Power distribution network investment method based on multi-source heterogeneous data mode
By establishing a centralized data platform and preprocessing technology in distribution network investment, extracting frequency domain features and value feedback index, and combining prediction models, the problems of data integration and decision-making feature extraction in the existing technology are solved, and the distribution network is realized Efficient decision-making support for investment and adaptation to market changes.
Patent Information
- Application Number
- CN202411220262.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-02
- Publication Date
- 2024-12-06
AI Technical Summary
The existing distribution network investment method based on multi-source heterogeneous data model is difficult to efficiently integrate and manage massive heterogeneous data, extract important characteristics for investment decision-making, and build an accurate decision-making model to ensure the timeliness and accuracy of decision-making suggestions. and reliability, it is impossible to deeply explore the implicit information in the data to adapt to the rapidly changing grid operating conditions and market demand.
Establish a centralized data platform to automatically collect and integrate real-time monitoring data, history, meteorological information, equipment maintenance logs and market operation data, extract frequency domain characteristics and value feedback index through pre-processing and Fourier calculations, and combine preset Prediction model predicts grid load and resource requirements, optimizes the model to generate future investment decision characteristics, and evaluates the scope of impact.
It realizes efficient data integration and feature extraction, provides comprehensive analysis and evaluation of distribution network investment decisions, improves the timeliness and accuracy of decisions, adapts to market changes, and supports smarter grid planning and operations.
Smart Images

Figure CN119090547A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of multi-source data analysis of distribution networks, and more specifically, to a method for distribution network investment based on a multi-source heterogeneous data model. Background Art
[0002] In the field of multi-source data analysis of distribution networks, it is of great significance and importance to study the distribution network investment method based on multi-source heterogeneous data mode. This method can integrate data from different sources and types such as power load, meteorological conditions, equipment operation status and market dynamics, and provide decision support for accurate investment in distribution networks through advanced data preprocessing, feature extraction and decision model construction. This multi-dimensional and in-depth analysis method can effectively improve the intelligence level of distribution network planning and operation, reduce operating costs, improve system stability and reliability, and promote the maximization of economic and social benefits of distribution networks. It plays an indispensable role in supporting the sustainable development of modern society.
[0003] Prior to the present invention, existing distribution network investment methods based on multi-source heterogeneous data models mainly rely on traditional data processing technology and simple decision support systems. These methods usually involve manually integrating data from different sources and formats, such as power load data, equipment operation data, meteorological information, etc., and then assisting investment decisions through basic data analysis methods (such as statistical analysis). This approach is not only time-consuming and labor-intensive, but also difficult to handle the growing amount and complexity of data, unable to deeply mine the implicit information in the data, and difficult to adapt to the rapidly changing power grid operating conditions and market demands. The technical difficulties and key points mainly include: 1) How to efficiently integrate and manage massive heterogeneous data from different data sources; 2) How to extract features that have an important impact on distribution network investment decisions from complex data; 3) How to build a decision model that can accurately reflect the operating characteristics of the distribution network and market changes; 4) How to ensure the timeliness, accuracy and reliability of decision recommendations; and 5) How to evaluate the comprehensive benefits of investment decisions, including economic benefits, system stability, and environmental impact. Summary of the invention
[0004] In view of the above problems, the present invention proposes a method for distribution network investment based on a multi-source heterogeneous data model. By quickly integrating and evaluating the distribution network investment decision-making characteristics and impact scope, it can efficiently integrate and manage massive heterogeneous data from different data sources and realize the output of different types of impacts.
[0005] According to a first aspect of an embodiment of the present invention, a method for distribution network investment based on a multi-source heterogeneous data mode is provided.
[0006] In one or more embodiments, preferably, the method for distribution network investment based on a multi-source heterogeneous data mode includes:
[0007] Establish a centralized data platform to automatically collect and integrate real-time monitoring data, historical records, weather information, equipment maintenance logs, and market operation data, and store them on the data platform;
[0008] Obtain the integrated data and extract the distribution network investment decision-making features;
[0009] Combined with preset forecasting models to predict future grid load and resource requirements;
[0010] Optimizing the prediction model using historical data to form a trained prediction model;
[0011] Input real-time data into the trained prediction model to generate features for future distribution network investment decisions;
[0012] The scope of the assessment impact is determined based on the characteristics of distribution network investment decisions.
[0013] In one or more embodiments, preferably, the establishment of a centralized data platform automatically collects and integrates real-time monitoring data, historical records, meteorological information, equipment maintenance logs, and market operation data, and stores them on the data platform, specifically including:
[0014] Design a centralized data platform to support the integration and storage of real-time monitoring data, historical records, weather information, equipment maintenance logs, and market operation data;
[0015] Automatically collect data, and conduct real-time data monitoring, equipment maintenance, and operational data collection through various communication protocols.
