CN115392380A - Method and device for identifying abnormal data of partial discharge online monitoring and storage medium - Google Patents
Method and device for identifying abnormal data of partial discharge online monitoring and storage medium Download PDFInfo
- Publication number
- CN115392380A CN115392380A CN202211044053.4A CN202211044053A CN115392380A CN 115392380 A CN115392380 A CN 115392380A CN 202211044053 A CN202211044053 A CN 202211044053A CN 115392380 A CN115392380 A CN 115392380A
- Authority
- CN
- China
- Prior art keywords
- data
- abnormal data
- abnormal
- partial discharge
- tree
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
- G01R31/1209—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing using acoustic measurements
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
- G01R31/1227—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials
Abstract
The invention provides a method, a device and a storage medium for recognizing and alarming abnormality in partial discharge online monitoring, which relate to the field of partial discharge online monitoring. The identification detection model has large processing capacity and high speed, has the capability of quickly and accurately distinguishing the abnormal partial discharge data, can accurately position the time period of abnormal data according to the data acquisition time, and quickly eliminates interference; on the basis, the service life of a battery of the warning device is greatly prolonged by combining an optimized warning strategy. The invention can effectively reduce the expenditure of manpower and material resources, improve the economic benefit and the service quality, increase the management level of the power grid and improve the safe operation capability of the power grid.
Description
Technical Field
The invention relates to the field of partial discharge online monitoring, in particular to a method and a device for identifying partial discharge online monitoring abnormal data and a storage medium.
Background
Partial discharge is a key characteristic for measuring the insulation quality of equipment, and if the equipment generates partial discharge, certain loss and damage can be caused, so that the equipment needs to be monitored on line by partial discharge. In daily monitoring, due to internal defects of equipment or interference of external condition factors, abnormal data are easily generated when the local discharge monitoring is carried out by using an ultrasonic method and a transient earth voltage detection method, and the abnormal data can influence the accuracy of state evaluation of monitoring equipment. When abnormal data is detected, frequent alarms ultimately lead to a significant reduction in battery life.
Disclosure of Invention
In view of the above problems in the prior art, the present invention provides a method, an apparatus, and a storage medium for identifying abnormal data of partial discharge online monitoring, so as to identify and alarm the abnormal data of partial discharge online monitoring.
In order to achieve the above object, on one hand, a method for identifying and alarming an abnormal condition in an online partial discharge monitoring is provided, and the method includes:
s1, constructing a model for detecting abnormity of monitoring data based on an isolated forest algorithm, and carrying out abnormity detection on partial discharge online monitoring data by using the model to identify abnormal data, wherein the monitoring data comprises earth electric wave data or ultrasonic wave data;
s2, judging the abnormal data by using a fixed value discrimination method to obtain effective abnormal data, and the method comprises the following steps:
s21, detecting a background value of the ground electric wave or the ultrasonic wave of the switch cabinet;
s22, obtaining a difference value p between the background value and each abnormal data;
s23, when p meets a preset condition, determining the corresponding abnormal data as effective abnormal data;
and S3, analyzing the effective abnormal data, and giving an alarm according to the analysis result and a preset alarm strategy.
Further, in step S1, the step of constructing the model includes:
s11, collecting the monitoring data of the switch cabinet as a training data set, and randomly selecting a subset x with the size of n from the training data set as a root node of a random binary tree;
s12, randomly appointing the dimensionality of the current node, and determining a cutting point q between the minimum value and the maximum value of the dimensionality;
s13, according to the cutting point q, dividing the data of which the data attribute value is smaller than the characteristic numerical value q in the node into left sub-tree nodes, and dividing the data of the node divided into the left sub-trees into right sub-tree nodes;
s14, constructing new nodes for recursion steps S12-S13 of each left sub-tree node and each right sub-tree node, and ending recursion when the current isolated tree reaches a preset height or leaf nodes cannot be cut continuously to obtain a complete isolated tree;
and S15, repeating the steps S12-S14 to obtain a preset number of isolated trees to form an isolated forest.
Further, in step S1, the step of identifying abnormal data includes:
s16, calculating the average length of the subset x in the isolated tree;
and S17, calculating an abnormal index according to the path length, and identifying abnormal data.
