CN116777281A - ARIMA model-based power equipment quality trend prediction method and device - Google Patents
ARIMA model-based power equipment quality trend prediction method and device Download PDFInfo
- Publication number
- CN116777281A CN116777281A CN202310753599.5A CN202310753599A CN116777281A CN 116777281 A CN116777281 A CN 116777281A CN 202310753599 A CN202310753599 A CN 202310753599A CN 116777281 A CN116777281 A CN 116777281A
- Authority
- CN
- China
- Prior art keywords
- model
- sequence
- value
- data set
- prediction
- 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
- 238000000034 method Methods 0.000 title claims abstract description 57
- YHXISWVBGDMDLQ-UHFFFAOYSA-N moclobemide Chemical compound C1=CC(Cl)=CC=C1C(=O)NCCN1CCOCC1 YHXISWVBGDMDLQ-UHFFFAOYSA-N 0.000 title claims abstract description 48
- 238000001514 detection method Methods 0.000 claims abstract description 53
- 238000012360 testing method Methods 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000007781 pre-processing Methods 0.000 claims description 10
- 238000003860 storage Methods 0.000 claims description 10
- 230000009191 jumping Effects 0.000 claims description 9
- 238000000819 phase cycle Methods 0.000 claims description 9
- 230000000087 stabilizing effect Effects 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 238000010276 construction Methods 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 5
- 238000012935 Averaging Methods 0.000 claims description 4
- 238000007689 inspection Methods 0.000 claims description 3
- 239000000463 material Substances 0.000 abstract description 11
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 238000006243 chemical reaction Methods 0.000 abstract 1
- 238000011158 quantitative evaluation Methods 0.000 abstract 1
- 238000004590 computer program Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000005457 optimization Methods 0.000 description 4
- 238000001303 quality assessment method Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000013441 quality evaluation Methods 0.000 description 2
- 238000012502 risk assessment Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The application relates to a method and a device for predicting the quality trend of power equipment based on an ARIMA model, which mainly comprise two major contents. One is to provide a method for scoring the quality of electric power materials. And (3) performing score scale conversion on the original detection data by using a threshold value given by the electric power material detection report, and calculating the quality score of the electric power material by combining the weight value of the detection index, thereby realizing quantitative evaluation of the quality of the electric power material. And secondly, an ARIMA prediction model with self-adaptive optimal order is provided. By utilizing the obtained electric power material score data, an ARIMA self-adaptive order optimizing method is designed, different optimal parameters are obtained for different suppliers, accurate prediction of future batches of different suppliers is realized, and the method is beneficial to helping technicians to know the future change condition of the supply quality of the suppliers in time.
Description
Technical Field
The application relates to a power equipment quality trend prediction method and device based on an ARIMA model, and belongs to the technical field of quality prediction.
Background
The traditional electric power material quality assessment method mainly depends on experience of technicians to judge, lacks quantitative analysis, has man-made subjectivity, lacks related data support, and is easy to cause misjudgment and other problems. In addition, the model parameters of the existing cargo quality prediction model are all fixed parameters, the parameters cannot be adjusted according to the characteristics of suppliers or products, and the prediction performance cannot be always kept in an optimal state.
The conventional quality evaluation method at present mainly evaluates the quality of the power equipment according to the related experience of technicians, and has certain subjectivity. At present, only relevant supervision methods are used for the quality of the electric power materials, quality assessment and prediction methods are lacking, the quality supervision methods can ensure the safety of the existing electric power materials in the transportation and storage processes, but risk assessment cannot be carried out on the supply quality of suppliers.
Therefore, in view of the above problems, there is a need for a quantized power quality assessment method and a power quality trend parameter prediction model with adaptive parameter optimization, which are used for helping related technicians to better judge the quality of power quality and timely make risk early warning.
Disclosure of Invention
The application aims to overcome the defects in the prior art, and provides a power equipment quality trend prediction method and device based on an ARIMA model, which are used for helping related technicians to better predict the quality of power supplies of suppliers and timely make risk early warning.
