Power marketing inspection index correlation analysis method and system based on data mining

Abstract

The invention relates to the technical field of power systems and data mining, and particularly discloses a power marketing inspection index correlation analysis method and system based on data mining, and the method comprises the steps: carrying out the real-time monitoring of the real-time power utilization data of a user side and the energy loss data in a power transmission process in a power marketing monitoring period, and carrying out the real-time monitoring of the real-time power utilization data; an intelligent electric meter and a sensor network are utilized to ensure the accuracy and timeliness of data, fast Fourier transform is applied to calculate electricity consumption abnormal characteristic values of a user, a k-means clustering algorithm is used to analyze energy loss variation amplitude and calculate energy loss characteristic values, the characteristic values are constructed into comprehensive characteristic vectors to serve as input of a random forest model, and the random forest model is constructed into a random forest model. Based on the influence score, the system automatically generates an optimization instruction to guide an electric power company to adjust an electricity price structure or improve the service response speed.

Description

Technical Field

[0001] This invention relates to the fields of power system and data mining technology, specifically to a method and system for correlation analysis of power marketing audit indicators based on data mining. Background Technology

[0002] With the rapid development of the power industry and the widespread application of smart grid technology, power marketing management faces increasing challenges. Traditional power marketing audits rely primarily on manual inspections and simple data analysis methods, which are not only inefficient but also ill-equipped to handle the complexity and diversity of massive amounts of data. Especially in areas such as real-time monitoring of user electricity consumption behavior, assessing energy losses during transmission, and predicting the effectiveness of power marketing, traditional methods fall short. To improve the operational efficiency and service quality of power companies, a more intelligent and systematic solution is urgently needed to address these complex issues.

[0003] The existing technology has the following shortcomings: Most existing systems lack the ability to deeply analyze user electricity consumption behavior and energy loss during power transmission, typically remaining at a superficial data statistics level and failing to effectively identify potential abnormal patterns or energy loss problems. Secondly, existing technologies often face challenges in processing large-scale data, such as high computational resource consumption and slow processing speed, limiting real-time monitoring and rapid response capabilities. Furthermore, most existing decision support systems rely on fixed rules or simple models, lacking adaptability and flexibility, making it difficult to dynamically adjust marketing strategies based on actual conditions. Therefore, existing technologies still have significant room for improvement in enhancing electricity marketing efficiency, reducing operating costs, and improving customer satisfaction. This invention addresses these issues by utilizing advanced technologies such as Fast Fourier Transform, k-means clustering algorithm, and random forest model to provide a comprehensive, efficient, and intelligent correlation analysis method and system for electricity marketing audit indicators. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for correlation analysis of electricity marketing audit indicators based on data mining, so as to solve the problems mentioned above.

[0005] The objective of this invention can be achieved through the following technical solutions: A correlation analysis method for electricity marketing audit indicators based on data mining includes the following steps: S1: During the monitoring period of electricity marketing, real-time monitoring is conducted on the real-time electricity consumption data at the user end and the energy loss data during the power transmission process. S2: Based on the degree of change in real-time electricity consumption data, determine the abnormal electricity consumption characteristics of users and assess the stability of users' electricity consumption behavior; S3: Analyze the variation of energy loss at different time periods during power transmission, calculate the characteristic value of energy loss, and evaluate the stability of power transmission efficiency; S4: Based on the abnormal electricity consumption characteristics of users and the energy loss characteristics during power transmission, establish a model to assess the impact of electricity marketing effectiveness, and determine the degree of impact of real-time electricity consumption and energy loss during power transmission on the effectiveness of electricity marketing based on the model output. S5: Based on the impact analysis results, determine whether the current power system needs to adjust its marketing strategy.

[0006] As a further aspect of the present invention: the evaluation of the stability of user electricity consumption behavior specifically includes: During the monitoring period of electricity marketing, real-time electricity consumption data of users is collected according to the time series. Based on the fluctuation range of real-time electricity consumption data, abnormal electricity consumption characteristic values ​​are calculated. It is determined whether the abnormal electricity consumption characteristic values ​​are greater than or equal to the preset threshold. If so, the user's electricity consumption behavior is unstable; otherwise, the user's electricity consumption behavior is stable.

