hierarchy process (AHP).

[0128] It should be noted that the data in the analysis data are all related influence data of the expected life of the power grid facility. The larger each data in the analysis data is, the smaller the expected life of the power grid facility will be, and vice versa.

[0129] The early warning module is used to generate early warning instructions on the platform side, and sends the early warning instructions, analysis data, real-time running data and residual life to the corresponding facility side, so that the relevant operation and maintenance personnel of the corresponding facility side can timely know the abnormal conditions of the power grid facility and related various data, and can effectively maintain and repair the power grid facility.

[0130] The embodiment can monitor the running state of the power grid facility in real time, and timely discover abnormal conditions by collecting long-term running data of the facility side and analyzing the analysis data. In-depth analysis by collecting real-time running data in combination with the analysis data can in-depthly judge whether the life of the power grid facility is reduced, so as to early warn and generate corresponding instructions. In addition, the individual residual life diagnosis of the power grid facility is realized by the genetic algorithm, which greatly improves the detection accuracy and diagnosis effect, can provide a scientific basis for the maintenance and update of the power grid facility, and improves the reliability and stability of the power grid operation. In-depth learning and dynamic decision of the life management of the power grid facility is realized, and the management efficiency and operation safety of the power grid facility are improved.

[0131] Embodiment 2

[0132] Please refer to Figure 3As shown, the embodiment is further improved on the basis of the embodiment 1, in the embodiment 1, only the early warning instruction is generated, that is, the early warning instruction generated under different life change conditions of the power grid facility is consistent, which cannot help the relevant operation and maintenance personnel to accurately master the life change condition of the power grid facility, in order to assist the relevant operation and maintenance personnel to more intuitively and clearly understand the life decline condition of the power grid facility, therefore, the embodiment provides a power supply facility diagnosis and early warning system for a smart grid, and further comprises an early warning instruction matching module;

[0133] The early warning instruction matching module is used for analyzing the residual life of the power grid facility on the platform side and matching the corresponding early warning instruction.

[0134] The initial operation life and the importance of the power grid facility are obtained; the initial operation life is the expected life of the power grid facility under normal operation conditions, and the initial operation life is obtained according to the model parameter table provided by the power grid facility manufacturer; the importance is the importance of the power grid facility in the power grid, and the importance is evaluated by the technical personnel in the art according to the influence of the safety of each power grid facility in the power grid on the power grid in combination with the power system analysis software, and the influence area of each power grid facility when the power grid fails is analyzed through the circuit simulation, and the importance of each power grid facility is comprehensively evaluated; the higher the importance, the higher the importance set for the power grid facility, and the lower the importance, the lower the importance set for the power grid facility; the importance ranges from 0 to 1. J is an integer greater than 1, and in the embodiment, J is preferably 5.

[0135] The residual life is divided by the initial operation life to obtain the life change rate; according to the importance of the power grid facility, the corresponding early warning instruction is divided into V different levels of early warning instruction, that is, the corresponding early warning instruction includes the first early warning instruction, …, and the Vth early warning instruction; then 1 is divided by the importance to obtain the matching division degree, and the life change rate range corresponding to each level of early warning instruction is , wherein t is the early warning instruction level, , is the matching division degree; then the life change rate of the power grid facility is compared with the life change rate range corresponding to each level of early warning instruction, and the early warning instruction corresponding to the life change rate range in which the life change rate is located and the highest level of early warning instruction are taken as the matching early warning instruction of the power grid facility.

[0136] For example, the importance of a power grid facility is 4, so the corresponding early warning instruction of the power grid facility is divided into 4 levels of early warning instruction, that is, the corresponding early warning instruction includes the first early warning instruction, the second early warning instruction, the third early warning instruction and the fourth early warning instruction; the matching division degree is , wherein the first early warning instruction corresponds to the life change rate range , the second early warning instruction corresponds to the life change rate range The third early warning instruction corresponds to a life change rate range of The fourth early warning instruction corresponds to a life change rate range of And the residual life of the power grid facility is 20, and the initial operation life is 30, so the life change rate of the power grid facility is And Therefore, the early warning instruction matched by the power grid facility is the third early warning instruction.