In order to achieve the above purpose, the application is realized by adopting the following technical scheme:
in a first aspect, the present application provides a method for predicting a quality trend of a power device based on an ARIMA model, including the steps of:
step 1: acquiring raw data of a provider;
step 2: preprocessing the provider original data to obtain a provider score data set;
step 3: according to the provider score data set, constructing a prediction model of the provider based on an ARIMA model;
step 4: and obtaining future prediction data of the current moment of the supplier based on the prediction model.
Further, obtaining the raw data of the supplier includes:
and obtaining detection index values of all equipment of at least one annotation of the supplier to form an original data set.
Further, obtaining detection index values of all devices of at least one annotation of the supplier to form an original data set, including:
detection type co-detection for each deviceB, counting the seeds; a total of N batches of equipment are provided, and an original data set is set as R= { R 1 ,R 2 ,...,R i ,...,R N A set of detection index values R of the ith device i ＝{r 1 ,r 2 ,...,r j ,...,r b J=1, 2,., b, b is the total number of detection indicators; i=1, 2, N is the total number of devices in a batch;
each R represents a set of detection index values of each device, R 1 ,r 2 ,...,r j ,...,r b Representing the detection index value corresponding to each device.
Further, preprocessing the provider original data to obtain a data set reflecting equipment quality information, including:
performing scale transformation on the original data to enable the original data to contain quality information capable of reflecting equipment, and obtaining a data set X reflecting the quality information of the equipment;
and (3) performing score calculation on the data set X reflecting the equipment quality information to obtain a score data set of each batch of the supplier.
Further, scaling the original data to include quality information reflecting the device to obtain a data set X reflecting the quality information of the device, including:
for each detection index, acquiring a corresponding threshold value c given in a detection report;
subtracting a threshold value from each detection index value of each power device, and taking an absolute value:
x j ＝|r j -c|
processing all detection index values in sequence, and obtaining a data set X reflecting device quality information of the ith device i ＝{x 1 ,x 2 ,...,x j ,...,x b J=1, 2,., b, b is the total number of detection indicators; obtaining a data set X reflecting the quality information of the device:
X＝{X 1 ，X 2 …X i ，…X N }
wherein i=1, 2.
Further, performing score calculation on the data set X reflecting the quality information of the device to obtain the provider score data set, including:
let the weight corresponding to b indexes be w= { w 1 ,w 2 ,...,w j ,...,w b B is the total number of metrics, and a specific set of quality scores S for N devices is specified as:
S＝w*X∈R N×1 ＝{s 1 ，...，s N }
wherein R is N×1 Representing a real matrix of N rows and columns, s i A specific quality score representing the ith device;
averaging the scoring matrix S, namely:
providing the supplier with N W The score data set of each batch of the supplier can be obtained according to the method as follows:
further, constructing a predictive model of the vendor based on the ARIMA model according to the vendor score dataset, including:
the mathematical expression of the ARIMA model is as follows:
the above represents the current time sequence value x t Is the previous p-phase sequence value x t-1 ,…,x t-p And a first q phase error value ε t-1 ,…,ε t-q Is jointly affected by the past p-phase sequence value and the past q-phase error term.Representing d-order differential operation symbol, y t Representing the sequence x t Is a differential sequence of (a).
By comparison with sequence X S Performing ADF test to determine the optimal differential order d and the sequence after the differential;
the sequences after the difference are optimized in order p and q by BIC criterion. Determining the optimal order by searching a BIC minimum model to obtain optimal parameters p and q;
and building a prediction model of a specific supplier according to the determined optimal input parameters p, q and d of the ARIMA.