[0007] As a further aspect of the present invention: the process for obtaining the abnormal electricity consumption characteristic value is as follows: During the electricity marketing monitoring period, real-time electricity consumption data from the user end is collected according to time series. Data collection was conducted, including... Representing discrete points in time, This represents the total number of data collection points. The real-time electricity consumption data is subjected to mean-reduction processing to eliminate the influence of DC component, resulting in preprocessed electricity consumption data. The calculation expression is: ,in, This represents the average of real-time electricity consumption data; Preprocessed electricity consumption data The frequency domain representation is obtained by applying the Fast Fourier Transform. ,in, Represents a frequency index, ranging from 0 to... ; The energy of each frequency component is calculated using the following expression: ;in, Represents frequency The power spectral density at a given point is calculated by summing the energy of all frequency components to obtain the total energy. The sum of the mean and standard deviation of the energy of all frequency components is then calculated to obtain the abnormal energy threshold. The ratio of the sum of the energy of all frequency components greater than the abnormal energy threshold to the total energy is then calculated to obtain the abnormal power consumption characteristic value.

[0008] As a further aspect of the present invention: the evaluation of the stability of power transmission efficiency specifically includes: During the monitoring period of electricity marketing, energy loss data in the power transmission process is collected according to the time series. Based on the change in energy loss in different time periods during the power transmission process, the energy loss characteristic value is calculated. It is then determined whether the energy loss characteristic value is greater than or equal to a preset threshold. If it is, the power transmission efficiency is unstable; otherwise, the power transmission efficiency is stable.

[0009] As a further aspect of the present invention: the process for obtaining the energy loss characteristic value is as follows: During the electricity marketing monitoring period, energy loss data during the power transmission process are analyzed in time series. Data collection was conducted, including... Representing discrete points in time, This represents the total number of data collection points. The real-time energy loss data is subjected to mean removal processing to eliminate the influence of offset, resulting in preprocessed energy loss data. The calculation expression is: ;in, This represents the mean of real-time energy loss data; Determine the total number of clusters Initialize cluster centers and select them randomly. 1 sample is used as the initial cluster center; For each energy loss data point, calculate its distance to all cluster centers and assign it to the nearest cluster center. The expression for calculating the distance to the cluster center is as follows: ;in, Indicates the first Cluster centers, Indicates the number of cluster centers. Indicates energy loss data and the first The distance between cluster centers Represents energy loss data The 1 eigenvalue, Indicates the number of eigenvalues. Representing cluster centers The One eigenvalue; Update the cluster centers by recalculating the new center position for each cluster, which is then used as the average of all samples in that cluster. The calculation expression is as follows: ;in, Indicates the updated number Cluster centers, Indicates the first The set of all energy loss data for each cluster. This indicates the number of energy loss data in the set; the allocation and update steps are repeated until the cluster centers reach the preset maximum number of iterations. Based on the final clustering results, the standard deviation of each cluster is calculated, and the standard deviations of all clusters are summed to obtain the energy loss characteristic value.

[0010] As a further aspect of the present invention: the establishment of an electricity marketing effectiveness impact model based on the user's abnormal electricity consumption characteristics and the energy loss characteristics during power transmission specifically includes: The system acquires the abnormal electricity consumption characteristics of users and the energy loss characteristics during power transmission. It constructs a comprehensive feature vector from the abnormal electricity consumption characteristics and the energy loss characteristics, which serves as the input to the power marketing effect impact model. The training objective of the power marketing effect impact model is to minimize the error between the predicted impact score and the actual impact score. Based on the trained power marketing effect impact model, the system outputs the impact score. The power marketing effect impact model is a random forest model.

[0011] As a further aspect of the present invention: the training process of the power marketing effectiveness impact model is as follows: Historical comprehensive feature vectors and influence scores are obtained as training datasets to train the random forest model. In each iteration, the model generates multiple decision trees, each learning based on different sample subsets and feature subsets. The final prediction result is obtained by voting or averaging the results of these decision trees. The model performance is optimized by adjusting model parameters such as the number of trees and the maximum depth to ensure that it can accurately predict the influence score of electricity marketing effectiveness.

[0012] As a further aspect of the present invention: the determination of the impact of energy loss during real-time electricity consumption and transmission on the effectiveness of electricity marketing specifically includes: The impact score of the electricity marketing effect output by the electricity marketing effect impact model is compared with a preset threshold. If the impact score is greater than or equal to the preset threshold, it is determined whether the impact score is greater than or equal to the preset threshold. If it is, the energy loss in real-time electricity consumption and transmission process has a serious impact on the electricity marketing effect. If not, the energy loss in real-time electricity consumption and transmission process has a slight impact on the electricity marketing effect.

