Power transmission and transformation engineering quantity dynamic prediction method, device, equipment, medium and product under power demand fluctuation
By combining the LSTM and BP neural network methods, the problems of low efficiency and insufficient accuracy of traditional power transmission and transformation engineering quantities are solved, and dynamic prediction of engineering quantities under fluctuations in power demand is realized, which improves the scientificity and practicality of the prediction.
Patent Information
- Application Number
- CN202510473673.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-16
- Publication Date
- 2025-08-01
AI Technical Summary
The traditional power transmission and transformation project volume prediction methods are inefficient and insufficiently accurate, and cannot adapt to the dynamic fluctuations of power demand, resulting in the inconsistent project volume evaluation and actual demand, and data processing is limited, making it difficult to characterize the nonlinear relationship between project volume and multiple factors.
A combination of long and short-term memory neural network (LSTM) and BP neural network is used to perform dynamic predictions by selecting multiple factors that affect power demand and engineering volume.
It improves the efficiency and accuracy of power transmission and transformation project volume prediction, can adapt to the dynamic fluctuations of power demand, and achieve accurate prediction of project volume.
Smart Images
Figure CN120410047A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of engineering quantity prediction, and in particular, to a dynamic prediction method, device, equipment, medium and product for transmission and transformation engineering quantity under power demand fluctuations. Background Art
[0002] The prediction of engineering quantity is an important part of the management of transmission and transformation engineering projects. Accurate prediction of engineering quantity helps transmission and transformation engineering enterprises make correct decisions in the early stage and improve project management level. However, the prediction of engineering quantity has not received enough attention in the traditional engineering quantity final accounts work, and there are problems such as information lag, time-consuming and laborious, and static prediction in the means of engineering quantity prediction. Especially for transmission and transformation projects with a long cycle, there will be a certain difference between the engineering quantity at the initial stage of the project and the engineering quantity during actual construction. As the most important component of transmission and transformation engineering quantity, power demand has the characteristics of large proportion and large fluctuation. Accurate prediction of the power demand time series is of great significance for the dynamic prediction of transmission and transformation engineering quantity.
[0003] Traditional engineering quantity prediction relies on artificial experience and static formulas, and there are two core defects. The first is low efficiency. It takes a lot of time for manual analysis of construction drawings and statistics of engineering quantity. Especially when dealing with complex projects, data extraction is prone to errors and it is difficult to scale up. The second is insufficient accuracy. Empirical estimation is easily affected by subjective factors, resulting in large budget deviations. For example, the impact of geographical environment complexity (such as terrain, vegetation coverage) on construction costs is ignored, resulting in inconsistent engineering quantity assessment and actual requirements. Secondly, data processing has limitations. Historical engineering data records are incomplete, there are redundancies or missing values between indicators, and the fragmentation of data makes it difficult to be directly used for model training. Traditional methods do not consider the dynamic changes of time series (such as material price fluctuations, power demand growth) and the impact of external environments (such as policies, climate) on engineering quantity, resulting in static prediction results and inability to adapt to actual engineering adjustments. Linear models have limitations. Traditional models such as simple linear regression and moving average are difficult to describe the non-linear relationship between engineering quantity and multiple factors (such as the length of the external line, electricity consumption capacity), and they have poor adaptability to high-dimensional data. It can be seen that traditional engineering quantity prediction methods have deficiencies in terms of efficiency, accuracy and dynamic adaptability. Summary of the Invention
[0004] The purpose of the present application is to provide a dynamic prediction method, device, equipment, medium and product for transmission and transformation engineering quantity under power demand fluctuations, which can improve the efficiency and accuracy of dynamic prediction of transmission and transformation engineering quantity.
[0005] To achieve the above object, the present application provides the following solutions:
[0006] In a first aspect, the present application provides a dynamic prediction method for the quantity of power transmission and transformation projects under power demand fluctuations, including: selecting multiple factors affecting power demand fluctuations as independent variables for power demand impact; obtaining future power demand time series prediction values using a trained long short-term memory neural network (Long Short-Term Memory, LSTM) based on the historical power load data of the independent variables for power demand impact; using the SelectKBest method to select multiple factors affecting the quantity of power transmission and transformation projects as independent variables for the quantity of power transmission and transformation projects impact; the independent variables for the quantity of power transmission and transformation projects impact including power demand; and inputting the future power demand time series prediction values and the future values of the independent variables for the quantity of power transmission and transformation projects impact other than power demand into a trained BP (BackPropagation) neural network to obtain future prediction values for the quantity of power transmission and transformation projects.
