Electric power productivity prediction method and system based on artificial intelligence

Abstract

The invention provides an electric power productivity prediction method and system based on artificial intelligence, and belongs to the technical field of electric power. The method comprises the steps of obtaining actual productivity data of a first period of a target area, determining a plurality of reference areas of the target area according to the actual productivity data of the first period, and determining a first productivity prediction model of the target area according to productivity prediction models corresponding to the plurality of reference areas, obtaining first power load data of a first period of the target area and second power load data of a plurality of reference areas, constructing a second productivity prediction model of the target area according to the first power load data and the second power load data, and performing bidirectional model fusion according to the first productivity prediction model and the second productivity prediction model, and obtaining a third productivity prediction model of the target area, and determining a power productivity prediction result of the second period of the target area according to the third productivity prediction model. The objective of the invention is to improve the problem that the power productivity prediction of the newly-built area is not accurate enough.

Description

Technical Field

[0001] The present application relates to the field of electric power technology, and in particular to an artificial intelligence-based electric power production capacity prediction method and system. Background Art

[0002] Power capacity forecasting is the process of estimating and analyzing future power production capacity. By using statistical models, machine learning algorithms and expert evaluations, it helps relevant departments understand the future power supply and demand, so as to formulate reasonable production and investment strategies and ensure the reliability and economy of power supply. This process not only helps prevent power shortages and surpluses, but also supports the optimization of energy structure and the rational layout of renewable energy, thereby promoting the sustainable development of energy.

[0003] In related technologies, when predicting the power production capacity of a target area, a huge amount of data support is often required. However, when predicting the production capacity of a newly built area, traditional data-driven methods are difficult to apply effectively because the newly built areas often have no historical data as a reference. Summary of the invention

[0004] The embodiments of the present application provide an artificial intelligence-based power production capacity prediction method and system for solving the problems in the above-mentioned background technology.

[0005] To achieve the above objectives, the embodiments of the present application adopt the following technical solutions:

[0006] In a first aspect, a method for predicting power production capacity based on artificial intelligence is provided, the method comprising:

[0007] Acquire actual capacity data of the target area in the first cycle, and determine multiple reference areas of the target area according to the actual capacity data of the first cycle;

[0008] Determine a first capacity prediction model for the target area according to the capacity prediction models corresponding to the multiple reference areas, where the first capacity prediction model is used to characterize the power capacity supply-side relationship of the target area;

[0009] Acquire first power load data of a first period of a target area and second power load data of a plurality of reference areas, and construct a second capacity prediction model of the target area according to the first power load data and the second power load data, wherein the first capacity prediction model is used to characterize the power production and demand side relationship of the target area;

[0010] According to the bidirectional model fusion of the first capacity prediction model and the second capacity prediction model, a third capacity prediction model of the target area is obtained;

[0011] According to the third capacity prediction model, the power capacity prediction result of the second period of the target area is determined.

[0012] The power capacity prediction method based on artificial intelligence provided in the present application determines multiple reference areas by using the actual capacity data of the first cycle of the target area. This method provides a benchmark by comparing the target area with the area with rich existing data, so as to use the reference area with similarity to the target area to complete the data improvement and model construction of the target area. Then, the first capacity prediction model of the target area is established using the capacity prediction models corresponding to the multiple reference areas to characterize the supply side relationship. Then, the second capacity prediction model is constructed using the power load data of the first cycle of the target area and the second power load data of the reference area to characterize the demand side relationship. This method not only considers the supply side factors, but also the demand side factors, so that the prediction is more comprehensive. Finally, the third capacity prediction model of the target area is obtained by bidirectionally fusing the first capacity prediction model and the second capacity prediction model. This fusion takes into account the interaction between different factors, so that the prediction results are more accurate. This method comprehensively considers factors from different aspects and makes up for the problem of missing data by fusion of data models, so that it can also be effectively applied when predicting power capacity in newly built areas.

[0013] In a possible implementation, determining multiple reference areas of the target area according to the actual production capacity data of the first cycle includes:

[0014] Taking the location factor and weather factor of the target area as the first evaluation target, determining the first similarity score of each candidate reference area;

[0015] Taking the change rate and stability of the actual capacity data as the second evaluation target, determining the second similarity score of each candidate reference area;

[0016] A plurality of reference regions of the target region are determined according to the first similarity score and the second similarity score.

[0017] In a possible implementation, determining a first capacity prediction model for a target area according to capacity prediction models corresponding to a plurality of reference areas includes:

[0018] Obtaining equipment stability parameters of the power equipment in the target area in the first cycle, and adjusting the capacity prediction models corresponding to the multiple reference areas according to the equipment stability parameters to obtain multiple optimized capacity prediction models;

[0019] A stability assessment is performed on multiple optimized capacity prediction models, and a first capacity prediction model for the target area is determined based on the stability assessment results.

[0020] In a possible implementation, obtaining first power load data of a first period of a target area and second power load data of a plurality of reference areas, and constructing a second capacity prediction model of the target area according to the first power load data and the second power load data includes:

[0021] Performing data processing on the first power load data and the plurality of second power load data to obtain a first data set and a plurality of second data sets;

[0022] Based on the first data set and multiple second data sets, the preset random forest model is trained and cross-validated to obtain a second capacity prediction model.

