Charging station and power grid interaction state monitoring method based on multi-dimensional data fusion

Abstract

The invention discloses a charging station and power grid interaction state monitoring method based on multi-dimensional data fusion, and the method comprises the following steps: S1, collecting first interaction state data of a charging pile and a power grid in an SCADA system, synchronously collecting second interaction state data of the charging pile and the power grid in a PMU system, and obtaining target interaction state data; s2, a multi-input CNN-LSTM network model is constructed; and S3, based on the multi-input CNN-LSTM network model, according to the target interaction state data, monitoring the interaction state of the charging pile and the power grid, according to a monitoring result, determining whether an abnormal state occurs, and after determining that the abnormal state occurs, sending an early warning signal. Therefore, the interaction state data collected in the SCADA system and the PMU system are synchronously fused, and the interaction state of the charging station and the power grid is monitored according to the fused interaction state data based on the multi-input CNN-LSTM network model, so that the interaction state of the charging station and the power grid can be accurately and effectively monitored based on multi-dimensional data fusion.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of charging pile and power grid interaction monitoring, and particularly relates to a charging station and power grid interaction state monitoring method based on multi-dimensional data fusion. BACKGROUND

[0002] With the rapid popularization of new energy vehicles, as the key infrastructure connecting the power grid and users, the safety and stability of the operation state of the charging station are crucial to the reliability of the power grid and the user experience. However, in the actual operation process of the charging station, affected by multiple factors such as power grid fluctuations, load changes, equipment aging and external environmental interference, faults such as voltage abnormalities, frequency deviations and communication interruptions are prone to occur. If not discovered and warned in time, it may cause equipment damage, power grid disturbance and even safety accidents.

[0003] In related technologies, monitoring is usually dependent on a single data source or simple threshold, which is difficult to fully capture the complex dynamic characteristics in the interaction process of the charging station and the power grid. Therefore, the interaction state of the charging pile and the power grid cannot be accurately and effectively monitored. SUMMARY

[0004] To solve the above technical problems, the present application provides a charging station and power grid interaction state monitoring method based on multi-dimensional data fusion, which synchronously fuses the interaction state data collected in the SCADA (data acquisition and supervisory control system) system and the PMU (Phasor Measurement Unit, synchronous phasor measurement device) system, and monitors the interaction state of the charging station and the power grid based on the multi-input CNN-LSTM (Convolutional Neural Network-Long Short-Term Memory) network model according to the fused interaction state data, so that the interaction state of the charging station and the power grid can be accurately and effectively monitored based on multi-dimensional data fusion.

[0005] The technical scheme adopted by the present application is as follows: A charging station and power grid interaction state monitoring method based on multi-dimensional data fusion, comprising the following steps: S1, collecting first interaction state data of a charging pile and a power grid in a SCADA system, and synchronously collecting second interaction state data of the charging pile and the power grid in a PMU system, and performing data fusion processing on the first interaction state data and the second interaction state data to obtain target interaction state data; S2, constructing a multi-input CNN-LSTM network model; S3, monitoring the interaction state of the charging pile and the power grid based on the multi-input CNN-LSTM network model according to the target interaction state data, and determining whether an abnormal state occurs according to the monitoring result, and issuing a warning signal after determining that an abnormal state occurs.

[0006] In an embodiment of the present application, step S1 specifically comprises: S11, updating the first interaction state data and the second interaction state data at the same time based on a reference time point of simultaneous collection of the first interaction state data and the second interaction state data, and maintaining the collection period of the first interaction state data and the second interaction state data respectively until the next simultaneous update time point; S12, calculating the average value of the first interaction state data and the second interaction state data between the reference time point and the next simultaneous update time point, and obtaining the target interaction state data according to the average value of the first interaction state data and the average value of the second interaction state data.