[0007] In a second aspect, the present application provides a dynamic prediction device for the quantity of power transmission and transformation projects under power demand fluctuations, including: a power demand impact factor selection module, a power demand prediction module, a project quantity impact factor selection module, and a project quantity prediction module.
[0008] The power demand impact factor selection module is used to select multiple factors affecting power demand fluctuations as independent variables for power demand impact;
[0009] The power demand prediction module is used to obtain future power demand time series prediction values using a trained long short-term memory neural network based on the historical power load data of the independent variables for power demand impact;
[0010] The project quantity impact factor selection module is used to use the SelectKBest method to select multiple factors affecting the quantity of power transmission and transformation projects as independent variables for the quantity of power transmission and transformation projects impact; the independent variables for the quantity of power transmission and transformation projects impact including power demand;
[0011] The project quantity prediction module is used to input the future power demand time series prediction values and the future values of the independent variables for the quantity of power transmission and transformation projects impact other than power demand into a trained BP neural network to obtain future prediction values for the quantity of power transmission and transformation projects.
[0012] In a third aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the computer program to implement the dynamic prediction method for the quantity of power transmission and transformation projects under power demand fluctuations described in any one of the above.
[0013] In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the dynamic prediction method for the quantity of power transmission and transformation projects under power demand fluctuations described in any one of the above.
[0014] In a fifth aspect, the present application provides a computer program product, including a computer program which, when executed by a processor, implements the dynamic prediction method for the power transmission and transformation project quantity under power demand fluctuations described in any one of the above.
[0015] According to the specific embodiments provided by the present application, the present application has the following technical effects:
[0016] The present application provides a dynamic prediction method, device, equipment, medium and product for the power transmission and transformation project quantity under power demand fluctuations. The long short-term memory neural network is used to dynamically predict the power demand, and the predicted value of the power transmission and transformation project quantity can be obtained by using the BP neural network based on the dynamically predicted power demand. By introducing the power demand time series prediction and the BP neural network into the prediction of the power transmission and transformation project quantity and organically combining the two, the efficiency and accuracy of the dynamic prediction of the power transmission and transformation project quantity are improved under power demand fluctuations. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.
[0018] Figure 1 It is a schematic flowchart of a dynamic prediction method for the power transmission and transformation project quantity under power demand fluctuations provided by an embodiment of the present application;
[0019] Figure 2 It is a schematic overall framework diagram of a dynamic prediction method for the power transmission and transformation project quantity under power demand fluctuations provided by an embodiment of the present application;
[0020] Figure 3 It is a schematic structural diagram of a long short-term memory neural network provided by an embodiment of the present application;
[0021] Figure 4 It is a schematic LSTM layer structure diagram provided by another embodiment of the present application;
[0022] Figure 5 It is a schematic structural diagram of a BP neural network provided by another embodiment of the present application;
[0023] Figure 6 It is a schematic structural diagram of a computer device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0024] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.
[0025] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.
[0026] The deficiencies of traditional engineering quantity prediction methods in terms of efficiency, accuracy, and dynamic adaptability have given rise to intelligent prediction technologies based on image recognition, machine learning, and big data. These technologies have significantly improved the scientificity and practicality of engineering quantity prediction by integrating multi-source data, optimizing the model structure (such as the LSTM+BP fusion), and dynamic correction mechanisms, providing technical support for the efficient management and cost control of power projects.
[0027] In view of this, in an exemplary embodiment, as Figure 1 shown, a dynamic prediction method for transmission and transformation engineering quantity under power demand fluctuations is provided, including the following steps 101 to 104. Among them:
[0028] Step 101: Select multiple factors that affect power demand fluctuations as independent variables for power demand impact.
[0029] Step 102: According to the historical power load data of the independent variables for power demand impact, use the trained long short-term memory neural network to obtain the predicted values of the future power demand time series.
[0030] Step 103: Use the SelectKBest method to select multiple factors that affect the transmission and transformation engineering quantity as independent variables for transmission and transformation engineering quantity impact; the independent variables for transmission and transformation engineering quantity impact include power demand.