[0013] As a further aspect of the present invention: the step of determining whether the current power system needs to adjust its marketing strategy based on the impact analysis results specifically includes: Given that real-time electricity consumption and energy losses during power transmission have a significant impact on the effectiveness of electricity marketing, the current power system needs to adjust its marketing strategies. This includes the system automatically generating optimization instructions, adjusting the electricity price structure, and improving service response speed.

[0014] A data mining-based correlation analysis system for electricity marketing audit indicators includes: The data acquisition module monitors real-time electricity consumption data at the user end and energy loss data during power transmission during the power marketing monitoring period. The electricity consumption behavior assessment module determines the abnormal electricity consumption characteristics of users and assesses the stability of users' electricity consumption behavior based on the degree of change in real-time electricity consumption data. The power transmission efficiency evaluation module analyzes the energy loss variation at different time points during the power transmission process, calculates energy loss characteristic values, and evaluates the stability of power transmission efficiency. The impact degree analysis module establishes an impact model on the effectiveness of electricity marketing based on the abnormal electricity consumption characteristics of users and the energy loss characteristics during power transmission. Based on the model output, it judges the degree of impact of real-time electricity consumption and energy loss during power transmission on the effectiveness of electricity marketing. The adjustment module determines whether the current power system needs to adjust its marketing strategy based on the results of the impact analysis.

[0015] The beneficial effects of this invention are: (1) By applying Fast Fourier Transform (FFT) to perform frequency domain analysis on real-time electricity consumption data at the user end and using the k-means clustering algorithm to process energy loss data during power transmission, this invention can accurately identify abnormal patterns in user electricity consumption behavior and fluctuations in energy loss over different time periods. Specifically, the FFT technology not only eliminates the influence of the DC component, ensuring the consistency and accuracy of electricity consumption data, but also reveals potential abnormal electricity consumption behavior by quantifying the energy contribution of each frequency component. At the same time, the k-means clustering algorithm identifies fluctuation characteristics within different patterns through mean-reduction processing and cluster analysis of energy loss data during power transmission, providing a quantitative assessment of the stability of power transmission efficiency. Based on these detailed data analysis results, this invention further constructs a comprehensive feature vector, which is used as input to a random forest model to predict the impact score of electricity marketing effectiveness. This method not only significantly improves the sensitivity and accuracy of anomaly detection, enabling power companies to promptly identify and address potential problems, such as abnormal electricity consumption behavior on the user side or high energy loss areas during power transmission, but also provides a scientific basis for optimizing power grid facilities and reducing energy loss. Ultimately, this systematic approach helps improve the overall operational efficiency and service quality of the power system, promotes efficient resource allocation and sustainable development, and provides a solid foundation for power companies to make quick and accurate decisions in a complex and ever-changing market environment.

[0016] (2) This invention is not limited to data collection and analysis, but also has the ability to automatically generate optimization instructions based on the results of the impact analysis, thereby guiding power companies to implement efficient and targeted adjustments to marketing and service strategies. In areas where abnormal electricity consumption behavior is detected, the system can intelligently adjust the electricity price structure to incentivize more reasonable electricity consumption patterns; and when areas with high energy loss are identified, it will recommend improving service response speed and carrying out necessary maintenance and upgrades. This decision support mechanism based on deep data analysis provides power companies with a fast and accurate basis for decision-making in a complex and ever-changing market environment. In addition, for different abnormal situations, this invention can also provide personalized energy-saving suggestions or strengthen supervision, effectively reducing default risks and improving electricity bill collection rates. Through these refined strategy adjustments, this invention not only significantly improves the operational efficiency and customer satisfaction of power companies, but also promotes the optimal allocation and sustainable development of power resources. Specifically, by accurately analyzing real-time electricity consumption data and transmission loss data, and combining advanced machine learning models to predict the impact score of electricity marketing effectiveness, this invention achieves early detection and timely intervention of potential problems, enhancing the reliability and stability of power grid facilities. Ultimately, this approach not only improved the market competitiveness and adaptability of power companies, but also laid a solid foundation for a more intelligent and efficient power management system, thus promoting the progress and development of the entire industry. Attached Figure Description

[0017] The invention will now be further described with reference to the accompanying drawings.