Further, step S16 includes:
calculating the average path length C (n) of the subset x for normalizing the data x in the subset x i Normalized E (h (x)) and path length of (d) i ) Wherein, h (x) i ) From root node to leaf node x i The path length of E (h (x)) i ) Is the node x i The expected value of (d);
the calculation formula of C (n) is as follows:
wherein H (i) = Ln (i) + γ, i ∈ (1, 2,3, \8230;, n), γ =0.5772156649 is an Euler constant.
Further, step S17 includes:
s1701, calculating each data x in the subset x i Abnormality index S (x) i N), the abnormality index calculation formula is:
s1702, judging the data x i Whether the data is abnormal data or not comprises the following steps:
when E (h (x) i ) S (x) is close to n-1 i N) is close to 0, sample data x i Is judged to be normal;
when E (h (x) i ) S (x) is close to 0 i N) is close to 1, sample data x i Is determined to be abnormal;
when E (h (x) i ) S (x) approaches the average path length C (n) i N) is 0.5, sample data x i It can be determined to be normal.
Further, step S23 includes:
when the monitoring data is earth electric wave data, determining the abnormal data meeting the condition that 20dB is more than p and less than or equal to 25dB as general effective abnormal data, and determining the abnormal data in p more than 25dB as dangerous data;
when the monitoring data is ultrasonic data, determining the abnormal data meeting the condition that p is more than 20dB and less than or equal to 30dB as general effective abnormal data, and determining the abnormal data in p is more than 30dB as dangerous data.
Further, the step S3 includes one or more of the following steps:
s31, when the dangerous data in the effective abnormal data are detected, alarming and eliminating interference;
s32, when the intermittent alarm is needed,
if the general effective abnormal data or the general effective abnormal data is not detected to be sporadic and meets the condition of the preset occurrence times, setting to alarm once in the morning and afternoon of each day or alarm once every 3 hours in the time period of the morning and afternoon;
if the general effective abnormal data continuously appear, setting a fixed interval for alarming according to the actual bearing time and the use cost of the equipment;
s33, when continuous alarm is needed, selecting one time within 5-30 minutes as an alarm period. And calling the model once to detect abnormal data and alarm in each alarm period.
The invention also provides a device for identifying abnormal data of the on-line detection of the partial discharge, which comprises a memory and a processor, wherein the memory stores at least one section of program, and the at least one section of program is executed by the processor to realize the method for identifying the abnormal data of the on-line detection of the partial discharge.
The invention also provides a device for identifying abnormal data of on-line detection of partial discharge, which is characterized in that the device comprises a memory and a processor, wherein the memory stores at least one program, and the at least one program is executed by the processor to implement the method for identifying abnormal data of on-line detection of partial discharge as described above.
A computer readable storage medium, wherein at least one program is stored, and the at least one program is executed by a processor to implement the method for identifying abnormal data in office discharge online detection as described above.
The technical scheme has the following technical effects:
the technical scheme of the embodiment of the invention firstly adopts an isolated forest algorithm to establish a model for detecting the partial discharge online monitoring data abnormity, after the abnormal data is identified, effective abnormal data in the abnormal data is further distinguished by using a fixed value distinguishing method, the obtained effective abnormal data is analyzed, and the alarm is further carried out by combining a preset alarm strategy. The identification detection model has large processing capacity and high speed, has the capability of quickly and accurately distinguishing the abnormal partial discharge data, and can accurately position the time period of abnormal data according to the data acquisition time so as to quickly eliminate interference; on the basis, the service life of the battery of the warning device can be greatly prolonged by combining an optimized warning strategy, such as reasonably selecting test data in different time ranges. The invention can effectively reduce the indication of manpower and material resources, improve the economic benefit and the service quality, increase the management level of the power grid and improve the safe operation capability of the power grid.
Drawings
Fig. 1 is a flowchart of a method for identifying abnormal data of partial discharge online monitoring according to an embodiment of the present invention;
fig. 2 is a flowchart of abnormal partial discharge data monitoring in a method for identifying abnormal partial discharge online monitoring data according to an embodiment of the present invention;
fig. 3 is a comparison graph of battery service life of the method for identifying abnormal data of partial discharge online monitoring according to an embodiment of the present invention at different calling periods;
fig. 4 is a schematic diagram of an apparatus for identifying abnormal data of partial discharge online monitoring according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The invention will now be further described with reference to the drawings and the detailed description.