[0023] In a possible implementation, based on the first data set and the plurality of second data sets, a preset random forest model is trained and cross-validated to obtain a second capacity prediction model, including:

[0024] Constructing multiple sets of training data sets, each set of training data sets includes a first data set and at least two different second data sets;

[0025] The preset random forest model is trained based on the training data set, and the random forest model is verified based on different training data sets to obtain a second capacity prediction model.

[0026] In a possible implementation, the bidirectional model fusion according to the first capacity prediction model and the second capacity prediction model includes:

[0027] According to the positive correlation results between the first capacity prediction model and the second capacity prediction model, a positive correlation relationship between the first capacity prediction model and the second capacity prediction model is established;

[0028] According to the negative correlation result between the first capacity prediction model and the second capacity prediction model, a negative correlation relationship between the first capacity prediction model and the second capacity prediction model is established;

[0029] According to the positive correlation and the negative correlation, the first capacity prediction model and the second capacity prediction model are integrated.

[0030] In a possible implementation, determining the power capacity prediction result of the second period of the target area according to the third capacity prediction model includes:

[0031] Constructing a first input feature based on the estimated weather data of the second period of the target area;

[0032] Constructing a second input feature according to the equipment stability parameters of the electric power equipment in the first period;

[0033] The first input feature and the second input feature are input into a third capacity prediction model to obtain a power capacity prediction result of a second period in the target area.

[0034] In a possible implementation, the method further includes:

[0035] Obtain the demand-side growth characteristics and supply-side growth characteristics of the target area, and determine the supply and demand change characteristics of the target area based on the demand-side growth characteristics and supply-side growth characteristics;

[0036] The third capacity forecasting model is dynamically adjusted based on the changing characteristics of supply and demand.

[0037] In the second aspect, a power production capacity prediction system based on artificial intelligence is provided, the system comprising:

[0038] A first acquisition module, used to acquire actual capacity data of a target area in a first period, and determine a plurality of reference areas of the target area according to the actual capacity data of the first period;

[0039] A first model building module is used to determine a first capacity prediction model for a target area according to capacity prediction models corresponding to a plurality of reference areas, wherein the first capacity prediction model is used to characterize a power capacity supply-side relationship of the target area;

[0040] A second model building module is used to obtain the first power load data of the first period of the target area and the second power load data of multiple reference areas, and to build a second capacity prediction model of the target area according to the first power load data and the second power load data, wherein the first capacity prediction model is used to characterize the power capacity demand side relationship of the target area;

[0041] A model fusion module, used for bidirectionally fusing the first capacity prediction model and the second capacity prediction model to obtain a third capacity prediction model for the target area;

[0042] The prediction module is used to determine the power production capacity prediction result of the second period of the target area according to the third production capacity prediction model.

[0043] In a possible implementation, the first acquisition module includes:

[0044] A first evaluation submodule, for determining a first similarity score for each candidate reference area by taking the location factor and weather factor of the target area as a first evaluation target;

[0045] A second evaluation submodule is used to determine a second similarity score for each candidate reference area by taking the change rate and stability of the actual capacity data as a second evaluation target;

[0046] The screening submodule is used to determine a plurality of reference regions of the target region according to the first similarity score and the second similarity score.

[0047] In a possible implementation, the first model building module includes:

[0048] An optimization submodule, used to obtain equipment stability parameters of the power equipment in the target area in the first cycle, and adjust the capacity prediction models corresponding to multiple reference areas according to the equipment stability parameters to obtain multiple optimized capacity prediction models;

[0049] The evaluation submodule is used to perform stability evaluation on multiple optimized capacity prediction models and determine the first capacity prediction model of the target area according to the stability evaluation results.

[0050] In a possible implementation, the second model building module includes:

[0051] A data processing submodule, used for performing data processing on the first power load data and the plurality of second power load data to obtain a first data set and a plurality of second data sets;

[0052] The model training module is used to train and cross-validate a preset random forest model based on the first data set and multiple second data sets to obtain a second capacity prediction model.

[0053] In a possible implementation, the model training module includes:

[0054] A data set construction unit, used to construct multiple sets of training data sets, each set of training data sets includes a first data set and at least two different second data sets;

[0055] The verification unit is used to train a preset random forest model based on a training data set, verify the random forest model based on different training data sets, and obtain a second capacity prediction model.

[0056] In a possible implementation, the model fusion module includes:

[0057] A first relationship building submodule is used to build a positive correlation relationship between the first capacity prediction model and the second capacity prediction model according to the positive correlation association results between the first capacity prediction model and the second capacity prediction model;

[0058] A second relationship building submodule is used to build a negative correlation relationship between the first capacity prediction model and the second capacity prediction model according to the negative correlation association result between the first capacity prediction model and the second capacity prediction model;

[0059] The fusion module is used to fuse the first capacity prediction model and the second capacity prediction model according to the positive correlation relationship and the negative correlation relationship.

