CN107730117B - Cable maintenance early warning method and system based on heterogeneous data comprehensive analysis - Google Patents
Cable maintenance early warning method and system based on heterogeneous data comprehensive analysis Download PDFInfo
- Publication number
- CN107730117B CN107730117B CN201710966945.2A CN201710966945A CN107730117B CN 107730117 B CN107730117 B CN 107730117B CN 201710966945 A CN201710966945 A CN 201710966945A CN 107730117 B CN107730117 B CN 107730117B
- Authority
- CN
- China
- Prior art keywords
- power cable
- data
- cable data
- neural network
- convolutional neural
- 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
- 238000000034 method Methods 0.000 title claims abstract description 49
- 238000012423 maintenance Methods 0.000 title claims abstract description 48
- 238000004458 analytical method Methods 0.000 title claims abstract description 20
- 238000012545 processing Methods 0.000 claims abstract description 66
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 57
- 238000005065 mining Methods 0.000 claims abstract description 8
- 238000011156 evaluation Methods 0.000 claims description 17
- 230000004913 activation Effects 0.000 claims description 13
- 238000011176 pooling Methods 0.000 claims description 13
- 238000002372 labelling Methods 0.000 claims description 10
- 238000010606 normalization Methods 0.000 claims description 10
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 7
- 238000006467 substitution reaction Methods 0.000 claims description 6
- 238000012351 Integrated analysis Methods 0.000 claims 1
- 238000007689 inspection Methods 0.000 abstract description 4
- 238000004519 manufacturing process Methods 0.000 abstract description 4
- 238000003908 quality control method Methods 0.000 abstract description 3
- 238000009412 basement excavation Methods 0.000 abstract 1
- 230000006870 function Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004870 electrical engineering Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000011158 quantitative evaluation Methods 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 238000004148 unit process Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Physics & Mathematics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Monitoring And Testing Of Transmission In General (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a cable maintenance early warning method based on heterogeneous data comprehensive analysis, which comprises the following steps: processing power cable data containing heterogeneous data to obtain standardized and identified power cable data; performing supervised learning by using the standardized and identified power cable data and adopting a convolutional neural network, internalizing the overhaul decision knowledge into parameters in the convolutional network, and acquiring a trained convolutional neural network model; and analyzing and mining the power cable data acquired in real time by using the trained convolutional neural network model, evaluating the running state of the power cable, and pushing maintenance early warning information. The invention adopts the convolutional neural network framework to carry out supervised learning excavation, utilizes the trained convolutional neural network model to automatically evaluate the running state of the cable and push the maintenance early warning information, can provide technical support for the flexible maintenance work of a power cable operation and inspection unit, and can also provide certain reference for the quality control of a cable production enterprise.
Description
Technical Field
The invention relates to the technical field of power cable operation and maintenance, in particular to a cable overhaul early warning method and system based on heterogeneous data comprehensive analysis.
Background
In recent years, the importance of the power cable in the urban power grid is increasingly remarkable, and the scientific overhaul of the power cable is related to the reliability of the operation of the power grid. The maintenance method adopted in the industry at present mainly depends on classical models in the electrical engineering subject, such as partial discharge experiments, aging analysis and the like, and the time is the main basis for the establishment of a maintenance plan. Because the model is susceptible to noise and has a plurality of uncertain factors, the commonly adopted strategy highly depends on the service quality of the maintainers and has limited reliability. In addition, maintenance strategies such as partial discharge experiments need to be performed off-line and off-line, which affects normal production and life. Recently, a plurality of cable accidents also show that the traditional maintenance mode has problems in the aspects of reliability and maintenance efficiency, so that a maintenance early warning method with more intelligent level is urgently needed.
Disclosure of Invention
The invention provides a cable maintenance early warning method and system based on heterogeneous data comprehensive analysis, and aims to solve the problem of how to carry out maintenance early warning on a power cable.
