A Method for Monitoring the State of Drop-out Fuses in Alpine Regions

Abstract

The present invention relates to the technical field of power equipment monitoring. The present invention discloses a method for monitoring the state of a drop-out fuse in alpine regions, which includes constructing a fuse distribution map, dividing the monitoring period into sub-sections, collecting the comprehensive operation data of the drop-out fuse in the sub-sections, predicting the micro-motion state value of the next sub-section based on a neural network model, marking the fault points on the fuse distribution map, and planning the maintenance path. Compared with the prior art, the present invention collects multi-dimensional data such as vibration, temperature, humidity, and current of the drop-out fuse, and combines with the neural network model to be able to predict in advance the operating state of the drop-out fuse at a future moment, that is, to issue a maintenance and replacement prompt before the drop-out fuse is about to drop, so that power maintenance personnel can perform maintenance and replacement operations on the faulty drop-out fuse, avoiding the lag of maintenance and replacement after the drop-out fuse fails.

Description

Technical Field

[0001] The present invention relates to the technical field of power equipment monitoring, and more specifically, to a method for monitoring the state of a drop-out fuse in alpine regions. Background Art

[0002] The drop-out fuse is one of the most commonly used short-circuit protection switches for the branch lines of 10kV distribution lines and distribution transformers. However, in the actual operation process of the power system, faults such as abnormal fusing of high-voltage fuses in the distribution network often occur, which seriously affects the safety and reliability of the power grid. Especially in the power systems of alpine regions, the drop-out fuse will be affected by the harsh external environment, resulting in more prone to failure phenomena. Therefore, it is necessary to accurately monitor the operating state of the drop-out fuse.

[0003] The patent application with the publication number CN107238774B discloses a state monitoring system for drop-out fuses, which uses wireless communication technology for data transmission and control, avoiding cable construction brought by traditional data transmission methods, greatly reducing the construction difficulty and system installation cost. Through telemetry and tele-signaling technology, maintenance personnel can query the operating state of the drop-out fuse in real time on the mobile phone. Not only in the hardware design, the whole machine is designed with high shielding and high sealing, having good high-temperature resistance and corrosion resistance, and using watchdog circuits and equipotential grounding and other methods to enhance its anti-interference performance and effectively prevent the system from crashing;

[0004] When the existing drop-out fuses are in state monitoring, it is necessary to manually check the states of each fuse in the area one by one to find the fuse fault point and perform subsequent operation and maintenance replacement. Since the method of real-time monitoring the state of the fuse can only analyze the state of the fuse at the current moment, when the fuse fault point is analyzed, the fuse has already had a drop-out fault phenomenon at this time, resulting in a lag in subsequent fuse inspection and replacement operations, which is not conducive to the safe and stable operation of the power system.

[0005] In view of this, the present invention proposes a method for monitoring the state of a drop-out fuse in alpine regions to solve the above problems. Summary of the Invention

[0006] In order to overcome the above defects of the prior art and to achieve the above object, the present invention provides the following technical solution: A method for monitoring the state of a drop-out fuse in alpine regions, which is applied to a power monitoring server, includes:

[0007] S1: Obtain the geographical data of the drop-out fuses in the selected area and construct a fuse distribution map. The geographical data includes the positions of the electric towers and the station numbers;

[0008] S2: Calculate the monitoring period of the drop-out fuse, and divide the monitoring period into sub-sections based on the section division criterion;

[0009] S3: Collect the comprehensive operation data of the drop-out fuse in the sub-section. The comprehensive operation data includes the ratio of over-limit temperature, the frequency of abnormal current, the over-limit value of environmental humidity, and the ratio of dangerous vibration;

[0010] S4: Input the collected comprehensive operation data into a pre-trained neural network model to predict the micro-motion state value of the next sub-section, and determine whether to issue a drop warning prompt. The micro-motion state value includes the triggered state and the un-triggered state; if a drop warning prompt is issued, execute S5; if a drop warning prompt is not issued, repeat the execution of S3 - S4;

[0011] S5: Mark the fault point in the fuse distribution map, and plan the maintenance path of the fault point based on the fault maintenance criterion.

[0012] Furthermore, the construction method of the fuse distribution map includes:

[0013] Query the position data of the drop-out fuses in the selected area one by one through the positioning system to obtain pieces of position data; Mark the interval points of

[0014] pieces of position data, and take the interval points as the splitting positions to split pieces of position data into the first data group and the second data group one by one; Identify the text part and the digital part of the first data group and the second data group one by one through natural language processing technology;

[0015] Record the digital part in the first data group as the position of the electric tower to obtain

[0016] pieces of electric tower positions, and record the digital part in the second data group as the work station number to obtain pieces of work station numbers; Mark the location points of

[0017] pieces of electric tower positions on the electronic map of the selected area one by one to obtain pieces of electric tower points;

[0018] Establish a note box at each of pieces of electric tower points, and import pieces of work station numbers into the note boxes of the electric tower points corresponding to the work station numbers one by one to obtain the breaker distribution map.

[0019] Furthermore, the calculation method of the monitoring period includes:

[0020] Mark the locations of power maintenance stations one by one on the fuse distribution map to obtain maintenance points;

[0021] Draw auxiliary lines between adjacent maintenance points to obtain auxiliary lines, and draw vertical dividing lines through the midpoints of auxiliary lines;

[0022] Extend the two ends of dividing lines respectively to divide the fuse distribution map into maintenance areas;

[0023] Query one by one through the maintenance management system the time taken for maintenance points to reach the power tower points within the same maintenance area to obtain the first time values, and after accumulating the first time values, generate the first sub-cycle;

[0024] The expression of the first sub-cycle is:

[0025] ;

[0026] In the formula, is the first sub-cycle, is the th first time value;

[0027] Query one by one through the maintenance management system the maximum time taken for drop-out fuses for maintenance and replacement to obtain the second time values, and after accumulating the second time values, generate the second sub-cycle;

[0028] The expression of the second sub-cycle is:

[0029] ;

[0030] In the formula, is the second sub-cycle, is the th second time value;

[0031] Add the first sub-cycle and the second sub-cycle to obtain the monitoring cycle;

[0032] The expression of the monitoring cycle is:

[0033] ;

[0034] In the formula, is the monitoring period.