[0031] Step 104: Input the predicted values of the future power demand time series and the future values of the independent variables for transmission and transformation engineering quantity impact other than power demand into the trained BP neural network to obtain the predicted values of the future transmission and transformation engineering quantity.
[0032] The transmission and transformation projects applicable to the dynamic prediction method for transmission and transformation engineering quantity under power demand fluctuations in the present application specifically refer to: transmission and transformation projects with a voltage of 35 kV or below.
[0033] Figure 2Overall framework diagram of the dynamic prediction method for the transmission and transformation project volume under power demand fluctuations in this application. To implement the above steps 101 to 104, power demand time series prediction and BP neural network are introduced into the prediction of the transmission and transformation project volume below 35 kV, and the two are organically combined to construct a dynamic prediction method for the transmission and transformation project volume under power demand fluctuations.
[0034] In another exemplary embodiment of this application, the above step 101 can be replaced by the following steps 201 to 202:
[0035] Step 201: Select multiple initial factors affecting power demand fluctuations.
[0036] Power demand fluctuations are mainly affected by economic development, environmental climate, population density, etc. Therefore, on the basis of synthesizing relevant research results at home and abroad, the following 13 groups of independent variables are initially selected, including: from the dimension of economic development, industrial electricity consumption intensity, manufacturing PMI index (Purchasing Managers' Index), and power consumption of data centers; from the dimension of climate environment, daily maximum temperature in summer, number of days of extreme weather, and heating degree days (HDD); from the dimension of energy transformation, electric vehicle penetration rate and photovoltaic installed capacity; from the dimension of social population, population density of megacities and household air conditioner ownership; from the dimension of market policy, carbon trading price, peak-valley electricity price difference, and power rationing rate for high-energy-consuming industries.
[0037] Step 202: Conduct Pearson correlation analysis on the multiple initial factors, and screen out the key factors affecting power demand fluctuations as the independent variables of power demand impact.
[0038] Through the analysis of the correlation between each index and Pearson correlation, the correlation between each index and power demand is obtained. Pearson correlation is an important index to measure the relationship between two variables, and its relationship range is from 0 to 1. When the relationship between the Pearson correlation factor and the independent variable is closer to 0 or 1, the degree of association with the factor is higher. By calculating the Pearson correlation coefficient between each of the 13 features and power demand, the feature with the highest correlation is found.
[0039]
[0040] In the formula, P xy is the Pearson correlation factor; x i represents the i-th component of the sequence x, y i represents the i-th component of the sequence y,
[0041] Eight groups of features with relatively high correlations were selected according to the Pearson correlation coefficient, namely industrial electricity intensity, manufacturing PMI index, daily maximum temperature in summer, number of consecutive days of extreme weather, electric vehicle penetration rate, photovoltaic installed capacity, average household air conditioner ownership, and carbon trading price.
[0042] In another exemplary embodiment of the present application, as Figure 3 shown, the long short-term memory neural network includes: an input layer (the first input layer), a hidden layer (the first hidden layer), and an output layer (the first output layer). The input layer has 8 neurons (should be 8) for receiving one-dimensional input features. The dimension of the input features here is 8. The hidden layer consists of LSTM layers and contains 50 neurons. These neurons are used for feature extraction and operation and can capture the time-dependent relationships in the input sequence. The input dimension of the LSTM layer is (None, 1, 8), and the output dimension is (None, 1). Here, "1" means outputting one feature value at each time step, and "8" means the dimension of the input features. Output layer (Dense layer): The output layer is a fully connected layer (Dense layer) that contains 1 neuron. This neuron is fully connected to the 50 neurons of the LSTM layer and performs a weighted sum operation on the output results of the LSTM layer to obtain the final prediction result. The input dimension of the Dense layer is (None, 50), and the output dimension is (None, 1). Here, "50" means receiving 50 feature values from the LSTM layer, and "1" means the dimension of the final prediction result.
[0043] The LSTM layer structure is as Figure 4 shown.
[0044] In another exemplary embodiment of the present application, the training process of the long short-term memory neural network can be replaced by the following steps 301 to 305:
[0045] Step 301: Obtain multiple pieces of electricity demand data for a historical time period; the electricity demand data includes historical electricity load data and historical electricity demand values of independent variables affecting electricity demand.
