CN108009750A - A kind of power grid O&M efficiency rating method based on DEA and SVM - Google Patents
A kind of power grid O&M efficiency rating method based on DEA and SVM Download PDFInfo
- Publication number
- CN108009750A CN108009750A CN201711422113.0A CN201711422113A CN108009750A CN 108009750 A CN108009750 A CN 108009750A CN 201711422113 A CN201711422113 A CN 201711422113A CN 108009750 A CN108009750 A CN 108009750A
- Authority
- CN
- China
- Prior art keywords
- msub
- mrow
- efficiency
- dea
- represent
- 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 28
- 238000011156 evaluation Methods 0.000 claims abstract description 45
- 230000006870 function Effects 0.000 claims abstract description 20
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 19
- 230000007613 environmental effect Effects 0.000 claims abstract description 14
- 238000012549 training Methods 0.000 claims abstract description 5
- 238000007621 cluster analysis Methods 0.000 claims abstract description 4
- 239000013598 vector Substances 0.000 claims description 15
- 238000012706 support-vector machine Methods 0.000 claims description 8
- 238000004458 analytical method Methods 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 5
- 230000000694 effects Effects 0.000 claims description 4
- 238000012937 correction Methods 0.000 claims description 3
- 238000006073 displacement reaction Methods 0.000 claims description 3
- 238000004519 manufacturing process Methods 0.000 abstract description 4
- 238000011160 research Methods 0.000 abstract description 2
- 238000005457 optimization Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 238000013210 evaluation model Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 230000008439 repair process Effects 0.000 description 2
- 238000003786 synthesis reaction Methods 0.000 description 2
- 206010010356 Congenital anomaly Diseases 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 108010001267 Protein Subunits Proteins 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000003344 environmental pollutant Substances 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000001802 infusion Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 231100000572 poisoning Toxicity 0.000 description 1
- 230000000607 poisoning effect Effects 0.000 description 1
- 231100000719 pollutant Toxicity 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 230000000452 restraining effect Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Tourism & Hospitality (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Educational Administration (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Public Health (AREA)
- Primary Health Care (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of power grid O&M efficiency rating method based on DEA and SVM, all data of power grid decision package DMU are divided into input (input), output (output) and environmental data first, environmental data is used for cluster analysis, by being evaluated again after fuzzy clustering, so as to ensure the similar comparativity of decision package DMU.Data will be output and input to be used in DEA evaluations, unexpected input and output are innovatively considered and it is expected input/output relation, optimization tradition DEA algorithms are limited only for low the tactful of input high production.Importance in view of time dimension to evaluation, by the forgetting function for adding the more time dimensions of full size with thoroughly evaluating DMU efficiency absolute score and relative score.The efficiency label of DEA evaluation results is used to instruct SVM algorithm to carry out supervised learning training, training result model can be used for simplifying evaluation rubric.The method of the present invention has important scientific meaning and application value to the research of efficiency rating related direction.
Description
Technical field
The present invention relates to a kind of power grid O&M efficiency rating method based on DEA and SVM, determines for all kinds of in power grid O&M
Plan unit efficiencies evaluation problem.
Background technology
As increasingly fierce market environment and the interconnection networking digitization development trend quickly changed, power grid are wanted
Future electrical energy obtains vantage point in market, it is necessary to efficiently uses data driven analysis instrument, the daily operation benefits of lifting company
With enterprise's input-output ratio.Wherein, efficiency rating is the effective of a kind of measurement development of company level and same period intratype competition power
Means.By efficiency rating, self poisoning can be specified in enterprise, and quantitative analysis colleague's gap, abandons blindly to put into the past and disregard
The thick spacious Management Pattern of Enterprises of ROI.