Further, by comparison of sequence X S Performing ADF test to determine the optimal differential order d, including:
step A: based on the current provider N W Scoring sequence for each batchObtaining a unit root P-value through ADF test;
and (B) step (B): taking a judgment threshold value as epsilon, if the P-value is smaller than epsilon, stabilizing the sequence, and taking d=0 at the moment; f, jumping to the step F;
step C: if the P-value is larger than epsilon, the sequence is not stable, and the difference is carried out once, and d=1 is taken at the moment; step D, jumping to the step D;
step D: performing ADF inspection on the differential sequence, if the P-value is smaller than epsilon, stabilizing the sequence, skipping to the step F, and if the P-value is larger than epsilon, not stabilizing the sequence, skipping to the step E;
step E: performing difference again on the sequence, updating d=d+1, and jumping to the step D;
step F: outputting the value of d and the sequence after d times of difference
Further, the sequences after the difference are subjected to BIC criterionAnd carrying out p-order and q-order optimization. Determining the optimal order by searching the minimum BIC model to obtain the most optimalThe preference parameters p, q include:
assuming that the best model orders p and q are located at [ p ], respectively l ，p m ]And [ q ] l ，q m ]Is within the range of the value of (2);
wherein p is l And q l Representing the minimum value of the values of p and q, p m And q m Represents the maximum value of p and q values;
decomposing BIC values corresponding to all p and q combinations in the calculation range, taking the p and q combination with the smallest BIC value as the optimal order of the model, and marking asAnd->The method of calculating BIC is as follows:
wherein the method comprises the steps ofLikelihood function representing estimates for p and q, < >>Indicating that at the current time k, i and j are integers, at p i ∈[p l ，p m ]And q i ∈[q l ,q m ]N samples.
Through the steps, the optimal input parameters p, q and d of ARIMA can be determined, and a prediction model of a specific supplier can be established. A specific algorithm flow chart is shown in fig. 3.
Further, based on the prediction model, obtaining future prediction data of the current moment of the provider includes:
in the constructed prediction model, a score data set and a prediction step length n of a provider are input, and data of n steps in the future of the current moment of the sequence are output.
In a second aspect, the present application provides an ARIMA model-based power equipment quality trend prediction apparatus, including:
an input module: for obtaining vendor raw data;
and a pretreatment module: the supplier score data set is used for preprocessing the supplier original data to obtain the supplier score data set;
model construction module: the method comprises the steps of establishing a predictive model of a provider based on an ARIMA model according to the provider score data set;
and a prediction module: and the method is used for obtaining future prediction data of the current moment of the supplier based on the prediction model.
In a third aspect, the present application provides an ARIMA model-based power equipment quality trend prediction apparatus, including a processor and a storage medium;
the storage medium is used for storing instructions;
the processor is configured to operate in accordance with the instructions to perform the method of the first aspect.
Compared with the prior art, the application has the beneficial effects that:
the method provided by the application provides a quantized electric power material quality evaluation method, and overcomes subjectivity of human judgment.
The application provides a prediction method for the quality trend of electric power materials, which can help related technicians to make risk assessment in time and avoid economic loss caused by receiving electric power equipment with poor quality.
Drawings
FIG. 1 is an AC arrester raw test data type;
fig. 2 is a lightning arrester device detection report (0.75 times dc reference voltage leakage current experiment);
fig. 3 is an algorithm flow chart.
Detailed Description
The application is further described below with reference to the accompanying drawings. The following examples are only for more clearly illustrating the technical aspects of the present application, and are not intended to limit the scope of the present application.
Embodiment one:
the embodiment provides a power equipment quality trend prediction method based on an ARIMA model, which comprises the following steps:
step 1: acquiring raw data of a provider;
obtaining vendor raw data, comprising:
and obtaining detection index values of all equipment of at least one annotation of the supplier to form an original data set.
The detection index value category of the original data is as shown in FIG. 1, and comprises
Specifically, obtaining the detection index values of all devices of at least one annotation of the supplier to form the original data set includes:
the detection type of each device is b in total; a total of N batches of equipment are provided, and an original data set is set as R= { R 1 ,R 2 ,...,R i ,...,R N A set of detection index values R of the ith device i ＝{r 1 ,r 2 ,...,r j ,...,r b J=1, 2,., b, b is the total number of detection indicators; i=1, 2, N is the total number of devices in a batch;
each R represents a set of detection index values of each device, R 1 ,r 2 ,...,r j ,...,r b Representing the detection index value corresponding to each device.
Step 2: preprocessing the provider original data to obtain a provider score data set;
preprocessing the original data of the provider to obtain a data set reflecting equipment quality information, wherein the data set comprises:
performing scale transformation on the original data to enable the original data to contain quality information capable of reflecting equipment, and obtaining a data set X reflecting the quality information of the equipment;
and (3) performing score calculation on the data set X reflecting the equipment quality information to obtain a score data set of each batch of the supplier.
