CN118003961B - Intelligent charging pile group control system and method - Google Patents
Intelligent charging pile group control system and method Download PDFInfo
- Publication number
- CN118003961B CN118003961B CN202410058193.XA CN202410058193A CN118003961B CN 118003961 B CN118003961 B CN 118003961B CN 202410058193 A CN202410058193 A CN 202410058193A CN 118003961 B CN118003961 B CN 118003961B
- Authority
- CN
- China
- Prior art keywords
- charging pile
- time sequence
- training
- power
- temperature time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 30
- 230000002159 abnormal effect Effects 0.000 claims abstract description 26
- 238000004458 analytical method Methods 0.000 claims abstract description 11
- 238000004891 communication Methods 0.000 claims abstract description 11
- 239000013598 vector Substances 0.000 claims description 163
- 238000012549 training Methods 0.000 claims description 107
- 239000011159 matrix material Substances 0.000 claims description 59
- 238000013527 convolutional neural network Methods 0.000 claims description 28
- 238000003062 neural network model Methods 0.000 claims description 27
- 238000000605 extraction Methods 0.000 claims description 23
- 238000001514 detection method Methods 0.000 claims description 15
- 238000010586 diagram Methods 0.000 claims description 12
- 238000010219 correlation analysis Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000009826 distribution Methods 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 3
- 238000012098 association analyses Methods 0.000 claims description 2
- 238000012545 processing Methods 0.000 abstract description 14
- 230000006870 function Effects 0.000 description 20
- 238000011176 pooling Methods 0.000 description 10
- 230000004913 activation Effects 0.000 description 6
- 238000012546 transfer Methods 0.000 description 6
- 230000005856 abnormality Effects 0.000 description 5
- 230000002776 aggregation Effects 0.000 description 5
- 238000004220 aggregation Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 230000001419 dependent effect Effects 0.000 description 4
- 230000003993 interaction Effects 0.000 description 4
- 238000007726 management method Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000000547 structure data Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 239000000306 component Substances 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000004931 aggregating effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 239000008358 core component Substances 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007479 molecular analysis Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003012 network analysis Methods 0.000 description 1
- 238000006116 polymerization reaction Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
- B60L53/00—Methods of charging batteries, specially adapted for electric vehicles; Charging stations or on-board charging equipment therefor; Exchange of energy storage elements in electric vehicles
- B60L53/60—Monitoring or controlling charging stations
- B60L53/67—Controlling two or more charging stations
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
- B60L53/00—Methods of charging batteries, specially adapted for electric vehicles; Charging stations or on-board charging equipment therefor; Exchange of energy storage elements in electric vehicles
- B60L53/30—Constructional details of charging stations
- B60L53/31—Charging columns specially adapted for electric vehicles
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
- B60L53/00—Methods of charging batteries, specially adapted for electric vehicles; Charging stations or on-board charging equipment therefor; Exchange of energy storage elements in electric vehicles
- B60L53/60—Monitoring or controlling charging stations
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
- B60L53/00—Methods of charging batteries, specially adapted for electric vehicles; Charging stations or on-board charging equipment therefor; Exchange of energy storage elements in electric vehicles
- B60L53/60—Monitoring or controlling charging stations
- B60L53/62—Monitoring or controlling charging stations in response to charging parameters, e.g. current, voltage or electrical charge
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses an intelligent charging pile cluster control system and method, which are characterized in that the operation state parameters of charging piles, such as a charging power value and a charging temperature value, are transmitted to a cluster controller through communication modules of each charging pile, and in the cluster controller, time sequence collaborative analysis of the operation state parameters of the charging piles is carried out by utilizing a data processing and analysis algorithm, so that the operation state of each charging pile is monitored and analyzed in real time by utilizing the time sequence association of the operation states of the whole charging piles, the abnormal situation is found and processed in time, the working performance and the reliability of the charging piles are improved, and the whole efficiency of a charging network is optimized.
Description
Technical Field
The application relates to the field of intelligent control, in particular to an intelligent charging pile cluster control system and method.
Background
With the popularization of electric vehicles, the demand for charging piles is also increasing. However, the operation state of the charging pile is affected by various factors, such as charging power, charging temperature, and interaction between the charging piles, which may cause abnormal operation of the charging piles, affecting charging efficiency and safety.
However, the conventional management and control manner of the charging piles is generally decentralized, each charging pile independently operates and manages, and lacks overall coordination and optimization capability, which results in insufficient information exchange and resource sharing among the charging piles, and unified management and scheduling of the charging pile clusters cannot be realized, so that the working performance and reliability of the charging piles are affected.
Accordingly, an intelligent charging pile cluster control system is desired.
Disclosure of Invention
The present application has been made to solve the above-mentioned technical problems. The embodiment of the application provides an intelligent charging pile cluster control system and method, which are used for transmitting the operation state parameters of charging piles, such as a charging power value and a charging temperature value, to a cluster controller through communication modules of each charging pile, and carrying out time sequence collaborative analysis on the operation state parameters of the charging piles in the cluster controller by utilizing a data processing and analysis algorithm, so that the operation state of each charging pile is monitored and analyzed in real time by utilizing the time sequence association of the operation states of the whole charging piles, thereby being convenient for finding and processing abnormal conditions in time, improving the working performance and the reliability of the charging piles, and optimizing the whole efficiency of a charging network.
According to an aspect of the present application, there is provided an intelligent charging pile cluster control system, including:
the data acquisition and transmission module is used for transmitting the operation state parameters of each charging pile at a plurality of preset time points in a preset time period to the cluster controller through the communication module of each charging pile, wherein the operation state parameters comprise a charging power value and a charging temperature value;
The operation state parameter data time sequence association module is used for respectively carrying out association coding on operation state parameters of a plurality of preset time points of each charging pile in a preset time period in the cluster controller so as to obtain a plurality of charging pile power-temperature time sequence association matrixes;
The charging pile operation state parameter time sequence feature extraction module is used for respectively carrying out feature extraction on the charging pile power-temperature time sequence correlation matrixes through a charging state parameter time sequence correlation feature extractor based on a deep neural network model so as to obtain a plurality of charging pile power-temperature time sequence correlation feature vectors;
The charging pile operation state parameter time sequence feature correlation analysis module is used for carrying out correlation analysis on each charging pile power-temperature time sequence correlation feature vector in the charging pile power-temperature time sequence correlation feature vectors so as to obtain a correlation topology feature matrix;
the diagram structure association coding module is used for carrying out association coding based on a diagram structure on the plurality of charging pile power-temperature time sequence association characteristic vectors and the association topology characteristic matrix to obtain a plurality of context association topology charging pile power-temperature time sequence association characteristics;
and the detection module of the working state of the charging pile to be analyzed is used for extracting the power-temperature time sequence correlation characteristic of the charging pile to be analyzed from the power-temperature time sequence correlation characteristics of the plurality of context correlation topology charging piles to determine whether the working state of the charging pile to be analyzed is abnormal.