[0018] Figure 1 This is a flowchart illustrating the specific steps of the correlation analysis method for electricity marketing audit indicators based on data mining, as described in this invention. Figure 2 This is a flowchart of the correlation analysis system for electricity marketing audit indicators based on data mining in this invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see Figure 1 As shown, this invention is a correlation analysis method for electricity marketing audit indicators based on data mining, including the following steps: S1: During the monitoring period of electricity marketing, real-time monitoring is conducted on the real-time electricity consumption data at the user end and the energy loss data during the power transmission process. S2: Based on the degree of change in real-time electricity consumption data, determine the abnormal electricity consumption characteristics of users and assess the stability of users' electricity consumption behavior; S3: Analyze the variation of energy loss at different time periods during power transmission, calculate the characteristic value of energy loss, and evaluate the stability of power transmission efficiency; S4: Based on the abnormal electricity consumption characteristics of users and the energy loss characteristics during power transmission, establish a model to assess the impact of electricity marketing effectiveness, and determine the degree of impact of real-time electricity consumption and energy loss during power transmission on the effectiveness of electricity marketing based on the model output. S5: Based on the impact analysis results, determine whether the current power system needs to adjust its marketing strategy.

[0021] In S1, during the electricity marketing monitoring period, real-time monitoring is conducted on user-end electricity consumption data and energy loss data during power transmission, specifically including: During the monitoring period of electricity marketing, real-time electricity consumption data at the user end is monitored. Specifically, smart meters deployed at the user end automatically record the user's real-time electricity consumption at preset time intervals (such as every 15 minutes), forming a complete time series from the beginning to the end of the monitoring period. These smart meters not only accurately measure the user's total electricity consumption but also support two-way communication, enabling data to be transmitted to the data center in real time, ensuring data accuracy and timeliness.

[0022] Simultaneously, energy loss data during power transmission is monitored in real time within the same monitoring cycle. This step primarily relies on a grid sensor network installed at key nodes, which can capture the energy loss of transmission lines in real time and generate energy loss data. Sensors are distributed at key locations in substations, transformers, and important transmission lines to ensure coverage of the entire transmission path, comprehensively reflecting the energy loss status in different regions and time periods. The collected energy loss data is also stored in time-series format and uploaded to the central control system via a secure data transmission channel for subsequent analysis and processing.

[0023] In S2, based on the degree of change in real-time electricity consumption data, abnormal electricity consumption characteristic values ​​of users are determined and the stability of users' electricity consumption behavior is assessed, specifically including: During the monitoring period of electricity marketing, real-time electricity consumption data of users is collected according to the time series. Based on the fluctuation range of real-time electricity consumption data, abnormal electricity consumption characteristic values ​​are calculated. It is determined whether the abnormal electricity consumption characteristic values ​​are greater than or equal to the preset threshold. If so, the user's electricity consumption behavior is unstable; otherwise, the user's electricity consumption behavior is stable.

[0024] The process for obtaining the abnormal electricity consumption characteristic values ​​is as follows: During the electricity marketing monitoring period, real-time electricity consumption data from the user end is collected according to time series. Data collection was conducted, including... Representing discrete points in time, This represents the total number of data collection points. The real-time electricity consumption data is subjected to mean-reduction processing to eliminate the influence of DC component, resulting in preprocessed electricity consumption data. The calculation expression is: ,in, This represents the average of real-time electricity consumption data; Preprocessed electricity consumption data The frequency domain representation is obtained by applying the Fast Fourier Transform. ,in, Represents a frequency index, ranging from 0 to... ; The energy of each frequency component is calculated using the following expression: ;in, Represents frequency The power spectral density at a given point is calculated by summing the energy of all frequency components to obtain the total energy. The sum of the mean and standard deviation of the energy of all frequency components is then calculated to obtain the abnormal energy threshold. The ratio of the sum of the energy of all frequency components greater than the abnormal energy threshold to the total energy is then calculated to obtain the abnormal power consumption characteristic value.