[0046] Step 302: Normalize each piece of the electricity demand data, and form an electricity demand data set with all the normalized electricity demand data.
[0047] The normalization is linear normalization, and the formula is:
[0048]
[0049] where x i ′ is the normalized data, x i is the original data, minx i is the minimum value in the original data, and maxx iis the maximum value in the original data.
[0050] Step 303: Divide the power demand data set into a training data set and a test set.
[0051] Step 304: Set training parameters; the training parameters include: a step size of 1, a batch size of 12, and a number of training rounds of 250.
[0052] Step 305: Train the long short-term memory neural network according to the training data set and the training parameters, and use the test set to verify the accuracy of the trained long short-term memory neural network. Determine the long short-term memory neural network with an accuracy greater than or equal to the accuracy threshold as the trained long short-term memory neural network.
[0053] Exemplarily, this application selects a total of 94 pieces of historical load data of the power grid company's transmission and transformation projects from December 2019 to September 2023, and uses 61 of these projects as the training data set to train the LSTM model. In the hyperparameter setting of the model, the step size is set to 1, the batch size is 12, and 250 rounds of training are performed. After the training is completed, the convergence of the model is checked by visualizing the loss during testing and training. From the trend of the training results of the loss value, the model has approached fitting after the 150th round, the training loss is about 0.05, and the test loss also remains at about 0.1. Use the remaining 33 pieces of power load data as the test set to verify the accuracy of the model prediction.
[0054] In another exemplary embodiment of this application, the above step 103 can be replaced by the following steps 401 to 405:
[0055] Step 401: Obtain the historical data of transmission and transformation projects with a voltage of 35 kV and below within the historical time period.
[0056] Step 402: Organize the historical data into multiple attribute indicators.
[0057] Step 403: Combine the attribute indicators with the same content to obtain multiple initial indicators affecting the amount of transmission and transformation work.
[0058] Step 404: Use the F-test mathematical method to perform a correlation test on multiple initial indicators to obtain a P value.
[0059] Step 405: According to the P value, use the SelectKBest method to select multiple indicators affecting the amount of transmission and transformation work from multiple initial indicators as the independent variables affecting the amount of transmission and transformation work; the independent variables affecting the amount of transmission and transformation work include project type, outside line length, electricity consumption capacity, number of users, inside line length, power demand, geographical complexity, and government subsidies.
[0060] Exemplarily, the F-test yields index data related to power transmission and transformation projects. The selected data are the historical data of the actual power transmission and transformation projects of a regional power grid company from 2019 to 2023. After preliminary summarization and collation of the historical data for these 6 years, it is found that there are a total of 217 distribution projects from 2017 to 2023. Due to the long time span, the company's records of historical data are not completely the same, and there are a total of 60 various historical budget indicators. There are intersections and overlaps among the indicators, and some indicator data are seriously missing. Therefore, there is a large amount of work in data preprocessing. All data need to be regularized, and the overlapping indicators need to be merged and sorted. A total of 169 relatively complete projects are selected from the historical data. The indicators sorted for the total of 169 groups are as follows: 1 project type, 2 geographical scope, 3 outside line type, 4 outside line length, 5 pole model, 6 number of poles, 7 conductor model, 8 conductor unit price, 9 electricity consumption type, 10 cable unit price, 11 length of two-hole pipe row, 12 cable model, 13 electricity consumption capacity (kVA), 14 inside line length, 15 transformer model, 16 number of transformers, 17 transformer size, 18 number of distribution cabinets, 19 high-voltage incoming cabinet model, 20 high-voltage outgoing cabinet model, 21 metering cabinet model, 22 number of low-voltage distribution cabinets, 23 number of low-voltage distribution loops, 24 DC panel model, 25 communication system mode, 26 whether there is a fire monitoring system, 27 whether there is a tie busbar cabinet, 28 busbar model, 29 busbar length, 30 number of household meters, 31 power load, etc., a total of 31 indicator historical data. Different projects have different attribute indicators. For two different attribute indicators with the same content meaning, we merge them to reduce the number of indicators. There are obvious overlaps in the indicators of high- and low-voltage distribution cabinets. For 18 number of distribution cabinets, 19 high-voltage incoming cabinet model, 20 high-voltage outgoing cabinet model, 21 metering cabinet model, 23 number of low-voltage distribution loops, 24 DC panel model, the number of distribution cabinets actually includes the data information of the other 6 items, and the information about the model is included in the three characteristic indicators of total electricity consumption capacity, electricity consumption type, and project type. The five data indicators of 15 transformer model, 16 number of transformers, 17 transformer size, 28 busbar model, and 29 busbar length can be merged. The project type and total electricity consumption capacity have defined the transformer size, and in most cases, each transformer