[0016] In one or more embodiments, preferably, the acquiring of the integrated data and extracting the distribution network investment decision-making features therein specifically includes:
[0017] Perform preprocessing operations on raw data collected from multiple data sources, including removing outliers and noise, and standardizing data to eliminate the impact of different dimensions;
[0018] The frequency domain features are obtained by Fourier calculation on the preprocessed data;
[0019] Calculate the rate of change of the preprocessed data;
[0020] Extract frequency domain features, and use the first calculation formula to calculate the correlation index between the frequency domain features and the value feedback index;
[0021] Extract the data change rate and use the second calculation formula to calculate the value feedback index correlation;
[0022] When the correlation index between the frequency domain feature and the value feedback index satisfies the third calculation formula, the corresponding frequency domain feature of this feature is retained as the distribution network investment decision feature;
[0023] When the value feedback index correlation satisfies the fourth calculation formula, this feature is retained in the distribution network investment decision feature;
[0024] The first calculation formula is:
[0025] G=(HJ)÷(TD-TX)
[0026] Among them, G is the correlation index between the frequency domain feature and the value feedback index, H is the frequency domain feature value corresponding to the maximum value of the value feedback index, J is the frequency domain feature value corresponding to the average value of the value feedback index, TD is the maximum value of the frequency domain feature value, and TX is the minimum value of the frequency domain feature value;
[0027] The second calculation formula is:
[0028] P=ZS
[0029] Among them, P is the correlation of the value feedback index, Z is the normalized index of the change rate of the value feedback index, and S is the normalized index of the data change rate;
[0030] The third calculation formula is:
[0031] G>y1
[0032] Wherein, y1 is the first contrast margin;
[0033] The fourth calculation formula is:
[0034] P <y2
[0035] Wherein, y2 is the first contrast margin.
[0036] In one or more embodiments, preferably, the predicting future power grid load and resource demand by combining a preset prediction model specifically includes:
[0037] Select a preset model suitable for grid load and resource demand forecasting;
[0038] Preprocess the collected real-time monitoring data, historical records, meteorological information, equipment maintenance logs, and market operation data to extract and construct features useful for the prediction model;
[0039] The pre-processed data is used to train the preset prediction model;
[0040] Integrate the prediction model into the data platform to achieve continuous processing of real-time data streams and load forecasting.
[0041] In one or more embodiments, preferably, the using historical data to optimize the prediction model to form a trained prediction model specifically includes:
[0042] Use historical data to perform forecast bias analysis to identify patterns and sources of error;
[0043] According to the result of the error analysis, the prediction model is adjusted to a mode with the smallest prediction deviation for the historical data as the trained prediction model.
[0044] In one or more embodiments, preferably, the step of inputting the real-time data into the trained prediction model to generate future distribution network investment decision features specifically includes:
[0045] Fusion and synchronization based on real-time data to ensure the timeliness and consistency of data and form input data for predictive analysis;
[0046] Based on the predictive analysis input data, changes in key features related to distribution network investment decisions are extracted from the output of the trained predictive model.
[0047] In one or more embodiments, preferably, judging the scope of the evaluation impact according to the distribution network investment decision characteristics specifically includes:
[0048] For each data parameter, a preset impact ratio of economic benefit, system stability and social impact is set, and the sum of the impact ratios of economic benefit, system stability and social impact corresponding to each parameter is 1;
[0049] According to the changes in the characteristics of future distribution network investment decisions, multiply them by the preset impact ratios of the corresponding economic benefits, system stability and social impacts to obtain the main impact of the corresponding investment.
[0050] According to a second aspect of an embodiment of the present invention, a system for distribution network investment based on a multi-source heterogeneous data mode is provided.
[0051] In one or more embodiments, preferably, the system for distribution network investment based on a multi-source heterogeneous data mode includes:
[0052] The data collection and integration module is used to establish a centralized data platform to automatically collect and integrate real-time monitoring data, historical records, meteorological information, equipment maintenance logs and market operation data, and store them on the data platform;
[0053] Extract key feature module, which is used to obtain the integrated data and extract the distribution network investment decision-making features;
[0054] A preliminary forecasting module is used to forecast future grid load and resource requirements in combination with a preset forecasting model;
[0055] A prediction correction module, used to optimize the prediction model using historical data to form a trained prediction model;
[0056] A feature correction module is used to input real-time data into the trained prediction model to generate features for future distribution network investment decisions;
[0057] The impact display module is used to determine the scope of the assessment impact based on the characteristics of distribution network investment decisions.
[0058] According to a third aspect of an embodiment of the present invention, there is provided a computer-readable storage medium on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method as described in any one of the first aspect of the embodiment of the present invention is implemented.