Specifically, scaling the original data to include quality information reflecting the device to obtain a data set X reflecting the quality information of the device, including:
for each detection index, acquiring a corresponding threshold value c given in a detection report;
subtracting a threshold value from each detection index value of each power device, and taking an absolute value:
x j ＝|r j -c|
processing all detection index values in sequence, and obtaining a data set X reflecting device quality information of the ith device i ＝{x 1 ,x 2 ,...,x j ,...,x b J=1, 2,., b, b is the total number of detection indicators; obtaining a data set X reflecting the quality information of the device:
X＝{X 1 ，X 2 ...X i ，...X N }
wherein i=1, 2.
Specifically, the scoring calculation is performed on the data set X reflecting the quality information of the equipment to obtain the provider scoring data set, which includes:
let the weight corresponding to b indexes be w= { w 1 ,w 2 ,...,w j ,...,w b And b is the total index number, and specific mass scores of N lightning arresters are specified as follows:
S＝w*X∈R N×1 ＝{s 1 ，...，s N }
averaging the scoring matrix S, namely:
providing the supplier with N W The score data set of each batch of the supplier can be obtained according to the method as follows:
step 3: according to the provider score data set, constructing a prediction model of the provider based on an ARIMA model;
the mathematical expression of the ARIMA model is as follows:
the above represents the current time sequence value x t Is the previous p-phase sequence value x t-1 ,…,x t-p And a first q phase error value ε t-1 ,…,ε t-q Is jointly affected by the past p-phase sequence value and the past q-phase error term.Representing d-order differential operation symbol, y t Representing the sequence x t Is a differential sequence of (a).
By comparison with sequence X S Performing ADF test to determine the optimal differential order d and the sequence after the differential;
the sequences after the difference are optimized in order p and q by BIC criterion. Determining the optimal order by searching a BIC minimum model to obtain optimal parameters p and q;
and building a prediction model of a specific supplier according to the determined optimal input parameters p, q and d of the ARIMA.
Specifically, by comparison of sequence X S Performing ADF test to determine the optimal differential order d, including:
step A: based on the current provider N W Scoring sequence for each batchObtaining a unit root P-value through ADF test;
and (B) step (B): taking a judgment threshold value as epsilon, if the P-value is smaller than epsilon, stabilizing the sequence, and taking d=0 at the moment; f, jumping to the step F;
step C: if the P-value is larger than epsilon, the sequence is not stable, and the difference is carried out once, and d=1 is taken at the moment; step D, jumping to the step D;
step D: performing ADF inspection on the differential sequence, if the P-value is smaller than epsilon, stabilizing the sequence, skipping to the step F, and if the P-value is larger than epsilon, not stabilizing the sequence, skipping to the step E;
step E: performing difference again on the sequence, updating d=d+1, and jumping to the step D;
step F: outputting the value of d and the sequence after d times of difference
Specifically, the sequence after the difference is subjected to BIC criterionAnd carrying out p-order and q-order optimization. Determining the optimal order by searching the minimum BIC model to obtain optimal parameters p and q, wherein the method comprises the following steps:
assuming that the best model orders p and q are located at [ p ], respectively l ，p m ]And [ q ] l ，q m ]Is within the range of the value of (2);
wherein p is l And q l Representing the minimum value of the values of p and q, p m And q m Represents the maximum value of p and q values;
decomposing BIC values corresponding to all p and q combinations in the calculation range, taking the p and q combination with the smallest BIC value as the optimal order of the model, and marking asAnd->The method of calculating BIC is as follows:
wherein the method comprises the steps ofLikelihood function representing estimates for p and q, < >>Indicating that at the current time k, i and j are integers, at p i ∈[p l ，p m ]And q i ∈[q l ，q m ]N samples.
Through the steps, the optimal input parameters p, q and d of ARIMA can be determined, and a prediction model of a specific supplier can be established. A specific algorithm flow chart is shown in fig. 3.
Step 4: and obtaining future prediction data of the current moment of the supplier based on the prediction model.
Based on the prediction model, obtaining future prediction data of the current moment of the supplier comprises the following steps:
in the constructed prediction model, a score data set and a prediction step length n of a provider are input, and data of n steps in the future of the current moment of the sequence are output.
