Electric power field operation guidance interaction system and method

Abstract

The invention discloses an electric power field operation guidance interaction system and method, and belongs to the technical field of electric power construction. The method comprises the steps that double-end communication connection is constructed, continuous supervision data and operation abnormal performance of target power equipment are acquired, a blank analysis space is constructed, and a suspended potential discharge abnormal signal is generated; according to the method, the core problems that a guidance scheme flows on the surface and interaction communication is low in efficiency and delayed are fundamentally solved, and sustainable and optimized intelligent supporting capacity of power field operation is formed through data driving and knowledge precipitation in the whole process; the early warning capability, the handling efficiency and the safety management and control level of potential faults of complex power equipment are remarkably improved, an innovative development direction of a power field operation guidance interaction technology is represented, and multi-source deep diagnosis, dynamic risk assessment and intelligent decision interaction are organically fused to achieve the purpose of improving the safety of the power field operation guidance interaction technology. And a closed-loop intelligent interaction system integrating monitoring, analysis, decision making and guidance is constructed.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of electric power construction, in particular to an electric power on-site operation guidance interaction system and method. BACKGROUND

[0002] Electric power on-site operation refers to a series of production activities such as installation, maintenance, test, debugging, operation and maintenance and accident handling performed by professional personnel at the site or surrounding area of power facilities such as power generation, power transmission, power transformation and power distribution; electric power operation guidance interaction communication refers to the information transmission and feedback process between on-site operators and experts, work supervisors or guardians in the rear for the purpose of precise guidance, collaborative operation, safety supervision and decision support in the process of electric power on-site operation.

[0003] In combination with the above content, it needs to be explained that the Chinese patent with the application number CN2021110136883 discloses a system and device for electric power grid on-site operation guidance, which obtains the electric power grid on-site operation guidance scheme according to the actual situation of the electric power grid on-site operation and in combination with the training of the neural network model, so as to avoid the emergence of electric power grid hazards caused by the insufficient technical level and ability of the operators when they perform the electric power grid on-site operation according to experience;

[0004] However, the above-mentioned guidance scheme generated according to the actual situation lacks the root cause exploration of the actual problems faced by the user electric power operation site, as well as the high-precision analysis of the problem causes and the accompanying associated abnormalities, for example, during the electric power on-site operation of the GIS device, the generated guidance scheme can only solve the surface problems, or cannot accurately, quickly and efficiently interact with the user electric power operation, resulting in low electric power on-site operation efficiency and delayed construction period.

[0005] In view of the above technical defects, a solution is proposed. SUMMARY

[0006] The purpose of the present application is to provide an electric power on-site operation guidance interaction system and method to solve the problems.

[0007] To achieve the above purpose, the present application provides the following technical scheme: an electric power on-site operation guidance interaction method, comprising the following steps:

[0008] S1, a double-end communication connection is constructed, continuous supervision data and operation abnormal performance of a target electric power device are obtained, a blank analysis space is constructed, and a suspended potential discharge abnormal signal is generated;

[0009] S2, a GIS device structure digital model is constructed based on the suspended potential discharge abnormal signal, a UHF signal intensity matrix and an acoustic signal intensity matrix are obtained, a defect area comprehensive judgment report and a confidence level rating are generated;

[0010] S3, constructing a discharge mode feature vector according to the electric signal pulse, the signal amplitude stable peak value and the signal pulse number sequence, and jointly analyzing with a historical accident case library to generate a load-discharge correlation analysis conclusion;

[0011] S4, sorting the key of the multiple types of data to obtain four core parameters and defining risk contribution weights, constructing a risk quantification model to compare to obtain a risk level identifier, and jointly constructing a field operation guidance scheme combining a decision tree knowledge base.

[0012] Further, the hierarchical analysis process of the suspended potential discharge abnormal signal in the S1 step is as follows:

[0013] Based on the electric signal pulse screened along the cycle node based on multiple types of data, the signal amplitude of the electric signal pulse is analyzed in multiple dimensions, and the signal amplitude stable peak value, continuity, and phase distribution area of the electric signal pulse are compared along the cycle node: if there is an abnormal peak value node that exceeds the average value of several cycle nodes by inverse extrapolation, or there is a peak value fracture, or the electric signal pulse exists in any one of the dispersed arrangement, it is determined that the electric signal pulse is abnormal, and a partial discharge signal is generated.

[0014] Based on the noise vibration data screened along the cycle node based on multiple types of data, the noise monomer sub-data and the vibration monomer sub-data are compared frame by frame, and if both exist, the similarity of the noise monomer sub-data and the pre-stored current preset audio is further analyzed, and it is determined that the noise vibration data with a similarity exceeding a preset value is abnormal, a sound wave signal is generated, and is marked as a discharge point according to the detection position.