[0059] According to a fourth aspect of an embodiment of the present invention, there is provided an electronic device, comprising a memory and a processor, wherein the memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement any one of the methods described in the first aspect of the embodiment of the present invention.
[0060] The technical solution provided by the embodiments of the present invention may have the following beneficial effects:
[0061] In the solution of the present invention, data from different sources and in different formats are unified together to create a comprehensive data view, thereby achieving efficient distribution data fusion analysis.
[0062] In the solution of the present invention, the key data features that affect the distribution network investment decision are automatically identified and extracted, and the impact range is analyzed to determine the impact range.
[0063] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.
[0064] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0065] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.
[0066] Figure 1 It is a flow chart of a method for distribution network investment based on a multi-source heterogeneous data mode according to an embodiment of the present invention.
[0067] Figure 2 It is a flowchart of an embodiment of the present invention, in which a method for distribution network investment based on a multi-source heterogeneous data model is used to establish a centralized data platform, automatically collect and integrate real-time monitoring data, historical records, meteorological information, equipment maintenance logs and market operation data, and store them on the data platform.
[0068] Figure 3 It is a flowchart of acquiring integrated data and extracting distribution network investment decision features in a distribution network investment method based on a multi-source heterogeneous data mode in one embodiment of the present invention.
[0069] Figure 4 It is a flowchart of a method for distribution network investment based on a multi-source heterogeneous data mode in accordance with an embodiment of the present invention, which combines a preset prediction model to predict future power grid load and resource demand.
[0070] Figure 5 It is a flowchart of using historical data to optimize the prediction model in a method for distribution network investment based on a multi-source heterogeneous data mode in an embodiment of the present invention to form a trained prediction model.
[0071] Figure 6 It is a flowchart of inputting real-time data into a trained prediction model to generate future distribution network investment decision features in a method for distribution network investment based on a multi-source heterogeneous data mode in an embodiment of the present invention.
[0072] Figure 7 It is a flowchart of judging the scope of evaluation influence according to the distribution network investment decision-making characteristics in a distribution network investment method based on a multi-source heterogeneous data mode in one embodiment of the present invention.
[0073] Figure 8 It is a structural diagram of a system for distribution network investment based on a multi-source heterogeneous data mode according to an embodiment of the present invention.
[0074] Fig. 9 It is a structural diagram of an electronic device in one embodiment of the present invention. DETAILED DESCRIPTION
[0075] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.
[0076] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.
[0077] In the field of multi-source data analysis of distribution networks, it is of great significance and importance to study the distribution network investment method based on multi-source heterogeneous data mode. This method can integrate data from different sources and types such as power load, meteorological conditions, equipment operation status and market dynamics, and provide decision support for accurate investment in distribution networks through advanced data preprocessing, feature extraction and decision model construction. This multi-dimensional and in-depth analysis method can effectively improve the intelligence level of distribution network planning and operation, reduce operating costs, improve system stability and reliability, and promote the maximization of economic and social benefits of distribution networks. It plays an indispensable role in supporting the sustainable development of modern society.
[0078] Prior to the present invention, existing distribution network investment methods based on multi-source heterogeneous data models mainly rely on traditional data processing technology and simple decision support systems. These methods usually involve manually integrating data from different sources and formats, such as power load data, equipment operation data, meteorological information, etc., and then assisting investment decisions through basic data analysis methods (such as statistical analysis). This approach is not only time-consuming and labor-intensive, but also difficult to handle the growing amount and complexity of data, unable to deeply mine the implicit information in the data, and difficult to adapt to the rapidly changing power grid operating conditions and market demands. The technical difficulties and key points mainly include: 1) How to efficiently integrate and manage massive heterogeneous data from different data sources; 2) How to extract features that have an important impact on distribution network investment decisions from complex data; 3) How to build a decision model that can accurately reflect the operating characteristics of the distribution network and market changes; 4) How to ensure the timeliness, accuracy and reliability of decision recommendations; and 5) How to evaluate the comprehensive benefits of investment decisions, including economic benefits, system stability, and environmental impact.
[0079] In an embodiment of the present invention, a method for distribution network investment based on a multi-source heterogeneous data model is provided. The solution can efficiently integrate and manage massive heterogeneous data from different data sources and output different types of impacts by quickly integrating and evaluating the characteristics and impact range of distribution network investment decisions.
[0080] According to a first aspect of an embodiment of the present invention, a method for distribution network investment based on a multi-source heterogeneous data mode is provided.
[0081] Figure 1 It is a flow chart of a method for distribution network investment based on a multi-source heterogeneous data mode according to an embodiment of the present invention.