The technical solution of the present application is explained below from practical cases:
taking an ac lightning arrester of an electrical device as an example. The types of the detection data of the original power equipment are shown in fig. 1, and the total number of the detection types of each lightning arrester equipment is 15. Let a certain batch of lightning arresters number total N, let the original data set be R= { R 1 ,R 2 ,...,R i ,...,R N A set of detection index values R of the ith device i ＝{r 1 ,r 2 ,...,r j ,...,r 15 J=1, 2,. }, 15, 15 is the total number of detection indicators; i=1, 2, N is the total number of devices in a batch; each R represents a set of detection index values of each device, R 1 ,r 2 ,...,r j ,...,r 15 Representing the detection index value corresponding to each device. The electric power detection data such as current and voltage in the original data does not contain equipment quality information, namely, the original data cannot be used for evaluating the quality of materials, so that the original data needs to be subjected to scale transformation to contain quality information capable of reflecting equipment. The specific transformation process is as follows:
as shown in figure 2, the test report of the lightning arrester equipment is that the obtained data is a leakage current test under the reference voltage of 0.75 timesAnd data, wherein the required value for the lightning arrester device is 50uA or less. Thus, corresponding to the accompanying drawings, a lightning arrester detection data type: r1, R2 and R3, namely "leakage current under 0.75 times reference voltage" detection data are taken: x is x i ＝|r i -50|(i＝1，2，3).
For the rest of the data class data of the arrester (r= { R 4 ，r 5 ...r 15 Pre-processing is done according to the above formula calculation rules according to the threshold ranges given in the corresponding and detection reports (for each detection indicator, there is a corresponding threshold given in the detection report). Subtracting the threshold value from each detection index value of each power equipment, and taking the absolute value. ) Finally, a data set X reflecting the quality information of the equipment is obtained:
X＝{X 1 ，X 2 ...X i ，...X N }
wherein X is i ＝{x 1 ，x 2 ...x 15 }
Score calculation:
let the weight corresponding to 15 indexes be w= { w 1 ，w 2 ，...，w 15 -specifying a specific mass score for the N arresters as:
S＝w*X∈R N×1 ＝{s 1 ,…,s N }
averaging the scoring matrix S, namely:
providing the supplier with N W The score data set of each batch of the supplier can be obtained according to the method as follows:
after obtaining the scoring dataset for each batch of the supplier, the optimal differential order d is determined:
the mathematical expression of the ARIMA model is as follows:
the above represents the current time sequence value x t Is the previous p-phase sequence value x t-1 ,…,x t-p And a first q phase error value ε t-1 ,…,ε t-q Is jointly affected by the past p-phase sequence value and the past q-phase error term.Representing d-order differential operation symbol, y t Representing the sequence x t Is a differential sequence of (a). p, d, q are three key parameters for constructing ARIMA. The process of constructing an ARIMA prediction model using the power plant dataset is also referred to as the process of determining the parameters p, d, q.
Based on the current provider N W Scoring sequence for each batchIs provided with->Is the optimal estimate of d for ARIMA (p, d, q). For sequence X S ADF test is performed for the purpose of determining whether the sequence is a smooth sequence. The ADF test is carried out to obtain a unit root P-value, a judgment threshold value is generally taken to be 0.05, if the judgment threshold value is larger than 0.05, the sequence is unstable, d=1 is taken at the moment, the ADF test is carried out on the sequence after the difference again, if the sequence is not unstable, d=d+1 is carried out, otherwise, the value of d at the moment is output, and the value of the model d parameter of ARIMA is obtained.
Specifically, if P-value is less than 0.05, sequence X is specified S Smooth and stable takingIf the P-value is larger than 0.05, performing primary differential treatment on the sequence, and if the calculated P-value is smaller than 0.05 after primary differential treatment, takingSimilarly, for->And the like can also be obtained by the above treatment. By the above process, the optimal estimation of d value can be completed, and the obtained +.>Sequence after sub-differentiation->
(2) Determining optimal parameters p, q
p, q are two parameters of ARIMA (p, d, q) model, and represent the current time sequence value x used in model construction t Is the previous p-phase sequence value x t-1 ,…,x t-p And a first q phase error value ε t-1 ,…,ε t-q To perform model construction.