The first embodiment is as follows:
fig. 1 is a flowchart of a method for identifying abnormal data of partial discharge online monitoring according to an embodiment of the present invention. As shown in fig. 1, first, normal values and abnormal values are detected using an isolated forest algorithm. Secondly, firstly, the data of the ultrasonic waves and the TEV are judged by using a fixed value judgment method, which comprises the following steps: detecting the difference p between the background value and the measured value, detecting the relative value according to the technical specification of electrified detection of the power equipment, giving an early warning, and if the difference exceeds a specified value, namely dangerous data is detected, checking the reason and the occurrence position of the overhigh ultrasonic measured value and the transient earth voltage value of the switch cabinet, and if necessary, combining the two to detect the reason of the interference according to the signal characteristics and eliminating the reason. And finally, giving an alarm, and taking a corresponding alarm strategy by considering the frequency generated according to normal and abnormal data, the severity of partial discharge and the forest tree detection frequency.
The specific method comprises the following steps:
s1, constructing a model for detecting abnormity of monitoring data based on an isolated forest algorithm, and carrying out abnormity detection on partial discharge online monitoring data by using the model to identify abnormal data, wherein the monitoring data comprises earth electric wave data or ultrasonic wave data;
specifically, fig. 2 is a flowchart of abnormal partial discharge data monitoring in the method for identifying abnormal partial discharge online monitoring data according to an embodiment of the present invention, and as shown in fig. 2, monitoring data of a switch cabinet is collected as a training data set, and a subset x with a size of n is randomly selected from the training data set as a root node of a random binary tree iTree; randomly appointing the dimension of the current node, and determining a cutting point q between the minimum value and the maximum value of the dimension; according to the cutting point q, dividing data of which the data attribute value is smaller than the characteristic value q in the node into left sub-tree root nodes, and dividing data of the node except the left sub-tree into right sub-tree root nodes; recursion is carried out on each left sub-tree root node and each right sub-tree root node, a new node is constructed, and the recursion is ended when the current isolated tree reaches a preset height or the leaf nodes cannot be cut continuously, so that a complete isolated tree is obtained; repeating the steps to obtain a preset number of isolated trees to form an isolated forest. Illustratively, the collected monitoring data has an association with the collection time. For example, the number of the isolated trees in the isolated forest is generally set to 100 or more.
Specifically, for the isolated forest, calculating the average length of the subset x in the isolated tree, calculating an abnormal index according to the path length, and identifying abnormal data, including:
calculating the average path length C (n) of the subset x for normalizing the data x in the subset x i Path length of (d) and normalization E (h (x) i ) Wherein, h (x) i ) From root node to leaf node x i E (h (x)) of i ) Is the node x i The expected value of (d);
the calculation formula of C (n) is as follows:
wherein H (i) = Ln (i) + γ, i ∈ (1, 2,3, \8230;, n), γ =0.5772156649 is an Euler constant.
Calculating each data x in the subset x i Is abnormal index S (x) i N), the abnormality index calculation formula is:
judging the data x i Whether the data is abnormal data or not comprises the following steps:
when E (h (x) i ) S (x) is close to n-1 i N) is close to 0, sample data x i Is judged to be normal;
when E (h (x) i ) S (x) is close to 0 i N) is close to 1, sample data x i Is determined to be abnormal;
when E (h (x) i ) S (x) approaches the average path length C (n) i N) is 0.5, sample data x i It can be determined to be normal.
S2, judging the abnormal data by using a fixed value discrimination method to obtain effective abnormal data, wherein the method comprises the following steps:
s21, detecting a background value of ground electric waves or ultrasonic waves of the switch cabinet;
s22, obtaining a difference value p between the background value and each abnormal data;
s23, when p meets a preset condition, determining the abnormal data as effective abnormal data; specifically, the invalid anomaly is an anomaly point calculated by an algorithm, but is mistakenly detected as an anomaly number within a range in which the domestic detection standard is determined to be normal. The valid exception is the number which is detected by the algorithm to be abnormal and is detected by the difference value to be abnormal and consistent in the standard.
As shown in FIG. 1, generally, 20dB is taken as a boundary; p is greater than 20 is valid exception data; p is 20 or less as invalid exception data. Specifically, when the monitored data is earth electric wave data, determining abnormal data meeting the condition that 20dB is more than p and less than or equal to 25dB as general effective abnormal data, and determining the abnormal data in p more than 25dB as dangerous data; when the monitored data is ultrasonic data, determining abnormal data meeting the condition that p is more than 20dB and less than or equal to 30dB as general effective abnormal data, and determining the abnormal data with p more than 30dB as dangerous data.