[0082] In one or more embodiments, preferably, the method for distribution network investment based on a multi-source heterogeneous data mode includes:
[0083] S101. Establish a centralized data platform to automatically collect and integrate real-time monitoring data, historical records, weather information, equipment maintenance logs and market operation data, and store them on the data platform;
[0084] S102, obtaining the integrated data and extracting the distribution network investment decision-making features therein;
[0085] S103, predicting future power grid load and resource demand in combination with a preset prediction model;
[0086] S104, optimizing the prediction model using historical data to form a trained prediction model;
[0087] S105, inputting the real-time data into the trained prediction model to generate future distribution network investment decision features;
[0088] S106. Determine the scope of the assessment impact based on the characteristics of the distribution network investment decision.
[0089] In the embodiment of the present invention, the core performs the following six steps: collecting and integrating data, extracting key features, preliminary prediction, prediction correction, feature correction and impact display. Finally, the analysis, evaluation and display of the investment decision of the distribution network are realized through the above steps.
[0090] Figure 2 It is a flowchart of an embodiment of the present invention, in which a method for distribution network investment based on a multi-source heterogeneous data model is used to establish a centralized data platform, automatically collect and integrate real-time monitoring data, historical records, meteorological information, equipment maintenance logs and market operation data, and store them on the data platform.
[0091] like Figure 2 As shown, in one or more embodiments, preferably, a centralized data platform is established to automatically collect and integrate real-time monitoring data, historical records, meteorological information, equipment maintenance logs and market operation data, and store them on the data platform, specifically including:
[0092] S201. Design a centralized data platform to support the integration and storage of real-time monitoring data, historical records, meteorological information, equipment maintenance logs, and market operation data;
[0093] S202: Automatically collect data and conduct real-time data monitoring, equipment maintenance and operation data collection through various communication protocols.
[0094] In an embodiment of the present invention, a data platform based on cloud computing technology is constructed, which has high availability and scalability and can process and store large amounts of data. The data platform will adopt a distributed database system to ensure the consistency and integrity of the data. An automated data collection system is deployed, which can be seamlessly connected to the existing monitoring system, maintenance management system, and market operation system through various communication protocols. An advanced data processing and analysis module is integrated into the data platform, which uses tools and data cleaning algorithms to automatically identify and correct errors and inconsistencies in the data. At the same time, data mining and machine learning algorithms are applied to extract useful information from the data to provide support for predictive maintenance and market analysis. An intuitive user interface is created to allow users to access the data platform according to their permissions for data query, analysis, and report generation.
[0095] Figure 3It is a flowchart of acquiring integrated data and extracting distribution network investment decision features in a distribution network investment method based on a multi-source heterogeneous data mode in one embodiment of the present invention.
[0096] like Figure 3 As shown, in one or more embodiments, preferably, the acquiring of the integrated data and extracting the distribution network investment decision-making features therein specifically includes:
[0097] S301, performing preprocessing operations on raw data collected from multiple data sources, the process of which includes removing outliers and noise, and standardizing data to eliminate the influence of different dimensions;
[0098] S302, obtaining frequency domain features by Fourier calculation on the preprocessed data;
[0099] S303, calculating the rate of change of the preprocessed data;
[0100] S304, extracting frequency domain features, and calculating the correlation index between the frequency domain features and the value feedback index using the first calculation formula;
[0101] S305, extracting the data change rate, and calculating the value feedback index correlation using the second calculation formula;
[0102] S306. When the correlation index between the frequency domain feature and the value feedback index satisfies the third calculation formula, retain the corresponding frequency domain feature of this feature as the distribution network investment decision feature;
[0103] S307. When the value feedback index correlation satisfies the fourth calculation formula, retain this feature in the distribution network investment decision feature;
[0104] The first calculation formula is:
[0105] G=(HJ)÷(TD-TX)
[0106] Among them, G is the correlation index between the frequency domain feature and the value feedback index, H is the frequency domain feature value corresponding to the maximum value of the value feedback index, J is the frequency domain feature value corresponding to the average value of the value feedback index, TD is the maximum value of the frequency domain feature value, and TX is the minimum value of the frequency domain feature value;
[0107] The second calculation formula is:
[0108] P=ZS
[0109] Among them, P is the correlation of the value feedback index, Z is the normalized index of the change rate of the value feedback index, and S is the normalized index of the data change rate;
[0110] The third calculation formula is:
[0111] G > y1
[0112] where y1 is the first comparison margin;
[0113] The fourth calculation formula is:
[0114] P < y2
[0115] where y2 is the first comparison margin.
