CN113837473A - Charging equipment fault rate analysis system and method based on BP neural network - Google Patents
Charging equipment fault rate analysis system and method based on BP neural network Download PDFInfo
- Publication number
- CN113837473A CN113837473A CN202111133536.7A CN202111133536A CN113837473A CN 113837473 A CN113837473 A CN 113837473A CN 202111133536 A CN202111133536 A CN 202111133536A CN 113837473 A CN113837473 A CN 113837473A
- Authority
- CN
- China
- Prior art keywords
- charging equipment
- data
- charging
- fault rate
- equipment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 73
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 25
- 238000003062 neural network model Methods 0.000 claims abstract description 42
- 238000007781 pre-processing Methods 0.000 claims abstract description 26
- 238000010276 construction Methods 0.000 claims abstract description 18
- 238000012549 training Methods 0.000 claims description 38
- 210000002569 neuron Anatomy 0.000 claims description 20
- 238000012360 testing method Methods 0.000 claims description 18
- 238000012545 processing Methods 0.000 claims description 16
- 230000008569 process Effects 0.000 claims description 13
- 238000007619 statistical method Methods 0.000 claims description 10
- 230000006870 function Effects 0.000 claims description 9
- 230000002159 abnormal effect Effects 0.000 claims description 8
- 238000012423 maintenance Methods 0.000 claims description 7
- 238000012986 modification Methods 0.000 claims description 7
- 230000004048 modification Effects 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 5
- 238000004140 cleaning Methods 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- 238000004891 communication Methods 0.000 claims description 3
- 238000004141 dimensional analysis Methods 0.000 claims description 2
- 238000012706 support-vector machine Methods 0.000 description 7
- 238000012544 monitoring process Methods 0.000 description 6
- 238000011161 development Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 229910052799 carbon Inorganic materials 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 230000008030 elimination Effects 0.000 description 2
- 238000003379 elimination reaction Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000006855 networking Effects 0.000 description 2
- 239000003208 petroleum Substances 0.000 description 2
- 230000001737 promoting effect Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000009960 carding Methods 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 239000000295 fuel oil Substances 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000006386 neutralization reaction Methods 0.000 description 1
- 239000003921 oil Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000000007 visual effect Effects 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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- 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
- G06Q10/20—Administration of product repair or maintenance
-
- 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
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Game Theory and Decision Science (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Educational Administration (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention belongs to the field of power grid informatization, and particularly relates to a charging equipment fault rate analysis system and method based on a BP (back propagation) neural network. The data preprocessing module is used for preprocessing data derived from the Internet of vehicles platform; the index system building module comprehensively constructs indexes influencing the fault rate of the charging equipment from multiple dimensions to form a charging equipment fault rate analysis index system; the model construction module constructs multi-dimensional wide-table data of the charging equipment according to the fault rate analysis index system, and constructs and trains a BP neural network model to predict the fault rate of the charging equipment; and the analysis and decision module analyzes factors influencing the failure rate of the charging equipment in a multi-dimensional manner and provides data support for decision making. The invention can fully utilize the multi-dimensional information of the operation of the charging equipment, comprehensively and effectively predict the failure rate of the charging equipment and provide effective support for improving the equipment investment yield.
Description
Technical Field
The invention belongs to the field of power grid informatization, and particularly relates to a charging equipment fault rate analysis system and method based on a BP neural network.
Background
At present, energy and environment become the most concerned problem in the world, and with the development of 'carbon peak reaching' and 'carbon neutralization' new state, the traditional fuel oil automobile is continuously impacted by novel clean energy automobiles as a large household consuming petroleum resources and polluting the environment. The electric automobile replaces oil with electricity, can realize zero emission and low noise, is an important means for relieving the problems of shortage of petroleum resources and serious urban atmospheric pollution, and is an effective carrier for promoting the conversion of traffic development modes.
At present, the research and development force of the electric automobile is continuously increased, and the development of a charging facility is considered while the electric automobile is developed. The construction of an electric vehicle charging station realizes energy conversion between an electric vehicle and a power grid, and is a foundation for promoting the development of the electric vehicle industry. The government provides a strategic target of 'new infrastructure' for the demand, wherein the charging pile is used as a 'gas station' of the electric automobile and is one of important fields of 'new infrastructure'.
Along with the rapid increase of the number of charging piles, the monitoring of the operating state of the charging equipment becomes a problem which has to be paid major attention to, relevant data of the operated charging piles are collected and are subjected to data sorting, screening and summarizing, a corresponding index system is established by combining specific services through a data statistical analysis and calculation method on the basis of data acquisition, a monitoring analysis model and a prediction model are established, the failure rate of the charging equipment is analyzed and predicted, abnormal factors influencing the stable operation of the charging piles are comprehensively analyzed, and the method has important significance for improving the operating efficiency of the equipment and providing better charging service and development for users. However, research in the field of analyzing and predicting failure rates of charging devices is still in a starting stage, most of the prior art performs device failure analysis and prediction based on single device operation data, multi-dimensional information of the charging devices cannot be fully utilized, the charging device failure and failure-causing factors cannot be comprehensively perceived, and relevant research is urgently needed to fill the blank.