According to another aspect of the present application, there is provided an intelligent charging pile cluster control method, including:
Transmitting operation state parameters of each charging pile at a plurality of preset time points in a preset time period to a cluster controller through a communication module of each charging pile, wherein the operation state parameters comprise a charging power value and a charging temperature value;
respectively carrying out association coding on operation state parameters of a plurality of preset time points of each charging pile in a preset time period in the cluster controller to obtain a plurality of charging pile power-temperature time sequence association matrixes;
Respectively extracting features of the power-temperature time sequence correlation matrixes of the plurality of charging piles by a charging state parameter time sequence correlation feature extractor based on a deep neural network model so as to obtain power-temperature time sequence correlation feature vectors of the plurality of charging piles;
Performing relevance analysis on each charging pile power-temperature time sequence associated feature vector in the plurality of charging pile power-temperature time sequence associated feature vectors to obtain an associated topology feature matrix;
Performing association coding based on a graph structure on the plurality of charging pile power-temperature time sequence association feature vectors and the association topology feature matrix to obtain a plurality of context association topology charging pile power-temperature time sequence association features;
And extracting the power-temperature time sequence related characteristics of the charging pile of the context-related topology from the power-temperature time sequence related characteristics of the charging pile of the context-related topology to determine whether the working state of the charging pile to be analyzed is abnormal.
Compared with the prior art, the intelligent charging pile cluster control system and the intelligent charging pile cluster control method provided by the application have the advantages that the communication modules of the charging piles are used for transmitting the operation state parameters of the charging piles, such as the charging power value and the charging temperature value, to the cluster controller, and the time sequence collaborative analysis of the operation state parameters of the charging piles is carried out by utilizing the data processing and analysis algorithm in the cluster controller, so that the operation state of each charging pile is monitored and analyzed in real time by utilizing the time sequence association of the operation states of the whole charging piles, the abnormal situation can be found and processed in time, the working performance and the reliability of the charging piles are improved, and the whole efficiency of a charging network is optimized.
Drawings
The above and other objects, features and advantages of the present application will become more apparent by describing embodiments of the present application in more detail with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of embodiments of the application and are incorporated in and constitute a part of this specification, illustrate the application and together with the embodiments of the application, and not constitute a limitation to the application. In the drawings, like reference numerals generally refer to like parts or steps.
Fig. 1 is a block diagram of an intelligent charging pile cluster control system according to an embodiment of the present application;
Fig. 2 is a system architecture diagram of an intelligent charging pile cluster control system according to an embodiment of the present application;
FIG. 3 is a block diagram of a plurality of charging pile operating state parameter time sequence feature correlation analysis modules in an intelligent charging pile cluster control system according to an embodiment of the application;
Fig. 4 is a block diagram of a detection module for a working state of a charging pile to be analyzed in an intelligent charging pile cluster control system according to an embodiment of the present application;
fig. 5 is a flowchart of an intelligent charging pile cluster control method according to an embodiment of the present application.
Detailed Description
Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
As used in the specification and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Although the present application makes various references to certain modules in a system according to embodiments of the present application, any number of different modules may be used and run on a user terminal and/or server. The modules are merely illustrative, and different aspects of the systems and methods may use different modules.
A flowchart is used in the present application to describe the operations performed by a system according to embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in order precisely. Rather, the various steps may be processed in reverse order or simultaneously, as desired. Also, other operations may be added to or removed from these processes.
Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
The traditional charging pile management control mode is generally decentralized, each charging pile independently operates and manages, and the overall coordination and optimization capability is lacked, so that the information exchange and resource sharing among the charging piles are insufficient, and unified management and scheduling of the charging pile clusters cannot be realized, thereby influencing the working performance and reliability of the charging piles. Accordingly, an intelligent charging pile cluster control system is desired.
In the technical scheme of the application, an intelligent charging pile cluster control system is provided. Fig. 1 is a block diagram of an intelligent charging pile cluster control system according to an embodiment of the present application. Fig. 2 is a system architecture diagram of an intelligent charging pile cluster control system according to an embodiment of the present application. As shown in fig. 1 and 2, an intelligent charging pile cluster control system 300 according to an embodiment of the present application includes: the data acquisition and transmission module 310 is configured to transmit, to the cluster controller, operation state parameters of each charging pile at a plurality of predetermined time points within a predetermined time period through a communication module of each charging pile, where the operation state parameters include a charging power value and a charging temperature value; the operation state parameter data time sequence association module 320 is configured to perform association encoding on operation state parameters of each charging pile at a plurality of predetermined time points in a predetermined time period in the cluster controller, so as to obtain a plurality of charging pile power-temperature time sequence association matrices; the charging pile operation state parameter time sequence feature extraction module 330 is configured to perform feature extraction on the plurality of charging pile power-temperature time sequence correlation matrices by using a charging state parameter time sequence correlation feature extractor based on a deep neural network model, so as to obtain a plurality of charging pile power-temperature time sequence correlation feature vectors; the charging pile operation state parameter time sequence feature correlation analysis module 340 is configured to perform correlation analysis on each charging pile power-temperature time sequence correlation feature vector in the charging pile power-temperature time sequence correlation feature vectors to obtain a correlation topology feature matrix; the graph structure association coding module 350 is configured to perform association coding based on a graph structure on the plurality of charging pile power-temperature time sequence association feature vectors and the association topology feature matrix to obtain a plurality of context association topology charging pile power-temperature time sequence association features; the charging pile to be analyzed working state detection module 360 is configured to extract a context-related topology charging pile power-temperature time sequence related characteristic of the charging pile to be analyzed from the context-related topology charging pile power-temperature time sequence related characteristics to determine whether an abnormality exists in the working state of the charging pile to be analyzed.
In particular, the data acquisition and transmission module 310 is configured to transmit, to the cluster controller, operation state parameters of each charging pile at a plurality of predetermined time points within a predetermined time period, where the operation state parameters include a charging power value and a charging temperature value, through a communication module of each charging pile. In one example, the charging power values of the respective charging piles at a plurality of predetermined time points within a predetermined period of time may be acquired by an electric power sensor; and acquiring, by a temperature sensor, charging temperature values at the plurality of predetermined time points. It is worth mentioning that an electric power sensor is a device for measuring power in an electric circuit. It is commonly used to monitor power consumption, power quality, and power distribution in electrical power systems. The electric power sensor can measure parameters such as current, voltage, power factor and the like, and calculate the actual power in the circuit. A temperature sensor is a device for measuring the temperature of an environment or object. It can convert the temperature into an electrical or digital signal for real-time monitoring and control of the temperature.
In particular, the operation state parameter data timing correlation module 320 is configured to perform correlation encoding on operation state parameters of each charging pile at a plurality of predetermined time points within a predetermined time period in the cluster controller, so as to obtain a plurality of charging pile power-temperature timing correlation matrices. Considering that the operation state parameters of each charging pile, such as the charging power value and the charging temperature value, have a time sequence dynamic change rule in the time dimension, and have a time sequence cooperative association relationship between the charging power value and the charging temperature value. Therefore, in order to analyze the operation state of each charging pile more fully and accurately, in the cluster controller, the operation state parameters, such as the charging power value and the charging temperature value, of each charging pile at a plurality of predetermined time points within a predetermined time period are subjected to time-series related encoding. In other words, in the technical scheme of the application, the operation state parameters of each charging pile at a plurality of preset time points in a preset time period are respectively associated and encoded to obtain a plurality of charging pile power-temperature time sequence association matrixes.