In order to solve the above problem, according to an aspect of the present invention, there is provided a cable repair warning method based on heterogeneous data comprehensive analysis, the method including:
processing power cable data containing heterogeneous data to obtain standardized and identified power cable data, wherein the power cable data comprises: cable attribute and state data, channel environment data and cable state evaluation data;
performing supervised learning by using the standardized and identified power cable data and adopting a convolutional neural network, internalizing the overhaul decision knowledge into parameters in the convolutional network, and acquiring a trained convolutional neural network model;
and analyzing and mining the power cable data acquired in real time by using the trained convolutional neural network model to acquire probability information of an output layer, evaluating the running state of the power cable according to the probability information, and pushing maintenance early warning information to a manager according to an evaluation result.
Preferably, the processing of the power cable data including heterogeneous data to obtain standardized and identified power cable data includes:
filling power cable data containing heterogeneous data by using a mean filling method;
carrying out continuous processing on the filled power cable data, and converting the discrete power cable data into continuous power cable data by using an integer substitution method;
normalizing the continuous power cable data to obtain normalized power cable data;
and performing self-labeling processing on the normalized power cable data, associating the maintenance data information of the power cable with the normalized power cable data, and acquiring standardized and identified power cable data.
Preferably, the normalizing the continuous power cable data includes:
x_new=(x-x_min)/(x_max-x_min),
wherein, x _ new is the value of the current attribute after normalization processing; x _ min is the minimum value of the current attribute; x _ max is the maximum value of the current attribute; x is the value of the current attribute.
Preferably, the performing supervised learning by using the standardized and identified power cable data by using a convolutional neural network, internalizing the overhaul decision knowledge into parameters in the convolutional network, and acquiring a trained convolutional neural network model, includes:
the method comprises the steps of adopting a convolutional neural network to carry out supervised learning, inputting standardized power cable data with an identifier into an input layer, carrying out convolution, PReLU activation and pooling processing on the standardized power cable data with the identifier for a first preset time threshold, carrying out full-connection processing on the power cable data subjected to the convolution, PReLU activation and pooling processing for a second preset time threshold, outputting a processing result by using a Softmax classifier, internalizing maintenance decision knowledge into parameters in the convolutional network, and obtaining a trained convolutional neural network model.
Preferably, the Adam algorithm is used for optimizing link weight parameters between each processing process in supervised learning by adopting a convolutional neural network.
Preferably, the pushing mode of the overhaul early warning information comprises the following steps: system internal messages, short messages and calls.
According to another aspect of the invention, a cable overhaul early warning system based on heterogeneous data comprehensive analysis is provided, which is characterized by comprising:
the data processing unit is used for processing power cable data containing heterogeneous data to obtain standardized and identified power cable data, wherein the power cable data comprises: cable attribute and state data, channel environment data and cable state evaluation data;
the model acquisition unit is used for performing supervised learning by using the standardized and identified power cable data through a convolutional neural network, internalizing the overhaul decision knowledge into parameters in the convolutional network, and acquiring a trained convolutional neural network model;
and the maintenance early warning unit is used for analyzing and mining power cable data acquired in real time by using the trained convolutional neural network model, acquiring probability information of an output layer, evaluating the running state of the power cable according to the probability information, and pushing maintenance early warning information to managers according to evaluation results.
Preferably, the data processing unit processes the power cable data including the heterogeneous data to obtain standardized and identified power cable data, and includes:
filling power cable data containing heterogeneous data by using a mean filling method;
carrying out continuous processing on the filled power cable data, and converting the discrete power cable data into continuous power cable data by using an integer substitution method;
normalizing the continuous power cable data to obtain normalized power cable data;
and performing self-labeling processing on the normalized power cable data, associating the maintenance data information of the power cable with the normalized power cable data, and acquiring standardized and identified power cable data.
Preferably, the normalizing the continuous power cable data includes:
x_new=(x-x_min)/(x_max-x_min),
wherein, x _ new is the value of the current attribute after normalization processing; x _ min is the minimum value of the current attribute; x _ max is the maximum value of the current attribute; x is the value of the current attribute.