Disclosure of Invention
In view of this, the present invention aims to provide a charging device failure rate analysis system based on a BP neural network, so as to solve the problems in the prior art that most of the charging devices perform corresponding failure recognition and prediction based on single hardware device operation data, multi-dimensional information of the operation of the charging devices cannot be fully utilized, the failure rate of the charging devices cannot be comprehensively and effectively predicted, the true cause of the failure rate of the devices cannot be accurately analyzed, the failure rate of the devices is high, and the service quality is poor, thereby improving the usage efficiency of the devices in the charging transaction, reducing the failure rate of the devices and customer complaints, and improving the investment yield of the devices.
The invention further aims to provide a charging equipment fault rate analysis method based on the BP neural network.
The invention relates to a charging equipment fault rate analysis system based on a BP neural network, which comprises:
the data preprocessing module is used for preprocessing related data derived from the Internet of vehicles platform;
the index system building module is used for associating the failure rate of the charging equipment with an actual service scene according to a preprocessing result of the data preprocessing module, comprehensively constructing indexes influencing the failure rate of the charging equipment from multiple dimensions according to actual service conditions and forming a failure rate analysis index system of the charging equipment;
and the model construction module is used for constructing multi-dimensional wide-table data of the charging equipment according to the fault rate analysis index system, constructing and training a BP neural network model to predict the fault rate of the charging equipment, and obtaining a prediction result of the fault rate of the charging equipment.
In a preferred embodiment, the system for analyzing a failure rate of a charging device of the present invention further includes:
and the analysis decision module is used for comprehensively sensing the failure rate of the charging equipment according to the failure rate of the charging equipment predicted by the BP neural network model, perfecting a failure rate analysis index system of the charging equipment by combining factors influencing the failure rate of the charging equipment through multi-dimensional analysis and providing data support for decision making.
In the preferred analysis system of the invention, an index system construction module constructs a derivative index by combining a plurality of dimensions of basic attributes, equipment attributes, self attributes and time characteristics of charging equipment; the data dimension of the constructed fault rate analysis index system comprises the following steps: the method comprises the steps of obtaining the geographical position, the equipment commissioning life, weather factors, charging station types, manufacturers, holidays and statistical indexes, carrying out fuzzy processing on the commissioning life according to data distribution characteristics of the equipment after statistical analysis is carried out on the equipment commissioning life, and carrying out fuzzy processing on the charging station according to the charging station types. The BP neural network model comprises an input layer, a hidden layer and an output layer, wherein neurons of adjacent layers are fully interconnected, neurons of the same layer are not connected, the number of nodes of the input layer is consistent with the data dimension in a fault rate analysis index system, the number of the neurons of the hidden layer is determined by a traversal method according to the mean square error of a test set and a training set, and the number of the output nodes is 1.
The invention relates to a charging equipment fault rate analysis method based on a BP neural network, which comprises the following steps:
s1, preprocessing data such as charging pile information and charging pile maintenance derived from the Internet of vehicles platform, including data cleaning, data specification, data missing value filling and data error format modification;
s2, according to the preprocessing result, the failure rate of the charging equipment is associated with a specific business process, various indexes which influence the failure rate of the charging equipment are comprehensively depicted and constructed by combining a data statistical analysis method according to the actual business situation, and a failure rate analysis index system of the charging equipment is formed;
s3, constructing corresponding multi-dimensional wide-table data according to the charging equipment fault rate analysis index system, and constructing a BP neural network model to predict the charging equipment fault rate.
In a preferred embodiment, the method for analyzing the failure rate of the charging device further includes the following steps:
s4, according to the prediction result of the BP neural network model, namely the predicted failure rate of the charging equipment, the failure rate of the charging equipment is comprehensively sensed, factors influencing the failure rate of the charging equipment are analyzed in a multi-dimensional mode, an analysis index system of the failure rate of the charging equipment is perfected, and data support is provided for decision making.
Preferably, the BP neural network model constructed in step S3 includes an input layer, a hidden layer, and an output layer, neurons in adjacent layers are all interconnected, neurons in the same layer are not connected, the number of nodes in the input layer is consistent with the data dimension in the fault rate analysis index system, the number of neurons in the hidden layer is determined by an traversal method according to the mean square error of the test set and the training set, and the number of output nodes is 1.
Further preferably, the size of the network structure of the BP neural network model is (d, q, l), where d is the number of nodes of the input layer, q is the number of nodes of the hidden layer, and l is the number of nodes of the output layer; the construction process of the BP neural network model comprises the following steps:
301, constructing sample data of a corresponding relation between wide table data and a fault rate of the charging equipment, carrying out normalization processing on the sample data to form a fault rate sample set, and dividing the fault rate sample set into a training set and a test set according to a proportion, wherein the training set is
xk∈Rn,yk∈R，(xk,yk) Inputting training samples, wherein m is the total number of samples, and R represents a real number space;
step 302, respectively giving different hidden layer node numbers q to determine different network structures, and respectively performing step 303-308 on the networks with different hidden layer node numbers;
step 303, setting each weight and threshold of each network for initialization,
(1) calculating a scaling factor:
(2) initializing any layerWeight omegaijIs [ -0.5,0.5 [)]A random value in between;
(3) reinitializing the weight ωij：
(4) For the ith neuron in the hidden layer, set the bias to one [ - ωij,ωij]A random value in between;
setting an activation function f, wherein neurons of the hidden layer and the output layer use Sigmoid functions;
step 304, taking the multidimensional wide table data corresponding to the fault rate as the input of a neural network, obtaining output data after weighting processing of a hidden layer and an output layer, and calculating the error square sum of expected output;
step 305, calculating gradient items of weights and thresholds among layers of the hidden layer and the output layer;
step 306, correcting weights and thresholds among all layers of the hidden layer and the output layer;
step 307, after all samples in the failure rate sample set are subjected to the step 304-306, calculating a performance index;
step 308, if the performance index meets the precision requirement, the training is finished, otherwise, the step 304 is carried out to continue the next training period;
and 309, determining the number of the nodes of the final hidden layer according to the mean square errors of the test set and the training set to obtain a trained BP neural network model.