[0025] It should be noted that by collecting real-time electricity consumption data from users during the electricity marketing monitoring cycle and applying Fast Fourier Transform to calculate abnormal electricity consumption characteristic values, this invention can accurately identify abnormal patterns and fluctuation amplitudes in user electricity consumption behavior, thereby effectively assessing the stability of user electricity consumption behavior. This method not only eliminates the influence of DC components, ensuring the accuracy of data processing, but also quantifies the energy contribution of different frequency components through frequency domain analysis, identifying key frequency components that may represent abnormal behavior. Furthermore, by setting an abnormal energy threshold and calculating the proportion of energy components exceeding this threshold to the total energy as an abnormal electricity consumption characteristic value, this invention provides a scientific and quantitative method to determine whether user electricity consumption behavior is stable. The technical advantage of this method is that it improves the sensitivity and accuracy of anomaly detection, enabling power companies to promptly identify potential problems and take corresponding measures.

[0026] In S3, the variation in energy loss at different time points during power transmission is analyzed, energy loss characteristic values ​​are calculated, and the stability of transmission efficiency is evaluated. Specifically, this includes: During the monitoring period of electricity marketing, energy loss data in the power transmission process is collected according to the time series. Based on the change in energy loss in different time periods during the power transmission process, the energy loss characteristic value is calculated. It is then determined whether the energy loss characteristic value is greater than or equal to a preset threshold. If it is, the power transmission efficiency is unstable; otherwise, the power transmission efficiency is stable.

[0027] The process for obtaining the energy loss characteristic value is as follows: During the electricity marketing monitoring period, energy loss data during the power transmission process are analyzed in time series. Data collection was conducted, including... Representing discrete points in time, This represents the total number of data collection points. The real-time energy loss data is subjected to mean removal processing to eliminate the influence of offset, resulting in preprocessed energy loss data. The calculation expression is: ;in, This represents the mean of real-time energy loss data; The specific steps involved in calculating energy loss characteristic values ​​using the k-means clustering algorithm include: determining the total number of clusters. Initialize cluster centers and select them randomly. 1 sample is used as the initial cluster center; For each energy loss data point, calculate its distance to all cluster centers and assign it to the nearest cluster center. The expression for calculating the distance to the cluster center is as follows: ;in, Indicates the first Cluster centers, Indicates the number of cluster centers. Indicates energy loss data and the first The distance between cluster centers Represents energy loss data The 1 eigenvalue, Indicates the number of eigenvalues. Representing cluster centers The One eigenvalue; Update the cluster centers by recalculating the new center position for each cluster, which is then used as the average of all samples in that cluster. The calculation expression is as follows: ;in, Indicates the updated number Cluster centers, Indicates the first The set of all energy loss data for each cluster. This indicates the number of energy loss data in the set; the allocation and update steps are repeated until the cluster centers reach the preset maximum number of iterations. Based on the final clustering results, the standard deviation of each cluster is calculated, and the standard deviations of all clusters are summed to obtain the energy loss characteristic value.

[0028] It should be noted that by collecting energy loss data during power transmission within the power marketing monitoring cycle and applying the k-means clustering algorithm to calculate energy loss characteristic values, this invention can accurately analyze the variation range of energy loss in different time periods, thereby effectively assessing the stability of transmission efficiency. By removing the mean from the original energy loss data, the influence of offset is eliminated, ensuring data consistency and accuracy. Next, the k-means clustering algorithm is used to identify different patterns of energy loss. The standard deviation of each cluster is calculated to quantify the fluctuation within each pattern, and finally, the standard deviations of all clusters are summed to obtain the energy loss characteristic values. This step not only identifies abnormal fluctuation patterns in energy loss but also allows for the determination of transmission efficiency stability by setting reasonable thresholds, enabling power companies to promptly identify and resolve potential transmission loss problems.

[0029] In S4, an impact model on electricity marketing effectiveness is established based on users' abnormal electricity consumption characteristics and energy loss characteristics during power transmission. Based on the model output, the degree of impact of real-time electricity consumption and energy loss during transmission on electricity marketing effectiveness is determined, specifically including: The system acquires the abnormal electricity consumption characteristics of users and the energy loss characteristics during power transmission. It constructs a comprehensive feature vector from the abnormal electricity consumption characteristics and the energy loss characteristics, which serves as the input to the power marketing effect impact model. The training objective of the power marketing effect impact model is to minimize the error between the predicted impact score and the actual impact score. Based on the trained power marketing effect impact model, the system outputs the impact score. The power marketing effect impact model is a random forest model.