[0137] It should be noted that the reason why the early warning instruction matched by the power grid facility includes the highest level early warning instruction is that different power grid facilities have different importance, and the number of early warning instruction levels corresponding to the early warning instruction is also different, so the same level early warning instruction represents different life change rates in power grid facilities with different importance. If the highest level early warning instruction is not included, the operation and maintenance personnel cannot understand the meaning of the generated early warning instruction, for example: the importance of one power grid facility is 2, and the importance of one power grid facility is 5. At this time, both power grid facilities generate the first early warning instruction. For the power grid facility with importance 5, the life change rate is And for the power grid facility with importance 2, the life change rate is Therefore, it is necessary to combine the highest level early warning instruction to help operation and maintenance personnel accurately understand the meaning of the early warning instruction matched by the power grid facility.

[0138] In this embodiment, the early warning instruction is divided into multiple levels according to the importance of the power grid facility, and the life change rate is calculated to match the early warning instruction of the corresponding level. This can improve the accuracy and practicality of early warning, provide personalized and target-oriented intelligent decision support for operation and maintenance personnel, help relevant operation and maintenance personnel accurately master the life change of the power grid facility, optimize resource allocation and management, and provide more effective support for the management and operation of the power grid facility.

[0139] Embodiment 3

[0140] Please refer to Figure 4 The part not described in detail in this embodiment can be seen from the description of embodiment 1 and embodiment 2. A power supply facility diagnosis and early warning method for a smart grid is provided, and the method comprises the following steps:

[0141] The platform end receives the long-term operation data collected by the facility end;

[0142] The platform end analyzes the long-term operation data, obtains analysis data, and determines whether to generate an abnormal instruction;

[0143] If an abnormal instruction is generated, the platform end receives the real-time operation data collected by the facility end;

[0144] The platform end analyzes the analysis data and the real-time running data to determine whether to generate a descending instruction;

[0145] If the descending instruction is generated, the platform end diagnoses the residual life of the power grid facility according to the analysis data and the real-time running data;

[0146] The platform end generates a warning instruction and sends the warning instruction, the analysis data, the real-time running data and the residual life to the corresponding facility end.

[0147] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0148] Finally, the above is only a preferred embodiment of the present application and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for diagnosing and warning power supply facilities for smart grids, characterized in that, include: The platform receives long-term operational data collected from the facilities. The platform analyzes long-term operational data, obtains analytical data, and determines whether abnormal instructions are generated. If an abnormal command is generated, the platform will receive real-time operational data collected by the facility. The platform analyzes the analytical data and real-time operational data to determine whether to generate a descent command; The method for determining whether a descent instruction has been generated includes: The analytical data and real-time operational data are used as test data. The test data is then input into the trained lifespan degradation analysis model to determine whether the lifespan of the power grid facilities has decreased. Based on the predicted judgment labels, obtain the corresponding judgment results to determine whether the lifespan of the power grid facilities has decreased under the test data conditions. If the judgment result is a decrease in lifespan, then a decrease instruction is generated; If the judgment result is that the lifetime is constant, then no descent instruction will be generated; If a descent command is generated, the platform will diagnose the remaining lifespan of the power grid facilities based on the analysis data and real-time operating data. The platform analyzes the remaining lifespan of power grid facilities and matches corresponding early warning instructions. Obtain the initial operational life and importance of power grid facilities; initial operational life is the expected number of years a power grid facility can operate under normal operating conditions; importance is the degree of importance of the power grid facility within the power grid, with an importance range of [missing information]. J is an integer greater than 1; Divide the remaining lifetime by the initial operating lifetime to obtain the lifetime change rate; based on the importance of the power grid facilities, classify the corresponding early warning instructions into V different levels of early warning instructions. That is, the corresponding warning instructions include the first warning instruction, ..., the Vth warning instruction; then divide 1 by the importance to obtain the matching degree, and the range of lifespan change rate corresponding to each level of warning instruction is... Where t represents the warning instruction level, , To match the classification degree; then compare the life change rate of the power grid facilities with the life change rate range corresponding to each level of warning instructions, and take the warning instructions corresponding to the life change rate within the life change rate range and the highest level warning instructions as the warning instructions that match the power grid facilities. The platform generates early warning commands and sends the commands, analysis data, real-time operating data, and remaining lifespan to the corresponding facilities.