[0060] In a possible implementation, the prediction module includes:

[0061] A first feature construction submodule is used to construct a first input feature according to the estimated weather data of the second period of the target area;

[0062] A second feature construction submodule is used to construct a second input feature according to the device stability parameter of the power equipment in the first period;

[0063] The prediction submodule is used to input the first input feature and the second input feature into a third capacity prediction model to obtain a power capacity prediction result of a second period in the target area.

[0064] In a possible implementation, the system further includes an optimization module:

[0065] It is used to obtain the demand-side growth characteristics and supply-side growth characteristics of the target area, and determine the supply and demand change characteristics of the target area based on the demand-side growth characteristics and supply-side growth characteristics;

[0066] It is used to dynamically adjust the third capacity forecasting model according to the changing characteristics of supply and demand.

[0067] According to a third aspect, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor implements a method as described in any one of the above aspects when executing the program.

[0068] According to a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method according to any one of the above aspects is implemented.

[0069] The technical effects of the second to fourth aspects refer to the technical effects of the aspect and any of its embodiments, which will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 It is a flowchart of the steps of a power production capacity prediction method based on artificial intelligence provided by an embodiment of the present invention;

[0071] Figure 2 It is a functional module diagram of an artificial intelligence-based power production capacity prediction system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0072] In order to make the purpose, technical solutions and advantages of this application more clear, the following Figure 1-Figure 2 It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0073] The terms “first”, “second”, etc. involved in the embodiments of the present application are only used to distinguish features of the same type and cannot be understood as indicating relative importance, quantity, order, etc.

[0074] The terms "exemplary" or "for example" and the like in the embodiments of the present application are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the terms "exemplary" or "for example" is intended to present the related concepts in a specific way.

[0075] The terms "coupling" and "connection" involved in the embodiments of the present application should be understood in a broad sense. For example, they may refer to a direct physical connection, or an indirect connection achieved through electronic devices, such as a connection achieved through resistors, inductors, capacitors or other electronic devices.

[0076] At present, when predicting the power capacity of a target area, it is usually necessary to rely on a large amount of historical data and real-time data to support the prediction model. These data include but are not limited to past power load data, power generation data, weather data, demographic data, economic activity data, etc. By analyzing these data, the laws of regional power demand and supply can be identified, and a reliable power capacity prediction model can be constructed. However, when predicting power capacity in newly built areas, the main challenge is that there is no historical data for reference in these areas. For example, when analyzing the power capacity of newly built wind power stations and hydropower stations, the lack of sufficient data support will have an impact on the power construction in the newly built areas.

[0077] Reference Figure 1 , an embodiment of the present invention provides a method for predicting power production capacity based on artificial intelligence, which may include the following steps:

[0078] S101: Acquire actual production capacity data of a target area in a first period, and determine a plurality of reference areas of the target area according to the actual production capacity data of the first period.

[0079] In this implementation, the first cycle is usually a very short period of time. During the first cycle, the data obtained in the target area is very limited. The main reason for determining the reference area is that when the target area lacks sufficient actual capacity data, by selecting reference areas that are similar to the target area in terms of geography, climate, power demand and supply structure, the capacity data of these areas can be used for analysis and prediction, thereby making up for the lack of data in the target area and providing more accurate power capacity assessment and decision support, which specifically includes:

[0080] S1011: Taking the location factor and weather factor of the target area as the first evaluation target, determining a first similarity score for each candidate reference area;

[0081] S1012: Taking the change rate and stability of the actual capacity data as the second evaluation target, determining a second similarity score for each candidate reference area;

[0082] S1013: Determine a plurality of reference regions of the target region according to the first similarity score and the second similarity score.

[0083] In the implementation of S1011 to S1013, the location factors may include: Geographic location: geographic coordinates (longitude and latitude) of the target area. Terrain features: terrain features of the target area, such as plains, mountainous areas, coastal areas, etc. Weather factors may include: Temperature: average temperature, maximum temperature and minimum temperature of the target area. Precipitation: precipitation of the target area, including annual precipitation, seasonal precipitation, etc. Solar radiation: solar radiation of the target area, especially for areas that rely on solar energy. Wind speed: average wind speed of the target area, which is particularly important for areas that rely on wind energy.

[0084] Obtain location and weather data for each candidate reference area. Normalize the location and weather factor data to be in the same dimension. Calculate the similarity of the location and weather data between the target area and each candidate reference area using Euclidean distance, Manhattan distance, or other suitable distance metrics. Assign a first similarity score to each candidate reference area based on the calculated similarity values, with higher scores indicating higher similarity.

[0085] Taking the rate of change and stability of actual capacity data as the second evaluation target can ensure that the selected reference area is similar to the target area in terms of the dynamic and fluctuation characteristics of power capacity, thereby improving the accuracy of the prediction and the comparability of the data.

[0086] The rate of change reflects the change in power production capacity at different points in time. By comparing the rate of change of the target area and the candidate reference area, it is ensured that the reference area is similar to the target area in terms of power supply and demand patterns. For example, if the power demand in the target area has a significant increase or decrease in certain time periods, selecting reference areas with similar rates of change can more accurately reflect these changes. Stability measures the degree of fluctuation in power production capacity. If the power production capacity of the target area is relatively stable (small fluctuations) and the power production capacity of the candidate area fluctuates greatly (large fluctuations), then these areas do not match in terms of power supply and demand characteristics. Therefore, by comparing the stability, reference areas that are similar to the target area in terms of power production capacity fluctuations can be screened out.