Using BIC criteria for the sequence after differencingAnd carrying out p-order and q-order optimization. And determining the optimal order by searching the minimum BIC model. The method comprises the following specific steps:
assuming that the best model orders p and q are located at [ p ], respectively l ，p m ]And [ q ] l ，q m ]Is within the range of values of (2). Wherein p is l And q l Representing the minimum value of the values of p and q, p m And q m The maximum value of p and q is represented, then the BIC values corresponding to all p and q combinations in the calculation range are decomposed, the p and q combination with the minimum BIC value is taken as the optimal order of the model, and the optimal order is recorded asAnd->The method of calculating BIC is as follows:
wherein the method comprises the steps ofLikelihood function representing estimates for p and q, < >>At the current time k, i and j are integers, and p i ∈[p l ，p m ]And q i ∈[q l ，q m ]N samples.
Through the steps, the optimal input parameters p, q and d of ARIMA can be determined, and a prediction model of a specific supplier can be established. A specific algorithm flow chart is shown in fig. 3.
The built model is input into the scoring sequence obtained in the step one
Three parameters p, d, q of the ARIMA model of the sequence can be calculated based on step one and step two using the input data.
After the model is built, the sequence and the prediction step length n are input, and data of n steps in the future of the current moment of the sequence can be output.
Embodiment two:
the embodiment provides a power equipment quality trend prediction device based on ARIMA model, which comprises:
an input module: for obtaining vendor raw data;
and a pretreatment module: the supplier score data set is used for preprocessing the supplier original data to obtain the supplier score data set;
model construction module: the method comprises the steps of establishing a predictive model of a provider based on an ARIMA model according to the provider score data set;
and a prediction module: and the method is used for obtaining future prediction data of the current moment of the supplier based on the prediction model.
The apparatus of this embodiment may be used to implement the method described in embodiment one.
Embodiment III:
the embodiment provides a power equipment quality trend prediction device based on an ARIMA model, which comprises a processor and a storage medium;
the storage medium is used for storing instructions;
the processor is configured to operate in accordance with the instructions to perform the method of embodiment one.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The foregoing is merely a preferred embodiment of the present application, and it should be noted that modifications and variations could be made by those skilled in the art without departing from the technical principles of the present application, and such modifications and variations should also be regarded as being within the scope of the application.
Claims (10)
1. The method for predicting the quality trend of the power equipment based on the ARIMA model is characterized by comprising the following steps of:
acquiring raw data of a provider;
preprocessing the provider original data to obtain a provider score data set;
according to the provider score data set, constructing a prediction model of the provider based on an ARIMA model;
and obtaining future prediction data of the current moment of the supplier based on the prediction model.
2. The ARIMA model-based power equipment quality trend prediction method according to claim 1, wherein obtaining vendor raw data comprises:
the detection type of each device is b in total; the number of the batch of equipment is N in total, and the original data set is R= { R 1 ,R 2 ,...,R i ,...,R N A set of detection index values R of the ith device i ＝{r 1 ,r 2 ,...,r j ,...,r b J=1, 2,., b, b is the total number of detection indicators; i=1, 2, N is the total number of devices in a batch;
each R represents a set of detection index values of a device, R 1 ,r 2 ,...,r j ,...,r b Representing the detection index value corresponding to each device.
3. The ARIMA model-based power equipment quality trend prediction method according to claim 2, wherein preprocessing the vendor raw data to obtain a data set reflecting equipment quality information comprises:
performing scale transformation on the original data to enable the original data to contain quality information capable of reflecting equipment, and obtaining a data set X reflecting the quality information of the equipment;
and (3) performing score calculation on the data set X reflecting the equipment quality information to obtain a score data set of each batch of the supplier.
4. The ARIMA model-based power equipment quality trend prediction method according to claim 3, wherein scaling the raw data to include quality information reflecting the equipment to obtain a data set X reflecting the equipment quality information comprises:
for each detection index, acquiring a corresponding threshold value c given in a detection report;
subtracting a threshold value from each detection index value of each power device, and taking an absolute value:
x j ＝|r j -c|
processing all detection index values in sequence, and obtaining a data set X reflecting device quality information of the ith device i ＝{x 1 ,x 2 ,...,x j ,...,x b J=1, 2,., b, b is the total number of detection indicators; obtaining a data set X reflecting the quality information of the device:
X＝{X 1 ，X 2 ...X i ，...X N }
wherein i=1, 2.