2. The method for diagnosing and warning power supply facilities for smart grids according to claim 1, characterized in that, The long-term operating data includes current values ​​and oil quality indicators; The current value is collected according to the preset sampling interval. The calculation method for oil product indicators is as follows: Oil product impact data, including total acid value, mineral content, and lead content, is collected at the collection interval. A lead content threshold is preset, and the lead content is compared with this threshold. If the lead content is less than or equal to the threshold, the lead content in the oil product impact data is adjusted to 0. If the lead content is greater than the threshold, the lead content is subtracted from the threshold to obtain the lead content difference, and the lead content in the oil product impact data is adjusted to reflect this difference. The oil product indicators are then calculated based on the oil product impact data.

3. The method for diagnosing and warning power supply facilities for smart grids according to claim 2, characterized in that, The method for acquiring and analyzing data includes: A current variation curve is constructed based on the obtained n current values, and an oil quality indicator variation curve is constructed based on the obtained n oil quality indicators. The current variation curve represents the continuous change of current during the operation of the power grid facilities, and the oil quality indicator variation curve represents the continuous change of oil quality indicators during the operation of the power grid facilities. The time corresponding to the nth current value in the current variation curve is marked as the prediction time. Inputting a current value into a pre-trained current prediction model, the model predicts the current value corresponding to the prediction time and marks it as the predicted current value; the preceding current value in the oil product index change curve is then used to predict the current value. Each oil product indicator is input into the trained indicator prediction model, which predicts the oil product indicator corresponding to the prediction time and marks it as the predicted oil product indicator. Divide the nth current value in the current change curve by the predicted current value to obtain the current error rate; divide the nth oil index in the oil index change curve by the predicted oil index to obtain the index error rate. Subtract the nth current value from the current change curve. The current value is used to obtain the current change amplitude. The nth oil product index in the oil product index change curve is subtracted from the nth current value. The magnitude of change in each oil product indicator was obtained. The current variation range, the index variation range, the current error rate, and the index error rate are used as the analysis data.

4. The method for diagnosing and warning power supply facilities for smart grids according to claim 3, characterized in that, The method for determining whether an abnormal instruction has been generated includes: The analyzed data is input into a trained probability analysis model to predict the probability of anomalies in power grid facilities. Preset probability threshold ; the probability of anomalies With probability threshold Compare; like If so, no exception instruction will be generated; like If so, an exception instruction will be generated.

5. A method for diagnosing and warning power supply facilities for smart grids according to claim 4, characterized in that, The real-time operating data includes polarization rate, number of hot spots, and number of defects; The polarization rate is the degree of imbalance in the vibration of power grid facilities in different directions. The number of hotspots refers to the number of abnormal temperature points on the surface of power grid facilities. The number of defects refers to the number of physical defects on the surface of the power grid facilities.