Preferably, the model obtaining unit performs supervised learning by using the normalized and identified power cable data through a convolutional neural network, internalizes the overhaul decision knowledge into parameters in the convolutional network, and obtains the trained convolutional neural network model, including:
the method comprises the steps of adopting a convolutional neural network to carry out supervised learning, inputting standardized power cable data with an identifier into an input layer, carrying out convolution, PReLU activation and pooling processing on the standardized power cable data with the identifier for a first preset time threshold, carrying out full-connection processing on the power cable data subjected to the convolution, PReLU activation and pooling processing for a second preset time threshold, outputting a processing result by using a Softmax classifier, internalizing maintenance decision knowledge into parameters in the convolutional network, and obtaining a trained convolutional neural network model.
Preferably, the Adam algorithm is used for optimizing link weight parameters between each processing process in supervised learning by adopting a convolutional neural network.
Preferably, the pushing mode of the overhaul early warning information comprises the following steps: system internal messages, short messages and calls.
The invention provides a cable maintenance early warning method and system based on heterogeneous data comprehensive analysis, which are characterized in that three heterogeneous data, namely cable attribute and state data, channel environment data and cable state evaluation data, are processed, then a convolutional neural network framework designed aiming at the characteristics of the three types of data is adopted for supervised learning and mining, maintenance decision knowledge is internalized into parameters in a convolutional network, a trained convolutional neural network model is determined, and the trained convolutional neural network model is utilized to automatically carry out uninterrupted quantitative evaluation on the running state of a cable and timely push maintenance early warning information. The method can provide technical support for flexible maintenance work of a power cable operation and inspection unit, and also can provide certain reference for quality control of cable production enterprises.
Drawings
A more complete understanding of exemplary embodiments of the present invention may be had by reference to the following drawings in which:
fig. 1 is a flowchart of a cable repair warning method 100 based on heterogeneous data comprehensive analysis according to an embodiment of the present invention; and
fig. 2 is a schematic structural diagram of a cable repair early warning system 200 based on heterogeneous data comprehensive analysis according to an embodiment of the present invention.
Detailed Description
The exemplary embodiments of the present invention will now be described with reference to the accompanying drawings, however, the present invention may be embodied in many different forms and is not limited to the embodiments described herein, which are provided for complete and complete disclosure of the present invention and to fully convey the scope of the present invention to those skilled in the art. The terminology used in the exemplary embodiments illustrated in the accompanying drawings is not intended to be limiting of the invention. In the drawings, like units/elements are identified with like reference numerals.
Unless otherwise defined, terms (including 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. Further, it will be understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense.
Fig. 1 is a flowchart of a cable repair warning method 100 based on heterogeneous data comprehensive analysis according to an embodiment of the present invention. As shown in fig. 1, in the cable overhaul warning method based on heterogeneous data comprehensive analysis according to the embodiment of the present invention, power cable data including heterogeneous data is processed to obtain standardized and identified power cable data; then, performing supervised learning by using the standardized and identified power cable data and adopting a convolutional neural network, internalizing the overhaul decision knowledge into parameters in the convolutional network, and acquiring a trained convolutional neural network model; and finally, analyzing and mining the power cable data acquired in real time by using the trained convolutional neural network model to push maintenance early warning information to managers, so that technical support can be provided for flexible maintenance work of a power cable operation and inspection unit, and certain reference can be provided for quality control of cable production enterprises. The cable overhaul early warning method 100 based on heterogeneous data comprehensive analysis starts at step 101, and power cable data including heterogeneous data is processed at step 101 to obtain standardized and identified power cable data, where the power cable data includes: cable attribute and status data, channel environment data, and cable status evaluation data.
Preferably, the processing of the power cable data including heterogeneous data to obtain standardized and identified power cable data includes:
filling power cable data containing heterogeneous data by using a mean filling method;
carrying out continuous processing on the filled power cable data, and converting the discrete power cable data into continuous power cable data by using an integer substitution method;
normalizing the continuous power cable data to obtain normalized power cable data;
and performing self-labeling processing on the normalized power cable data, associating the maintenance data information of the power cable with the normalized power cable data, and acquiring standardized and identified power cable data.