The algorithm can position the specific time of the abnormal point, and then, in combination with the relevant regulations of the technical specification for detecting the electrification of the power equipment, the method can further deduce the reason of the abnormal data of the outgoing station according to the background value of the current environment measurement, thereby eliminating the interference.
And S3, analyzing the effective abnormal data, and giving an alarm according to the analysis result and a preset alarm strategy. Specifically, one or a combination of the following strategies is selected as an alarm strategy for partial discharge online monitoring, and the strategies include:
and when dangerous data in the effective abnormal data is detected, alarming and eliminating interference. Specifically, according to the relevance of monitoring data and practice, the algorithm can position the specific time of occurrence of abnormal points, and then, in combination with relevant regulations of power equipment live detection technical specifications, the background value is measured according to the current environment, so that the reason of occurrence of abnormal data can be further deduced, and further, interference is eliminated. Specifically, the isolated forest algorithm has the characteristic of high detection speed, and in the process of obtaining abnormal data and carrying out judgment of a fixed value judgment method, when the monitored data is earth electric wave data, dangerous data with p being more than 25dB exists in the detected abnormal data, or when the monitored data is ultrasonic wave data, dangerous data with p being more than 30dB exists in the detected abnormal data, an alarm is carried out, and interference is immediately eliminated so as to avoid the influence on subsequent monitoring.
When a break-in-time warning is required,
if the general effective abnormal data or the general effective abnormal data is not detected as the sporadic condition meeting the preset occurrence frequency condition, the alarm is set to be performed once in the morning and in the afternoon of each day or to be performed once every 3 hours in the time period of the morning and in the afternoon. Specifically, the ultrasonic wave and TEV partial discharge monitoring number analysis of each day finds that if most data of a certain area are normal or are sporadically abnormal, a fixed time point in the morning and in the afternoon of each day can be selected for carrying out alarm once or every 3 hours in the morning or in the afternoon time period can be selected for carrying out alarm once.
If the common effective abnormal data continuously appears, setting a fixed interval for alarming according to the actual bearing time and the use cost of the equipment. Specifically, in the case of continuous abnormality, if abnormal data continuously appears in a certain area within a certain period of time, in order to prolong the service life of the battery to the maximum extent, identify the abnormal data as much as possible and infer the cause of the abnormal data, the alarm may be performed at fixed intervals, and the frequency of the alarm may be selected in accordance with the duration and the use cost of the device.
When a continuous alarm is required, the alarm is sent,
and (3) taking the cooperative game relation between the battery alarm times and the forest algorithm running state into consideration, and selecting a time within 5-30 minutes as an alarm period, wherein in each alarm period, the detection model is called once to detect normal and abnormal data, and alarm is performed. In an exemplary manner, in an alarm period, a fixed value discrimination method can be used to obtain general valid abnormal data and dangerous data, so as to perform the alarm.
Specifically, algorithm power consumption consumed by an isolated forest algorithm model is predicted once a day ahead, the power consumption for transmitting and judging the algorithm in a calling period is estimated once, an optimal time is selected between 5 minutes and half an hour for calling the algorithm once, wherein in order to prevent overlarge monitoring data or undersize data of a forest tree, if a data set for 7 days exists, the forest tree algorithm is called once in each certain interval time, previous data are removed, the data set is updated, and the data in the data set are guaranteed to be the data of the latest 7 days. Fig. 3 is a comparison diagram of battery service life of the method for identifying abnormal data of partial discharge online monitoring according to the embodiment of the invention in different call cycles. As shown in fig. 3, for the comparison of the service lives of the batteries of the intelligent monitoring platform of the ring main unit in different calling periods, it is proved that the service life of the batteries can be greatly prolonged by adjusting the forest tree algorithm and the alarm frequency to 5-30 minutes for centralized judgment and alarm. The specific comparison process is as follows:
the ring main unit intelligent monitoring platformThe pool alarm system adopts a No. 5 battery, and the capacity of one No. 5 battery is 2000mAh under the general condition, which is equivalent to 7.2 multiplied by 10 6 And (6) mAs. Assuming that the alarm is transmitted once, namely, the signal is transmitted once, the power consumption is 100mA, the forest tree algorithm needs to be called once, namely, the working state, and assuming that the forest tree algorithm needs to be called once, the power consumption is 10mA. The dormant state, i.e. when not in operation, is 50 muA. Assuming that the duration of transmission and work is 100ms, calculation can be carried out, and within a period, the alarm needs to consume 10mAs of electricity, and the abnormal diagnosis needs to consume 1mAs of electricity.