corresponds to a standard length of busbar. The information of 10 cable unit price, 12 cable model, 7 conductor model, and 8 conductor unit price is included in the total project capacity, electricity consumption type, and project type information, so the indicators are merged. For 5 pole model and 6 number of poles, the information about the number of pole models with standard length intervals is included in the outside line length and type. Information such as 25 communication system mode and 26 whether there is a fire monitoring system is included in the project type and electricity consumption type information according to the basic requirements, so the indicators are merged.After initially merging overlapping information indicators, 31 statistical indicators are initially merged into 12 power transmission and transformation project volume prediction indicators, which still need to be further optimized compared to the number of indicators used for modeling. For these 12 project volume prediction indicators, the following uses the F-test mathematical method to conduct a correlation test to obtain the indicators with the highest correlation. One is the between-group difference, that is, the mean difference between different groups, which is represented by the sum of the squared deviations of the variable's mean in each group from the overall mean, denoted as SSA. If there are r groups, its degrees of freedom is r - 1; the other is the within-group difference, that is, the difference between samples within the same group, which is represented by the sum of the squared deviations of the variable's mean in each group from the variable values within that group, denoted as SSE. If there are a total of n samples and r groups, its degrees of freedom is n - r.
[0061]
[0062] The essence of the F-test is to find the linear relationship between two sets of data. Its null hypothesis is that there is no significant linear relationship in the data. It returns two statistics, the F value and the P value. When the P value of a feature is larger, it indicates that the feature has a significant linear correlation with the label (power transmission and transformation project volume). Conversely, when the P value is smaller, the feature is considered to have no significant linear relationship with the label and should be deleted. Similar to the chi-square test, the F test also requires the use of the SelectKBest method to complete. The SelectKBest method can select features based on the top k scores. Users can directly judge the correlation between features and labels through the output scores, and then select the value of k. In this patent, k is selected as 8. According to the feature scores of SelectKBest, this patent can select the eight indicators with the highest correlation as features, namely project type, outside line length, electricity consumption capacity, number of users, inside line length, power demand, geographical complexity, and government subsidy.
[0063] In another exemplary embodiment of the present application, as Figure 5 shown, after determining that the input features of the BP neural network are 8-dimensional and the output is 1-dimensional, the BP neural network includes: an input layer (the second input layer), a hidden layer (the second hidden layer), and an output layer (the second output layer); the second input layer includes 8 neurons, the second hidden layer includes 15 neurons, and the second output layer includes 1 neuron.
[0064] In another exemplary embodiment of the present application, when training the BP neural network, after obtaining the historical data of the power transmission and transformation project volume, scaling and normalization processing is performed. The batch size is set to 8, and 200 rounds of training are carried out. After the training is completed, the convergence of the model is checked by printing the losses during testing and training. From the changing trend of the loss values, it can be seen that the model has approached fitting after the 100th round. The training loss is around 0.01, and the testing loss also remains around 0.03.
[0065] In another exemplary embodiment of the present application, the power transmission and transformation project quantity includes the number of transformer installations, the cable laying length, etc.
[0066] The present application introduces power demand time series prediction and BP neural network into the power transmission and transformation project quantity below 35 kV, and organically combines the two to construct a prediction model for the power transmission and transformation project quantity (such as the number of transformer installations, the cable laying length, etc.) under power demand fluctuations. By analyzing the influencing factors of power demand fluctuations, finally 8 important independent variables affecting power demand are selected. Based on the LSTM long short-term memory neural network, a prediction model for the future power demand time series value is established, and the prediction result of the power demand time series value is obtained. Secondly, statistical analysis is carried out on the data characteristics of the power transmission and transformation projects below 35 kV. Through the SelectKBest method, the most important 8 independent variables of the power transmission and transformation projects are selected, and a model for predicting the power transmission and transformation project quantity is established and trained based on the BP neural network, realizing the dynamic prediction of the project quantity of the power transmission and transformation projects. Finally, the power demand time series model and the BP neural network power transmission and transformation project quantity prediction model are combined in series. The predicted value of the power demand time series is used as the input value, and the dynamic predicted project quantity in the power transmission and transformation project quantity prediction model is output.