Preferably, the normalizing the continuous power cable data includes:
x_new=(x-x_min)/(x_max-x_min),
wherein, x _ new is the value of the current attribute after normalization processing; x _ min is the minimum value of the current attribute; x _ max is the maximum value of the current attribute; x is the value of the current attribute.
In the implementation mode of the invention, three kinds of source data of cable attribute and state data, channel environment data and cable state evaluation data are firstly acquired, wherein the cable attribute and state data comprise power cable running state online monitoring data and cable standing book data, the channel environment data comprise geological conditions, meteorological conditions and surrounding environment data, and the cable state evaluation data comprise cable accident related record information. The three data sources can be directly exported from the power enterprise information system by a power cable operation and inspection department. Because the original data has different formats and ranges, for example: the voltage is a continuous numerical value of 110kV, the current is a continuous numerical value of 1000A, and the accident severity recorded in the cable maintenance record is discrete numerical values of high, medium and low. Before analysis, filling missing values according to an averaging mode, carrying out continuous processing on discrete numerical values, and then sequentially carrying out normalization preprocessing and self-labeling processing on the three data.
The continuous processing is to convert discrete values into continuous values by using an integer replacement method, for example, three levels of high, medium and low can be converted into values 3, 2 and 1.
The normalization process is to normalize different value ranges to the [0,1] interval. A specific normalization method is to calculate using the following formula x _ new ═ x-x _ min)/(x _ max-x _ min), i.e., subtracting the current attribute minimum value from the current attribute value and then dividing by the maximum difference value of the current attribute. For example: if one attribute value is: 3. 2, 1, performing normalization processing on the current value 2, wherein the calculation method comprises the following steps: x _ new ═ (2-1)/(3-1) ═ 0.5.
And finally, self-labeling the normalized numerical value, wherein in the self-labeling process, cable maintenance records in the cable state evaluation information and the normalized data are matched according to record numbers during data acquisition, so that automatic labeling of characteristic data is realized, namely, the maintenance records and corresponding cable attributes are associated with state data and channel environment data according to information such as cable number sections recorded in the maintenance records, and the associated retrieval can simply use SQL retrieval statements based on keywords. The preprocessed data only comprise continuous numerical data, and the numerical data are all in the range of [0,1], so that the subsequent learning and mining are facilitated.
Preferably, in step 102, the standardized and identified power cable data is used for supervised learning by using a convolutional neural network, and the overhaul decision knowledge is internalized into parameters in the convolutional network to obtain a trained convolutional neural network model.
Preferably, the performing supervised learning by using the standardized and identified power cable data by using a convolutional neural network, internalizing the overhaul decision knowledge into parameters in the convolutional network, and acquiring a trained convolutional neural network model, includes:
the method comprises the steps of adopting a convolutional neural network to carry out supervised learning, inputting standardized power cable data with an identifier into an input layer, carrying out convolution, PReLU activation and pooling processing on the standardized power cable data with the identifier for a first preset time threshold, carrying out full-connection processing on the power cable data subjected to the convolution, PReLU activation and pooling processing for a second preset time threshold, outputting a processing result by using a Softmax classifier, internalizing maintenance decision knowledge into parameters in the convolutional network, and obtaining a trained convolutional neural network model.
Preferably, the Adam algorithm is used for optimizing link weight parameters between each processing process in supervised learning by adopting a convolutional neural network.
In the implementation mode of the invention, in consideration of the characteristics of multiple cable data types and complex potential relations, the overall structure of the convolutional network designed in the method is as follows: input → convolution → PReLU → pooling 5 → full connection 4 → Softmax, the input layer receives the preprocessed data directly. In the first convolutional layer, 3 sets of filters are designed to correspond to the features of 3 source data types, each set of filters is connected to only one data type feature, and each set of filters contains 20 filters, thus strengthening targeted filtering and reducing the number of parameters. The activation function of the neurons in the network adopts a PReLU function which has no saturation characteristic and is easy to have better convergence characteristic, the output unit adopts Softmax, and the cost function adopts corresponding Softmax classifier cost. Considering the mass of cable monitoring data, the parameter training algorithm adopts an efficient Adam algorithm to optimize link weight parameters among layers in the network, complete a supervised learning process, internalize maintenance decision knowledge into parameters in a convolutional network, and acquire a trained convolutional neural network model.