[0116] In an embodiment of the present invention, raw data from real-time monitoring, historical records, meteorological information, equipment maintenance logs, and market operations is collected. Preprocessing operations are performed on this data, including removing outliers and noise, and data standardization to eliminate the influence of different dimensions. For example, for voltage time series data, obvious abnormal peaks are removed, and the remaining data is processed using the z-score standardization method. Fourier transform is performed on the preprocessed data to obtain frequency domain features. These frequency domain features reveal the periodic change characteristics in the data, such as the day-night fluctuation pattern of the daily load curve. The change rate of the preprocessed data is calculated, which helps to capture the dynamic characteristics of the data, such as the rapid increase or decrease trend of the load. The correlation index G between the frequency domain features and the value feedback index is calculated using the first calculation formula G = (H - J) ÷ (TD - TX). Here, H is the frequency domain feature value corresponding to the maximum value of the value feedback index, J is the frequency domain feature value corresponding to the average value of the value feedback index, TD is the maximum value of the frequency domain feature value, and TX is the minimum value of the frequency domain feature value. This step aims to evaluate the influence degree of the frequency domain features on the value feedback index. The value feedback index correlation degree P is calculated using the second calculation formula P = Z - S. Here, Z is the normalized index of the change rate of the value feedback index, and S is the normalized index of the data change rate. This step is used to measure the correlation between the value feedback index and the data change rate. The third calculation formula G > y1 and the fourth calculation formula P < y2 are set as the conditions for feature screening. Where y1 and y2 are the preset first comparison margin and second comparison margin respectively, used to determine which features have a significant impact on the distribution network investment decision. Investment decision feature extraction: When the correlation index G between the frequency domain features and the value feedback index satisfies G > y1, the corresponding frequency domain features of this feature are retained as the distribution network investment decision features. At the same time, when the value feedback index correlation degree P satisfies P < y2, this feature is also retained in the distribution network investment decision features. Through the above steps, features that have a significant impact on the distribution network investment decision can be extracted from the integrated data, providing valuable information for decision-makers.
[0117] Figure 4 It is a flowchart of predicting future grid load and resource demand by combining a preset prediction model in a method for distribution network investment based on a multi-source heterogeneous data model in an embodiment of the present invention.
[0118] like Figure 4 As shown, in one or more embodiments, preferably, the combining of a preset prediction model to predict future power grid load and resource demand specifically includes:
[0119] S401, selecting a preset model suitable for grid load and resource demand forecasting;
[0120] S402, pre-processing the collected real-time monitoring data, historical records, meteorological information, equipment maintenance logs, and market operation data to extract and construct features useful for the prediction model;
[0121] S403, training a preset prediction model using the preprocessed data;
[0122] S404. Integrate the prediction model into the data platform to achieve continuous processing of real-time data streams and load prediction.
[0123] In an embodiment of the present invention, a preset model suitable for grid load and resource demand forecasting is selected, such as a neural network or a linear regression model, and the model is customized according to the specific characteristics and operating data of the grid. For example, for a grid with significant seasonal load changes, a recurrent neural network model that can capture seasonal patterns is selected. The collected real-time monitoring data, historical records, meteorological information, equipment maintenance logs, and market operation data are preprocessed, including data cleaning, missing value processing, outlier detection and correction. Then, features useful for the prediction model are extracted and constructed, such as time series features of historical load data, meteorological conditions, calendar events, equipment operating status, etc. The preset prediction model is trained using the preprocessed data. The prediction accuracy of the model is evaluated by using methods such as cross-validation, and the performance is optimized by adjusting the model parameters. For example, the optimal number of nodes and learning rate of the hidden layer of the neural network are found by the grid search method. The trained prediction model is integrated into the data platform to realize continuous processing and load forecasting of real-time data streams. The model regularly receives new real-time data and dynamically adjusts the prediction results to ensure the accuracy and timeliness of the prediction. A model performance monitoring system is established to regularly evaluate the consistency of the prediction results with the actual grid load. By analyzing prediction errors and patterns, feedback is continuously fed back into the model training process to improve the accuracy and reliability of predictions.
[0124] Figure 5 It is a flowchart of using historical data to optimize the prediction model in a method for distribution network investment based on a multi-source heterogeneous data mode in an embodiment of the present invention to form a trained prediction model.
[0125] like Figure 5As shown, in one or more embodiments, preferably, the use of historical data to optimize the prediction model to form a trained prediction model specifically includes:
[0126] S501. Use historical data to analyze forecast deviations and identify patterns and sources of errors;
[0127] S502: According to the result of the error analysis, the prediction model is adjusted to a mode with the smallest prediction deviation for the historical data as the trained prediction model.