Several cases of the initial strategy are assumed, a 1 minute strategy, a 2 minute strategy and a 4 minute strategy, respectively. The 1-minute strategy refers to setting 1 minute as a period, transmitting, working and sleeping within 1 minute, namely performing forest tree calling and alarming on ultrasonic waves and earth electric wave partial discharge data detected by the ring main unit monitoring system, and so on, and the 2-minute and 4-minute strategies are the same. After improvement, the forest tree algorithm and the alarm frequency are adjusted to 8 minutes for centralized judgment and alarm. The following are the calculation ideas under the several strategies: if the original strategy is 1 minute, the power consumption of the sleep state in 1 minute is 2.99mAs, and the alarm, the abnormality diagnosis and the power consumption in the standby state are added, so that the total power consumption in one period is 14mAs. The remaining battery level y can then be obtained as a function of the service time t (min): y =7.2 × 10 6 -14t。
By the same token, the power consumption in the use period is 17mAs and 23mAs respectively under the time period of 2 minutes and 4 minutes, and the function relationship is as follows:
if the improved 8-minute strategy is adopted, 8 minutes is an optimized scheduling period, the power consumption is 23.99mAs in the sleep state within one period, the power consumption in the sleep state is increased compared with that in the original strategy, and the transmission and working power quantity is unchanged. The sum then gives a total power consumption of 25mAs. The function relationship is as follows:
as shown in FIG. 3, the data diagnosis and alarm are performed under the original strategy, the battery can only be used for 358 days under the 1 minute strategy, the using days of the battery under the 2 minutes and 4 minutes strategies are 589 days and 870 days respectively, and the optimized battery strategy can be used for 1143 days after adjustment. It is clear that the optimized strategy does have a significant effect over the previous strategy, extending the two year usage time over the 1 minute battery alarm strategy. The battery alarm strategy based on the isolated forest algorithm can actually and effectively prolong the service life of the battery.
The method comprises the steps of firstly establishing a model for detecting the partial discharge online monitoring data abnormality by adopting an isolated forest algorithm, identifying abnormal data, further distinguishing effective abnormal data by using a fixed value distinguishing method, analyzing the obtained effective abnormal data, and further combining a preset alarm strategy to alarm. The recognition detection model is large in processing capacity and high in speed, has the capability of quickly and accurately distinguishing the abnormal partial discharge data, and can accurately position the time period of abnormal data according to the data acquisition time and quickly eliminate interference. On the basis, the service life of a battery of the warning device is greatly prolonged by combining with an optimized warning strategy.
The second embodiment:
the present invention further provides a device for identifying abnormal data in partial discharge online monitoring, as shown in fig. 4, the device includes a processor 401, a memory 402, a bus 403, and a computer program stored in the memory 402 and capable of running on the processor 401, the processor 401 includes one or more processing cores, the memory 402 is connected to the processor 401 through the bus 403, the memory 402 is used for storing program instructions, and the processor implements the steps in the foregoing method embodiment according to the first embodiment of the present invention when executing the computer program.
Further, as an executable scheme, the device for identifying the partial discharge online monitoring abnormal data may be a computer unit, and the computer unit may be a computing device such as a desktop computer, a notebook, a palmtop computer, and a cloud server. The computer unit may include, but is not limited to, a processor, a memory. It will be appreciated by those skilled in the art that the above-described constituent structures of the computer unit are merely examples of the computer unit, and do not constitute a limitation of the computer unit, and may include more or less components than those described above, or combine some components, or different components. For example, the computer unit may further include an input/output device, a network access device, a bus, and the like, which is not limited in this embodiment of the present invention.
Further, as an executable solution, the Processor may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, a discrete hardware component, and the like. The general purpose processor may be a microprocessor or the processor may be any conventional processor or the like, the processor being the control center for the computer unit, various interfaces and lines connecting the various parts of the overall computer unit.
The memory may be used to store the computer programs and/or modules, and the processor may implement the various functions of the computer unit by running or executing the computer programs and/or modules stored in the memory, as well as by invoking data stored in the memory. The memory can mainly comprise a program storage area and a data storage area, wherein the program storage area can store an operating system and an application program required by at least one function; the storage data area may store data created according to the use of the mobile phone, and the like. In addition, the memory may include high speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), at least one magnetic disk storage device, a Flash memory device, or other volatile solid state storage device.
