CN114638555B - Power consumption behavior detection method and system based on multilayer regularization extreme learning machine - Google Patents
Power consumption behavior detection method and system based on multilayer regularization extreme learning machine Download PDFInfo
- Publication number
- CN114638555B CN114638555B CN202210536401.3A CN202210536401A CN114638555B CN 114638555 B CN114638555 B CN 114638555B CN 202210536401 A CN202210536401 A CN 202210536401A CN 114638555 B CN114638555 B CN 114638555B
- Authority
- CN
- China
- Prior art keywords
- learning machine
- extreme learning
- multilayer
- factor
- state
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 108
- 230000006399 behavior Effects 0.000 title claims abstract description 42
- 238000000034 method Methods 0.000 claims abstract description 35
- 230000005611 electricity Effects 0.000 claims abstract description 30
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 29
- 238000012549 training Methods 0.000 claims abstract description 27
- 230000002159 abnormal effect Effects 0.000 claims abstract description 26
- 230000007704 transition Effects 0.000 claims abstract description 20
- 238000009826 distribution Methods 0.000 claims abstract description 17
- 238000005457 optimization Methods 0.000 claims abstract description 16
- 230000008569 process Effects 0.000 claims abstract description 13
- 230000009466 transformation Effects 0.000 claims description 74
- 238000013519 translation Methods 0.000 claims description 59
- 230000006870 function Effects 0.000 claims description 38
- 239000011159 matrix material Substances 0.000 claims description 27
- 210000002569 neuron Anatomy 0.000 claims description 20
- 230000003044 adaptive effect Effects 0.000 claims description 15
- 239000013598 vector Substances 0.000 claims description 15
- 238000003860 storage Methods 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 abstract 2
- 150000001875 compounds Chemical class 0.000 description 11
- 238000012360 testing method Methods 0.000 description 6
- 238000012545 processing Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013501 data transformation Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 210000004205 output neuron Anatomy 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000009827 uniform distribution Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Data Mining & Analysis (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- General Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Educational Administration (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Operations Research (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Game Theory and Decision Science (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses a power utilization behavior detection method and a system based on a multilayer regularization extreme learning machine, wherein the method comprises the following steps: acquiring original power consumption data of power users of a power distribution network system, and training a preset multilayer regularization extreme learning machine based on the original power consumption data to obtain a multilayer extreme learning machine detection model; carrying out network parameter optimization on the multilayer extreme learning machine detection model based on a novel self-adaptive state transition algorithm to output an optimal network structure parameter; and inputting the online detection data into a multilayer extreme learning machine detection model established based on the optimal network structure parameters, so as to output users with abnormal electricity utilization. The network structure parameters of the detection model of the multilayer regularization extreme learning machine are adjusted and optimized according to a novel self-adaptive state transition algorithm, and the conversion factors of the state transition algorithm are adjusted to enable the conversion factors to have the nonlinear self-adaptive characteristic, so that the network structure parameter optimizing process of the multilayer regularization extreme learning machine is simple and easy to implement.
Description
Technical Field
The invention belongs to the technical field of abnormal electricity utilization analysis, and particularly relates to an electricity utilization behavior detection method and system based on a multilayer regularization extreme learning machine.
Background
With the rapid development of economy, the power consumption demand of users is continuously increased, if the power consumption behavior of the users is abnormal, the non-technical loss of a power grid is increased, and the operation cost of a power company is increased. The traditional method for detecting the abnormal electricity utilization behaviors of the users is that field personnel regularly patrol circuits, regularly check electric meters, report users and the like, the means has high dependence on people, a large amount of labor cost needs to be invested, and meanwhile, the electricity utilization behaviors are long in detection time and low in efficiency.
The research on abnormal electricity utilization behavior detection is mainly divided into two methods based on states and artificial intelligence. The state-based analysis method is used for detecting abnormality by comparing changes of a large amount of data such as power, voltage, current and the like of the power distribution network in real time; the abnormal electricity consumption behavior detection model based on artificial intelligence firstly extracts indexes capable of reflecting abnormal electricity consumption behaviors through data analysis, and then completes construction of the abnormal electricity consumption behavior detection model by training a mapping relation between the indexes and an electricity consumption behavior detection result through an artificial intelligence method. However, the time of the current model in the parameter optimization and training process is long, and the current model cannot be suitable for detecting abnormal electricity consumption of users in different scenes.
Disclosure of Invention
The invention provides a power utilization behavior detection method and system based on a multilayer regularization extreme learning machine, which are used for solving at least one of the technical problems.
In a first aspect, the invention provides a power consumption behavior detection method based on a multilayer regularization extreme learning machine, which comprises the following steps: the method comprises the steps of obtaining original power consumption data of power users of a power distribution network system, training a preset multilayer regularization extreme learning machine based on the original power consumption data, and enabling a multilayer extreme learning machine detection model to be achieved, wherein a target function of the multilayer regularization extreme learning machine is as follows:
in the formula (I), wherein,
to adjust the parameters of empirical risk and structural risk,
weighting coefficients for the L2 regularization and the L1 regularization,
in order to minimize the objective function,
in order to output a set of data samples,
in order to have a hidden layer output matrix,
in order to have the hidden layer output weights,
the normalized output weight vector norm for L2,
a vector norm normalized to L1; carrying out network parameter optimization on the multilayer extreme learning machine detection model based on a novel self-adaptive state transition algorithm to enable the optimal network structure parameters to be output, wherein the process of outputting the optimal network structure parameters comprises the following steps: updating transformation factors based on a nonlinear adaptive adjustment strategy, wherein the transformation factors comprise rotation factors, translation factors, scaling factors and axial factors, and the expression of the nonlinear adaptive adjustment strategy is as follows:
in the formula (I), wherein,
、
、
、
respectively as the maximum value of a rotation factor, the maximum value of a translation factor, the maximum value of a stretching factor and the maximum value of an axial factor,
for the current number of iterations,
、
、
、
the maximum iteration times of the rotation factor satisfying the termination condition, the maximum iteration times of the translation factor satisfying the termination condition, the maximum iteration times of the expansion factor satisfying the termination condition and the maximum iteration times of the axial factor satisfying the termination condition are respectively,
、
、
、
respectively, a rotation factor, a translation factor, a scale factor and an axial factor; selecting a group from the current population, the fitness function F of which reaches a minimum value
Value, is recorded as
The corresponding fitness is
Will be
The number of the initial population is copied as the number of the individuals
Group of (1), as
Performing telescopic transformation according to a telescopic transformation operator, a rotary transformation operator or an axial transformation operator to obtain a new population, wherein the optimal individuals in the population after the telescopic transformation are
The corresponding fitness is
If, if
Then, the individuals are subjected to translation transformation operators
Performing translation transformation and updating the translation transformed
And
otherwise, no translation transformation is performed, wherein,
the number of neurons in the first layer of the multilayer extreme learning machine detection model,
the number of neurons in the second layer of the multilayer extreme learning machine detection model,
detecting the number of neurons in the third layer of the model for the multilayer extreme learning machine; judging whether the fitness function meets the minimum requirement or reaches the maximum iteration times, and if the fitness function meets the minimum requirement or reaches the maximum iteration times, outputting the optimal individuals in the population as optimal network structure parameters; and inputting the online detection data into a multilayer extreme learning machine detection model established based on the optimal network structure parameters, so as to output users with abnormal electricity utilization.
In a second aspect, the present invention provides a power consumption behavior detection system based on a multi-layer regularization extreme learning machine, including: the training module is configured to acquire original power consumption data of power users of the power distribution network system, and train a preset multilayer regularization extreme learning machine based on the original power consumption data to obtain a multilayer extreme learning machine detection model, wherein a target function of the multilayer regularization extreme learning machine is as follows:
in the formula (I), wherein,
to adjust the parameters of empirical risk and structural risk,
weighting coefficients for the L2 regularization and the L1 regularization,
in order to minimize the objective function,
in order to output a set of data samples,
in order to have a hidden layer output matrix,
in order to have the hidden layer output weights,
the normalized output weight vector norm for L2,
a vector norm normalized to L1; the optimizing module is configured to perform network parameter optimizing on the multilayer extreme learning machine detection model based on a novel self-adaptive state transition algorithm so as to output optimal network structure parameters, wherein the process of outputting the optimal network structure parameters comprises the following steps: updating transformation factors based on a nonlinear adaptive adjustment strategy, wherein the transformation factors comprise rotation factors, translation factors, expansion factors and axial factors, and the expression of the nonlinear adaptive adjustment strategy is as follows:
in the formula (I), wherein,
、
、
、
respectively as the maximum value of a rotation factor, the maximum value of a translation factor, the maximum value of a stretching factor and the maximum value of an axial factor,
the number of times of the current iteration is,
、
、
、
the maximum iteration times of the rotation factor satisfying the termination condition, the maximum iteration times of the translation factor satisfying the termination condition, the maximum iteration times of the expansion factor satisfying the termination condition and the maximum iteration times of the axial factor satisfying the termination condition are respectively,
、
、
、
respectively, a rotation factor, a translation factor, a scale factor and an axial factor; selecting a group of fitness functions F reaching a minimum value from the current population
Value, is recorded as
The corresponding fitness is
Will be
The number of the initial population is copied as the number of the individuals
Group of (1), as
Performing telescopic transformation according to a telescopic transformation operator, a rotation transformation operator or an axial transformation operator to obtain a new population, wherein the optimal individuals in the population after the telescopic transformation are
The corresponding fitness is
If, if
Then, the individuals are subjected to translation transformation operators
Performing translation transformation and updating the translation transformed
And
otherwise, no translation transformation is performed, wherein,
the number of neurons in the first layer of the multilayer extreme learning machine detection model,
the number of neurons in the second layer of the multilayer extreme learning machine detection model,
detecting the number of neurons in the third layer of the model for the multilayer extreme learning machine; judging whether the fitness function meets the minimum requirement or reaches the maximum superpositionGenerating times, and outputting the optimal individual in the population as the optimal network structure parameter if the fitness function meets the minimum requirement or reaches the maximum iteration times; and the output module is configured to input the online detection data into a multilayer extreme learning machine detection model established based on the optimal network structure parameters, so that users with abnormal electricity utilization can be output.
In a third aspect, an electronic device is provided, comprising: the power usage behavior detection method comprises at least one processor and a memory which is in communication connection with the at least one processor, wherein the memory stores instructions which can be executed by the at least one processor, and the instructions are executed by the at least one processor so as to enable the at least one processor to execute the steps of the power usage behavior detection method based on the multi-layer regularized limit learning machine according to any embodiment of the invention.
In a fourth aspect, the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the program instructions, when executed by a processor, cause the processor to execute the steps of the power usage behavior detection method based on a multi-layered regularized limit learning machine according to any one of the embodiments of the present invention.
According to the power utilization behavior detection method and system based on the multilayer regularization extreme learning machine, accurate detection of power utilization abnormity of the user can be achieved, time of the model in parameter optimization and training processes is greatly reduced, and in addition, the power utilization abnormity detection method and system based on the multilayer regularization extreme learning machine also have good adaptability and high efficiency for power utilization abnormity detection of the user in different scenes.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.
Fig. 1 is a flowchart of a power consumption behavior detection method based on a multi-layer regularization extreme learning machine according to an embodiment of the present invention;
fig. 2 is a flowchart of a power consumption behavior detection method based on a multi-layer regularization extreme learning machine according to an embodiment of the present invention;
fig. 3 is a block diagram of a structure of a power consumption behavior detection system based on a multi-layer regularization extreme learning machine 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
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, a flow chart of a power consumption behavior detection method based on a multi-layer regularization extreme learning machine according to the present application is shown.
As shown in fig. 1, step S101, original power consumption data of a power consumer in a power distribution network system is obtained, and a preset multilayer regularization extreme learning machine is trained based on the original power consumption data, so that a multilayer extreme learning machine detection model is obtained.
In the embodiment, the multilayer regularization extreme learning machine introduces the weighted sum of L1 regularization and L2 regularization to reduce the model structural risk, fully utilizes the advantages of L1 regularization and L2 regularization, can generate a sparse weight matrix by L1 regularization, extracts different attention degrees for hidden layer features of the multilayer extreme learning machine, plays a role in feature selection, and is beneficial to model learning to obtain better feature representation; the L2 regularization can effectively limit the number of weight parameters of the multilayer extreme learning machine model, thereby effectively reducing the complexity of the model and improving the stability of the model. By combining the advantages of the two, the optimization of the network structure of the multilayer extreme learning machine model can be realized, the functions of simplifying the model and preventing over-fitting can be achieved, and the learning ability and generalization performance are excellent.
In conclusion, the method of the embodiment adopts the multilayer extreme learning machine detection model, can fully learn the hidden characteristics of the power consumption behaviors in the data, and simultaneously plays a role in feature screening, thereby simplifying the detection model, improving the detection accuracy and saving the time cost.
And S102, optimizing network parameters of the multilayer extreme learning machine detection model based on a novel self-adaptive state transition algorithm, so that optimal network structure parameters are output.
In the embodiment, the nonlinear adaptive adjustment strategy is used for carrying out nonlinear adaptive processing on the conversion factor, so that the conversion factor can be rapidly reduced in the initial stage of iteration, and the change of the conversion factor tends to be stable in the later stage of iteration, and therefore, the optimization of the hyperparameter of the multilayer extreme learning machine model can be rapidly, accurately and efficiently realized, and the multilayer extreme learning machine model has excellent user power utilization abnormity detection capability.
And step S103, inputting online detection data into a multilayer extreme learning machine detection model established based on the optimal network structure parameters, and outputting users with abnormal electricity utilization.
In conclusion, the method can realize accurate detection of the power utilization abnormity of the user, greatly reduces the time of the model in parameter optimization and training processes, and has good adaptability and high efficiency for detection of the power utilization abnormity of the user in different scenes.
Referring to fig. 2, a flowchart of a power consumption behavior detection method based on a multi-layer regularized extreme learning machine according to an embodiment of the present application is shown.
As shown in fig. 2, the method for detecting power consumption behavior based on the multi-layer regularization extreme learning machine specifically includes the following steps:
step 1: data acquisition
The method comprises the steps of obtaining original power consumption data of power users of a power distribution network system from a power consumption acquisition system and an energy management system, wherein the original power consumption data comprises basic power consumption information data of the users, alarm information data of a terminal and electricity stealing information data of the users in the area.
And 2, step: data pre-processing
Data cleaning: data cleansing refers to the removal of redundant, irrelevant data from the original data to smooth out data noise. Non-resident users such as utilities and the like generally do not have abnormal electricity utilization behaviors, and electricity utilization data of the non-resident users can be deleted.
Missing value processing: data recorded by the power utilization acquisition system can be partially lost due to acquisition equipment faults, transmission packet loss and other reasons, and if lost samples are directly ignored, the data error of the daily loss rate is larger, so that the accuracy of detecting abnormal power utilization behaviors is reduced. In order to avoid the influence of the missing value, the missing value is processed by an interpolation method.
Data transformation: the data is normalized, that is, the data format is converted to be suitable for the detection technology provided by the invention. According to the data characteristics, data change can be carried out from two aspects of normalized processing and attribute construction. The normalization process converts data having different dimensions to the same dimension, and specifies the data to a smaller extent. The aim of normalization processing can be achieved by adopting minimum-maximum normalization, and the formula is as follows:
in the formula (I), the compound is shown in the specification,
in order to obtain the normalized sample data,
is the actual value of the sample data,
is the minimum value of the sample data,
is the maximum value of the sample data;
and step 3: multilayer regularization extreme learning machine-based multilayer extreme learning machine detection model
1) Model input
Dividing the preprocessed sample data set into a training set and a test set according to the proportion of 8:2, training the multilayer extreme learning machine based on the training set, and using the test set as input data of model performance evaluation.
2) Constructing a multi-layer regularization extreme learning machine
An Extreme Learning Machine (ELM) is a single hidden layer feedforward neural network, an ELM algorithm randomly generates a connection weight of an input and a hidden layer and a threshold value of the hidden layer neural network, and a global optimal solution of a solution problem can be obtained only by setting the number of neurons of the hidden layer without adjustment in a training process. Compared with the traditional feedforward neural network, the method has the advantages of high learning speed, good generalization performance and difficulty in falling into local optimal solution. For N training samples
The basic ELM algorithm is expressed as follows:
in the formula (I), the compound is shown in the specification,
to imply the number of layer neurons,
is the number of training samples that are to be trained,
is input into
To a corresponding second
The output of each of the hidden layer neurons,
is as follows
The connection weight vector between each hidden layer neuron and the output neuron,
a hidden layer activation function is represented that,
for the j-th input sample,
for the input weights connected to the jth input sample and the ith implicit node,
for the threshold of the ith hidden node, the ELM algorithm minimizes the output weight
The generalization capability of the neural network is ensured, and a least square solution is usually taken.
The matrix of equation (2) is represented as:
in the formula (I), the compound is shown in the specification,
in order to have a hidden layer output matrix,
in order to have the hidden layer output weights,
is a set of output data samples;
training the ELM is equivalent to solving
The expression of the minimum standard two multiplier solution of (1) is:
In deep learning, overfitting phenomena of a training model can occur due to too many network parameters, so in order to obtain a better training model, cost functions of weighted regularization terms of L2 and L1 are introduced to solve output weights, and thus the following formula is obtained:
in the formula (I), the compound is shown in the specification,
to adjust the parameters of empirical risk and structural risk,
weighting coefficients for the L2 regularization and L1 regularization,
in order to minimize the objective function,
in order to output a set of data samples,
in order to have a hidden layer output matrix,
in order to have the hidden layer output weights,
the normalized output weight vector norm for L2,
a vector norm normalized to L1;
the derivation is carried out on the objective function (5) to obtain the output weight
The following formula shows:
the main difference between ELM-AE (extreme learning machine-auto encoder) and conventional ELM is that ELM is a supervised learning algorithm, the output of which is a corresponding label. And the ELM-AE is an unsupervised learning algorithm, the output of the ELM-AE is the mapping of the input of the ELM-AE, the hidden layer output of the ELM-AE can be represented by the formula (7) to the formula (8).
In the formula (I), the compound is shown in the specification,
respectively weight vectors and bias vectors between the input layer and the hidden layer,
in order to be transposed, the device is provided with a plurality of groups of parallel connection terminals,
is a matrix of the unit diagonal,
is an input sample;
the ELM-AE hidden layer parameters need to be orthogonalized after random generation. The input data is mapped to a random subspace. Compared with ELM random initialization input weight and hidden layer bias, orthogonalization can better capture various edge features of input data, so that a model can effectively learn the nonlinear structure of the data. Output weights
The calculation can be performed by equation (6).
When the ML-ELM (multi layer extreme learning machine) utilizes ELM-AE training, the input of the (i + 1) th hidden layer is the output on the (i) th hidden layer, which is expressed by the formula (9).
Wherein, the first and the second end of the pipe are connected with each other,
is as follows
An output of the hidden layer when
When the value is 1, the input is the whole model,
for ELM-AE pair
A hidden layer and the second
And (4) a weight matrix during hidden layer training.
And 4, step 4: carrying out parameter optimization on the detection model of the multilayer extreme learning machine by using a novel self-adaptive state transition algorithm to determine the optimal detection model
Firstly, a state transformation operator, a neighborhood and sampling are used for generating a candidate solution, then the current optimal solution is replaced by selection and updating, and finally, an alternate rotation strategy is adopted for realizing the calling of different state transformation operators. The state transformation operator mainly has four transformation modes, namely a rotation transformation operator, a translation transformation operator, a telescopic transformation operator and an axial transformation operator.
1) And (3) a rotation transformation operator:
in the formula (I), the compound is shown in the specification,
in order to be a factor of rotation,
the state at the moment of time when the parameter-exceeding variable k, i.e. the current state,
obeying an element to [ -1,1 [ ]]A random matrix of uniform distribution of the random number,
the state at the moment of the over-parameter variable k +1,
is a random matrix
Dimension (d) of (a). The rotation transformation operator can be generated in
Is a candidate solution within the radius hypersphere.
2) Translation transformation operator:
in the formula (I), the compound is shown in the specification,
the state at the moment of the over-parameter variable k +1,
the state at the moment of time when the parameter-exceeding variable k, i.e. the current state,
the state at the moment of the over-parameter variable k-1,
is a 2 norm of the difference between the k time and the k-1 time of the hyper-parametric variable,
obey an element by 0,1]The random numbers are distributed evenly and the random numbers are distributed evenly,
is a translation factor. The translation transform operator can be implemented in
To
In the linear range to a maximumHas a length of
A function of performing a search.
3) Scaling transform operator
In the formula (I), the compound is shown in the specification,
the state at the moment of time when the parameter-exceeding variable k, i.e. the current state,
the state at the moment of the over-parameter variable k +1,
in order to be a translation factor, the translation factor,
the elements are subjected to a random diagonal matrix of gaussian distribution. The scaling transform operator will
Each element of (1) to
Scaling within a range.
4) Axial transformation operator
In the formula (I), the compound is shown in the specification,
the state at the moment of the over-parameter variable k +1,
the state at the moment of time when the parameter-exceeding variable k, i.e. the current state,
is a factor in the axial direction and is,
a sparse random diagonal matrix obeying a gaussian distribution for non-zero elements. The function of the axial transform operator is to enhance single-dimensional searches.
The conversion factor is adjusted, so that a larger value is taken at the early stage to obtain a larger reduction rate, and a smaller value is taken at the later stage to increase the success rate of algorithm optimization. The novel state transition algorithm with the adaptive conversion factor is added, so that the optimization process is accelerated, and the optimization algorithm is prevented from falling into a local optimal solution. The nonlinear adaptive adjustment strategy expression of the conversion factor is as follows:
in the formula (I), the compound is shown in the specification,
、
、
、
respectively as the maximum value of a rotation factor, the maximum value of a translation factor, the maximum value of a stretching factor and the maximum value of an axial factor,
for the current number of iterations,
、
、
、
the maximum iteration times of the rotation factor satisfying the termination condition, the maximum iteration times of the translation factor satisfying the termination condition, the maximum iteration times of the expansion factor satisfying the termination condition and the maximum iteration times of the axial factor satisfying the termination condition are respectively,
、
、
、
respectively, a rotation factor, a translation factor, a scaling factor, and an axial factor.
In the multilayer regularization extreme learning machine, the number of hidden layers is set to 3, because the number of neurons in each hidden layer
A regularization factor and
and
regularized weight coefficients
Influence on the abnormal use of the multi-layer kernel extreme learning machine to the usersThe invention adopts a novel self-adaptive state transition algorithm to carry out parameter optimization on a detection model of a multilayer extreme learning machine, and finds out the optimal hyperparameter
And the detection capability of the multilayer regularization extreme learning machine model on the abnormal electricity utilization behavior of the user is optimal.
The problem of optimizing the network parameters of the multilayer regularization extreme learning machine by adopting a novel self-adaptive state transition algorithm can be represented by the following formula:
in the formula (I), the compound is shown in the specification,
for the current state in the variable space,
in order to be a state transition matrix,
in order to train the total number of samples,
in order to correctly detect the number of samples,
the detection error rate is a fitness function, namely the detection error rate of the abnormal electricity utilization behavior of the user.
The process of optimizing the network structure parameters of the multilayer regularization extreme learning machine by adopting a novel self-adaptive state transition algorithm is as follows:
step A: the number of the initialization population is
STA Algorithm initialization parameters, rotation factorsSeed of Japanese apricot
Translation factor
Factor of expansion
Axial factor of
The maximum iteration number is 100, and random uniform initialization is carried out in a feasible domain
5 variables, generate the initial population, generate
Set initial feasible solutions.
And B: the transform factor is updated by equation (14).
And C: selecting a fitness function from a current population
Group reaching minimum
Value, is recorded as
The corresponding fitness is
Will be
Is replicated into a unit number of
Group of (1), as
Performing scaling transformation according to a formula (12) to obtain a new population, wherein the optimal individuals in the population after scaling transformation are
The corresponding fitness is
If, if
Then pressing formula (11) to the individuals
Performing translation transformation and updating the translation transformed
And
otherwise, no translation transformation is performed.
Step D: will be provided with
Is replicated into a unit number of
Then carrying out rotation transformation according to the formula (10) to obtain a new population, and selecting the optimal individual in the new population
Calculating its corresponding fitness
If, if
The translation conversion is performed according to equation (11), and the translation-converted data is updated
And
otherwise, no translation transformation is performed.
Step E: a population selection and updating process similar to that of the step C is adopted, except that a new population is generated through the axial transformation of the formula (13), and then the translational transformation is updated through the same method as that of the step C
And
。
step F: and (4) judging whether the fitness function meets the minimum requirement or whether the maximum iteration number is reached, and otherwise, repeating the steps (B) to (E). And when the termination condition is reached, outputting the optimal individual in the population as a network structure parameter of the multilayer regularization extreme learning machine.
And 5: model evaluation
After network structure parameters of the multilayer regularization extreme learning machine are searched by adopting a novel self-adaptive state transition algorithm, a multilayer extreme learning machine detection model is established according to the optimal network structure parameters, then the detection model is trained again through a training set, and finally the detection performance of the model is verified by utilizing a test set. The accuracy test is carried out on the multi-layer extreme learning machine detection model with the optimal performance on the divided test set, and the result shows that the detection model optimized by the novel self-adaptive state transition algorithm provided by the invention has remarkable improvement on the comprehensive evaluation indexes such as precision, f1 score (f 1 score) and AUC (area Under cut). The effectiveness of the multi-layer extreme learning machine detection model optimized based on the novel self-adaptive state transition algorithm in the abnormal power utilization detection of the user is shown on the performance and the time efficiency of the model on the test set.
And preprocessing the data acquired on line, inputting the data into the trained detection model, acquiring a model detection result, and judging whether abnormal power utilization occurs.
Referring to fig. 3, a block diagram of a power consumption behavior detection system based on a multi-layer regularized extreme learning machine according to the present application is shown.
As shown in fig. 3, the electricity consumption behavior detection system 200 includes a training module 210, an optimizing module 220, and an output module 230.
The training module 210 is configured to acquire original power consumption data of power users of a power distribution network system, and train a preset multilayer regularization extreme learning machine based on the original power consumption data, so as to obtain a multilayer extreme learning machine detection model, where a target function of the multilayer regularization extreme learning machine is:
in the formula (I), wherein,
to adjust the parameters of empirical risk and structural risk,
weighting coefficients for the L2 regularization and the L1 regularization,
in order to minimize the objective function,
in order to output a set of data samples,
in order to imply the layer output matrix,
in order to have the hidden layer output weights,
the normalized output weight vector norm for L2,
a vector norm normalized to L1; an optimizing module 220 configured to perform network parameter optimizing on the multi-layer extreme learning machine detection model based on a novel adaptive state transition algorithm, so as to output an optimal network structure parameter, wherein a process of outputting the optimal network structure parameter includes: updating transformation factors based on a nonlinear adaptive adjustment strategy, wherein the transformation factors comprise rotation factors, translation factors, scaling factors and axial factors, and the expression of the nonlinear adaptive adjustment strategy is as follows:
in the formula (I), wherein,
、
、
、
respectively as the maximum value of a rotation factor, the maximum value of a translation factor, the maximum value of a stretching factor and the maximum value of an axial factor,
the number of times of the current iteration is,
、
、
、
the maximum iteration times of the rotation factor satisfying the termination condition, the maximum iteration times of the translation factor satisfying the termination condition, the maximum iteration times of the expansion factor satisfying the termination condition and the maximum iteration times of the axial factor satisfying the termination condition are respectively,
、
、
、
respectively, a rotation factor, a translation factor, a scale factor and an axial factor; selecting a group from the current population, the fitness function F of which reaches a minimum value
Value, is recorded as
The corresponding fitness is
Will be
The number of the initial population is copied as the number of the individuals
Group of (1), as
According to scaling, rotation or axial transformation operatorsPerforming scaling transformation on the seeds to obtain a new population, wherein the optimal individuals in the population after scaling transformation are
The corresponding fitness is
If, if
Then, the individuals are subjected to translation transformation operators
Performing translation transformation and updating the translation transformed
And
otherwise, no translation transformation is performed, wherein,
the number of neurons in the first layer of the detection model of the multilayer extreme learning machine,
the number of neurons in the second layer of the multilayer extreme learning machine detection model,
detecting the number of neurons in the third layer of the model for the multilayer extreme learning machine; judging whether the fitness function meets the minimum requirement or reaches the maximum iteration times, and if the fitness function meets the minimum requirement or reaches the maximum iteration times, outputting the optimal individuals in the population as optimal network structure parameters; and the output module 230 is configured to input the online detection data into a multi-layer extreme learning machine detection model established based on the optimal network structure parameters, so that users with abnormal electricity utilization are output.
It should be understood that the modules depicted in fig. 3 correspond to various steps in the method described with reference to fig. 1. Thus, the operations and features described above for the method and the corresponding technical effects are also applicable to the modules in fig. 3, and are not described again here.
In still other embodiments, the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the program instructions, when executed by a processor, cause the processor to execute the power usage behavior detection method based on a multi-layer regularized limit learning machine in any of the above method embodiments;
as one embodiment, the computer-readable storage medium of the present invention stores computer-executable instructions configured to:
acquiring original power consumption data of power users of a power distribution network system, and training a preset multilayer regularization extreme learning machine based on the original power consumption data to obtain a multilayer extreme learning machine detection model;
carrying out network parameter optimization on the multilayer extreme learning machine detection model based on a novel self-adaptive state transition algorithm to output an optimal network structure parameter;
and inputting the online detection data into a multilayer extreme learning machine detection model established based on the optimal network structure parameters, so as to output users with abnormal electricity utilization.
The computer-readable storage medium may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created from use of the electricity usage behavior detection system based on the multilayer regularization limit learning machine, and the like. Further, the computer-readable storage medium may include high speed random access memory, and may also include memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some embodiments, the computer readable storage medium optionally includes memory remotely located from the processor, and the remote memory may be connected to the multi-layer regularized limit learning machine based power usage behavior detection system via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
Fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 4, the electronic device includes: a processor 310 and a memory 320. The electronic device may further include: an input device 330 and an output device 340. The processor 310, the memory 320, the input device 330, and the output device 340 may be connected by a bus or other means, such as the bus connection in fig. 4. The memory 320 is the computer-readable storage medium described above. The processor 310 executes various functional applications and data processing of the server by running the nonvolatile software programs, instructions and modules stored in the memory 320, namely, implementing the electricity usage behavior detection method based on the multi-layer regularization limit learning machine of the above method embodiment. The input device 330 may receive input numeric or character information and generate key signal inputs related to user settings and function control of the multi-layer regularized extreme learning machine based electricity usage behavior detection system. The output device 340 may include a display device such as a display screen.
The electronic device can execute the method provided by the embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in this embodiment, reference may be made to the method provided by the embodiment of the present invention.
As an embodiment, the electronic device is applied to a power consumption behavior detection system based on a multi-layer regularization extreme learning machine, and is used for a client, and the electronic device includes: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to cause the at least one processor to:
acquiring original power consumption data of power users of a power distribution network system, and training a preset multilayer regularization extreme learning machine based on the original power consumption data to obtain a multilayer extreme learning machine detection model;
carrying out network parameter optimization on the multilayer extreme learning machine detection model based on a novel self-adaptive state transition algorithm to output optimal network structure parameters;
and inputting the online detection data into a multilayer extreme learning machine detection model established based on the optimal network structure parameters, so as to output users with abnormal electricity utilization.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods of the various embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present 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 solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (4)
1. A power utilization behavior detection method based on a multilayer regularization extreme learning machine is characterized by comprising the following steps:
the method comprises the steps of obtaining original power consumption data of power users of a power distribution network system, training a preset multilayer regularization extreme learning machine based on the original power consumption data, and enabling a multilayer extreme learning machine detection model to be achieved, wherein a target function of the multilayer regularization extreme learning machine is as follows:
in the formula, C is a parameter for adjusting the empirical risk and the structural risk, α is a weighting coefficient for L2 regularization and L1 regularization, min L is a minimization target function, Y is an output data sample set, H is a hidden layer output matrix, β is a hidden layer output weight, | | β | | Y 2 Is the L2 normalized output weight vector norm, | | β | | luminance 1 A vector norm normalized to L1;
carrying out network parameter optimization on the multilayer extreme learning machine detection model based on a novel self-adaptive state transfer algorithm to enable the output of an optimal network structure parameter, wherein an expression for carrying out network parameter optimization on the multilayer extreme learning machine detection model based on the novel self-adaptive state transfer algorithm is as follows:
in the formula, v k Being the current state in the variable space, A k Is a state transition matrix, N total For the total number of training samples, N actual F (v) is the number of correctly detected samples k+1 ) The user abnormal electricity consumption behavior detection error rate is a fitness function;
the process of outputting the optimal network structure parameters comprises the following steps:
updating transformation factors based on a nonlinear adaptive adjustment strategy, wherein the transformation factors comprise rotation factors, translation factors, scaling factors and axial factors, and the expression of the nonlinear adaptive adjustment strategy is as follows:
in the formula, S a.max 、S b.max 、S c.max 、S d.max Respectively the maximum value of the rotation factor, the maximum value of the translation factor, the maximum value of the expansion factor and the maximum value of the axial factorT is the current iteration number, T a . max 、T b.max 、T c.max 、T d.max Respectively, the maximum iteration times of the rotation factor satisfying the termination condition, the maximum iteration times of the translation factor satisfying the termination condition, the maximum iteration times of the expansion factor satisfying the termination condition and the maximum iteration times of the axial factor satisfying the termination condition, wherein a, b, c and d are respectively the rotation factor, the translation factor, the expansion factor and the axial factor;
selecting a group l with fitness function F reaching minimum value from current population 1 ，l 2 ，l 3 Value of C, alpha, noted as v best With a corresponding fitness of F best V is to be best The number N of the initial population is copied as the number of individuals SE Group of (1), denoted as v k Performing telescopic transformation according to a telescopic transformation operator, a rotary transformation operator or an axial transformation operator to obtain a new population, wherein the optimal individual in the population after the telescopic transformation is v newbest The corresponding fitness is F newbest If F is newbest ＜F best Then, the individual v is subjected to the translation transformation operator newbest Performing translation transformation, and updating v after translation transformation best And F best Otherwise, no translation transformation is performed, wherein 1 For the first layer neuron number, l, of the multilayer extreme learning machine detection model 2 Detecting the number of neurons in the second layer of the model for a multi-layer extreme learning machine 3 Calculating the number of neurons in the third layer of the detection model of the multilayer extreme learning machine, wherein the expression of the expansion transformation operator is as follows:
v k+1 ＝v k +cR e v k ，
in the formula, V k For the state at the moment k of the hyper-parameter variable, i.e. the current state, v k+1 The state at the moment of the over-parameter variable k +1, c is a translation factor, R e A random diagonal matrix with elements obeying a gaussian distribution;
calculating the expression of the rotation transformation operator as follows:
wherein a is a twiddle factor, v k For the state at the moment k of the hyper-parameter variable, i.e. the current state, R r Obeying an element to [ -1,1 [ ]]Uniformly distributed random matrix, v k+1 Is the state of the moment of the hyper-parameter variable k +1, and n is a random matrix R r Dimension, | | v k || 2 Is a 2 norm at the moment of exceeding a parameter variable k;
calculating an expression of the axial transformation operator as follows:
v k+1 ＝v k +dR a v k ，
in the formula, v k+1 For the state at the moment of the hyper-parametric variable k +1, v k Is the state at the moment of the over-parameter variable k, i.e. the current state, d is the axial factor, R a A sparse random diagonal matrix with non-zero elements obeying Gaussian distribution; calculating the expression of the translation transformation operator as follows:
in the formula, v k+1 Is the state at the moment of the hyperparametric variable k +1, v k For the state at the moment k of the hyper-parameter variable, i.e. the current state, v k-1 Is the state of the hyper-parameter variable k-1 at the moment, | | v k -v k-1 || 2 2 norm, R, of the difference between the k time and the k-1 time of the hyper-parametric variable t Obey [0,1 ] to an element]Uniformly distributed random numbers, b is a translation factor;
judging whether the fitness function meets the minimum requirement or reaches the maximum iteration times, and if the fitness function meets the minimum requirement or reaches the maximum iteration times, outputting the optimal individuals in the population as optimal network structure parameters;
and inputting the online detection data into a multilayer extreme learning machine detection model established based on the optimal network structure parameters, so as to output users with abnormal electricity utilization.
2. A power consumption behavior detection system based on a multilayer regularization extreme learning machine is characterized by comprising:
the training module is configured to acquire original power consumption data of power users of the power distribution network system, and train a preset multilayer regularization extreme learning machine based on the original power consumption data to obtain a multilayer extreme learning machine detection model, wherein a target function of the multilayer regularization extreme learning machine is as follows:
in the formula, C is a parameter for adjusting the empirical risk and the structural risk, α is a weighting coefficient for L2 regularization and L1 regularization, min L is a minimization target function, Y is an output data sample set, H is a hidden layer output matrix, β is a hidden layer output weight, | | β | | Y 2 Vector norm of output weight normalized for L2, | | β | | luminance 1 A vector norm normalized to L1;
the optimizing module is configured to perform network parameter optimizing on the multilayer extreme learning machine detection model based on a novel self-adaptive state transfer algorithm to output optimal network structure parameters, wherein an expression for performing network parameter optimizing on the multilayer extreme learning machine detection model based on the novel self-adaptive state transfer algorithm is as follows:
in the formula, v k Being the current state in the variable space, A k Is a state transition matrix, N total For the total number of training samples, N actual F (v) is the number of correctly detected samples k+1 ) The user abnormal electricity consumption behavior detection error rate is a fitness function;
the process of outputting the optimal network structure parameters comprises the following steps:
updating transformation factors based on a nonlinear adaptive adjustment strategy, wherein the transformation factors comprise rotation factors, translation factors, scaling factors and axial factors, and the expression of the nonlinear adaptive adjustment strategy is as follows:
in the formula, S a.max 、S b.max 、S c.max 、S d.max Respectively the maximum value of a rotation factor, the maximum value of a translation factor, the maximum value of a stretching factor and the maximum value of an axial factor, T is the current iteration number, T a.max 、T b.max 、T c.max 、T d.max Respectively, the maximum iteration times of the rotation factor satisfying the termination condition, the maximum iteration times of the translation factor satisfying the termination condition, the maximum iteration times of the expansion factor satisfying the termination condition and the maximum iteration times of the axial factor satisfying the termination condition, wherein a, b, c and d are respectively the rotation factor, the translation factor, the expansion factor and the axial factor;
selecting a group l with fitness function F reaching minimum value from current population 1 ，l 2 ，l 3 Value of C, alpha, noted as v best The corresponding fitness is F best V is to be best The number N of the initial population is copied as the number of individuals SE Group of (1), denoted as v k Performing telescopic transformation according to a telescopic transformation operator, a rotary transformation operator or an axial transformation operator to obtain a new population, wherein the optimal individual in the population after the telescopic transformation is v newbest The corresponding fitness is F rewbest If F is newbest ＜F best Then, the individual v is subjected to the translation transformation operator newbest Performing translation transformation and updating v after translation transformation best And F best Otherwise, no translation transformation is performed, wherein 1 For the first layer neuron number, l, of the multilayer extreme learning machine detection model 2 Detecting the number of neurons in the second layer of the model for the multi-layer extreme learning machine 3 Calculating the number of neurons in the third layer of the detection model of the multilayer extreme learning machine, wherein the expression of the expansion transformation operator is as follows:
v k+1 ＝v k +cR e v k ，
in the formula, v k For the state at the moment k of the hyper-parameter variable, i.e. the current state, v k+1 Is the state at the moment of the hyper-parametric variable k +1, c is the translation factor, R e A random diagonal matrix with elements obeying a gaussian distribution;
calculating the expression of the rotation transformation operator as follows:
wherein a is a twiddle factor, v k For the state at the moment k of the hyper-parameter variable, i.e. the current state, R r Obeying an element to [ -1,1 [ ]]Uniformly distributed random matrix, v k+1 Is the state of the moment of the hyper-parameter variable k +1, and n is a random matrix R r Dimension, | | v k || 2 2 norm at the moment of exceeding the parameter variable k;
calculating an expression of the axial transformation operator as follows:
v k+1 ＝v k +dR a v k ，
in the formula, v k+1 Is the state at the moment of the hyperparametric variable k +1, v k Is the state at the moment of the over-parameter variable k, i.e. the current state, d is the axial factor, R a A sparse random diagonal matrix with non-zero elements obeying Gaussian distribution; calculating the expression of the translation transformation operator as follows:
in the formula, v k+1 Is the state at the moment of the hyperparametric variable k +1, v k For the state at the moment of the hyper-parameter variable k, i.e. the current state, v k-1 Is the state of the hyper-parameter variable k-1 at the moment, | | v k -v k-1 || 2 2 norm, R, of the difference between the k time and the k-1 time of the hyper-parametric variable t Obey [0,1 ] to an element]Uniformly distributed random numbers, b is a translation factor;
judging whether the fitness function meets the minimum requirement or reaches the maximum iteration times, and if the fitness function meets the minimum requirement or reaches the maximum iteration times, outputting the optimal individuals in the population as optimal network structure parameters;
and the output module is configured to input the online detection data into a multilayer extreme learning machine detection model established based on the optimal network structure parameters, so that users with abnormal electricity utilization can be output.
3. An electronic device, comprising: at least one processor, and a memory communicatively coupled to the at least one processor, wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of claim 1.
4. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the method of claim 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210536401.3A CN114638555B (en) | 2022-05-18 | 2022-05-18 | Power consumption behavior detection method and system based on multilayer regularization extreme learning machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210536401.3A CN114638555B (en) | 2022-05-18 | 2022-05-18 | Power consumption behavior detection method and system based on multilayer regularization extreme learning machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114638555A CN114638555A (en) | 2022-06-17 |
| CN114638555B true CN114638555B (en) | 2022-09-16 |
Family
ID=81953066
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210536401.3A Active CN114638555B (en) | 2022-05-18 | 2022-05-18 | Power consumption behavior detection method and system based on multilayer regularization extreme learning machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114638555B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106650920A (en) * | 2017-02-19 | 2017-05-10 | 郑州大学 | Prediction model based on optimized extreme learning machine (ELM) |
Family Cites Families (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103942453A (en) * | 2014-05-07 | 2014-07-23 | 华北电力大学 | Intelligent electricity utilization anomaly detection method for non-technical loss |
| CN104537033B (en) * | 2014-12-23 | 2017-08-11 | 清华大学 | Interval type indices prediction method based on Bayesian network and extreme learning machine |
| CN105203869A (en) * | 2015-09-06 | 2015-12-30 | 国网山东省电力公司烟台供电公司 | Microgrid island detection method based on extreme learning machine |
| CN105160437A (en) * | 2015-09-25 | 2015-12-16 | 国网浙江省电力公司 | Load model prediction method based on extreme learning machine |
| CN105550744A (en) * | 2015-12-06 | 2016-05-04 | 北京工业大学 | Nerve network clustering method based on iteration |
| CN105976051A (en) * | 2016-04-29 | 2016-09-28 | 武汉大学 | Wavelet transformation and improved firefly-optimized extreme learning machine-based short-term load prediction method |
| US20180240018A1 (en) * | 2016-05-19 | 2018-08-23 | Jiangnan University | Improved extreme learning machine method based on artificial bee colony optimization |
| CN106650797B (en) * | 2016-12-07 | 2020-12-04 | 广东电网有限责任公司江门供电局 | Power distribution network electricity stealing suspicion user intelligent identification method based on integrated ELM |
| CN106790248B (en) * | 2017-01-23 | 2020-07-03 | 中南大学 | Network intrusion detection method based on double-self-adaptive regularization online extreme learning machine |
| US20200063665A1 (en) * | 2018-01-25 | 2020-02-27 | Dalian University Of Technology | Aero-engine full flight envelope model adaptive modification method based on deep learning algorithm |
| CN108664990B (en) * | 2018-03-29 | 2020-09-18 | 清华大学 | Electricity stealing detection method and device based on comprehensive entropy method and density clustering method |
| CN109146705B (en) * | 2018-07-02 | 2022-04-12 | 昆明理工大学 | Method for detecting electricity stealing by using electricity characteristic index dimension reduction and extreme learning machine algorithm |
| CN109359723A (en) * | 2018-11-20 | 2019-02-19 | 北京科技大学 | Based on the converter terminal manganese content prediction technique for improving regularization extreme learning machine |
| CN109657147B (en) * | 2018-12-21 | 2022-11-11 | 岭南师范学院 | Microblog abnormal user detection method based on firefly and weighted extreme learning machine |
| CN110084324B (en) * | 2019-05-10 | 2021-05-04 | 杭州电子科技大学 | Kalman filtering parameter self-adaptive updating method based on extreme learning machine |
| CN110147890A (en) * | 2019-05-13 | 2019-08-20 | 湖北工业大学 | A kind of method and system based on lion group's algorithm optimization extreme learning machine integrated study |
| CN110363384A (en) * | 2019-06-03 | 2019-10-22 | 杭州电子科技大学 | Exception electric detection method based on depth weighted neural network |
| CN111416797B (en) * | 2020-02-25 | 2022-07-05 | 江西理工大学 | Intrusion detection method for optimizing regularization extreme learning machine by improving longicorn herd algorithm |
| CN111541237B (en) * | 2020-04-02 | 2021-08-27 | 浙江大学 | Wind power nonparametric interval prediction method based on opportunity constraint extreme learning machine |
| CN111539492B (en) * | 2020-07-08 | 2020-11-20 | 武汉格蓝若智能技术有限公司 | Abnormal electricity utilization judgment system and method based on reinforcement learning |
| CN111860638B (en) * | 2020-07-17 | 2022-06-28 | 湖南大学 | Parallel intrusion detection method and system based on unbalanced data deep belief network |
| CN113408610B (en) * | 2021-06-18 | 2022-10-25 | 北京理工大学 | Image identification method based on adaptive matrix iteration extreme learning machine |
| CN113435595B (en) * | 2021-07-08 | 2024-02-06 | 南京理工大学 | Two-stage optimization method for network parameters of extreme learning machine based on natural evolution strategy |
| CN113627075B (en) * | 2021-07-19 | 2024-04-09 | 南京理工大学 | Projectile pneumatic coefficient identification method based on adaptive particle swarm optimization extreme learning |
| CN113392594B (en) * | 2021-08-13 | 2021-11-30 | 北京科技大学 | Mechanical property interval prediction method and device based on ABC extreme learning machine |
| CN113723517A (en) * | 2021-08-31 | 2021-11-30 | 北京理工大学 | Image classification method based on state transition extreme learning machine |
-
2022
- 2022-05-18 CN CN202210536401.3A patent/CN114638555B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106650920A (en) * | 2017-02-19 | 2017-05-10 | 郑州大学 | Prediction model based on optimized extreme learning machine (ELM) |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114638555A (en) | 2022-06-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tian et al. | An intrusion detection approach based on improved deep belief network | |
| CN115018021B (en) | Machine room abnormity detection method and device based on graph structure and abnormity attention mechanism | |
| Pan et al. | An approach for HVCB mechanical fault diagnosis based on a deep belief network and a transfer learning strategy | |
| CN112350876A (en) | Network flow prediction method based on graph neural network | |
| CN114970774B (en) | Intelligent transformer fault prediction method and device | |
| Tan et al. | Multi-node load forecasting based on multi-task learning with modal feature extraction | |
| CN114118150A (en) | Power distribution network single-phase earth fault line selection method and system | |
| CN116245033A (en) | Artificial intelligent driven power system analysis method and intelligent software platform | |
| CN116484740A (en) | Line parameter identification method based on space topology characteristics of excavated power grid | |
| CN111695611A (en) | Bee colony optimization kernel extreme learning and sparse representation mechanical fault identification method | |
| Kim et al. | Real-time power system transient stability prediction using convolutional layer and long short-term memory | |
| CN114006411A (en) | Wind power prediction method and system based on LSTM-CNN combined model | |
| Xu et al. | An improved ELM-WOA–based fault diagnosis for electric power | |
| Chakraborty et al. | A dual‐tree complex wavelet transform‐based approach for recognition of power system transients | |
| CN114638555B (en) | Power consumption behavior detection method and system based on multilayer regularization extreme learning machine | |
| CN117421571A (en) | Topology real-time identification method and system based on power distribution network | |
| CN117407675A (en) | Lightning arrester leakage current prediction method based on multi-variable reconstruction combined dynamic weight | |
| CN117439069A (en) | Electric quantity prediction method based on neural network | |
| Wang et al. | Short term wind speed forecasting based on feature extraction by cnn and mlp | |
| CN115271198A (en) | Net load prediction method and device of photovoltaic equipment | |
| CN114372418A (en) | Wind power space-time situation description model establishing method | |
| Lin et al. | Network security situation prediction based on combining 3D-CNNs and Bi-GRUs | |
| CN113808071A (en) | Non-invasive load monitoring method and system based on deep learning | |
| Shen et al. | Learning to prune in training via dynamic channel propagation | |
| CN111008584A (en) | Electric energy quality measurement deficiency repairing method of fuzzy self-organizing neural network |
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 |