DEA (DEA) method is that the one kind researched and proposed by Charnes et al. is based on low input high production
Evaluation decision package (DMU) method of the lifting entirety ROI of principle.Main research and produce may gather relevant input and output effect
Rate studies a question.The ratio that input and output are typically based in traditional DEA methods is object function, by solving the convex excellent of belt restraining
Change problem attempts to find production proportions face, by production proportions face, can divide decision package two for efficiency unit and non-effect
Rate unit.Further, calculated by evaluating the effective of efficiency score, the sensitivity point of decision package indices can be obtained
Analysis and leading surface distance analysis are as a result, effectively planning instructs decision package in following input and output resource distribution process.
When being evaluated using DEA models DMU, existing tradition DEA evaluation methods have the following disadvantages:First,
Usually evaluation lacks similar comparable consider.Illustrated with grid company, the province environment of coastal province and inland the Northwest
There are larger difference, long-term incommensurability caused by congenital environmental factor, if if therefore carry out efficiency rating when without advance
Environmental variance factor is rejected, ensures similar comparativity, it will cause the evaluation of inland the Northwest to be chronically at unfavorable position, so that
Lose examination meaning.Secondly, traditional DEA methods are all necessary for non-negative in the selection of input/output variable, and universal defeated
Enter and the direction that exports has limitation, when considering under some new scenes, Model suitability is limited.Furthermore script DEA models lack
In the evaluation model of considering for time dimension, mostly quiet hour, time of fusion scale considers the efficiency rating mould of variation tendency
Type needs to be established.Finally, traditional method for making evaluation model analyzes N number of decision package and needs to solve N*N-1 convex Optimized model,
Need to calculate the efficiency score of multiple DMU after the completion of solution to obtain the judgement in whether ahead of the curve face, the operation of process numerous and complicated
Step is more.
The content of the invention
In view of the above-mentioned deficiencies in the prior art, it is an object of the present invention to provide a kind of power grid O&M efficiency based on DEA and SVM
Evaluation method, including for the DEA Modifying models that special scenes carry out, consider the full size evaluation of time dimension, utilize
Algorithm of support vector machine carries out classification eases to evaluation procedure.
The purpose of the present invention is what is be achieved through the following technical solutions:A kind of power grid O&M efficiency based on DEA and SVM
Evaluation method, this method comprise the following steps:
Step 1, the similar comparativity for guarantee decision package DMU, using fuzzy clustering algorithm, by decision package DMU roots
According to cluster result divide into several classes, influence of the environmental factor to efficiency rating is rejected.Fuzzy clustering algorithm is preset in cluster
Calculation mesh c and Fuzzy Exponential m, passes through subordinated-degree matrix μijWith cluster centre cjIteration renewal, until loss function JmConvergence.
Subordinated-degree matrix μijWith cluster centre cjIteration renewal rule meet the following formula:
Loss function JmConvergence rule meets the following formula:
‖Jm‖ < ε
Wherein, wherein ε is a dimensionless;xiRepresent sample to be clustered, d (xi,cj) represent sample x to be clusterediWith gathering
Class center cjEuclidean distance;I represents the subscript of sample data, and k represents the subscript of cluster centre;N represents total sample number, K tables
Show the sum of cluster centre.
Step 2, the diversity scene for considering efficiency rating, the unexpected input and output of comprehensive study and expectation inputoutput pair
The amendment of model influences, final it is expected to repair with the DEA of unexpected input and output after non-linear linear variable displacement equivalent processes
The object function of positive model meets the following formula:
WhereinRepresent the efficiency score of expectation input direction,Represent the efficiency score in desired output direction,
Represent the efficiency score in non-universal gauge bosons Z' direction, object function summation represents that gross efficiency score is maximum.And the pact of DEA correction models
Beam condition meets the following formula:
∑k(λk+μk)Xik≤Xir
∑k(λk+μk)Apk≥Apr+θap
∑kλkYjk≥Yjr+θyj
∑kλkBqk≥Bqr-θbq
∑k(λk+μk)=1
Wherein, r represents the subscript of decision package to be evaluated, and k represents the subscript of all decision packages, and K represents all and determines
The sum of plan unit;XikRepresent the i dimensional vectors of unexpected input, ApkRepresent it is expected the p dimensional vectors inputted, YjkRepresent it is expected defeated
The j dimensional vectors gone out, BqkRepresent the q dimensional vectors of non-universal gauge bosons Z'.Decision variable λkIt is the linear combination weight of k-th of decision package
The factor, μkIt is weak Podinovskis parameters.
