CN113406438B - Intelligent fault diagnosis method suitable for low-voltage transformer area and operation and maintenance system thereof - Google Patents
Intelligent fault diagnosis method suitable for low-voltage transformer area and operation and maintenance system thereof Download PDFInfo
- Publication number
- CN113406438B CN113406438B CN202110694891.5A CN202110694891A CN113406438B CN 113406438 B CN113406438 B CN 113406438B CN 202110694891 A CN202110694891 A CN 202110694891A CN 113406438 B CN113406438 B CN 113406438B
- Authority
- CN
- China
- Prior art keywords
- fault
- low
- model
- voltage
- characteristic
- 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.)
- Active
Links
- 238000003745 diagnosis Methods 0.000 title claims abstract description 101
- 238000012423 maintenance Methods 0.000 title claims abstract description 83
- 238000000034 method Methods 0.000 title claims abstract description 72
- 230000002159 abnormal effect Effects 0.000 claims abstract description 99
- 230000008447 perception Effects 0.000 claims abstract description 69
- 238000012544 monitoring process Methods 0.000 claims abstract description 10
- 238000012549 training Methods 0.000 claims description 37
- 230000008569 process Effects 0.000 claims description 36
- 238000007726 management method Methods 0.000 claims description 35
- 230000005856 abnormality Effects 0.000 claims description 21
- 238000012545 processing Methods 0.000 claims description 17
- 238000009826 distribution Methods 0.000 claims description 16
- 238000004458 analytical method Methods 0.000 claims description 15
- 230000005611 electricity Effects 0.000 claims description 14
- 230000006870 function Effects 0.000 claims description 12
- 238000010606 normalization Methods 0.000 claims description 11
- 210000002569 neuron Anatomy 0.000 claims description 8
- 238000012706 support-vector machine Methods 0.000 claims description 8
- 238000012795 verification Methods 0.000 claims description 8
- 230000004044 response Effects 0.000 claims description 7
- 238000004422 calculation algorithm Methods 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 4
- 238000013528 artificial neural network Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000007635 classification algorithm Methods 0.000 claims description 4
- 238000004140 cleaning Methods 0.000 claims description 4
- 238000001514 detection method Methods 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 2
- 238000004891 communication Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000007935 neutral effect Effects 0.000 description 1
- 238000005191 phase separation Methods 0.000 description 1
- 238000011112 process operation Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
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
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/081—Locating faults in cables, transmission lines, or networks according to type of conductors
- G01R31/086—Locating faults in cables, transmission lines, or networks according to type of conductors in power transmission or distribution networks, i.e. with interconnected conductors
-
- 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
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/088—Aspects of digital computing
Abstract
The invention relates to the field of power systems, in particular to an intelligent fault diagnosis method suitable for a low-voltage transformer area and an operation and maintenance system thereof. The intelligent fault diagnosis method comprises the following steps: step one: determining each fault type of the low-voltage transformer area and a plurality of characteristic quantities associated with each fault type; defining abnormal characteristic states reflected by the characteristic quantities as fault labels; step two: constructing a fault perception model and a fault diagnosis model; step three: characteristic data of the low-voltage transformer area are collected in real time, and corresponding fault labels are generated; step four: inputting the characteristic data into a fault perception model, and judging whether a fault exists in the current low-voltage transformer area or not; step five: and inputting the generated characteristic value of the fault label into a fault diagnosis model, and diagnosing the cause of the fault of the current low-voltage transformer area. The invention solves the problems that the operation monitoring diagnosis and operation maintenance of the existing low-voltage transformer area depend on manual work, and the operation maintenance system of power management is not enough in automation.
Description
Technical Field
The invention relates to the field of power systems, in particular to an intelligent fault diagnosis method suitable for a low-voltage transformer area and an operation and maintenance system thereof.
Background
In the operation and maintenance management process of the existing low-voltage transformer area, the operation state of the system is mainly managed by technicians, and the fault state is analyzed and processed, so that the stable operation of the power system is maintained. The management method is highly dependent on the capability and experience of management staff, is unfavorable for the normalized management of the power system, and can influence the fault monitoring, fault cause diagnosis and system operation and maintenance level of the power system when staff change of first-line technicians.
In addition, the grid comprehensive service with the end service fusion as the target is one of important works of power management reform and exploration in the future, the information technology development and the big data application provide technical support for developing the grid comprehensive service, and simultaneously, higher requirements are also provided for comprehensive quality of basic staff. The internet of things is advanced deeply, the quantity and the variety of metering, collecting and sensing equipment at the tail end of the low-voltage power supply business are more and more, business skills and expertise mastered by a traditional platform manager are difficult to meet the development needs of modern management businesses, the business automation level of the low-voltage platform is improved through an informatization means, and the system is suitable for grid business operation and maintenance management and control.
Disclosure of Invention
Based on the fault intelligent diagnosis method and the operation and maintenance system thereof, which are applicable to the low-voltage transformer area, are provided in the invention to solve the problems that the operation monitoring diagnosis and the operation and maintenance of the existing low-voltage transformer area are dependent on manual work and the operation and maintenance system of power management is not sufficient in automation.
The invention provides a fault intelligent diagnosis method suitable for a low-voltage transformer area, which comprises the following steps:
step one: determining each fault type of the low-voltage transformer area and a plurality of characteristic quantities associated with each fault type; and defines the abnormal characteristic state reflected by each characteristic quantity as a corresponding fault label.
Step two: and constructing a fault perception model for identifying the fault type of the low-voltage area and a fault diagnosis model for analyzing the fault occurrence cause. The fault perception model is a label model based on a classification algorithm; the input of the fault perception model is the characteristic value of each characteristic quantity in the running process of the low-voltage transformer area, and the output of the fault perception model is the prediction result of different fault types. The fault diagnosis model is a trained neural network model; the failure diagnosis model includes a plurality of sub-models corresponding to respective failure types. The input of the fault diagnosis model is the characteristic value of each fault label corresponding to the fault prediction result output by the fault perception model, and the output of the fault diagnosis model is the occurrence reason corresponding to the predicted fault.
Step three: and collecting the characteristic data of various characteristic quantities related to various fault types in the low-voltage transformer area in real time, and cleaning the characteristic data. And generating a corresponding fault label based on the cleaned characteristic data, and completing assignment of the fault label.
Step four: and (3) inputting the characteristic data cleaned in the previous step into a fault perception model, judging whether a fault exists in the current low-voltage station area, and outputting a corresponding fault prediction result.
Step five: when the low-voltage transformer area fails, the generated characteristic value of the failure label is input into a failure diagnosis model, and the cause of the failure of the current low-voltage transformer area is diagnosed.
Further, in the first step, the fault types of the low-voltage transformer area include acquisition faults, metering faults, power consumption faults, cost control faults, distribution transformer faults and transformer area line loss anomalies.
Further, in the first step, the correspondence relationship between the abnormal feature states associated with the respective fault types includes:
(a) Abnormal signature status associated with an acquisition failure, comprising: the terminal is offline, the electricity meter under the concentrator has no data, the collector has no data, and the electric energy meter has no meter reading data for a plurality of days.