[0035] Furthermore, the section division criterion is that the durations of any two sub - sections are the same;

[0036] The method for dividing sub - sections includes:

[0037] After adding the maximum value of the first duration value and the maximum value of the second duration value, a division duration value is generated;

[0038] The expression of the division duration value is:

[0039] ;

[0040] In the formula, is the division duration value, is the maximum value of the first duration value, is the maximum value of the second duration value;

[0041] Taking the duration corresponding to the division duration value as the division standard, the monitoring period is divided into continuous sub - sections, and sub - sections are obtained.

[0042] Furthermore, the method for obtaining the over - limit temperature occupancy ratio includes:

[0043] In sub - sections, the temperature inside the fuse tube is detected in real - time through a temperature sensor to obtain the real - time temperature value;

[0044] The real - time temperature value greater than the preset upper temperature limit value is recorded as the first abnormal value, and the continuous duration of the first abnormal value in sub - sections is counted to obtain the first abnormal duration;

[0045] The real - time temperature value less than the preset lower temperature limit value is recorded as the second abnormal value, and the continuous duration of the second abnormal value in sub - sections is counted to obtain the second abnormal duration;

[0046] After adding the first abnormal duration and the corresponding the second abnormal duration, and comparing it with the duration of the sub - section, the over - limit temperature occupancy ratio is obtained;

[0047] The expression of the over - limit temperature occupancy ratio is:

[0048] ;

[0049] In the formula, is the over - limit temperature occupancy ratio of the th sub - section, is the first abnormal duration of the th sub-segment, is the second abnormal duration of the th sub-segment.

[0050] Furthermore, the method for obtaining the abnormal current frequency includes:

[0051] A1: Taking the preset marking interval as the standard, respectively mark s current moments within the th sub-segment, and sequentially number the s current moments in ascending order according to the marking sequence;

[0052] A2: Taking the interval of one number as the first standard, record the current moment corresponding to the next number as the first moment;

[0053] A3: After the first moment, taking the interval of two numbers as the second standard, record the current moment corresponding to the next number as the second moment;

[0054] A4: Repeat the steps of A2 - A3 for r times to obtain r first moments and r second moments;

[0055] A5: At the r first moments and r second moments, detect the real-time current of the drop-out fuse through a current sensor, and record it as the detected current value;

[0056] A6: Record the detected current value greater than the standard current value as the abnormal current value, and count the number of times the abnormal current value appears within the th sub-segment to obtain the

[0057] A7; Compare each of the number of abnormal times with the duration of the sub-segment to obtain the abnormal current frequency;

[0058] The expression of the abnormal current frequency is:

[0059] ;

[0060] In the formula, is the abnormal current frequency of the th sub-segment, is the number of abnormal times of the th sub-segment.

[0061] Furthermore, the method for obtaining the environmental humidity overlimit value includes:

[0062] Within the th sub-segment, respectively detect the real-time humidity of the fuse tube at s current moments through a humidity sensor to obtain s humidity values;

[0063] Humidity values greater than a preset safe humidity value are recorded as over-limit humidity values, and over-limit humidity values are obtained;

[0064] After subtracting each of the over-limit humidity values from the safe humidity value one by one, the differences are accumulated and averaged to obtain the environmental humidity over-limit value;

[0065] The expression for the environmental humidity over-limit value is:

[0066] ;

[0067] In the formula, is the environmental humidity over-limit value of the th sub-segment, is the th over-limit humidity value of the th sub-segment, is the safe humidity value.

[0068] Furthermore, the method for obtaining the ratio of dangerous vibrations includes:

[0069] Taking the preset vibration period as the standard, each of the sub-segments is equally divided into adjacent vibration time periods;

[0070] The starting time, midpoint time, and ending time of each of the vibration time periods are respectively marked, and the vibration frequencies at the starting time, midpoint time, and ending time are respectively detected by a vibration sensor to obtain starting frequency values, midpoint frequency values, and ending frequency values;

[0071] After adding the starting frequency values, midpoint frequency values, and ending frequency values respectively and averaging them, time period frequency values are obtained;

[0072] The expression for the time period frequency value is:

[0073] ;

[0074] In the formula, is the time period frequency value of the th vibration time period of the th sub-segment, is the starting frequency value of the th vibration time period of the th sub-segment, is the The midpoint frequency value of the vibration period of the sub-segment, is the midpoint frequency value of the vibration period of the sub-segment;

[0075] Record the period frequency value greater than the upper limit value of the vibration frequency as the dangerous frequency value, and record the vibration period where the dangerous frequency value is located as the dangerous period, and obtain the number of dangerous periods;

[0076] After accumulating the durations of the dangerous periods one by one and comparing with the duration of the sub-segment, obtain the ratio of dangerous vibrations;

[0077] The expression for the ratio of dangerous vibrations is:

[0078] ;

[0079] In the formula, is the ratio of dangerous vibrations of the sub-segment, is the duration of the th dangerous period of the

[0080] Further, the training method of the neural network model includes:

[0081] Pre-collect multiple sets of comprehensive operation data and the corresponding micro-motion state values;

[0082] Convert the comprehensive operation data into multiple feature vectors using the sliding window method, convert the micro-motion state values into labels corresponding to the comprehensive operation data according to the sliding step, convert the untriggered state into 0, convert the triggered state into 1, one feature vector corresponds to one label, and form a set of training data. Multiple sets of training data form a training set. Arrange the comprehensive operation data in the order of collection time, and preset the prediction time step Z, sliding step Q, and sliding window length N;

[0083] Use the feature vectors as the input of the neural network model, use the micro-motion state value of the next sub-segment after the prediction time step Z as the output, use the subsequent micro-motion state values of each training set as the prediction target, and use the sum of minimized prediction errors as the training target to train the neural network model to generate the neural network model.