[0030] The training process for the model influencing the effectiveness of electricity marketing is as follows: Historical comprehensive feature vectors and influence scores are obtained as training datasets to train the random forest model. In each iteration, the model generates multiple decision trees, each learning based on different sample subsets and feature subsets. The final prediction result is obtained by voting or averaging the results of these decision trees. The model performance is optimized by adjusting model parameters such as the number of trees and the maximum depth to ensure that it can accurately predict the influence score of electricity marketing effectiveness.

[0031] The assessment of the impact of energy loss during real-time electricity consumption and transmission on the effectiveness of electricity marketing specifically includes: The impact score of the electricity marketing effect output by the electricity marketing effect impact model is compared with a preset threshold. If the impact score is greater than or equal to the preset threshold, it is determined whether the impact score is greater than or equal to the preset threshold. If it is, the energy loss in real-time electricity consumption and transmission process has a serious impact on the electricity marketing effect. If not, the energy loss in real-time electricity consumption and transmission process has a slight impact on the electricity marketing effect.

[0032] In S5, based on the impact analysis results, it is determined whether the current power system needs to adjust its marketing strategy, specifically including: Given that real-time electricity consumption and energy losses during power transmission have a significant impact on the effectiveness of electricity marketing, the current power system needs to adjust its marketing strategies. This includes the system automatically generating optimization instructions, adjusting the electricity price structure, and improving service response speed.

[0033] For areas with high abnormal electricity consumption characteristic values, it indicates that there is significant instability in electricity consumption behavior or potential electricity theft on the user side. The system will adjust the electricity price structure of the area to improve the electricity consumption pattern. At the same time, if the energy loss characteristic values ​​of some areas are found to be outside the normal range, it may indicate that the power grid facilities are aging or the power distribution plan is unreasonable, and it is necessary to improve the service response speed for maintenance and upgrades.

[0034] It should be noted that, based on the impact analysis results, determining whether the current power system needs to adjust its marketing strategy can not only automatically identify key areas requiring attention but also generate specific action guidelines to guide power companies in implementing effective adjustments to their marketing and service strategies. For example, in high-energy-loss areas, energy loss can be reduced by improving grid infrastructure and optimizing distribution paths, thereby improving overall power transmission efficiency and service quality. Furthermore, for abnormal user electricity consumption behavior, the risk of default can be reduced and electricity bill collection rates improved by strengthening supervision and providing personalized energy-saving suggestions. These strategic adjustments aim to address potential problems promptly, ensure the efficient operation of the power system and customer satisfaction, and ultimately achieve optimal allocation of power resources and sustainable development.

[0035] The working principle of this invention is as follows: By monitoring electricity consumption data at the user end and energy loss data during power transmission in real time, and applying advanced data analysis techniques (such as Fast Fourier Transform and k-means clustering algorithm) to calculate electricity consumption anomaly characteristic values ​​and energy loss characteristic values, accurate assessment of the stability of user electricity consumption behavior and the stability of power transmission efficiency is achieved. These characteristic values ​​are constructed into a comprehensive feature vector, which serves as input to a random forest model to predict the impact score of electricity marketing effectiveness, thereby determining whether marketing strategies need to be adjusted. The technical effects of this method are significant: First, by performing mean-reduction processing on the raw data and combining it with frequency domain analysis, the sensitivity and accuracy of anomaly detection are improved, enabling timely discovery of abnormal user electricity consumption behavior and energy loss problems during power transmission. Second, the k-means clustering algorithm is used to identify energy loss patterns in different time periods and quantify the fluctuations within each pattern, helping power companies to optimize grid facilities and distribution schemes in a targeted manner. In addition, the application of the random forest model not only improves the accuracy of prediction but also provides a scientific basis to guide the adjustment of electricity price structure and the improvement of service response speed. This method can not only automatically identify key areas that require attention, but also generate specific action guidelines to help power companies make quick and accurate decisions in complex market environments, thereby achieving more efficient resource utilization and higher customer satisfaction.