5. The ARIMA model-based power equipment quality trend prediction method according to claim 3, wherein performing score calculation on the data set X reflecting equipment quality information to obtain the provider score data set includes:
let the weight corresponding to b indexes be w= { w 1 ,w 2 ,...,w j ,...,w b B is the total number of metrics, and a specific set of quality scores S for N devices is specified as:
s＝w*X∈R N×1 ＝{s 1 ，...，s N }
wherein R is N×1 Representing a real matrix of N rows and columns, s i A specific quality score representing the ith device;
averaging the scoring matrix S, namely:
providing the supplier with N W Each batch, the score dataset for each batch of the supplier is obtained as:
6. the ARIMA model-based power plant quality trend prediction method according to claim 1, wherein constructing a prediction model of a provider based on the ARIMA model according to the provider score data set comprises:
the mathematical expression of the ARIMA model is as follows:
the above represents the current time sequence value x t Is the previous p-phase sequence value x t-1 ,…,x t-p And a first q phase error value ε t-1 ,…,ε t-q Is jointly affected by the past p-phase sequence value and the past q-phase error term;representing d-order differential operation symbol, y t Representing the sequence x t Is a differential sequence of (a);
by comparison with sequence X S Performing ADF test to determine the optimal differential order d and the sequence after the differential;
optimizing the sequences after difference in p and q orders through BIC criteria; determining the optimal order by searching a BIC minimum model to obtain optimal parameters p and q;
and building a prediction model of a specific supplier according to the determined optimal input parameters p, q and d of the ARIMA.
7. The ARIMA model-based power plant quality trend prediction method according to claim 6, wherein the power plant quality trend is predicted by comparing the sequence X S Performing ADF test to determine the optimal differential order d, including:
step A: based on the current provider N W Scoring sequence for each batchObtaining a unit root P-value through ADF test;
and (B) step (B): taking a judgment threshold value as epsilon, if the P-value is smaller than epsilon, stabilizing the sequence, and taking d=0 at the moment; f, jumping to the step F;
step C: if the P-value is larger than epsilon, the sequence is not stable, and the difference is carried out once, and d=1 is taken at the moment; step D, jumping to the step D;
step D: performing ADF inspection on the differential sequence, if the P-value is smaller than epsilon, stabilizing the sequence, skipping to the step F, and if the P-value is larger than epsilon, not stabilizing the sequence, skipping to the step E;
step E: performing difference again on the sequence, updating d=d+1, and jumping to the step D;
step F: outputting the value of d and the sequence after d times of difference
For sequences after differencing by BIC criterionOptimizing the p and q orders; determining the optimal order by searching the minimum BIC model to obtain optimal parameters p and q, wherein the method comprises the following steps:
assuming that the best model orders p and q are located at [ p ], respectively l ，p m ]And [ q ] l ，q m ]Is within the range of the value of (2);
wherein p is l And q l Representing the minimum value of the values of p and q, p m And q m Represents the maximum value of p and q values;
decomposing BIC values corresponding to all p and q combinations in the calculation range, taking the p and q combination with the smallest BIC value as the optimal order of the model, and marking asAnd->The method of calculating BIC is as follows:
wherein the method comprises the steps ofLikelihood function representing estimates for p and q, < >>Indicating that at the current time k, i and j are integers, at p i ∈[p l ，p m ]And q i ∈[q l ，q m ]N samples;
through the steps, the optimal input parameters p, q and d of ARIMA can be determined, and a prediction model of a specific provider is established; a specific algorithm flow chart is shown in fig. 3.
8. The ARIMA model-based power equipment quality trend prediction method according to claim 1, wherein obtaining predicted data of the future of the current time of the provider based on the prediction model comprises:
in the constructed prediction model, a score data set and a prediction step length n of a provider are input, and data of n steps in the future of the current moment of the sequence are output.