(b) An abnormal signature state associated with a metering fault, comprising: the method comprises the steps of flying or suddenly changing an electric energy meter, reversing the electric energy meter, stopping the electric energy meter indicating value, exceeding the clock tolerance of the electric energy meter, uncovering the electric energy meter, covering an electric energy meter button, zero voltage of the electric energy meter and undervoltage of a battery of the electric energy meter.
(c) An abnormal signature state associated with an electrical fault, comprising: voltage phase loss, current loss, zero sequence current abnormality, voltage phase interruption, current three-phase imbalance, voltage three-phase imbalance, power factor abnormality and overload.
(d) An abnormal signature status associated with a fee control fault, comprising: carrier overtime, task issuing failure, authentication failure, terminal response overtime, denial message, task overtime, issuing times or overtime, uplink message error, ammeter file error, task issuing failure and task environment abnormality.
(e) Abnormal signature status associated with a configuration failure, comprising: current three-phase imbalance, A/B/C phase current loss, A/B/C phase overload, A/B/C phase voltage open-phase, voltage three-phase imbalance, power factor abnormality, low voltage, platform load rate >80%, A/B/C phase overload, A/B/C phase light load and no load.
(f) An abnormal feature state associated with a station area line loss abnormality, comprising: and collecting an operation and maintenance management type state, a platform area metering management type state, a platform area diagnosis analysis type state and a high-loss abnormal type state. Wherein, the collecting operation and maintenance management class state includes: the continuous collection fails for more than three days, the meter is in clock error, the electric energy meter flies away and suddenly changes, the voltage is lost, the current is lost, the phase is lost and the overload is carried out; the platform area metering management class state comprises: total surface uncovered, magnification abnormal, reference table abnormal, metering point level abnormal; the diagnostic analysis class status of the area includes: the three-phase unbalance of the transformer area, the power factor monitoring of the transformer area, the reverse active power being more than 0, the stop of the electric energy meter indicating value, the reverse current, the cover opening of the electric energy meter, the lower limit of the voltage, the zero voltage and the zero live wire out of tolerance; the high-loss exception class state includes: and (5) analyzing the month continuous high loss, the day continuous high loss, the month line loss qualified day line loss unqualified and unmonitored station areas.
Further, in the second step, the construction method of the fault perception model is as follows:
and (i) extracting characteristic values of abnormal characteristic states of all fault types in the low-voltage transformer area according to fault labels associated with all fault types, and forming a complete fault portrait based on the fault labels.
(ii) carrying out manual category distinction and hierarchical relation establishment on various fault labels in the step, designing weights of different fault labels in various fault types, defining a calculation formula of the fault labels and the fault types, and forming a prediction model of a training data set.
(iii) collecting feature data of abnormal feature states associated with various fault types occurring in the low-voltage transformer area in real time, and taking the feature data as a training data set; and carrying out preliminary prediction on the training data set by combining a prediction model and an artificial verification.
(iv) using a Gaussian kernel function as a kernel function of a fault perception model, adopting a one-to-many multi-label fault diagnosis algorithm, taking characteristic data in a training data set as an input sample, and taking collected faults, metering faults, power consumption faults, cost control faults, distribution transformer faults and station area line loss anomalies as output fault types; constructing six support vector machine classifiers for six fault types; and forming a required fault perception model by using six support vector machine classifiers.
(v) setting super parameters of a training process, obtaining a preliminary prediction result obtained through comprehensive judgment of a prediction model and artificial verification, training a constructed fault perception model by utilizing a training data set, comparing the preliminary prediction result with a result output by the fault perception model, and further calculating the minimum error of the fault perception model; and finishing the training process of the fault perception model until the minimum error requirement of the training stage is met.
Further, the fault diagnosis model is a trained three-layer neural network, comprising an input layer, an hidden layer and an output layer. The number of neurons of the input layer corresponds to the number of fault labels associated with each fault type, and the input characteristic data is the characteristic value of each fault label. The hidden layer adopts a ReLU activation function; the number of neurons of the output layer is equal to the number of reasons associated with each fault type, the output layer obtains the probability of each reason associated with the fault type, and the reason with the highest probability among the reasons is used as the output diagnosis result.
Further, in the third step, the fault label generation and assignment process is as follows: judging whether the corresponding collected characteristic quantity is in an abnormal interval or not, and if so, generating a fault label associated with the current characteristic quantity; and then carrying out normalization processing on the detection value of the current feature quantity, and assigning the normalization processing result to the fault label as the feature value of the fault label.
Further, the invention provides a fault intelligent diagnosis method suitable for the low-voltage transformer area, which further comprises the following steps:
step six: and according to the determined faults and the diagnosed fault reasons, inquiring a fault and solution comparison table established by expert experience to obtain the solution of the current faults, and sending the solution to corresponding operation and maintenance personnel.
Wherein, the mapping relation between the fault types caused by different reasons and the corresponding fault solutions is established in the fault and solution comparison table.
The invention also comprises an operation and maintenance system suitable for the low-voltage station, wherein the operation and maintenance system adopts the fault intelligent diagnosis method suitable for the low-voltage station, realizes the processes of real-time monitoring, fault sensing and fault cause diagnosis of the low-voltage station, and sends corresponding operation and maintenance requirements and solutions to corresponding responsible personnel for processing. The operation and maintenance system comprises: the system comprises a characteristic acquisition module, a fault label generation module, a fault perception module, a fault diagnosis module, a solution inquiry module and an operation and maintenance requirement dispatch module.
The characteristic acquisition module is used for acquiring values of all characteristic quantities related to faults possibly occurring in the low-voltage station area.
The fault label generation module is used for judging whether the value of each characteristic quantity acquired by the characteristic acquisition module is in an abnormal interval, generating a fault label aiming at the characteristic quantity in the abnormal interval, carrying out normalization processing on the characteristic value of the corresponding characteristic quantity, and assigning the value to the fault label.
The fault sensing module is used for judging whether the current low-voltage station area has faults or not according to the values of the characteristic quantities acquired by the characteristic acquisition module; the fault type judged by the fault perception model comprises acquisition faults, metering faults, electricity consumption faults, cost control faults, distribution and transformation faults and abnormal line loss of the transformer area.
The fault diagnosis module is used for acquiring the value of a fault label corresponding to the current fault when the fault diagnosis module judges that the current low-voltage transformer area has the fault, and analyzing and obtaining the fault occurrence reason corresponding to the current fault according to the characteristic value of the fault label.
The solution inquiry module is used for acquiring the faults obtained by the fault perception module and the fault occurrence reasons obtained by the fault diagnosis module; and inquiring a 'fault and solution comparison table' established according to expert experience to obtain the solution of the current fault.
The operation and maintenance requirement dispatching module is used for acquiring a current fault solution obtained by the solution query model; and inquiring an operation and maintenance personnel responsibility table to obtain operation and maintenance personnel corresponding to the current fault, and finally, sending the solution corresponding to the current fault to the operation and maintenance personnel corresponding to the current fault.