Example three:
the present invention also provides a computer-readable storage medium, which stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned method of an embodiment of the present invention.
The computer unit integrated module/unit, if implemented in the form of a software functional unit and sold or used as a separate product, may be stored in a computer readable storage medium. Based on such understanding, all or part of the flow of the method according to the embodiments of the present invention may also be implemented by a computer program, which may be stored in a computer-readable storage medium, and when the computer program is executed by a processor, the steps of the method embodiments may be implemented. Wherein the computer program comprises computer program code, which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer Memory, read-Only Memory (ROM), random Access Memory (RAM), software distribution medium, and the like. It should be noted that the computer readable medium may contain content that is appropriately increased or decreased as required by legislation and patent practice in the jurisdiction.
While the invention has been particularly shown and described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (9)
1. A method for identifying abnormal data of partial discharge online monitoring is used for online monitoring and alarming partial discharge data of a switch cabinet, and is characterized by comprising the following steps:
s1, constructing a model for detecting abnormity of monitoring data based on an isolated forest algorithm, and carrying out abnormity detection on partial discharge online monitoring data by using the model to identify abnormal data, wherein the monitoring data comprises earth electric wave data or ultrasonic wave data;
s2, judging the abnormal data by using a fixed value discrimination method to obtain effective abnormal data, wherein the method comprises the following steps:
s21, detecting a background value of the ground electric wave or the ultrasonic wave of the switch cabinet;
s22, obtaining a difference value p between the background value and each abnormal data;
s23, when p meets a preset condition, determining the corresponding abnormal data as effective abnormal data;
and S3, analyzing the effective abnormal data, and giving an alarm according to the analysis result and a preset alarm strategy.
2. The method of claim 1, wherein in step S1, the step of constructing the model comprises:
s11, collecting the monitoring data of the switch cabinet as a training data set, and randomly selecting a subset x with the size of n from the training data set as a root node of a random binary tree;
s12, randomly appointing the dimensionality of the current node, and determining a cutting point q between the minimum value and the maximum value of the dimensionality;
s13, according to the cutting point q, dividing the data of which the data attribute value is smaller than the characteristic value q in the node into left sub-tree nodes, and dividing the data of the node except the left sub-tree into right sub-tree nodes;
s14, constructing new nodes for recursion steps S12-S13 of each left sub-tree node and each right sub-tree node, and ending recursion when the current isolated tree reaches a preset height or leaf nodes cannot be cut continuously to obtain a complete isolated tree;
and S15, repeating the steps S12-S14 to obtain a preset number of isolated trees to form an isolated forest.
3. The method of claim 2, wherein in step S1, the step of identifying anomalous data comprises:
s16, calculating the average length of the subset x in the isolated tree;
and S17, calculating an abnormal index according to the path length, and identifying abnormal data.
4. The method according to claim 3, wherein step S16 comprises:
calculating the average path length C (n) of the subset x for normalizing the data x in the subset x i Path length of (d) and normalization E (h (x) i ) Wherein, h (x) i ) From root node to leaf node x i E (h (x)) of i ) Is the node x) i The expected value of (d);
the calculation formula of C (n) is as follows:
wherein H (i) = Ln (i) + gamma, i epsilon (1, 2,3, \8230; n), gamma =0.5772156649 is Euler constant.
5. The method according to claim 3, wherein step S17 comprises:
s1701, calculating each data x in the subset x i Abnormality index S (x) i N), the abnormality index calculation formula is:
s1702, judging the data x i Whether the data is abnormal data or not comprises the following steps:
when E (h (x) i ) S (x) is close to n-1 i N) is close to 0, sample data x i Is judged to be normal;
when E (h (x) i ) S (x) is close to 0 i N) is close to 1, sample data x i Is determined to be abnormal;
when E (h (x) i ) S (x) approaches the average path length C (n) i N) is 0.5, sample data x i It can be determined to be normal.
6. The method according to claim 1, wherein step S23 comprises:
determining abnormal data with the p value larger than 20 as valid abnormal data;
further, when the monitoring data is earth electric wave data, determining the abnormal data meeting the condition that 20dB is more than p and less than or equal to 25dB as general effective abnormal data, and determining the abnormal data in p more than 25dB as dangerous data;
and when the monitoring data are ultrasonic data, determining the abnormal data meeting the condition that p is more than 20dB and less than or equal to 30dB as general effective abnormal data, and determining the abnormal data with p more than 30dB as dangerous data.