[0067] Based on the same inventive concept, the embodiment of the present application also provides a device for dynamically predicting the power transmission and transformation project quantity under power demand fluctuations for implementing the above-mentioned method for dynamically predicting the power transmission and transformation project quantity under power demand fluctuations. The implementation solutions provided by this device to solve problems are similar to the implementation solutions described in the above method. Therefore, the specific limitations in one or more embodiments of the device for dynamically predicting the power transmission and transformation project quantity under power demand fluctuations provided below can refer to the limitations on the method for dynamically predicting the power transmission and transformation project quantity under power demand fluctuations in the above text, and will not be repeated here.
[0068] In an exemplary embodiment, a device for dynamically predicting the power transmission and transformation project quantity under power demand fluctuations is provided, including: a power demand influencing factor selection module, a power demand prediction module, a project quantity influencing factor selection module, and a project quantity prediction module.
[0069] The power demand influencing factor selection module is used to select multiple factors affecting power demand fluctuations as independent variables of power demand influence.
[0070] The power demand prediction module is used to obtain the future power demand time series prediction value by using the trained long short-term memory neural network according to the historical power load data of the independent variables of power demand influence.
[0071] The engineering quantity influencing factor selection module is used to select multiple factors influencing the power transmission and transformation engineering quantity by using the SelectKBest method as the independent variables influencing the power transmission and transformation engineering quantity. The independent variables influencing the power transmission and transformation engineering quantity include power demand.
[0072] The engineering quantity prediction module is used to input the future predicted value of the power demand time series and the future values of the independent variables influencing the power transmission and transformation engineering quantity other than the power demand into the trained BP neural network to obtain the future predicted value of the power transmission and transformation engineering quantity.
[0073] In an exemplary embodiment, a computer device is provided. The computer device can be a server or a terminal, and its internal structure diagram can be as Figure 6 shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the future predicted value of the power transmission and transformation engineering quantity. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a dynamic prediction method for the power transmission and transformation engineering quantity under power demand fluctuations.
[0074] Those skilled in the art can understand that Figure 6 the structure shown in
[0075] is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.
[0076] In an exemplary embodiment, a computer program product is provided, including a computer program which, when executed by a processor, implements the steps in the above method embodiments.
[0077] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data that have been authorized by the user or fully authorized by all parties.
[0078] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above method embodiments. Among them, any reference to a memory, database, or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memories. Non-volatile memories can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memories can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.
[0079] The databases involved in the embodiments provided in this application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in this application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.
[0080] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.
[0081] Specific examples are used in this article to elaborate on the principles and implementation manners of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those of ordinary skill in the art, according to the idea of the present application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present application.
Claims
1. A dynamic prediction method for the amount of power transmission and transformation projects under power demand fluctuations, characterized in that, Including: Select multiple factors that affect the fluctuation of power demand as the independent variables affecting power demand; Based on the historical power load data of the independent variables affecting power demand, use the trained long short-term memory neural network to obtain the predicted values of the future power demand time series; Use the SelectKBest method to select multiple factors that affect the quantity of transmission and transformation projects as the independent variables affecting the quantity of transmission and transformation projects; the independent variables affecting the quantity of transmission and transformation projects include power demand; Input the predicted values of the future power demand time series and the future values of the independent variables affecting the quantity of transmission and transformation projects other than power demand into the trained BP neural network to obtain the predicted values of the future quantity of transmission and transformation projects.