Compared with the prior art, the invention has the following advantages and beneficial effects:
the method analyzes factors influencing the failure rate of the equipment from different dimensions, constructs an index system influencing the failure rate of the charging equipment by calculation, constructs a BP neural network model on the basis, analyzes and predicts the failure rate of the charging equipment on the basis of the BP neural network, perfects the failure rate analysis system of the charging equipment on the basis of the predicted failure rate of the charging equipment, solves the problem of incomplete sensing of the failure rate of the charging equipment, improves the use efficiency of the equipment in the subsequent charging transaction, reduces the failure rate of the equipment and customer complaints, supports service management decisions, improves the quality of power supply service and improves the satisfaction degree of customers.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a schematic structural diagram of a charging device fault rate analysis system based on a BP neural network in an embodiment of the present invention;
fig. 2 is a flowchart of a charging device fault rate analysis method based on a BP neural network in the embodiment of the present invention.
Detailed Description
In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are a part of the embodiments of the present invention, but not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Examples
Referring to fig. 1, the charging device failure rate analysis system based on the BP neural network of the present embodiment includes a data preprocessing module 100, an index system building module 200, a model building module 300, and an analysis decision module 400. The analysis system of the embodiment can evaluate the failure rate of the charging equipment, before related data acquired by the internet of vehicles platform are deeply mined, data exploration and analysis, data missing value filling, data error format modification and data abnormal value elimination are performed through the data preprocessing module, a data base is provided for subsequent model construction through the construction of an index system construction module on a derivative index, and then the failure rate prediction of the charging equipment is analyzed and predicted through the analysis decision module.
The analysis system data collection of this embodiment selects the charging network operation data, the stake information of charging, fill data such as stake maintenance of the ji north area that derive from the car networking platform, and the relevant data that the car networking platform derived in the data preprocessing module promptly includes charging network operation data, fills the stake information, fills data such as stake maintenance. Since the north charging station is put into operation in large scale in 2016-2019, and the influence of epidemic situations on charging data is also considered, in order to make the model more accurate, in this embodiment, data from month 1 in 2019 to month 12 in 2019 is selected, a charging station with the number of days of operation greater than 60 is selected, and a middle database is adopted to perform data caching, and the flow of source configuration, data extraction and warehousing of the source service system is developed according to the ETL mode, so as to perform the whole-table import of the source service table.
The data preprocessing module 100 performs data cleaning, data stipulation, data abnormal value repair and data missing value filling on charging network operation data, charging pile information, charging pile maintenance data and charging transaction data which are derived from the internet of vehicles platform. In other words, the data preprocessing module is used for carrying out exploration analysis, filling of missing data values, modification of data error formats and elimination of abnormal data values on the original data. Missing value filling is also called null value repairing, and is mainly used for processing data missing caused by communication, computer crash or other reasons in the data exploration process. If the name of the charging station in the charging transaction data is lost, the name of the charging station is filled by associating the charging pile number in the charging transaction data with a charging pile information table; and the transaction time and the number of the charging pile are lost, and the transaction serial number in the charging transaction data comprises the number of the charging pile and the transaction time, and the transaction serial number is analyzed to be filled.
Further, the data preprocessing module also gives a definition of the failure rate of the charging device: defining abnormal power failure, charging equipment failure, TCU (automatic transmission control unit) failure or BMS (battery management system) communication abnormal failure in the transaction end reasons as failure, and defining the failure times divided by the total transaction times as the failure rate of the charging equipment; and calculating the equipment failure rate of the charging station according to months, and deleting the monthly charging times less than 10 and the monthly charging electric quantity less than 50 kwh. Meanwhile, the carding charging equipment analyzes scenes in the aspects of equipment asset profile, operation region, asset life, fault profile, manufacturer, operation environment and the like, and extracts key service indexes from dimensional information such as basic attribute, equipment attribute, self attribute, time characteristic and the like. The data error format modification includes modifying a time format error that results from the charging transaction data being exported from the internet of vehicles foreground.
The index system building module 200 associates the failure rate of the charging device with an actual service scene according to a preprocessing result of the data preprocessing module on the basis of the given failure rate of the charging device, comprehensively constructs indexes affecting the failure rate of the charging device from multiple dimensions according to actual service conditions, forms a charging device failure rate analysis index system, comprehensively describes the failure rate of the field device, and is applied to building a subsequent model. Specifically, the index system construction module constructs a derivative index by combining multiple dimensions such as basic attributes, equipment attributes, self attributes and time characteristics of the charging equipment; the data dimension of the constructed fault rate analysis index system comprises the following steps: geographical position, equipment operation life, weather factor, charging station type, manufacturer, holidays, statistical indexes and the like, wherein the indexes are introduced as follows:
geographic location: longitude and latitude were chosen for the measurement.