9. An ARIMA model-based power equipment quality trend prediction device, comprising:
an input module: for obtaining vendor raw data;
and a pretreatment module: the supplier score data set is used for preprocessing the supplier original data to obtain the supplier score data set;
model construction module: the method comprises the steps of establishing a predictive model of a provider based on an ARIMA model according to the provider score data set;
and a prediction module: and the method is used for obtaining future prediction data of the current moment of the supplier based on the prediction model.
10. The power equipment quality trend prediction device based on the ARIMA model is characterized by comprising a processor and a storage medium;
the storage medium is used for storing instructions;
the processor is operative to perform the method of any one of claims 1-8 in accordance with the instructions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310753599.5A CN116777281A (en) | 2023-06-26 | 2023-06-26 | ARIMA model-based power equipment quality trend prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310753599.5A CN116777281A (en) | 2023-06-26 | 2023-06-26 | ARIMA model-based power equipment quality trend prediction method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116777281A true CN116777281A (en) | 2023-09-19 |
Family
ID=87990926
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310753599.5A Pending CN116777281A (en) | 2023-06-26 | 2023-06-26 | ARIMA model-based power equipment quality trend prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116777281A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117688464A (en) * | 2024-02-04 | 2024-03-12 | 国网上海市电力公司 | Hidden danger analysis method and system based on multi-source sensor data |
-
2023
- 2023-06-26 CN CN202310753599.5A patent/CN116777281A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117688464A (en) * | 2024-02-04 | 2024-03-12 | 国网上海市电力公司 | Hidden danger analysis method and system based on multi-source sensor data |
| CN117688464B (en) * | 2024-02-04 | 2024-04-19 | 国网上海市电力公司 | Hidden danger analysis method and system based on multi-source sensor data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Collin et al. | Extending approximate Bayesian computation with supervised machine learning to infer demographic history from genetic polymorphisms using DIYABC Random Forest | |
| Lu et al. | An MDL approach to the climate segmentation problem | |
| CN116777281A (en) | ARIMA model-based power equipment quality trend prediction method and device | |
| CN111967655A (en) | Short-term load prediction method and system | |
| CN112150237A (en) | Multi-model fused order overdue early warning method, device, equipment and storage medium | |
| CN111025041A (en) | Electric vehicle charging pile monitoring method and system, computer equipment and medium | |
| CN111737249A (en) | Abnormal data detection method and device based on Lasso algorithm | |
| US11373285B2 (en) | Image generation device, image generation method, and image generation program | |
| CN114330090A (en) | Defect detection method and device, computer equipment and storage medium | |
| Bordes et al. | Partially observed competing degradation processes: modeling and inference | |
| CN116629918A (en) | User consumption prediction method and system based on cross-border electronic commerce | |
| Ding et al. | Degradation analysis with nonlinear exponential‐dispersion process: Bayesian offline and online perspectives | |
| CN115238583B (en) | Business process remaining time prediction method and system supporting incremental log | |
| CN111080037A (en) | Short-term power load prediction method and device based on deep neural network | |
| CN114363004B (en) | Risk assessment method, risk assessment device, computer equipment and storage medium | |
| CN114648178A (en) | Operation and maintenance strategy optimization method of electric energy metering device based on DDPG algorithm | |
| Boyko et al. | Methodology for Estimating the Cost of Construction Equipment Based on the Analysis of Important Characteristics Using Machine Learning Methods | |
| Pedrielli et al. | Metamodel-based quantile estimation for hedging control of manufacturing systems | |
| Galatro et al. | Data Analytics for Process Engineers: Prediction, Control and Optimization | |
| CN117370824B (en) | Intelligent monitoring method and system for air inlet state of air compressor | |
| CN117422314B (en) | Enterprise data evaluation method and equipment based on big data analysis | |
| CN117764276A (en) | Method and system based on general model for calculating carbon footprint of power grid material product | |
| CN118395200A (en) | Method, device, computer equipment and storage medium for determining matching user | |
| CN117931883A (en) | Abnormality detection method, abnormality detection device, abnormality detection apparatus, and computer-readable storage medium | |
| CN117745343A (en) | Message delivery effect diagnosis method and device and computer readable storage medium |
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 |