Further, the fault diagnosis module comprises an acquisition fault sub-model, a metering fault sub-model, an electricity consumption fault sub-model, a cost control fault sub-model, a distribution transformer fault sub-model and a transformer area line loss abnormal sub-model; when the fault diagnosis model obtains the fault prediction result judged by the fault perception model, the corresponding sub-models are respectively called to complete the analysis process of fault occurrence reasons corresponding to different fault types.
Further, the operation and maintenance personnel responsibility table is a responsibility classification table of responsible personnel established according to the current state of low-voltage station area management; the operation and maintenance personnel responsibility table is provided with different fault types and corresponding relations between fault occurrence reasons and each operation and maintenance personnel; and the content in the operation and maintenance personnel responsibility table is updated in real time according to the management current situation of the current low-voltage station area.
The intelligent fault diagnosis method and the operation and maintenance system thereof suitable for the low-voltage transformer area have the following beneficial effects:
the intelligent fault diagnosis method provided by the invention can monitor the real-time running condition of each device in the current low-voltage transformer area according to the collected characteristic data, evaluate the fault state of each device, timely predict possible faults, and accurately judge the fault type and the formation reason of the faults. The invention adopts a label model based on a clustering algorithm and a neural network model as a fault perception model and a fault diagnosis model in the scheme; the self-learning performance, high reliability and high availability of the models are utilized, the accuracy and the rapidity of fault diagnosis results are greatly improved, and more timely and more accurate fault classification and analysis effects are achieved. The dependency on manual experience in the fault analysis and operation and maintenance management process is reduced.
According to the invention, a comparison table of faults and solutions is established according to expert experience, so that an optimal operation and maintenance scheme is provided for faults caused by different reasons of each type, and guidance is provided for relevant technicians to process operation and maintenance requirements. Meanwhile, the technical scheme of the invention has high automation level, higher operability and no need of additional education and training for management staff in the implementation process; the influence of inconsistent capability levels of different technicians on operation and maintenance stability is solved.
Drawings
The foregoing and other objects, features and advantages of the invention will be apparent from the following more particular description of preferred embodiments of the invention, as illustrated in the accompanying drawings.
FIG. 1 is a flow chart of a fault intelligent diagnosis method suitable for a low-voltage transformer area in the embodiment 1 of the invention;
FIG. 2 is a flowchart of a fault intelligent diagnosis method applicable to a low-voltage transformer area, in which an operation and maintenance requirement dispatch process is added in embodiment 2 of the present invention;
fig. 3 is a schematic block diagram of an operation and maintenance system suitable for a low-voltage transformer area in embodiment 3 of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "or/and" as used herein includes any and all combinations of one or more of the associated listed items.
Example 1
The embodiment provides a fault intelligent diagnosis method suitable for a low-voltage transformer area, which comprises the following steps:
step one: determining each fault type of the low-voltage transformer area and a plurality of characteristic quantities associated with each fault type; and defines the abnormal characteristic state reflected by each characteristic quantity as a corresponding fault label.
According to faults occurring in the actual running process of the low-voltage transformer area, the fault types occurring in the low-voltage transformer area are divided into six main categories, namely acquisition faults, metering faults, electricity consumption faults, cost control faults, distribution transformer faults and transformer area line loss anomalies.
In this embodiment, the correspondence relationship of the abnormal feature states associated with each fault type includes:
(a) Abnormal signature status associated with an acquisition failure, comprising: the terminal is offline, the electricity meter under the concentrator has no data, the collector has no data, and the electric energy meter has no meter reading data for a plurality of days.
(b) An abnormal signature state associated with a metering fault, comprising: the method comprises the steps of flying or suddenly changing an electric energy meter, reversing the electric energy meter, stopping the electric energy meter indicating value, exceeding the clock tolerance of the electric energy meter, uncovering the electric energy meter, covering an electric energy meter button, zero voltage of the electric energy meter and undervoltage of a battery of the electric energy meter.
(c) An abnormal signature state associated with an electrical fault, comprising: voltage phase loss, current loss, zero sequence current abnormality, voltage phase interruption, current three-phase imbalance, voltage three-phase imbalance, power factor abnormality and overload.
(d) An abnormal signature status associated with a fee control fault, comprising: carrier overtime, task issuing failure, authentication failure, terminal response overtime, denial message, task overtime, issuing times or overtime, uplink message error, ammeter file error, task issuing failure and task environment abnormality.
(e) Abnormal signature status associated with a configuration failure, comprising: current three-phase imbalance, A/B/C phase current loss, A/B/C phase overload, A/B/C phase voltage open-phase, voltage three-phase imbalance, power factor abnormality, low voltage, platform load rate >80%, A/B/C phase overload, A/B/C phase light load and no load.
(f) An abnormal feature state associated with a station area line loss abnormality, comprising: and collecting an operation and maintenance management type state, a platform area metering management type state, a platform area diagnosis analysis type state and a high-loss abnormal type state. Wherein, the collecting operation and maintenance management class state includes: the continuous collection fails for more than three days, the meter is in clock error, the electric energy meter flies away and suddenly changes, the voltage is lost, the current is lost, the phase is lost and the overload is carried out; the platform area metering management class state comprises: total surface uncovered, magnification abnormal, reference table abnormal, metering point level abnormal; the diagnostic analysis class status of the area includes: the three-phase unbalance of the transformer area, the power factor monitoring of the transformer area, the reverse active power being more than 0, the stop of the electric energy meter indicating value, the reverse current, the cover opening of the electric energy meter, the lower limit of the voltage, the zero voltage and the zero live wire out of tolerance; the high-loss exception class state includes: and (5) analyzing the month continuous high loss, the day continuous high loss, the month line loss qualified day line loss unqualified and unmonitored station areas.
Step two: and constructing a fault perception model for identifying the fault type of the low-voltage area and a fault diagnosis model for analyzing the fault occurrence cause. The fault perception model is a label model based on a classification algorithm; the input of the fault perception model is the characteristic value of each characteristic quantity in the running process of the low-voltage transformer area, and the output of the fault perception model is the prediction result of different fault types. The fault diagnosis model is a trained neural network model; the failure diagnosis model includes a plurality of sub-models corresponding to respective failure types. The input of the fault diagnosis model is the characteristic value of each fault label corresponding to the fault prediction result output by the fault perception model, and the output of the fault diagnosis model is the occurrence reason corresponding to the predicted fault.
In this embodiment, the method for constructing the fault perception model is as follows:
and (i) extracting characteristic values of abnormal characteristic states of all fault types in the low-voltage transformer area according to fault labels associated with all fault types, and forming a complete fault portrait based on the fault labels.
(ii) carrying out manual category distinction and hierarchical relation establishment on various fault labels in the step, designing weights of different fault labels in various fault types, defining a calculation formula of the fault labels and the fault types, and forming a prediction model of a training data set.