Weather factors: weather is classified into 3 categories according to the IEEE standards: normal weather, severe weather, heavy disaster weather, and statistics of the proportion of the normal weather and the severe weather. Most weather conditions are classified as both normal and severe because of the minimal chance of catastrophic weather occurrence. The weather with small influence on the charging equipment is normal weather, the weather with large influence on the charging equipment is classified into general severe weather such as thunderstorm, strong wind, ice, snow, rain, fog and the like, the proportion of the normal weather and the severe weather is counted, and weather factor indexes are constructed.
Equipment commissioning life: after statistical analysis is carried out on the equipment commissioning life, fuzzy processing is carried out on the commissioning life according to the data distribution characteristics, and the fuzzification rule is as follows: the operation year is blurred to 1 when the operation year is less than 0.5, 2 when the operation year is not less than 0.5 and less than 1, 3 when the operation year item is more than 1 and less than 1.5, and 4 when the operation year item is more than 1.5.
Charging station type: the charging infrastructure mainly includes two types: centralized charging station and distributed electric pile that fills. Wherein, centralized charging station: the charging device is mainly used for charging special vehicles, and is mainly applied to buses, taxies, environmental sanitation logistics, urban public, intercity quick charging and the like; the centralized charging station consists of four systems, namely a power distribution system, a charging system, a monitoring system, civil engineering and the like, wherein the power distribution system comprises a power distribution cabinet, a transformer, active filtering, a direct-current power supply and a metering cabinet, and the charging system comprises a charging pile and a cable; the monitoring system mainly ensures the charging safety and charging functions and comprises four aspects of current and voltage monitoring, power distribution protection, security monitoring and charging system. Distributed charging pile: the electric pile is a consumer electric pile special for users, and mainly solves the problem of difficult endurance of the electric passenger car. In this embodiment, the charging station is fuzzified according to the charging station type, and the fuzzification rule is as follows: the centralized charging station is fuzzified to be 1, and the distributed charging pile is fuzzified to be 2.
Festival and holiday: through analyzing the existing charging transaction records, the fact that the highest average daily charging electric quantity is a holiday, the next is a holiday, and the least is a working day is found; the average daily charging amount of the holiday is more than 2 times of the average daily charging amount of the working day, and the average daily charging amount of the holiday is close to 1.5 times of the average daily charging amount of the working day, so that the influence of the holidays on the charging capacity is shown, and the charging frequency has direct relevance on the charging equipment. The day ratios of the festival, the holiday and the working day are respectively calculated.
In order to avoid the influence on the effectiveness and the convergence rate of subsequent data mining due to the large difference of the dimension of the independent variable in the data, the data is normalized and limited to be within the range of 0-1, in this embodiment, zero-mean normalization is adopted, and the conversion formula is as follows:
in the formula (I), the compound is shown in the specification,
is the mean of the original data (i.e. the mean of the processed data), σiIs the standard deviation of the original data (i.e., the standard deviation of the processed data). After the above formula processing, the mean value of the data is 0 and the standard deviation is 1.
The model building module 300 builds multi-dimensional wide-table data of the charging device according to the fault rate analysis index system, and builds and trains a BP (Back propagation) neural network model; and the trained BP neural network model is used for predicting the failure rate of the charging equipment to obtain the prediction result of the failure rate of the charging equipment.
On the basis of the constructed index system, in order to evaluate the effect of the BP neural network model, a model construction module divides data into a training set and a test set according to a ratio of 4:1, and constructs and trains the BP neural network model by using the training set; and predicting the failure rate of the charging equipment in the test set by using the trained BP neural network model.
The model construction module also carries out information statistics on the modeling process of the index data from different angles to obtain corresponding statistical reports, wherein the statistical reports comprise a data quality report, a data preprocessing result report, an index system distribution characteristic report, a model result report and an analysis strategy report.
The analysis decision module 400 comprehensively senses the failure rate of the charging equipment according to the failure rate of the charging equipment predicted by the BP neural network model, improves a failure rate analysis index system of the charging equipment by combining factors influencing the failure rate of the charging equipment through multidimensional analysis, provides data support for decision making, solves the problem of incomprehensive sensing of the failure rate of the charging equipment, supports service management decisions, improves the utilization rate and the power supply service quality of the charging equipment, and improves the overall satisfaction degree of customers
Specifically, the analysis decision module analyzes factors influencing the failure rate of the charging equipment according to the multi-dimensional statistics of the prediction result, wherein the factors comprise the dimensions of equipment installation area distribution (namely geographic position), equipment operation life, equipment type, weather conditions, failure types, holidays, manufacturers and the like, a perfect charging equipment failure rate state detection and evaluation system is constructed, the specific dimensions and reasons causing the failure rate of the charging equipment are efficiently and accurately analyzed, data support is provided for the decision of the related work of a subsequent business department, the charging service quality is purposefully improved, the equipment operation efficiency is effectively improved, the equipment operation and maintenance service system is perfected, corresponding suggestions are provided for the actual work, and the equipment operation reliability and the customer charging service satisfaction are improved.
The multidimensional statistical analysis of the analysis decision module is used for carrying out information statistics on the modeling process of the index data from different angles to obtain a corresponding statistical report. The statistical report comprises a data quality report, a data preprocessing result report, an index system distribution characteristic report, a model result report and an analysis strategy report. The statistical analysis process specifically performs information statistics on the modeling process of the index data from different angles according to various daily report information of the charging network equipment fault rate analysis system based on the BP neural network model to obtain corresponding statistical reports, so that the statistical requirements of different business departments can be met, and a basis is provided for subsequent information analysis work.