Accordingly, in one possible implementation manner, the cluster controller may perform association coding on the operation state parameters of each charging pile at a plurality of predetermined time points within a predetermined period of time to obtain a plurality of charging pile power-temperature time sequence association matrices, for example: acquiring running state data of each charging pile in a preset time period; and preprocessing the acquired data. The method comprises the steps of data cleaning, missing value processing, abnormal value processing and the like; the power and temperature data of each charging pile is encoded in association. Associative coding is a method of converting a relationship between a plurality of variables into a matrix form. Different encoding methods may be used, such as correlation coefficients, mutual information, etc. For the relationship between power and temperature, a correlation coefficient can be used to measure the linear correlation between them; and constructing a plurality of charging pile power-temperature time sequence correlation matrixes according to the correlation coding result. Each of the correlation matrices represents a relationship between power and temperature of one of the charging piles. The rows of the matrix represent points in time, and the columns represent power and temperature. Each element in the matrix represents the degree of association of power and temperature at that point in time; and carrying out data analysis and application on the incidence matrix.
In particular, the charging pile operation state parameter time sequence feature extraction module 330 is configured to perform feature extraction on the plurality of charging pile power-temperature time sequence correlation matrices by using a charging state parameter time sequence correlation feature extractor based on a deep neural network model, so as to obtain a plurality of charging pile power-temperature time sequence correlation feature vectors. In other words, in the technical scheme of the application, the plurality of charging pile power-temperature time sequence correlation matrices are respectively subjected to feature mining through a charging state parameter time sequence correlation feature extractor based on a convolutional neural network model, so as to extract time sequence collaborative correlation feature information of the charging power value and the charging temperature value in a time dimension, thereby obtaining a plurality of charging pile power-temperature time sequence correlation feature vectors. Specifically, each layer using the charge state parameter time sequence correlation feature extractor based on the convolutional neural network model performs input data in forward transfer of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; pooling the convolution feature images based on a feature matrix to obtain pooled feature images; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the charge state parameter time sequence correlation feature extractor based on the convolutional neural network model is the power-temperature time sequence correlation feature vectors of the plurality of charge piles, and the input of the first layer of the charge state parameter time sequence correlation feature extractor based on the convolutional neural network model is the power-temperature time sequence correlation matrix of the plurality of charge piles.
It is noted that convolutional neural network (Convolutional Neural Network, CNN for short) is a deep learning model, mainly used for image processing and computer vision tasks. CNNs perform well in processing data (such as images) having a mesh structure, and are capable of automatically learning feature representations. The core idea of CNN is to extract and classify features in images through convolution, pooling and full-connection layers. The following are the general structures and major components of CNN: convolution layer: the convolutional layer is the core component of the CNN. It uses a set of learnable filters (also called convolution kernels) to scan the input image and extract the local features of the image by convolution operations. Each filter slides on the input image and a convolution operation corresponding to the filter weight is calculated to generate a feature map. Different characteristics such as edges, textures and the like can be extracted through the combination of a plurality of filters; pooling layer: the pooling layer is used for reducing the dimension of the feature map, reducing the number of parameters and extracting main features. Common pooling operations have maximum pooling and average pooling. The pooling layer reduces the size of the feature map by selecting a maximum or average value within the local region while retaining the primary features; full tie layer: the full connection layer connects the feature map output by the pooling layer to one or more full connection layers for classification or regression tasks. Each neuron in the full-connection layer is connected with all neurons of the previous layer, and a mapping relation between input characteristics and output categories is established through learning weights and biases; activation function: between the convolutional layer and the fully-connected layer, an activation function is typically added to introduce nonlinear transformations. Common activation functions include ReLU, sigmoid, tanh, and the like. The introduction of the activation function can increase the expression capacity of the model and improve the nonlinear fitting capacity of the model; loss function: during training, the CNN uses the loss function to measure the difference between the model output and the real label. Common loss functions include cross entropy loss and mean square error loss. By minimizing the loss function, the CNN can learn a more accurate classification or regression model. The training process of CNNs typically uses back-propagation algorithms and optimization algorithms (e.g., gradient descent) to update network parameters so that the model can gradually optimize and improve performance. CNN has achieved great success in the field of computer vision, and is widely applied to tasks such as image classification, object detection, semantic segmentation, face recognition, and the like.
In particular, the multiple charging pile operation state parameter time sequence feature correlation analysis module 340 is configured to perform correlation analysis on each charging pile power-temperature time sequence correlation feature vector in the multiple charging pile power-temperature time sequence correlation feature vectors to obtain a correlation topology feature matrix. In particular, in one specific example of the present application, as shown in fig. 3, the plurality of charging pile operation state parameter time sequence feature association analysis modules 340 include: a correlation calculating unit 341, configured to calculate a correlation between any two charging pile power-temperature time sequence correlation feature vectors in the plurality of charging pile power-temperature time sequence correlation feature vectors to obtain a correlation topology matrix; and the correlation topological feature extraction unit 342 is used for passing the correlation topological matrix through a topological feature extractor based on a convolutional neural network model to obtain the correlation topological feature matrix.
Specifically, the correlation calculating unit 341 is configured to calculate a correlation between any two power-temperature time sequence correlation feature vectors of the plurality of power-temperature time sequence correlation feature vectors of the charging pile to obtain a correlation topology matrix. Considering that the power and temperature conditions between charging piles may interact, for example, high power usage of one charging pile may cause the temperature of surrounding charging piles to rise, connection and interaction between charging piles are critical for cluster control and optimal scheduling. In order to analyze and describe the relevance and interaction between the charging piles, in the technical scheme of the application, the relevance between any two charging pile power-temperature time sequence relevance eigenvectors in the plurality of charging pile power-temperature time sequence relevance eigenvectors is further calculated to obtain a relevance topology matrix. By calculating the correlation between the power-temperature time sequence correlation feature vectors of the charging piles, the mutual influence relationship among the charging piles can be revealed, and the overall behavior of the charging pile clusters can be understood.
Specifically, the correlation topological feature extraction unit 342 is configured to pass the correlation topological matrix through a topological feature extractor based on a convolutional neural network model to obtain the correlation topological feature matrix. In other words, in the technical scheme of the application, feature mining is performed on the association topology matrix through a topology feature extractor based on a convolutional neural network model, so as to extract correlation association feature distribution information between power and temperature time sequence cooperative association features of each charging pile, thereby obtaining an association topology feature matrix. More specifically, each layer using the topological feature extractor based on the convolutional neural network model performs the following steps on input data in forward transfer of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; pooling the convolution feature map along a channel dimension to obtain a pooled feature map; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the topological feature extractor based on the convolutional neural network model is the association topological feature matrix, and the input of the first layer of the topological feature extractor based on the convolutional neural network model is the association topological matrix.