[0007] In an embodiment of the present application, step S12 specifically comprises obtaining the target interaction state data by the following formula: , Wherein, represents the average value of the first interaction state data, represents the reference time point, represents the next simultaneous update time point corresponding to the reference time point, represents the first time point, represents the first time point, represents the second interaction state data collected at the first time point, represents the target interaction state data.

[0008] In an embodiment of the present application, the multi-input CNN-LSTM network model comprises an image encoding module, a CNN module, a full connection layer and an LSTM module connected in sequence, wherein the LSTM module comprises a forgetting gate, an input gate and an output gate.

[0009] In an embodiment of the present application, step S3 specifically comprises the following steps: S31, converting the target interaction state data into a corresponding RGB image based on the image encoding module; S32, extracting time sequence feature data in the RGB image based on the CNN module; and S33, processing the time sequence feature data based on the LSTM module to capture time sequence features of the interaction between the charging station and the power grid.

[0010] In an embodiment of the present application, the output of the LSTM module is represented by the following formula: , wherein, t represents the time step of the output gate, represents the output vector of the output gate at the time step, represents the weight matrix of the output gate, represents the bias vector of the output gate, represents the hidden state of the current time step, represents the hidden state of the previous time step, represents a normalization function, represents the memory cell state of the current time step, represents the input vector of the current time step, represents an activation function.

[0011] In an embodiment of the present application, the method for monitoring the interaction state between the charging station and the power grid based on multi-dimensional data fusion further comprises: supplementing the target interaction state data by combining the monitoring result with Spline interpolation to obtain supplemented interaction state data, and monitoring the interaction state between the charging pile and the power grid based on the multi-input CNN-LSTM network model according to the supplemented interaction state data.

[0012] A computer device comprises a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the computer program to implement the method for monitoring the interaction state between the charging station and the power grid based on multi-dimensional data fusion.

[0013] A non-transitory computer readable storage medium having a computer program stored thereon, the program being executable by a processor to implement the method for monitoring the interaction state between the charging station and the power grid based on multi-dimensional data fusion.

[0014] The present application has the following advantages: The application synchronously fuses the interactive state data collected in the SCADA system and the PMU system, and monitors the interactive state of the charging station and the power grid based on the multi-input CNN-LSTM network model according to the fused interactive state data, so that the interactive state of the charging station and the power grid can be accurately and effectively monitored based on multi-dimensional data fusion. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 A flowchart of the charging station and power grid interactive state monitoring method based on multi-dimensional data fusion of the embodiment of the application; Figure 2 A structural schematic diagram of the CNN module of one embodiment of the application. Figure 3 A structural schematic diagram of the LSTM module of one embodiment of the application. DETAILED DESCRIPTION

[0016] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the application.

[0017] Figure 1 A flowchart of the charging station and power grid interactive state monitoring method based on multi-dimensional data fusion of the embodiment of the application.

[0018] As shown in Figure 1 The charging station and power grid interactive state monitoring method based on multi-dimensional data fusion of the embodiment of the application can include the following steps: S1, collecting first interactive state data of the charging pile and the power grid in the SCADA system, and synchronously collecting second interactive state data of the charging pile and the power grid in the PMU system, and performing data fusion processing on the first interactive state data and the second interactive state data to obtain target interactive state data.

[0019] The remote terminal unit of the SCADA system stores real-time data used in charging station state estimation and dispatching processes, and the PMU system, as an important power system measurement means, can provide high-precision synchronous phasor measurement.

[0020] In one embodiment of the application, step S1 specifically includes the following steps: S11, updating the first interaction state data and the second interaction state data simultaneously based on a reference time point of a time point of simultaneous collection of the first interaction state data and the second interaction state data, and maintaining a collection period of the first interaction state data and the second interaction state data respectively until a next time point of simultaneous updating.

[0021] S12, calculating an average value of the first interaction state data and the second interaction state data between the reference time point and the next time point of simultaneous updating respectively, and obtaining target interaction state data according to the average value of the first interaction state data and the average value of the second interaction state data.