(iii) collecting feature data of abnormal feature states associated with various fault types occurring in the low-voltage transformer area in real time, and taking the feature data as a training data set; and carrying out preliminary prediction on the training data set by combining a prediction model and an artificial verification.
(iv) using a Gaussian kernel function as a kernel function of a fault perception model, adopting a one-to-many multi-label fault diagnosis algorithm, taking characteristic data in a training data set as an input sample, and taking collected faults, metering faults, power consumption faults, cost control faults, distribution transformer faults and station area line loss anomalies as output fault types; constructing six support vector machine classifiers for six fault types; and forming a required fault perception model by using six support vector machine classifiers.
(v) setting super parameters of a training process, obtaining a preliminary prediction result obtained through comprehensive judgment of a prediction model and artificial verification, training a constructed fault perception model by utilizing a training data set, comparing the preliminary prediction result with a result output by the fault perception model, and further calculating the minimum error of the fault perception model; and finishing the training process of the fault perception model until the minimum error requirement of the training stage is met.
The fault diagnosis model constructed in this embodiment is a three-layer neural network that has been trained, including an input layer, an hidden layer, and an output layer. The number of neurons of the input layer corresponds to the number of fault labels associated with each fault type, and the input characteristic data is the characteristic value of each fault label. The hidden layer adopts a ReLU activation function; the number of neurons of the output layer is equal to the number of reasons associated with each fault type, the output layer obtains the probability of each reason associated with the fault type, and the reason with the highest probability among the reasons is used as the output diagnosis result.
Step three: and collecting the characteristic data of various characteristic quantities related to various fault types in the low-voltage transformer area in real time, and cleaning the characteristic data. And generating a corresponding fault label based on the cleaned characteristic data, and completing assignment of the fault label.
In this embodiment, the fault label generation and assignment process is as follows: judging whether the corresponding collected characteristic quantity is in an abnormal interval or not, and if so, generating a fault label associated with the current characteristic quantity; and then carrying out normalization processing on the detection value of the current feature quantity, and assigning the normalization processing result to the fault label as the feature value of the fault label.
Step four: and (3) inputting the characteristic data cleaned in the previous step into a fault perception model, judging whether a fault exists in the current low-voltage station area, and outputting a corresponding fault prediction result.
Step five: when the low-voltage transformer area fails, the generated characteristic value of the failure label is input into a failure diagnosis model, and the cause of the failure of the current low-voltage transformer area is diagnosed.
In this embodiment, the diagnosed fault causes mainly include the following:
when the current fault type is judged to be the acquisition fault: 1. for an abnormal state that the terminal is offline, a concentrator that appears as if the terminal is offline for a certain zone is not online, which may be caused by one or more factors, such as a field outage, a terminal failure, a communication failure, etc. 2. For the abnormal state that the electricity meter under the concentrator has no data, when the phenomenon occurs, the specific problem of the abnormal phenomenon cannot be directly judged, and the problems of abnormal collection of the main station, disconnection of the concentrator, meter reading failure of the concentrator, abnormal clock of the concentrator, meter reading failure of the CCO module and meter reading of the on-site electric energy meter are likely to occur. 3. For the abnormal state that the electricity meter has no data under the collector, the fault can be caused by the disconnection of a 485 line, the fault of an STA module and the fault of the collector, and is generally difficult to be caused by the fault of all the electricity meters. 4. For the abnormal state that the electric energy meter continuously has no meter reading data for a plurality of days, the abnormal state may be caused by hardware faults, out-of-tolerance clocks and under-voltage batteries of the electric energy meter, or may be caused by interruption of connection between an electric energy meter collector or carrier module and an electric meter, or may be caused by failure of a terminal to acquire the data of the collector or the carrier module.
When the current fault type is judged to be metering fault: the abnormal state of the electric energy meter with uneven representation value can be caused by a concentrator, a collector or a carrier module, or can be caused by storage of the electric energy meter or hardware faults.
When the current fault type is judged to be the power failure: for abnormal states of unbalanced three phases of voltage, the abnormal states may be caused by more users of phase separation bands, and disconnection or grounding may also cause power failure.
When the current fault type is judged to be a cost control fault: 1. for abnormal states of carrier timeout, possible reasons include that the terminal issues instructions to the meter, and the meter is not timely or successfully returned with the carrier environment influence data to cause task timeout. 2. For abnormal states of task delivery failure, there may be a reason that a channel is occupied (multiple tasks are simultaneously executed, and when a concentrator receives a fee control instruction, other tasks are executed, such as tasks of collecting data, delivering parameters, and the like). 3. For the abnormal state of authentication failure, the possible reason is that the master station issues a fee control instruction, the identity authentication of the encryption machine needs to be called first, if the identity authentication fails, the secret key cannot be downloaded, and the subsequent instruction cannot execute the command to the terminal. 4. For the abnormal state of the response timeout of the terminal, the possible reason is that the terminal does not return the result to the main station within 180 seconds after receiving the fee control instruction, the channel is closed for more than 180 seconds, and the response timeout of the terminal is primarily judged. 5. For the abnormal state of the denial message, the main station issues a fee control instruction to the terminal, the terminal does not successfully recognize the instruction, the terminal directly feeds back the denial message of the main station, the command is not executed, and the channel is closed. 6. And for the abnormal state of the overtime of the task, the abnormal state is expressed as that the master station issues a fee control instruction to the terminal, the execution result is returned when the time for the terminal to receive and execute the task is not within the stipulation, and the overtime of the task is judged to be unsuccessful in execution. 7. For the abnormal state of the issuing times or overtime, the abnormal state is represented by the fact that the execution times of the same ammeter exceeds 10 times, and the issuing times are judged; the multi-fee control instruction execution (a plurality of users simultaneously and concurrently execute fee control instructions under one concentrator, so that response time of the concentrator is overtime.8. For the abnormal state of the uplink message error, the problem that the terminal executes the task successfully is shown, an execution result is returned to the main station, the type of the terminal return message does not accord with the type of the main station message, the main station cannot identify the message, and then judges that the uplink message is wrong, and cannot identify the abnormal state of the electric meter file is shown as that the main station issues the instruction to the field terminal, the electric meter file (asset number, communication address and the like) is not found by the field terminal, the issuing process is inconsistent, the command cannot be executed, the problem that the electric meter file is wrong is primarily judged, and the task environment is abnormal is mainly caused by the reasons that the electric meter file is abnormal, the parameters of the terminal are lost, the 4G signal problem of the field, the online state of the terminal and the like, and the task cannot be issued to the field.
When the current fault type is judged to be the distribution transformer fault: for abnormal states of three-phase unbalance of current, voltage fluctuation is possible, so that the voltage unbalance is caused, and the three-phase current is unbalanced; three-phase and single-phase loads are unbalanced, so that three-phase current is unbalanced; it is also possible that a short circuit between the phase and neutral leads to a three-phase voltage, current imbalance.