[0087] Similar change rates and stability ensure that the data in the reference area is consistent with the target area in terms of dynamic characteristics, which helps improve the accuracy of the prediction model based on the reference area data. If the capacity change rate and stability of the reference area are consistent with those of the target area, the prediction model can better capture the future power capacity change trend of the target area.

[0088] The similarity of the actual capacity data change rate and stability between the target area and each candidate reference area is calculated using a similarity measurement method, such as Pearson correlation coefficient, Euclidean distance, etc. Based on the calculated similarity value, a second similarity score is assigned to each candidate reference area, and a higher score indicates a higher similarity.

[0089] According to the comprehensive score, all candidate reference areas are sorted. Several areas with the highest comprehensive scores are selected as reference areas for the target area. According to the comprehensive score sorting, several areas with the highest scores are selected as reference areas.

[0090] Through the above steps, the location and weather factors as well as the rate of change and stability of actual capacity data can be used to scientifically determine multiple reference areas in the target area, helping to conduct effective power capacity analysis and forecasting when data is insufficient.

[0091] S102: Determine a first capacity prediction model of a target area according to capacity prediction models corresponding to a plurality of reference areas, where the first capacity prediction model is used to characterize a supply-side relationship of electric power capacity in the target area.

[0092] In this embodiment, the problem of insufficient data in the target area can be effectively compensated by the relevant data and prediction model of the reference area, and the accuracy and reliability of the prediction can be improved. According to the capacity prediction models corresponding to multiple reference areas, the prediction results and characteristics of these models can be integrated to determine the first capacity prediction model of the target area. This model characterizes the power capacity supply-side relationship of the target area by drawing on and integrating the capacity change law and supply-side characteristics of the reference area, so that in the absence of sufficient historical data, the power capacity supply of the target area can still be accurately predicted, providing reliable data support for the planning and management of the power capacity area. The specific steps include:

[0093] S1021: Obtain equipment stability parameters of the power equipment in the target area in the first period, and adjust the capacity prediction models corresponding to multiple reference areas according to the equipment stability parameters to obtain multiple optimized capacity prediction models;

[0094] S1022: Perform stability evaluation on multiple optimized capacity prediction models, and determine a first capacity prediction model for the target area based on the stability evaluation results.

[0095] In the implementation of S1021 to S1022, the equipment stability parameters may include equipment reliability, failure rate, maintenance frequency, service life, performance fluctuations, etc. By obtaining the equipment stability parameters of the power equipment in the target area in the first cycle, the operating stability and reliability of the equipment in the cycle can be understood, and then these stability parameters are used to adjust the capacity prediction models corresponding to multiple reference areas to obtain multiple optimized capacity prediction models. Then, through statistical analysis and modeling, the relationship between the equipment stability parameters and the power production capacity is determined. For example, the failure rate and maintenance frequency of the equipment may have a negative impact on the power production capacity, while the load rate may affect the volatility of the production capacity. Extract the key stability parameters and their influencing factors, and quantify the degree of influence of these parameters on the power production capacity. According to the analysis results, the equipment stability parameters of the target area are applied to the capacity prediction model of the reference area. This can be achieved by adjusting the model's input parameters, adding characteristic variables related to stability parameters, or directly modifying the model structure, conducting stability assessments on these optimized capacity forecasting models, analyzing their prediction accuracy and consistency under different conditions, and ultimately selecting the most stable and accurate model based on the stability assessment results to determine the first capacity forecasting model for the target area, thereby ensuring that the model can more accurately characterize the target area's power capacity supply relationship and provide reliable data support for the planning and management of the power system.

[0096] S103: Acquire first power load data of a first period of a target area and second power load data of a plurality of reference areas, and construct a second capacity prediction model for the target area according to the first power load data and the second power load data.

[0097] In this embodiment, the load data of the reference area is used to make up for the deficiency of the target area data, and through the similarity and correlation analysis between the load data, a second capacity prediction model that can reflect the change trend and characteristics of the power demand in the target area is constructed, so as to achieve more accurate prediction and effective management of the power capacity supply in the target area. By integrating the actual power load characteristics of the target area and the load data of the reference area, more information sources can be fully utilized to improve the accuracy and robustness of the model. The first power load data of the first period of the target area and the second power load data of multiple reference areas are obtained, and the second capacity prediction model of the target area is constructed according to the first power load data and the second power load data. It includes:

[0098] S1031: performing data processing on the first power load data and the plurality of second power load data to obtain a first data set and a plurality of second data sets;

[0099] S1032: Based on the first data set and the multiple second data sets, train and cross-validate the preset random forest model to obtain a second capacity prediction model.