Step 3, the efficiency score for obtaining step 2 are denoted as absolute efficiency score θ, and it is special to introduce the evaluation changed over time
Sign, obtains relative evaluation efficiency score θdelta.Comprehensive absolute evaluation and relative evaluation are calculated final comprehensive as a result, using weight beta
Close efficiency score θ*Meet the following formula:
θ*=θ+β θdelta,β∈(0,+∞)
Step 4, the overall efficiency score θ for obtaining step 3*, attenuation coefficient α and attenuation function are introduced, characterizes full-time
Dimension Synthesis efficiency score θ**, meet the following formula:
Wherein, t0Represent Evaluation: Current time point, τ represents historical time point.
Step 5, obtain the label whether decision package DMU is in efficiency leading surface according to step 1-4, using the label as
The tag along sort of support vector machines, carries out learning training to support vector machines model by historical data, is simplified
Classification of assessment model;By the simplified classification of assessment model of decision package DMU data input to be evaluated, immediately arriving at DMU is
The no classification results in efficiency leading surface.
Further, the step 1, the data preprocessing phase of step 2 are directed to DMU data acquisition systems, division input data,
Output data and environmental data, wherein environmental data are used for fuzzy cluster analysis, and input data and output data are used to improve DEA
Algorithm evaluation analysis.
Further, DMU is ranked up by super efficiency SE-DEA on efficiency leading surface, ensures the gradient of evaluation result
And validity.
Further, the algorithm of support vector machine in the step 5 is used for two classification problems, and classification results can interpolate that
Whether DMU is effective, and the numerical value of estimated efficiency score is capable of by SVR regression models.
The innovative point of the present invention is:The present invention is public to power grid using a kind of improved DEA (DEA) method
Department carries out evaluation analysis.Specifically, the present invention is innovatively polymerize using fuzzy clustering method sub-unit, improves evaluation
It is deficient in model time dimension, consider the full size evaluation problem of absolute evaluation and relative evaluation, consider that mode input is defeated
Go out the Model suitability under the new scene that multi-direction vector is brought, finally innovatively attempt fused data sorting technique to reach letter
Change the effect of cumbersome evaluation rubric.
Accompanying drawing content
Fig. 1 is the power grid O&M efficiency rating method flow diagram of the invention based on DEA and SVM;
Fig. 2 represents the double cluster centre degree of membership distribution maps of fuzzy clustering algorithm;
Fig. 3 represents loss function convergent tendency figure;
Fig. 4 represents first kind power grid decision package absolute efficiency and relative efficiency evaluation figure;
Fig. 5 represents the second class power grid decision package absolute efficiency and relative efficiency evaluation figure;
Fig. 6 represents first kind power grid decision package SE-DEA absolute efficiencies and relative efficiency evaluation figure;
Fig. 7 represents the second class power grid decision package SE-DEA absolute efficiencies and relative efficiency evaluation figure.
Embodiment
The present invention is described in further detail with specific embodiment below in conjunction with the accompanying drawings.
A kind of power grid O&M efficiency rating method based on DEA and SVM provided in this embodiment, comprises the following steps:
Step 1, the similar comparativity for guarantee decision package DMU, using fuzzy clustering algorithm, by decision package DMU roots
According to cluster result divide into several classes, influence of the environmental factor to efficiency rating is rejected.Fuzzy clustering algorithm is preset in cluster
Calculation mesh c and Fuzzy Exponential m, passes through subordinated-degree matrix μijWith cluster centre cjIteration renewal, until loss function JmConvergence.
Subordinated-degree matrix μijWith cluster centre cjIteration renewal rule meet the following formula:
Loss function JmConvergence rule meets the following formula:
‖Jm‖ < ε
Wherein, wherein ε is a dimensionless;xiRepresent sample to be clustered, d (xi,cj) represent sample x to be clusterediWith gathering
Class center cjEuclidean distance;I represents the subscript of sample data, and k represents the subscript of cluster centre;N represents total sample number, K tables
Show the sum of cluster centre.
Step 2, the diversity scene for considering efficiency rating, the unexpected input and output of comprehensive study and expectation inputoutput pair
The amendment of model influences, final it is expected to repair with the DEA of unexpected input and output after non-linear linear variable displacement equivalent processes
The object function of positive model meets the following formula:
WhereinRepresent the efficiency score of expectation input direction,Represent the efficiency score in desired output direction,
Represent the efficiency score in non-universal gauge bosons Z' direction, object function summation represents that gross efficiency score is maximum.And the pact of DEA correction models
Beam condition meets the following formula:
∑k(λk+μk)Xik≤Xir
∑k(λk+μk)Apk≥Apr+θap
∑kλkYjk≥Yjr+θyj
∑kλkBqk≥Bqr-θbq
∑k(λk+μk)=1
Wherein, r represents the subscript of decision package to be evaluated, and k represents the subscript of all decision packages, and K represents all and determines
The sum of plan unit;XikRepresent the i dimensional vectors of unexpected input, ApkRepresent it is expected the p dimensional vectors inputted, YjkRepresent it is expected defeated
The j dimensional vectors gone out, BqkRepresent the q dimensional vectors of non-universal gauge bosons Z'.Decision variable λkIt is the linear combination weight of k-th of decision package
The factor, μkIt is weak Podinovskis parameters.
Step 3, the efficiency score for obtaining step 2 are denoted as absolute efficiency score θ, and it is special to introduce the evaluation changed over time
Sign, obtains relative evaluation efficiency score θdelta.Comprehensive absolute evaluation and relative evaluation are calculated final comprehensive as a result, using weight beta
Close efficiency score θ*Meet the following formula:
θ*=θ+β θdelta,β∈(0,+∞)
Step 4, the overall efficiency score θ for obtaining step 3*, attenuation coefficient α and attenuation function are introduced, characterizes full-time
Dimension Synthesis efficiency score θ**, meet the following formula:
Wherein, t0Represent Evaluation: Current time point, τ represents historical time point.
Step 5, obtain the label whether decision package DMU is in efficiency leading surface according to step 1-4, using the label as
The tag along sort of support vector machines, carries out learning training to support vector machines model by historical data, is simplified
Classification of assessment model;By the simplified classification of assessment model of decision package DMU data input to be evaluated, immediately arriving at DMU is
The no classification results in efficiency leading surface.
Further, the step 1, the data preprocessing phase of step 2 are directed to DMU data acquisition systems, division input data,
Output data and environmental data, wherein environmental data are used for fuzzy cluster analysis, and input data and output data are used to improve DEA
Algorithm evaluation analysis.
Further, DMU is further sequence on the efficiency leading surface of the step 2, can be with using super efficiency SE-DEA
Effectively overcome the sequencing problem of DMU on leading surface, ensure the gradient and validity of evaluation result.
Further, the algorithm of support vector machine of affiliated step 5 is used for two classification problems, and classification results can interpolate that DMU
Whether effectively, the numerical value of estimated efficiency score is capable of by SVR regression models.
The present embodiment utility data dimension is as follows:Environmental data mainly includes generated energy (GN), electricity consumption number of users (CN),
Electric energy sales volume (SA), generator installation number (GR), power infrastructures number (FA), the total installed capacity of generator (NC) and power generation
Machine summer total capacity (SC).Five input input datas are mainly by total infusion of financial resources (TC), generator quantity (GR) and electric power
Three unexpected inputs of facility facility number (FA), and the total installed capacity of generator (NC) and generator summer total capacity (SC) two
It is a it is expected to input composition.And mainly include three desired outputs in terms of output output data:It is generated energy (GN) respectively, electricity pin
Measure (SA), total income (RE);Three non-universal gauge bosons Z's are mainly disposal of pollutants correlation:It is CO2 (CO), SO2 (SO) and NOx respectively
(NO) discharge capacity.
