A Personalized Electricity Package Recommendation Method and System Based on User Profiles

Abstract

This application relates to the technical field of personalized electricity package recommendation, providing a method and system for personalized electricity package recommendation based on user profiles to solve the problem that recommended packages in existing technologies often do not match users' actual electricity consumption scenarios. This application collects data on household appliance types, historical electricity consumption, package price selection preferences, and current variation characteristics to obtain equipment operation categories, power consumption distribution, and cost sensitivity assessment data. Based on the equipment operation categories, it matches current characteristics to obtain associated data, combines collaborative filtering to mine group electricity consumption characteristics, and integrates cost data to construct user profiles. Then, it uses the analytic hierarchy process (AHP) to evaluate the cost-effectiveness of packages, ranks them, and generates personalized recommendation results. By integrating household appliance electricity consumption characteristics with user cost-sensitive information to construct profiles and combining algorithms to evaluate package cost-effectiveness, it achieves personalized and accurate electricity package recommendations for users with multiple household appliances.

Description

Technical Field

[0001] This application relates to the technical field of personalized electricity sales package recommendations, and in particular to a method and system for personalized electricity sales package recommendations based on user profiles. Background Technology

[0002] In household electricity use, with the increasing variety of electrical appliances such as refrigerators, air conditioners, and washing machines, the power consumption characteristics of different devices vary significantly. Users' demands for electricity packages are becoming increasingly personalized, requiring both matching the actual power consumption of various devices and aligning with their acceptable price range. Therefore, combining the power consumption characteristics of household appliances with users' cost-sensitive information to provide accurate electricity package recommendations for multi-device users has become a crucial requirement for improving user experience.

[0003] Currently, existing technologies mostly recommend electricity packages based on users' overall electricity consumption data. By statistically analyzing users' historical total electricity consumption, they recommend corresponding tiered electricity price packages; or by collecting basic user information, such as family size and living area, they divide users into groups and push general packages; some methods combine users' historical package selection records and use simple association rules to match the package preferences of similar users.

[0004] However, the biggest drawback of existing technologies is that they fail to deeply integrate the specific operating characteristics of various household electrical appliances with users' cost-sensitive details. They cannot distinguish the impact of different types of equipment, such as the difference in power consumption between high-power equipment and frequently used equipment, on package demand. They also cannot accurately link users' price acceptance with the actual power consumption of equipment, resulting in recommended packages that often do not match users' actual power consumption scenarios and lack personalization. Summary of the Invention

[0005] The purpose of this application is to provide a method and system for personalized electricity package recommendation based on user profiles, in order to solve the problem that the recommended packages in the prior art often do not match the user's actual electricity consumption scenario.

[0006] To address the aforementioned technical problems, in a first aspect, this application provides a method for personalized electricity package recommendation based on user profiles, comprising:

[0007] Collect information on the type of electrical appliances in the home, historical power consumption data of the appliances, user preference data for different price range packages, and current variation characteristics of each electrical appliance during operation;

[0008] The equipment type information is classified and labeled to obtain the equipment operation category corresponding to each electrical device. The historical power consumption data of the equipment is statistically analyzed to obtain power consumption distribution data. The selection preference data is divided according to price acceptance level to obtain cost-sensitive assessment data.

[0009] Based on the equipment operation category, the current change characteristics are matched and analyzed to obtain the correlation data between equipment type information and power load. Based on the power consumption distribution data and the correlation data, a collaborative filtering algorithm is used to mine the power consumption characteristics of users using similar equipment to obtain the group power consumption pattern characteristics.

[0010] By integrating the cost-sensitivity assessment data with the electricity consumption pattern characteristics of the group, a user demand profile is constructed that includes the proportion of equipment electricity consumption, power demand characteristics, and cost sensitivity level.

[0011] Based on the user demand profile, and combined with the analytic hierarchy process, the cost-effectiveness of packages in different price ranges is evaluated to obtain the comprehensive scores corresponding to the packages in different price ranges. The comprehensive scores are then sorted from high to low to generate personalized electricity package recommendations.

[0012] Optionally, the step of fusing the cost-sensitivity assessment data with the group's electricity consumption pattern characteristics to construct a user demand profile including equipment electricity consumption ratio, power demand characteristics, and cost sensitivity level includes:

[0013] The electricity consumption ratio of each electrical device during peak hours is extracted from the characteristics of the group's electricity consumption patterns to obtain the device's electricity consumption ratio.

[0014] Based on the device activation time patterns and load distribution patterns in the characteristics of the group's electricity consumption patterns, the power level required by users during peak hours is calculated to obtain the power demand characteristics.

[0015] Based on the price acceptance level preference in the cost sensitivity assessment data, the user's cost acceptance range is determined, and the cost sensitivity level is obtained;

[0016] By integrating the power consumption ratio of the equipment, the power demand characteristics, and the cost sensitivity level, a user demand profile is constructed.

[0017] Optionally, the step of integrating the device's power consumption ratio, power demand characteristics, and cost sensitivity level to construct a user demand profile includes:

[0018] Establish a profile structure template that includes three dimensions: type of electrical equipment, power demand range, and price acceptance range;

[0019] Fill the power consumption ratio of each device in the power consumption ratio of the device into the power device type dimension of the portrait structure template;

[0020] Fill the peak-hour power level from the power demand characteristics into the power demand interval dimension of the profile structure template;

[0021] Fill the cost acceptance range in the cost sensitivity level into the price acceptance range dimension of the portrait structure template;

[0022] The entered dimensions of electrical equipment type, power demand range, and price acceptance range are linked and stored to form a user demand profile.

[0023] Optionally, based on the user demand profile and combined with the analytic hierarchy process (AHP), the cost-effectiveness of packages in different price ranges is evaluated to obtain comprehensive scores corresponding to packages in different price ranges, including:

[0024] Using the analytic hierarchy process, an evaluation index system is established that includes price level, service content and demand matching degree. Based on the user demand profile, an importance judgment matrix corresponding to the evaluation index system is constructed. The importance judgment matrix includes the weight coefficient of each evaluation index in the evaluation index system.

[0025] Based on the evaluation index system, the degree of fit between the different price range packages and the user demand profile is calculated, and a quantitative score for each price range package for each evaluation index is obtained.

[0026] The quantitative score is weighted and calculated with the corresponding weight coefficient to obtain the comprehensive score of each price range package.

[0027] Optionally, the analytic hierarchy process (AHP) is used to establish an evaluation index system that includes price level, service content, and demand matching degree. Based on the user demand profile, an importance judgment matrix corresponding to the evaluation index system is constructed, including:

[0028] Using the Analytic Hierarchy Process (AHP), a first-level evaluation index is established, which includes price level, service content, and demand matching degree. The demand matching degree is divided into equipment matching degree, power matching degree, and cost matching degree to obtain a second-level evaluation index. The first-level evaluation index and the second-level evaluation index are integrated to obtain an evaluation index system.

[0029] Based on the electricity consumption ratio data of the devices in the user demand profile, the relative importance of electrical devices in package selection is determined. Based on the power demand characteristics in the user demand profile, the contribution of peak-hour power demand to package selection is analyzed. Based on the cost sensitivity level in the user demand profile, the user's acceptance of the range of electricity price fluctuations is assessed.

[0030] Based on the relative importance, the degree of contribution, and the degree of acceptance, the evaluation index system is systematically compared to construct an importance judgment matrix.

[0031] Optionally, based on the equipment operation category, the current change characteristics are matched and analyzed to obtain the correlation data between equipment type information and power load. Based on the power consumption distribution data and the correlation data, a collaborative filtering algorithm is used to mine the power consumption characteristics of users using similar equipment to obtain group power consumption pattern characteristics, including:

[0032] Based on the equipment operation category, the current change characteristics are matched with a preset standard current waveform to obtain waveform matching results. The frequency of occurrence of each equipment type information in the electrical load is statistically analyzed in the waveform matching results to generate correlation data between equipment type information and electrical load.

[0033] The power consumption distribution data and the associated data are combined and processed to construct user power consumption feature vectors. A collaborative filtering algorithm is used to calculate the similarity between the user power consumption feature vectors to obtain user groups using the same type of equipment.

[0034] Extract common electricity consumption habits from the user group, wherein the electricity consumption habits include the regularity of device activation time and load distribution pattern;

[0035] The electricity consumption habit characteristics are integrated with the corresponding equipment type information to form a group electricity consumption pattern characteristic.

[0036] Optionally, the process of classifying and labeling the equipment type information to obtain the equipment operation category corresponding to each electrical device, statistically analyzing the historical equipment power consumption data to obtain power consumption distribution data, and dividing the selection preference data according to price acceptance to obtain cost-sensitive assessment data, includes:

[0037] Based on the power level and start / stop frequency contained in the equipment type information, all electrical equipment is divided into basic operation equipment and intermittent operation equipment, forming equipment operation categories;

[0038] According to the preset time period division rules, the historical power consumption data of the equipment is statistically analyzed, and the power consumption ratio of each equipment during peak and off-peak periods is calculated to obtain power consumption distribution data.

[0039] Based on users' selection records in different price ranges, multiple price acceptance levels are set, and the selection preference data is divided according to the degree of price acceptance to obtain cost-sensitive assessment data.

[0040] Secondly, this application provides a personalized electricity package recommendation system based on user profiles, including:

[0041] The data acquisition module is used to collect information on the type of each electrical appliance in the home, historical power consumption data of the appliances, user preference data for different price range packages, and current variation characteristics of each electrical appliance during operation.