[0100] In the implementation of S1031 to S1032, the first power load data and the plurality of second power load data are first processed to make them suitable for the machine learning model. Steps such as data cleaning, missing value processing, and feature engineering may be included. The ultimate goal is to generate a first data set and a plurality of second data sets, which will be used for subsequent model training. The model is trained using the first data set and the plurality of second data sets. The preset model is a random forest model, which is an integrated learning method consisting of multiple decision trees. The model will be trained based on the provided data, and cross-validation will be used to evaluate the performance of the model. Cross-validation is to divide the data into multiple subsets, and use one part of them as a test set and the other parts as a training set in turn. The ultimate goal is to obtain a second capacity prediction model that can predict capacity based on the input power load data.

[0101] The overall goal of this process is to build a model that can predict power generation based on power load data. Through data processing and model training, the accuracy and reliability of the model can be improved, so that it can better respond to actual power demand and supply conditions.

[0102] In a feasible implementation, based on the first data set and the plurality of second data sets, a preset random forest model is trained and cross-validated to obtain a second capacity prediction model, including:

[0103] Constructing multiple sets of training data sets, each set of training data sets includes a first data set and at least two different second data sets;

[0104] The preset random forest model is trained based on the training data set, and the random forest model is verified based on different training data sets to obtain a second capacity prediction model.

[0105] In this embodiment, the first data set refers to the actual power production capacity data of the target area in the first period. The second data set refers to the power production capacity data from multiple reference areas. Each training data set should include data of the target area (first data set) and data of at least two different reference areas (second data sets).

[0106] For example, assuming there are N second data sets, the following multiple groups of training data sets can be constructed: the first group of training data sets: the first data set + the second data set 1 + the second data set 2; the second group of training data sets: the first data set + the second data set 1 + the second data set 3; the third group of training data sets: the first data set + the second data set 2 + the second data set 3.

[0107] For each set of training data sets, the random forest algorithm is used to train the model. The training steps are as follows: Input: training data set (data of the target area and data of the reference area). Output: trained random forest model. Use cross-validation technology to verify the trained model to evaluate its performance. The specific steps are as follows: Divide each set of training data sets into K subsets (such as K-fold cross-validation). Use K-1 subsets for model training each time, and the remaining 1 subset for verification. Repeat K times to ensure that each subset is used as a verification set once. Record the results of each verification and calculate the average performance indicators of the model, such as mean square error (MSE), R 2 wait.

[0108] After multiple training and verification based on different training data sets, the model with the best performance is selected as the final second capacity prediction model. Through the above steps, the data of the target area and the reference area can be used to train and verify the random forest model, thereby obtaining an accurate power capacity prediction model. By integrating multiple data, the reliability and accuracy of the prediction can be improved.

[0109] S104: Obtain a third capacity prediction model for the target area by bidirectionally fusing the first capacity prediction model and the second capacity prediction model.

[0110] In this embodiment, after obtaining the first capacity forecasting model and the second capacity forecasting model, it is necessary to merge the two models from both positive and negative directions to obtain a third capacity forecasting model that can more comprehensively reflect the supply and demand relationship. The specific steps include:

[0111] S1041: constructing a positive correlation relationship between the first capacity prediction model and the second capacity prediction model according to the positive correlation results between the first capacity prediction model and the second capacity prediction model;

[0112] S1042: constructing a negative correlation relationship between the first capacity prediction model and the second capacity prediction model according to the negative correlation result between the first capacity prediction model and the second capacity prediction model;

[0113] S1043: According to the positive correlation relationship and the negative correlation relationship, the first capacity prediction model and the second capacity prediction model are integrated.

[0114] In the implementation of S1041 to S1042, first, it is necessary to collect the prediction results of the first capacity prediction model and the second capacity prediction model. These data may include the predicted capacity values ​​at various time points. Next, the positive correlation between the two models is analyzed. A positive correlation means that when the predicted value of one model increases, the predicted value of the other model will also increase. This relationship can be measured by calculating the correlation coefficient. Although a specific formula is not required here, the positive correlation can be confirmed by observing data trends or using statistical tools. Once the positive correlation is confirmed, a positive correlation relationship can be constructed. This means that when using the prediction model, the positive correlation of the two models can be taken into account. For example, when the predicted value of the first model increases, it can be reasonably expected that the predicted value of the second model will also increase accordingly. This positive correlation can be used to adjust and optimize the prediction results of the model and improve the accuracy of the prediction.

[0115] Similar to a positive correlation, a negative correlation means that when the predicted value of one model increases, the predicted value of the other model decreases. By observing the data or using statistical analysis, it is possible to determine whether this negative correlation exists between the two models. Once a negative correlation is determined, a negative correlation can be constructed. This means that during the forecasting process, this inverse relationship between the two models needs to be taken into account. For example, when the predicted value of the first model increases, it can be expected that the predicted value of the second model may decrease. This allows for better adjustment of the forecasting model to ensure accurate forecasts of capacity changes. After constructing positive and negative correlations, the next step is to integrate these relationships into the forecasting model. This means that when making forecasts, both the impact of positive correlations on the forecast results and the adjustment of negative correlations on the forecast results should be considered. By fusing positive and negative correlations, the forecast results of the two models can be optimized. Specific methods can include weighted averaging of the forecast results of the two models, adjusting model parameters, etc. to achieve more accurate forecasts. This fusion can improve the accuracy and reliability of the overall forecasting model.