6. A method for diagnosing and warning power supply facilities for a smart grid according to claim 5, characterized in that, The method for obtaining the polarization rate is as follows: Obtain triaxial vibration values, including X-axis, Y-axis, and Z-axis vibration values; mark the vibration value with the largest value as the maximum vibration value and the vibration value with the smallest value as the minimum vibration value; obtain the vibration deviation value based on the difference between the maximum and minimum vibration values; sum each vibration value in the triaxial direction sequentially and take the average to obtain the vibration average value; obtain the polarization rate based on the ratio of the vibration deviation value to the vibration average value. The method for obtaining the number of hotspots is as follows: acquire a temperature distribution image; preset a temperature threshold, compare the temperature value of each pixel in the temperature distribution image with the temperature threshold, mark the pixels with temperature values ​​greater than the temperature threshold as hotspots, and do not mark the pixels with temperature values ​​less than or equal to the temperature threshold. The number of hot topics is counted to obtain the total number of hot topics; The method for obtaining the number of defects includes: A three-dimensional coordinate system is established for the power grid facilities to obtain m coordinate points corresponding to the surface of the power grid facilities. A laser sensor is used to scan the surface of the power grid facilities at multiple coordinate points, that is, all m coordinate points are scanned. The laser sensor includes a laser emitter and a laser receiver. The laser emitter emits a laser beam, which is a short-pulse laser beam, and the laser receiver receives the laser beam reflected back from the surface of the power grid facilities. The reflection time of each coordinate point is obtained, which is the duration from laser emission to laser reception. The emission angle and reception angle of each coordinate point are also obtained. The emission angle is the angle between the laser emitter and the horizontal plane when the laser is emitted, and the reception angle is the angle between the laser receiver and the horizontal plane when the laser is received. Based on the reflection time and speed of light at each coordinate point, the laser path at each coordinate point is calculated. The laser path is the distance the laser travels from emission to reception. Based on the laser path, emission angle, and reception angle at each coordinate point, the straight-line distance at each coordinate point is calculated. The straight-line distance is the perpendicular distance between each coordinate point and the laser sensor. Point cloud data is generated based on the straight-line distances between m coordinate points. The point cloud data is then reconstructed using a polygonal mesh method to obtain the three-dimensional surface structure of the power grid facility, which is then marked as the real-time structure. The real-time structure is compared with the preset power grid facility structure, and the number of areas with different structures is counted to obtain the number of defects.

7. A method for diagnosing and warning power supply facilities for a smart grid according to claim 6, characterized in that, The steps for diagnosing the remaining life of power grid facilities include: Step 1: Encode the remaining lifetime into chromosomes, and randomly generate N chromosomes to form the initial population; Step 2: Fitness is the same as accuracy. Accuracy is determined by the reciprocal of the lifetime error value, which is the absolute value of the difference between the expected lifetime and the remaining lifetime. Step 3: Retain the top K chromosomes with the highest fitness and directly copy them to the next generation. Randomly select U chromosomes according to fitness probability for crossover. After generating new chromosomes, replace individuals in the original population with the best ones. Select the remaining chromosomes to make up the population size N. Step 4: Randomly select U chromosomes for crossover to generate new chromosomes. Then, select the best individuals in the original population to replace them and randomly exchange gene values ​​in the chromosomes according to the preset mutation probability. Step 5: When the preset maximum number of iterations is reached, or when a chromosome with fitness exceeding the threshold appears in the population, terminate the loop and output the optimal solution. The optimal solution is the residual life of the chromosome with the highest fitness.

8. A power supply facility diagnosis and early warning system for smart grids, implementing the power supply facility diagnosis and early warning method for smart grids as described in any one of claims 1-7, characterized in that, include: The long-term data receiving module is used by the platform to receive long-term operational data collected by the facility. The long-term data analysis module is used by the platform to analyze long-term running data, obtain the analysis data, and determine whether abnormal instructions are generated. If the real-time data receiving module generates an abnormal command, the platform will receive the real-time operation data collected by the facility. The lifespan degradation analysis module is used by the platform to analyze the analysis data and real-time operation data to determine whether to generate a degradation command. If the lifespan diagnosis module generates a decline command, the platform will diagnose the remaining lifespan of the power grid facilities based on the analysis data and real-time operating data. The early warning module is used to generate early warning commands on the platform and send the early warning commands, analysis data, real-time operating data, and remaining lifespan to the corresponding facilities.

Images