Preferably, in step 103, the trained convolutional neural network model is used to analyze and mine power cable data collected in real time, to obtain probability information of an output layer, to evaluate the operation state of the power cable according to the probability information, and to push maintenance early warning information to a manager according to the evaluation result.
Preferably, the pushing mode of the overhaul early warning information comprises the following steps: system internal messages, short messages and calls.
In the embodiment of the invention, the trained convolutional neural network model is utilized to analyze and mine the actual measurement data collected by a certain section of cable, the running state of the cable is evaluated according to the probability information of a Softmax output layer, and the evaluation is marked by five risk levels ABCDE. A indicates the highest risk level and may fail within 1 hour, B24 hours, C72 hours, D one week, and E one month. Based on the danger level, maintenance early warning information is pushed to operation and maintenance units and related responsible persons belonging to different cable sections in various modes of managing system internal information, mobile phone short messages, telephones and the like.
Fig. 2 is a schematic structural diagram of a cable repair early warning system 200 based on heterogeneous data comprehensive analysis according to an embodiment of the present invention. As shown in fig. 2, a cable repair early warning system 200 based on heterogeneous data comprehensive analysis according to an embodiment of the present invention includes: a data processing unit 201, a model acquisition unit 202 and a maintenance early warning unit 203. Preferably, in the data processing unit 201, the power cable data including heterogeneous data is processed to obtain standardized and identified power cable data, wherein the power cable data includes: cable attribute and status data, channel environment data, and cable status evaluation data.
Preferably, the data processing unit 201, for processing the power cable data containing heterogeneous data to obtain standardized and identified power cable data, includes:
filling power cable data containing heterogeneous data by using a mean filling method;
carrying out continuous processing on the filled power cable data, and converting the discrete power cable data into continuous power cable data by using an integer substitution method;
normalizing the continuous power cable data to obtain normalized power cable data;
and performing self-labeling processing on the normalized power cable data, associating the maintenance data information of the power cable with the normalized power cable data, and acquiring standardized and identified power cable data.
Preferably, the normalizing the continuous power cable data includes:
x_new=(x-x_min)/(x_max-x_min),
wherein, x _ new is the value of the current attribute after normalization processing; x _ min is the minimum value of the current attribute; x _ max is the maximum value of the current attribute; x is the value of the current attribute.
Preferably, in the model obtaining unit 202, the standardized and identified power cable data is used to perform supervised learning by using a convolutional neural network, and the overhaul decision knowledge is internalized into parameters in the convolutional network, so as to obtain a trained convolutional neural network model.
Preferably, the model obtaining unit 202, using the standardized and identified power cable data to perform supervised learning by using a convolutional neural network, internalizes the overhaul decision knowledge into parameters in the convolutional network, and obtains the trained convolutional neural network model, including:
the method comprises the steps of adopting a convolutional neural network to carry out supervised learning, inputting standardized power cable data with an identifier into an input layer, carrying out convolution, PReLU activation and pooling processing on the standardized power cable data with the identifier for a first preset time threshold, carrying out full-connection processing on the power cable data subjected to the convolution, PReLU activation and pooling processing for a second preset time threshold, outputting a processing result by using a Softmax classifier, internalizing maintenance decision knowledge into parameters in the convolutional network, and obtaining a trained convolutional neural network model.
Preferably, the Adam algorithm is used for optimizing link weight parameters between each processing process in supervised learning by adopting a convolutional neural network.
Preferably, in the overhaul warning unit 203, the trained convolutional neural network model is used to analyze and mine power cable data acquired in real time, probability information of an output layer is acquired, the operating state of the power cable is evaluated according to the probability information, and overhaul warning information is pushed to a manager according to an evaluation result.