[0042] The statistics module is used to classify and label the equipment type information to obtain the equipment operation category corresponding to each electrical device, to statistically analyze the historical equipment power consumption data to obtain power consumption distribution data, and to divide the selection preference data according to price acceptance to obtain cost-sensitive assessment data.

[0043] The mining module is used to perform matching analysis on the current change characteristics based on the equipment operation category to obtain the correlation data between equipment type information and power load. Based on the power consumption distribution data and the correlation data, a collaborative filtering algorithm is used to mine the power consumption characteristics of users using similar equipment to obtain the group power consumption pattern characteristics.

[0044] The fusion module is used to fuse the cost-sensitive assessment data with the group's electricity consumption pattern characteristics to construct a user demand profile that includes the proportion of equipment electricity consumption, power demand characteristics, and cost sensitivity level.

[0045] The evaluation module is used to evaluate the cost-effectiveness of different price range packages based on the user demand profile and combined with the analytic hierarchy process, to obtain the comprehensive score corresponding to each price range package, and to sort the comprehensive scores from high to low to generate personalized electricity package recommendation results.

[0046] Thirdly, this application provides an electronic device, comprising:

[0047] Memory, used to store computer programs;

[0048] A processor, configured to execute the computer program to implement the steps of the personalized electricity package recommendation method based on user profiles as described in the first aspect above.

[0049] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps of the personalized electricity package recommendation method based on user profiles as described in the first aspect above.

[0050] The personalized electricity package recommendation method based on user profiles provided in this application collects information on the device type of each electrical appliance in the household, historical electricity consumption data, user preference data for packages in different price ranges, and current variation characteristics of each electrical appliance during operation. The device type information is categorized and labeled to obtain the corresponding device operation category. The historical electricity consumption data is statistically analyzed to obtain electricity consumption distribution data. The preference data is divided according to price acceptance to obtain cost-sensitive assessment data. Based on the device operation category, the current variation characteristics are matched and analyzed to obtain device categories. Based on the correlation data between electricity consumption information and electricity load, and using the collaborative filtering algorithm, the electricity consumption characteristics of users using similar equipment are mined to obtain group electricity consumption pattern characteristics. The cost-sensitive assessment data is then fused with the group electricity consumption pattern characteristics to construct a user demand profile that includes equipment electricity consumption ratio, power demand characteristics, and cost sensitivity level. Based on the user demand profile, and combined with the analytic hierarchy process, the cost-effectiveness of different price range packages is evaluated to obtain the comprehensive score corresponding to each price range package. The comprehensive scores are then sorted from high to low to generate personalized electricity sales package recommendation results.

[0051] The technical solution of this application has the following beneficial effects:

[0052] This application collects data on the types of household electrical appliances, historical electricity consumption, price preferences, and current variation characteristics. This provides multi-dimensional raw data covering equipment characteristics, electricity consumption history, user preferences, and real-time operating status, ensuring the comprehensiveness of the data foundation. By classifying, labeling, statistically analyzing, and segmenting the collected data, the raw data is transformed into structured data on equipment operation categories, electricity consumption distribution, and cost-sensitive assessment, making the data analyzable and providing clear input for subsequent correlation matching and feature mining. Based on the current characteristics matched with equipment operation categories, correlation data is obtained. Combined with electricity consumption distribution, a collaborative filtering algorithm is used to mine group electricity consumption characteristics. This approach establishes a direct link between equipment and load, while also incorporating common electricity consumption patterns among similar users, providing dual support for user demand analysis based on both individual and group characteristics. It integrates cost-sensitivity assessment data with group electricity consumption pattern characteristics to construct user demand profiles, encompassing equipment electricity consumption ratios, power demand characteristics, and cost sensitivity levels. This organically combines user electricity consumption characteristics with cost preferences, providing precise user characteristic basis for package recommendations. Based on these user demand profiles and combined with the analytic hierarchy process (AHP), the cost-effectiveness of packages is evaluated and ranked, directly linking package recommendations to user equipment electricity consumption needs, power characteristics, and cost sensitivity, thus improving the personalization and accuracy of recommendations.

[0053] Furthermore, this application extracts the proportion of electricity consumption of each device during peak hours from the characteristics of group electricity consumption patterns to obtain the device electricity consumption ratio. Based on the device activation time pattern and load distribution pattern, it calculates the power level during peak hours to obtain the power demand characteristics. Based on the price acceptance level preference of cost sensitivity assessment data, it determines the cost acceptance range to obtain the cost sensitivity level. Then, it integrates these three factors to construct a user demand profile.

[0054] This application clarifies the specific sources of equipment electricity consumption ratio, power demand characteristics, and cost sensitivity level by refining the extraction dimensions of group electricity consumption pattern characteristics and the analysis methods of cost sensitivity assessment data. This makes the constructed user demand profile more accurately reflect the details of users' peak electricity consumption allocation, power demand intensity, and cost acceptance boundaries, providing more detailed user characteristic support for subsequent package cost-effectiveness assessment and improving the matching degree between the profile and actual needs. Attached Figure Description

[0055] To more clearly illustrate the technical solutions of the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0056] Figure 1 A flowchart illustrating a method for personalized electricity package recommendation based on user profiles, provided in an embodiment of this application;

[0057] Figure 2 A schematic diagram illustrating a specific implementation of a personalized electricity package recommendation method based on user profiles, provided in this application embodiment;

[0058] Figure 3 A schematic diagram of the structure of a personalized electricity package recommendation system based on user profiles provided in this application embodiment;

[0059] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0060] Existing methods for recommending electricity packages have significant limitations when dealing with households with multiple electrical appliances. These methods rely solely on the user's overall electricity consumption or basic information to recommend packages. They fail to differentiate the electricity consumption characteristics of different types of appliances, such as refrigerators and air conditioners, for example, the impact of the difference in power consumption between high-power and frequently used appliances on package demand. Furthermore, they fail to accurately link the user's willingness to accept a price range with the actual power consumption of the appliances. As a result, the recommended packages often do not match the user's actual electricity usage scenarios and fail to meet personalized needs.

[0061] To address the aforementioned issues, this application provides a personalized electricity package recommendation method based on user profiles. This method first collects data on the types of electrical appliances in a household, historical electricity consumption, user preferences for price ranges, and current characteristics during appliance operation. This data is then processed to obtain data on appliance operation categories, power consumption distribution, and cost sensitivity assessment. Next, combining this data, algorithms are used to mine the electricity consumption characteristics of users with similar appliances, and cost information is integrated to construct a user profile including the appliance's electricity consumption percentage, power demand, and cost sensitivity level. Finally, the cost-effectiveness of the packages is evaluated using the analytic hierarchy process (AHP) based on the user profile, and the packages are ranked and recommended. This solution, by deeply integrating specific appliance operating characteristics with user cost-sensitivity details, accurately correlates appliance power consumption needs with user price preferences, effectively solving the problems of insufficient personalization and mismatch between recommendations and actual scenarios in existing methods.

[0062] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0063] The core of this application is to provide a method for personalized electricity package recommendations based on user profiles, and a flowchart of one specific implementation is shown below. Figure 1 As shown, the method includes:

[0064] S101. Collect information on the type of each electrical appliance in the household, historical power consumption data of the appliances, user preference data for different price range packages, and current variation characteristics of each electrical appliance during operation.

[0065] In the above scheme, equipment type information refers to the specific types of electrical appliances in the home, such as refrigerators, air conditioners, washing machines, water heaters, etc., used to distinguish the power consumption characteristics of different appliances; historical equipment power consumption data refers to the power consumption records of each appliance over a period of time, used to reflect the long-term power consumption patterns of the appliances; user preference data for different price range packages refers to user preference information for different price range electricity packages, used to reflect the user's cost sensitivity; current change characteristics refer to the fluctuation of current of each appliance over time, used to reflect the real-time operating status of the appliances.

[0066] In this application example, firstly, device type information for each household appliance is collected through manual input by the user on the electricity sales platform or automatic identification by the system based on device access parameters. This information will be used to differentiate the power consumption characteristics of different devices. Next, using intelligent monitoring modules connected to the power supply lines of each device, the power consumption of each device over a period of time is continuously recorded and summarized to form historical device power consumption data, providing a basis for analyzing long-term power consumption patterns. Then, through online questionnaires or package selection preference pop-ups pushed by the electricity sales platform, user preference information for different price ranges of electricity packages is collected, i.e., user preference data for packages in different price ranges, reflecting the user's cost sensitivity. Finally, using current sensors deployed on the power supply lines of each device, the fluctuation of current during device operation over time is captured in real time, obtaining the current change characteristics of each appliance during operation, which is used to reflect the real-time operating status of the equipment. The four types of data collected above will serve as the raw input for subsequent data processing steps.

[0067] In practical applications, in family A, the equipment information registration function of the electricity sales platform is used to collect data on household electrical appliances, including a single-door refrigerator, a 2-horsepower cabinet air conditioner, a top-loading washing machine, and an electric water heater.

[0068] Then, using smart monitors installed on the lines of each device, the power consumption over the past three months was recorded: the refrigerator consumed a total of 75 kWh, and the monthly power consumption was 75 kWh divided by 3, which equals 25 kWh; the air conditioner consumed a total of 540 kWh over the three summer months, and the monthly power consumption was 540 kWh divided by 3, which equals 180 kWh; the washing machine consumed a total of 90 kWh over the three months, and the monthly power consumption was 90 kWh divided by 3, which equals 30 kWh; and the electric water heater consumed a total of 360 kWh over the three months, and the monthly power consumption was 360 kWh divided by 3, which equals 120 kWh. This data will be used for subsequent statistical analysis of power consumption distribution.

[0069] Next, through a price preference survey pushed by the platform, users chose to prioritize packages with monthly electricity costs in the range of 100-200 yuan. This data will be used for subsequent cost-sensitive assessment data.

[0070] Finally, by connecting current sensors to the circuits of each device, it was found that the refrigerator's current was stable at 0.4A when it was running, the air conditioner's current reached 6A when it was started and remained at 4A during operation, the washing machine's current fluctuated between 1.2-2A when it was working, and the electric water heater's current was stable at 8A when it was heating. These current variation characteristics will be used to match the operating categories of the devices in the future.

[0071] The aforementioned S101 overall solution comprehensively covers information from four dimensions: equipment characteristics, electricity consumption history, user needs, and operating status. It collects information on the type of household electrical appliances, historical power consumption, price preferences, and real-time current characteristics through multiple channels. This provides a complete and original data source for subsequent data processing, feature mining, and user profile construction, ensuring that subsequent steps are based on real and detailed information and avoiding deviations from actual user needs due to missing or incomplete data.

[0072] S102. Classify and label the equipment type information to obtain the equipment operation category corresponding to each electrical device. Statistically analyze the historical power consumption data of the equipment to obtain power consumption distribution data. Divide the selection preference data according to price acceptance level to obtain cost-sensitive assessment data.

[0073] Optionally, step S102 involves classifying and labeling the equipment type information to obtain the equipment operation category corresponding to each electrical device, statistically analyzing the historical equipment power consumption data to obtain power consumption distribution data, and dividing the selection preference data according to price acceptance to obtain cost-sensitive assessment data, including:

[0074] Step 1021: Based on the power level and start / stop frequency contained in the equipment type information, classify all electrical equipment into basic operation equipment and intermittent operation equipment to form equipment operation categories.

[0075] Step 1022: According to the preset time period division rules, the historical power consumption data of the equipment is statistically analyzed, and the power consumption ratio of each equipment during peak hours and off-peak hours is calculated to obtain power consumption distribution data.

[0076] Step 1023: Based on the user's selection records in different price ranges, set multiple price acceptance levels, and divide the selection preference data according to the degree of price acceptance to obtain cost-sensitive assessment data.

[0077] In the above scheme, equipment operation category refers to the category based on the power level and start-stop frequency of electrical equipment, including basic operation equipment and intermittent operation equipment, used to clarify the operating characteristics of different equipment; power consumption distribution data refers to the proportion of power consumption of each device in different time periods according to preset time periods, used to reflect the power consumption pattern of equipment in different time periods; cost sensitivity assessment data refers to multiple price acceptance levels based on users' selection records of different price range packages, used to quantify the user's sensitivity to package prices; preset time period division rules refer to the pre-set time range standards used to distinguish between peak and off-peak electricity consumption periods; peak period refers to the time period with high electricity demand; off-peak period refers to the time period with low electricity demand; price acceptance level refers to the grading standard used to classify the user's price acceptance level.

[0078] In this application example, firstly, through step 1021, all electrical equipment is classified based on the power level and start / stop frequency of each device included in the equipment type information collected in S101. Devices with lower power and continuous operation for long periods are classified as basic operation devices, and devices with higher power and non-continuous operation are classified as intermittent operation devices, thus forming the corresponding equipment operation category for each device.

[0079] Next, in step 1022, according to the preset time period division rules, 8:00-22:00 is set as the peak period, and 22:00-8:00 the next day is set as the off-peak period. The historical device power consumption data collected by S101 is statistically analyzed by time period. The power consumption of each device during the peak period and the off-peak period are calculated separately, and the power consumption ratio of each device in different time periods is obtained by the ratio of the two, forming power consumption distribution data.

[0080] Finally, in step 1023, based on the user selection records of different price range packages collected in S101, three price acceptance levels are pre-set: low, medium, and high, corresponding to price ranges of less than 100 yuan, 100-200 yuan, and more than 200 yuan per month, respectively. User selection preference data is matched with these levels to obtain cost-sensitive assessment data based on price acceptance levels. This data will be used to subsequently build user demand profiles.

[0081] In practical applications, in household A, firstly, based on the device type information collected in S101, the power level and start / stop frequency of each device are analyzed: Since the single-door refrigerator has a power of 100W and runs continuously for 24 hours, it meets the characteristics of low power and long-term continuous operation, and its device operation category is determined to be a basic operating device; Since the 2-horsepower cabinet air conditioner has a power of 1500W and only runs during the high temperature of summer, the pulsator washing machine has a power of 500W and runs 3-4 times a week, and the electric water heater has a power of 2000W and runs 1-2 times a day, they all meet the characteristics of high power and non-continuous operation, and the device operation category of these three devices is determined to be an intermittent operating device. These categories will be used to match the current change characteristics in S103.

[0082] Next, according to the preset time period division rules, the historical electricity consumption data of S101 is statistically analyzed: Since the ratio of electricity consumption during peak and off-peak periods needs to reflect the distribution of electricity consumption of the equipment in different time periods, the ratio of 15 degrees Celsius during the peak period and 10 degrees Celsius during the off-peak period for refrigerators is calculated. 15 divided by 10 equals 1.5, so the electricity consumption distribution ratio of refrigerators is 1.5. Similarly, the peak temperature of air conditioners is 160 degrees Celsius divided by the off-peak temperature of 20 degrees Celsius, which equals 8, so the electricity consumption distribution ratio of air conditioners is 8; the peak temperature of washing machines is 25 degrees Celsius divided by the off-peak temperature of 5 degrees Celsius, which equals 5, so the electricity consumption distribution ratio of washing machines is 5; the peak temperature of electric water heaters is 100 degrees Celsius divided by the off-peak temperature of 20 degrees Celsius, which equals 5, so the electricity consumption distribution ratio of electric water heaters is 5. The resulting electricity consumption distribution data will be used in S103 to mine the characteristics of group electricity consumption.

[0083] Finally, the preset price acceptance levels are low (less than 100 yuan per month), medium (100-200 yuan per month), and high (more than 200 yuan per month). Since users tend to choose packages in the 100-200 yuan range, which matches the price range of the medium level, the cost sensitivity assessment data is medium level.

[0084] The aforementioned S102 overall solution transforms raw data into structured data on equipment operation categories, power consumption distribution, and cost sensitivity assessment by classifying, statistically analyzing, and segmenting equipment type information, historical power consumption data, and selection preference data. This makes the data more targeted and analyzable. The processed data analysis reveals equipment operating characteristics, time-of-use power consumption patterns, and user price preferences. This provides standardized and accurate input for subsequent steps such as matching current characteristics based on equipment operation categories and mining group power consumption patterns by combining power consumption distribution. This ensures that subsequent analysis focuses on key features and improves the reliability of the overall solution.

[0085] S103. Based on the equipment operation category, perform matching analysis on the current change characteristics to obtain the correlation data between equipment type information and power load. Based on the power consumption distribution data and the correlation data, use a collaborative filtering algorithm to mine the power consumption characteristics of users using similar equipment to obtain the group power consumption pattern characteristics.

[0086] In this context, "user" refers to a specific user who needs personalized electricity package recommendations, such as family A. This user is the target of the entire recommendation process, and all data collection, such as the user's device type and electricity consumption, revolves around them. The data is only associated with the user themselves, with the aim of building a unique demand profile for them. "Users using similar devices" refers to a group of users who own the same or similar types of electrical appliances as the specific user mentioned above, such as families B and C, who use appliances like refrigerators and air conditioners as family A. This group is a reference group selected through collaborative filtering algorithms. Its role is to provide common electricity consumption characteristics of users with similar devices, such as the activation patterns of similar devices, to supplement the demand profile of the specific user with common group information. The data is associated with the common electricity consumption information of all users in the group, rather than individual users.

[0087] Optionally, in step S103, based on the equipment operation category, a matching analysis is performed on the current change characteristics to obtain the correlation data between equipment type information and power load. Based on the power consumption distribution data and the correlation data, a collaborative filtering algorithm is used to mine the power consumption characteristics of users using similar equipment to obtain group power consumption pattern characteristics, including:

[0088] Step 1031: Based on the equipment operation category, match the current change characteristics with a preset standard current waveform to obtain waveform matching results. Statistically analyze the frequency of occurrence of each equipment type information in the electrical load within the waveform matching results to generate correlation data between equipment type information and electrical load. Frequency of occurrence is the core basis for generating correlation data between equipment type information and electrical load. Frequency of occurrence and correlation data are directly correlated through the correspondence of "equipment type - load state - frequency of occurrence": First, determine the electrical load state corresponding to a certain type of equipment, such as basic operating equipment or intermittent operating equipment, such as stable low load or intermittent high load, through waveform matching. Then, statistically analyze the frequency of occurrence of that equipment type under the corresponding load state. For example, basic operating equipment appears continuously every day under stable low load, resulting in a high frequency of occurrence; intermittent operating equipment appears only during specific periods under intermittent high load, resulting in a low frequency of occurrence. Finally, bind and integrate "specific equipment type, corresponding load state, and frequency of occurrence under that state" to form correlation data. That is, the higher the frequency of occurrence, the closer the binding relationship between the equipment type and the corresponding load state, and the more accurately the correlation data reflects the impact of the equipment type on the load in actual electricity consumption.

[0089] Step 1032: Combine the power consumption distribution data with the associated data to construct user power consumption feature vectors. Use a collaborative filtering algorithm to calculate the similarity between each user power consumption feature vector to obtain a user group using the same device type. The term "each user power consumption feature vector" is used because the collaborative filtering algorithm needs to compare the power consumption characteristics of multiple users to filter similar user groups. Here, "each user" includes not only the target user for whom a package recommendation is currently needed, but also other users already stored in the system who also possess complete power consumption data such as device type, power consumption distribution, and associated data. Combining each user's power consumption distribution data with the associated data into a feature vector transforms multi-dimensional power consumption information into a quantifiable and computable mathematical form, facilitating subsequent similarity calculations. The number of user power consumption feature vectors is determined by the "total number of users with complete power consumption data" in the system. The number of users who can provide the complete data needed to construct the feature vectors will result in the number of corresponding user power consumption feature vectors. These vectors collectively form the computational basis of the collaborative filtering algorithm, used to filter out user groups that are highly similar to the target user vectors and use the same device type.

[0090] Step 1033: Extract common electricity consumption habit characteristics from the user group, wherein the electricity consumption habit characteristics include the regularity of equipment activation time and load distribution pattern.

[0091] Step 1034: Integrate the electricity consumption habit characteristics with the corresponding equipment type information to form a group electricity consumption pattern characteristic.

[0092] In the above scheme, the data relating equipment type information to power load refers to data obtained by matching equipment operating categories with current change characteristics, reflecting the correspondence between equipment type and its power load, used to clarify the actual power load of different equipment; the collaborative filtering algorithm is an algorithm that mines common patterns by analyzing the behavioral characteristics of similar users in a user group, used here to find common power consumption characteristics of users using the same type of equipment; the group power consumption pattern characteristics refer to the common power consumption patterns extracted from the user group of the same type of equipment, including the equipment activation time pattern and load distribution pattern, used to reflect the common power consumption characteristics of the same type of users; the preset standard current waveform refers to the typical current change curves of different types of equipment in operation that are stored in advance, used as a reference benchmark for matching current change characteristics; the waveform matching result refers to the comparison result between the actual collected current change characteristics and the preset standard current waveform, used to determine the correspondence between equipment type and current characteristics; the user power consumption feature vector refers to the vector data formed by combining power consumption distribution data and associated data, used to characterize the user power consumption characteristics, facilitating the calculation of similarity between users; the user group refers to the set of users using the same type of equipment and with high similarity in power consumption feature vectors, used to extract common power consumption habits.

[0093] In this application example, firstly, through step 1031, based on the equipment operation category obtained in S102, the current change characteristics of each equipment collected in S101 are compared with the preset standard current waveform to obtain waveform matching results. The frequency of occurrence of each equipment type information in the electrical load is statistically analyzed to generate correlation data between equipment type information and electrical load.

[0094] Next, in step 1032, the power consumption distribution data obtained in S102 is combined with the generated associated data to form a user power consumption feature vector that can characterize the user's power consumption characteristics. Then, a collaborative filtering algorithm is used to filter out users who use the same type of equipment and have similar power consumption characteristics by calculating the similarity between the power consumption feature vectors of different users, thus obtaining a user group using the same type of equipment.

[0095] Next, through step 1033, the common electricity consumption patterns of all users are extracted from the obtained user group, including the common activation time periods of each device, i.e. the device activation time pattern, and the proportion of different devices in the total electricity consumption, i.e. the load distribution pattern, to obtain electricity consumption habit characteristics.

[0096] Finally, in step 1034, the obtained electricity consumption habit characteristics are integrated with the corresponding equipment type information to form a group electricity consumption pattern characteristic that includes both equipment type and reflects the common electricity consumption patterns of the group. This characteristic will be used to subsequently build user demand profiles.

[0097] In practical applications, in household A, firstly, based on the device operation category obtained in S102, the current change characteristics collected in S101 are compared with the preset standard current waveform: Since the current of the single-door refrigerator is stable at 0.4A, it is consistent with the standard waveform characteristics of the basic operating equipment, with a high degree of matching. Moreover, it runs continuously for 24 hours a day, and its occurrence time in the daily electricity load is 24 hours, with a frequency of 24 hours divided by 24 hours equaling 1, i.e., 100%. Thus, the association data between the refrigerator type and stable low load is obtained. Since the starting current of the 2-horsepower cabinet air conditioner is 6A and the running current is 4A, it matches the standard waveform characteristics of intermittently operating equipment. Moreover, it runs from 18:00 to 22:00 every day in summer, with a daily occurrence time of 4 hours, and a frequency of 4 hours divided by 24 hours approximately equaling 0.17, i.e., 17%, the association data between the air conditioner type and intermittent high load is obtained. Similarly, the washing machine and electric water heater also obtain corresponding association data by matching their respective current characteristics with the corresponding standard waveform. These association data will be used in step 1032 to construct the feature vector.

[0098] Next, the power consumption distribution data of S102 is combined with the above-mentioned associated data according to the corresponding device type to construct the power consumption feature vector of household A. Subsequently, the cosine similarity algorithm in collaborative filtering is used to calculate the similarity of this vector with other users. The cosine similarity formula is as follows: Where A is the electricity consumption characteristic vector of family A, and B is the electricity consumption characteristic vector of family B (1.6, 7.8, 4.9, 5.2). The dot product of two vectors. Let A be the modulus. Let B be the modulus; then calculate the dot product: ;calculate : ;calculate : Calculate similarity: Since the similarity is close to 1, it indicates that the electricity consumption characteristics of the two households are highly similar, thus obtaining a user group that includes households A, B, etc.

[0099] Then, it was observed from this group that air conditioners are generally turned on between 18:00 and 22:00, refrigerators run 24 hours a day, air conditioners account for the highest proportion of total electricity load, and washing machines are mostly turned on on weekend mornings. Based on these commonalities, the operating time patterns and load distribution patterns of the equipment were extracted.

[0100] Finally, these habitual characteristics are integrated with the corresponding refrigerator, air conditioner, washing machine, and electric water heater types to form a group electricity consumption pattern characteristic. This characteristic will be used to construct a user demand profile for family A in S104.

[0101] The aforementioned S103 overall solution establishes a direct correlation between equipment type and electrical load by matching equipment operating categories with current variation characteristics, clarifying the actual electrical load characteristics of different equipment. Furthermore, by combining power consumption distribution data and utilizing collaborative filtering algorithms, it mines the group electricity consumption pattern characteristics of users with similar equipment, extracting common equipment activation patterns and load allocation patterns. These processes not only establish a correlation between individual equipment characteristics and load but also introduce common electricity consumption patterns among similar users, providing a rich foundation of electricity consumption characteristics for subsequent integration of cost-sensitive information to construct user demand profiles. This ensures that the profiles simultaneously reflect individual equipment characteristics and common group patterns.

[0102] S104. Integrate the cost-sensitive assessment data with the group's electricity consumption pattern characteristics to construct a user demand profile that includes the proportion of equipment electricity consumption, power demand characteristics, and cost sensitivity level.

[0103] Optionally, step S104 involves fusing the cost-sensitivity assessment data with the group's electricity consumption pattern characteristics to construct a user demand profile that includes the proportion of equipment electricity consumption, power demand characteristics, and cost sensitivity level, including:

[0104] Step 1041: Extract the electricity consumption ratio data of each electrical device during peak hours from the characteristics of the group's electricity consumption pattern to obtain the device's electricity consumption ratio.

[0105] The physical meaning of peak hours refers to the specific time period of the day when household users have the most frequent daily activities, electrical equipment is used intensively, and the overall electricity demand and load are high. It usually corresponds to the time when people's electricity consumption is concentrated during the day for work and study, cooking and laundry in the evening, and lighting and entertainment at night.

[0106] Step 1042: Based on the device activation time pattern and load distribution pattern in the characteristics of the group power consumption pattern, calculate the power level required by the user during peak hours to obtain the power demand characteristics.

[0107] The physical meaning of the pattern in device activation time refers to the distribution of specific time periods during the day when various electrical devices are concentratedly used, based on the actual electricity consumption behavior of a group of users with similar devices. Essentially, it's the combination of users' daily routines and electricity needs that creates observable patterns in device activation times. For example, air conditioners are often used between 6:00 PM and 10:00 PM, and washing machines are often used on weekend mornings. These specific time distributions constitute its physical meaning. The relationship between users and electrical devices is that users are the initiators of electricity needs, while electrical devices are the hardware carriers that meet those needs. Users' lifestyles directly determine the activation time, frequency, and duration of device use, and the operating status of the devices directly reflects users' electricity needs. The difference lies in the fact that users are demand subjects with autonomous decision-making capabilities; their core function is to generate electricity needs and operate the devices. Electrical devices are tools with specific energy consumption functions; their core function is to achieve the functions required by users through energy consumption. One is the source of demand, and the other is the carrier of function; their roles and attributes are completely different.

[0108] Step 1043: Based on the price acceptance level preference in the cost sensitivity assessment data, determine the user's cost acceptance range and obtain the cost sensitivity level.

[0109] Among them, cost sensitivity level and cost acceptance range are directly related as "general description" and "specific quantitative range". The former is a qualitative classification of the latter, and the latter is the specific data support for the former. The cost acceptance range is the specific range of electricity package prices that users can accept. It is a quantifiable price boundary extracted from cost sensitivity assessment data. The cost sensitivity level is a classification summary of users' price preferences based on this specific range and according to preset standards.

[0110] Step 1044: Integrate the power consumption ratio of the equipment, the power demand characteristics, and the cost sensitivity level to construct a user demand profile.

[0111] Step 1044 can specifically include the following steps: establishing a profile structure template containing three dimensions: type of electrical equipment, power demand range, and price acceptance range; filling the proportion of electricity consumption of each device in the power consumption ratio of the equipment into the electrical equipment type dimension of the profile structure template; filling the peak power level in the power demand characteristics into the power demand range dimension of the profile structure template; filling the cost acceptance range in the cost sensitivity level into the price acceptance range dimension of the profile structure template; associating and storing the filled electrical equipment type dimension, the filled power demand range dimension, and the filled price acceptance range dimension to form a user demand profile. The core of establishing this profile structure template is to focus on the goal of "accurately matching electricity sales packages," combining the core elements of user electricity demand with the key dimensions of package evaluation. The specific process is as follows: First, it is clarified that the template must cover three core dimensions: "user's electricity hardware foundation, electricity capacity requirements, and electricity cost preferences." These three dimensions correspond to the evaluation points of "equipment compatibility," "power carrying capacity," and "price compatibility" of electricity sales packages, respectively, and are crucial to ensuring that the subsequent profile can support package recommendations; then... For each dimension, specific content and data formats are defined: For the "Electrical Equipment Type" dimension, combining previously collected equipment type information and group electricity consumption pattern characteristics, it is clear that the specific types of all user's electrical equipment and the electricity consumption percentage of each equipment during peak hours must be recorded, with the data format set as "Equipment Name - Peak Electricity Consumption Percentage"; For the "Power Demand Range" dimension, referring to the power division standards of common electricity packages, multiple power ranges are preset, and the specific range corresponding to the user's required power level during peak hours must be filled in, with the data format set as "Power Range - Corresponding Power Value"; For the "Price Acceptance Range" dimension, the price range associated with cost-sensitive assessment data must be recorded, specifying the acceptable monthly electricity package price range for the user and marking the matching cost sensitivity level, with the data format set as "Price Range - Cost Sensitivity Level"; Finally, the three dimensions are integrated in the logical order of "Equipment - Power - Price" to form a structured template framework containing dimension names, data items, and data formats. This ensures that the data such as equipment electricity consumption percentage, power demand level, and cost acceptance range generated in subsequent steps can be accurately filled into the corresponding dimensions, providing a standardized structure for associated storage to form a user demand profile.

[0112] In the above scheme, the equipment power consumption ratio refers to the power consumption ratio of each electrical device during peak hours extracted from the characteristics of group power consumption patterns, used to reflect the weight of different devices in peak power consumption; the power demand characteristics refer to the total power level required by users during peak hours calculated based on the characteristics of group power consumption patterns, used to reflect the power scale of users' peak power consumption; the cost sensitivity level refers to the user's cost acceptance range determined based on cost sensitivity assessment data, used to quantify the user's sensitivity to package prices; the profile structure template refers to a pre-established framework containing three dimensions: electrical device type, power demand range, and price acceptance range, used to standardize the data storage structure of user demand profiles; the user demand profile refers to a multi-dimensional user power consumption and cost preference model formed by integrating the equipment power consumption ratio, power demand characteristics, and cost sensitivity level, used to provide accurate user characteristic basis for subsequent package evaluation.

[0113] In this application example, firstly, through step 1041, the electricity consumption records of each electrical device during peak hours are filtered from the group electricity consumption pattern characteristics obtained in S103, and the proportion of electricity consumption of each device to the total electricity consumption during peak hours is extracted to obtain the device electricity consumption ratio.

[0114] Then, through step 1042, based on the device activation time pattern and load distribution pattern in the characteristics of group power consumption patterns, the types of devices operating simultaneously during peak hours and the rated power of each device are counted, and the sum of the rated power of these devices is calculated to obtain the power level required by the user during peak hours, i.e., the power demand characteristics.

[0115] Next, through step 1043, referring to the obtained cost sensitivity assessment data, based on the user's price acceptance level preference, the upper and lower limits of the package cost that the user can accept are determined, and the cost sensitivity level is obtained.

[0116] Finally, in step 1044, a user profile template is first established, comprising three dimensions: equipment type, power demand range, and price acceptance range. Then, the equipment power consumption percentage obtained in step 1041 is filled into the equipment type dimension of the template; the peak-hour power level obtained in step 1042 is filled into the power demand range dimension; and the cost acceptance range obtained in step 1043 is filled into the price acceptance range dimension. Finally, the data from these three dimensions is correlated and stored to form a complete user demand profile. This user demand profile will be used in subsequent step S105 to evaluate the cost-effectiveness of packages in different price ranges.

[0117] In practical application, within household A, firstly, based on the group electricity consumption pattern characteristics obtained from S103, the total electricity consumption during peak hours is calculated to be 100 units. Of this, the air conditioner consumes 50 units, the refrigerator 10 units, the washing machine 20 units, and the electric water heater 20 units during peak hours. Then, the electricity consumption percentage of each device is calculated using the following formula: ,in, Let i be the percentage of electricity consumption for the i-th type of device. Let be the electricity consumption of the i-th device during peak hours. This represents the total electricity consumption during peak hours. Based on this, the percentage of electricity consumed by air conditioning is calculated. That is, 50%, the proportion of electricity used by refrigerators. That is, 10%, the proportion of electricity used by washing machines. That is, 20%, the proportion of electricity used by electric water heaters. That is, 20%, which gives the percentage of electricity used by household A's devices. This data will be used to fill in the device type dimension of the profile structure template.

[0118] Next, based on the characteristics of the group's electricity consumption patterns, it was determined that the devices that family A typically operates simultaneously during peak hours are the air conditioner and the electric water heater, with the air conditioner having a rated power of [missing information]. Rated power of electric water heater The formula for calculating the power level required during peak hours is: ,in, This represents the total power demand during peak hours. This refers to the rated power of the air conditioner. This is the rated power of the electric water heater. Substituting the data, we get... =1500+2000=3500W, so the power demand characteristic is 3500W. This data will be used to fill in the power demand range dimension of the portrait structure template.

[0119] Then, referring to the cost sensitivity assessment data obtained in S102, it is known that the price acceptance level of family A in this data is medium. Combining the statistical results of the price range corresponding to the level among similar users, it is determined that the cost acceptance range corresponding to the medium level is 100-200 yuan per month. Thus, family A's cost sensitivity level is moderately sensitive and they accept a monthly package of 100-200 yuan. This data will be used to fill in the price acceptance range dimension of the profile structure template.

[0120] Finally, first, establish a profile structure template that includes three dimensions: type of electrical equipment, power demand range, and price acceptance range; then, fill in the proportion of electricity consumption of the equipment into the electrical equipment type dimension, fill in 3500W into the power demand range dimension, and fill in 100-200 yuan per month into the price acceptance range dimension; finally, link and store the data of the three dimensions to form a user demand profile of family A. This profile will be used in S105 to evaluate the cost-effectiveness of electricity sales packages in different price ranges.

[0121] The aforementioned S104 overall solution integrates cost-sensitivity assessment data with group electricity consumption pattern characteristics to construct a user demand profile from three key dimensions: equipment electricity consumption ratio, power demand characteristics, and cost sensitivity level. This process combines common electricity consumption references of similar users with the personalized cost preferences of target users, avoiding the one-sidedness of profiles caused by relying solely on individual data or referring only to group data. The resulting user demand profile comprehensively and accurately reflects users' electricity needs and cost acceptance boundaries, providing a core basis for subsequent evaluation of package cost-effectiveness based on the analytic hierarchy process (AHP) that aligns with users' actual needs.

[0122] S105. Based on the user demand profile, and combined with the analytic hierarchy process, the cost-effectiveness of packages in different price ranges is evaluated to obtain the comprehensive scores corresponding to the packages in different price ranges. The comprehensive scores are sorted from high to low to generate personalized electricity package recommendation results.

[0123] Optionally, step S105 involves evaluating the cost-effectiveness of packages in different price ranges based on the user demand profile and using the analytic hierarchy process (AHP) to obtain comprehensive scores for each package in different price ranges, including:

[0124] Step 1051: Using the Analytic Hierarchy Process (AHP), establish an evaluation index system that includes price level, service content and demand matching degree. Based on the user demand profile, construct an importance judgment matrix corresponding to the evaluation index system. The importance judgment matrix includes the weight coefficient of each evaluation index in the evaluation index system.

[0125] Step 1051 may specifically include the following steps: Using the Analytic Hierarchy Process (AHP), establish a first-level evaluation index that includes price level, service content, and demand matching degree; divide the demand matching degree into equipment matching degree, power matching degree, and cost matching degree to obtain a second-level evaluation index; integrate the first-level evaluation index with the second-level evaluation index to obtain an evaluation index system; based on the equipment electricity consumption ratio data in the user demand profile, determine the relative importance of electrical equipment in package selection; based on the power demand characteristics in the user demand profile, analyze the contribution of peak-hour power demand to package selection; based on the cost sensitivity level in the user demand profile, assess the user's acceptance of the electricity price fluctuation range; based on the relative importance, the contribution degree, and the acceptance degree, systematically compare the evaluation index system to construct an importance judgment matrix.

[0126] The weight coefficients of the first-level evaluation indicators are calculated using their corresponding 3×3 importance judgment matrices. First, based on user demand profiles (e.g., cost sensitivity reflecting price importance, power demand characteristics reflecting demand matching importance), pairwise importance comparisons are made between price level, service content, and demand matching, such as "demand matching is more important than price level" and "price level is more important than service content." A 1-9 scale is used to assign values ​​to the comparison results, forming a 3×3 matrix. Then, the eigenvector corresponding to the largest eigenvalue of this matrix is ​​calculated, and the normalized eigenvector yields the weight coefficient for each first-level indicator. This coefficient directly reflects the relative importance of each indicator in the overall package evaluation. The weight coefficients of the second-level evaluation indicators... The importance of each indicator needs to be calculated using its own 3×3 importance judgment matrix, combined with the weight of the first-level indicators: First, based on the user demand profile, such as the proportion of equipment power consumption reflecting the importance of equipment matching degree and power demand characteristics reflecting the importance of power matching degree, the importance of equipment matching degree, power matching degree, and cost matching degree are compared pairwise and assigned values ​​to construct a 3×3 matrix; the normalized feature vector of this matrix is ​​calculated to obtain the "relative weight coefficient" of the second-level indicators relative to the first-level indicators of "demand matching degree"; then, the relative weight coefficient of each second-level indicator is multiplied by the first-level weight coefficient of "demand matching degree" itself to obtain the "absolute weight coefficient" of each second-level indicator in the overall evaluation, and finally, its influence in the overall evaluation is reflected through the absolute weight coefficient.

[0127] The evaluation index system corresponds to two importance judgment matrices. The matrix dimensions are determined by the number of evaluation indicators at each level: the first level of evaluation indicators includes three items: price level, service content, and demand matching degree. Therefore, a 3×3 importance judgment matrix is ​​constructed for the first level of indicators. The second level of evaluation indicators includes three items under demand matching degree: equipment matching degree, power matching degree, and cost matching degree. Therefore, a 3×3 importance judgment matrix is ​​constructed for the second level of indicators. The importance relationship between the two levels of indicators is reflected by two 3×3 matrices, rather than a single matrix.

[0128] Step 1052: Based on the evaluation index system, calculate the degree of fit between the different price range packages and the user demand profile, and obtain the quantitative score of each price range package for each evaluation index.

[0129] Step 1053: Calculate the weighted score by weighting the quantitative score with the corresponding weight coefficient to obtain the comprehensive score of each price range package.

[0130] In the above scheme, the Analytic Hierarchy Process (AHP) refers to a systematic analysis method that decomposes complex decision-making problems into multiple levels, such as the target level and the indicator level, and determines the indicator weights by constructing a judgment matrix and conducting a comprehensive evaluation. Here, it is used to scientifically quantify the evaluation process of package cost-effectiveness. The evaluation indicator system refers to a package evaluation framework with a two-layer structure. The first-layer evaluation indicators are price level, service content and demand matching degree. The second-layer evaluation indicators further divide the demand matching degree into equipment matching degree, power matching degree and cost matching degree, which are used to comprehensively cover the key dimensions of package evaluation. The importance judgment matrix is ​​a matrix constructed based on user demand profiles that reflects the relative importance of each evaluation indicator. It includes the weight coefficients of each indicator, which are used to reflect the degree of influence of different indicators in package evaluation. The degree of fit refers to the degree of matching between packages in different price ranges and user demand profiles. It needs to be judged in combination with the equipment power consumption ratio, power demand characteristics and cost sensitivity level in the profile. The quantitative score is the specific score converted into the degree of fit, which is used to intuitively reflect the performance of the package on each evaluation indicator. The comprehensive score is the score calculated by weighting the quantitative score with the corresponding weight coefficients. It is used to comprehensively measure the cost-effectiveness of the package and is the core basis for package ranking and recommendation.

[0131] In this application example, firstly, through step 1051, an evaluation index system is constructed using the Analytic Hierarchy Process (AHP): the first-level evaluation indexes are determined as price level and service content matching degree. Then, the matching degree is refined into equipment matching degree, power matching degree, and cost matching degree as the second-level evaluation indexes. The two levels of indicators are integrated to form a complete evaluation index system. Based on the electricity consumption ratio of equipment in the user demand profile, the relative importance of equipment in package selection is analyzed. Combined with power demand characteristics, the contribution of peak power demand to package selection is judged. Based on the cost sensitivity level, the user's acceptance of electricity price fluctuations is assessed. An importance judgment matrix is ​​constructed through a systematic comparison of these dimensions, and the weight coefficients of each evaluation index are extracted from the matrix.

[0132] Next, through step 1052, the matching of different price range packages with user demand profiles is compared one by one based on the indicator system: the suitability of package prices with user cost acceptance range is judged to assess the price level, the correspondence between package service content and user potential needs is checked to assess the service content, and the coverage of package for user device types and the satisfaction of peak power demand are compared to assess device matching degree and power matching degree respectively. These matching results are converted into quantitative scores of 0-100 points to obtain the specific score of each package for each evaluation indicator.

[0133] Finally, in step 1053, the quantitative score of each package on each evaluation indicator is multiplied by the weight coefficient corresponding to that indicator, and then all the product results are summed to obtain the comprehensive score of each price range package; the packages are sorted in descending order of comprehensive score to generate the final personalized electricity package recommendation result.

[0134] In practical application, for family A, firstly, based on their user needs profile—that is, the electricity consumption of appliances is 50% for air conditioner, 10% for refrigerator, 20% for washing machine, and 20% for electric water heater, with a power requirement of 3500W, and a cost sensitivity level of accepting a monthly package of 100-200 yuan—an evaluation index system is constructed using the Analytic Hierarchy Process (AHP). The first-level indicators are price level, service content, and demand matching degree; the second-level indicators are appliance matching degree, power matching degree, and cost matching degree. Considering that the user's appliance electricity consumption characteristics and power requirements directly affect the electricity experience, the demand matching degree weight coefficient is set to 0.5; the user's sensitivity to package prices is given a price level weight coefficient of 0.3; service content is a secondary need, and its weight coefficient is set to 0.2; under the demand matching degree, power requirements affect electricity stability, so the power matching degree weight coefficient is set to 0.4; appliance matching degree affects appliance adaptability, so its weight coefficient is set to 0.3; and cost matching degree is related to the price acceptance range, so its weight coefficient is set to 0.3. An importance judgment matrix is ​​constructed through systematic comparison, and the above weight coefficients are determined.

[0135] Next, Package X and Package Y were selected: Package X costs 120-180 yuan per month, supports a maximum power of 4000W, provides 24-hour electricity consultation service, and is compatible with common household appliances; Package Y costs 180-230 yuan per month, supports a maximum power of 3500W, provides electricity data analysis service, and is compatible with common household appliances. Based on the assessment and quantitative scoring of family profile A: Package X, priced between 100-200 yuan, scores 90 points for price level; includes basic consultation service, so service content scores 80 points; is compatible with all appliances, so equipment matching scores 90 points; supports power above 3500W, so power matching scores 95 points; and fully within the cost range, so cost matching scores 90 points. Package Y partially exceeds the 100-200 yuan range, so price level scores 70 points; includes advanced analysis service, so service content scores 90 points; is compatible with all appliances, so equipment matching scores 90 points; just meets power requirements, so power matching scores 95 points; and partially exceeds the cost range, so cost matching scores 70 points.

[0136] Finally, the comprehensive score calculation formula is adopted. Where S is the overall score of the package. Let be the weight coefficient of the i-th evaluation indicator. This is the quantitative score for the package on the i-th evaluation indicator. Calculate the overall score for Package X: First, calculate the demand matching weighted score: 0.3×90 + 0.4×95 + 0.3×90 = 27 + 38 + 27 = 92; then calculate the total score: 0.3×90 + 0.2×80 + 0.5×92 = 27 + 16 + 46 = 89. Calculate the overall score for Package Y: First, calculate the demand matching weighted score: 0.3×90 + 0.4×95 + 0.3×70 = 27 + 38 + 21 = 86; then calculate the total score: 0.3×70 + 0.2×90 + 0.5×86 = 21 + 18 + 43 = 82. Then, sort the packages from highest to lowest overall score as Package X > Package Y, generating personalized electricity package recommendations. This result will be directly used to push suitable electricity packages to family A.

[0137] The aforementioned S105 overall solution constructs a structured evaluation index system using the analytic hierarchy process (AHP), upgrading package evaluation from a single dimension to a multi-dimensional analysis covering price, service, and demand matching. It determines index weights based on user demand profiles, avoiding subjectivity in the evaluation. Quantitative scoring transforms the fit between the package and the user profile into intuitive scores, which are then weighted to obtain a comprehensive score, ensuring the scientific and objective nature of the package cost-effectiveness evaluation. Finally, the recommendation results generated according to the score ranking accurately match the user's equipment power consumption characteristics, power requirements, and cost acceptance range, effectively improving the personalization and accuracy of electricity package recommendations and helping users quickly select the package best suited to their needs.

[0138] The following is a complete example for steps S101 to S105, such as Figure 2 As shown, data was first collected from household A. The equipment types included a single-door refrigerator, a 2HP floor-standing air conditioner, a top-loading washing machine, and a storage-type electric water heater. Historical electricity consumption data for the past month showed peak consumption from 8:00 AM to 10:00 PM: refrigerator 18 kWh, air conditioner 150 kWh, washing machine 30 kWh, and water heater 90 kWh; during the off-peak period from 10:00 PM to 8:00 AM the next day: refrigerator 12 kWh, air conditioner 30 kWh, washing machine 6 kWh, and water heater 30 kWh. User preference data for different price range packages showed that users explicitly preferred packages costing 100-200 yuan per month, and would carefully consider packages exceeding 200 yuan. Current variation characteristics for each appliance were as follows: refrigerator current stable at 0.45A, air conditioner starting current 5.8A and operating current 3.9A, washing machine starting current 2.1A and operating current 1.7A, and water heater starting current 8.8A and operating current 7.2A. All of this collected data will be used in subsequent data processing steps. Equipment type information will be used for classification and labeling, historical electricity consumption data will be used to calculate power consumption distribution data, selection bias data will be used to segment cost-sensitive assessment data, and current change characteristics will be used for subsequent matching analysis.

[0139] Next, the equipment type information was categorized and labeled: a single-door refrigerator with a power of 110W and running continuously for 24 hours meets the characteristics of low power and long-term continuous operation, and is labeled as a basic operating device; a 2-horsepower cabinet air conditioner with a power of 1450W runs only during the high-temperature period in summer, a pulsator washing machine with a power of 480W runs 3 times a week, and a storage-type electric water heater with a power of 1900W runs for 1.5 hours a day all meet the characteristics of high power and non-continuous operation, and are labeled as intermittent operating devices. This classification result will be used for subsequent current change characteristic matching analysis. Historical equipment power consumption data was statistically analyzed using a formula. Calculate the power consumption distribution ratio of each device, where R is the power consumption distribution ratio of the devices. This refers to the power consumption of the equipment during peak hours. The power consumption of the equipment during off-peak hours; substituting the data into the calculation, refrigerator ,air conditioner ,washing machine Electric water heater The power consumption distribution data was obtained, which will be used to construct user electricity consumption feature vectors. Selection preference data was categorized according to price acceptance levels, with three preset levels: low acceptance (<100 yuan / month), medium acceptance (100-200 yuan / month), and high acceptance (>200 yuan / month). Since users prefer the 100-200 yuan range, the cost-sensitive assessment data was classified as medium acceptance, which will be used to construct user demand profiles.

[0140] Then, based on the equipment operation category, the current change characteristics were matched and analyzed: the current change characteristics of each device were compared with the preset standard current waveform. The current characteristic of the refrigerator with a stable current of 0.45A matched the standard waveform of continuous low current of the basic operating equipment with a 98% match. The frequency of its occurrence in the power load was 24 hours divided by 24 hours, which equals 100%. The current characteristics of the air conditioner, washing machine, and electric water heater all matched the standard waveform of high starting current and stable operating current of intermittently operating equipment with a match of over 90%. The frequencies of occurrence were approximately 21% for 5 hours per day divided by 24 hours in summer, approximately 6% for 10 hours per week divided by 168 hours, and approximately 6% for 1.5 hours per day divided by 24 hours, respectively. The correlation data between equipment type information and power load was obtained. This data was combined with the aforementioned power consumption distribution data to construct a feature vector. Subsequently, the R values ​​(1.5, 5, 5, 3) in the power consumption distribution data were combined with the frequency of occurrence (100%, 21%, 6%, 6%) in the correlation data. The frequency of occurrence was converted into a decimal to construct the power consumption feature vector of household A. Simultaneously, obtain the electricity consumption feature vector of household B in the system. Using the cosine similarity formula Calculate the similarity, where, for The length of the mold, for Substituting the modulus into the data, we can calculate the similarity: Subsequently, users with similarity scores exceeding 0.95 (A, B, C, etc.) were selected to form user groups with similar devices. From these groups, electricity consumption habits were extracted, such as air conditioners being used mostly between 17:30 and 22:30, refrigerators running 24 hours a day, air conditioners accounting for the highest proportion of total electricity load, and washing machines being used mostly between 9:00 and 11:00 on weekends. These characteristics were integrated with device type information to obtain the group's electricity consumption pattern features, which will be used to build user demand profiles in the future.

[0141] Next, the total peak-hour electricity consumption of 100 units was extracted from the characteristics of group electricity consumption patterns, including 12 units for refrigerators, 55 units for air conditioners, 18 units for washing machines, and 15 units for electric water heaters. The formula was then used to calculate the peak-hour electricity consumption. Calculate the proportion of electricity consumption by equipment, where P represents the proportion of electricity consumption by equipment. This refers to the power consumption of the equipment during peak hours. This represents the total electricity consumption during peak hours; substituting the data into the calculation, the refrigerator... ,air conditioner ,washing machine Electric water heater The electricity consumption percentage of each device was obtained. Then, based on the pattern of simultaneous use of air conditioners (17:30-22:30) and electric water heaters (18:00-19:30) in the group's electricity consumption characteristics, and given that the air conditioner's rated power is 1450W and the electric water heater's rated power is 1900W, the peak power requirement was calculated as 1450W + 1900W = 3350W, yielding the power demand characteristic. Next, based on the acceptance level in the cost sensitivity assessment data obtained from the data processing steps, and combined with the statistical results of the price range corresponding to the acceptance level among similar users, the cost acceptance range corresponding to the medium acceptance level was determined to be 100-200 yuan per month, yielding the cost sensitivity level. Finally, the electricity consumption percentage, power demand characteristic, and cost sensitivity level were integrated to construct a user profile for family A, including an electricity consumption percentage of 55% for air conditioners, 12% for refrigerators, 18% for washing machines, and 15% for electric water heaters, a power demand characteristic of 3350W, and a medium acceptance level cost sensitivity corresponding to 100-200 yuan per month. This profile will be used for subsequent package cost-effectiveness evaluation.

[0142] Finally, based on the user needs profile of family A, an evaluation index system was established using the analytic hierarchy process (AHP). The first-level evaluation indexes are price level, service content, and demand matching degree. Through analysis of the user needs profile, cost sensitivity reflects the importance of price, and power demand characteristics reflect the importance of demand matching degree. The weights for the first level are determined as follows: price level 0.3, service content 0.2, and demand matching degree 0.5. The second-level evaluation indexes are equipment matching degree, power matching degree, and cost matching degree. Based on the importance comparison of equipment electricity consumption ratio, power demand characteristics, and cost sensitivity level, the weights for the second level are determined as follows: equipment matching degree 0.3, power matching degree 0.4, and cost matching degree 0.3. This index system and weights will be used for quantitative scoring of the packages. Two alternative packages were selected: Package X costs 110-180 yuan per month, supports a maximum power of 3500W, provides 24-hour electricity consultation service, and is compatible with common household appliances; Package Y costs 180-230 yuan per month, supports a maximum power of 3300W, provides electricity data analysis service, and is compatible with common household appliances. Based on user needs profiles, a quantitative scoring system is used: Package X, priced between 100-200 yuan, scores 90 points for price level; includes basic consultation services, scoring 80 points for service content; compatible with all devices, scoring 95 points for device compatibility; supports power exceeding 3350W, scoring 90 points for power compatibility; and within cost range, scoring 90 points for cost compatibility. Package Y, partially priced over 200 yuan, scores 70 points for price level; includes advanced analysis services, scoring 90 points for service content; compatible with all devices, scoring 95 points for device compatibility; supports power exactly 3350W, scoring 85 points for power compatibility; and partially exceeds cost range, scoring 70 points for cost compatibility. This score will be used to calculate the overall score. The overall score formula is used. Calculate, where S is the overall score of the package. Let be the weight coefficient of the i-th indicator. This is the quantitative score for the package on the i-th indicator. Calculate Package X: First, calculate the demand matching score: 0.3×95+0.4×90+0.3×90=28.5+36+27=91.5, then calculate the comprehensive score: 0.3×90+0.2×80+0.5×91.5=27+16+45.75=88.75. Calculate Package Y: First, calculate the demand matching score: 0.3×95+0.4×85+0.3×70=28.5+34+21=83.5, then calculate the comprehensive score: 0.3×70+0.2×90+0.5×83.5=21+18+41.75=80.75. Then, sort the packages by comprehensive score from highest to lowest, with Package X being higher than Package Y. Generate a personalized electricity package recommendation result for Family A, prioritizing Package X.

[0143] Figure 3 This is a schematic diagram illustrating a specific implementation of a personalized electricity package recommendation system based on user profiles, as provided in this application. (Refer to...) Figure 3The system may include:

[0144] The data acquisition module 31 is used to collect information on the type of each electrical appliance in the home, historical power consumption data of the appliances, user preference data for different price range packages, and current change characteristics of each electrical appliance during operation.

[0145] The statistics module 32 is used to classify and label the equipment type information to obtain the equipment operation category corresponding to each electrical device, to statistically analyze the historical equipment power consumption data to obtain power consumption distribution data, and to divide the selection preference data according to price acceptance to obtain cost-sensitive assessment data.

[0146] The mining module 33 is used to perform matching analysis on the current change characteristics based on the equipment operation category to obtain the correlation data between equipment type information and power load. Based on the power consumption distribution data and the correlation data, a collaborative filtering algorithm is used to mine the power consumption characteristics of users using similar equipment to obtain the group power consumption pattern characteristics.

[0147] The fusion module 34 is used to fuse the cost-sensitive assessment data with the group's electricity consumption pattern characteristics to construct a user demand profile that includes the proportion of equipment electricity consumption, power demand characteristics, and cost sensitivity level.

[0148] The evaluation module 35 is used to evaluate the cost-effectiveness of different price range packages based on the user demand profile and combined with the analytic hierarchy process, to obtain the comprehensive score corresponding to each price range package, and to sort the comprehensive scores from high to low to generate personalized electricity package recommendation results.

[0149] The personalized electricity package recommendation system based on user profiles in this application is used to implement the aforementioned personalized electricity package recommendation method based on user profiles. Therefore, the specific implementation of the personalized electricity package recommendation system based on user profiles can be found in the embodiment section of the personalized electricity package recommendation method based on user profiles above. The specific implementation can be referred to the description of the corresponding embodiments, which will not be repeated here.

[0150] like Figure 4 As shown, this application also provides an electronic device, including: a memory for storing a computer program; and a processor for executing the computer program to implement the steps of any of the above-described methods for personalized electricity package recommendations based on user profiles.

[0151] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the above-described methods for personalized electricity package recommendations based on user profiles.

[0152] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory, random access memory, portable hard drives, magnetic disks, or optical disks.

[0153] Embodiments of the present invention also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the above embodiments of the personalized recommendation method for electricity sales packages based on user profiles.

[0154] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0155] The above provides a detailed description of the personalized electricity package recommendation method and system based on user profiles provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A method for personalized electricity package recommendation based on user profiles, characterized in that, The method comprises the following steps: Collecting device type information, historical device power consumption data, user selection preference data for different price interval packages, and current variation characteristics of each electrical device in the family; Classifying and labeling the device type information to obtain the corresponding device operation category of each electrical device, statistically analyzing the historical device power consumption data to obtain power consumption distribution data, and dividing the selection preference data according to the price acceptance degree to obtain cost sensitivity evaluation data; Based on the device operation category, the current variation characteristics are matched and analyzed to obtain the association data of device type information and power load, based on the power consumption distribution data and the association data, the power consumption characteristics of users using the same type of device are mined using a collaborative filtering algorithm to obtain group power consumption mode characteristics; Fusing the cost sensitivity evaluation data and the group power consumption mode characteristics to construct a user demand portrait containing device power consumption proportion, power demand characteristics, and cost sensitivity level; Based on the user demand portrait, the cost-effectiveness of different price interval packages is evaluated by combining the analytic hierarchy process to obtain the comprehensive scores of different price interval packages, and the comprehensive scores are sorted from high to low to generate a personalized electricity sales package recommendation result; The evaluation of the cost-effectiveness of different price interval packages based on the user demand portrait and combining the analytic hierarchy process to obtain the comprehensive scores of different price interval packages comprises: Using the analytic hierarchy process to establish an evaluation index system containing price level, service content, and demand matching degree, based on the user demand portrait, an importance judgment matrix corresponding to the evaluation index system is constructed, and the importance judgment matrix includes the weight coefficients of each evaluation index in the evaluation index system; Based on the evaluation index system, the fit degree of the different price interval packages and the user demand portrait is calculated to obtain the quantitative scores of each price interval package for each evaluation index; The quantitative scores and the corresponding weight coefficients are weighted to obtain the comprehensive scores of each price interval package. 2.The user portrait-based electricity sales package personalization recommendation method of claim 1, wherein, The fusion of the cost sensitivity evaluation data and the group power consumption mode characteristics to construct a user demand portrait containing device power consumption proportion, power demand characteristics, and cost sensitivity level comprises: Extracting power consumption proportion data of each electrical device in the peak period from the group power consumption mode characteristics to obtain the device power consumption proportion; Based on the device activation time rule and the load distribution mode in the group power consumption mode characteristics, the power level required by the user in the peak period is calculated to obtain the power demand characteristics; According to the price acceptance level preference degree in the cost sensitivity evaluation data, the cost acceptance range of the user is determined to obtain the cost sensitivity level; The device power consumption proportion, the power demand characteristics, and the cost sensitivity level are fused to construct the user demand portrait. 3.The user portrait-based electricity sales package personalization recommendation method of claim 2, characterized in that, The fusion of the device power consumption proportion, the power demand characteristics, and the cost sensitivity level to construct the user demand portrait comprises: Establishing a portrait structure template containing three dimensions of electrical device type, power demand interval, and price acceptance range; Fill the proportion of power consumption of each device in the device power consumption proportion in the power consumption device type dimension of the portrait structure template; Fill the peak period power level in the power demand feature in the power demand interval dimension of the portrait structure template; Fill the cost acceptance range in the cost sensitivity level in the price acceptance range dimension of the portrait structure template; After filling, the power consumption device type dimension, the filled power demand interval dimension, and the filled price acceptance range dimension are stored in association to form a user demand portrait. 4.The user portrait-based electricity sales package personalization recommendation method of claim 1, wherein, The evaluation index system including the price level, service content, and demand matching degree is established by using the analytic hierarchy process, and based on the user demand portrait, an importance judgment matrix corresponding to the evaluation index system is constructed, including: The first layer evaluation index including the price level, service content, and demand matching degree is established by using the analytic hierarchy process, the demand matching degree is divided into device matching degree, power matching degree, and cost matching degree to obtain the second layer evaluation index, and the first layer evaluation index and the second layer evaluation index are integrated to obtain the evaluation index system; Based on the device power consumption proportion data in the user demand portrait, the relative importance of power consumption devices in the selection of the package is determined, based on the power demand feature in the user demand portrait, the contribution degree of the peak period power demand to the selection of the package is analyzed, and based on the cost sensitivity level in the user demand portrait, the acceptance degree of the user to the price floating range is evaluated; Based on the relative importance, the contribution degree, and the acceptance degree, the evaluation index system is compared systematically to construct the importance judgment matrix. 5.The user portrait-based electricity sales package personalization recommendation method of claim 1, wherein, Based on the device running category, the current change feature is matched and analyzed to obtain the association data of the device type information and the power load, based on the power consumption distribution data and the association data, the power consumption features of users using the same type of device are mined by using the collaborative filtering algorithm to obtain the group power consumption mode feature, including: Based on the device running category, the current change feature is matched with the standard current waveform to obtain the waveform matching result, and the occurrence frequency of each device type information in the power load is counted to generate the association data of the device type information and the power load; The power consumption distribution data and the association data are combined and processed to construct a user power consumption feature vector, the similarity between each user power consumption feature vector is calculated by using the collaborative filtering algorithm to obtain a user group using the same device type information; Common power consumption habit features are extracted from the user group, wherein the power consumption habit features include device activation time rules and load distribution patterns; The power consumption habit features and the corresponding device type information are integrated to form the group power consumption mode feature. 6.The user portrait-based electricity sales package personalization recommendation method of claim 1, wherein, The device type information is classified and labeled to obtain the device running category corresponding to each power consumption device, the historical device power consumption data is counted to obtain the power consumption distribution data, and the selection tendency data is divided according to the price acceptance degree to obtain the cost sensitivity evaluation data, including: Based on the power level and start-stop frequency contained in the device type information, all electrical equipment is divided into basic running equipment and intermittent running equipment, forming a device running category; According to the preset time period division rule, the historical device power consumption data is counted, the power consumption proportion of each electrical equipment in the peak period and the flat period is calculated, and the power consumption distribution data is obtained; Based on the selection records of users in different price intervals, multiple price acceptance levels are set, the selection tendency data is divided according to the price acceptance degree, and the cost sensitivity evaluation data is obtained. 7.A user portrait-based electricity sales package personalization recommendation system, characterized in that, A user portrait-based electricity sales package personalized recommendation method for realizing any one of claims 1-6, comprising: An acquisition module for acquiring device type information, historical device power consumption data, user selection tendency data for different price interval packages, and current change characteristics of each electrical equipment in the home; A statistical module for classifying and labeling the device type information to obtain the corresponding device running category of each electrical equipment, counting the historical device power consumption data to obtain the power consumption distribution data, and dividing the selection tendency data according to the price acceptance degree to obtain the cost sensitivity evaluation data; A mining module for matching and analyzing the current change characteristics based on the device running category to obtain the association data between the device type information and the power load, using a collaborative filtering algorithm to mine the power consumption characteristics of users using similar devices based on the power consumption distribution data and the association data, and obtaining the group power consumption mode characteristics; A fusion module for fusing the cost sensitivity evaluation data and the group power consumption mode characteristics to construct a user demand portrait containing device power consumption proportion, power demand characteristics, and cost sensitivity level; An evaluation module for evaluating the performance-price ratio of different price interval packages based on the user demand portrait and combining the analytic hierarchy process to obtain the comprehensive scores of different price interval packages, sorting the comprehensive scores from high to low to generate a personalized electricity sales package recommendation result.

8. An electronic device, comprising: It includes: A memory for storing a computer program; A processor for executing the computer program to realize the steps of the user portrait-based electricity sales package personalized recommendation method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to realize the user portrait-based electricity sales package personalized recommendation method according to any one of claims 1 to 6.

Images