[0036] Please see Figure 2 As shown, the correlation analysis system for electricity marketing audit indicators based on data mining includes: The data acquisition module monitors real-time electricity consumption data at the user end and energy loss data during power transmission during the power marketing monitoring period. The electricity consumption behavior assessment module determines the abnormal electricity consumption characteristics of users and assesses the stability of users' electricity consumption behavior based on the degree of change in real-time electricity consumption data. The power transmission efficiency evaluation module analyzes the energy loss variation at different time points during the power transmission process, calculates energy loss characteristic values, and evaluates the stability of power transmission efficiency. The impact degree analysis module establishes an impact model on the effectiveness of electricity marketing based on the abnormal electricity consumption characteristics of users and the energy loss characteristics during power transmission. Based on the model output, it judges the degree of impact of real-time electricity consumption and energy loss during power transmission on the effectiveness of electricity marketing. The adjustment module determines whether the current power system needs to adjust its marketing strategy based on the results of the impact analysis.

[0037] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A correlation analysis method for electricity marketing audit indicators based on data mining, characterized in that, Includes the following steps: S1: During the monitoring period of electricity marketing, real-time monitoring is conducted on the real-time electricity consumption data at the user end and the energy loss data during the power transmission process. S2: Based on the degree of change in real-time electricity consumption data, determine the abnormal electricity consumption characteristics of users and assess the stability of users' electricity consumption behavior; S3: Analyze the variation of energy loss at different time periods during power transmission, calculate the characteristic value of energy loss, and evaluate the stability of power transmission efficiency; S4: Based on the abnormal electricity consumption characteristics of users and the energy loss characteristics during power transmission, establish a model to assess the impact of electricity marketing effectiveness, and determine the degree of impact of real-time electricity consumption and energy loss during power transmission on the effectiveness of electricity marketing based on the model output. S5: Based on the impact analysis results, determine whether the current power system needs to adjust its marketing strategy.

2. The correlation analysis method for electricity marketing audit indicators based on data mining according to claim 1, characterized in that, The assessment of the stability of user electricity consumption behavior specifically includes: During the monitoring period of electricity marketing, real-time electricity consumption data of users is collected according to the time series. Based on the fluctuation range of real-time electricity consumption data, abnormal electricity consumption characteristic values ​​are calculated. It is determined whether the abnormal electricity consumption characteristic values ​​are greater than or equal to the preset threshold. If so, the user's electricity consumption behavior is unstable; otherwise, the user's electricity consumption behavior is stable.

3. The correlation analysis method for electricity marketing audit indicators based on data mining according to claim 2, characterized in that, The process for obtaining the abnormal electricity consumption characteristic values ​​is as follows: During the electricity marketing monitoring period, real-time electricity consumption data from the user end is collected according to time series. Data collection was conducted, including... Representing discrete points in time, This represents the total number of data collection points. The real-time electricity consumption data is subjected to mean-reduction processing to eliminate the influence of DC component, resulting in preprocessed electricity consumption data. The calculation expression is: ,in, This represents the average of real-time electricity consumption data; Preprocessed electricity consumption data The frequency domain representation is obtained by applying the Fast Fourier Transform. ,in, Represents a frequency index, ranging from 0 to... ; The energy of each frequency component is calculated using the following expression: ;in, Represents frequency The power spectral density at a given point is calculated by summing the energy of all frequency components to obtain the total energy. The sum of the mean and standard deviation of the energy of all frequency components is then calculated to obtain the abnormal energy threshold. The ratio of the sum of the energy of all frequency components greater than the abnormal energy threshold to the total energy is then calculated to obtain the abnormal power consumption characteristic value.

4. The method for correlation analysis of electricity marketing audit indicators based on data mining according to claim 1, characterized in that, The evaluation of the stability of transmission efficiency specifically includes: During the monitoring period of electricity marketing, energy loss data in the power transmission process is collected according to the time series. Based on the change in energy loss in different time periods during the power transmission process, the energy loss characteristic value is calculated. It is then determined whether the energy loss characteristic value is greater than or equal to a preset threshold. If it is, the power transmission efficiency is unstable; otherwise, the power transmission efficiency is stable.

5. The correlation analysis method for electricity marketing audit indicators based on data mining according to claim 4, characterized in that, The process for obtaining the energy loss characteristic value is as follows: During the electricity marketing monitoring period, energy loss data during the power transmission process are analyzed in time series. Data collection was conducted, including... Representing discrete points in time, This represents the total number of data collection points. The real-time energy loss data is mean-removed to eliminate the influence of offset, resulting in preprocessed energy loss data. The calculation expression is: ;in, This represents the mean of real-time energy loss data; Determine the total number of clusters Initialize cluster centers and select them randomly. 1 sample is used as the initial cluster center; For each energy loss data point, calculate its distance to all cluster centers and assign it to the nearest cluster center. The expression for calculating the distance to the cluster center is as follows: ;in, Indicates the first Cluster centers, Indicates the number of cluster centers. Indicates energy loss data and the first The distance between cluster centers Represents energy loss data The 1 eigenvalue, Indicates the number of eigenvalues. Representing cluster centers The One eigenvalue; Update the cluster centers by recalculating the new center position for each cluster, which is then used as the average of all samples in that cluster. The calculation expression is as follows: ;in, Indicates the updated number Cluster centers, Indicates the first The set of all energy loss data for each cluster. This indicates the number of energy loss data in the set; the allocation and update steps are repeated until the cluster centers reach the preset maximum number of iterations. Based on the final clustering results, the standard deviation of each cluster is calculated, and the standard deviations of all clusters are summed to obtain the energy loss characteristic value.

6. The method for correlation analysis of electricity marketing audit indicators based on data mining according to claim 1, characterized in that, The aforementioned model for influencing the effectiveness of electricity marketing is established based on the abnormal electricity consumption characteristics of users and the energy loss characteristics during power transmission. Specifically, it includes: The system acquires the abnormal electricity consumption characteristics of users and the energy loss characteristics during power transmission. It constructs a comprehensive feature vector from the abnormal electricity consumption characteristics and the energy loss characteristics, which serves as the input to the power marketing effect impact model. The training objective of the power marketing effect impact model is to minimize the error between the predicted impact score and the actual impact score. Based on the trained power marketing effect impact model, the system outputs the impact score. The power marketing effect impact model is a random forest model.

7. The correlation analysis method for electricity marketing audit indicators based on data mining according to claim 6, characterized in that, The training process for the model influencing the effectiveness of electricity marketing is as follows: Historical comprehensive feature vectors and influence scores are obtained as training datasets to train the random forest model. In each iteration, the model generates multiple decision trees, each learning based on different sample subsets and feature subsets. The final prediction result is obtained by voting or averaging the results of these decision trees. The model performance is optimized by adjusting model parameters such as the number of trees and the maximum depth to ensure that it can accurately predict the influence score of electricity marketing effectiveness.

8. The method for correlation analysis of electricity marketing audit indicators based on data mining according to claim 1, characterized in that, The determination of the impact of energy loss during real-time electricity consumption and transmission on the effectiveness of electricity marketing specifically includes: The impact score of the electricity marketing effect output by the electricity marketing effect impact model is compared with a preset threshold. If the impact score is greater than or equal to the preset threshold, it is determined whether the impact score is greater than or equal to the preset threshold. If it is, the energy loss in real-time electricity consumption and transmission process has a serious impact on the electricity marketing effect. If not, the energy loss in real-time electricity consumption and transmission process has a slight impact on the electricity marketing effect.

9. The method for correlation analysis of electricity marketing audit indicators based on data mining according to claim 1, characterized in that, The determination of whether the current power system needs to adjust its marketing strategy based on the impact analysis results specifically includes: Given that real-time electricity consumption and energy losses during power transmission have a significant impact on the effectiveness of electricity marketing, the current power system needs to adjust its marketing strategies. This includes the system automatically generating optimization instructions, adjusting the electricity price structure, and improving service response speed.

10. A correlation analysis system for electricity marketing audit indicators based on data mining, characterized in that, The method for correlation analysis of electricity marketing audit indicators based on data mining as described in any one of claims 1-9 includes: The data acquisition module monitors real-time electricity consumption data at the user end and energy loss data during power transmission during the power marketing monitoring period. The electricity consumption behavior assessment module determines the abnormal electricity consumption characteristics of users and assesses the stability of users' electricity consumption behavior based on the degree of change in real-time electricity consumption data. The power transmission efficiency evaluation module analyzes the energy loss variation at different time points during the power transmission process, calculates energy loss characteristic values, and evaluates the stability of power transmission efficiency. The impact degree analysis module establishes an impact model on the effectiveness of electricity marketing based on the abnormal electricity consumption characteristics of users and the energy loss characteristics during power transmission. Based on the model output, it judges the degree of impact of real-time electricity consumption and energy loss during power transmission on the effectiveness of electricity marketing. The adjustment module determines whether the current power system needs to adjust its marketing strategy based on the results of the impact analysis.

Images