[0015] Further, based on the SF6 gas data extracted along the cycle node based on multiple types of data, the key feature micro-occupancy of sulfur dioxide, hydrogen sulfide and sulfur fluoride is compared based on the detection result: if there is key feature micro-occupancy data, it is determined that the internal discharge of the power equipment exists and has caused chemical evidence of gas ionization, and a gas decomposition signal is generated.

[0016] The partial discharge signal, the sound wave signal and the gas decomposition signal generated by the multiple types of data analysis are summarized, which are consistent with the abnormal performance of the operation, and if there is any one or more of the same symptoms, it is determined that the accuracy of the continuous supervision data and the operation abnormal performance are consistent, and a suspended potential discharge abnormal signal is generated.

[0017] Further, the analysis process of the defect area comprehensive judgment report in the S2 step is as follows:

[0018] Based on the definition summary of the discharge anomaly signal of the suspension potential, key sensitive words are extracted to extract features from the data covered in the partial discharge signal and acoustic signal, to obtain the signal strength, pulse number covered in the partial discharge signal, acoustic intensity, and discharge point coordinate covered in the acoustic signal. The above data are labeled as UHF signal intensity sequence, signal pulse number sequence, acoustic signal intensity sequence, and discharge point coordinate, respectively. A three-dimensional coordinate grid is established by digitizing the GIS device structure model.

[0019] Further, in the three-dimensional coordinate grid, first, the spatial position coordinates of all UHF sensors are labeled, and the time difference positioning algorithm is used to calculate the possible source coordinates of the pulse signal in the three-dimensional grid. The number of signal sources calculated in each grid unit within the supervision period is counted, and the amplitudes of these signals are accumulated. For each grid unit, the signal number and accumulated amplitude are normalized to generate a UHF signal intensity matrix.

[0020] On the surface of the three-dimensional coordinate grid model, the accurate position coordinates of each detection point during the on-site ultrasonic detection are labeled. The value of the acoustic signal determined as abnormal in S1 is assigned to the surface grid unit where the corresponding detection point is located. For surface grid units without direct detection data, the adjacent point interpolation algorithm is used to calculate the value of the unit according to the intensity of the grid units with existing data in the surrounding area. Finally, an acoustic signal intensity matrix is generated.

[0021] Further, the two matrices are superimposed and analyzed on the three-dimensional coordinate grid. The matching degree of the signal intensity of each grid unit is calculated to obtain the overlapping area where both signals show high intensity. A collaborative positioning confidence heat map is generated, which visually indicates the possible defect area with color depth, and marks the highest confidence area as the suspected defect core area.

[0022] The gas micro-occupancy data covered in the gas decomposition signal are associated with the time sequence to generate a gas concentration time trend line, which is superimposed on the collaborative positioning confidence heat map. It is determined whether the positioning area is consistent with the upward direction of the gas sampling flow channel, and whether the signal active period coincides with the rapid growth period of gas concentration. A defect area comprehensive judgment report is generated to clearly indicate the 1-2 gas chambers or components where the defect is most likely to exist, and give a confidence rating.

[0023] Further, the generation process of the load-discharge correlation analysis conclusion in S3 is as follows:

[0024] Analyze the phase distribution area, calculate the phase distribution aggregation degree of the first and third quadrants of the power frequency voltage, the fluctuation coefficient of the stable peak value of the signal amplitude, and the time distribution uniformity of the pulse number sequence, generate a discharge mode feature vector containing three key indicators: phase distribution aggregation degree, amplitude fluctuation coefficient, and pulse uniformity. Calculate the similarity between the generated discharge mode feature vector and each historical defect feature vector in the historical accident case library to generate a historical matching degree ranking list listing the top three most likely defect types and their matching probabilities.

[0025] According to the continuous monitoring data provided by the user end, extract the line load current historical accident case library, and draw the superimposed curve graph of the partial discharge signal activity and the line load current with the cycle node, and calculate the correlation coefficient; generate the load-discharge correlation analysis conclusion.

[0026] Further, the construction process of the S4 step about the field operation guidance scheme is as follows:

[0027] In the blank analysis space, call the historical accident case library, analyze the key parameter sequence that caused the fault escalation in the past, combine the expert experience rule library, and sort the multi-type data generated in S1 and S3 stages, select the current concentration of sulfur dioxide, the concentration growth rate of hydrogen sulfide, the maximum amplitude of the ultra-high frequency signal, and the discharge activity stability index four core parameters and assign risk contribution weights;

[0028] According to the preset weight proportion, the parameter values are weighted and summed: the measured value of each parameter is multiplied by the corresponding weight, and all the product results are added, and finally a device risk index between zero and one hundred percent is obtained; compare the calculated device risk index with the threshold interval preset in the model: if the index is located in the zero to forty percent interval, it is classified as low risk; forty percent to sixty percent is classified as medium risk; sixty percent to eighty percent is classified as high risk; eighty percent to one hundred percent is classified as emergency risk, and the final risk level identifier is generated.