[0022] Specifically, in order to fuse the multi-dimensional data collected by the SCADA and PMU systems, that is, the first interaction state data and the second interaction state data, a time point of simultaneous collection of the first interaction state data and the second interaction state data is taken as a reference time point, the data of the two is updated simultaneously, and the collection period of the two is maintained until the next time point of simultaneous updating. The average value of the data measured between the two time points is processed to obtain a dynamic measurement result.

[0023] Specifically, step S12 specifically includes obtaining the target interaction state data by the following formula: , wherein, the average value of the first interaction state data is represented by, the reference time point is represented by, the next time point of simultaneous updating corresponding to the reference time point is represented by, the reference time point is represented by, the reference time point is represented by, the second interaction state data collected at the reference time point is represented by, the reference time point is represented by, the target interaction state data is represented by.

[0024] S2, constructing a multi-input CNN-LSTM network model.

[0025] In an embodiment of the present application, the multi-input CNN-LSTM network model includes an image encoding module, a CNN (Convolutional Neural Network) module, a full connection layer and an LSTM (Long Short-Term Memory) module connected in sequence, wherein the LSTM module includes a forgetting gate, an input gate and an output gate.

[0026] S3, monitoring the interaction state of the charging pile and the power grid according to the target interaction state data based on the multi-input CNN-LSTM network model, and judging whether an abnormal state occurs according to a monitoring result, and sending a warning signal after judging that an abnormal state occurs.

[0027] In an embodiment of the present application, step S3 specifically comprises the following steps: S31, converting the target interaction state data into a corresponding RGB image based on the image coding module.

[0028] S32, extracting time sequence feature data in the RGB image based on the CNN module.

[0029] Specifically, in order to enhance image feature extraction, a residual network is used as the CNN component of the multi-input CNN-LSTM network. As shown in Figure 2 The CNN module is composed of a convolution layer, a max-pooling layer, residual modules (residual module 1, residual module 2 and residual module 3) and a full connection layer. The working principle of each residual module is similar to that of a traditional CNN, but an identity mapping function is added, which allows the input signal to directly jump through several layers in the network and fuse with the output of the subsequent layer through a jump connection. This structure allows the network to maintain the input unchanged when necessary, thereby avoiding training difficulties caused by too many layers.

[0030] At the same time, in order to enhance the non-linear expression ability of the network and prevent overfitting, a BN (Batch Normalization) layer and a RELU (Rectified Linear Unit) activation function are added to the CNN module. The BN layer can improve the training and convergence speed of the network, reduce the possibility of overfitting, and alleviate gradient explosion and disappearance. The residual network used in the present application effectively deepens the CNN module by connecting the residual modules and solves the problem of gradient disappearance. In addition, due to the unique architecture of the CNN module, it has shown excellent performance in extracting frequency domain information from two-dimensional images.

[0031] S33, processing the time sequence feature data based on the LSTM module to capture the time sequence features of the interaction between the charging station and the power grid.

[0032] Specifically, the LSTM module has significant advantages in processing time sequence feature data, which can effectively capture time sequence features such as power grid voltage fluctuations and load changes, and improve the accuracy of data fusion. As shown in Figure 3As shown, the core structure of the LSTM module can include a forget gate, an input gate and an output gate, wherein the forget gate evaluates the retention degree of the memory unit at the last time; the input gate controls the update strength of the current input information, and generates a candidate memory value; the combination of the two updates the memory unit, realizes the dynamic fusion of historical information and current input. The output gate adjusts the information output after the nonlinear transformation of the memory unit, and forms the hidden state at the current time.

[0033] In an embodiment of the application, the output of the LSTM module is represented by the following formula: wherein, t represents the time step of the output gate, y t represents the output vector of the output gate at the time step, W o represents the weight matrix of the output gate, b o represents the bias vector of the output gate, h t represents the hidden state at the current time step, h t-1 represents the hidden state at the last time step, σ represents a normalization function, c t represents the memory unit state at the current time step, x t represents the input vector at the current time step, f represents an activation function.