[0128] In an embodiment of the present invention, the pre-processed historical data is used to perform preliminary training on the preset prediction model. For example, if a neural network model is used, the network weights are first randomly initialized, and then trained by the back propagation algorithm so that the model can learn the basic patterns and relationships in the data. The prediction model obtained by the training is used to make predictions on the validation set, and the deviation between the prediction results and the actual values is analyzed to identify the pattern and source of the error. For example, by comparing the predicted load and the actual load, it is found that in hot weather, the predicted load is generally lower than the actual load, indicating that the model may not fully consider the impact of temperature on the load. According to the results of the error analysis, the structure or parameters of the prediction model are adjusted to reduce the prediction deviation. Continue to use historical data for training and verification until the prediction performance of the model is optimal. For example, temperature can be added as an input feature in the model, or the number of layers and nodes of the neural network can be adjusted to improve the model's ability to capture complex relationships. After the model has been trained and adjusted for multiple rounds, the model that performs best on the validation set is selected as the trained prediction model. This model will be able to more accurately predict future power grid loads and resource requirements, providing more reliable decision support for power grid operations.
[0129] Figure 6 It is a flowchart of inputting real-time data into a trained prediction model to generate future distribution network investment decision features in a method for distribution network investment based on a multi-source heterogeneous data mode in an embodiment of the present invention.
[0130] like Figure 6 As shown, in one or more embodiments, preferably, the step of inputting real-time data into a trained prediction model to generate future distribution network investment decision features specifically includes:
[0131] S601. Fusion and synchronization are performed based on real-time data to ensure the timeliness and consistency of data and form input data for prediction analysis;
[0132] S602: Based on the prediction analysis input data, extract changes in key features related to distribution network investment decisions from the output of the trained prediction model.
[0133] In an embodiment of the present invention, the processed predictive analysis input data is input into the trained predictive model. The model has been trained and optimized with historical data and can capture key patterns and trends in grid load and resource demand. For example, if the trained predictive model is a deep learning network, it will output a forecast of grid load for a period of time in the future based on the input real-time data. The changes in key features related to distribution network investment decisions are extracted from the output of the predictive model. These features may include future load growth points, potential overload areas, equipment that needs to be upgraded or replaced, etc. For example, by analyzing the output of the predictive model, specific distribution lines where loads may exceed equipment capacity in the next 6 months are identified, thereby determining the grid areas that need priority investment upgrades. Using the extracted investment decision features, combined with the actual needs of distribution network planning and operation, investment decision analysis is performed. For example, based on the predicted future load distribution and the capacity limitations of the existing distribution network, an optimal distribution network expansion plan is formulated to meet future needs.
[0134] Figure 7 It is a flowchart of judging the scope of evaluation influence according to the distribution network investment decision-making characteristics in a distribution network investment method based on a multi-source heterogeneous data mode in one embodiment of the present invention.
[0135] like Figure 7 As shown, in one or more embodiments, preferably, the scope of the evaluation impact is determined according to the distribution network investment decision characteristics, specifically including:
[0136] S701. For each data parameter, a preset impact ratio of economic benefit, system stability and social impact is set, and the sum of the impact ratios of economic benefit, system stability and social impact corresponding to each parameter is 1;
[0137] S702. Based on the changes in the characteristics of future distribution network investment decisions, multiply them by the preset impact ratios of the corresponding economic benefits, system stability and social impact to take them as the main impact of the corresponding investment.
[0138] In an embodiment of the present invention, a corresponding preset impact ratio of economic benefit, system stability and social impact is set for each data parameter in the distribution network investment decision feature. For example, for the decision feature of the newly added distribution line, the economic benefit impact ratio may be set to 0.4, the system stability impact ratio to 0.3, and the social impact ratio to 0.3. Ensure that the sum of the impact ratios of economic benefit, system stability and social impact corresponding to each parameter is 1, so as to facilitate quantitative analysis. According to the changes in the future distribution network investment decision features, identify which parameters will change. For example, the prediction model may indicate that the load in a certain area will increase by 20% in the next 5 years, and new distribution lines will be needed to cope with the load growth. The predicted changes in the distribution network investment decision features are multiplied by the preset impact ratios of the corresponding economic benefits, system stability and social impact, and the main impacts expected to be generated by each investment decision are calculated. For example, if the decision feature change of the newly added distribution line is 1 (indicating that a large amount of investment is required), its impact on economic benefit is 1*0.4=0.4, its impact on system stability is 1*0.3=0.3, and its impact on social impact is 1*0.3=0.3. The impacts calculated from the changes in all investment decision characteristics are summed up to obtain the expected impact of the overall investment decision on economic benefits, system stability, and social impact. For example, if in addition to adding new distribution lines, other decision characteristics such as upgrading substations are also included, the impact of each decision characteristic is accumulated to obtain the final comprehensive impact assessment. Through the above steps, the scope of the impact can be judged and evaluated based on the distribution network investment decision characteristics, including economic benefits, system stability, and social impact, thereby providing decision makers with a comprehensive analysis and quantitative impact assessment to support more informed and reasonable investment decisions.