[0084] Further, the fault repair criterion is: the fault points in the same repair area are only connected once;

[0085] The method for planning the repair path includes:

[0086] Measure the distances from the maintenance points to the fault points in the same maintenance area one by one, and record the fault point corresponding to the minimum value of the distances as the starting point of the path;

[0087] Taking the starting point of the path as the starting point, connect the remaining fault points in the same maintenance area in sequence according to the laying direction of the transmission line to obtain the initial path;

[0088] According to the sequence of connection of the fault points, number the fault points on the initial path in ascending order in turn, and record the fault point corresponding to the maximum value of the numbers as the end point of the path to obtain the maintenance path.

[0089] The technical effects and advantages of a method for monitoring the state of a drop-out fuse in a high-cold region according to the present invention:

[0090] By obtaining the geographical data of the drop-out fuses in the selected area, constructing a fuse distribution map, calculating the monitoring period of the drop-out fuses, and based on the section division criterion, dividing the monitoring period into sub-sections, collecting the comprehensive operation data of the drop-out fuses in the sub-sections, inputting the collected comprehensive operation data into a pre-trained neural network model, predicting the micro-motion state value of the next sub-section, determining whether to issue a drop warning prompt, marking the fault points on the fuse distribution map, and based on the fault maintenance criterion, planning the maintenance path of the fault points; compared with the prior art, by obtaining the geographical data of the drop-out fuses, the distribution map of the drop-out fuses can be accurately constructed and used as the basis for guiding the position of subsequent maintenance and replacement of the drop-out fuses. At the same time, by collecting multi-dimensional data such as vibration, temperature, humidity and current of the drop-out fuses, and combining with the neural network model, the operation state of the drop-out fuses at future moments can be predicted in advance, that is, a maintenance and replacement prompt can be issued before the drop-out fuses are about to drop, and a reasonable maintenance path can be planned, so that the power maintenance personnel can quickly perform maintenance and replacement operations on the drop-out fuses with fault phenomena, avoiding the lag caused by the maintenance and replacement after the drop-out fuses have dropped, and thus ensuring the safe and stable operation state of the power supply system in the high-cold region. Description of the Drawings

[0091] Figure 1 It is a schematic flow chart of a method for monitoring the state of a drop-out fuse in a high-cold region provided by Embodiment 1 of the present invention;

[0092] Figure 2 It is a schematic diagram of a system for monitoring the state of a drop-out fuse in a high-cold region provided by Embodiment 2 of the present invention. Detailed Embodiments

[0093] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0094] Embodiment 1: Please refer to Figure 1 As shown in the figure, a method for monitoring the state of a drop-out fuse in an alpine region, which is applied to a power monitoring server, includes:

[0095] S1: Obtain the geographical data of the drop-out fuses in the selected area and construct a fuse distribution map;

[0096] Geographical data refers to the data that can accurately represent the location of the drop-out fuse in the selected area, that is, the location of each drop-out fuse in the selected area can be located, and an accurate data basis is provided for the subsequent construction of the fuse distribution map to ensure that the fuse distribution map can locate the drop-out fuses in the selected area;

[0097] The geographical data includes the position of the power tower and the work station number; the position of the power tower refers to the specific geographical location of the power tower where the drop-out fuse is located, which can overall locate the power tower position of the drop-out fuse, and the work station number refers to the number of the specific line where the drop-out fuse is located on the power tower, which can finely locate the line position of the drop-out fuse;

[0098] The construction method of the fuse distribution map includes:

[0099] Through the positioning system, query the position data of each drop-out fuse in the selected area one by one to obtain pieces of position data, and obtain pieces of position data;

[0100] Mark the interval points of pieces of position data, and take the interval points as the splitting positions to split pieces of position data into the first data group and the second data group one by one; the interval point refers to the point used to distinguish different types of data in the position data, that is, the power tower data and the work station data in the position data can be effectively distinguished, so as to ensure the accuracy of the subsequent splitting of the first data group and the second data group;

[0101] Through natural language processing technology, identify the text part and the digital part of the first data group and the second data group one by one;

[0102] Record the digital part in the first data group as the power tower position to obtain pieces of power tower positions, and record the digital part in the second data group as the work station number to obtain Station numbers;

[0103] Mark the locations of electric towers one by one on the electronic map of the selected area to obtain the locations of the electric towers;

[0104] At the locations of the electric towers, establish remark boxes respectively, and import the station numbers one by one into the remark boxes corresponding to the electric tower locations to obtain the circuit breaker distribution map. The remark box refers to the text box used to summarize and bind the station numbers at the electric tower locations, so that the corresponding relationship can be maintained between the electric tower locations and the station numbers.

[0105] It should be noted that since there are multiple transmission lines on the same electric tower, and a drop fuse can be deployed on each transmission line, and the station number is a specific representation of the line where the drop fuse is located, and the electric tower location is a representation of the location of the electric tower where the drop fuse is located, therefore, the number of electric tower locations will be less than the number of station numbers, that is less than , and the constructed fuse distribution map can not only locate the location of the electric tower, but also represent the number of the line where each drop fuse is located, so as to ensure that the specific location of each drop fuse can be accurately and independently located and marked subsequently.