In the actual application process, the fault perception model and the fault diagnosis model in the embodiment can realize autonomous learning according to the perception or diagnosis result feedback in the actual operation process, thereby improving the self prediction success rate and further better exerting the requirement of guaranteeing the operation stability of the low-voltage station area.
According to the intelligent fault diagnosis method provided by the embodiment, the real-time running condition of each device in the current low-voltage transformer area can be monitored according to the collected characteristic data, the fault state of each device is evaluated, possible faults are predicted in time, and the fault type and the fault forming reason are accurately judged. The invention adopts a label model based on a clustering algorithm and a neural network model as a fault perception model and a fault diagnosis model in the scheme; the self-learning performance, high reliability and high availability of the models are utilized, the accuracy of fault diagnosis results is greatly improved, and the more timely and more accurate fault classification and analysis effects are achieved; the dependency on manual experience in the fault analysis and operation and maintenance management process is reduced.
Example 2
The present embodiment provides a fault intelligent diagnosis method applicable to a low-voltage transformer area, and the difference between the present embodiment and embodiment 1 is that: in the order sending process for increasing the operation and maintenance requirements in this embodiment, a specific flow of a method provided in this embodiment is shown in fig. 2, and the specific process is as follows:
step one: determining each fault type of the low-voltage transformer area and a plurality of characteristic quantities associated with each fault type; and defines the abnormal characteristic state reflected by each characteristic quantity as a corresponding fault label.
According to faults occurring in the actual running process of the low-voltage transformer area, the fault types occurring in the low-voltage transformer area are divided into six main categories, namely acquisition faults, metering faults, electricity consumption faults, cost control faults, distribution transformer faults and transformer area line loss anomalies.
In this embodiment, the correspondence relationship of the abnormal feature states associated with each fault type includes:
(a) Abnormal signature status associated with an acquisition failure, comprising: the terminal is offline, the electricity meter under the concentrator has no data, the collector has no data, and the electric energy meter has no meter reading data for a plurality of days.
(b) An abnormal signature state associated with a metering fault, comprising: the method comprises the steps of flying or suddenly changing an electric energy meter, reversing the electric energy meter, stopping the electric energy meter indicating value, exceeding the clock tolerance of the electric energy meter, uncovering the electric energy meter, covering an electric energy meter button, zero voltage of the electric energy meter and undervoltage of a battery of the electric energy meter.
(c) An abnormal signature state associated with an electrical fault, comprising: voltage phase loss, current loss, zero sequence current abnormality, voltage phase interruption, current three-phase imbalance, voltage three-phase imbalance, power factor abnormality and overload.
(d) An abnormal signature status associated with a fee control fault, comprising: carrier overtime, task issuing failure, authentication failure, terminal response overtime, denial message, task overtime, issuing times or overtime, uplink message error, ammeter file error, task issuing failure and task environment abnormality.
(e) Abnormal signature status associated with a configuration failure, comprising: current three-phase imbalance, A/B/C phase current loss, A/B/C phase overload, A/B/C phase voltage open-phase, voltage three-phase imbalance, power factor abnormality, low voltage, platform load rate >80%, A/B/C phase overload, A/B/C phase light load and no load.
(f) An abnormal feature state associated with a station area line loss abnormality, comprising: and collecting an operation and maintenance management type state, a platform area metering management type state, a platform area diagnosis analysis type state and a high-loss abnormal type state. Wherein, the collecting operation and maintenance management class state includes: the continuous collection fails for more than three days, the meter is in clock error, the electric energy meter flies away and suddenly changes, the voltage is lost, the current is lost, the phase is lost and the overload is carried out; the platform area metering management class state comprises: total surface uncovered, magnification abnormal, reference table abnormal, metering point level abnormal; the diagnostic analysis class status of the area includes: the three-phase unbalance of the transformer area, the power factor monitoring of the transformer area, the reverse active power being more than 0, the stop of the electric energy meter indicating value, the reverse current, the cover opening of the electric energy meter, the lower limit of the voltage, the zero voltage and the zero live wire out of tolerance; the high-loss exception class state includes: and (5) analyzing the month continuous high loss, the day continuous high loss, the month line loss qualified day line loss unqualified and unmonitored station areas.
Step two: and constructing a fault perception model for identifying the fault type of the low-voltage area and a fault diagnosis model for analyzing the fault occurrence cause. The fault perception model is a label model based on a classification algorithm; the input of the fault perception model is the characteristic value of each characteristic quantity in the running process of the low-voltage transformer area, and the output of the fault perception model is the prediction result of different fault types. The fault diagnosis model is a trained neural network model; the failure diagnosis model includes a plurality of sub-models corresponding to respective failure types. The input of the fault diagnosis model is the characteristic value of each fault label corresponding to the fault prediction result output by the fault perception model, and the output of the fault diagnosis model is the occurrence reason corresponding to the predicted fault.
In this embodiment, the method for constructing the fault perception model is as follows:
and (i) extracting characteristic values of abnormal characteristic states of all fault types in the low-voltage transformer area according to fault labels associated with all fault types, and forming a complete fault portrait based on the fault labels.
(ii) carrying out manual category distinction and hierarchical relation establishment on various fault labels in the step, designing weights of different fault labels in various fault types, defining a calculation formula of the fault labels and the fault types, and forming a prediction model of a training data set.
(iii) collecting feature data of abnormal feature states associated with various fault types occurring in the low-voltage transformer area in real time, and taking the feature data as a training data set; and carrying out preliminary prediction on the training data set by combining a prediction model and an artificial verification.
(iv) using a Gaussian kernel function as a kernel function of a fault perception model, adopting a one-to-many multi-label fault diagnosis algorithm, taking characteristic data in a training data set as an input sample, and taking collected faults, metering faults, power consumption faults, cost control faults, distribution transformer faults and station area line loss anomalies as output fault types; constructing six support vector machine classifiers for six fault types; and forming a required fault perception model by using six support vector machine classifiers.
(v) setting super parameters of a training process, obtaining a preliminary prediction result obtained through comprehensive judgment of a prediction model and artificial verification, training a constructed fault perception model by utilizing a training data set, comparing the preliminary prediction result with a result output by the fault perception model, and further calculating the minimum error of the fault perception model; and finishing the training process of the fault perception model until the minimum error requirement of the training stage is met.
The fault diagnosis model constructed in this embodiment is a three-layer neural network that has been trained, including an input layer, an hidden layer, and an output layer. The number of neurons of the input layer corresponds to the number of fault labels associated with each fault type, and the input characteristic data is the characteristic value of each fault label. The hidden layer adopts a ReLU activation function; the number of neurons of the output layer is equal to the number of reasons associated with each fault type, the output layer obtains the probability of each reason associated with the fault type, and the reason with the highest probability among the reasons is used as the output diagnosis result.
Step three: and collecting the characteristic data of various characteristic quantities related to various fault types in the low-voltage transformer area in real time, and cleaning the characteristic data. And generating a corresponding fault label based on the cleaned characteristic data, and completing assignment of the fault label.