[0139] According to a second aspect of an embodiment of the present invention, a system for distribution network investment based on a multi-source heterogeneous data mode is provided.
[0140] Figure 8 It is a structural diagram of a system for distribution network investment based on a multi-source heterogeneous data mode according to an embodiment of the present invention.
[0141] In one or more embodiments, preferably, the system for distribution network investment based on a multi-source heterogeneous data mode includes:
[0142] The data collection and integration module 801 is used to establish a centralized data platform to automatically collect and integrate real-time monitoring data, historical records, meteorological information, equipment maintenance logs and market operation data, and store them on the data platform;
[0143] Extraction of key features module 802, used to obtain the integrated data and extract the distribution network investment decision-making features therein;
[0144] A preliminary prediction module 803 is used to predict future grid load and resource requirements in combination with a preset prediction model;
[0145] A prediction correction module 804 is used to optimize the prediction model using historical data to form a trained prediction model;
[0146] The feature correction module 805 is used to input the real-time data into the trained prediction model to generate the future distribution network investment decision features;
[0147] The impact display module 806 is used to determine the scope of the assessment impact according to the distribution network investment decision-making characteristics.
[0148] In the embodiment of the present invention, a system suitable for different structures is realized through a series of modular designs. The system can achieve closed-loop, reliable and efficient execution through collection, analysis and control.
[0149] According to a third aspect of an embodiment of the present invention, there is provided a computer-readable storage medium on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method as described in any one of the first aspect of the embodiment of the present invention is implemented.
[0150] According to a fourth aspect of an embodiment of the present invention, an electronic device is provided. Fig. 9 It is a structural diagram of an electronic device in one embodiment of the present invention. Fig. 9 The electronic device shown is a general distribution network investment device based on a multi-source heterogeneous data mode. The electronic device can be a smart phone, a tablet computer and other devices. As shown, the electronic device 900 includes a processor 901 and a memory 902. Among them, the processor 901 is electrically connected to the memory 902. The processor 901 is the control center of the terminal 900, and uses various interfaces and lines to connect various parts of the entire terminal. By running or calling a computer program stored in the memory 902, and calling the data stored in the memory 902, the various functions of the terminal are executed and the data is processed, thereby monitoring the terminal as a whole.
[0151] In this embodiment, the processor 901 in the electronic device 900 will load the instructions corresponding to the processes of one or more computer programs into the memory 902 according to the following steps, and the processor 901 will run the computer program stored in the memory 902 to achieve various functions: establish a centralized data platform to automatically collect and integrate real-time monitoring data, historical records, meteorological information, equipment maintenance logs and market operation data, and store them in the data platform; obtain the integrated data and extract the distribution network investment decision characteristics therein; combine the preset prediction model to predict future power grid load and resource requirements; use historical data to optimize the prediction model to form a trained prediction model; input real-time data into the trained prediction model to generate future distribution network investment decision characteristics; judge the scope of evaluation impact based on the distribution network investment decision characteristics.
[0152] The memory 902 may be used to store computer programs and data. The computer programs stored in the memory 902 include instructions that can be executed in the processor. The computer programs may constitute various functional modules. The processor 901 executes various functional applications and data processing by calling the computer programs stored in the memory 902.
[0153] The technical solution provided by the embodiments of the present invention may have the following beneficial effects:
[0154] In the solution of the present invention, data from different sources and in different formats are unified together to create a comprehensive data view, thereby achieving efficient distribution data fusion analysis.
[0155] In the solution of the present invention, the key data features that affect the distribution network investment decision are automatically identified and extracted, and the impact range is analyzed to determine the impact range.
[0156] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.
[0157] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0158] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0159] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0160] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. A method for distribution network investment based on multi-source heterogeneous data mode, It is characterized in that The method includes: Establish a centralized data platform to automatically collect and integrate real-time monitoring data, historical records, weather information, equipment maintenance logs, and market operation data, and store them on the data platform; Obtain the integrated data and extract the distribution network investment decision-making features; Combined with preset forecasting models to predict future grid load and resource requirements; Optimizing the prediction model using historical data to form a trained prediction model; Input real-time data into the trained prediction model to generate features for future distribution network investment decisions; The scope of the assessment impact is determined based on the characteristics of distribution network investment decisions.
2. A method for distribution network investment based on a multi-source heterogeneous data model as claimed in claim 1, It is characterized in that The establishment of a centralized data platform automatically collects and integrates real-time monitoring data, historical records, weather information, equipment maintenance logs and market operation data, and stores them on the data platform, including: Design a centralized data platform to support the integration and storage of real-time monitoring data, historical records, weather information, equipment maintenance logs, and market operation data; Automatically collect data, and conduct real-time data monitoring, equipment maintenance, and operational data collection through various communication protocols.