Preferably, the pushing mode of the overhaul early warning information comprises the following steps: system internal messages, short messages and calls.
The cable overhaul warning system 200 based on heterogeneous data comprehensive analysis according to the embodiment of the present invention corresponds to the cable overhaul warning method 100 based on heterogeneous data comprehensive analysis according to another embodiment of the present invention, and details thereof are not repeated herein.
The invention has been described with reference to a few embodiments. However, other embodiments of the invention than the one disclosed above are equally possible within the scope of the invention, as would be apparent to a person skilled in the art from the appended patent claims.
Generally, all terms used in the claims are to be interpreted according to their ordinary meaning in the technical field, unless explicitly defined otherwise herein. All references to "a/an/the [ device, component, etc ]" are to be interpreted openly as referring to at least one instance of said device, component, etc., unless explicitly stated otherwise. The steps of any method disclosed herein do not have to be performed in the exact order disclosed, unless explicitly stated.
Claims (6)
1. A cable maintenance early warning method based on heterogeneous data comprehensive analysis is characterized by comprising the following steps:
the method for processing power cable data containing heterogeneous data to obtain standardized and identified power cable data comprises the following steps: filling power cable data containing heterogeneous data by using a mean filling method; carrying out continuous processing on the filled power cable data, and converting the discrete power cable data into continuous power cable data by using an integer substitution method; normalizing the continuous power cable data to obtain normalized power cable data; performing self-labeling processing on the normalized power cable data, associating overhaul data information of the power cable with the normalized power cable data, and acquiring standardized and identified power cable data; wherein the power cable data comprises: cable attribute and state data, channel environment data and cable state evaluation data;
the standardized power cable data with the identification is utilized to perform supervised learning by adopting a convolutional neural network, the maintenance decision knowledge is internalized into parameters in the convolutional network, and a trained convolutional neural network model is obtained, and the method comprises the following steps: performing supervised learning by adopting a convolutional neural network, inputting the standardized and identified power cable data in an input layer, performing convolution, PReLU activation and pooling processing on the standardized and identified power cable data by using a first preset time threshold, performing full-connection processing on the power cable data subjected to convolution, PReLU activation and pooling processing by using a second preset time threshold, outputting a processing result by using a Softmax classifier, internalizing maintenance decision knowledge into parameters in the convolutional network, and acquiring a trained convolutional neural network model; optimizing link weight parameters among all processing processes in supervised learning by adopting a convolutional neural network by using an Adam algorithm;
and analyzing and mining the power cable data acquired in real time by using the trained convolutional neural network model to acquire probability information of an output layer, evaluating the running state of the power cable according to the probability information, and pushing maintenance early warning information to a manager according to an evaluation result.
2. The method of claim 1, wherein the normalizing the continuous power cable data comprises:
x_new=(x-x_min)/(x_max-x_min),
wherein, x _ new is the value of the current attribute after normalization processing; x _ min is the minimum value of the current attribute; x _ max is the maximum value of the current attribute; x is the value of the current attribute.
3. The method of claim 1, wherein the pushing of the overhaul warning information comprises: system internal messages, short messages and calls.
4. The utility model provides a cable overhauls early warning system based on heterogeneous data integrated analysis which characterized in that, the system includes:
the data processing unit is used for processing power cable data containing heterogeneous data and acquiring standardized and identified power cable data, and comprises: filling power cable data containing heterogeneous data by using a mean filling method; carrying out continuous processing on the filled power cable data, and converting the discrete power cable data into continuous power cable data by using an integer substitution method; normalizing the continuous power cable data to obtain normalized power cable data; performing self-labeling processing on the normalized power cable data, associating overhaul data information of the power cable with the normalized power cable data, and acquiring standardized and identified power cable data; wherein the power cable data comprises: cable attribute and state data, channel environment data and cable state evaluation data;
the model acquisition unit is used for performing supervised learning by using the standardized and identified power cable data and adopting a convolutional neural network, internalizes the overhaul decision knowledge into parameters in the convolutional network, and acquires a trained convolutional neural network model, and comprises: performing supervised learning by adopting a convolutional neural network, inputting the standardized and identified power cable data in an input layer, performing convolution, PReLU activation and pooling processing on the standardized and identified power cable data by using a first preset time threshold, performing full-connection processing on the power cable data subjected to convolution, PReLU activation and pooling processing by using a second preset time threshold, outputting a processing result by using a Softmax classifier, internalizing maintenance decision knowledge into parameters in the convolutional network, and acquiring a trained convolutional neural network model; optimizing link weight parameters among all processing processes in supervised learning by adopting a convolutional neural network by using an Adam algorithm;
and the maintenance early warning unit is used for analyzing and mining power cable data acquired in real time by using the trained convolutional neural network model, acquiring probability information of an output layer, evaluating the running state of the power cable according to the probability information, and pushing maintenance early warning information to managers according to evaluation results.