Based on the same inventive concept, the embodiment further provides a charging device fault rate analysis method based on the BP neural network, which includes the following steps:
s1, preprocessing data such as charging pile information and charging pile maintenance derived from the Internet of vehicles platform, including data cleaning, data specification, data missing value filling and data error format modification;
s2, according to the preprocessing result, the failure rate of the charging equipment is associated with a specific business process, various indexes which influence the failure rate of the charging equipment are comprehensively depicted and constructed by combining a data statistical analysis method according to the actual business situation, and a failure rate analysis index system of the charging equipment is formed;
s3, constructing corresponding multi-dimensional wide-table data according to the charging equipment fault rate analysis index system, and constructing a BP neural network model to predict the charging equipment fault rate;
s4, according to the prediction result of the BP neural network model, namely the predicted failure rate of the charging equipment, the failure rate of the charging equipment is comprehensively sensed, factors influencing the failure rate of the charging equipment are analyzed in a multi-dimensional mode, an analysis index system of the failure rate of the charging equipment is perfected, and data support is provided for decision making.
In this embodiment, the BP neural network model constructed in step S3 is widely applied to solving the nonlinear problem as a method for simulating human intelligence and visual thinking ability. The neural network extracts the characteristics of the input mode through learning the input mode, thereby achieving the memory of the input mode. The most mature and widely applied artificial neural network model in the current research is the BP neural network, which adopts the neural network of the error back propagation learning algorithm and is a supervised learning algorithm. The BP neural network comprises three layers, namely an input layer, a hidden layer and an output layer, wherein neurons in adjacent layers are fully interconnected, and neurons in the same layer are not connected; the number of the input layer nodes is consistent with the data dimension in the fault rate analysis index system. In this embodiment, because the output result is the failure rate, the number of output nodes is 1, and in the construction process of the BP neural network model, the most important thing is to determine the number of nodes of the hidden layer, the number of nodes of the hidden layer may be more or less, but too many may cause overfitting, and too few may cause poor fitting. The specific construction process of the BP neural network model is described below by taking a given network structure size as an example.
Training set
k∈Rn,kE is R, m is the total number of samples, and R represents a real number space; the network structure size of the three-layer BP neural network model is (d, q, l), and in this embodimentThe number d of nodes of the middle input layer is 5, the number q of nodes of the hidden layer is selected by adopting a traversal method, and the number l of nodes of the output layer is 1; wherein the input layer node is represented by xiRepresenting, hidden layer nodes by bhIndicating that the output layer node is represented by yjRepresenting, input layer to hidden layer node weights by vihExpressed, the threshold is expressed by gamma h, and the weight of the node from the hidden layer to the output layer is whjThreshold value of betaj(ii) a The learning rate η is 0.1 and the regularization parameter λ. The specific construction process comprises the following steps:
step 301, constructing sample data of a corresponding relation between wide table data and a fault rate of the charging equipment, performing normalization processing on the sample data to form a fault rate sample set, and performing normalization processing on the fault rate sample set according to a ratio of 4:1 into a training set and a test set, wherein the training set
xk∈Rn,yk∈R。
Step 302, respectively giving different hidden layer node numbers q to determine different network structures, and respectively performing step 303-308 on networks with different hidden layer node numbers.
Step 303, setting each weight and threshold of the neural network model according to an Nguyen-Widrow method, and initializing;
(1) calculating a scaling factor:
wherein d is the number of nodes of the input layer, and q is the number of nodes of the hidden layer;
(2) initializing the weight omega of any layerijIs [ -0.5,0.5 [)]A random value in between;
(3) the weight is reinitialized according to the following formula:
(4) for the ith neuron in the hidden layer, set the bias to one [ - ωij,ωij]A random value in between.
Setting an activation function f, wherein neurons of the hidden layer and the output layer use Sigmoid functions;
and step 304, taking the multidimensional wide table data corresponding to the fault rate as the input of the neural network, obtaining output data after weighting processing of the hidden layer and the output layer, and calculating the error square sum of expected output.
For the current input training sample (x)k,yk) E, calculating the output of each network layer of the current input training sample:
bh＝f(γh-θh) Wherein
Calculating the sum of the squared errors of the output data and the expected output:
step 305, calculating gradient items of weights and thresholds between layers of the hidden layer and the output layer:
Step 306, correcting weights and thresholds between the hidden layer and each output layer:
whj＝whj+Δwhjthe value is the output layer weight;
θj＝θj+Δθjoutput layer threshold;
vih＝vih+Δvihthe weight value is hidden layer weight value;
γh＝γh+Δγhthe hidden layer threshold is used.
Step 307, after all samples in the failure rate sample set go through step 304-306, a training period (Epoch) is completed, and after the training period Epoch is completed, the performance index is calculated:
the subscript k here indicates the mean square error value corresponding to the input kth sample.
And 308, if the performance index meets the precision requirement, namely E is less than or equal to epsilon, ending the training, otherwise, turning to the step 304 and continuing the next training period. Epsilon is a positive number used to measure the accuracy requirement, depending on the actual situation.
And 309, determining the number of nodes of the final hidden layer according to the mean square errors of the test set and the training set to obtain the trained BP neural network model. And predicting the failure rate of the charging equipment by using the trained improved BP neural network model.
Based on the description, the traversal method is adopted during model construction, the number of hidden layer nodes of the BP network is determined according to the mean square error of the test set and the training set, the Nguyen-Widrow initialization algorithm is adopted during model training to improve the training speed of the network, and meanwhile, the square sum term of the weight is added in the target error function to inhibit the overfitting phenomenon of the BP neural network.
After the model is established based on the index system and the corresponding multi-dimensional wide table data, in order to make a BP neural network model and a Support Vector Machine (SVM) model have comparability so as to select the model most suitable for predicting the equipment failure rate, the data are divided into a training set and a test set according to a ratio of 4:1, and the BP neural network model is established by using the training set. For predicting the failure rate of the charging device in the test set.
And selecting a standard mean square error and a mean square error as model evaluation standards, wherein the standard mean square error is the mean square error of the predicted fault rate divided by the mean square error of the actual fault rate, and is the comparison of the fitting value of the dependent variable of the model and the original value of the dependent variable, when the standard mean square error is more than 1, the result shows that the fault rate predicted by the model is not as good as the mean effect of the actual fault rate, and the mean square error reflects the fitting effect of the model, and the smaller the mean square error is, the closer the fault rate predicted by the model is to the actual fault rate is. And defining the qualified prediction records as the difference between the predicted fault rate and the actual fault rate of 0.1 according to the fault rate distribution, and the accuracy of model prediction as the number of the qualified prediction records divided by the total number.
TABLE 1 analytical results
The analysis results in table 1 represent support vector machine regression, BP neural network regression, and random sampling mean square error and normalized error divided into training and test sets in a 4:1 ratio. The standard mean square error of a training set and a test set of a support vector machine model of a radial kernel function and a BP neural network model is smaller than 1, which shows that the mean effect of the model prediction on the failure rate is better than that of the actual failure rate; the mean square error of the test set and the training set of the BP neural network model is smaller than that of the support vector machine, which shows that the fitting capability of the neural network model to the failure rate is better than that of the support vector machine, and the qualification rate of the model is higher than that of the support vector machine, so that the neural network model is selected.
From the above description of the embodiments, it is clear to those skilled in the art that the present invention can be implemented by software plus necessary general hardware platform. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which may be stored in a storage medium, such as a ROM/RAM, an optical disc, etc., and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method according to the embodiments or some parts of the embodiments of the present application.
The foregoing represents a preferred embodiment of the present invention and it will be appreciated by those skilled in the art that modifications and adaptations to those embodiments may be made without departing from the principles of the invention and are intended to be within the scope of the invention.
Claims (10)
1. Charging equipment fault rate analysis system based on BP neural network, its characterized in that includes:
the data preprocessing module is used for preprocessing related data derived from the Internet of vehicles platform;
the index system building module is used for associating the failure rate of the charging equipment with an actual service scene according to a preprocessing result of the data preprocessing module, comprehensively constructing indexes influencing the failure rate of the charging equipment from multiple dimensions according to actual service conditions and forming a failure rate analysis index system of the charging equipment;
and the model construction module is used for constructing multi-dimensional wide-table data of the charging equipment according to the fault rate analysis index system, constructing and training a BP neural network model to predict the fault rate of the charging equipment, and obtaining a prediction result of the fault rate of the charging equipment.
2. The charging device failure rate analysis system of claim 1, further comprising:
and the analysis decision module is used for comprehensively sensing the failure rate of the charging equipment according to the failure rate of the charging equipment predicted by the BP neural network model, perfecting a failure rate analysis index system of the charging equipment by combining factors influencing the failure rate of the charging equipment through multi-dimensional analysis and providing data support for decision making.
3. The charging device failure rate analysis system of claim 1, wherein a charging device failure rate is defined as: and defining abnormal power failure, charging equipment failure, TCU failure or BMS communication abnormal failure in the transaction end reasons as failures, and defining the failure times divided by the total transaction times as the failure rate of the charging equipment.
4. The charging equipment fault rate analysis system of claim 1, wherein the index system construction module constructs a derivative index by combining multiple dimensions of basic attributes, equipment attributes, self attributes and time characteristics of the charging equipment; the data dimension of the constructed fault rate analysis index system comprises the following steps: the method comprises the steps of obtaining the geographical position, the equipment commissioning life, weather factors, charging station types, manufacturers, holidays and statistical indexes, carrying out fuzzy processing on the commissioning life according to data distribution characteristics of the equipment after statistical analysis is carried out on the equipment commissioning life, and carrying out fuzzy processing on the charging station according to the charging station types.
5. The charging device fault rate analysis system of claim 1, wherein the BP neural network model comprises an input layer, a hidden layer and an output layer, neurons in adjacent layers are fully interconnected, neurons in the same layer are not connected, the number of nodes in the input layer is consistent with the data dimension in a fault rate analysis index system, the number of the neurons in the hidden layer is determined by a traversal method according to the mean square error of the test set and the training set, and the number of the output nodes is 1.