It should be noted that, in other specific examples of the present application, the correlation analysis may be performed on each of the plurality of charging pile power-temperature time sequence correlation feature vectors in other manners to obtain a correlation topology feature matrix, for example: acquiring power-temperature time sequence associated feature vectors of a plurality of charging piles; for each pair of the power-temperature time sequence correlation eigenvectors of the charging pile, calculating a correlation coefficient between them. Common correlation coefficients include pearson correlation coefficients and spearman correlation coefficients. The correlation coefficient measures the linear correlation between two variables, and the value range of the correlation coefficient is-1 to 1, wherein-1 represents the complete negative correlation, 1 represents the complete positive correlation, and 0 represents no correlation; and constructing a correlation topology feature matrix according to the calculated correlation coefficient. The rows and columns of the matrix represent the indices of the charging posts, respectively, and each element in the matrix represents a correlation coefficient between corresponding charging posts. The absolute value of the correlation coefficient may be used as the value of the matrix element to represent the strength of the correlation; and analyzing the associated topology feature matrix. The associated topological features in the matrix, such as the degree of nodes, cluster coefficients, median centrality, etc., can be analyzed using a graph theory approach. These features may reveal the degree of association and topology between charging piles; the association topological feature matrix is visualized so as to more intuitively understand the association relation between the charging piles.
In particular, the graph structure association encoding module 350 is configured to perform graph structure-based association encoding on the plurality of charging pile power-temperature time sequence association feature vectors and the association topology feature matrix to obtain a plurality of context association topology charging pile power-temperature time sequence association features. That is, the power-temperature time sequence association characteristic vector of each charging pile is used as the characteristic representation of the node, the association topology characteristic matrix is used as the characteristic representation of the edge between the nodes, and the power-temperature time sequence association characteristic matrix of the charging piles and the association topology characteristic matrix obtained by two-dimensional arrangement of the power-temperature time sequence association characteristic vectors of the plurality of charging piles are used for obtaining a plurality of context association topology charging pile power-temperature time sequence association characteristic vectors through a graph neural network model. Specifically, the graph neural network model performs graph structure data coding on the charging pile power-temperature time sequence association characteristic matrix and the association topology characteristic matrix through a learnable neural network parameter to obtain the context association topology charging pile power-temperature time sequence association characteristic vectors containing irregular charging pile mutual association topology characteristics and power-temperature time sequence cooperative association characteristic information of each charging pile.
Notably, the graph neural network (Graph Neural Network, GNN for short) is a type of machine learning model for processing graph structure data. The core idea of GNN is to learn the representation of nodes through information transfer and aggregation between nodes. The method gradually merges the characteristics of the node and the characteristics of the neighbor nodes thereof by iteratively updating the characteristic vector of the node, thereby capturing the local and global information in the graph structure. The following are the general structures and major components of GNNs: node representation learning: GNNs represent each node by defining a feature vector of the node. Initially, each node has an initial feature vector. The GNN then enables it to capture local and global information of the node by iteratively updating the feature vectors of the node. The process of node representation learning typically includes information transfer and aggregation operations; information transfer: information transfer is one of the core operations in GNN. The method updates the characteristics of the nodes by transmitting the characteristic vectors of the nodes to the neighbor nodes and combining the characteristics of the neighbor nodes. Information delivery may use a messaging mechanism in which each node generates messages from the feature vectors of its neighboring nodes and aggregates these messages into its own feature vector; polymerization: aggregation is a complementary operation of information delivery for aggregating messages received by a node. Common aggregation modes include summation, averaging, maximization, etc. The purpose of the aggregation operation is to incorporate the information of the neighbor nodes into the feature vector of the current node to obtain a more comprehensive node representation; output prediction: after the node representation learning is completed, further task prediction, such as node classification, graph classification, link prediction and the like, can be performed according to the feature vector of the node. Output prediction may be performed using a fully connected layer or other suitable model structure. The training process of GNNs typically uses back-propagation and optimization algorithms to update network parameters so that the model can gradually optimize and improve performance. GNNs have achieved significant results in many areas, such as social network analysis, recommendation systems, image annotation, chemical molecular analysis, and the like. The method can effectively process the graph structure data and capture the relation and the dependence among the nodes.
In particular, the to-be-analyzed charging pile working state detection module 360 is configured to extract a context-related topology charging pile power-temperature time sequence related characteristic of the to-be-analyzed charging pile from the plurality of context-related topology charging pile power-temperature time sequence related characteristics to determine whether an abnormality exists in the working state of the to-be-analyzed charging pile. In particular, in one specific example of the present application, as shown in fig. 4, the charging pile operation state detection module 360 to be analyzed includes: a charging pile operation state feature extraction unit 361 to be analyzed is configured to extract a context-related topological charging pile power-temperature time sequence related feature vector of a charging pile to be analyzed from the context-related topological charging pile power-temperature time sequence related feature vectors as a context-related topological charging pile power-temperature time sequence related feature of the charging pile to be analyzed; the charging pile working state to be analyzed judging unit 362 is configured to use the context-related topology charging pile power-temperature time sequence related feature of the charging pile to be analyzed as a charging pile working state feature vector to be analyzed, and pass through a classifier to obtain a classification result, where the classification result is used to indicate whether the working state of the charging pile to be analyzed is abnormal.
Specifically, the charging pile running state feature extraction unit 361 is configured to extract, from the context-associated topology charging pile power-temperature time sequence associated feature vectors, a context-associated topology charging pile power-temperature time sequence associated feature vector of the charging pile to be analyzed as a context-associated topology charging pile power-temperature time sequence associated feature of the charging pile to be analyzed. In order to be able to analyze the operational state of the charging pile to be analyzed, it is necessary to extract a context-dependent topology charging pile power-temperature time-sequence-dependent feature vector of the charging pile to be analyzed from the plurality of context-dependent topology charging pile power-temperature time-sequence-dependent feature vectors as the operational state feature vector of the charging pile to be analyzed.
Specifically, the to-be-analyzed charging pile working state judging unit 362 is configured to use the context-related topology charging pile power-temperature time sequence related feature of the to-be-analyzed charging pile as the to-be-analyzed charging pile working state feature vector to pass through a classifier to obtain a classification result, where the classification result is used to indicate whether the to-be-analyzed charging pile working state is abnormal. In particular, in one specific example of the present application, the charging pile operation state determination unit 362 to be analyzed includes: the working state detection classification subunit is used for enabling the feature vector of the working state of the charging pile to be analyzed to pass through the classifier to obtain a classification result, and the classification result is used for indicating whether the working state of the charging pile to be analyzed is abnormal or not.