In this embodiment, the fault label generation and assignment process is as follows: judging whether the corresponding collected characteristic quantity is in an abnormal interval or not, and if so, generating a fault label associated with the current characteristic quantity; and then carrying out normalization processing on the detection value of the current feature quantity, and assigning the normalization processing result to the fault label as the feature value of the fault label.
Step four: and (3) inputting the characteristic data cleaned in the previous step into a fault perception model, judging whether a fault exists in the current low-voltage station area, and outputting a corresponding fault prediction result.
Step five: when the low-voltage transformer area fails, the generated characteristic value of the failure label is input into a failure diagnosis model, and the cause of the failure of the current low-voltage transformer area is diagnosed.
Step six: and according to the determined faults and the diagnosed fault reasons, inquiring a fault and solution comparison table established by expert experience to obtain the solution of the current faults, and sending the solution to corresponding operation and maintenance personnel.
Wherein, the mapping relation between the fault types caused by different reasons and the corresponding fault solutions is established in the fault and solution comparison table.
In this embodiment, for the diagnosed fault, a corresponding solution can be obtained by querying a "fault and solution comparison table", where the solution is generally a standard flow of processing related faults summarized by on-site operation and maintenance personnel according to working experience, and for new technicians, the detailed flow of the solution can be used as a specific working guide, so that they can solve the problem in time, thus reducing the dependency of experience and/or personal ability of the technicians in operation and maintenance management, and achieving a more stable operation and maintenance level.
Typically, different faults in the operation and maintenance process are handled by corresponding technicians, and during operation and maintenance requirement processing, these specific operation and maintenance requirements are also typically sent to specific operation and maintenance management technicians.
Example 3
The embodiment provides an operation and maintenance system suitable for a low-voltage station, which adopts the fault intelligent diagnosis method suitable for the low-voltage station as in embodiment 2, realizes the processes of real-time monitoring, fault sensing and fault cause diagnosis of the low-voltage station, and dispatches corresponding operation and maintenance requirements and solutions to corresponding responsible personnel for processing. As shown in fig. 3, the operation and maintenance system includes: the system comprises a characteristic acquisition module, a fault label generation module, a fault perception module, a fault diagnosis module, a solution inquiry module and an operation and maintenance requirement dispatch module.
The characteristic acquisition module is used for acquiring values of all characteristic quantities related to faults possibly occurring in the low-voltage station area.
The fault label generation module is used for judging whether the value of each characteristic quantity acquired by the characteristic acquisition module is in an abnormal interval, generating a fault label aiming at the characteristic quantity in the abnormal interval, carrying out normalization processing on the characteristic value of the corresponding characteristic quantity, and assigning the value to the fault label.
The fault sensing module is used for judging whether the current low-voltage station area has faults or not according to the values of the characteristic quantities acquired by the characteristic acquisition module; the fault type judged by the fault perception model comprises acquisition faults, metering faults, electricity consumption faults, cost control faults, distribution and transformation faults and abnormal line loss of the transformer area.
The fault diagnosis module is used for acquiring the value of a fault label corresponding to the current fault when the fault diagnosis module judges that the current low-voltage transformer area has the fault, and analyzing and obtaining the fault occurrence reason corresponding to the current fault according to the characteristic value of the fault label.
The solution inquiry module is used for acquiring the faults obtained by the fault perception module and the fault occurrence reasons obtained by the fault diagnosis module; and inquiring a 'fault and solution comparison table' established according to expert experience to obtain the solution of the current fault.
The operation and maintenance requirement dispatching module is used for acquiring a current fault solution obtained by the solution query model; and inquiring an operation and maintenance personnel responsibility table to obtain operation and maintenance personnel corresponding to the current fault, and finally, sending the solution corresponding to the current fault to the operation and maintenance personnel corresponding to the current fault.
In this embodiment, the fault diagnosis module includes an acquisition fault sub-model, a measurement fault sub-model, an electricity consumption fault sub-model, a cost control fault sub-model, a distribution transformer fault sub-model and a transformer area line loss abnormal sub-model; when the fault diagnosis model obtains the fault prediction result judged by the fault perception model, the corresponding sub-models are respectively called to complete the analysis process of fault occurrence reasons corresponding to different fault types.
Wherein, the operation and maintenance personnel responsibility table is a responsibility classification table of responsible personnel established according to the current state of low-voltage station area management; the operation and maintenance personnel responsibility table establishes the correspondence between different fault types and fault occurrence reasons and each operation and maintenance personnel. And the content in the operation and maintenance personnel responsibility table is updated in real time according to the management current situation of the current low-voltage station area. Different operation and maintenance requirements are guaranteed to be directly delivered to corresponding responsible persons.
The operation and maintenance requirement dispatch module in the embodiment can send out corresponding operation and maintenance requirements to the PC end and also send out corresponding operation and maintenance requirements to the mobile end. So as to realize one-key dispatch to a district manager and provide an effective solution and operation guidance; when receiving the corresponding operation and maintenance requirements, the platform manager can process all faults in the platform on site, and the fault processing efficiency is greatly improved.
The technical features of the above-described embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above-described embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples merely represent a few embodiments of the present invention, which are described in more detail and are not to be construed as limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of the invention should be assessed as that of the appended claims.
Claims (9)
1. The intelligent fault diagnosis method suitable for the low-voltage transformer area is characterized by comprising the following steps of:
step one: determining each fault type of a low-voltage station area and a plurality of characteristic quantities associated with each fault type; defining the abnormal characteristic state reflected by each characteristic quantity as a corresponding fault label;
Step two: constructing a fault perception model for identifying the fault type of the low-voltage transformer area and a fault diagnosis model for analyzing the fault occurrence cause; the fault perception model is a label model based on a classification algorithm; the input of the fault perception model is the characteristic value of each characteristic quantity in the running process of the low-voltage transformer area, and the output of the fault perception model is the prediction result of different fault types; the fault diagnosis model is a trained neural network model; the fault diagnosis model comprises a plurality of sub-models corresponding to the fault types; the input of the fault diagnosis model is the characteristic value of each fault label corresponding to the fault prediction result output by the fault perception model, and the output of the fault diagnosis model is the occurrence reason corresponding to the predicted fault;
the construction method of the fault perception model comprises the following steps:
extracting characteristic values of abnormal characteristic states of all fault types in the low-voltage transformer area according to fault labels associated with all fault types, and forming a complete fault image based on the fault labels;
(ii) carrying out manual category distinction and hierarchical relation establishment on various fault labels in the step (i), designing weights of different fault labels in various fault types, defining a calculation formula of the fault labels and the fault types, and forming a prediction model of a training data set;
(iii) collecting in real time feature data of abnormal feature states associated with each of the fault types occurring in the low-voltage transformer area, and taking the feature data as a training data set; carrying out preliminary prediction on the training data set by combining a prediction model and an artificial verification;
(iv) using a Gaussian kernel function as a kernel function of the fault perception model, adopting a one-to-many multi-label fault diagnosis algorithm, and taking characteristic data in the training data set as an input sample to acquire fault types of output, namely, metering fault, power consumption fault, cost control fault, distribution transformer fault and station area line loss abnormality; constructing six support vector machine classifiers for six fault types; constructing a required fault perception model by using six support vector machine classifiers;
(v) setting super parameters of a training process, obtaining a preliminary prediction result obtained through comprehensive judgment of the prediction model and artificial verification, training the constructed fault perception model by utilizing the training data set, comparing the preliminary prediction result with a prediction result output by the fault perception model, and further calculating the minimum error of the fault perception model; until the minimum error requirement of the training stage is met, completing the training process of the fault perception model;
Step three: collecting characteristic data of abnormal characteristic states associated with the fault types in the low-voltage transformer area in real time, and cleaning the characteristic data; generating a corresponding fault label based on the cleaned characteristic data, and completing assignment of the fault label;
step four: inputting the characteristic data cleaned in the previous step into the fault perception model, judging whether a fault exists in the current low-voltage station area, and outputting a corresponding fault prediction result;
step five: when the low-voltage transformer area fails, the generated characteristic value of the failure label is input into the corresponding failure diagnosis model, and the cause of the failure of the current low-voltage transformer area is diagnosed.