[0116] Through the above steps, the correlation between the two capacity forecasting models can be effectively utilized to build a more accurate and reliable comprehensive forecasting model. This not only improves the accuracy of the forecast, but also provides stronger support for decision-making.

[0117] S105: Determine the power capacity prediction result of the second period of the target area according to the third capacity prediction model.

[0118] In this embodiment, after obtaining the third capacity prediction model that satisfies the supply side and the demand side, a long-term prediction of the power capacity of the target area can be performed based on the third capacity prediction model. The specific steps include:

[0119] Constructing a first input feature based on the estimated weather data of the second period of the target area;

[0120] Constructing a second input feature according to the equipment stability parameters of the electric power equipment in the first period;

[0121] The first input feature and the second input feature are input into a third capacity prediction model to obtain a power capacity prediction result of a second period in the target area.

[0122] In this embodiment, weather forecast data for the second period of the target area is obtained, including temperature, humidity, wind speed, sunshine duration, precipitation, etc. These weather data are organized into structured input features for input into the third capacity prediction model. The equipment stability parameters of the power equipment in the first period are obtained, and these equipment parameters are organized into input features for the capacity prediction model. The stability and efficiency of the equipment have a direct impact on the power supply. The first input feature (weather data) and the second input feature (equipment stability parameter) are integrated into a comprehensive input feature set. The integrated input feature set is input into the third capacity prediction model. Through model calculation, the power capacity prediction result of the second period of the target area is obtained. The second period is a longer time period relative to the first period.

[0123] In a feasible implementation, the method further includes:

[0124] Obtain the demand-side growth characteristics and supply-side growth characteristics of the target area, and determine the supply and demand change characteristics of the target area based on the demand-side growth characteristics and supply-side growth characteristics;

[0125] The third capacity forecasting model is dynamically adjusted based on the changing characteristics of supply and demand.

[0126] In this embodiment, the demand growth characteristics may include factors such as population growth, economic development, technological progress, and seasonal changes. By analyzing these factors, the trend and fluctuation of future electricity demand can be understood. For example, electricity demand usually increases in summer and winter due to the use of air conditioning and heating equipment. The electricity demand in industrial areas may fluctuate with the production plan and capacity changes of the factory. The supply growth characteristics involve factors such as power plant construction, energy policy changes, and equipment upgrades. By analyzing these factors, changes in electricity supply capacity can be understood. For example, an increase in the proportion of renewable energy sources such as wind power and solar power may lead to an increase in the volatility of electricity supply, and other power sources need to be flexibly adjusted to maintain a balance between supply and demand.

[0127] By integrating the growth characteristics of the demand side and the supply side, the changing characteristics of electricity supply and demand are determined to help identify the supply and demand imbalance that may occur in the future. According to the changing characteristics of supply and demand, the power capacity forecasting model is dynamically adjusted to more accurately predict future power demand and supply. For example, if it is predicted that the demand for electricity will increase significantly in a certain period of time in the future, it is possible to plan and increase the power generation capacity in advance, or strengthen the demand side management measures, such as implementing the peak power price policy to guide users to use electricity during the off-peak period.

[0128] By comprehensively analyzing the growth characteristics of the demand and supply sides, we can more accurately predict changes in electricity demand and supply, reduce forecast errors, and improve the planning and management level of the power system. Identifying supply and demand change trends in advance and taking corresponding measures can enhance the flexibility and reliability of the power system, avoid power shortages or surpluses, and ensure the stability of power supply.

[0129] The power capacity prediction method based on artificial intelligence provided in the present application determines multiple reference areas by using the actual capacity data of the first cycle of the target area. This method provides a benchmark by comparing the target area with the area with rich existing data. Then, the first capacity prediction model of the target area is established using the capacity prediction models corresponding to the multiple reference areas to characterize the supply-side relationship. Next, the second capacity prediction model is constructed using the power load data of the first cycle of the target area and the second power load data of the reference area to characterize the demand-side relationship. This method not only considers the supply-side factors, but also the demand-side factors, making the prediction more comprehensive. Finally, the third capacity prediction model of the target area is obtained by bidirectionally fusing the first capacity prediction model and the second capacity prediction model. This fusion takes into account the interaction between different factors, making the prediction result more accurate. This method comprehensively considers factors from different aspects and makes up for the problem of missing data by fusing data models, so that it can also be effectively applied when predicting power capacity in newly built areas.

[0130] The embodiment of the present invention also provides an artificial intelligence-based power production capacity prediction system, referring to Figure 2 , shows a functional module diagram of an artificial intelligence-based power production capacity prediction system of the present invention, which may include the following modules:

[0131] A first acquisition module 201 is used to acquire actual capacity data of a target area in a first period, and determine a plurality of reference areas of the target area according to the actual capacity data of the first period;

[0132] A first model building module 202 is used to determine a first capacity prediction model for a target area according to capacity prediction models corresponding to a plurality of reference areas, wherein the first capacity prediction model is used to characterize a power capacity supply-side relationship of the target area;

[0133] A second model building module 203 is used to obtain the first power load data of the first period of the target area and the second power load data of multiple reference areas, and to build a second capacity prediction model of the target area according to the first power load data and the second power load data, wherein the first capacity prediction model is used to characterize the power capacity demand side relationship of the target area;

[0134] A model fusion module 204 is used to obtain a third capacity prediction model for the target area according to bidirectional model fusion of the first capacity prediction model and the second capacity prediction model;

[0135] The prediction module 205 is used to determine the power production capacity prediction result of the second period of the target area according to the third production capacity prediction model.

