A power data fusion method and device

Abstract

The application relates to a power data fusion method, comprising the following steps: acquiring time series data from multiple data sources respectively; calculating the similarity between each time series data to obtain a similarity matrix; accumulating the matrix elements in the same row or the same column in the similarity matrix to obtain the similarity accumulation value of each data source and other data sources; performing longitudinal parameter fusion on each data source with a similarity accumulation value higher than a first threshold value, comprising the following steps: performing data preprocessing on the time series data; calculating the grey correlation degree between each two time series data of the time series data after the data preprocessing; filtering out the time series data with a grey correlation degree lower than a third threshold value; performing data fusion on the time series data by using a BP neural network to obtain a first fusion result; and performing horizontal parameter fusion on each data source with a similarity accumulation value lower than the first threshold value.

Description

Technical Field

[0001] This invention relates to a power data fusion method, belonging to the field of data fusion. Background Technology

[0002] The power grid system employs various types of data acquisition equipment in different power monitoring areas, while the power grid dispatch center, relay protection, and other departments independently construct their own databases and possess independent data models. These data are stored using different data models with varying types and encoding rules, resulting in inconsistencies in the operating parameters of the diverse power grids and an inability to share parameter information. Therefore, it is necessary to fuse power grid data to improve its quality.

[0003] The publication CN111950627A, titled "A Multi-Source Information Fusion Method and Its Application," discloses a method that starts from sensor data sources in environmental monitoring, standardizes and preprocesses the collected data, and introduces a support-corrected iterative fusion approach for multi-source isomorphic data. It compares the differences between the fused evidence and the original evidence to evaluate the support of the original evidence, and iterates multiple times until convergence to obtain the final fusion result. Summary of the Invention

[0004] To overcome the problems existing in the prior art, this invention designs a power data fusion method and device, which performs weighted horizontal data fusion and neural network-based vertical data fusion according to the similarity between data sources, thereby eliminating the differences between multi-source power grid data and optimizing the quality of power grid data.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: Technical Solution 1 A power data fusion method includes the following steps: Time series data are obtained from multiple data sources. Calculate the similarity between each time series data to obtain a similarity matrix, wherein the matrix elements are... This represents the similarity between time series data from data source i and time series data from data source j; by summing the matrix elements in the same row or column of the similarity matrix, the cumulative similarity value between each data source and other data sources is obtained. For each data source with a cumulative similarity value higher than the first threshold, longitudinal parameter fusion is performed, including: data preprocessing of time series data; calculation of grey relational degree between pairs of time series data for the preprocessed time series data; filtering out time series data with grey relational degree lower than the third threshold; and data fusion of time series data using a BP neural network to obtain the first fusion result. For data sources with a cumulative similarity value below the first threshold, horizontal parameter fusion is performed.

[0006] Furthermore, the similarity calculation formula is as follows:

[0007] In the formula, This represents time-series data from data source i; This represents time-series data from data source j.

[0008] Furthermore, the lateral parameter fusion includes: A weight is assigned to each time series data; the time series data are then weighted and summed according to the weights to obtain the second fusion result.

[0009] Furthermore, it also includes: filtering out the corresponding time series data for each data source whose cumulative similarity value is lower than the second threshold.

[0010] Furthermore, the parameters of the BP neural network are determined using an improved firefly algorithm: S1. Map the parameters to individuals in the population; initialize individual positions, set the maximum attraction, light intensity absorption coefficient, step size factor, search precision, and maximum number of iterations; S2. Calculate the relative brightness and attractiveness of each individual firefly. Determine the direction of movement based on the relative brightness and the distance traveled based on the attractiveness. This can be expressed by the following formula:

[0011] In the formula, γ is the light intensity absorption coefficient; r ij βi represents the spatial distance between fireflies i and j; β0 represents the maximum attraction. S3. Update the individual position, expressed by the formula:

[0012] In the formula, x i α represents the individual's position; α is the step size factor; rand is a random factor that follows a uniform distribution on [0, 1]. S4. Update individual brightness, expressed by the formula:

[0013] In the formula, I0 is the maximum fluorescence brightness of the firefly; γ is the light intensity absorption coefficient; If the search precision is met or the maximum number of searches is reached, proceed to the next step; otherwise, increase the number of searches by 1 and jump to step S3. S5. The local optimal solution is the weights and threshold of the BP neural network. S6. Determine if the termination condition is met; if the number of iterations reaches the preset maximum value of the given BP neural network and meets the minimum error requirement, the network outputs the optimal solution and stops counting; otherwise, proceed to step S2.

[0014] Technical Solution Two A power data fusion device includes a processor and a memory for storing executable instructions of the processor; the processor is configured to read the executable instructions from the memory and execute the instructions to implement the steps described in technical solution one.

[0015] Compared with the prior art, the present invention has the following features and beneficial effects: This invention performs weighted horizontal data fusion and neural network-based vertical data fusion based on the similarity between data sources to eliminate differences in multi-source power grid data and optimize power grid data quality. Furthermore, this invention utilizes an improved firefly algorithm to determine BP neural network parameters and adds a perturbation α(rand-1 / 2) during the position update process to increase the search area and avoid prematurely getting trapped in local optima. Attached Figure Description

[0016] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0017] The present invention will now be described in more detail with reference to the embodiments.