More specifically, the working state detection classifying subunit is configured to pass the feature vector of the working state of the charging pile to be analyzed through the classifier to obtain a classification result, where the classification result is used to indicate whether the working state of the charging pile to be analyzed is abnormal. That is, in a specific example of the present application, the feature vector of the operation state of the charging pile to be analyzed is passed through a classifier to obtain a classification result, where the classification result is used to indicate whether there is an abnormality in the operation state of the charging pile to be analyzed. That is, the running state characteristics of the charging piles to be analyzed in the whole charging pile cluster are utilized to carry out classification processing, so that the running states of the charging piles to be analyzed are monitored and analyzed in real time by utilizing the time sequence correlation of the running states of the whole charging piles, abnormal situations can be found and processed in time, the working performance and the reliability of the charging piles are improved, and the whole efficiency of a charging network is optimized. Specifically, the feature vector of the operation state of the charging pile to be analyzed is passed through the classifier to obtain a classification result, where the classification result is used to indicate whether the working state of the charging pile to be analyzed is abnormal, and the method includes: performing full-connection coding on the running state feature vector of the charging pile to be analyzed by using a plurality of full-connection layers of the classifier to obtain a coding classification feature vector; and passing the coding classification feature vector through a Softmax classification function of the classifier to obtain the classification result.
It should be noted that, in other specific examples of the present application, the context-related topology charging pile power-temperature time sequence related feature of the charging pile to be analyzed may be used as the running state feature vector of the charging pile to be analyzed by other ways to obtain a classification result through a classifier, where the classification result is used to indicate whether the working state of the charging pile to be analyzed is abnormal, for example: collecting power-temperature time sequence associated characteristic data comprising a charging pile to be analyzed and surrounding charging piles; extracting the power-temperature time sequence related characteristics of the context related topology charging pile of the charging pile to be analyzed from the collected data; combining the extracted features into a feature vector of the charging pile to be analyzed; adding a label to the characteristic vector of the charging pile to be analyzed to indicate whether the working state of the charging pile is abnormal; training a classifier model using feature vectors with labeling data; and predicting the classification result of the feature vector of the charging pile to be analyzed by using the trained classifier. Inputting the feature vector into a classifier model to obtain a classification result, wherein the classification result indicates whether the working state of the charging pile to be analyzed is abnormal or not; and analyzing the classification result and taking corresponding decisions according to the need.
It should be noted that, in other specific examples of the present application, the context-related topology charging pile power-temperature time sequence related features of the charging pile to be analyzed may be extracted from the plurality of context-related topology charging pile power-temperature time sequence related features in other manners to determine whether the working state of the charging pile to be analyzed is abnormal, for example: collecting power-temperature time sequence associated characteristic data comprising a plurality of charging piles; and extracting the context information of the charging pile to be analyzed from the collected data. The context information may include the number of charging piles around the charging pile to be analyzed, a positional relationship, a connection relationship, etc.; and extracting the power-temperature time sequence related characteristics of the charging pile from the collected data. Common characteristics include maximum power, minimum power, average power, power fluctuation, rate of temperature change, etc.; and determining whether the working state of the charging pile to be analyzed is abnormal or not by using a proper abnormality detection method. Common anomaly detection methods include threshold-based methods, statistical-based methods, machine-learning-based methods, and the like; analyzing the result of abnormality detection, and taking corresponding measures as required.
Further, in a specific example of the present application, the intelligent charging pile cluster control system 300 further includes: and the training module is used for training the charge state parameter time sequence associated feature extractor based on the deep neural network model, the topological feature extractor based on the convolution neural network model, the graph neural network model and the classifier. Wherein, training module includes: the training data acquisition unit is used for acquiring training operation state parameters of each charging pile at a plurality of preset time points in a preset time period and judging whether the working state of the charging pile to be analyzed has an abnormal true value or not; the training operation state parameter data time sequence association unit is used for respectively carrying out association coding on the training operation state parameters of a plurality of preset time points of each charging pile in a preset time period so as to obtain a plurality of training charging pile power-temperature time sequence association matrixes; The training charging pile operation state parameter time sequence feature extraction unit is used for respectively carrying out feature extraction on the plurality of training charging pile power-temperature time sequence correlation matrixes through a charging state parameter time sequence correlation feature extractor based on a deep neural network model so as to obtain a plurality of training charging pile power-temperature time sequence correlation feature vectors; the training association topology matrix acquisition unit is used for calculating the correlation between any two training charging pile power-temperature time sequence association feature vectors in the plurality of training charging pile power-temperature time sequence association feature vectors to obtain a training association topology matrix; the training correlation topological feature extraction unit is used for enabling the training correlation topological matrix to pass through a topological feature extractor based on a convolutional neural network model to obtain a training correlation topological feature matrix; the training association coding unit is used for obtaining a plurality of training context association topology charging pile power-temperature time sequence association characteristic vectors through a graph neural network model by using the plurality of training charging pile power-temperature time sequence association characteristic vectors and the training association topology characteristic matrix; the training charging pile operation state feature extraction unit is used for extracting training context associated topology charging pile power-temperature time sequence associated feature vectors of the charging pile to be analyzed from the plurality of training context associated topology charging pile power-temperature time sequence associated feature vectors to serve as training context associated topology charging pile power-temperature time sequence associated features of the charging pile to be analyzed; The classification loss function value acquisition unit is used for taking the context-associated topology charging pile power-temperature time sequence associated characteristic of the charging pile to be analyzed as the running state characteristic vector of the charging pile to be analyzed through a classifier to obtain a classification loss function value; the loss function value calculation unit is used for calculating the loss function value of the training charging pile running state feature vector to be analyzed and the training charging pile power-temperature time sequence correlation feature vector corresponding to the training charging pile running state feature vector; and the back propagation training unit is used for training the deep neural network model-based charge state parameter time sequence correlation feature extractor, the convolutional neural network model-based topological feature extractor, the graph neural network model and the classifier by means of back propagation of gradient descent by means of weighted sum of the training charge pile running state feature vector to be analyzed and the corresponding training charge pile power-temperature time sequence correlation feature vector.
In particular, in the technical scheme of the application, the plurality of training charging pile power-temperature time sequence associated feature vectors respectively express local time sequence high-order associated features under the global time domain source data association of the power values and the temperature values of the corresponding charging piles, so that after the plurality of training charging pile power-temperature time sequence associated feature vectors and the training associated topology feature matrix pass through a graph neural network model, the topology associated features of the power-temperature time sequence high-order associated features of each charging pile under the sample space feature associated topology are further extracted, that is, the training charging pile running state feature vector to be analyzed is equivalent to the interpolation topology associated feature mixture of the training charging pile power-temperature time sequence associated feature vectors corresponding to the training charging pile running state feature vector. However, the feature group density representation between the running state feature vector of the charging pile to be analyzed and the corresponding power-temperature time sequence associated feature vector of the training charging pile is inconsistent, so that iteration imbalance exists between the power-temperature time sequence associated feature extraction based on the convolutional neural network model and the topological association of the graph neural network model during the overall training of the model, and the overall training efficiency of the model is affected.
Therefore, the application considers improving the consistency of the feature group density representation between the running state feature vector of the charging pile to be analyzed and the corresponding power-temperature time sequence associated feature vector of the charging pile to be analyzed, thereby further introducing a loss function for the running state feature vector of the charging pile to be analyzed and the corresponding power-temperature time sequence associated feature vector of the charging pile to be analyzed.
Specifically, in one embodiment of the present application, the loss function value calculation unit is configured to: and calculating the loss function value of the training charging pile running state feature vector to be analyzed and the training charging pile power-temperature time sequence correlation feature vector corresponding to the training charging pile running state feature vector according to the following loss function value calculation formula.