[0136] In a possible implementation, the first acquisition module includes:

[0137] A first evaluation submodule, for determining a first similarity score for each candidate reference area by taking the location factor and weather factor of the target area as a first evaluation target;

[0138] A second evaluation submodule is used to determine a second similarity score for each candidate reference area by taking the change rate and stability of the actual capacity data as a second evaluation target;

[0139] The screening submodule is used to determine a plurality of reference regions of the target region according to the first similarity score and the second similarity score.

[0140] In a possible implementation, the first model building module includes:

[0141] An optimization submodule, used to obtain equipment stability parameters of the power equipment in the target area in the first cycle, and adjust the capacity prediction models corresponding to multiple reference areas according to the equipment stability parameters to obtain multiple optimized capacity prediction models;

[0142] The evaluation submodule is used to perform stability evaluation on multiple optimized capacity prediction models and determine the first capacity prediction model of the target area according to the stability evaluation results.

[0143] In a possible implementation, the second model building module includes:

[0144] A data processing submodule, used for performing data processing on the first power load data and the plurality of second power load data to obtain a first data set and a plurality of second data sets;

[0145] The model training module is used to train and cross-validate a preset random forest model based on the first data set and multiple second data sets to obtain a second capacity prediction model.

[0146] In a possible implementation, the model training module includes:

[0147] A data set construction unit, used to construct multiple sets of training data sets, each set of training data sets includes a first data set and at least two different second data sets;

[0148] The verification unit is used to train a preset random forest model based on a training data set, verify the random forest model based on different training data sets, and obtain a second capacity prediction model.

[0149] In a possible implementation, the model fusion module includes:

[0150] A first relationship building submodule is used to build a positive correlation relationship between the first capacity prediction model and the second capacity prediction model according to the positive correlation association results between the first capacity prediction model and the second capacity prediction model;

[0151] A second relationship building submodule is used to build a negative correlation relationship between the first capacity prediction model and the second capacity prediction model according to the negative correlation association result between the first capacity prediction model and the second capacity prediction model;

[0152] The fusion module is used to fuse the first capacity prediction model and the second capacity prediction model according to the positive correlation relationship and the negative correlation relationship.

[0153] In a possible implementation, the prediction module includes:

[0154] A first feature construction submodule is used to construct a first input feature according to the estimated weather data of the second period of the target area;

[0155] A second feature construction submodule is used to construct a second input feature according to the device stability parameter of the power equipment in the first period;

[0156] The prediction submodule is used to input the first input feature and the second input feature into a third capacity prediction model to obtain a power capacity prediction result of a second period in the target area.

[0157] In a possible implementation, the system further includes an optimization module:

[0158] It is used to obtain the demand-side growth characteristics and supply-side growth characteristics of the target area, and determine the supply and demand change characteristics of the target area based on the demand-side growth characteristics and supply-side growth characteristics;

[0159] It is used to dynamically adjust the third capacity forecasting model according to the changing characteristics of supply and demand.

[0160] Based on the same inventive concept, another embodiment of the present invention provides an electronic device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus.

[0161] Memory, used to store computer programs;

[0162] The processor is used to implement the power production capacity prediction method based on artificial intelligence of the present invention when executing the program stored in the memory.

[0163] The communication bus mentioned in the above terminal can be a peripheral component interconnect standard (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. The communication interface is used for communication between the above terminal and other devices. The memory may include a random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Optionally, the memory may also be at least one storage system located away from the aforementioned processor.

[0164] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0165] In addition, to achieve the above-mentioned purpose, an embodiment of the present invention further proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the artificial intelligence-based power production capacity prediction method of an embodiment of the present invention.

[0166] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, embodiments of the present invention may take the form of complete hardware embodiments, complete software embodiments, or embodiments combining software and hardware aspects. Moreover, embodiments of the present invention may take the form of a computer program product implemented on one or more computer-usable vehicles (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0167] The embodiments of the present invention are described with reference to flowcharts and / or block diagrams of methods, terminal devices (systems), and computer program products according to the 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 the 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 terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal 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 system that specifies the functions of a box or multiple boxes.

[0168] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction system, which is implemented in the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0169] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement 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.

[0170] Finally, it should be noted that, in this article, relational terms such as and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. "" and / or "" indicate that either one of the two can be selected, or both can be selected. Moreover, the terms "" include "" "" contain "" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "" include a..."" does not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0171] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for predicting power production capacity based on artificial intelligence, characterized in that: The method comprises: Acquire actual capacity data of a target area in a first cycle, and determine a plurality of reference areas of the target area according to the actual capacity data of the first cycle; Determining a first capacity prediction model for the target area according to the capacity prediction models corresponding to the multiple reference areas, wherein the first capacity prediction model is used to characterize the power capacity supply-side relationship of the target area; Acquire first power load data of a first period of a target area and second power load data of a plurality of reference areas, and construct a second capacity prediction model of the target area according to the first power load data and the second power load data, wherein the first capacity prediction model is used to characterize the relationship between power production and demand of the target area; Obtaining a third capacity prediction model for the target area by bidirectionally fusing the first capacity prediction model and the second capacity prediction model; According to the third capacity prediction model, the power capacity prediction result of the second period of the target area is determined.

2. The power production capacity prediction method based on artificial intelligence according to claim 1 is characterized in that: The determining of the plurality of reference areas of the target area according to the actual production capacity data of the first cycle includes: Taking the location factor and weather factor of the target area as the first evaluation target, determining a first similarity score for each candidate reference area; Taking the change rate and stability of the actual capacity data as the second evaluation target, determining the second similarity score of each candidate reference area; A plurality of reference regions of the target region are determined according to the first similarity score and the second similarity score.

3. The power production capacity prediction method based on artificial intelligence according to claim 1 is characterized in that: The determining, according to the capacity prediction models corresponding to the multiple reference areas, a first capacity prediction model of the target area includes: Acquire equipment stability parameters of the target area power equipment in the first cycle, and adjust the capacity prediction models corresponding to the multiple reference areas according to the equipment stability parameters to obtain multiple optimized capacity prediction models; A stability assessment is performed on the multiple optimized capacity prediction models, and a first capacity prediction model for the target area is determined according to the stability assessment result.

4. The method for predicting power production capacity based on artificial intelligence according to claim 1, characterized in that: The acquiring first power load data of a first period of a target area and second power load data of a plurality of reference areas, and constructing a second capacity prediction model of the target area according to the first power load data and the second power load data comprises: Performing data processing on the first power load data and the plurality of second power load data to obtain a first data set and a plurality of second data sets; Based on the first data set and multiple second data sets, a preset random forest model is trained and cross-validated to obtain the second capacity prediction model.

5. The power production capacity prediction method based on artificial intelligence according to claim 1 is characterized in that: The step of training and cross-validating a preset random forest model based on the first data set and the plurality of second data sets to obtain the second capacity prediction model includes: Constructing multiple sets of training data sets, each set of the training data sets includes the first data set and at least two different second data sets; A preset random forest model is trained based on the training data set, and the random forest model is verified based on different training data sets to obtain the second capacity prediction model.

6. The method for predicting power production capacity based on artificial intelligence according to claim 1, characterized in that: The bidirectional model fusion according to the first capacity prediction model and the second capacity prediction model includes: According to the positive correlation result between the first capacity prediction model and the second capacity prediction model, a positive correlation relationship between the first capacity prediction model and the second capacity prediction model is established; According to the negative correlation result between the first capacity prediction model and the second capacity prediction model, a negative correlation relationship between the first capacity prediction model and the second capacity prediction model is established; The first capacity prediction model and the second capacity prediction model are fused according to the positive correlation and the negative correlation.

7. The power production capacity prediction method based on artificial intelligence according to claim 1 is characterized in that: Determining the power capacity prediction result of the second period of the target area according to the third capacity prediction model includes: constructing a first input feature according to the estimated weather data of the target area in the second period; constructing a second input feature according to the device stability parameter of the electric power equipment in the first cycle; The first input feature and the second input feature are input into the third capacity prediction model to obtain the power capacity prediction result of the second period of the target area.

8. The method for predicting power production capacity based on artificial intelligence according to claim 1, characterized in that: The method further comprises: Obtaining demand-side growth characteristics and supply-side growth characteristics of the target area, and determining supply and demand change characteristics of the target area based on the demand-side growth characteristics and the supply-side growth characteristics; The third capacity forecasting model is dynamically adjusted according to the supply and demand change characteristics.

9. An artificial intelligence-based power production capacity prediction system, characterized in that: The system comprises: A first acquisition module, configured to acquire actual capacity data of a target area in a first period, and determine a plurality of reference areas of the target area according to the actual capacity data of the first period; A first model building module, configured to determine a first capacity prediction model for the target area according to the capacity prediction models corresponding to the multiple reference areas, wherein the first capacity prediction model is used to characterize a power capacity supply-side relationship of the target area; A second model building module, used for acquiring first power load data of a first period of a target area and second power load data of a plurality of reference areas, and building a second capacity prediction model of the target area according to the first power load data and the second power load data, wherein the first capacity prediction model is used for characterizing a power capacity demand-side relationship of the target area; A model fusion module, configured to obtain a third capacity prediction model for the target area by bidirectionally fusing the first capacity prediction model and the second capacity prediction model; The prediction module is used to determine the power production capacity prediction result of the second period of the target area according to the third production capacity prediction model.

10. The power production capacity prediction system based on artificial intelligence according to claim 9, characterized in that: The first acquisition module includes: A first evaluation submodule, configured to determine a first similarity score for each candidate reference area by taking the location factor and weather factor of the target area as a first evaluation target; A second evaluation submodule is used to determine a second similarity score for each candidate reference area by taking the change rate and stability of the actual capacity data as a second evaluation target; A screening submodule is used to determine a plurality of reference areas of the target area according to the first similarity score and the second similarity score.

Images