2. The dynamic prediction method for the power transmission and transformation project volume under power demand fluctuations according to claim 1, characterized in that, Select multiple factors that affect the fluctuation of power demand as the independent variables affecting power demand, specifically including: Select multiple initial factors that affect the fluctuation of power demand; the multiple initial factors include: industrial electricity consumption intensity, manufacturing PMI index, power consumption of data centers, daily maximum temperature in summer, number of days of extreme weather, heating degree days, electric vehicle penetration rate, photovoltaic installed capacity, urban population density, household air conditioner ownership, carbon trading price, peak-valley electricity price difference, and power rationing rate for high-energy-consuming industries; Conduct Pearson correlation analysis on the multiple initial factors, and screen out the key factors that affect the fluctuation of power demand as the independent variables affecting power demand; the independent variables affecting power demand include: industrial electricity consumption intensity, manufacturing PMI index, daily maximum temperature in summer, number of days of extreme weather, electric vehicle penetration rate, photovoltaic installed capacity, household air conditioner ownership, and carbon trading price.
3. The dynamic prediction method for the power transmission and transformation project volume under power demand fluctuations according to claim 1, characterized in that The long short-term memory neural network includes: a first input layer, a first hidden layer, and a first output layer; The first input layer includes 8 neurons, the first hidden layer includes 50 neurons, and the first output layer includes 1 neuron.
4. The dynamic prediction method for the power transmission and transformation project volume under power demand fluctuations according to claim 1, wherein The training process of the long short-term memory neural network specifically includes: Obtain multiple power demand data for a historical time period; the power demand data includes the historical power load data of the independent variables affecting power demand and the historical power demand values; Normalize each piece of the power demand data, and form a power demand data set with all the normalized power demand data; Divide the power demand data set into a training data set and a test set; Set training parameters; the training parameters include: a step size of 1, a batch size of 12, and a training epoch of 250; According to the training data set and the training parameters, train the long short-term memory neural network, and use the test set to verify the accuracy of the trained long short-term memory neural network. Determine the long short-term memory neural network with an accuracy greater than or equal to the accuracy threshold as the trained long short-term memory neural network.
5. The dynamic prediction method for the transmission and transformation project quantity under power demand fluctuations according to claim 1, wherein, Use the SelectKBest method to select multiple factors that affect the quantity of transmission and transformation projects as the independent variables affecting the quantity of transmission and transformation projects, specifically including: Obtain the historical data of transmission and transformation projects with a voltage of 35 kV and below within a historical time period; Organize the historical data into multiple attribute indicators; Merge the attribute indicators with the same content to obtain multiple initial indicators that affect the quantity of transmission and transformation projects; Use the F-test mathematical method to conduct a correlation test on the multiple initial indicators to obtain the P value; According to the P-value, the SelectKBest method is used to select multiple indicators that affect the quantity of transmission and transformation engineering from multiple initial indicators as the independent variables affecting the quantity of transmission and transformation engineering; the independent variables affecting the quantity of transmission and transformation engineering include project type, outdoor line length, electricity consumption capacity, number of users, indoor line length, power demand, geographical complexity, and government subsidies.
6. The dynamic prediction method for the power transmission and transformation project quantity under power demand fluctuations according to claim 1, wherein The BP neural network includes: a second input layer, a second hidden layer, and a second output layer; The second input layer includes 8 neurons, the second hidden layer includes 15 neurons, and the second output layer includes 1 neuron.
7. A dynamic prediction device for the quantity of power transmission and transformation projects under power demand fluctuations, characterized in that, It includes: A power demand influencing factor selection module, configured to select multiple factors that affect the power demand fluctuation as the independent variables affecting the power demand; A power demand prediction module, configured to obtain the predicted value of the future power demand time series by using the trained long short-term memory neural network according to the historical power load data of the independent variables affecting the power demand; A project quantity influencing factor selection module, configured to use the SelectKBest method to select multiple factors that affect the quantity of transmission and transformation engineering as the independent variables affecting the quantity of transmission and transformation engineering; the independent variables affecting the quantity of transmission and transformation engineering include power demand; A project quantity prediction module, configured to input the predicted value of the future power demand time series and the future values of the independent variables affecting the quantity of transmission and transformation engineering other than the power demand into the trained BP neural network to obtain the predicted value of the future quantity of transmission and transformation engineering.
8. A computer device, comprising: 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 dynamic prediction method for the quantity of transmission and transformation engineering under power demand fluctuation according to any one of claims 1-6.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the dynamic prediction method for the quantity of transmission and transformation engineering under power demand fluctuation according to any one of claims 1-6.
10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the dynamic prediction method for the quantity of transmission and transformation engineering under power demand fluctuation according to any one of claims 1-6.