5. The system of claim 4, wherein the normalizing the continuous power cable data comprises:
x_new=(x-x_min)/(x_max-x_min),
wherein, x _ new is the value of the current attribute after normalization processing; x _ min is the minimum value of the current attribute; x _ max is the maximum value of the current attribute; x is the value of the current attribute.
6. The system of claim 4, wherein the pushing manner of the overhaul warning information comprises: system internal messages, short messages and calls.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710966945.2A CN107730117B (en) | 2017-10-17 | 2017-10-17 | Cable maintenance early warning method and system based on heterogeneous data comprehensive analysis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710966945.2A CN107730117B (en) | 2017-10-17 | 2017-10-17 | Cable maintenance early warning method and system based on heterogeneous data comprehensive analysis |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107730117A CN107730117A (en) | 2018-02-23 |
CN107730117B true CN107730117B (en) | 2021-12-21 |
Family
ID=61211622
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710966945.2A Active CN107730117B (en) | 2017-10-17 | 2017-10-17 | Cable maintenance early warning method and system based on heterogeneous data comprehensive analysis |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107730117B (en) |
Families Citing this family (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108427986A (en) * | 2018-02-26 | 2018-08-21 | 中车青岛四方机车车辆股份有限公司 | A kind of production line electrical fault prediction technique and device |
CN108537415A (en) * | 2018-03-20 | 2018-09-14 | 深圳市泰和安科技有限公司 | A kind of distribution method, the apparatus and system of online safety utilization of electric power |
CN109325537A (en) * | 2018-09-26 | 2019-02-12 | 深圳供电局有限公司 | Power utilization management method and device, computer equipment and storage medium |
CN109359866B (en) * | 2018-10-17 | 2024-05-03 | 塔比星信息技术(深圳)有限公司 | Risk hidden danger monitoring method and device based on leasing equipment and computer equipment |
CN109921515B (en) * | 2019-03-12 | 2020-11-03 | 上海荷福人工智能科技(集团)有限公司 | Comprehensive power distribution management system |
CN111445103B (en) * | 2020-02-25 | 2023-01-31 | 国网河南省电力公司电力科学研究院 | Power transmission cable production quality management feedback system based on industrial internet |
CN113032458A (en) * | 2021-03-23 | 2021-06-25 | 中国人民解放军63920部队 | Method and device for determining abnormality of spacecraft |
CN113657437B (en) * | 2021-07-08 | 2024-04-19 | 中国南方电网有限责任公司 | Power grid overhaul alarm confirmation method and system |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105426908A (en) * | 2015-11-09 | 2016-03-23 | 国网冀北电力有限公司信息通信分公司 | Convolutional neural network based substation attribute classification method |
CN106203741A (en) * | 2016-08-10 | 2016-12-07 | 国家电网公司 | Multi-element heterogeneous Data Cleaning Method for network load prediction |
CN106251059A (en) * | 2016-07-27 | 2016-12-21 | 中国电力科学研究院 | A kind of cable status appraisal procedure based on probabilistic neural network algorithm |
CN106651188A (en) * | 2016-12-27 | 2017-05-10 | 贵州电网有限责任公司贵阳供电局 | Electric transmission and transformation device multi-source state assessment data processing method and application thereof |
CN106844425A (en) * | 2016-12-09 | 2017-06-13 | 国网北京市电力公司 | The data handling system of pipeline and cable |
CN107145675A (en) * | 2017-05-17 | 2017-09-08 | 国网天津市电力公司 | Diagnosing fault of power transformer device and method based on BP neural network algorithm |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9430829B2 (en) * | 2014-01-30 | 2016-08-30 | Case Western Reserve University | Automatic detection of mitosis using handcrafted and convolutional neural network features |
-
2017
- 2017-10-17 CN CN201710966945.2A patent/CN107730117B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105426908A (en) * | 2015-11-09 | 2016-03-23 | 国网冀北电力有限公司信息通信分公司 | Convolutional neural network based substation attribute classification method |