3. A method for distribution network investment based on a multi-source heterogeneous data model as claimed in claim 1, It is characterized in that The step of obtaining the integrated data and extracting the distribution network investment decision-making features therein specifically includes: Perform preprocessing operations on raw data collected from multiple data sources, including removing outliers and noise, and standardizing data to eliminate the impact of different dimensions; The frequency domain features are obtained by Fourier calculation on the preprocessed data; Calculate the rate of change of the preprocessed data; Extract frequency domain features, and use the first calculation formula to calculate the correlation index between the frequency domain features and the value feedback index; Extract the data change rate and use the second calculation formula to calculate the value feedback index correlation; When the correlation index between the frequency domain feature and the value feedback index satisfies the third calculation formula, the corresponding frequency domain feature of this feature is retained as the distribution network investment decision feature; When the value feedback index correlation satisfies the fourth calculation formula, this feature is retained in the distribution network investment decision feature; The first calculation formula is: G=(HJ)÷(TD-TX) Among them, G is the correlation index between the frequency domain feature and the value feedback index, H is the frequency domain feature value corresponding to the maximum value of the value feedback index, J is the frequency domain feature value corresponding to the average value of the value feedback index, TD is the maximum value of the frequency domain feature value, and TX is the minimum value of the frequency domain feature value; The second calculation formula is: P=ZS Among them, P is the correlation of the value feedback index, Z is the normalized index of the change rate of the value feedback index, and S is the normalized index of the data change rate; The third calculation formula is: G>y1 Wherein, y1 is the first contrast margin; The fourth calculation formula is: P <y2 Wherein, y2 is the first contrast margin.
4. A method for distribution network investment based on a multi-source heterogeneous data model as claimed in claim 1, It is characterized in that The method of combining a preset prediction model to predict future power grid load and resource demand specifically includes: Select a preset model suitable for grid load and resource demand forecasting; Preprocess the collected real-time monitoring data, historical records, meteorological information, equipment maintenance logs, and market operation data to extract and construct features useful for the prediction model; The pre-processed data is used to train the preset prediction model; Integrate the prediction model into the data platform to achieve continuous processing of real-time data streams and load forecasting.
5. A method for distribution network investment based on multi-source heterogeneous data mode as claimed in claim 1, It is characterized in that The using historical data to optimize the prediction model to form a trained prediction model specifically includes: Use historical data to perform forecast bias analysis to identify patterns and sources of error; According to the result of the error analysis, the prediction model is adjusted to a mode with the smallest prediction deviation for the historical data as the trained prediction model.
6. A method for distribution network investment based on a multi-source heterogeneous data model as claimed in claim 1, It is characterized in that The real-time data is input into the trained prediction model to generate future distribution network investment decision features, specifically including: Fusion and synchronization based on real-time data to ensure the timeliness and consistency of data and form input data for predictive analysis; Based on the predictive analysis input data, changes in key features related to distribution network investment decisions are extracted from the output of the trained predictive model.
7. A method for distribution network investment based on a multi-source heterogeneous data model as claimed in claim 1, It is characterized in that The scope of the assessment impact determined based on the distribution network investment decision-making characteristics specifically includes: For each data parameter, a preset impact ratio of economic benefit, system stability and social impact is set, and the sum of the impact ratios of economic benefit, system stability and social impact corresponding to each parameter is 1; According to the changes in the characteristics of future distribution network investment decisions, multiply it by the preset impact ratio of the corresponding economic benefits, system stability and social impact to obtain the main impact of the corresponding investment.
8. A distribution network investment system based on multi-source heterogeneous data mode, It is characterized in that The system is used to implement the method according to any one of claims 1 to 7, and the system comprises: The data collection and integration module is used to establish a centralized data platform to automatically collect and integrate real-time monitoring data, historical records, meteorological information, equipment maintenance logs and market operation data, and store them on the data platform; Extract key feature module, which is used to obtain the integrated data and extract the distribution network investment decision-making features; A preliminary forecasting module is used to forecast future grid load and resource requirements in combination with a preset forecasting model; A prediction correction module, used to optimize the prediction model using historical data to form a trained prediction model; A feature correction module is used to input real-time data into the trained prediction model to generate features for future distribution network investment decisions; The impact display module is used to determine the scope of the assessment impact based on the characteristics of distribution network investment decisions.
9. A computer-readable storage medium having computer program instructions stored thereon, It is characterized in that The computer program instructions implement the method according to any one of claims 1 to 7 when executed by a processor.
10. An electronic device comprising a memory and a processor, It is characterized in that The memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method according to any one of claims 1-7.