[0029] Further, call the decision tree knowledge base in the guidance end cloud database; take the current device risk level as the primary matching condition, and the primary defect type determined in S3 as the secondary matching condition, and search in the decision tree knowledge base; for example, when the condition combination is high risk level and loose shielding cover, the decision tree automatically matches to the planned power outage maintenance main path, and further fills in the specific maintenance steps, required tool list and safety control measures according to the defect area information; integrate all matching results to generate a structured field operation guidance scheme.

[0030] A power field operation guidance interaction system, comprising the following steps:

[0031] Multi-terminal data acquisition module: based on online communication, user terminal and guidance terminal are constructed, continuous supervision data of the target power equipment of the user terminal and abnormal performance of the work feedback of the user terminal power field work are received, and the cloud database of the guidance terminal is combined to jointly construct a blank analysis space;

[0032] Multi-source signal collaborative analysis module: integrate multi-source data, and perform spatial positioning and type preliminary judgment on defects;

[0033] Root cause depth analysis module: in-depth analysis of the physical properties of defects, and comparison with historical accident case library to accurately determine the root cause of defects;

[0034] Risk assessment and disposal decision joint module: based on all previous analysis, the device risk is quantitatively evaluated, and a concrete field operation guidance scheme is generated.

[0035] The beneficial effects of the present application are:

[0036] 1. The present application is to fuse and analyze multi-type data in multi-dimensions, which can sensitively identify the early weak signal of the latent insulation defect of the suspended potential discharge, and use three-dimensional coordinate grid and collaborative positioning algorithm to convert the abstract signal anomaly into the specific defect space position inside the device, effectively solving the technical pain points of unclear cause diagnosis and fuzzy positioning caused by data fragmentation and surface analysis in the prior art, and improving the depth and accuracy of problem analysis from the root cause, and realizing the depth diagnosis and accurate positioning of the latent defect inside the GIS device.

[0037] 2. The present application is to generate a quantitative device risk index based on multi-parameter weighted calculation, and combine the historical case library to perform intelligent matching and load correlation analysis of the defect type, which can dynamically evaluate the development trend and risk level of the defect, and automatically generate a structured field operation guidance scheme containing specific operation steps and safety measures through the decision tree knowledge base, completely changing the traditional communication mode relying on manual experience and low interaction efficiency, realizing the change from passive response to active early warning and accurate decision, greatly improving the efficiency and safety of field operation, and realizing the prospective evaluation of the device state and the precise guidance of field operation. BRIEF DESCRIPTION OF DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0039] Figure 1A flow chart of the method of the present application;

[0040] Figure 2 A flow chart of the system of the present application. DETAILED DESCRIPTION

[0041] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0042] Embodiment one: please refer to Figure 1 - Figure 2 As shown in the figure, the embodiment is an electric power field operation guidance interaction system and method, comprising the following steps:

[0043] S1, a multi-terminal data acquisition module, based on online communication, constructs a user terminal and a guidance terminal, receives continuous supervision data of the target electric power equipment from the user terminal, and receives operation abnormal performance of the user terminal electric power field operation feedback, combines the cloud database of the guidance terminal, and jointly constructs a blank analysis space;

[0044] In the blank analysis space, the continuous supervision data is disassembled along the time line and the multi-type data about the operation abnormal performance is performed according to the data type, so as to construct a supervision period from the supervision starting moment to the current supervision moment, uniformly process the multi-type data acquisition time and the time trend of the supervision period, divide them at equal intervals, and construct a plurality of cycle nodes;

[0045] Based on the electric signal pulse screened along the cycle node of the multi-type data, the signal amplitude of the electric signal pulse is analyzed in multiple dimensions, the signal amplitude stable peak value, continuity of the electric signal pulse along the cycle node trend is compared, and the phase distribution area of the electric signal pulse is compared: if there is an abnormal peak node exceeding the average value of the inverse 5 cycle nodes in the latest signal amplitude stable peak value, or there is a peak fracture, or the electric signal pulse exists in any one of the dispersed arrangement, it is judged that the electric signal pulse exists abnormally, a partial discharge signal is generated, and the electric signal pulse data can be detected by a UHF sensor to obtain a GIS cavity;

[0046] Based on the noise vibration data screened along the cycle node of the multi-type data, the noise monomer sub-data and the vibration monomer sub-data exist, then the noise monomer sub-data and the pre-stored current preset audio similarity are further analyzed, it is judged that the noise vibration data with a similarity exceeding 80% is abnormal, a sound wave signal is generated, and the detection position is marked as a discharge point according to the preset value, which can be 80%, and the preset value is judged by historical experience;

[0047] When the noise monomer sub-data and the vibration monomer sub-data are not synchronized, it is determined that the invalid interference data exists, and the noise and vibration signal data corresponding to the discharge can be detected by the ultrasonic detector on the GIS shell;