2. The intelligent fault diagnosis method suitable for the low-voltage transformer area according to claim 1, wherein the intelligent fault diagnosis method comprises the following steps: in the first step, the fault types of the low-voltage transformer area include acquisition faults, metering faults, power consumption faults, cost control faults, distribution transformer faults and transformer area line loss anomalies.
3. The intelligent fault diagnosis method suitable for the low-voltage transformer area according to claim 2, wherein: in the first step, the correspondence relationship between the abnormal feature states associated with the fault types includes:
(a) An abnormal feature state associated with the acquisition fault, comprising: the terminal is offline, the electricity meter under the concentrator has no data, the collector has no data, and the electric energy meter has no meter reading data for a plurality of days;
(b) An abnormal signature state associated with the metering fault, comprising: the method comprises the steps of flying or suddenly changing an electric energy meter, rewinding the electric energy meter, stopping the electric energy meter indicating value, exceeding the clock tolerance of the electric energy meter, uncovering the electric energy meter, covering an electric energy meter button, zero voltage of the electric energy meter and undervoltage of a battery of the electric energy meter;
(c) An abnormal feature state associated with the power failure, comprising: voltage phase loss, current loss, zero sequence current abnormality, voltage phase interruption, current three-phase imbalance, voltage three-phase imbalance, power factor abnormality and overload;
(d) An abnormal signature status associated with the fee control fault, comprising: carrier overtime, authentication failure, terminal response overtime, denial message, task overtime, issuing overtime times or overtime, uplink message error, ammeter file error, task issuing failure and task environment abnormality;
(e) An abnormal feature state associated with the distribution transformer fault, comprising: current three-phase imbalance, A/B/C phase current loss, A/B/C phase overload, A/B/C phase voltage open-phase, voltage three-phase imbalance, power factor abnormality, low voltage, platform load rate >80%, A/B/C phase overload, A/B/C phase light load and no load;
(f) An abnormal characteristic state associated with the site line loss abnormality, comprising: collecting an operation and maintenance management type state, a platform area metering management type state, a platform area diagnosis analysis type state and a high-loss abnormal type state; wherein the collecting operation and maintenance management class state comprises: the continuous collection fails for more than three days, the meter is in clock error, the electric energy meter flies away and suddenly changes, the voltage is lost, the current is lost, the phase is lost and the overload is carried out; the platform area metering management type state comprises: total surface uncovered, magnification abnormal, reference table abnormal, metering point level abnormal; the platform area diagnosis analysis type state comprises: the three-phase unbalance of the transformer area, the power factor monitoring of the transformer area, the reverse active power being more than 0, the stop of the electric energy meter indicating value, the reverse current, the cover opening of the electric energy meter, the lower limit of the voltage, the zero voltage and the zero live wire out of tolerance; the high-loss exception class state includes: and (5) analyzing the month continuous high loss, the day continuous high loss, the month line loss qualified day line loss unqualified and unmonitored station areas.
4. The intelligent fault diagnosis method suitable for the low-voltage transformer area according to claim 1, wherein the intelligent fault diagnosis method comprises the following steps: the fault diagnosis model is a trained three-layer neural network and comprises an input layer, an hidden layer and an output layer; the number of neurons of the input layer corresponds to the number of fault labels associated with each fault type, and the input characteristic data is the characteristic value of each fault label; the hidden layer adopts a ReLU activation function; the number of the neurons of the output layer is equal to the number of the reasons related to each fault type, the probability of each reason related to the fault type is obtained by the output layer, and the reason with the highest probability among the reasons is used as the output diagnosis result.
5. The intelligent fault diagnosis method for low-voltage transformer areas according to claim 4, wherein in the third step, the fault label generation and assignment process is as follows: judging whether the corresponding acquired characteristic quantity is in an abnormal interval or not, and if so, generating a fault label associated with the current characteristic quantity; and then carrying out normalization processing on the detection value of the current feature quantity, and assigning the normalization processing result to the fault label as the feature value of the fault label.
6. The intelligent fault diagnosis method suitable for the low-voltage transformer area according to claim 5, wherein the intelligent fault diagnosis method comprises the following steps: the intelligent fault diagnosis method further comprises the following steps:
step six: inquiring a fault and solution comparison table established by expert experience according to the determined fault and the diagnosed fault cause to obtain a solution of the current fault, and sending the solution to corresponding operation and maintenance personnel; wherein, the mapping relation between the fault types caused by different reasons and the corresponding fault solutions is established in the fault and solution comparison table.
7. The operation and maintenance system suitable for the low-voltage station area is characterized in that the operation and maintenance system adopts the fault intelligent diagnosis method suitable for the low-voltage station area according to any one of claims 1-6 to realize the processes of real-time monitoring, fault perception and fault cause diagnosis of the low-voltage station area, and sends corresponding operation and maintenance requirements and solutions to corresponding responsible personnel for processing; the operation and maintenance system comprises:
A feature collection module for collecting values of all feature quantities associated with faults that may occur in the low-voltage transformer area;
the fault label generation module is used for judging whether the value of each characteristic quantity acquired by the characteristic acquisition module is in an abnormal interval or not, generating a fault label aiming at the characteristic quantity in the abnormal interval, carrying out normalization processing on the characteristic value of the corresponding characteristic quantity, and assigning the normalized value to the fault label;
the fault sensing module is used for judging whether the current low-voltage station area has faults or not according to the values of the characteristic quantities acquired by the characteristic acquisition module; the fault type judged by the fault perception model comprises acquisition faults, metering faults, power consumption faults, cost control faults, distribution transformer faults and abnormal line loss of the transformer area;
a fault diagnosis module; the fault diagnosis module is used for acquiring the value of the fault label corresponding to the current fault when judging that the current low-voltage transformer area has the fault, and analyzing and obtaining the fault occurrence reason corresponding to the current fault according to the characteristic value of the fault label;
a solution query module; the fault diagnosis module is used for obtaining faults obtained by the fault sensing module and the fault occurrence reasons obtained by the fault diagnosis module; inquiring a fault and solution comparison table established according to expert experience to obtain a current fault solution; and
An operation and maintenance requirement dispatch module; the method is used for acquiring the current fault solution obtained by the solution query model; and inquiring an operation and maintenance personnel responsibility table to obtain operation and maintenance personnel corresponding to the current fault, and sending a solution corresponding to the current fault to the operation and maintenance personnel corresponding to the current fault.