Fig. 1 is power grid O&M efficiency rating method flow diagram of the present embodiment based on DEA and SVM.Fig. 2 represents to be directed to example
Degree of membership comparison diagram after being clustered using fuzzy clustering FCM algorithms, when first kind degree of membership is more than the second class degree of membership, then
Decision package belongs to the first kind;It is on the contrary then belong to the second class.Fig. 3 represents that clustering algorithm is updated iteration according to cluster goodness
Loss function convergent tendency figure.As iteration carries out, loss function Fast Convergent, power grid decision package data is divided into some
Class, it is two classes to select cluster number after the cluster goodness criterion optimizing of this example.
Fig. 4, Fig. 5 represent that the first kind and the second class power grid decision package use the overall merit after correcting DEA algorithms respectively
As a result.Fig. 6, Fig. 7 are obtained to distinguish the further sequence of power grid decision package on efficiency leading surface and using SE-DEA models
The evaluation evaluation result obtained.
The foregoing is merely the preferred embodiment of the invention, is not intended to limit the invention creation, all at this
All any modification, equivalent and improvement made within the spirit and principle of innovation and creation etc., should be included in the invention
Protection domain within.
Claims (4)
- A kind of 1. power grid O&M efficiency rating method based on DEA and SVM, it is characterised in that this method comprises the following steps:Step 1, using fuzzy clustering algorithm, by decision package DMU according to cluster result divide into several classes, reject environmental factor pair The influence of efficiency rating;Fuzzy clustering algorithm presets cluster centre number c and Fuzzy Exponential m, passes through subordinated-degree matrix μij With cluster centre cjIteration renewal, until loss function JmConvergence;Subordinated-degree matrix μijWith cluster centre cjIteration renewal rule Then meet the following formula:<mrow> <msub> <mi>c</mi> <mi>j</mi> </msub> <mo>=</mo> <mfrac> <mrow> <msubsup> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>N</mi> </msubsup> <msubsup> <mi>&mu;</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mi>m</mi> </msubsup> <msub> <mi>x</mi> <mi>i</mi> </msub> </mrow> <mrow> <msubsup> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>N</mi> </msubsup> <msubsup> <mi>&mu;</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mi>m</mi> </msubsup> </mrow> </mfrac> </mrow><mrow> <msub> <mi>&mu;</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mo>=</mo> <mfrac> <mn>1</mn> <mrow> <msubsup> <mi>&Sigma;</mi> <mrow> <mi>k</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>C</mi> </msubsup> <mfrac> <mrow> <mi>d</mi> <msup> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mi>i</mi> </msub> <mo>,</mo> <msub> <mi>c</mi> <mi>j</mi> </msub> <mo>)</mo> </mrow> <mfrac> <mn>2</mn> <mrow> <mi>m</mi> <mo>-</mo> <mn>1</mn> </mrow> </mfrac> </msup> </mrow> <mrow> <mi>d</mi> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mi>i</mi> </msub> <mo>,</mo> <msub> <mi>c</mi> <mi>k</mi> </msub> <mo>)</mo> </mrow> </mrow> </mfrac> </mrow> </mfrac> </mrow>Loss function JmConvergence rule meets the following formula:<mrow> <msub> <mi>J</mi> <mi>m</mi> </msub> <mo>=</mo> <munderover> <mo>&Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>N</mi> </munderover> <munderover> <mo>&Sigma;</mo> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>C</mi> </munderover> <msubsup> <mi>&mu;</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mi>m</mi> </msubsup> <msup> <mi>d</mi> <mn>2</mn> </msup> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mi>i</mi> </msub> <mo>,</mo> <msub> <mi>c</mi> <mi>j</mi> </msub> <mo>)</mo> </mrow> </mrow>‖Jm‖ < εWherein ε is a dimensionless;xiRepresent sample to be clustered, d (xi,cj) represent sample x to be clusterediWith cluster centre cj's Euclidean distance;I represents the subscript of sample data, and k represents the subscript of cluster centre;N represents total sample number, and K represents cluster centre Sum;Step 2, the diversity scene for considering efficiency rating, the unexpected input and output of comprehensive study and expectation inputoutput pair model Amendment influence, it is final it is expected and the DEA of unexpected input and output corrects mould after non-linear linear variable displacement equivalent processes The object function of type meets the following formula:<mrow> <mi>&theta;</mi> <mo>=</mo> <msub> <mi>Max&Sigma;</mi> <mi>p</mi> </msub> <msub> <mi>&theta;</mi> <msub> <mi>a</mi> <mi>p</mi> </msub> </msub> <mo>+</mo> <msub> <mi>&Sigma;</mi> <mi>j</mi> </msub> <msub> <mi>&theta;</mi> <msub> <mi>y</mi> <mi>j</mi> </msub> </msub> <mo>+</mo> <msub> <mi>&Sigma;</mi> <mi>q</mi> </msub> <msub> <mi>&theta;</mi> <msub> <mi>b</mi> <mi>q</mi> </msub> </msub> </mrow>WhereinRepresent the efficiency score of expectation input direction,Represent the efficiency score in desired output direction,Represent non- The efficiency score in desired output direction, object function summation represent that gross efficiency score is maximum;The constraints of DEA correction models meets the following formula:∑k(λk+μk)Xik≤Xir<mrow> <msub> <mi>&Sigma;</mi> <mi>k</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>&lambda;</mi> <mi>k</mi> </msub> <mo>+</mo> <msub> <mi>&mu;</mi> <mi>k</mi> </msub> <mo>)</mo> </mrow> <msub> <mi>A</mi> <mrow> <mi>p</mi> <mi>k</mi> </mrow> </msub> <mo>&GreaterEqual;</mo> <msub> <mi>A</mi> <mrow> <mi>p</mi> <mi>r</mi> </mrow> </msub> <mo>+</mo> <msub> <mi>&theta;</mi> <msub> <mi>a</mi> <mi>p</mi> </msub> </msub> </mrow><mrow> <msub> <mi>&Sigma;</mi> <mi>k</mi> </msub> <msub> <mi>&lambda;</mi> <mi>k</mi> </msub> <msub> <mi>Y</mi> <mrow> <mi>j</mi> <mi>k</mi> </mrow> </msub> <mo>&GreaterEqual;</mo> <msub> <mi>Y</mi> <mrow> <mi>j</mi> <mi>r</mi> </mrow> </msub> <mo>+</mo> <msub> <mi>&theta;</mi> <msub> <mi>y</mi> <mi>j</mi> </msub> </msub> </mrow><mrow> <msub> <mi>&Sigma;</mi> <mi>k</mi> </msub> <msub> <mi>&lambda;</mi> <mi>k</mi> </msub> <msub> <mi>B</mi> <mrow> <mi>q</mi> <mi>k</mi> </mrow> </msub> <mo>&GreaterEqual;</mo> <msub> <mi>B</mi> <mrow> <mi>q</mi> <mi>r</mi> </mrow> </msub> <mo>-</mo> <msub> <mi>&theta;</mi> <msub> <mi>b</mi> <mi>q</mi> </msub> </msub> </mrow>∑k(λk+μk)=1<mrow> <msub> <mi>&lambda;</mi> <mi>k</mi> </msub> <mo>,</mo> <msub> <mi>&mu;</mi> <mi>k</mi> </msub> <mo>&GreaterEqual;</mo> <mn>0</mn> <mo>,</mo> <mo>&ForAll;</mo> <mi>k</mi> <mo>&Element;</mo> <mi>K</mi> </mrow>Wherein, r represents the subscript of decision package to be evaluated, and k represents the subscript of all decision packages, and K represents all decision packages Sum;XikRepresent the i dimensional vectors of unexpected input, ApkRepresent it is expected the p dimensional vectors inputted, YjkRepresent the j dimensions of desired output Vector, BqkRepresent the q dimensional vectors of non-universal gauge bosons Z';Decision variable λkIt is the linear combination weight factor of k-th of decision package, μk It is weak Podinovskis parameters;Step 3, by the efficiency score θ that step 2 obtains be denoted as absolute efficiency score, introduces the evaluating characteristic changed over time, obtains Obtain relative evaluation efficiency score θdelta;Comprehensive absolute evaluation and relative evaluation calculate final comprehensive effect as a result, using weight beta Rate score θ*:θ*=θ+β θdelta,β∈(0,+∞)Step 4, the overall efficiency score θ for obtaining step 3*, attenuation coefficient α and attenuation function are introduced, it is comprehensive to characterize full time scale Close efficiency score θ**:<mrow> <msup> <mi>&theta;</mi> <mrow> <mo>*</mo> <mo>*</mo> </mrow> </msup> <mo>=</mo> <munder> <mo>&Sigma;</mo> <mi>&tau;</mi> </munder> <msup> <mi>e</mi> <mrow> <mo>-</mo> <mi>&alpha;</mi> <mrow> <mo>(</mo> <msub> <mi>t</mi> <mn>0</mn> </msub> <mo>-</mo> <mi>&tau;</mi> <mo>)</mo> </mrow> </mrow> </msup> <mrow> <mo>(</mo> <mi>&theta;</mi> <mo>+</mo> <msup> <mi>&beta;&theta;</mi> <mrow> <mi>d</mi> <mi>e</mi> <mi>l</mi> <mi>t</mi> <mi>a</mi> </mrow> </msup> <mo>)</mo> </mrow> </mrow>Wherein, t0Represent Evaluation: Current time point, τ represents historical time point;Step 5, obtain the label whether decision package DMU is in efficiency leading surface according to step 1-4, using the label as support The tag along sort of vector machine SVM, carries out learning training, what is be simplified comments by historical data to support vector machines model Valency disaggregated model;By the simplified classification of assessment model of decision package DMU data to be evaluated input, immediately arrive at whether DMU is in The classification results of efficiency leading surface.
- A kind of 2. power grid O&M efficiency rating method based on DEA and SVM according to claim 1, it is characterised in that number The Data preprocess stage is directed to DMU data acquisition systems, and division input data, output data and environmental data, wherein environmental data are used for Fuzzy cluster analysis, input data and output data are used to improve DEA algorithm evaluation analysis.
- A kind of 3. power grid O&M efficiency rating method based on DEA and SVM according to claim 1, it is characterised in that institute State DMU on efficiency leading surface to be ranked up by super efficiency SE-DEA, ensure the gradient and validity of evaluation result.
- A kind of 4. power grid O&M efficiency rating method based on DEA and SVM according to claim 1, it is characterised in that branch Hold vector machine algorithm and be used for two classification problems, classification results can interpolate that whether DMU is effective, can be estimated by SVR regression models Count the numerical value of efficiency score.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711422113.0A CN108009750A (en) | 2017-12-25 | 2017-12-25 | A kind of power grid O&M efficiency rating method based on DEA and SVM |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711422113.0A CN108009750A (en) | 2017-12-25 | 2017-12-25 | A kind of power grid O&M efficiency rating method based on DEA and SVM |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108009750A true CN108009750A (en) | 2018-05-08 |
Family
ID=62061021
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711422113.0A Pending CN108009750A (en) | 2017-12-25 | 2017-12-25 | A kind of power grid O&M efficiency rating method based on DEA and SVM |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108009750A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109670640A (en) * | 2018-12-17 | 2019-04-23 | 北京化工大学 | The efficiency evaluation method of DEA ethylene unit based on AP algorithm |
CN111274527A (en) * | 2020-01-20 | 2020-06-12 | 北京赛博贝斯数据科技有限责任公司 | Department decision evaluation method based on expectation and non-expectation |
CN114638499A (en) * | 2022-03-14 | 2022-06-17 | 西安工程大学 | Public cultural efficiency assessment method based on hesitation fuzzy four-stage DEA |
CN115860510A (en) * | 2022-10-31 | 2023-03-28 | 浙江淏瀚信息科技有限公司 | Production efficiency analysis and evaluation method based on big data |
-
2017
- 2017-12-25 CN CN201711422113.0A patent/CN108009750A/en active Pending
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109670640A (en) * | 2018-12-17 | 2019-04-23 | 北京化工大学 | The efficiency evaluation method of DEA ethylene unit based on AP algorithm |
CN111274527A (en) * | 2020-01-20 | 2020-06-12 | 北京赛博贝斯数据科技有限责任公司 | Department decision evaluation method based on expectation and non-expectation |
CN114638499A (en) * | 2022-03-14 | 2022-06-17 | 西安工程大学 | Public cultural efficiency assessment method based on hesitation fuzzy four-stage DEA |
CN115860510A (en) * | 2022-10-31 | 2023-03-28 | 浙江淏瀚信息科技有限公司 | Production efficiency analysis and evaluation method based on big data |
CN115860510B (en) * | 2022-10-31 | 2023-08-15 | 浙江淏瀚信息科技有限公司 | Production efficiency analysis and evaluation method based on big data |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Wu et al. | An intuitionistic fuzzy multi-criteria framework for large-scale rooftop PV project portfolio selection: Case study in Zhejiang, China | |
CN108009750A (en) | A kind of power grid O&M efficiency rating method based on DEA and SVM | |
CN104573879A (en) | Photovoltaic power station output predicting method based on optimal similar day set | |
Kaboli et al. | An expression-driven approach for long-term electric power consumption forecasting | |
CN110135643A (en) | Consider the Short-term Load Forecast method of steel forward price and Spot Price factor | |
CN109066819A (en) | A kind of idle work optimization method of the power distribution network based on case reasoning | |
CN111562541A (en) | Software platform for realizing electric energy meter detection data management by applying CART algorithm | |
Ni et al. | A review of line loss analysis of the low-voltage distribution system | |
Farhudi et al. | Multi-objective modelling of day ahead scheduling of MMG-based distribution networks accompanied by DEA considering economics, emissions and power quality | |
CN114880754B (en) | BIM-based building energy consumption management method and system | |
Cai et al. | Short-term forecasting of user power load in China based on XGBoost | |
CN114862363A (en) | Evaluation analysis method based on substation project cost data analysis | |
Zhang et al. | Neutrosophic Adaptive Clustering Optimization in Genetic Algorithm and Its Application in Cubic Assignment Problem. | |
Zhu et al. | On-site energy consumption technologies and prosumer marketing for distributed poverty alleviation photovoltaic linked to agricultural loads in china | |
Wang et al. | Short-term wind power forecasting based on BOMLS K-means similar hours Clustering method | |
Jankovic et al. | Classification of Energy Consumption in the Balkans using Ensemble Learning Methods | |
Chen et al. | Intelligent identification of potential customers for electricity substitution | |
Sala et al. | Smart multi-model approach based on adaptive Neuro-Fuzzy Inference Systems and Genetic Algorithms | |
CN112016631A (en) | Improvement scheme related to low-voltage treatment | |
Cho et al. | Application of Parallel ANN-PSO to Hourly Solar PV Estimation | |
Xie et al. | Comprehensive evaluation of urban economic development in Yangtze river Delta based on Cluster-Principal Component analysis | |
Zhang et al. | Topology identification method of low voltage aea based on topological data analysis | |
Sun et al. | Identification of Generating Units That Abuse Market Power in Electricity Spot Market Based on AdaBoost‐DT Algorithm | |
Liang et al. | Sustainability evaluation of modern photovoltaic agriculture based on interval Type-2 fuzzy AHP-TOPSIS and least squares support vector machine optimized by fireworks algorithm | |
Luo et al. | Prediction of Photovoltaic Power Generation Based on PSO-RNN and SVR Model |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180508 |