[0034] Therefore, the combination of the CNN module and the LSTM module enables the multi-input CNN-LSTM network model to effectively distinguish the features of various charging station early warning events, thereby obtaining a high-precision classification result. In addition, the charging station early warning event is often a dynamic evolution process, and the information at a single time point is not enough to make an accurate judgment, which is the core problem solved by the LSTM module. After the CNN module extracts the static spatial features at each time step, the LSTM module, as the core of the network time series analysis, receives these feature sequences, and through its unique gating mechanism, can capture the dependence relationship and evolution law of the features in the time dimension, and learn the complete dynamic pattern of the event from occurrence, development to continuous change. Therefore, the spatial feature extraction capability of the CNN module and the time series modeling capability of the LSTM module are complementary to each other, so that the multi-input CNN-LSTM network model can fully distinguish the essential features of the charging station early warning event.

[0035] It should be noted that in an embodiment of the application, the charging station and power grid interaction state monitoring method based on multi-dimensional data fusion further comprises: combining the monitoring result with Spline interpolation to supplement the target interaction state data, to obtain supplemented interaction state data, and monitoring the interaction state of the charging pile and the power grid based on the multi-input CNN-LSTM network model according to the supplemented interaction state data. ​

[0036] Specifically, after obtaining the monitoring results using the LSTM module in the multi-input CNN-LSTM network model, the monitoring results can be applied to piecewise interpolation. The interpolation results are then used as pseudo-measurement data and fused with the actual measurement data to improve the accuracy and reliability of the measurement data.

[0037] Specifically, the interpolation process can be represented by the following formula: , In the formula: Indicates interpolation. , and Denote the coefficients of each term; Indicates the start time of interpolation. and These are the current time and the predicted mixed measurement vectors, respectively.

[0038] In summary, the charging station and power grid interaction status monitoring method based on multi-dimensional data fusion according to embodiments of the present invention collects first interaction status data between the charging pile and the power grid in the SCADA system and simultaneously collects second interaction status data between the charging pile and the power grid in the PMU system. The first and second interaction status data are then fused to obtain target interaction status data. A multi-input CNN-LSTM network model is constructed, and based on this model, the interaction status between the charging pile and the power grid is monitored according to the target interaction status data. The monitoring results are used to determine whether an abnormal state has occurred, and an early warning signal is issued if an abnormal state is detected. Therefore, by simultaneously fusing the interaction status data collected from the SCADA system and the PMU system, and using the multi-input CNN-LSTM network model to monitor the interaction status between the charging station and the power grid based on the fused interaction status data, multi-dimensional data fusion can accurately and effectively monitor the interaction status between the charging station and the power grid.

[0039] Corresponding to the above embodiments, the present invention also proposes a computer device.

[0040] The computer device in this embodiment of the invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the charging station and power grid interaction status monitoring method based on multi-dimensional data fusion as described in the above embodiment.

[0041] According to the computer device of the present invention, by synchronously fusing the interaction status data collected by the SCADA system and the PMU system, and based on the multi-input CNN-LSTM network model, the interaction status between the charging station and the power grid is monitored according to the fused interaction status data, thereby enabling accurate and effective monitoring of the interaction status between the charging station and the power grid based on multi-dimensional data fusion.

[0042] Corresponding to the above embodiments, the present invention also proposes a non-transitory computer-readable storage medium.

[0043] The non-transitory computer-readable storage medium of this invention stores a computer program that, when executed by a processor, implements the above-described method for monitoring the interaction status between charging stations and the power grid based on multi-dimensional data fusion.

[0044] According to embodiments of the present invention, a non-transitory computer-readable storage medium synchronously fuses interactive status data collected from SCADA and PMU systems, and monitors the interaction status between charging stations and the power grid based on the fused interactive status data using a multi-input CNN-LSTM network model. This enables accurate and effective monitoring of the interaction status between charging stations and the power grid based on multi-dimensional data fusion.