6. The charging equipment fault rate analysis method based on the BP neural network is characterized by comprising the following steps of:
s1, preprocessing data such as charging pile information and charging pile maintenance derived from the Internet of vehicles platform, including data cleaning, data specification, data missing value filling and data error format modification;
s2, according to the preprocessing result, the failure rate of the charging equipment is related to a specific business process, and according to the actual business situation and a data statistical analysis method, various indexes which influence the failure rate of the charging equipment are constructed and comprehensively depicted, so that a failure rate analysis index system of the charging equipment is formed;
s3, constructing corresponding multi-dimensional wide-table data according to the charging equipment fault rate analysis index system, and constructing a BP neural network model to predict the charging equipment fault rate.
7. The charging device fault rate analysis method of claim 6, further comprising the steps of:
s4, according to the prediction result of the BP neural network model, namely the predicted failure rate of the charging equipment, the failure rate of the charging equipment is comprehensively sensed, factors influencing the failure rate of the charging equipment are analyzed in a multi-dimensional mode, an analysis index system of the failure rate of the charging equipment is perfected, and data support is provided for decision making.
8. The charging equipment fault rate analysis method according to claim 6, wherein step S2 is to construct a derivative index by combining multiple dimensions of basic attributes, equipment attributes, self attributes and time characteristics of the charging equipment; the data dimension of the constructed fault rate analysis index system comprises the following steps: the method comprises the steps of obtaining the geographical position, the equipment commissioning life, weather factors, charging station types, manufacturers, holidays and statistical indexes, carrying out fuzzy processing on the commissioning life according to data distribution characteristics of the equipment after statistical analysis is carried out on the equipment commissioning life, and carrying out fuzzy processing on the charging station according to the charging station types.
9. The charging device fault rate analysis method according to claim 6, wherein the BP neural network model constructed in step S3 includes an input layer, a hidden layer, and an output layer, neurons in adjacent layers are all interconnected, neurons in the same layer are not connected, the number of nodes in the input layer is consistent with the data dimension in the fault rate analysis index system, the number of neurons in the hidden layer is determined by a traversal method according to the mean square error of the test set and the training set, and the number of output nodes is 1.
10. The charging device fault rate analysis method according to claim 9, wherein the size of the network structure of the BP neural network model is (d, q, l), where d is the number of nodes of the input layer, q is the number of nodes of the hidden layer, and l is the number of nodes of the output layer; the construction process of the BP neural network model comprises the following steps:
301, constructing sample data of a corresponding relation between wide table data and a fault rate of the charging equipment, carrying out normalization processing on the sample data to form a fault rate sample set, and dividing the fault rate sample set into a training set and a test set according to a proportion, wherein the training set is
(xk,yk) Inputting training samples, wherein m is the total number of samples, and R represents a real number space;
step 302, respectively giving different hidden layer node numbers q to determine different network structures, and respectively performing step 303-308 on the networks with different hidden layer node numbers;
step 303, setting each weight and threshold of each network for initialization,
(1) calculating a scaling factor:
(2) initializing the weight omega of any layerijIs [ -0.5,0.5 [)]A random value in between;
(3) reinitializing the weight ωij：
(4) For the ith neuron in the hidden layer, set the bias to one [ - ωij,ωij]A random value in between;
setting an activation function f, wherein neurons of the hidden layer and the output layer use Sigmoid functions;
step 304, taking the multidimensional wide table data corresponding to the fault rate as the input of a neural network, obtaining output data after weighting processing of a hidden layer and an output layer, and calculating the error square sum of expected output;
step 305, calculating gradient items of weights and thresholds among layers of the hidden layer and the output layer;
step 306, correcting weights and thresholds among all layers of the hidden layer and the output layer;
step 307, after all samples in the failure rate sample set are subjected to the step 304-306, calculating a performance index;
step 308, if the performance index meets the precision requirement, the training is finished, otherwise, the step 304 is carried out to continue the next training period;
and 309, determining the number of the nodes of the final hidden layer according to the mean square errors of the test set and the training set to obtain a trained BP neural network model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111133536.7A CN113837473A (en) | 2021-09-27 | 2021-09-27 | Charging equipment fault rate analysis system and method based on BP neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111133536.7A CN113837473A (en) | 2021-09-27 | 2021-09-27 | Charging equipment fault rate analysis system and method based on BP neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113837473A true CN113837473A (en) | 2021-12-24 |
Family
ID=78970565
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111133536.7A Pending CN113837473A (en) | 2021-09-27 | 2021-09-27 | Charging equipment fault rate analysis system and method based on BP neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113837473A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114814408A (en) * | 2022-03-23 | 2022-07-29 | 国网北京市电力公司 | Method, device and equipment for monitoring running state of charging module and storage medium |
| CN115330281A (en) * | 2022-10-14 | 2022-11-11 | 成都秦川物联网科技股份有限公司 | Smart city new energy automobile charging service method, system, device and medium |
| CN116167200A (en) * | 2022-12-05 | 2023-05-26 | 上海蓝箭电控设备南通有限公司 | Service life detection method and device for power distribution cabinet |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106470112A (en) * | 2015-08-17 | 2017-03-01 | 中兴通讯股份有限公司 | The management method of charger and device |
| CN106650022A (en) * | 2016-11-25 | 2017-05-10 | 天津津航计算技术研究所 | Method for predicting fault of complex electronic device |
| CN109558979A (en) * | 2018-11-29 | 2019-04-02 | 广东电网有限责任公司 | Power equipment defect prediction method and device |
| CN112052966A (en) * | 2020-09-24 | 2020-12-08 | 佰聆数据股份有限公司 | Power customer satisfaction analysis system and method based on site emergency repair work order |
| WO2021028615A1 (en) * | 2019-08-15 | 2021-02-18 | Liikennevirta Oy / Virta Ltd | Charging station monitoring method and device |
| KR20210068164A (en) * | 2019-11-29 | 2021-06-09 | (주)제주전기자동차서비스 | System and method for diagnosing electric vehicle charger |
-
2021
- 2021-09-27 CN CN202111133536.7A patent/CN113837473A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106470112A (en) * | 2015-08-17 | 2017-03-01 | 中兴通讯股份有限公司 | The management method of charger and device |
| CN106650022A (en) * | 2016-11-25 | 2017-05-10 | 天津津航计算技术研究所 | Method for predicting fault of complex electronic device |
| CN109558979A (en) * | 2018-11-29 | 2019-04-02 | 广东电网有限责任公司 | Power equipment defect prediction method and device |
| WO2021028615A1 (en) * | 2019-08-15 | 2021-02-18 | Liikennevirta Oy / Virta Ltd | Charging station monitoring method and device |
| CN114514141A (en) * | 2019-08-15 | 2022-05-17 | 丽肯纳维尔塔有限公司 | Charging station monitoring method and device |
| KR20210068164A (en) * | 2019-11-29 | 2021-06-09 | (주)제주전기자동차서비스 | System and method for diagnosing electric vehicle charger |
| CN112052966A (en) * | 2020-09-24 | 2020-12-08 | 佰聆数据股份有限公司 | Power customer satisfaction analysis system and method based on site emergency repair work order |
Non-Patent Citations (2)
| Title |
|---|
| 华瑶汪欣卢亚楠徐梦瑶: "电动汽车充电站与太阳能、停车场结合的运营模式探究" * |
| 沈海燕;杨玲玲;程清波;: "铁路客运站设备故障诊断评价体系及多属性分析模型研究", 中国设备工程, no. 07 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114814408A (en) * | 2022-03-23 | 2022-07-29 | 国网北京市电力公司 | Method, device and equipment for monitoring running state of charging module and storage medium |
| CN115330281A (en) * | 2022-10-14 | 2022-11-11 | 成都秦川物联网科技股份有限公司 | Smart city new energy automobile charging service method, system, device and medium |
| CN115330281B (en) * | 2022-10-14 | 2023-01-24 | 成都秦川物联网科技股份有限公司 | Smart city new energy automobile charging service method, system, device and medium |
| US11798061B2 (en) | 2022-10-14 | 2023-10-24 | Chengdu Qinchuan Iot Technology Co., Ltd. | Method and internet of things system of charging service for new energy vehicle in smart city |
| CN116167200A (en) * | 2022-12-05 | 2023-05-26 | 上海蓝箭电控设备南通有限公司 | Service life detection method and device for power distribution cabinet |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110097297B (en) | Multi-dimensional electricity stealing situation intelligent sensing method, system, equipment and medium | |
| CN109285346B (en) | Urban road network traffic state prediction method based on key road sections | |
| CN110222991B (en) | Metering device fault diagnosis method based on RF-GBDT | |
| CN113837473A (en) | Charging equipment fault rate analysis system and method based on BP neural network | |
| CN105740975A (en) | Data association relationship-based equipment defect assessment and prediction method | |
| CN112688431A (en) | Power distribution network load overload visualization method and system based on big data | |
| CN114519514B (en) | Low-voltage transformer area reasonable line loss value measuring and calculating method, system and computer equipment | |
| Sun et al. | An ensemble system to predict the spatiotemporal distribution of energy security weaknesses in transmission networks | |
| CN114662563A (en) | Industrial electricity non-invasive load decomposition method based on gradient lifting algorithm | |
| CN115905319B (en) | Automatic identification method and system for abnormal electricity fees of massive users | |
| CN112070395A (en) | Energy internet reliability evaluation system, model establishing method and evaluation method | |
| CN113268709B (en) | Urban electric vehicle charging demand prediction method and system based on intelligent agent simulation | |
| CN111898822B (en) | Charging load interval prediction method based on multi-correlation-day scene generation | |
| CN114385403A (en) | Distributed cooperative fault diagnosis method based on double-layer knowledge graph framework | |
| CN118336678A (en) | Electric vehicle charging station medium-term load prediction method based on machine learning | |
| CN117952598A (en) | Energy system safety evaluation analysis method and device based on fault rate approximation | |
| CN114167837B (en) | Intelligent fault diagnosis method and system for railway signal system | |
| Keyan et al. | Anomaly detection method of distribution network line loss based on hybrid clustering and LSTM | |
| CN116151799A (en) | BP neural network-based distribution line multi-working-condition fault rate rapid assessment method | |
| CN115907719A (en) | Hierarchical operation and maintenance management method and device for charging station | |
| CN115423041A (en) | Edge cloud fault prediction method and system based on deep learning | |
| CN114168662A (en) | Power distribution network problem combing and analyzing method and system based on multiple data sources | |
| CN114266370A (en) | Method and system for generating fault handling plan of power grid equipment in typhoon meteorological environment on line and storage medium | |
| CN117131947B (en) | Overhead transmission line fault prediction method, device, equipment and storage medium | |
| CN113128895B (en) | Intelligent workshop management system based on big data |
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 |