8. The operation and maintenance system for low-voltage transformer areas according to claim 7, wherein the fault diagnosis module comprises an acquisition fault sub-model, a metering fault sub-model, an electricity consumption fault sub-model, a cost control fault sub-model, a distribution transformation fault sub-model and a transformer area line loss abnormal sub-model; when the fault diagnosis model obtains the fault prediction result judged by the fault perception model, the corresponding sub-models are respectively called to complete the analysis process of fault occurrence reasons corresponding to different fault types.
9. The operation and maintenance system for low-voltage transformer areas according to claim 7, wherein the operation and maintenance personnel responsibility table is a responsibility classification table of responsible personnel established according to the current state of low-voltage transformer area management; the operation and maintenance personnel responsibility table is provided with different fault types and corresponding relations between fault occurrence reasons and each operation and maintenance personnel; and the contents in the operation and maintenance personnel responsibility table are updated in real time according to the management current situation of the current low-voltage station area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110694891.5A CN113406438B (en) | 2021-06-23 | 2021-06-23 | Intelligent fault diagnosis method suitable for low-voltage transformer area and operation and maintenance system thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110694891.5A CN113406438B (en) | 2021-06-23 | 2021-06-23 | Intelligent fault diagnosis method suitable for low-voltage transformer area and operation and maintenance system thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113406438A CN113406438A (en) | 2021-09-17 |
| CN113406438B true CN113406438B (en) | 2023-11-24 |
Family
ID=77682471
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110694891.5A Active CN113406438B (en) | 2021-06-23 | 2021-06-23 | Intelligent fault diagnosis method suitable for low-voltage transformer area and operation and maintenance system thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113406438B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113923542B (en) * | 2021-10-28 | 2023-07-11 | 国网江苏省电力有限公司营销服务中心 | Low-voltage station all-link meter reading time delay detection method and system |
| CN114200243B (en) * | 2021-12-24 | 2023-10-24 | 广西电网有限责任公司 | Intelligent diagnosis method and system for faults of low-voltage transformer area |
| CN114528929B (en) * | 2022-02-14 | 2023-06-20 | 国网山东省电力公司营销服务中心(计量中心) | Multi-source data platform region measuring system and method |
| CN115859162B (en) * | 2023-02-21 | 2023-05-05 | 深圳市超业电力科技有限公司 | Power distribution system health diagnosis method and related device based on Internet of things |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108732528A (en) * | 2018-05-28 | 2018-11-02 | 国网福建省电力有限公司电力科学研究院 | A kind of digitalized electrical energy meter method for diagnosing faults based on depth confidence network |
| CN111985561A (en) * | 2020-08-19 | 2020-11-24 | 安徽蓝杰鑫信息科技有限公司 | Fault diagnosis method and system for intelligent electric meter and electronic device |
| CN111985558A (en) * | 2020-08-19 | 2020-11-24 | 安徽蓝杰鑫信息科技有限公司 | Electric energy meter abnormity diagnosis method and system |
| CN112749509A (en) * | 2020-12-30 | 2021-05-04 | 西华大学 | Intelligent substation fault diagnosis method based on LSTM neural network |
-
2021
- 2021-06-23 CN CN202110694891.5A patent/CN113406438B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108732528A (en) * | 2018-05-28 | 2018-11-02 | 国网福建省电力有限公司电力科学研究院 | A kind of digitalized electrical energy meter method for diagnosing faults based on depth confidence network |
| CN111985561A (en) * | 2020-08-19 | 2020-11-24 | 安徽蓝杰鑫信息科技有限公司 | Fault diagnosis method and system for intelligent electric meter and electronic device |
| CN111985558A (en) * | 2020-08-19 | 2020-11-24 | 安徽蓝杰鑫信息科技有限公司 | Electric energy meter abnormity diagnosis method and system |
| CN112749509A (en) * | 2020-12-30 | 2021-05-04 | 西华大学 | Intelligent substation fault diagnosis method based on LSTM neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113406438A (en) | 2021-09-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113406438B (en) | Intelligent fault diagnosis method suitable for low-voltage transformer area and operation and maintenance system thereof | |
| Bialek et al. | Benchmarking and validation of cascading failure analysis tools | |
| CN112131441B (en) | Method and system for rapidly identifying abnormal electricity consumption behavior | |
| CN102436226A (en) | Online monitoring and condition maintenance management system | |
| CN110928866A (en) | Line topology checking method and system based on power failure information big data analysis | |
| CN109873499A (en) | Intelligent distribution area management terminal | |
| Khalyasmaa et al. | High-voltage circuit breakers technical state patterns recognition based on machine learning methods | |
| CN110687473B (en) | Fault positioning method and system for relay protection test of intelligent substation | |
| CN110826228B (en) | Regional power grid operation quality limit evaluation method | |
| CN110311709A (en) | Power information acquisition system fault distinguishing method | |
| CN104091289A (en) | Large-scale power distribution network N-1 rapid verification method based on wiring mode rules | |
| CN110968703B (en) | Method and system for constructing abnormal metering point knowledge base based on LSTM end-to-end extraction algorithm | |
| CN116128467A (en) | Power utilization system maintenance method, device, computer equipment and storage medium | |
| CN108845285B (en) | Electric energy metering device detection method and system | |
| Jiang et al. | Quickest line outage detection and identification: Measurement placement and system partitioning | |
| CN116522746A (en) | Power distribution hosting method for high-energy-consumption enterprises | |
| CN109633381A (en) | A kind of electric network failure diagnosis intelligent analysis method | |
| CN112053086B (en) | Working method of operation diagnosis system based on big data power grid | |
| CN112052620B (en) | Intelligent monitoring and analyzing system for urban distribution transformer area | |
| CN113890013A (en) | Method and system for automatically identifying low-voltage line topology by utilizing network terminal power failure data | |
| Huang et al. | Extracting substation cyber-physical architecture through intelligent electronic devices' data | |
| Wu et al. | Survey on negative line loss rate of transformer region: Rectification measures and challenges | |
| Lingang et al. | Research on integrated calculation method of theoretical line loss of MV and LV distribution Network based on Adaboost integrated learning | |
| CN110261693A (en) | A kind of online thermal test method and system based on stable state sampling technique | |
| Kebe et al. | An experience of detection and classification of Quality-Of-Service problems in MV/LV distribution substations using artificial intelligence: Senegal case study |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |