CN112380759B - Smart electric meter service life prediction method based on deep learning and CoxPH model - Google Patents
Smart electric meter service life prediction method based on deep learning and CoxPH model Download PDFInfo
- Publication number
- CN112380759B CN112380759B CN201911185780.0A CN201911185780A CN112380759B CN 112380759 B CN112380759 B CN 112380759B CN 201911185780 A CN201911185780 A CN 201911185780A CN 112380759 B CN112380759 B CN 112380759B
- Authority
- CN
- China
- Prior art keywords
- ammeter
- meter
- electric meter
- abnormal
- data
- 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 61
- 238000013135 deep learning Methods 0.000 title claims abstract description 24
- 230000002159 abnormal effect Effects 0.000 claims abstract description 128
- 230000004083 survival effect Effects 0.000 claims abstract description 109
- 230000006870 function Effects 0.000 claims abstract description 82
- 238000012549 training Methods 0.000 claims abstract description 49
- 238000012217 deletion Methods 0.000 claims abstract description 16
- 230000037430 deletion Effects 0.000 claims abstract description 16
- 230000005611 electricity Effects 0.000 claims description 36
- 230000005856 abnormality Effects 0.000 claims description 35
- 238000012423 maintenance Methods 0.000 claims description 15
- 238000013528 artificial neural network Methods 0.000 claims description 13
- 238000003860 storage Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 7
- 230000003068 static effect Effects 0.000 abstract description 7
- 238000009826 distribution Methods 0.000 description 13
- 238000004458 analytical method Methods 0.000 description 10
- 238000004891 communication Methods 0.000 description 10
- 238000009434 installation Methods 0.000 description 7
- 230000006399 behavior Effects 0.000 description 5
- 238000010801 machine learning Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 206010042434 Sudden death Diseases 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000002301 combined effect Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
- 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
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
The embodiment of the invention provides a smart meter service life prediction method based on deep learning and CoxPH model, which comprises the following steps: inputting abnormal data of the electric meter to be predicted into an electric meter life prediction model, and outputting an electric meter survival curve corresponding to the abnormal data of the electric meter to be predicted, wherein the electric meter survival curve is a curve of the relation between the survival probability and time of the electric meter to be predicted; the ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deletion label, and a loss function of the ammeter life prediction model during training is formed by participation of logarithmic part of risk functions in a CoxPH model; and predicting the service life of the electric meter to be predicted based on the electric meter survival curve and a preset survival probability threshold. The method provided by the embodiment of the invention avoids the situation that the life prediction model of the intelligent ammeter in the prior art is too static, and improves the reliability of life prediction of the intelligent ammeter.
Description
Technical Field
The invention relates to the technical field of machine learning, in particular to a smart meter service life prediction method based on deep learning and CoxPH model.
Background
The intelligent ammeter is taken as a key metering device in a modern power system, and the higher reliability of the intelligent ammeter is an important guarantee for normal maintenance operation of a power grid system. Along with the access of a large amount of intelligent electric meters in a power distribution network, multi-source intelligent meter big data are formed, wherein the multi-source intelligent meter big data contain abundant user energy and equipment operation and maintenance information. And the operation and maintenance work of the intelligent ammeter in the power distribution network always has challenges. In view of the fact that there is no reasonably active method for evaluating the reliability of an electric meter, the grid operator can only adopt a strategy of expiring the whole replacement by making a uniform limit on the running time of the electric meter. Such a maintenance approach undoubtedly results in a large number of potential ammeter runtimes becoming margins for ensuring stable operation of the system, and thus being wasted. One way of operation and maintenance in parallel with this is for a field operation and maintenance technician to report to replace or maintain by detecting that the meter has a fault during regular maintenance. The passive operation and maintenance strategy faces the practical challenges of huge installation quantity of the ammeter, sporadic faults, service quality degradation of the electricity utilization end before the faults are found, and the like.
To date, researchers have made a series of studies on analysis of large data of electricity meters. For example, a method of establishing a mathematical model of a functional module of an electric energy meter is proposed, and by querying a component failure manual, reasonably estimating stress levels, and calculating failure rates of the respective functional modules, life and reliability analysis of the electric energy meter is completed. A method for presupposing that the survival function of the ammeter accords with the Weibull distribution is provided, so that parameters of the Weibull distribution are calculated by utilizing maximum likelihood estimation after actual production data and operation data of the ammeter are utilized for dividing batches; and according to the calculated parameters and a preset survival rate threshold value, obtaining a life expectancy node so as to establish an early warning mechanism, or adopting a Cox linear model in a survival analysis theory to fit a survival function.
However, the reliability estimation method based on the failure rate of the components lacks of combination with actual metering statistical data, and cannot reflect complex physical characteristics and operation environments of the ammeter; the formed model is static and cannot be adjusted and adapted according to the change of the running environment and the physical information; the proposed assumption is not reasonable. Some researches are based on stronger assumption, such as a highly simplified model, only partial functional modules are considered, and physical reality is broken away; or presuming that the survival function accords with a specific function distribution form; or the combined effect of the hypothetical covariates can be expressed as a linear combination of the values of the various covariates. These assumptions are so strong that it is difficult to characterize the real physical properties and operating environment of the electricity meter, and finally it is inevitable that research is flowing in form; based on the basic Cox linear model method, the covariate variety which can be influenced is needed to be screened manually in advance when fitting calculation is carried out, and the process is dependent on manpower analysis and experience, so that the obtained result is seriously influenced by screening working quality.
Therefore, how to avoid the excessively static prediction model of the life of the formed smart meter and improve the reliability of life prediction of the smart meter is still a problem to be solved by those skilled in the art.
Disclosure of Invention
The embodiment of the invention provides a smart meter life prediction method based on deep learning and CoxPH models, which is used for solving the problem that the reliability of the existing smart meter life prediction model is low in smart meter life prediction.
In a first aspect, an embodiment of the present invention provides a smart meter lifetime prediction method based on deep learning and a CoxPH model, including:
inputting abnormal data of the electric meter to be predicted into an electric meter life prediction model, and outputting an electric meter survival curve corresponding to the abnormal data of the electric meter to be predicted, wherein the electric meter survival curve is a curve of the relation between the survival probability and time of the electric meter to be predicted;
the ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deletion label, and a loss function of the ammeter life prediction model during training is formed by participation of logarithmic part of risk functions in a CoxPH model;
and predicting the service life of the electric meter to be predicted based on the electric meter survival curve and a preset survival probability threshold.
Preferably, in the method, the ammeter abnormal sample data is obtained by sorting the collected overhaul data of the ammeter and the abnormal alarm history record of the ammeter.
Preferably, in the method, the collected overhauling data of the electric meter and the abnormal alarming history record of the electric meter are collated to obtain abnormal sample data of the electric meter, and the method specifically comprises the following steps:
first, the maintenance data and the abnormality alarm history of the ammeter are described in the form of the following formulas:wherein i is E omega
In the above-mentioned formula(s),is the abnormal type of the nth meter abnormality occurring in the meter of meter i, +.>Is the survival time of the electricity meter i when the electricity meter of the t-th electricity meter is abnormal, d (i) The life of the meter, c, is that of meter i (i) Deleted tag of ammeter, meta, which is ammeter i (i) The object of the electricity meter is electricity meter iManagement information, T (i) The number of abnormal records in the observation period is omega, and omega represents the collection of the electric meters with the collected overhaul data and the abnormal alarm history records;
the data described by the formula are formed into a list, the main key is the number of the electric meter, the ith row corresponds to abnormal data of the electric meter i, the list comprises an abnormal column, a deleted column and a service life column, wherein the abnormal value of the mth column is the value obtained by normalizing the number of times that the observed electric meter of the electric meter i generates mth abnormal in the observation period, the deleted column takes a value of 0 or 1, the value of 0 represents that the electric meter has failed, the value of 1 represents that the electric meter still survives when the observation time is reached, and the service life column takes a value of the electric meter i.
Preferably, in the method, determining a value normalized by the number of times of occurrence of the m-th abnormality of the electric meter i observed in the observation period specifically includes:
the actual number of m-th abnormality occurrence of the ammeter i isThe normalized value of the number of times of occurrence of the m-th abnormality of the electric meter i is +.>min (m) and max (m) are the minimum and maximum values of the mth column anomaly respectively,
based on the formula, the normalized numerical value of the number of times of occurrence of the m-th abnormality of the ammeter i observed in the observation period is determined.
Preferably, in the method, the formula of the loss function l (θ) is as follows:
wherein h is θ (x) Representing the logarithmic portionRisk function, x represents physical information parameter vector of ammeter, x i Physical information parameter vector, x of ammeter representing ammeter i j A physical information parameter vector representing ammeter j, θ representing the parameters of the neural network training the logarithmic partial risk function, E i Is deleted, E i =1 means that the meter of meter i survives up to the observation time, T i Refers to the life of the electricity meter, R (T) i ) Is T i The set of meters that survive in time, j is R (T i ) Ammeter j in the collection.
In a second aspect, an embodiment of the present invention provides a smart meter lifetime prediction apparatus based on deep learning and a CoxPH model, including:
The survival curve determining unit is used for inputting abnormal data of the electric meter to be predicted into the electric meter life prediction model and outputting an electric meter survival curve corresponding to the abnormal data of the electric meter to be predicted, wherein the electric meter survival curve is a curve of the relation between the survival probability and time of the electric meter to be predicted; the ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deletion label, and a loss function of the ammeter life prediction model during training is formed by participation of logarithmic part of risk functions in a CoxPH model;
and the life prediction unit is used for predicting the life of the electric meter to be predicted based on the electric meter life curve and a preset life probability threshold.
Preferably, in the device, the ammeter abnormal sample data is obtained by sorting the collected overhaul data of the ammeter and the abnormal alarm history record of the ammeter.
Preferably, in the device, the collected overhauling data of the electric meter and the abnormal alarming history record of the electric meter are collated to obtain abnormal sample data of the electric meter, and the device specifically comprises:
first, the maintenance data and the abnormality alarm history of the ammeter are described in the form of the following formulas:
Wherein i is E omega
In the above-mentioned formula(s),is the abnormal type of the nth meter abnormality occurring in the meter of meter i, +.>Is the survival time of the electricity meter i when the electricity meter of the t-th electricity meter is abnormal, d (i) The life of the meter, c, is that of meter i (i) Deleted tag of ammeter, meta, which is ammeter i (i) Is the physical information of the ammeter, T (i) The number of abnormal records in the observation period is omega, and omega represents the collection of the electric meters with the collected overhaul data and the abnormal alarm history records;
the data described by the formula are formed into a list, the main key is the number of the electric meter, the ith row corresponds to abnormal data of the electric meter i, the list comprises an abnormal column, a deleted column and a service life column, wherein the abnormal value of the mth column is the value obtained by normalizing the number of times that the observed electric meter of the electric meter i generates mth abnormal in the observation period, the deleted column takes a value of 0 or 1, the value of 0 represents that the electric meter has failed, the value of 1 represents that the electric meter still survives when the observation time is reached, and the service life column takes a value of the electric meter i.
In a third aspect, an embodiment of the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the smart meter lifetime prediction method based on deep learning and CoxPH model as provided in the first aspect when the program is executed.
In a fourth aspect, embodiments of the present invention provide a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the smart meter lifetime prediction method based on deep learning and CoxPH model as provided in the first aspect.
According to the intelligent ammeter life prediction method based on the deep learning and CoxPH model, abnormal data of an ammeter to be predicted is input into an ammeter life prediction model, and an ammeter survival curve corresponding to the abnormal data of the ammeter to be predicted is output, wherein the ammeter survival curve is a curve of the survival probability and time relation of the ammeter to be predicted; the ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deleted label, a loss function of the ammeter life prediction model during training is composed of logarithmic part of risk functions in a CoxPH model, the ammeter life prediction model for intelligent ammeter life prediction is dynamically generated through deep learning, and the ammeter life prediction model can be retrained through updating training samples and labels to obtain a more accurate prediction model. Therefore, the method provided by the embodiment of the invention avoids the situation that the life prediction model of the intelligent ammeter in the prior art is too static, and improves the reliability of life prediction of the intelligent ammeter.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, a brief description will be given below of the drawings required for the embodiments or the prior art descriptions, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flow chart of a smart meter life prediction method based on deep learning and a CoxPH model according to an embodiment of the present invention;
FIG. 2 is a flow chart of a training method of an ammeter life prediction model based on machine learning and CoxPH model according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a smart meter life prediction device based on deep learning and a CoxPH model according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of 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 apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without any inventive effort, are intended to be within the scope of the invention.
The intelligent ammeter life prediction method based on the CoxPH model provided by the embodiment of the invention is based on a survival analysis theory, and survival analysis is a statistical analysis method for researching the distribution rule of survival time and the relationship between the survival time and related factors, and is widely applied to the fields of life analysis of patients, fault-time analysis of machine equipment and the like. First, a few basic terms are introduced in combination with the application scene of the intelligent ammeter:
event (Event): in the electricity meter life analysis, the electricity meter is replaced after failure/after reaching a prescribed maximum service time.
Time to live (survivinval Time): refers to the duration of time that the meter is removed from the time it is installed until the meter is determined to fail/cause for the maximum specified use time.
Deletion (Censoring) refers to the situation where the meter lifetime continues until the last observation time node, and the event has not yet occurred. In a broad sense, there are three cases of left deletion, right deletion and interval deletion, but for the smart meter, only right deletion is considered here, i.e. the actual lifetime of the meter is longer than the observed time.
Covariates (covariates): variable factors that affect the time of occurrence of an event, such as the manufacturer or the number of anomaly alarms for electricity meters.
Survival function (Survival Function) S (t): the Probability that an individual' S survival time exceeds T is defined as S (T) =probability (T > T).
Survival Curve (survivinal Curve): the survival at each time point was connected as a curve. The x-axis generally represents time to live and the y-axis represents probability of live.
Risk Function (Hazard Function) λ (t): the instantaneous death probability is characterized as follows:
wherein S (t) is a survival function.
Classical methods of survival analysis can be divided into three categories: parametric, non-parametric and semi-parametric methods. The parameter method needs to pre-presume or confirm the distribution model of the survival time, then estimate model parameters according to the data, finally calculate the survival rate by the distribution model, and the commonly used assumed distribution models are poisson distribution and Weibull distribution. The non-parametric model does not need to assume or calculate a survival time distribution model, the survival rate is directly estimated according to the sample statistics, and a common method is a Kaplan-Meier estimator, but the obtained survival function estimation cannot correct the influence caused by the change of the covariates. The semi-parametric method also does not need to know the distribution of survival time, but finally needs to evaluate the factors influencing the survival rate through a model, and most commonly, a Cox regression model can consider the influence of a plurality of covariates.
The Cox regression model is also called CoxPH (Cox Proportional Hazard, cox proportional hazards) model, and the defined risk function has the following form:
wherein lambda is 0 (t) is a reference risk function, x= (X) 1 ,...,x m ) H (X) is a logarithmic partial risk function for the vector of covariates. The CoxPH model assumes that the common influence of a plurality of covariates can be described by logarithm part of the risk function, so that the influence of the covariates is comprehensively considered, and the survival function of the product can be effectively described under the condition of assuming compliance. When h (X) =θ·x, θ= (θ) 1 ,...,θ m ) When the coefficients of the linear model, i.e., the logarithmic risk function, can be expressed as a linear combination of covariate values, the model at this time is referred to as a linear CoxPH model.
The relationship between risk and survival functions is known as:
where λ (t) is a risk function and S (t) is a survival function.
From the above relationship, it can be known that:
in the above-mentioned formula, the group of the compounds,known as cumulative benchmark risk; s is S 0 (t)＝exp(-H 0 (t)) is referred to as a reference survival function. For H 0 The most widely used method for solving (t) is to use a Breslow estimator, defined by:
in the above formula, if t i Event occurrence time (i.e. meter fault occurrence), then Wherein R (t) i ) At t i The set of individuals still under observation during the time of day, otherwise, λ 0 (t i )＝0。
Therefore, as long as the logarithmic partial risk function h (X) is determined, the survival function S (t) can be deduced, and thus the electricity meter lifetime can be predicted.
The existing intelligent ammeter life prediction method based on the CoxPH model generally has the problems that the ammeter life prediction model is too static and the prediction reliability is low. In view of the above, the embodiment of the invention provides a smart meter life prediction method based on a CoxPH model. Fig. 1 is a flow chart of a method for predicting service life of a smart meter based on a CoxPH model according to an embodiment of the present invention, as shown in fig. 1, the method includes:
step 110, inputting abnormal data of the electric meter to be predicted into an electric meter life prediction model, and outputting an electric meter survival curve corresponding to the abnormal data of the electric meter to be predicted, wherein the electric meter survival curve is a curve of the relation between the survival probability and time of the electric meter to be predicted; the ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deletion label, and a loss function of the ammeter life prediction model during training is formed by participation of logarithmic part of risk functions in the CoxPH model.
Specifically, abnormal data of the electric meter to be predicted is input into an electric meter life prediction model, and an electric meter survival curve corresponding to the abnormal data of the electric meter to be predicted is output. The abnormal data of the electric meter to be predicted includes types of abnormal electric meter, for example: specific abnormal types such as short circuit of a main board of the ammeter, abnormal display screen or abnormal battery power supply; also included are the installed times from the time of installation when each type of anomaly occurs, for example: the distance from the short circuit of the main board of the ammeter to the installation time of the ammeter is 567 days, the distance from the abnormal display screen to the installation time of the ammeter is 678 days, the distance from the abnormal battery power supply to the installation time of the ammeter is 345 days, and the like. The survival curve is a curve of the relationship between the survival Probability of the ammeter and time, and the survival curve S (T) is expressed by a formula, i.e., S (T) =probability (T > T).
The ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deletion label, and a loss function of the ammeter life prediction model during training is formed by participation of a logarithmic part of risk functions in the CoxPH model. From the foregoing, it can be seen that the ammeter life prediction model based on the CoxPH model derives the survival function S (t) by determining the logarithmic part risk function h (X), and the training of the ammeter life prediction model herein is to fit the nonlinear function h (X) to the logarithmic part risk function h (X) using the training of the neural network. Therefore, a loss function adopted in the ammeter life prediction model training process is constructed based on a logarithmic part risk function h (X), so that h (X) values output in the training process are more and more accurate, and the accuracy of the h (X) values is represented by the accuracy of S (t) deduced by the h (X) values.
And step 120, predicting the service life of the electric meter to be predicted based on the electric meter survival curve and a preset survival probability threshold.
Specifically, the service life of the electric meter to be predicted is predicted based on an electric meter survival curve output by an electric meter service life prediction model and a preset survival probability threshold. For example, the preset survival probability threshold is 80%, that is, the maximum survival time of the electric meter under the condition that the survival rate of the electric meter is about to reach 80% is determined according to the electric meter survival curve, for example, the electric meter survival curve output by the electric meter life prediction model obtains that the survival probability of the electric meter is just 80% when the survival time is 1000 days, and then 1000 days is the life of the electric meter under the condition that the preset survival probability threshold is 80%.
According to the method provided by the embodiment of the invention, abnormal data of the electric meter to be predicted is input into an electric meter life prediction model, and an electric meter survival curve corresponding to the abnormal data of the electric meter to be predicted is output, wherein the electric meter survival curve is a curve of the relation between the survival probability and time of the electric meter to be predicted; the ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deleted label, a loss function of the ammeter life prediction model during training is composed of logarithmic part of risk functions in a CoxPH model, the ammeter life prediction model for intelligent ammeter life prediction is dynamically generated through deep learning, and the ammeter life prediction model can be retrained through updating training samples and labels to obtain a more accurate prediction model. Therefore, the method provided by the embodiment of the invention avoids the situation that the life prediction model of the intelligent ammeter in the prior art is too static, and improves the reliability of life prediction of the intelligent ammeter.
Based on the above embodiment, in the method, the ammeter abnormal sample data is obtained by sorting the collected overhaul data of the ammeter and the abnormal alarm history record of the ammeter.
Specifically, the ammeter abnormal sample data is obtained from overhaul data of the ammeter and an abnormal alarm history of the ammeter, and the data is required to be classified and sorted.
Based on the above embodiment, in the method, the collected overhaul data of the electric meter and the abnormal alarm history record of the electric meter are collated to obtain the abnormal sample data of the electric meter, which specifically includes:
first, the maintenance data and the abnormality alarm history of the ammeter are described in the form of the following formulas:wherein i is E omega
In the above-mentioned formula(s),is the abnormal type of the nth meter abnormality occurring in the meter of meter i, +.>Is the survival time of the electricity meter i when the electricity meter of the t-th electricity meter is abnormal, d (i) The life of the meter, c, is that of meter i (i) Deleted tag of ammeter, meta, which is ammeter i (i) Is the physical information of the ammeter, T (i) The number of abnormal records in the observation period is omega, and omega represents the collection of the electric meters with the collected overhaul data and the abnormal alarm history records;
the data described by the formula are formed into a list, the main key is the number of the electric meter, the ith row corresponds to abnormal data of the electric meter i, the list comprises an abnormal column, a deleted column and a service life column, wherein the abnormal value of the mth column is the value obtained by normalizing the number of times that the observed electric meter of the electric meter i generates mth abnormal in the observation period, the deleted column takes a value of 0 or 1, the value of 0 represents that the electric meter has failed, the value of 1 represents that the electric meter still survives when the observation time is reached, and the service life column takes a value of the electric meter i.
Specifically, firstly, the overhaul data and the abnormal alarm history of the ammeter are uniformly described in the format, and the following formula is specifically adopted:
wherein i is E omega
In the above-mentioned formula(s),is the abnormal type of the nth meter abnormality occurring in the meter of meter i, +.>Is the survival time of the electricity meter i when the electricity meter of the t-th electricity meter is abnormal, d (i) The life of the meter, c, is that of meter i (i) Deleted tag of ammeter, meta, which is ammeter i (i) Is the physical information of the ammeter, T (i) The number of abnormal records in the observation period is omega, and omega represents the collection of the electric meters with the collected overhaul data and the abnormal alarm history records. The physical information of the electricity meter generally includes various physical parameters and specifications of the electricity meter, among others.
And then the data are arranged in a table form, wherein table 1 is a preliminary arrangement mode, and as shown in table 1, the main key of the table is { table number x abnormal time }, and each row represents an abnormal record of an ammeter corresponding to the table number.
Table 1 preliminary arrangement
In Table 1, each row corresponds to an anomaly of the meter corresponding to a table number, meta of the meter (i) The information comprises the attribute, manufacturer, specification, version and installation time of the electric meter, different anomaly types are represented by different anomaly codes, the anomaly time is the time from the installation time to the occurrence of the anomaly, the table 1 also comprises a deleted column and a service life column, wherein the deleted value is 0 or 1, the deleted value is 0, the electric meter has failed, the service life value is the service life of the electric meter when the deleted value of the electric meter is 0, the electric meter is still running when the deleted value is 1, and the service life is when the deleted value of the electric meter is 1 The value indicates the duration of continuous operation of the meter up to the observation time.
The data in table 1 were further processed so that the data could be subjected to a CoxPH model fit. Table 2 shows the final arrangement, and the primary key of Table 2 is { Table number }. As shown in table 2, one electric meter corresponds to one table number, one table number corresponds to one row in table 2, and the numerical value of each column of anomalies represents the normalized result of the number of anomalies observed during the observation period. For example, the value of anomaly 3 of the electric meter with the table number "00002" indicates the normalized value of the number of times the occurrence of anomaly of the electric meter with the table number "00002" is anomaly 3.
TABLE 2 final finishing mode
Based on any one of the above embodiments, in the method, determining a normalized value of the number of times of occurrence of the m-th type abnormality of the electric meter i observed in the observation period specifically includes:
the actual number of m-th abnormality occurrence of the ammeter i isThe normalized value of the number of times of occurrence of the m-th abnormality of the electric meter i is +.>min (m) and max (m) are the minimum and maximum values of the mth column anomaly respectively,
based on the formula, the normalized numerical value of the number of times of occurrence of the m-th abnormality of the ammeter i observed in the observation period is determined.
Specifically, the number of different types of anomalies occurring in each meter is normalized, e.g., for the mth type of anomaly, traversing all meters for the mth type of anomalyFinding out the maximum value max (m) and the minimum value min (m) from the constant times, wherein the actual times of occurrence of the m-th abnormality of the ammeter i are as followsThe normalized value of the number of times of occurrence of the m-th abnormality of the electric meter i is +.>The following formula is used:
obtainingTo->Is a transition of (2).
Based on any one of the above embodiments, in the method, the formula of the loss function l (θ) is as follows:
wherein h is θ (x) Representing the logarithmic part risk function, x represents the physical information parameter vector of the ammeter, x i Physical information parameter vector, x of ammeter representing ammeter i j A physical information parameter vector representing ammeter j, θ representing the parameters of the neural network training the logarithmic partial risk function, E i Is deleted, E i =1 means that the meter of meter i survives up to the observation time, T i Refers to the life of the electricity meter, R (T) i ) Is T i The set of meters that survive in time, j is R (T i ) Ammeter j in the collection.
Specifically, the formula for the loss function l (θ) is further specified herein as follows:
wherein h is θ (x) Representing the logarithmic part risk function, x represents the physical information parameter vector of the ammeter, x i Physical information parameter vector, x of ammeter representing ammeter i j A physical information parameter vector representing ammeter j, θ representing the parameters of the neural network training the logarithmic partial risk function, E i Is deleted, E i =1 means that the meter of meter i survives up to the observation time, T i Refers to the life of the electricity meter, R (T) i ) Is T i The set of meters that survive in time, j is R (T i ) Ammeter j in the collection. The training of the ammeter life prediction model in the embodiment of the invention is essentially that the neural network training is carried out on the logarithmic part risk function in the ammeter life prediction model, and h θ (x) Representing the logarithmic part risk function, wherein theta represents parameters of a neural network for training the logarithmic part risk function, and when the parameters theta of the neural network for training the logarithmic part risk function are obtained, the logarithmic part risk function h can be determined θ (x) The survival curve can then be deduced from the above-described transformation of the logarithmic partial risk function to the survival function.
Based on any one of the above embodiments, the embodiment of the present invention provides a method for training an ammeter life prediction model based on machine learning and a CoxPH model, and fig. 2 is a flow chart of the method for training an ammeter life prediction model based on machine learning and a CoxPH model provided by the embodiment of the present invention. Before the electric meter life prediction model shown in fig. 2 is trained, the abnormal data and fault data of the electric meter are cleaned and statistically analyzed to form a data table containing deleted labels, and the data table is used as a data sample set for subsequent machine learning training. Then, the flow shown in fig. 2 is entered, the deep neural network is first defined, the parameter structure is initialized, the data of the data sample set which is finished before is taken as input, specifically, the input covariate vector X is an n×1 vector formed by normalized values corresponding to N anomaly types in table 2, and the target output is the input label, namely, the two columns of data of life and deletion in table 2. The neural network training process is to train network parameters to fit nonlinear logarithmic part risk functions h (X), the output of each unit of the hidden layer and the output layer of the autumn neural network in the training process is to obtain an expression of h (X), a reference risk function is obtained by calculating the relation between the risk functions and the logarithmic part risk functions so as to obtain a complete survival function, then a consistency index C-index is calculated, if the value of the C-index is acceptable, the training is completed, the current electric meter life prediction model is used for electric meter life prediction, and if the value of the C-index is unacceptable, the electric meter life prediction is carried out by the following formula
The loss function is calculated and then the neural network parameters are adjusted using a gradient descent method.
Based on any one of the above embodiments, the embodiment of the present invention provides a smart meter lifetime prediction device based on deep learning and a CoxPH model, and fig. 3 is a schematic structural diagram of the smart meter lifetime prediction device based on deep learning and a CoxPH model provided by the embodiment of the present invention. As shown in fig. 3, the apparatus includes a survival curve determination unit 310 and a life prediction unit 320, wherein,
the survival curve determining unit 310 is configured to input abnormal data of the electric meter to be predicted into an electric meter life prediction model, and output an electric meter survival curve corresponding to the abnormal data of the electric meter to be predicted, where the electric meter survival curve is a curve of a relationship between a survival probability and time of the electric meter to be predicted; the ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deletion label, and a loss function of the ammeter life prediction model during training is formed by participation of logarithmic part of risk functions in a CoxPH model;
the life prediction unit 320 is configured to predict the life of the electric meter to be predicted based on the electric meter survival curve and a preset survival probability threshold.
The device provided by the embodiment of the invention inputs the abnormal data of the electric meter to be predicted into an electric meter life prediction model, and outputs an electric meter survival curve corresponding to the abnormal data of the electric meter to be predicted, wherein the electric meter survival curve is a curve of the relation between the survival probability and time of the electric meter to be predicted; the ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deleted label, a loss function of the ammeter life prediction model during training is composed of logarithmic part of risk functions in a CoxPH model, the ammeter life prediction model for intelligent ammeter life prediction is dynamically generated through deep learning, and the ammeter life prediction model can be retrained through updating training samples and labels to obtain a more accurate prediction model. Therefore, the device provided by the embodiment of the invention avoids the situation that the life prediction model of the intelligent ammeter in the prior art is too static, and improves the reliability of life prediction of the intelligent ammeter.
Based on any one of the above embodiments, in the device, the abnormal sample data of the electric meter is obtained by sorting the collected overhaul data of the electric meter and the abnormal alarm history record of the electric meter.
Based on any one of the above embodiments, in the device, the collecting maintenance data of the electric meter and the abnormal alarm history record of the electric meter are collated to obtain the abnormal sample data of the electric meter, specifically including:
first, the maintenance data and the abnormality alarm history of the ammeter are described in the form of the following formulas:
wherein i is E omega
In the above-mentioned formula(s),is the abnormal type of the nth meter abnormality occurring in the meter of meter i, +.>Is the survival time of the electricity meter i when the electricity meter of the t-th electricity meter is abnormal, d (i) The life of the meter, c, is that of meter i (i) Deleted tag of ammeter, meta, which is ammeter i (i) Is the physical information of the ammeter, T (i) The number of abnormal records in the observation period is omega, and omega represents the collection of the electric meters with the collected overhaul data and the abnormal alarm history records;
the data described by the formula are formed into a list, the main key is the number of the electric meter, the ith row corresponds to abnormal data of the electric meter i, the list comprises an abnormal column, a deleted column and a service life column, wherein the abnormal value of the mth column is the value obtained by normalizing the number of times that the observed electric meter of the electric meter i generates mth abnormal in the observation period, the deleted column takes a value of 0 or 1, the value of 0 represents that the electric meter has failed, the value of 1 represents that the electric meter still survives when the observation time is reached, and the service life column takes a value of the electric meter i.
Based on any one of the above embodiments, in the device, determining a normalized value of the number of times of occurrence of the m-th type abnormality of the electric meter i observed in the observation period specifically includes:
the actual number of m-th abnormality occurrence of the ammeter i isThe normalized value of the number of times of occurrence of the m-th abnormality of the electric meter i is +.>min (m) and max (m) are the minimum and maximum values of the mth column anomaly respectively,
based on the formula, the normalized numerical value of the number of times of occurrence of the m-th abnormality of the ammeter i observed in the observation period is determined.
Based on any one of the above embodiments, in the apparatus, the formula of the loss function l (θ) is as follows:
wherein h is θ (x) Representing the logarithmic part risk function, x represents the physical information parameter vector of the ammeter, x i Physical information parameter vector, x of ammeter representing ammeter i j A physical information parameter vector representing ammeter j, θ representing the parameters of the neural network training the logarithmic partial risk function, E i Is deleted, E i =1 means that the meter of meter i survives up to the observation time, T i Refers to the life of the electricity meter, R (T) i ) Is T i The set of meters that survive in time, j is R (T i ) Ammeter j in the collection.
Fig. 4 is a schematic physical structure diagram of an electronic device according to an embodiment of the present invention, as shown in fig. 4, where the electronic device may include: a processor (processor) 401, a communication interface (Communications Interface) 402, a memory (memory) 403 and a communication bus 404, wherein the processor 401, the communication interface 402 and the memory 403 complete communication with each other through the communication bus 404. The processor 401 may call a computer program stored in the memory 403 and executable on the processor 401 to perform the smart meter lifetime prediction method based on the deep learning and CoxPH model provided in the above embodiments, for example, includes: inputting abnormal data of the electric meter to be predicted into an electric meter life prediction model, and outputting an electric meter survival curve corresponding to the abnormal data of the electric meter to be predicted, wherein the electric meter survival curve is a curve of the relation between the survival probability and time of the electric meter to be predicted; the ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deletion label, and a loss function of the ammeter life prediction model during training is formed by participation of logarithmic part of risk functions in a CoxPH model; and predicting the service life of the electric meter to be predicted based on the electric meter survival curve and a preset survival probability threshold.
Further, the logic instructions in the memory 403 may be implemented in the form of software functional units and stored in a computer readable storage medium when sold or used as a stand alone product. Based on such understanding, the technical solution of the embodiments of the present invention may be embodied in essence or a part contributing to the prior art or a part of the technical solution, in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
Embodiments of the present invention also provide a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the smart meter lifetime prediction method based on deep learning and CoxPH model provided in the above embodiments, for example, including: inputting abnormal data of the electric meter to be predicted into an electric meter life prediction model, and outputting an electric meter survival curve corresponding to the abnormal data of the electric meter to be predicted, wherein the electric meter survival curve is a curve of the relation between the survival probability and time of the electric meter to be predicted; the ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deletion label, and a loss function of the ammeter life prediction model during training is formed by participation of logarithmic part of risk functions in a CoxPH model; and predicting the service life of the electric meter to be predicted based on the electric meter survival curve and a preset survival probability threshold.
The embodiments of the present invention also provide a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the authentication method provided in the above embodiments, for example, including: sending a verification request to a server to request a maze image to be verified to the server; wherein the maze image to be verified comprises a maze path in a disconnected state; receiving the maze image to be verified, and acquiring maze path communication behavior information corresponding to the maze image to be verified; the maze path communication behavior information characterizes the rotation behavior of a maze sub-image in the maze image to be verified; and returning the maze path communication behavior information to the server to request the server to determine an identity verification result based on the maze path communication behavior information.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (5)
1. The intelligent ammeter service life prediction method based on deep learning and CoxPH model is characterized by comprising the following steps:
inputting abnormal data of the electric meter to be predicted into an electric meter life prediction model, and outputting an electric meter survival curve corresponding to the abnormal data of the electric meter to be predicted, wherein the electric meter survival curve is a curve of the relation between the survival probability and time of the electric meter to be predicted;
the ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deletion label, and a loss function of the ammeter life prediction model during training is formed by participation of logarithmic part of risk functions in a CoxPH model;
predicting the service life of the electric meter to be predicted based on the electric meter survival curve and a preset survival probability threshold;
the ammeter abnormal sample data are obtained by sorting the acquired overhaul data of the ammeter and the abnormal alarm history record of the ammeter;
the method comprises the steps of obtaining the abnormal sample data of the electric meter by arranging the overhaul data of the collected electric meter and the abnormal alarm history record of the electric meter, and specifically comprises the following steps:
first, the maintenance data and the abnormality alarm history of the ammeter are described in the form of the following formulas:
Wherein i is E omega
In the above-mentioned formula(s),is the abnormal type of the nth meter abnormality occurring in the meter of meter i, +.>Is an ammeterThe survival time when the electricity meter of i is abnormal for the t-th electricity meter, d (i) The life of the meter, c, is that of meter i (i) Deleted tag of ammeter, meta, which is ammeter i (i) Is the physical information of the ammeter, T (i) The number of abnormal records in the observation period is omega, and omega represents the collection of the electric meters with the collected overhaul data and the abnormal alarm history records;
forming a list by the data described by the formula, wherein the main key is the number of the ammeter, the ith row corresponds to abnormal data of the ammeter i, the list comprises an abnormal column, a deleted column and a service life column, wherein the abnormal value of the mth column is the value obtained by normalizing the observed times of occurrence of the mth abnormal condition of the ammeter i in the observation period, the deleted column is 0 or 1, the value of 0 is used for indicating that the ammeter has failed, the value of 1 is used for indicating that the ammeter still survives when the ammeter is observed, and the service life column is the service life of the ammeter i;
the method for determining the normalized numerical value of the times of occurrence of the m-th type abnormality of the ammeter i in the observation period specifically comprises the following steps:
the actual number of m-th abnormality occurrence of the ammeter i is The normalized value of the number of times of occurrence of the m-th abnormality of the electric meter i is +.>min (m) and max (m) are the minimum and maximum values of the mth column anomaly respectively,
based on the formula, the normalized numerical value of the number of times of occurrence of the m-th abnormality of the ammeter i observed in the observation period is determined.
2. The smart meter life prediction method based on deep learning and CoxPH model of claim 1, wherein the formula of the loss function i (θ) is the following formula:
wherein h is θ (x) Representing the logarithmic part risk function, x represents the physical information parameter vector of the ammeter, x i Physical information parameter vector, x of ammeter representing ammeter i j A physical information parameter vector representing ammeter j, θ representing the parameters of the neural network training the logarithmic partial risk function, E i Is deleted, E i =1 means that the meter of meter i survives up to the observation time, T i Refers to the life of the electricity meter, R (T) i ) Is T i The set of meters that survive in time, j is R (T i ) Ammeter j in the collection.
3. Smart electric meter life prediction device based on deep learning and CoxPH model, characterized by comprising:
the survival curve determining unit is used for inputting abnormal data of the electric meter to be predicted into the electric meter life prediction model and outputting an electric meter survival curve corresponding to the abnormal data of the electric meter to be predicted, wherein the electric meter survival curve is a curve of the relation between the survival probability and time of the electric meter to be predicted; the ammeter life prediction model is obtained after training based on ammeter abnormal sample data, a preset ammeter life label and a preset deletion label, and a loss function of the ammeter life prediction model during training is formed by participation of logarithmic part of risk functions in a CoxPH model;
The life prediction unit is used for predicting the life of the electric meter to be predicted based on the electric meter survival curve and a preset survival probability threshold value;
the ammeter abnormal sample data are obtained by sorting the acquired overhaul data of the ammeter and the abnormal alarm history record of the ammeter;
the method comprises the steps of obtaining the abnormal sample data of the electric meter by arranging the overhaul data of the collected electric meter and the abnormal alarm history record of the electric meter, and specifically comprises the following steps:
first, the maintenance data and the abnormality alarm history of the ammeter are described in the form of the following formulas:
wherein i is E omega
In the above-mentioned formula(s),is the abnormal type of the nth meter abnormality occurring in the meter of meter i, +.>Is the survival time of the electricity meter i when the electricity meter of the t-th electricity meter is abnormal, d (i) The life of the meter, c, is that of meter i (i) Deleted tag of ammeter, meta, which is ammeter i (i) Is the physical information of the ammeter, T (i) The number of abnormal records in the observation period is omega, and omega represents the collection of the electric meters with the collected overhaul data and the abnormal alarm history records;
forming a list by the data described by the formula, wherein the main key is the number of the ammeter, the ith row corresponds to abnormal data of the ammeter i, the list comprises an abnormal column, a deleted column and a service life column, wherein the abnormal value of the mth column is the value obtained by normalizing the observed times of occurrence of the mth abnormal condition of the ammeter i in the observation period, the deleted column is 0 or 1, the value of 0 is used for indicating that the ammeter has failed, the value of 1 is used for indicating that the ammeter still survives when the ammeter is observed, and the service life column is the service life of the ammeter i;
The method for determining the normalized numerical value of the times of occurrence of the m-th type abnormality of the ammeter i in the observation period specifically comprises the following steps:
the actual number of m-th abnormality occurrence of the ammeter i isThe normalized value of the number of times of occurrence of the m-th abnormality of the electric meter i is +.>min (m) and max (m) are the minimum and maximum values of the mth column anomaly respectively,
based on the formula, the normalized numerical value of the number of times of occurrence of the m-th abnormality of the ammeter i observed in the observation period is determined.
4. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the smart meter lifetime prediction method based on deep learning and CoxPH model of any one of claims 1 to 2 when executing the program.
5. A non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor, implements the steps of the smart meter lifetime prediction method based on deep learning and CoxPH model of any one of claims 1 to 2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911185780.0A CN112380759B (en) | 2019-11-27 | 2019-11-27 | Smart electric meter service life prediction method based on deep learning and CoxPH model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911185780.0A CN112380759B (en) | 2019-11-27 | 2019-11-27 | Smart electric meter service life prediction method based on deep learning and CoxPH model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112380759A CN112380759A (en) | 2021-02-19 |
| CN112380759B true CN112380759B (en) | 2024-01-09 |
Family
ID=74586030
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911185780.0A Active CN112380759B (en) | 2019-11-27 | 2019-11-27 | Smart electric meter service life prediction method based on deep learning and CoxPH model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112380759B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112906994B (en) * | 2021-04-19 | 2023-04-07 | 拉扎斯网络科技（上海）有限公司 | Order meal delivery time prediction method and device, electronic equipment and storage medium |
| CN116302661B (en) * | 2023-05-15 | 2023-10-13 | 合肥联宝信息技术有限公司 | Abnormality prediction method and device, electronic equipment and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2602867A1 (en) * | 2007-09-18 | 2009-03-18 | Blue Line Innovations Inc. | System and method for reading power meters |
| CN110261811A (en) * | 2019-07-05 | 2019-09-20 | 北京志翔科技股份有限公司 | Intelligent electric meter batch method for early warning and system |
-
2019
- 2019-11-27 CN CN201911185780.0A patent/CN112380759B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2602867A1 (en) * | 2007-09-18 | 2009-03-18 | Blue Line Innovations Inc. | System and method for reading power meters |
| CN110261811A (en) * | 2019-07-05 | 2019-09-20 | 北京志翔科技股份有限公司 | Intelligent electric meter batch method for early warning and system |
Non-Patent Citations (3)
| Title |
|---|
| 基于贝叶斯的智能电能表可靠性评估方法;田正其;祝宇楠;刘建;徐晴;;计算机测量与控制(10);第274-277页 * |
| 智能电能表软件可靠性测试研究;严晶晶;王海巍;张卫欣;陈红芳;胡长安;;电测与仪表(07);第97-102页 * |
| 智能电表故障大数据分析探究;贺宁;;中小企业管理与科技(上旬刊)(07);第142-145页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112380759A (en) | 2021-02-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7107187B1 (en) | Method for modeling system performance | |
| CN111314173B (en) | Monitoring information abnormity positioning method and device, computer equipment and storage medium | |
| US9704382B2 (en) | Method for calculating error rate of alarm | |
| CN107992410B (en) | Software quality monitoring method and device, computer equipment and storage medium | |
| CN112380759B (en) | Smart electric meter service life prediction method based on deep learning and CoxPH model | |
| CN112381258B (en) | Meter operation life cycle prediction method and device | |
| CN111262750B (en) | Method and system for evaluating baseline model | |
| CN111796233B (en) | Method for evaluating secondary errors of multiple voltage transformers in double-bus connection mode | |
| CN111931334A (en) | Method and system for evaluating operation reliability of cable equipment | |
| CN112632179A (en) | Model construction method and device, storage medium and equipment | |
| CN110858812A (en) | Network element cutover and watching method and device | |
| CN111985816A (en) | Power equipment maintenance mode risk cost assessment method, device, terminal and medium | |
| CN111340287A (en) | Power distribution cabinet operation state prediction method and device | |
| CN111367782A (en) | Method and device for automatically generating regression test data | |
| CN115114124A (en) | Host risk assessment method and device | |
| CN110413482B (en) | Detection method and device | |
| CN111931323B (en) | Memory, hydrocracking equipment fault prediction method, device and equipment | |
| CN107832925A (en) | Internet content risk evaluating method, device and server | |
| CN115544803B (en) | Transformer residual life prediction method, device, equipment and storage medium | |
| CN113033845B (en) | Construction method and device for power transmission resource co-construction and sharing | |
| CN113298389A (en) | Method and device for evaluating cable running state | |
| CN109739210B (en) | Method and device for evaluating health state of equipment component | |
| Mun et al. | Optimal maintenance policy of repairable system with bathtub shaped intensity | |
| CN117194049B (en) | Cloud host intelligent behavior analysis method and system based on machine learning algorithm | |
| CN115981945A (en) | Nuclear power instrument control board card based hardware health parameter verification method and device |
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 |