[0018] Example 1 like Figure 1 As shown, a power data fusion method includes the following steps: Time series data are obtained from multiple data sources. Calculate the similarity between each time series data point to obtain a similarity matrix. The matrix elements in the similarity matrix are... This represents the similarity between time series data from data source i and time series data from data source j. The similarity calculation formula is as follows: In the formula, This represents time-series data from data source i; This represents time-series data from data source j.

[0019] By summing the matrix elements in the same row or column of the similarity matrix, the cumulative similarity value between each data source and other data sources can be obtained.

[0020] Data sources with cumulative similarity values ​​exceeding a first threshold are considered to be the same professional departments of power grid dispatch centers at different levels. To reduce parameter differences, vertical parameter fusion is performed to obtain the first fusion result.

[0021] Data sources with similarity cumulative values ​​below the first threshold are considered as different professional departments of the same level dispatch center. Horizontal parameter fusion is performed on them to obtain the second fusion result.

[0022] Example 2 This embodiment further provides a vertical parameter fusion method, including the following steps: For each data source with a cumulative similarity value below the second threshold, its time series data is filtered out.

[0023] Perform data preprocessing on the remaining time series data, such as normalization and noise reduction.

[0024] For time series data that have undergone preprocessing, calculate the grey correlation degree between each pair of time series data; filter out time series data with a grey correlation degree lower than the third threshold.

[0025] The time series data were fused as follows: Construct and train a data fusion model based on a backpropagation (BP) neural network. The BP neural network consists of an input layer, an output layer, and hidden layers. It takes time series data as input, processes the time series data using the activation function of the hidden layer, and outputs the data fusion result from the output layer.

[0026] Example 3 This embodiment further provides a method for lateral parameter fusion, including the following steps: The credibility of each data source is assessed, and a weight is assigned to each time series data. The time series data are then weighted and summed according to the weights to obtain the fusion result.

[0027] Example 4 Determining BP neural network parameters by improving the firefly algorithm: S1. Map the parameters to individuals in the population; randomly initialize individual positions, set the maximum attraction, light intensity absorption coefficient, step size factor, search precision, and maximum number of iterations.

[0028] S2. Calculate the relative brightness and attractiveness of each individual firefly. Determine the direction of movement based on the relative brightness and the distance traveled based on the attractiveness. This can be expressed by the following formula:

[0029] In the formula, γ is the light intensity absorption coefficient. Since fluorescence gradually weakens with increasing distance and absorption by the propagation medium, a light intensity absorption coefficient is set to reflect this characteristic; r ij Let β be the spatial distance between fireflies i and j. β0 is the maximum attraction, i.e., the attraction at the light source (r=0).

[0030] S3. Update the individual position, expressed by the formula:

[0031] In the formula, x i α is the individual position; α is the step size factor; rand is a random factor that follows a uniform distribution on [0, 1].

[0032] S4. Update individual brightness, expressed by the formula:

[0033] In the formula, I0 is the maximum fluorescence brightness of the firefly, that is, its own fluorescence brightness (at r=0); γ ​​is the light intensity absorption coefficient.

[0034] If the search precision is met or the maximum number of searches is reached, continue to the next step; otherwise, increase the number of searches by 1 and jump to step S3.

[0035] S5. Output the global extreme point and the optimal individual value. Use the global optimal solution of the output function as the weights and threshold of the BP neural network. S6. Determine if the termination condition is met. If the number of iterations reaches the preset maximum value of the given BP neural network and meets the minimum error requirement, the network outputs the optimal solution and stops counting; otherwise, proceed to step S2.

[0036] It should be noted that the aforementioned longitudinal parameter fusion device is also used to achieve the above-mentioned... Figure 1 The method steps corresponding to each embodiment of the power data fusion method shown are not repeated here.

[0037] It should be noted that the functional units / modules in the various embodiments of the present invention can be integrated into one processing unit / module, or each unit / module can exist physically separately, or two or more units / modules can be integrated into one unit / module. The integrated unit / module described above can be implemented in hardware or in the form of software functional units / modules.

[0038] From the above description of the embodiments, those skilled in the art will clearly understand that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code, or any suitable combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: Application-Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Digital Signal Processing Device (DSPD), Programmable Logic Device (PLD), Field-Programmable Gate Array (FPGA), processor, controller, microcontroller, microprocessor, other electronic units designed to implement the functions described herein, or combinations thereof. For software implementation, some or all of the processes of the embodiments can be implemented by a computer program instructing the associated hardware. During implementation, the program can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any medium that facilitates the transmission of a computer program from one place to another. Storage media can be any available medium accessible to a computer. Computer-readable media can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code having the form of instructions or data structures and accessible to a computer.