Wherein,Is the characteristic vector of the running state of the charging pile to be analyzed in training,Is the corresponding power-temperature time sequence associated characteristic vector of the training charging pile, and the training to-be-analyzed charging pile running state characteristic vectorAnd corresponding training charging pile power-temperature time sequence associated feature vectorThe length is the same as the length of the tube,Is the length of the feature vector and,Is the first training of the feature vector of the running state of the charging pile to be analyzedThe characteristic value of the individual position is used,Is the corresponding training charging pile power-temperature time sequence related characteristic vectorThe characteristic value of the individual position is used,Representing the per-position subtraction of vectors, anRepresenting the square of the two norms of the vector,And the loss function value of the training charging pile operation state characteristic vector to be analyzed and the training charging pile power-temperature time sequence correlation characteristic vector corresponding to the training charging pile operation state characteristic vector to be analyzed is represented.
Here, the loss function performs group count attention based on feature group density, and performs adaptive attention of different density representation modes between the training charging pile operation state feature vector to be analyzed and the training charging pile power-temperature time sequence associated feature vector corresponding to the training charging pile operation state feature vector by using group count as a recursive mapping of output feature group density. By taking the model as a loss function to train the model, the model can avoid overestimation and underestimation aiming at different density modes under the characteristic distribution of the running state characteristic vector of the charging pile to be analyzed and the corresponding power-temperature time sequence associated characteristic vector of the charging pile to be analyzed, and learn the corresponding relation between the characteristic value distribution and the group density distribution, thereby realizing the consistency optimization of the characteristic group density representation between the running state characteristic vector of the charging pile to be analyzed with different characteristic densities and the corresponding power-temperature time sequence associated characteristic vector of the charging pile to be analyzed, and improving the integral training efficiency of the model.
As described above, the intelligent charging pile cluster control system 300 according to the embodiment of the present application may be implemented in various wireless terminals, for example, a server having an intelligent charging pile cluster control algorithm, etc. In one possible implementation, the intelligent charging pile cluster control system 300 according to an embodiment of the present application may be integrated into the wireless terminal as a software module and/or hardware module. For example, the smart charge pile cluster control system 300 may be a software module in the operating system of the wireless terminal, or may be an application developed for the wireless terminal; of course, the intelligent charging pile cluster control system 300 may also be one of a plurality of hardware modules of the wireless terminal.
Alternatively, in another example, the intelligent charging pile cluster control system 300 and the wireless terminal may be separate devices, and the intelligent charging pile cluster control system 300 may be connected to the wireless terminal through a wired and/or wireless network and transmit interaction information in accordance with a agreed data format.
Further, an intelligent charging pile cluster control method is provided.
Fig. 5 is a flowchart of an intelligent charging pile cluster control method according to an embodiment of the present application. As shown in fig. 5, the intelligent charging pile cluster control method according to the embodiment of the application includes the steps of: s1, transmitting operation state parameters of each charging pile at a plurality of preset time points in a preset time period to a cluster controller through a communication module of each charging pile, wherein the operation state parameters comprise a charging power value and a charging temperature value; s2, respectively carrying out association coding on operation state parameters of each charging pile at a plurality of preset time points in a preset time period in the cluster controller to obtain a plurality of charging pile power-temperature time sequence association matrixes; s3, respectively extracting features of the power-temperature time sequence correlation matrixes of the plurality of charging piles by a charging state parameter time sequence correlation feature extractor based on a deep neural network model to obtain power-temperature time sequence correlation feature vectors of the plurality of charging piles; s4, performing relevance analysis on each charging pile power-temperature time sequence relevance feature vector in the plurality of charging pile power-temperature time sequence relevance feature vectors to obtain a relevance topology feature matrix; s5, carrying out association coding based on a graph structure on the power-temperature time sequence association characteristic vectors of the plurality of charging piles and the association topology characteristic matrix to obtain power-temperature time sequence association characteristics of the plurality of context association topology charging piles; and S6, extracting the power-temperature time sequence related characteristics of the topological charging pile to be analyzed from the power-temperature time sequence related characteristics of the topological charging pile to be analyzed to determine whether the working state of the charging pile to be analyzed is abnormal.
In summary, the intelligent charging pile cluster control method according to the embodiment of the application is explained, the communication module of each charging pile is used for transmitting the operation state parameters of the charging pile, such as the charging power value and the charging temperature value, to the cluster controller, and in the cluster controller, the time sequence collaborative analysis of the operation state parameters of the charging pile is performed by utilizing the data processing and analysis algorithm, so that the operation state of each charging pile is monitored and analyzed in real time by utilizing the time sequence association of the operation states of the whole charging piles, thereby being convenient for finding and processing abnormal situations in time, improving the working performance and the reliability of the charging pile, and optimizing the whole efficiency of the charging network.
The foregoing description of the embodiments of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the various embodiments described. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or the improvement of technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (7)
1. An intelligent charging pile cluster control system, comprising:
the data acquisition and transmission module is used for transmitting the operation state parameters of each charging pile at a plurality of preset time points in a preset time period to the cluster controller through the communication module of each charging pile, wherein the operation state parameters comprise a charging power value and a charging temperature value;
The operation state parameter data time sequence association module is used for respectively carrying out association coding on operation state parameters of a plurality of preset time points of each charging pile in a preset time period in the cluster controller so as to obtain a plurality of charging pile power-temperature time sequence association matrixes;
The charging pile operation state parameter time sequence feature extraction module is used for respectively carrying out feature extraction on the charging pile power-temperature time sequence correlation matrixes through a charging state parameter time sequence correlation feature extractor based on a deep neural network model so as to obtain a plurality of charging pile power-temperature time sequence correlation feature vectors;
The charging pile operation state parameter time sequence feature correlation analysis module is used for carrying out correlation analysis on each charging pile power-temperature time sequence correlation feature vector in the charging pile power-temperature time sequence correlation feature vectors so as to obtain a correlation topology feature matrix;
the diagram structure association coding module is used for carrying out association coding based on a diagram structure on the plurality of charging pile power-temperature time sequence association characteristic vectors and the association topology characteristic matrix to obtain a plurality of context association topology charging pile power-temperature time sequence association characteristics;
the charging pile working state detection module to be analyzed is used for extracting the context-associated topology charging pile power-temperature time sequence associated characteristics of the charging pile to be analyzed from the context-associated topology charging pile power-temperature time sequence associated characteristics to determine whether the working state of the charging pile to be analyzed is abnormal;
the time sequence characteristic association analysis module for the operation state parameters of the plurality of charging piles comprises:
The correlation calculation unit is used for calculating the correlation between any two charging pile power-temperature time sequence correlation characteristic vectors in the plurality of charging pile power-temperature time sequence correlation characteristic vectors so as to obtain a correlation topology matrix;
The correlation topological feature extraction unit is used for enabling the correlation topological matrix to pass through a topological feature extractor based on a convolutional neural network model to obtain the correlation topological feature matrix;
Wherein, still include: the training module is used for training the charge state parameter time sequence associated feature extractor based on the deep neural network model, the topological feature extractor based on the convolution neural network model, the graph neural network model and the classifier;
wherein, training module includes:
The training data acquisition unit is used for acquiring training operation state parameters of each charging pile at a plurality of preset time points in a preset time period and judging whether the working state of the charging pile to be analyzed has an abnormal true value or not;
the training operation state parameter data time sequence association unit is used for respectively carrying out association coding on the training operation state parameters of a plurality of preset time points of each charging pile in a preset time period so as to obtain a plurality of training charging pile power-temperature time sequence association matrixes;
The training charging pile operation state parameter time sequence feature extraction unit is used for respectively carrying out feature extraction on the plurality of training charging pile power-temperature time sequence correlation matrixes through a charging state parameter time sequence correlation feature extractor based on a deep neural network model so as to obtain a plurality of training charging pile power-temperature time sequence correlation feature vectors;
the training association topology matrix acquisition unit is used for calculating the correlation between any two training charging pile power-temperature time sequence association feature vectors in the plurality of training charging pile power-temperature time sequence association feature vectors to obtain a training association topology matrix;
the training correlation topological feature extraction unit is used for enabling the training correlation topological matrix to pass through a topological feature extractor based on a convolutional neural network model to obtain a training correlation topological feature matrix;
the training association coding unit is used for obtaining a plurality of training context association topology charging pile power-temperature time sequence association characteristic vectors through a graph neural network model by using the plurality of training charging pile power-temperature time sequence association characteristic vectors and the training association topology characteristic matrix;
The training charging pile operation state feature extraction unit is used for extracting training context associated topology charging pile power-temperature time sequence associated feature vectors of the charging pile to be analyzed from the plurality of training context associated topology charging pile power-temperature time sequence associated feature vectors to serve as training context associated topology charging pile power-temperature time sequence associated features of the charging pile to be analyzed;
The classification loss function value acquisition unit is used for taking the context-associated topology charging pile power-temperature time sequence associated characteristic of the charging pile to be analyzed as the running state characteristic vector of the charging pile to be analyzed through a classifier to obtain a classification loss function value;
The loss function value calculation unit is used for calculating the loss function value of the training charging pile running state feature vector to be analyzed and the training charging pile power-temperature time sequence correlation feature vector corresponding to the training charging pile running state feature vector;
The back propagation training unit is used for training the state of charge parameter time sequence correlation feature extractor based on the deep neural network model, the topological feature extractor based on the convolutional neural network model, the graph neural network model and the classifier by means of back propagation of gradient descent by means of weighted sum of the state of charge feature vector to be analyzed and the corresponding training charging pile power-temperature time sequence correlation feature vector;
wherein the loss function value calculation unit is configured to: calculating a loss function value of the training charging pile running state feature vector to be analyzed and the training charging pile power-temperature time sequence correlation feature vector corresponding to the training charging pile running state feature vector according to the following loss function value calculation formula;
Wherein, Is the characteristic vector of the running state of the charging pile to be analyzed in training,Is the corresponding power-temperature time sequence associated characteristic vector of the training charging pile, and the training to-be-analyzed charging pile running state characteristic vectorAnd corresponding training charging pile power-temperature time sequence associated feature vectorThe length is the same as the length of the tube,Is the length of the feature vector and,Is the first training of the feature vector of the running state of the charging pile to be analyzedThe characteristic value of the individual position is used,Is the corresponding training charging pile power-temperature time sequence related characteristic vectorThe characteristic value of the individual position is used,Representing the per-position subtraction of vectors, anRepresenting the square of the two norms of the vector,Representing the running state feature vector of the charging pile to be analyzed and the loss function value of the corresponding power-temperature time sequence associated feature vector of the charging pile;
In the loss function value calculation formula, the loss function performs group counting attention based on feature group density, and performs self-adaptive attention of different density representation modes between the training charging pile running state feature vector to be analyzed and the training charging pile power-temperature time sequence associated feature vector corresponding to the training charging pile running state feature vector by taking group counting as a recursive mapping of output feature group density; by taking the model as a loss function to train the model, the model can avoid overestimation and underestimation aiming at different density modes under the characteristic distribution of the running state characteristic vector of the charging pile to be analyzed and the corresponding power-temperature time sequence associated characteristic vector of the charging pile to be analyzed, and learn the corresponding relation between the characteristic value distribution and the group density distribution, thereby realizing the consistency optimization of the characteristic group density representation between the running state characteristic vector of the charging pile to be analyzed with different characteristic densities and the corresponding power-temperature time sequence associated characteristic vector of the charging pile to be analyzed, and improving the integral training efficiency of the model.
2. The intelligent charging stake cluster control system of claim 1, wherein the deep neural network model is a convolutional neural network model.
3. The intelligent charging pile cluster control system of claim 2, wherein the graph structure association encoding module is configured to: and the power-temperature time sequence association characteristic vectors of the plurality of charging piles and the association topology characteristic matrix are used for obtaining a plurality of context association topology charging pile power-temperature time sequence association characteristic vectors serving as the power-temperature time sequence association characteristics of the plurality of context association topology charging piles through a graph neural network model.
4. The intelligent charging pile cluster control system of claim 3, wherein the charging pile operating state detection module to be analyzed comprises:
the charging pile operation state feature extraction unit to be analyzed is used for extracting the context-associated topological charging pile power-temperature time sequence associated feature vector of the charging pile to be analyzed from the context-associated topological charging pile power-temperature time sequence associated feature vectors as the context-associated topological charging pile power-temperature time sequence associated feature of the charging pile to be analyzed;
The charging pile working state judging unit to be analyzed is used for taking the context-related topological charging pile power-temperature time sequence related characteristic of the charging pile to be analyzed as the charging pile running state characteristic vector to be analyzed to pass through the classifier so as to obtain a classification result, wherein the classification result is used for indicating whether the working state of the charging pile to be analyzed is abnormal or not.
5. The intelligent charging pile cluster control system according to claim 4, wherein the charging pile operation state determination unit to be analyzed includes:
The working state detection classification subunit is used for enabling the feature vector of the working state of the charging pile to be analyzed to pass through the classifier to obtain a classification result, and the classification result is used for indicating whether the working state of the charging pile to be analyzed is abnormal or not.
6. The intelligent charging pile cluster control system of claim 5, wherein the operating state detection classification subunit comprises:
The full-connection coding secondary subunit is used for carrying out full-connection coding on the running state feature vector of the charging pile to be analyzed by using a plurality of full-connection layers of the classifier so as to obtain a coding classification feature vector; and
And the classification result generation secondary subunit is used for passing the coding classification feature vector through a Softmax classification function of the classifier to obtain the classification result.
7. An intelligent charging pile cluster control method for the intelligent charging pile cluster control system according to claim 1, comprising:
Transmitting operation state parameters of each charging pile at a plurality of preset time points in a preset time period to a cluster controller through a communication module of each charging pile, wherein the operation state parameters comprise a charging power value and a charging temperature value;
respectively carrying out association coding on operation state parameters of a plurality of preset time points of each charging pile in a preset time period in the cluster controller to obtain a plurality of charging pile power-temperature time sequence association matrixes;
Respectively extracting features of the power-temperature time sequence correlation matrixes of the plurality of charging piles by a charging state parameter time sequence correlation feature extractor based on a deep neural network model so as to obtain power-temperature time sequence correlation feature vectors of the plurality of charging piles;
Performing relevance analysis on each charging pile power-temperature time sequence associated feature vector in the plurality of charging pile power-temperature time sequence associated feature vectors to obtain an associated topology feature matrix;
Performing association coding based on a graph structure on the plurality of charging pile power-temperature time sequence association feature vectors and the association topology feature matrix to obtain a plurality of context association topology charging pile power-temperature time sequence association features;
And extracting the power-temperature time sequence related characteristics of the charging pile of the context-related topology from the power-temperature time sequence related characteristics of the charging pile of the context-related topology to determine whether the working state of the charging pile to be analyzed is abnormal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410058193.XA CN118003961B (en) | 2024-01-16 | 2024-01-16 | Intelligent charging pile group control system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410058193.XA CN118003961B (en) | 2024-01-16 | 2024-01-16 | Intelligent charging pile group control system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118003961A CN118003961A (en) | 2024-05-10 |
| CN118003961B true CN118003961B (en) | 2024-08-20 |
Family
ID=90946545
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410058193.XA Active CN118003961B (en) | 2024-01-16 | 2024-01-16 | Intelligent charging pile group control system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118003961B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118405019B (en) * | 2024-07-01 | 2024-08-30 | 广东南缆电缆有限公司 | Fill electric pile system with intelligent operation maintenance function |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104029602A (en) * | 2014-03-05 | 2014-09-10 | 广州供电局有限公司 | Alternating current charging pile system on basis of TD-LTE (time division long term evolution) network |
| CN116639010A (en) * | 2023-07-24 | 2023-08-25 | 江西驴宝宝通卡科技有限公司 | Intelligent control system and method for charging pile |
| CN117124910A (en) * | 2023-09-20 | 2023-11-28 | 漳州建源电力工程有限公司 | Smart city charging pile node fault alarm system and method |
Family Cites Families (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11336662B2 (en) * | 2018-11-21 | 2022-05-17 | Abb Schweiz Ag | Technologies for detecting abnormal activities in an electric vehicle charging station |
| FI20195682A1 (en) * | 2019-08-15 | 2021-02-16 | Liikennevirta Oy / Virta Ltd | Charging station monitoring method and device |
| CN114898282A (en) * | 2021-01-27 | 2022-08-12 | 腾讯科技（深圳）有限公司 | Image processing method and device |
| CN113175947A (en) * | 2021-03-24 | 2021-07-27 | 北京中电飞华通信有限公司 | Charging station abnormity early warning method, intelligent operation and maintenance gateway and early warning system |
| CN113191301B (en) * | 2021-05-14 | 2023-04-18 | 上海交通大学 | Video dense crowd counting method and system integrating time sequence and spatial information |
| AU2022273313A1 (en) * | 2021-05-14 | 2023-11-30 | Veloce Energy, Inc. | Systems and methods for electric vehicle charging power distribution |
| CN114690701A (en) * | 2022-04-12 | 2022-07-01 | 池州学院 | Multifunctional intelligent charging pile and cluster control system |
| CN114898301A (en) * | 2022-06-08 | 2022-08-12 | 合肥市轨道交通集团有限公司 | Crowd density statistical method, device, equipment and medium based on detection and segmentation |
| CN115858609A (en) * | 2022-08-05 | 2023-03-28 | 国网天津市电力公司 | Electric vehicle charging pile state monitoring method, fault identification method and electronic equipment |
| CN116442836A (en) * | 2023-03-30 | 2023-07-18 | 厦门佳因特科技有限公司 | Distributed pile group management system |
| CN116494816B (en) * | 2023-06-30 | 2023-09-15 | 江西驴宝宝通卡科技有限公司 | Charging management system and method for charging pile |
| CN116552306B (en) * | 2023-07-12 | 2023-09-22 | 江西驴充充充电技术有限公司 | Monitoring system and method for direct current pile |
-
2024
- 2024-01-16 CN CN202410058193.XA patent/CN118003961B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104029602A (en) * | 2014-03-05 | 2014-09-10 | 广州供电局有限公司 | Alternating current charging pile system on basis of TD-LTE (time division long term evolution) network |
| CN116639010A (en) * | 2023-07-24 | 2023-08-25 | 江西驴宝宝通卡科技有限公司 | Intelligent control system and method for charging pile |
| CN117124910A (en) * | 2023-09-20 | 2023-11-28 | 漳州建源电力工程有限公司 | Smart city charging pile node fault alarm system and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118003961A (en) | 2024-05-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020077672A1 (en) | Method and device for training service quality evaluation model | |
| CN118003961B (en) | Intelligent charging pile group control system and method | |
| CN113794695B (en) | GRU-based network abnormal flow detection and identification method | |
| CN113821798B (en) | Method and system for detecting illegal accounts of Ethernet based on heterogeneous graph neural network | |
| CN117041017B (en) | Intelligent operation and maintenance management method and system for data center | |
| CN117156442B (en) | Cloud data security protection method and system based on 5G network | |
| CN116934304A (en) | Intelligent power distribution room equipment operation maintenance management system and method thereof | |
| CN110807518A (en) | Outlier detection method for power grid data | |
| CN114528755A (en) | Power equipment fault detection model based on attention mechanism combined with GRU | |
| CN112769869A (en) | SDN network security prediction method based on Bayesian attack graph and corresponding system | |
| CN118133203A (en) | Fault diagnosis method for electric energy metering detection information | |
| CN114048546A (en) | Graph convolution network and unsupervised domain self-adaptive prediction method for residual service life of aircraft engine | |
| CN118466413A (en) | Control method and device for unmanned equipment of power plant based on fuzzy logic | |
| CN118193954A (en) | Power distribution network abnormal data detection method and system based on edge calculation | |
| CN118316033A (en) | Power load prediction method based on CNN-GRU and ARIMA model | |
| CN118468186A (en) | Wind power data anomaly detection method and system based on federal learning mechanism | |
| CN117575685A (en) | Data analysis early warning system and method | |
| CN117676099A (en) | Security early warning method and system based on Internet of things | |
| Shen et al. | Long-term multivariate time series forecasting in data centers based on multi-factor separation evolutionary spatial–temporal graph neural networks | |
| CN113077038A (en) | Industrial data feature selection method and device, computer equipment and storage medium | |
| CN118378178B (en) | Transformer fault identification method and system based on residual map convolution neural network | |
| Xia et al. | Smart substation network fault classification based on a hybrid optimization algorithm | |
| CN118607886B (en) | Material supply and demand balance planning method and system | |
| CN118282876B (en) | Abnormal node detection method and system for incomplete heterogeneous Ethernet network | |
| CN117336187B (en) | Unmanned aerial vehicle communication network inference method based on inter-edge association |
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 |