[0045] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. "A plurality of" means two or more, unless otherwise explicitly specified.

[0046] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0047] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "over," and "on top" of the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0048] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0049] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0050] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for monitoring the interaction status between charging stations and the power grid based on multi-dimensional data fusion, characterized in that, Includes the following steps: S1, collect the first interaction status data between the charging pile and the power grid in the SCADA system, and simultaneously collect the second interaction status data between the charging pile and the power grid in the PMU system, and perform data fusion processing on the first interaction status data and the second interaction status data to obtain the target interaction status data. S2, Construct a multi-input CNN-LSTM network model; S3, based on the multi-input CNN-LSTM network model, monitor the interaction status between the charging pile and the power grid according to the target interaction status data, determine whether an abnormal state has occurred based on the monitoring results, and issue an early warning signal after determining that an abnormal state has occurred.

2. The method for monitoring the interaction status between charging stations and power grids based on multi-dimensional data fusion according to claim 1, characterized in that, Step S1 specifically includes: S11, using the moment when the first interactive state data and the second interactive state data are collected simultaneously as the reference time, update the first interactive state data and the second interactive state data simultaneously, and maintain the collection period of the first interactive state data and the second interactive state data respectively until the next simultaneous update time. S12, calculate the average value of the first interaction state data and the second interaction state data between the reference time and the next simultaneous update time, and obtain the target interaction state data based on the average value of the first interaction state data and the average value of the second interaction state data.

3. The method for monitoring the interaction status between charging stations and power grids based on multi-dimensional data fusion according to claim 2, characterized in that, Step S12 specifically includes obtaining the target interaction state data through the following formula: ， in, This represents the average value of the first interaction state data. Indicates the reference time, This indicates the time of the next simultaneous update corresponding to the reference time. Indicates the first time, Indicates the first The second interactive state data collected at all times, This represents the target interaction state data.

4. The method for monitoring the interaction status between charging stations and power grids based on multi-dimensional data fusion according to claim 1, characterized in that, The multi-input CNN-LSTM network model includes an image encoding module, a CNN module, a fully connected layer, and an LSTM module connected in sequence. The LSTM module includes a forget gate, an input gate, and an output gate.

5. The method for monitoring the interaction status between charging stations and the power grid based on multi-dimensional data fusion according to claim 4, characterized in that, Step S3 specifically includes the following steps: S31, the target interactive state data is converted into a corresponding RGB image based on the image encoding module; S32, extract temporal feature data from the RGB image based on the CNN module; S33, The time-series feature data is processed based on the LSTM module to capture the time-series features of the interaction between the charging station and the power grid.

6. The method for monitoring the interaction status between charging stations and the power grid based on multi-dimensional data fusion according to claim 5, characterized in that, The output of the LSTM module is represented by the following formula: ， in, This represents the time step of the output gate. This represents the output vector of the output gate at time step 1. This represents the weight matrix of the output gate. This represents the bias vector of the output gate. This indicates the hidden state at the current time step. This indicates the hidden state of the previous time step. Represents a normalized function. This indicates the state of the memory cell at the current time step. This represents the input vector at the current time step. This represents the activation function.

7. The method for monitoring the interaction status between charging stations and power grids based on multi-dimensional data fusion according to any one of claims 1-6, characterized in that, Also includes: The monitoring results are combined with Spline interpolation to supplement the target interaction state data to obtain supplemented interaction state data. Based on the multi-input CNN-LSTM network model, the interaction state between the charging pile and the power grid is monitored according to the supplemented interaction state data.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the charging station and power grid interaction status monitoring method based on multi-dimensional data fusion according to any one of claims 1-7.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method for monitoring the interaction status of charging stations and power grids based on multi-dimensional data fusion as described in any one of claims 1-7.

Images