[0048] Based on the SF6 gas data extracted along the periodic node based on multiple types of data, based on the detection result, the key characteristic micro-occupancy ratio of sulfur dioxide, hydrogen sulfide and sulfur fluoride is compared: if the key characteristic micro-occupancy ratio data exists, it is determined that the internal discharge of the power equipment exists and has caused the chemical evidence of gas ionization, and a gas decomposition signal is generated. SF6 gas data is used to analyze the composition of SF6 gas in the equipment;

[0049] The partial discharge signal, the sound wave signal and the gas decomposition signal generated by the summary of multiple types of data analysis are consistent with the abnormal performance of the operation, and if there is any one or more same symptoms, the accuracy of the continuous supervision data and the consistency of the abnormal performance of the operation are determined, and a suspended potential discharge abnormal signal is generated:

[0050] The guide end sends the text understanding of the suspended potential discharge abnormal signal to the user end according to the cloud data and the power equipment use guide, specifically: continuous discharge will continuously decompose SF6 gas to generate conductive and corrosive substances, reduce the insulation strength of the gas, and form a vicious cycle, and the gas insulation performance is reduced;

[0051] When the discharge energy accumulates to a certain extent, or after the carbonization channel of the solid insulation surface is formed, the surface flashover or breakdown of the main insulation will be triggered, which will cause the internal short circuit and grounding fault of the GIS equipment; At this time, the main protection such as differential protection and bus protection will trip, the internal pressure of the equipment will rise sharply, which may cause the rupture of the explosion-proof membrane and even the explosion of the shell, causing catastrophic power outage and equipment damage. The abnormal performance of the user's power field operation is summarized, and the problems encountered in the power field operation are summarized and defined.

[0052] S2, multi-source signal cooperative analysis module, integrating multi-source data, spatial positioning and type preliminary judgment of defects;

[0053] Based on the definition summary of the suspended potential discharge abnormal signal, the key sensitive words are extracted, and the data covered in the partial discharge signal and the sound wave signal are extracted, and the signal strength, pulse number covered in the partial discharge signal, sound wave intensity, discharge point coordinate covered in the sound wave signal are obtained. Marked as ultra-high frequency signal intensity sequence, signal pulse number sequence, sound wave signal intensity sequence and discharge point coordinate;

[0054] In the blank analysis space, the two-dimensional engineering design drawings and three-dimensional design models pre-stored in the corresponding equipment are called, the geometry, size and relative position of each component along the XYZ three-dimensional space axis of the equipment are analyzed, the three-dimensional entity modeling of the key components of the circuit breaker gas chamber, the disconnecting switch gas chamber, the busbar cylinder, the basin-type insulator and the conductor is carried out, and the equipment number and model attributes are added, the GIS equipment structure digital model is constructed, the spatial structure, component composition and assembly relationship of the physical equipment are converted into accurate digital models that can be recognized and calculated by the computer, and the model digitization is carried out based on the target electric power equipment original data stored in the guidance end cloud database;

[0055] Based on the outer contour of the GIS equipment structure digital model, a smallest cubic space completely wrapping the model is constructed as a reference space; along the X, Y and Z axes of the reference space, a series of continuous cubic units are generated by equidistant division according to the preset accuracy; the cubic units are associated and matched with the component information of the digital model, and the gas chamber or component information where each unit is located is marked, and finally the three-dimensional coordinate grid is constructed; the continuous equipment space is discretized into regular three-dimensional grid units, which provides a standard coordinate system for subsequent mapping of detection signals to spatial positions;

[0056] In the three-dimensional coordinate grid, first, the spatial position coordinates of all ultra-high frequency sensors are marked; for each abnormal electric signal pulse screened out in S1, based on the nanosecond-level time difference of the pulse signal arriving at different sensors, a time difference positioning algorithm is used to calculate the possible source coordinates of the pulse signal in the three-dimensional grid; the number of signal sources calculated in each grid unit within the supervision period is counted, and the amplitudes of these signals are accumulated; for each grid unit, the signal number and accumulated amplitude are normalized to generate an ultra-high frequency signal intensity matrix, which directly reflects the spatial distribution intensity of the ultra-high frequency signal inside the equipment; the time and space characteristics of the ultra-high frequency signal are converted into the intensity distribution of the signal in the three-dimensional space inside the equipment, and the spatial preliminary positioning of the discharge point is realized;

[0057] On the surface of the three-dimensional coordinate grid model, the accurate position coordinates of each detection point during the field ultrasonic detection are marked; the value of the abnormal acoustic signal determined in S1 is assigned to the surface grid unit where the corresponding detection point is located; for the surface grid unit without direct detection data, the adjacent point interpolation algorithm is used to calculate the value of filling the unit according to the intensity of the grid units with existing data around it; finally, an acoustic signal intensity matrix is generated, which describes the intensity distribution of the abnormal acoustic signal on the surface of the equipment shell; the detection result of the acoustic signal is positioned to the specific position on the surface of the equipment shell, which provides a basis for external positioning of defects;

[0058] The two matrices are superimposed on a three-dimensional coordinate grid, the matching degree of the two signal strengths is calculated for each grid cell, and the overlapping area where both show high strength is obtained; a collaborative positioning confidence heat map is generated, which visually indicates the possible defect area with color depth, and marks the highest confidence as the suspected defect core area; through cross verification of multi-source data, the false positives and ambiguous positioning of single means are solved, and the positioning reliability is improved;

[0059] The gas trace proportion data included in the gas decomposition signal is associated with the time sequence to generate a gas concentration time trend line, which is superimposed on the collaborative positioning confidence heat map; it is judged whether the positioning area is consistent with the upward direction of the gas sampling flow channel, and whether the signal active period coincides with the rapid growth period of gas concentration; a defect area comprehensive judgment report is generated to clearly indicate the 1-2 gas chambers or components where the defect is most likely to exist, and give a confidence rating;

[0060] The guidance end sends the defect area comprehensive judgment report to the user end, and the specific text guidance is: according to the collaborative analysis of multi-source signals, it is preliminarily determined that the discharge defect is located near the XX component of the XX gas chamber, with a high confidence; it is suggested that the on-site inspection personnel should preferentially perform appearance inspection and precision detection review on this area, and build an interactive and output guidance end.

[0061] S3, root cause analysis module, in-depth analysis of the physical characteristics of the defect, and comparison with the historical accident case library to accurately determine the root cause of the defect;

[0062] In the blank analysis space, the phase angle of each abnormal electric signal pulse corresponding to the power frequency voltage is obtained; ninety degrees is taken as one quadrant, and the power frequency cycle from zero to three hundred and sixty degrees is divided into four quadrants, the distribution number of all abnormal pulses in each quadrant is counted to form a phase distribution number statistical sequence, the sum of the pulse numbers of the two quadrants with the most pulses is calculated, and the ratio of the sum to the total number of pulses is calculated, and this ratio is defined as the phase distribution aggregation degree; a quantitative phase distribution aggregation degree value is generated, the higher the value, the more concentrated the phase distribution of discharge activity in the power frequency cycle; the abnormal electric signal pulses selected based on the ultra-high frequency sensor data in S1 and the phase reference obtained from the power frequency voltage reference signal identify the occurrence rule of discharge activity in the power frequency cycle, and different defect types will show different phase aggregation characteristics;

[0063] Read all signal amplitude stable peak value data in the supervision cycle, and constitute an amplitude stable peak value sequence; first calculate the arithmetic mean of all peak values in the sequence, and then calculate the standard deviation of the sequence, which represents the dispersion degree of each peak value relative to the average value; divide the standard deviation by the average value to obtain the fluctuation coefficient of the signal amplitude stable peak value; generate a fluctuation coefficient value, the larger the value, the more unstable the discharge signal intensity, the signal amplitude stable peak value sequence obtained by analyzing the cycle nodes in S1 measures the fluctuation of discharge intensity, and different amplitude fluctuations correspond to different physical characteristics of the discharge source;

[0064] Divide the supervision cycle into a plurality of time windows according to a fixed time length, and count the number of pulses in each window to form a pulse number sequence, calculate the average value of all elements in the sequence, and calculate the standard deviation of the sequence; divide the standard deviation by the average value to obtain the time distribution uniformity of the pulse number sequence; generate a uniformity value, the closer the value is to zero, the more uniform the discharge activity in time, and the larger the value, the more intermittent or explosive the discharge activity, the pulse number sequence extracted from the ultra-high frequency signal in S2 and used to construct the intensity matrix, judges whether the discharge occurs continuously, periodically or randomly, and the time mode is a key indicator for distinguishing defect types;

[0065] Create a data structure containing three fixed fields; fill the calculated phase distribution aggregation degree value into the first field, the amplitude fluctuation coefficient value into the second field, and the pulse uniformity value into the third field; generate a discharge mode feature vector containing the data of the three key indicators, which is the unique feature fingerprint of this discharge activity, and is output to the subsequent historical accident case library matching process; The phase distribution aggregation degree, amplitude fluctuation coefficient and pulse uniformity calculated in the foregoing steps condense complex discharge waveform features into a standardized, multi-dimensional data combination, facilitating fast matching and similarity calculation with the historical accident case library;

[0066] Retrieve the historical accident case library in the guidance end cloud database, and calculate the similarity between the generated discharge mode feature vector and each historical defect feature vector in the historical accident case library, to generate a historical matching degree sorting list listing the top 3 most likely defect types and their matching probabilities;

[0067] According to the line load current historical accident case library and the partial discharge signal activity provided by the user terminal, draw a superimposed curve graph of the partial discharge signal activity and the line load current changing with the cycle node, and calculate the correlation coefficient of the two; generate a load-discharge correlation analysis conclusion to indicate whether the defect is related to the flow guide circuit state, and to judge whether the discharge activity is related to the electrical / mechanical stress of the equipment operation, so as to distinguish the defect types;

[0068] The guidance end integrates and sends the historical matching degree ranking list and the load-discharge correlation analysis conclusion to the user end. The specific guidance in words is: comprehensive discharge mode and historical case matching result, the primary suspected is the shield loose, the matching degree is eighty-five percent, and the discharge activity is strongly positively correlated with the load current, which further confirms the judgment; the root cause may be the insufficient tightening torque during installation or loosening due to long-term vibration. It is recommended to focus on checking the fasteners at this part on site, and to build the guidance end interaction and output.

[0069] Example two:

[0070] S4, risk assessment and disposal decision joint module, based on all previous analysis, quantitatively assesses the device risk, and generates a concrete on-site operation guidance scheme;

[0071] In the blank analysis space, call the historical accident case library, analyze the key parameter sequence that led to the fault escalation in the past; combine the expert experience rule library to sort the multiple types of data generated in stages S1 and S3, and select four core parameters: current concentration of sulfur dioxide, concentration growth rate of hydrogen sulfide, maximum amplitude of ultra-high frequency signal, and discharge activity stability index;

[0072] Define the risk contribution weight for each parameter, where the concentration growth rate of hydrogen sulfide is given the highest weight because it represents defect acceleration, the current concentration of sulfur dioxide and the maximum amplitude of ultra-high frequency signal are given medium weight because they directly reflect the current hazard intensity, and the discharge activity stability index is given basic weight as a basis for trend judgment. Thus, the parameter system construction of the risk quantification model is completed, and the parameter definition of the risk quantification model is based on the historical accident case library and expert experience rules stored in the guidance end cloud database;

[0073] Read the latest values of the four core parameters from the specified data source; according to the preset weight proportion, calculate the weighted sum of the parameter values: multiply the measured value of each parameter by the corresponding weight, and then add all the product results to finally obtain a device risk index between zero and one hundred percent; compare the calculated device risk index with the threshold interval preset in the model:

[0074] If the index is in the zero to forty percent interval, it is classified as low risk;

[0075] Forty to sixty percent is classified as medium risk;

[0076] Sixty to eighty percent is classified as high risk;

[0077] Eighty to one hundred percent is classified as emergency risk, and the final risk level identifier is generated;

[0078] The current concentration of sulfur dioxide generated by S1, the concentration growth rate of hydrogen sulfide, the maximum amplitude of the ultra-high frequency signal generated by S2, and the discharge activity stability index generated by S3 are fused into a quantitative indicator, and the risk severity is classified, providing clear basis for decision-making.

[0079] The decision tree knowledge base in the guidance end cloud database is called. The current device risk level is used as the primary matching condition, and the primary defect type determined by S3 is used as the secondary matching condition. Search is performed in the decision tree knowledge base. For example, when the condition combination is high risk level and loose shielding cover, the decision tree automatically matches the planned power-off maintenance main path, and further fills in the specific maintenance steps, required tool list and safety control measures according to the defect area information. All matching results are integrated to generate a structured on-site operation guidance scheme, which clearly defines the action priority, operation core content and resource preparation requirements. The device risk index and risk level generated by the foregoing steps, the defect area comprehensive determination report of S2, and the historical matching degree sorting list of S3; combine the quantitative risk assessment conclusion with the specific defect diagnosis information to output an action plan that can directly guide the on-site operation.

[0080] All data and conclusions of this case are packaged and stored to create an independent case tracking file. When receiving the on-site operation completion information from the user end, the system triggers a re-inspection instruction, requiring the user end to re-collect the ultra-high frequency, ultrasonic and gas composition data according to the established period. The re-inspection data is automatically compared with the historical accident case library in the case file to generate a treatment effect verification report, which clearly marks whether each abnormal signal has disappeared or decreased below the safety threshold, thereby completing the full-process closed-loop management of this risk treatment. Analyze all raw data, intermediate processing results and final on-site operation guidance scheme generated in the whole process to form a management closed loop, ensuring the effectiveness of the treatment action and accumulating data for the continuous optimization of the knowledge base.

[0081] Combining example one and example two, not only fundamentally solves the core problem of superficial guidance scheme and inefficient interactive communication, but also forms a sustainable optimization of power on-site operation intelligent support capability through data-driven and knowledge sedimentation throughout the process, significantly improves the early warning ability, treatment efficiency and safety control level of potential faults of complex power equipment, represents an innovative development direction of power on-site operation guidance interaction technology, and through the organic integration of multi-source deep diagnosis, dynamic risk assessment and intelligent decision interaction, a closed-loop intelligent interaction system integrating monitoring, analysis, decision-making and guidance is built.

[0082] The above merely illustrates and describes the structure of the present application, and those skilled in the art can make various modifications or supplements to the specific embodiments described or use similar ways to replace, as long as the modifications or supplements do not deviate from the structure of the present application or exceed the scope defined by the claims, and should belong to the protection scope of the present application.

[0083] In the description of the present specification, the description referring to the terms "one embodiment", "an example", "a specific example" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0084] The preferred embodiments of the present application disclosed above are only used to help explain the present application. The preferred embodiments do not describe all the details and limit the present application to the specific embodiments. Obviously, many modifications and changes can be made according to the content of the present specification. The present specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present application, so that those skilled in the art can well understand and utilize the present application. The present application is limited only by the claims and their full scope and equivalents.

Claims

1. A method for interactive guidance of power field operations, characterized in that, Includes the following steps; S1. Establish a two-way communication connection to obtain continuous monitoring data and abnormal operation performance of the target power equipment, construct a blank analysis space, and generate a floating potential discharge abnormal signal. S2. Construct a digital model of GIS equipment structure based on the abnormal signal of floating potential discharge, obtain the ultra-high frequency signal intensity matrix and the acoustic signal intensity matrix, and generate a comprehensive judgment report and confidence rating of defect area. S3. Construct a discharge mode feature vector based on the electrical signal pulse, stable peak value of the signal amplitude, and the sequence of signal pulse counts, and analyze it in conjunction with the historical accident case database to generate load-discharge correlation analysis conclusions; S4. The key data of multiple types are sorted to obtain four core parameters and the risk contribution weight is defined. A risk quantification model is constructed and compared to obtain the risk level identifier. The on-site operation guidance scheme is jointly constructed by combining the decision tree knowledge base.

2. The interactive method for providing on-site guidance for power operations according to claim 1, characterized in that, The hierarchical analysis process for the floating potential discharge anomaly signal in step S1 is as follows: Based on the electrical signal pulses selected along the periodic nodes of multi-type data, the signal amplitude of the electrical signal pulses is analyzed in multiple dimensions. The stable peak value, continuity, and phase distribution area of ​​the electrical signal pulses are compared along the periodic node direction. If the latest stable peak value of the signal amplitude has an abnormal peak node that exceeds the average value of several periodic nodes in reverse, or if there is a peak break, or if the electrical signal pulse has any of the following characteristics, then it is determined that the electrical signal pulse is abnormal and a partial discharge signal is generated. Based on noise and vibration data filtered along the periodic nodes of multiple types of data, if both noise and vibration individual data are found in frame-by-frame comparison, the similarity between the noise individual data and the preset audio of the pre-stored current is further analyzed. If the similarity exceeds the preset value, the noise and vibration data is deemed abnormal, an acoustic signal is generated, and the detection location is marked as the discharge point.

3. The interactive method for providing on-site guidance for power operations according to claim 2, characterized in that, Based on SF6 gas data extracted along periodic nodes from multiple types of data, the key characteristics of sulfur dioxide, hydrogen sulfide, and sulfur fluoride are compared based on the detection results. If the key characteristics are found to be present, chemical evidence of internal discharge in the power equipment and gas ionization is determined, and a gas decomposition product signal is generated. The partial discharge signal, acoustic signal, and gas decomposition product signal generated by the analysis of multiple types of data are summarized and compared with the abnormal operation performance for consistency verification. If any one or more of the same symptoms are present, the accuracy of the continuous monitoring data and the degree of consistency with the abnormal operation performance are determined, and a floating potential discharge abnormal signal is generated.

4. The interactive method for providing on-site guidance for power operations according to claim 1, characterized in that, The analysis process for the comprehensive defect area assessment report in step S2 is as follows: Based on the definition and summary of floating potential discharge anomaly signals, key sensitive words are extracted to extract features from the data covered by partial discharge signals and acoustic signals. The signal intensity and pulse number of the partial discharge signal are obtained, and the acoustic intensity and discharge point coordinates of the acoustic signal are obtained. These are then labeled as UHF signal intensity sequence, signal pulse number sequence, acoustic signal intensity sequence and discharge point coordinates, respectively. The GIS equipment structure model is digitized and a three-dimensional coordinate grid is established.

5. The interactive method for providing on-site guidance for power operations according to claim 4, characterized in that, In the three-dimensional coordinate grid, the spatial coordinates of all UHF sensors are first marked, and the possible source coordinates of the pulse signal in the three-dimensional grid are calculated using the time difference positioning algorithm. The number of signal sources calculated in each grid cell during the monitoring period is counted, and the amplitudes of these signals are accumulated. For each grid cell, the signal quantity and cumulative amplitude are combined and normalized to generate an ultra-high frequency signal strength matrix. On the outer surface of the three-dimensional coordinate mesh model, the precise position coordinates of each detection point during the on-site ultrasonic testing are marked. The values ​​of the acoustic signals identified as abnormal in S1 are assigned to the surface mesh cells where the corresponding detection points are located. For surface mesh cells without direct detection data, the nearest neighbor interpolation algorithm is used to calculate the value to fill the cell based on the strength of the mesh cells with existing data in its surrounding area. Finally, an acoustic signal intensity matrix is ​​generated.

6. The interactive method for providing on-site guidance for power operations according to claim 5, characterized in that, The two matrices are superimposed on a three-dimensional coordinate grid for analysis. The degree of matching between the two signal intensities is calculated for each grid cell to obtain the overlapping area where both show high intensity. Generate a collaborative localization confidence heatmap, in which the possible defect areas are visually marked by color depth, and the area with the highest confidence is marked as the core area of ​​the suspected defect. The data on the proportion of trace gases included in the gas decomposition product signal are correlated with the time series to generate a gas concentration time trend line, which is then superimposed on the collaborative localization confidence heatmap. It is determined whether the localization area matches the direction of the gas sampling channel and whether the signal activity period coincides with the period of rapid gas concentration growth. A comprehensive defect area judgment report is generated, which clearly indicates the 1 to 2 gas chambers or components where the defect is most likely to exist, and gives a confidence rating.

7. The interactive method for providing on-site guidance for power operations according to claim 1, characterized in that, The process of generating the load-discharge correlation analysis conclusion in step S3 is as follows: Analyze the phase distribution region, calculate the phase distribution clustering degree, the fluctuation coefficient of the stable peak value of the signal amplitude, and the time distribution uniformity of the statistical pulse number sequence, which are concentrated in the first and third quadrants of the power frequency voltage. Generate a discharge mode feature vector containing three key indicators: phase distribution clustering degree, amplitude fluctuation coefficient, and pulse uniformity. Calculate the similarity between the generated discharge mode feature vector and each historical defect feature vector in the historical accident case library to generate a historical matching degree ranking list, listing the top three most likely defect types and their matching probabilities. Based on the continuous monitoring data provided by the user terminal, a historical accident case library of line load current and the activity of partial discharge signals are extracted. A superimposed curve of the activity of partial discharge signals and the change of line load current with the direction of the cycle node is plotted, and the correlation coefficient between the two is calculated. The results of the load-discharge correlation analysis are generated.

8. The interactive method for providing on-site guidance for power operations according to claim 1, characterized in that, The process of constructing the on-site operation guidance plan in step S4 is as follows: In the blank analysis space, the historical accident case library is called to analyze the key parameter sequence that led to the escalation of the fault in the past. Combined with the expert experience rule library, the various types of data generated in the S1 and S3 stages are sorted by key, and four core parameters are selected: current concentration of sulfur dioxide, hydrogen sulfide concentration growth rate, maximum amplitude of ultra-high frequency signal, and discharge activity stability index, and assigned risk contribution weights. Based on the preset weight ratio, the parameter values ​​are weighted and summed: the measured value of each parameter is multiplied by the corresponding weight, and all the product results are added together to obtain an equipment risk index between 0 and 100. The calculated equipment risk index is compared with the preset threshold range in the model: if the index is between 0 and 40, it is classified as low risk; 40 to 60 is classified as medium risk; 60 to 80 is classified as high risk; and 80 to 100 is classified as emergency risk, and the final risk level label is generated.

9. The interactive method for providing on-site guidance for power operations according to claim 8, characterized in that, The system calls upon the decision tree knowledge base in the cloud database of the guidance terminal; it uses the current equipment risk level as the primary matching condition and the primary defect type determined by S3 as the secondary matching condition, and performs a traversal search in the decision tree knowledge base; for example, when the condition combination is high risk level and loose shielding, the decision tree automatically matches the main path of planned power outage maintenance, and further fills in the specific maintenance steps, required tool list and safety control measures according to the defect area information; it integrates all matching results to generate a structured on-site operation guidance plan.

10. A power field operation guidance interaction system, used in the power field operation guidance interaction method according to any one of claims 1-9, characterized in that, Includes the following steps: Multi-terminal data acquisition module: Based on online communication, it builds a user terminal and a guidance terminal, receives continuous monitoring data of target power equipment from the user terminal, as well as abnormal operation feedback from the power field operation of the user terminal, and combines it with the cloud database of the guidance terminal to jointly build a blank analysis space; Multi-source signal collaborative analysis module: integrates multi-source data to spatially locate defects and make preliminary judgments on their types; Root Cause Analysis Module: This module provides in-depth analysis of the physical characteristics of defects and compares them with a database of historical accident cases to accurately determine the root cause of defects. Risk assessment and response decision-making module: Based on all preliminary analyses, it quantifies equipment risks and generates concrete on-site operation guidance plans.

Images