[0039] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should be able to analyze that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A power data fusion method, characterized in that, Includes the following steps: Time series data are obtained from multiple data sources. Calculate the similarity between each time series data to obtain a similarity matrix, wherein the matrix elements are... This represents the similarity between time series data from data source i and time series data from data source j; by summing the matrix elements in the same row or column of the similarity matrix, the cumulative similarity value between each data source and other data sources is obtained. For each data source with a cumulative similarity value higher than the first threshold, longitudinal parameter fusion is performed, including: data preprocessing of time series data; calculation of grey relational degree between pairs of time series data for the preprocessed time series data; filtering out time series data with grey relational degree lower than the third threshold; and data fusion of time series data using a BP neural network to obtain the first fusion result. For data sources with a cumulative similarity value below a first threshold, horizontal parameter fusion is performed. The horizontal parameter fusion includes: evaluating the credibility of each data source, setting a weight for each time series data, and performing a weighted summation of the time series data according to the weight to obtain the fusion result.

2. The power data fusion method according to claim 1, characterized in that, The similarity calculation formula is as follows: In the formula, This represents time-series data from data source i; This represents time-series data from data source j.

3. The power data fusion method according to claim 1, characterized in that, Also includes: For each data source with a cumulative similarity value lower than the second threshold, the corresponding time series data is filtered out.

4. The power data fusion method according to claim 1, characterized in that, Determining BP neural network parameters using an improved firefly algorithm: S1. Map the parameters to individuals in the population; initialize individual positions, set the maximum attraction, light intensity absorption coefficient, step size factor, search precision, and maximum number of iterations; S2. Calculate the relative brightness and attractiveness of each individual firefly. Determine the direction of movement based on the relative brightness and the distance traveled based on the attractiveness. This can be expressed by the following formula: In the formula, γ is the light intensity absorption coefficient; r ij βi represents the spatial distance between fireflies i and j; β0 represents the maximum attraction. S3. Update the individual position, expressed by the formula: In the formula, x i For individual location, x i ’ To update the individual position; α is the step size factor; rand is a random factor that follows a uniform distribution on [0, 1]; S4. Update individual brightness, expressed by the formula: In the formula, I0 is the maximum fluorescence brightness of the firefly; γ is the light intensity absorption coefficient; If the search precision is met or the maximum number of searches is reached, proceed to the next step; otherwise, increase the number of searches by 1 and jump to step S3. S5. The local optimal solution is the weights and threshold of the BP neural network. S6. Determine if the termination condition is met; if the number of iterations reaches the preset maximum value of the given BP neural network and meets the minimum error requirement, the network outputs the optimal solution and stops counting; otherwise, proceed to step S2.

5. A power data fusion device, characterized in that, The system includes a processor and a memory for storing executable instructions; the processor is configured to read the executable instructions from the memory and execute the instructions to perform the following steps: Time series data are obtained from multiple data sources. Calculate the similarity between each time series data to obtain a similarity matrix, wherein the matrix elements are... This represents the similarity between time series data from data source i and time series data from data source j; by summing the matrix elements in the same row or column of the similarity matrix, the cumulative similarity value between each data source and other data sources is obtained. For each data source with a cumulative similarity value higher than the first threshold, longitudinal parameter fusion is performed, including: data preprocessing of time series data; calculation of grey relational degree between pairs of time series data for the preprocessed time series data; filtering out time series data with grey relational degree lower than the third threshold; and data fusion of time series data using a BP neural network to obtain the first fusion result. For data sources with a cumulative similarity value below a first threshold, horizontal parameter fusion is performed. The horizontal parameter fusion includes: evaluating the credibility of each data source, setting a weight for each time series data, and performing a weighted summation of the time series data according to the weight to obtain the fusion result.

6. The power data fusion device according to claim 5, characterized in that, The similarity calculation formula is as follows: ； In the formula, This represents time-series data from data source i; This represents time-series data from data source j.

7. A power data fusion device according to claim 5, characterized in that, Also includes: For each data source with a cumulative similarity value lower than the second threshold, the corresponding time series data is filtered out.

8. A power data fusion device according to claim 5, characterized in that, Determining BP neural network parameters using an improved firefly algorithm: S1. Map the parameters to individuals in the population; initialize individual positions, set the maximum attraction, light intensity absorption coefficient, step size factor, search precision, and maximum number of iterations; S2. Calculate the relative brightness and attractiveness of each individual firefly. Determine the direction of movement based on the relative brightness and the distance traveled based on the attractiveness. This can be expressed by the following formula: In the formula, γ is the light intensity absorption coefficient; r ij βi represents the spatial distance between fireflies i and j; β0 represents the maximum attraction. S3. Update the individual position, expressed by the formula: In the formula, x i For individual location, x i ’ To update the individual position; α is the step size factor; rand is a random factor that follows a uniform distribution on [0, 1]; S4. Update individual brightness, expressed by the formula: In the formula, I0 is the maximum fluorescence brightness of the firefly; γ is the light intensity absorption coefficient; If the search precision is met or the maximum number of searches is reached, proceed to the next step; otherwise, increase the number of searches by 1 and jump to step S3. S5. The local optimal solution is the weights and threshold of the BP neural network. S6. Determine if the termination condition is met; if the number of iterations reaches the preset maximum value of the given BP neural network and meets the minimum error requirement, the network outputs the optimal solution and stops counting; otherwise, proceed to step S2.

Images