CN106251059A (en) * | 2016-07-27 | 2016-12-21 | 中国电力科学研究院 | A kind of cable status appraisal procedure based on probabilistic neural network algorithm |
CN106203741A (en) * | 2016-08-10 | 2016-12-07 | 国家电网公司 | Multi-element heterogeneous Data Cleaning Method for network load prediction |
CN106844425A (en) * | 2016-12-09 | 2017-06-13 | 国网北京市电力公司 | The data handling system of pipeline and cable |
CN106651188A (en) * | 2016-12-27 | 2017-05-10 | 贵州电网有限责任公司贵阳供电局 | Electric transmission and transformation device multi-source state assessment data processing method and application thereof |
CN107145675A (en) * | 2017-05-17 | 2017-09-08 | 国网天津市电力公司 | Diagnosing fault of power transformer device and method based on BP neural network algorithm |
Also Published As
Publication number | Publication date |
---|---|
CN107730117A (en) | 2018-02-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107730117B (en) | Cable maintenance early warning method and system based on heterogeneous data comprehensive analysis | |
CN110807550B (en) | Distribution transformer overload recognition and early warning method based on neural network and terminal equipment | |
CN110674189B (en) | Method for monitoring secondary state and positioning fault of intelligent substation | |
CN112348339A (en) | Power distribution network planning method based on big data analysis | |
CN111738462B (en) | Fault first-aid repair active service early warning method for electric power metering device | |
CN104410163B (en) | A kind of safety in production based on electric energy management system and power-economizing method | |
CN110516848B (en) | Power equipment maintenance cost optimization method based on survival analysis model | |
CN112488327B (en) | Self-learning power grid equipment fault defect early warning system and method thereof | |
CN107666148B (en) | Line fault studying and judging method based on distribution transformer power failure signal | |
CN103856339A (en) | Method and device for compressing alarm information | |
CN110751338A (en) | Construction and early warning method for heavy overload characteristic model of distribution transformer area | |
CN105548744A (en) | Substation equipment fault identification method based on operation-detection large data and system thereof | |
CN106570567A (en) | Main network maintenance multi-constraint multi-target evaluation expert system and optimization method | |
CN111881961A (en) | Power distribution network fault risk grade prediction method based on data mining | |
CN111680879A (en) | Power distribution network operation toughness evaluation method and device considering sensitive load failure | |
CN115409264A (en) | Power distribution network emergency repair stagnation point position optimization method based on feeder line fault prediction | |
CN115687969A (en) | Low-voltage transformer fault diagnosis method based on sound characteristic analysis | |
CN114548800A (en) | Future-state power grid maintenance risk identification method and device based on power grid knowledge graph | |
CN108062603A (en) | Based on distribution power automation terminal life period of an equipment life-span prediction method and system | |
CN117335570B (en) | Visual monitoring system and method for panoramic information of elastic power distribution network | |
CN105550791A (en) | Railway locomotive maintenance fault management information system | |
CN109523422A (en) | A kind of method for digging of distribution network failure influence factor | |
CN117614137A (en) | Power distribution network optimization system based on multi-source data fusion | |
CN112286987A (en) | Electric power internet of things abnormal alarm compression method based on Apriori algorithm | |
CN111931969A (en) | Merging unit equipment state prediction method based on time sequence analysis |
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 |