[0106] S2: Calculate the monitoring period of the drop fuse, and divide the monitoring period into sub-sections based on the section division criterion;

[0107] The monitoring period refers to the overall duration for the power maintenance station in the selected area to reach the electric tower location where the corresponding drop fuse is located and complete the maintenance operation of the drop fuse, so as to ensure that the power maintenance station can effectively maintain and process the corresponding drop fuse in the selected area;

[0108] The calculation method of the monitoring period includes:

[0109] Mark the locations of the power maintenance stations one by one on the fuse distribution map to obtain the maintenance locations;

[0110] Draw auxiliary lines between adjacent two maintenance locations to obtain the number of auxiliary lines, and draw the number of dividing lines vertically along the midpoints of the auxiliary lines;

[0111] Extend the two ends of the dividing lines respectively to divide the fuse distribution map into An overhaul area; By dividing the fuse distribution map with a dividing line, it can ensure that each overhaul area contains an independent power overhaul station, and at the same time, it can also divide the positions of power towers in different locations into the corresponding overhaul areas, so as to facilitate the subsequent overhaul and replacement of the faulty drop fuses;

[0112] Through the overhaul management system, query one by one the time taken for an overhaul point to reach the power tower points in the same overhaul area, and obtain a first duration value, and after accumulating the first duration values, generate a first sub-cycle;

[0113] The expression of the first sub-cycle is:

[0114] ;

[0115] In the formula, is the first sub-cycle, is the th first duration value;

[0116] Through the overhaul management system, query one by one the maximum time taken for overhaul and replacement of a drop fuse, and obtain a second duration value, and after accumulating the second duration values, generate a second sub-cycle;

[0117] The expression of the second sub-cycle is:

[0118] ;

[0119] In the formula, is the second sub-cycle, is the th second duration value;

[0120] After adding the first sub-cycle and the second sub-cycle, obtain the monitoring cycle;

[0121] The expression of the monitoring cycle is:

[0122] ;

[0123] In the formula, is the monitoring cycle.

[0124] After the monitoring period is generated, since the duration corresponding to the monitoring period is the duration for replacing all the drop-out fuses in the selected area, the duration span of the monitoring period is relatively large, and the amount of data corresponding to the change in the operating state of the drop-out fuses within the monitoring period is also excessive. In order to achieve a more detailed monitoring effect of the drop-out fuse state in a short time, it is necessary to divide the monitoring period with too long a duration span into sub-sections with a shorter duration span;

[0125] When dividing the monitoring period into sub-sections, it is necessary to be restricted by the section division criterion to ensure that each sub-section after division can maintain consistency and continuity;

[0126] The section division criterion is: the durations of any two sub-sections are the same; this can ensure that two adjacent sub-sections are in a continuous state on the time line, enabling the sub-sections to perform a complete and continuous division process on the monitoring period;

[0127] The division method of the sub-sections includes:

[0128] After adding the maximum value of the first duration value and the maximum value of the second duration value, a division duration value is generated;

[0129] The expression of the division duration value is:

[0130] ;

[0131] In the formula, is the division duration value, is the maximum value of the first duration value, is the maximum value of the second duration value;

[0132] Taking the duration corresponding to the division duration value as the division standard, the monitoring period is divided into continuous sub-sections, obtaining sub-sections.

[0133] When the monitoring period is divided into sub-sections, it can provide an accurate duration basis for the subsequent data collection of the operating state of the drop-out fuses, enabling each sub-section to represent the operating state of the drop-out fuses separately.

[0134] S3: Collect the comprehensive operating data of the drop-out fuses in the sub-sections. The comprehensive operating data includes the ratio of over-limit temperature, the frequency of abnormal current, the over-limit value of environmental humidity, and the ratio of dangerous vibration;

[0135] The comprehensive operating data refers to the data that can represent the change in the operating state of the drop-out fuses within the sub-sections, and thus can be used as the data basis for predicting the state quality of the drop-out fuses subsequently;

[0136] The comprehensive operation data includes the over-limit temperature occupancy ratio, abnormal current frequency, environmental humidity over-limit value, and dangerous vibration occupancy ratio;

[0137] The over-limit temperature occupancy ratio refers to the ratio of the duration during which the real-time temperature in the fuse tube of the drop-out fuse within a sub-section is outside the preset temperature range to the total duration, which can numerically represent the over-limit duration of the temperature in the fuse tube. When the over-limit temperature occupancy ratio is larger, it indicates that the ratio of the duration during which the real-time temperature in the fuse tube of the drop-out fuse within the sub-section is outside the preset temperature range to the total duration is larger. At this time, the operating state of the drop-out fuse is worse, and the probability of dropping is higher;

[0138] The method for obtaining the over-limit temperature occupancy ratio includes:

[0139] Within a sub-section, the temperature in the fuse tube is detected in real time through a temperature sensor to obtain the real-time temperature value;

[0140] The real-time temperature value greater than the preset upper temperature limit value is recorded as the first abnormal value, and the duration of the first abnormal value within the sub-section is statistically counted to obtain the first abnormal duration; the preset upper temperature limit value refers to the maximum temperature in the fuse tube when the drop-out fuse is within the normal range, which can distinguish between too high temperature and normal temperature, so as to accurately identify the too high temperature data;

[0141] The real-time temperature value less than the preset lower temperature limit value is recorded as the second abnormal value, and the duration of the second abnormal value within the sub-section is statistically counted to obtain the second abnormal duration; the preset lower temperature limit value refers to the minimum temperature in the fuse tube when the drop-out fuse is within the normal range, which can distinguish between too low temperature and normal temperature, so as to accurately identify the too low temperature data;

[0142] Add the first abnormal duration and the corresponding second abnormal duration, and compare it with the duration of the sub-section to obtain the over-limit temperature occupancy ratio;

[0143] The expression of the over-limit temperature occupancy ratio is:

[0144] ;

[0145] In the formula, is the over-limit temperature occupancy ratio of the th sub-section, is the first abnormal duration of the th sub-section, is the The second abnormal duration of the sub-section.

[0146] The abnormal current frequency refers to the number of times of abnormal current in the fuse tube of the drop-out fuse in the sub-section within a unit time, that is, it can represent the number of times of abnormal current in the fuse tube. When the abnormal current frequency is larger, it means that the number of times of abnormal current in the fuse tube of the drop-out fuse in the sub-section within a unit time is more. At this time, the operating state of the drop-out fuse is worse and the probability of dropping is higher.

[0147] The methods for obtaining the abnormal current frequency include:

[0148] A1: Taking the preset marking interval as the standard, respectively mark s current moments in the sub-section, and sequentially number the s current moments in ascending order according to the marking sequence; the preset marking interval refers to the interval duration between two adjacent current detection moments, that is, it can independently distinguish each current detection moment, preventing the phenomenon of consistent or repeated current data when detecting current at two adjacent current detection moments.

[0149] A2: Taking the interval of one number as the first standard, record the current moment corresponding to the next number as the first moment;

[0150] A3: After the first moment, taking the interval of two numbers as the second standard, record the current moment corresponding to the next number as the second moment;

[0151] A4: Repeat the steps of A2 - A3 for r times to obtain r first moments and r second moments;

[0152] A5: At the r first moments and r second moments, detect the real-time current of the drop-out fuse through the current sensor, and record it as the detected current value;

[0153] A6: Record the detected current value greater than the standard current value as the abnormal current value, and count the number of times of abnormal current value in the sub-section to obtain the number of abnormal times; the standard current value refers to the maximum value of the temperature when the drop-out fuse is in a normal condition, that is, it can accurately represent the excessive temperature data, so as to accurately identify the phenomenon of abnormal current.

[0154] A7; Compare each number of abnormal times with the duration of the sub-section to obtain the abnormal current frequency;

[0155] The expression of the abnormal current frequency is:

[0156] ;

[0157] In the formula, is the abnormal current frequency of the th sub-section, is the number of abnormalities in the th sub-section.

[0158] The environmental humidity over-limit value refers to the difference between the humidity of the fuse tube of the drop-out fuse in the sub-section and the safe humidity, which can represent the humidity of the environment where the fuse tube is located. When the environmental humidity over-limit value is larger, it indicates that the difference between the humidity of the environment where the fuse tube of the drop-out fuse in the sub-section and the safe humidity is larger. At this time, the operating state of the drop-out fuse is worse, and the probability of dropping is higher;

[0159] The method for obtaining the environmental humidity over-limit value includes:

[0160] In the th sub-section, the real-time humidity of the fuse tube is detected by a humidity sensor at s current moments respectively, and s humidity values are obtained;

[0161] The humidity values greater than the preset safe humidity value are recorded as over-limit humidity values, and over-limit humidity values are obtained; the preset safe humidity value refers to the maximum humidity of the external environment where the fuse tube of the drop-out fuse is in the normal range, which can distinguish between excessive humidity and normal humidity, so as to accurately identify excessive humidity data;

[0162] After subtracting each of the over-limit humidity values from the safe humidity value one by one, the differences are accumulated and averaged to obtain the environmental humidity over-limit value;

[0163] The expression of the environmental humidity over-limit value is:

[0164] ;

[0165] In the formula, is the environmental humidity over-limit value of the th sub-section, is the th over-limit humidity value of the th sub-section, is the safe humidity value.

[0166] The dangerous vibration occupancy ratio refers to the ratio of the duration of the fuse tube of the drop-out fuse in the sub-section at the dangerous vibration frequency to the total duration, which can represent the dangerous vibration phenomenon of the fuse tube. When the dangerous vibration occupancy ratio is larger, it indicates that the ratio of the duration of the fuse tube of the drop-out fuse in the sub-section at the dangerous vibration frequency to the total duration is larger. At this time, the operating state of the drop-out fuse is worse, and the probability of dropping is higher;

[0167] The method for obtaining the ratio of dangerous vibrations includes:

[0168] Taking a preset vibration period as the standard, each of the sub - sections is equally divided into

[0169] adjacent vibration time periods; the preset vibration period is the time interval marking two adjacent vibration moments, which can ensure the consistency of the time interval between two adjacent vibration moments, thus providing accurate time points for subsequent vibration frequency detection; Mark the starting moment, mid - point moment, and ending moment of each vibration time period respectively, and detect the vibration frequencies at the starting moment, mid - point moment, and ending moment through a vibration sensor to obtain starting frequency values, mid - point frequency values, and

[0170] ending frequency values; through the frequencies corresponding to the starting moment, mid - point moment, and ending moment, the frequency of each vibration time period can be represented in the whole stage of the early, middle, and late periods, improving the comprehensiveness of vibration data; Add up the starting frequency values, mid - point frequency values, and ending frequency values respectively and then take the average to obtain

[0171] the time - period frequency value;

[0172] ;

[0173] In the formula, is the time - period frequency value of the th vibration time period of the th sub - section, is the starting frequency value of the th sub - section and the th vibration time period, is the mid - point frequency value of the th sub - section and the th vibration time period, is the ending frequency value of the th sub - section and the th vibration time period;

[0174] Record the time - period frequency values greater than the vibration frequency upper limit value as dangerous frequency values, and record the vibration time periods where the dangerous frequency values are located as dangerous time periods to obtain The upper limit of vibration frequency refers to the maximum value of the vibration frequency of the fuse tube of the drop-out fuse in the normal range, which can distinguish between excessive vibration frequency and normal vibration frequency;

[0175] Will The duration of each dangerous period is accumulated one by one and compared with the duration of the sub-segment to obtain The percentage of dangerous vibrations;

[0176] The expression of dangerous vibration ratio is:

[0177] ;

[0178] In the formula, For the The percentage of dangerous vibration in each sub-segment, For the The sub-segment The length of the dangerous period.

[0179] S4: input the collected comprehensive operation data into the neural network model trained in advance, predict the micro-motion state value of the next sub-section, and determine whether to issue a fall warning prompt;

[0180] The micro-motion state value refers to the stroke state of the micro-switch in the drop-out fuse, which can indicate whether the drop-out fuse triggers the micro-switch action. The micro-motion state value includes a triggered state and an untriggered state. The triggered state indicates that the micro-switch in the drop-out fuse has a stroke and generates a drop-out warning signal. The untriggered state indicates that the micro-switch in the drop-out fuse has no stroke and generates no drop-out warning signal. The micro-motion state value is obtained by monitoring the warning signal of the micro-switch.

[0181] After obtaining the comprehensive operation data, the neural network model can be used to predict the comprehensive operation data, so that the neural network model can accurately predict the micro-motion state value of the next sub-section, thereby realizing the prediction effect of the micro-motion state value of the drop-out fuse at the future moment, ensuring that the power monitoring platform can be advanced and accurate, so as to achieve the effect of early prediction of the drop-out fuse state;

[0182] The training methods of neural network models include:

[0183] Pre-collecting multiple sets of comprehensive operation data and micro-motion state values ​​corresponding to the comprehensive operation data;

[0184] Convert the comprehensive operation data into multiple feature vectors using the sliding window method. Convert the micro-motion state values into labels corresponding to the comprehensive operation data according to the sliding step. Convert the untriggered state to 0 and the triggered state to 1. One feature vector corresponds to one label, and a set of training data is formed. Multiple sets of training data constitute a training set. Arrange the comprehensive operation data in the order of the acquisition time. Preset the prediction time step Z, the sliding step Q, and the sliding window length N.

[0185] Use the feature vectors as the input of the neural network model, and use the micro-motion state value of the next sub-section after the prediction time step Z as the output. Use the subsequent micro-motion state values of each training set as the prediction target, and use the sum of the minimized prediction errors as the training target to train the neural network model, and generate a neural network model that predicts the micro-motion state value of the next sub-section based on the comprehensive operation data of the previous sub-section.

[0186] Exemplarily, the neural network model is any one of CNN or AlexNet;

[0187] The calculation formula for the prediction error is:

[0188] ;

[0189] In the formula, is the prediction error, is the group number of the feature vectors; is the predicted state value corresponding to the th group of feature vectors, is the actual state value corresponding to the

[0190] th group of training data;

[0191] The drop warning prompt is a warning prompt issued when the drop fuse is about to enter the drop state at a future time, so as to provide early warning information for the power monitoring platform and perform subsequent processing operations on the drop fuse that issues the drop warning information;

[0192] The determination method for whether to issue a drop warning prompt includes:

[0193] When the output of the neural network model is 0, it means that the predicted micro-motion state value of the next sub-section is the untriggered state. At this time, the probability of the drop fuse dropping in the future is low, and no warning prompt is required. It is determined that no drop warning prompt is issued;

[0194] When the output of the neural network model is 1, it indicates that the predicted micro-motion state value of the next sub-section is the trigger state. At this time, the probability of the drop fuse falling in the future is high, and a warning prompt is required. Then, it is determined to issue a drop warning prompt.

[0195] S5: Mark the fault points in the fuse distribution map, and based on the fault repair criteria, plan the repair path for the fault points;

[0196] The fault point refers to the tower point where the drop fuse is located when the predicted micro-motion state value of the next sub-section by the neural network model is the trigger state, which can be used as the point for subsequent repair or replacement of the drop fuse that issues a drop warning prompt, so as to achieve the effect of early repair or replacement of the drop fuse with a relatively poor operating state;

[0197] When marking the fault points, it is necessary to obtain the station number of the drop fuse that issues a drop warning prompt and identify the note box where the work number is located, so that the tower point corresponding to the note box is the fault point;

[0198] It should be noted that the number of drop fuses that issue a drop warning prompt may be one or more, and the corresponding fault points of the drop fuses that issue a drop warning prompt may be one or more. When there are multiple fault points, it is necessary to arrange the multiple fault points in an orderly manner to facilitate subsequent repair. To ensure the rapidity and rationality of multiple fault points during repair, it is necessary to plan the route under the restriction of the fault repair criteria, so as to be able to plan the repair path, and the repair path can be used as the repair walking route for the drop fuses at the subsequent fault points;

[0199] The fault repair criteria are: the fault points in the same repair area are only connected once; it can ensure that the fault points in the same repair area can be connected in sequence, and ensure that the fault points in the same repair area can complete the repair and replacement operations in the shortest time;

[0200] The method for planning the repair path includes:

[0201] Measure the distances between the repair points and the fault points in the same repair area one by one, and record the fault point corresponding to the minimum distance as the starting point of the path;

[0202] Taking the starting point of the path as the starting point, in accordance with the laying direction of the transmission line, connect the remaining fault points in the same maintenance area in sequence to obtain the initial path; the method of connecting points along the laying direction of the transmission line can not only ensure the orderliness of the connection between each fault point, without the phenomenon of chaotic connection order, but also ensure that each fault point will only be connected once, without the phenomenon of repeated connection;

[0203] According to the sequence of connection of the fault points, sequentially number the fault points on the initial path in ascending order, and record the fault point corresponding to the maximum value of the number as the end point of the path to obtain the maintenance path.

[0204] It should be noted that since there is a power maintenance station corresponding to each maintenance area, when multiple fault points are distributed in different maintenance paths, only the maintenance paths in each maintenance area need to be planned separately, so as to ensure that each power maintenance station can perform maintenance on the fault location in its respective maintenance area, and ultimately achieve the maintenance effect of all fault points.

[0205] In this embodiment, by obtaining the geographical data of the drop fuses in the selected area, constructing a fuse distribution map, calculating the monitoring period of the drop fuses, and based on the section division criterion, dividing the monitoring period into sub-sections, collecting the comprehensive operation data of the drop fuses in the sub-sections, inputting the collected comprehensive operation data into a pre-trained neural network model, predicting the micro motion state value of the next sub-section, determining whether to issue a drop warning prompt, marking the fault points on the fuse distribution map, and based on the fault maintenance criterion, planning the maintenance path of the fault points; compared with the prior art, by obtaining the geographical data of the drop fuses, the distribution map of the drop fuses can be accurately constructed and used as the basis for guiding the position of subsequent maintenance and replacement of the drop fuses. At the same time, by collecting multi-dimensional data such as vibration, temperature, humidity and current of the drop fuses, and combining with the neural network model, the operation state of the drop fuses at future moments can be predicted in advance, so that a maintenance and replacement prompt can be issued before the drop phenomenon of the drop fuses occurs, and a reasonable maintenance path can be planned, so that the power maintenance personnel can quickly perform maintenance and replacement operations on the drop fuses with fault phenomena, avoiding the lag caused by the maintenance and replacement after the drop fault of the drop fuses, and thus ensuring the safe and stable operation state of the power supply system in alpine regions.

[0206] Embodiment 2: Please refer to Figure 2As shown, for the parts not described in detail in this embodiment, refer to the description of Embodiment 1. A state monitoring system for drop fuses in alpine regions is provided, which is applied to a power monitoring server and is used to implement a method for monitoring the state of drop fuses in alpine regions. It includes a distribution map construction module, a sub-section division module, a data acquisition module, a model prediction module, and a path planning module. Among them, each module is connected by a wired or wireless network;

[0207] The distribution map construction module is used to obtain the geographical data of the drop fuses within the selected area and construct a fuse distribution map. The geographical data includes the positions of power towers and station numbers;

[0208] The sub-section division module is used to calculate the monitoring period of the drop fuses and divide the monitoring period into sub-sections based on the section division criteria;

[0209] The data acquisition module is used to collect the comprehensive operation data of the drop fuses in the sub-sections. The comprehensive operation data includes the over-limit temperature occupancy ratio, abnormal current frequency, environmental humidity over-limit value, and dangerous vibration occupancy ratio;

[0210] The model prediction module is used to input the collected comprehensive operation data into a pre-trained neural network model, predict the micro-motion state value of the next sub-section, and determine whether to issue a drop warning prompt. The micro-motion state value includes a triggered state and an un-triggered state;

[0211] The path planning module is used to mark the fault points in the fuse distribution map and plan the maintenance path of the fault points based on the fault maintenance criteria.

[0212] As mentioned above, the above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention.

Claims

1. A method for monitoring the status of drop-out fuses in high-cold areas, applied to a power monitoring server, characterized in that: include: S1: Obtain geographic data of drop-out fuses in a selected area and construct a fuse distribution map. The geographic data includes tower locations and station numbers. S2: Calculate the monitoring period of the drop-out fuse and divide the monitoring period into sub-segments based on the segment division criterion; The calculation method of the monitoring period includes: Mark each fuse on the fuse distribution diagram. The location of the power maintenance station is obtained Maintenance points; Draw auxiliary lines between two adjacent maintenance points to obtain Auxiliary lines, respectively The midpoint of the auxiliary line is drawn in the vertical direction dividing line; Respectively The two ends of the dividing line extend outward, dividing the fuse distribution diagram into maintenance area; Through the maintenance management system, we can query The time it takes for each maintenance point to reach the tower point in the same maintenance area is obtained. The first duration value, and After the first duration values ​​are accumulated, a first sub-period is generated; Through the maintenance management system, we can query The maximum time for repairing and replacing a drop-out fuse is obtained. The second duration value, and After the second duration values ​​are accumulated, a second sub-period is generated; After adding the first sub-period and the second sub-period, a monitoring period is obtained; The criteria for segment division are: the duration of any two sub-segments are the same; The sub-segment division methods include: Adding the maximum value of the first duration value and the maximum value of the second duration value to generate a divided duration value; The duration corresponding to the division duration value is used as the division standard to divide the monitoring period into continuous sub-segments to obtain sub-segments; S3: Collect comprehensive operation data of the drop-out fuse in the sub-section, including the percentage of over-limit temperature, abnormal current frequency, over-limit value of ambient humidity and percentage of dangerous vibration; S4: Input the collected comprehensive operation data into the pre-trained neural network model, predict the micro-motion state value of the next sub-section, and determine whether to issue a fall warning prompt. The micro-motion state value includes a triggered state and a non-triggered state; if a fall warning prompt is issued, execute S5; if no fall warning prompt is issued, repeat S3-S4; S5: Mark the fault point in the fuse distribution map, and plan the maintenance path of the fault point based on the fault maintenance criteria.

2. The method for monitoring the status of drop-out fuses in high-cold areas according to claim 1 is characterized in that: The method for constructing the fuse distribution map includes: Through the positioning system, we can query the selected areas one by one. The position data of the drop-out fuse is obtained location data; Mark out The interval points of the position data are used as the splitting positions. The position data are split into a first data group and a second data group one by one; Using natural language processing technology, the text part and the number part of the first data group and the second data group are identified one by one; The digital part in the first data group is recorded as the tower position, and the The number in the second data group is recorded as the station number, and the Workstation number; Mark the selected areas one by one on the electronic map The location of the tower, get Tower locations; exist Create a remark box at each tower point and The station numbers are imported one by one into the remarks box of the tower point corresponding to the station number to obtain the circuit breaker distribution map.

3. A method for monitoring the status of drop-out fuses in high-cold areas according to claim 2, characterized in that: The expression of the first sub-period is: ； In the formula, is the first sub-period, For the The first duration value; The expression for the second sub-period is: ； In the formula, is the second sub-period, For the Second duration value; The expression of the monitoring period is: ； In the formula, For the monitoring cycle.

4. A method for monitoring the status of drop-out fuses in high-cold areas according to claim 3, characterized in that: The expression of the partition duration value is: ； In the formula, To divide the duration value, is the maximum value of the first duration value, The maximum value of the second duration.

5. A method for monitoring the status of drop-out fuses in high-cold areas according to claim 4, characterized in that: The method for obtaining the over-limit temperature ratio value includes: exist In each sub-section, the temperature in the fuse tube is detected in real time by the temperature sensor to obtain the real-time temperature value; The real-time temperature value greater than the preset temperature upper limit is recorded as the first abnormal value, and the The duration of the first abnormal value in the sub-segment is obtained The first abnormal duration; The real-time temperature value that is less than the preset lower limit of temperature is recorded as the second abnormal value, and the The duration of the second abnormal value in the sub-segment is obtained The second abnormal duration; Will The first abnormal duration and the corresponding After adding the second abnormal durations and comparing them with the duration of the sub-segment, we get The percentage of over-limit temperature; The expression of the over-limit temperature ratio is: ； In the formula, For the The percentage of over-limit temperature in each sub-section, For the The first abnormal duration of the sub-segment, For the The second abnormal duration of the sub-segment.

6. A method for monitoring the status of drop-out fuses in high-cold areas according to claim 5, characterized in that: The method for obtaining the abnormal current frequency includes: A1: Based on the preset marking interval, Mark s current moments in each sub-segment, and number the s current moments in ascending order according to the order of marking; A2: Taking the interval of one number as the first standard, the current moment corresponding to the next number is recorded as the first moment; A3: After the first moment, the current moment corresponding to the next number is recorded as the second moment, with the interval of two numbers as the second standard; A4: Repeat steps A2-A3 r times to obtain r first moments and r second moments; A5: At r first moments and r second moments, the real-time current of the drop-out fuse is detected by the current sensor, which is recorded as the detected current value; A6: Record the detected current value greater than the standard current value as an abnormal current value and count it The number of abnormal current values ​​in each sub-segment is obtained. Number of abnormalities; A7; The number of abnormalities is compared with the duration of the sub-segment one by one to obtain Abnormal current frequency; The expression of abnormal current frequency is: ； In the formula, For the The abnormal current frequency of each sub-section, For the The number of anomalies in each sub-segment.

7. A method for monitoring the status of drop-out fuses in high-cold areas according to claim 6, characterized in that: The method for obtaining the excess value of ambient humidity comprises: exist In each sub-section, the real-time humidity of the fuse tube at s current moments is detected by the humidity sensor to obtain s humidity values; The humidity value greater than the preset safe humidity value is recorded as the excess humidity value. Exceeding humidity value; Will After subtracting each excess humidity value from the safe humidity value one by one, the differences are accumulated and averaged to obtain the excess humidity value. The expression of the ambient humidity exceeding the limit is: ； In the formula, For the The ambient humidity of each sub-segment exceeds the limit value. For the The sub-segment Exceeding humidity value, It is a safe humidity value.

8. A method for monitoring the status of drop-out fuses in high-cold areas according to claim 7, characterized in that: The method for obtaining the dangerous vibration ratio value includes: Based on the preset vibration period, The sub-segments are divided into adjacent vibration periods; Mark out The starting time, midpoint time and end time of each vibration period are measured, and the vibration frequency at the starting time, midpoint time and end time is detected by a vibration sensor to obtain The starting frequency value, The midpoint frequency value and End point frequency value; Will The starting frequency value, The midpoint frequency value and The end frequency values ​​are added and averaged to obtain Frequency value of each period; The expression for the period frequency value is: ； In the formula, For the The sub-segment The period frequency value of each vibration period, For the The sub-segment The starting frequency value of each vibration period, For the The sub-segment The midpoint frequency value of a vibration period, For the The sub-segment The end frequency value of each vibration period; The frequency value of the period greater than the upper limit of the vibration frequency is recorded as the dangerous frequency value, and the vibration period where the dangerous frequency value is located is recorded as the dangerous period. Dangerous period; Will The duration of each dangerous period is accumulated one by one and compared with the duration of the sub-segment to obtain The percentage of dangerous vibrations; The expression of dangerous vibration ratio is: ； In the formula, For the The percentage of dangerous vibration in each sub-segment, For the The sub-segment The length of the dangerous period.

9. A method for monitoring the status of drop-out fuses in high-cold regions according to claim 8, characterized in that: The training method of the neural network model includes: Pre-collecting multiple sets of comprehensive operation data and micro-motion state values ​​corresponding to the comprehensive operation data; The comprehensive operation data is converted into multiple feature vectors using a sliding window method. The micro-motion state value is converted into a label corresponding to the comprehensive operation data according to the sliding step size. The untriggered state is converted into 0, and the triggered state is converted into 1. One feature vector corresponds to one label and constitutes a set of training data. Multiple sets of training data constitute a training set. The comprehensive operation data are arranged in the order of collection time, and the prediction time step size Z, sliding step size Q and sliding window length N are preset. The feature vector is used as the input of the neural network model, the micromotion state value of the next sub-segment after the predicted time step Z is used as the output, the subsequent micromotion state value of each training set is used as the prediction target, and the minimized sum of prediction errors is used as the training target to train the neural network model to generate a neural network model.

10. A method for monitoring the status of drop-out fuses in high-cold regions according to claim 9, characterized in that: The fault repair criteria are: the fault points in the same repair area are connected only once; The planning methods of maintenance path include: The distances between the maintenance points and the fault points in the same maintenance area are measured one by one, and the fault point corresponding to the minimum distance is recorded as the starting point of the path; Taking the path start point as the starting point, the remaining fault points in the same maintenance area are connected in sequence according to the laying direction of the transmission line to obtain the initial path; According to the order of the fault point connection, the fault points on the initial path are numbered in ascending order, and the fault point corresponding to the maximum number is recorded as the path end point to obtain the maintenance path.

Images