7. The method according to claim 6, wherein said step S3 comprises one or more of the following steps:
s31, when the dangerous data in the effective abnormal data is detected, alarming and eliminating interference;
s32, when the intermittent alarm is needed,
if the general effective abnormal data or the general effective abnormal data is not detected to be sporadic and meets the condition of the preset occurrence times, setting to alarm once in the morning and afternoon of each day or alarm once every 3 hours in the time period of the morning and afternoon;
if the general effective abnormal data continuously appear, setting a fixed interval for alarming according to the actual bearing time and the use cost of the equipment;
s33, when continuous alarming is needed, selecting one time within 5-30 minutes as an alarming period, wherein in each alarming period, the model is called once to detect abnormal data, and alarming is carried out.
8. An apparatus for identifying abnormal data in an office discharge online detection, the apparatus comprising a memory and a processor, the memory storing at least one program, the at least one program being executed by the processor to implement the method for identifying abnormal data in an office discharge online detection according to any one of claims 1 to 7.
9. A computer-readable storage medium, wherein at least one program is stored in the storage medium, and the at least one program is executed by a processor to implement the method for identifying partial discharge online detection abnormal data according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211044053.4A CN115392380A (en) | 2022-08-30 | 2022-08-30 | Method and device for identifying abnormal data of partial discharge online monitoring and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211044053.4A CN115392380A (en) | 2022-08-30 | 2022-08-30 | Method and device for identifying abnormal data of partial discharge online monitoring and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115392380A true CN115392380A (en) | 2022-11-25 |
Family
ID=84122586
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211044053.4A Pending CN115392380A (en) | 2022-08-30 | 2022-08-30 | Method and device for identifying abnormal data of partial discharge online monitoring and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115392380A (en) |
-
2022
- 2022-08-30 CN CN202211044053.4A patent/CN115392380A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109034244B (en) | Line loss abnormity diagnosis method and device based on electric quantity curve characteristic model | |
| CN111143102B (en) | Abnormal data detection method and device, storage medium and electronic equipment | |
| CN109977535B (en) | Line loss abnormality diagnosis method, device, equipment and readable storage medium | |
| CN112188531B (en) | Abnormality detection method, abnormality detection device, electronic apparatus, and computer storage medium | |
| CN108429649B (en) | System for comprehensive abnormity judgment based on multiple single-type acquisition results | |
| US10802057B2 (en) | Systems and methods for monitoring a power system | |
| CN110441584B (en) | Equipment running state monitoring method, device, storage medium and system | |
| CN108198408B (en) | Self-adaptive anti-electricity-stealing monitoring method and system based on electricity information acquisition system | |
| CN115865649B (en) | Intelligent operation and maintenance management control method, system and storage medium | |
| CN116112292B (en) | Abnormal behavior detection method, system and medium based on network flow big data | |
| CN111371188A (en) | Real-time early warning system for abnormal power consumption data | |
| CN112532435B (en) | Operation and maintenance method, operation and maintenance management platform, equipment and medium | |
| CN111539563A (en) | Energy consumption safety state prediction method, device, server and storage medium | |
| CN111865407A (en) | Intelligent early warning method, device, equipment and storage medium for optical channel performance degradation | |
| CN109656786B (en) | Log acquisition method and device | |
| CN113591393A (en) | Fault diagnosis method, device, equipment and storage medium of intelligent substation | |
| CN115049410A (en) | Electricity stealing behavior identification method and device, electronic equipment and computer readable storage medium | |
| CN116882804A (en) | Intelligent power monitoring method and system | |
| CN116345700B (en) | Energy consumption monitoring method and monitoring system for energy storage power station | |
| CN113326585B (en) | Energy efficiency abnormality early warning method and device for gas boiler and computer equipment | |
| CN114138601A (en) | Service alarm method, device, equipment and storage medium | |
| CN113988723A (en) | User behavior locking method and system based on power consumption data anomaly analysis | |
| CN116381542B (en) | Health diagnosis method and device of power supply equipment based on artificial intelligence | |
| CN115392380A (en) | Method and device for identifying abnormal data of partial discharge online monitoring and storage medium | |
| CN105741187B (en) | Data processing method and device for living resource consumption and monitoring method and device for living resource consumption |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |