Power load intelligent dispatching system and method based on user load portrait

Abstract

The application relates to the technical field of power dispatch systems, and discloses a power load intelligent dispatching system and method based on user load profiling, which comprises a user load profiling construction module, a user power load prediction module and a user power load dispatching module; the system is used for carrying out numerical statistics on the change trend of the historical prediction power dispatch load error ratio of the user end power load prediction time by combining numerical processing based on the historical prediction power dispatch load error ratio parameters of the user; the system is used for intelligently and accurately predicting the power theoretical actual dispatch load prediction result of the user end power load prediction time according to the historical prediction power dispatch load parameters of the user, the historical actual power dispatch load parameters of the user and the historical prediction power dispatch load error change weight parameters of the user; the system realizes scientific statistics of the power load dispatching parameters based on the change trend of the user power dispatch load error, and improves the accuracy and reliability of power load dispatching.

Description

Technical Field

[0001] This invention relates to the technical field of power dispatching systems, specifically to a power load intelligent dispatching system and method based on user load profiles. Background Technology

[0002] User load profiling is a refined descriptive label formed by analyzing user electricity consumption data, their electricity consumption behavior patterns, load characteristics, and potential trends. It helps power companies or energy service providers understand user needs, optimize service strategies, and provide data support for demand-side management and market transactions. Based on user load profiling, new power dispatching systems can be built, shifting from "load determined by source" to "load follows source," thus improving the intelligent management of power load dispatching. However, existing power dispatching systems cannot intelligently create user load profiling information, nor can they accurately predict user-end power dispatching parameters, reducing the refinement and reliability of user-end power dispatching.

[0003] Chinese invention patent CN116544934B discloses a power dispatching method and system for power load forecasting. This method involves acquiring production characteristic information and temperature characteristic information; acquiring predicted power load; acquiring idle power sources in a preset time zone; generating M power supply schemes; acquiring M first fitness values; acquiring M second fitness values; acquiring M third fitness values; and acquiring a recommended power supply scheme for power dispatching. However, this technical solution cannot correct power load dispatching errors, thus reducing the reliability of power dispatching. Summary of the Invention

[0004] (a) Technical problems to be solved

[0005] To address the issues that existing power dispatching systems cannot intelligently establish user load profiles or accurately predict user-side power dispatching parameters, thus reducing the precision and reliability of user-side power dispatching, this paper aims to achieve the goals of scientifically constructing a user load profile feature library, accurately predicting user-side power load dispatching parameters, and intelligently executing user-side power load dispatching operations.

[0006] (II) Technical Solution

[0007] This invention is achieved through the following technical solution: a power load intelligent dispatching method based on user load profiles, the method comprising the following steps:

[0008] S1. Collect historical power load dispatch feature text data from users;

[0009] S2. Based on the user's historical power load scheduling feature text data and user load profile, keywords are constructed to establish a user-end power load profile feature library, and the user load profile feature library is generated.

[0010] S3. Collect user power load forecast time data;

[0011] S4. Based on the user power load prediction time data and the user load profile feature library, perform historical predicted power dispatch load information, historical actual power dispatch load information and historical predicted power dispatch load error ratio information of the user end power load prediction time, and generate user historical predicted power dispatch load data, user historical actual power dispatch load data and user historical predicted power dispatch load error ratio data respectively.

[0012] S5. Based on the user's historical predicted power dispatch load error ratio data, perform numerical analysis and processing on the trend of the historical predicted power dispatch load error ratio change during the user's power load prediction time, and generate user's historical predicted power dispatch load error change weight data.

[0013] S6. Based on the user's historical predicted power dispatch load data, the user's historical actual power dispatch load data, and the user's historical predicted power dispatch load error change weight data, perform metering and processing of the power theoretical actual dispatch load prediction results for the user's power load prediction time, and generate user power theoretical actual dispatch load prediction data.

[0014] S7. Construct user power load scheduling prediction results data and execute user power load scheduling operations.

[0015] Preferably, the steps for collecting historical power load dispatch feature text data from users are as follows:

[0016] S11. Collect historical power load dispatching characteristic information of target users online through the power management platform, and generate text data of users' historical power load dispatching characteristics. The user's historical power load scheduling feature text data includes the target user's historical annual monthly power forecast scheduling load information, historical annual monthly actual power scheduling load information, historical annual monthly power forecast scheduling load error ratio, and user identity feature information;

[0017] Preferably, the steps for establishing a user-side power load profile feature library based on the user's historical power load scheduling feature text data and user load profile keywords, and generating the user load profile feature library are as follows:

[0018] S21. Establish user load profiles and construct keyword sets. , ;in Indicates the first User load profile construction keywords This represents the maximum number of keywords used to construct a user load profile; the keywords for constructing a user load profile include keywords for historical year and month-on-month power forecasting and dispatching load, keywords for historical year and month-on-month power actual dispatching load, keywords for historical year and month-on-month power forecasting and dispatching load error ratio, and keywords for user identity feature information.

[0019] S22. The KD-tree nearest neighbor search algorithm is used to process the user's historical power load scheduling feature text data. Construct a keyword set based on the user load profile. Keyword construction for user load profile as described in the article Historical power load dispatching feature information of users is searched, classified, and processed to construct a user load profile feature library. ,in Keywords representing the user load profile construction The corresponding user load profile feature data, wherein the user load profile feature data represents keywords constructed based on the set user load profile. We categorize and organize characteristic information of personal historical power load scheduling for each user.

[0020] Preferably, the steps for collecting user power load forecast time data are as follows:

[0021] S31. Collect the specific monthly time characteristics information of the current year's power load dispatch forecast from the target user terminal online through the power management platform, and generate user power load forecast time data. ,in The units consist of years and months.

[0022] Preferably, the steps for searching and processing historical predicted power dispatch load information, historical actual power dispatch load information, and historical predicted power dispatch load error ratio information based on the user power load prediction time data and the user load profile feature database, and generating user historical predicted power dispatch load data, user historical actual power dispatch load data, and user historical predicted power dispatch load error ratio data respectively, are as follows:

[0023] S41. The user power load prediction time data With the user load profile feature library User load profile feature data described in the document Based on time-related character matching, the system retrieves the predicted power dispatch load information, actual power dispatch load information, and predicted power dispatch load error ratio information for each month of the historical year corresponding to the user's power load prediction time. These data are then used to generate a user's historical predicted power dispatch load dataset after data identification. User historical actual power dispatch load data set Data set comparing the error ratio of historical power dispatching forecasts with user data The process generates the user's historical predicted power dispatch load data set. The user's historical actual power dispatch load data set and the user's historical predicted power dispatch load error ratio data set The specific operating steps are as follows:

[0024] S411. Initialize algorithm parameters: population size N, maximum number of iterations T;

[0025] S412. Initialize the population, calculate the fitness value, and determine the power load dispatching information search pathfinders and power load dispatching information search followers;

[0026] S413, According to the position formula In the user load profile feature library The search space is updated with power load dispatch information to search for the pathfinder's location, where t represents the current iteration number of the algorithm; This represents the pathfinder for searching power load dispatch information after the t-th iteration. In the user load profile feature library The location in the search space, This represents the pathfinder for searching power load dispatch information after the (t-1)th iteration. In the user load profile feature library The location in the search space, This represents the pathfinder for searching power load dispatch information after the (t+1)th iteration. In the user load profile feature library The location in the search space, This represents the step size factor for the pathfinder movement in the power load dispatch information search. It follows a uniform distribution in [0,1].

[0027] S414, According to the position formula Searching for updated power load dispatch information and following the user load profile feature database The location in the search space, where This represents the follower in the power load dispatch information search after the t-th iteration. In the user load profile feature library The location in the search space, This represents the follower in the power load dispatching information search after the (t+1)th iteration. In the user load profile feature library The position in the search space; This indicates the follower in the search for other power load scheduling information after the t-th iteration. In the user load profile feature library The position in the search space, the position of the follower in the search for power load dispatch information. Mobile search for Pathfinder locations is not only related to power load dispatching information. It is related to, and influenced by, the location of other power load dispatch information search followers. The impact, This indicates the search for power load dispatch information by followers within the user load profile feature database. The location distance parameter in the search space, This indicates the relationship between power load dispatching information search pioneers and power load dispatching information search followers in the user load profile feature database. The location distance parameter in the search space, , ; This represents the interaction coefficient between followers of the power load dispatch information search. This represents the attraction coefficient between a pathfinder for power load dispatching information search and a follower for power load dispatching information search. , All are uniformly distributed in [1,2]. The step size factor for the movement of power load dispatch information search followers and other power load dispatch information search followers. The step size factor for the movement of power load dispatching information search followers and power load dispatching information search pathfinders. , All are random numbers within the range [0,1].

[0028] S415. Calculate the user load profile feature library. All the user load profile feature data in the search space With the user power load forecast time data The fitness value, and in the user load profile feature library Update the search space with the time data of the user's power load prediction. The most matching user load profile feature data Global optimum;

[0029] S416. When the maximum number of iterations is met, the time data searched in step S415 that corresponds to the user's power load prediction will be output. Matching user load profile feature data The corresponding historical monthly forecast power dispatch load information, historical monthly actual power dispatch load information, and historical monthly forecast power dispatch load error ratio information are processed through data identification to generate user historical forecast power dispatch load datasets. User historical actual power dispatch load data set Data set comparing the error ratio of historical power dispatching forecasts with user data ;in , ,in This indicates the search for time data related to the user's electricity load forecast. Corresponding to Historical forecast power dispatch load data for users in the same month of a historical year. This represents the maximum number in a given historical year. The unit is kilowatt; among which The larger the number value, the more accurate the collection of the user's historical predicted power dispatch load data. Time data from the user's power load forecast The larger the corresponding time interval;

[0030] ,in This indicates the search for time data related to the user's electricity load forecast. Corresponding to Historical actual power dispatch load data for users in the same month of a historical year. The unit is kilowatt;

[0031] ,in This indicates the search for time data related to the user's electricity load forecast. Corresponding to Historical forecast power dispatch load error ratio data for the same month in historical years. ; The values ​​of include positive numbers, negative numbers, and zero; among them A positive value indicates that the user's historical predicted power dispatch load is greater than the user's historical actual power dispatch load. A negative value indicates that the user's historical predicted power dispatch load is less than the user's historical actual power dispatch load. A value of zero indicates that the user's historical predicted power dispatch load is equal to the user's historical actual power dispatch load.

[0032] Preferably, the steps for performing numerical analysis on the historical predicted power dispatch load error ratio of the user's power load forecast time based on the user's historical predicted power dispatch load error ratio data, and generating user's historical predicted power dispatch load error change weight data are as follows:

[0033] S51. Obtain the set of user historical predicted power dispatch load error ratio data. ;

[0034] S52. Use a linear regression algorithm to analyze the user's historical predicted power dispatch load error ratio data set. The user's historical predicted power dispatch load error ratio data to A linear regression numerical analysis and measurement process is performed on the trend of the historical predicted power dispatch load error ratio of users, and a weighted dataset of the historical predicted power dispatch load error change of users is generated. ,in Indicates the first The user's historical predicted power dispatch load error ratio data to The weighted data of the change in the historical predicted power dispatch load error ratio of users is generated by linear regression numerical analysis and metering processing. Indicates the first The user's historical predicted power dispatch load error ratio data to The weighted data of the change in the historical predicted power dispatch load error ratio of users is generated by linear regression numerical analysis and metering processing.

[0035] The weighted data of the user's historical predicted power dispatch load error change represents the user's historical predicted power dispatch load error ratio data. to Corresponding to The same month in historical years Numerical characteristic parameters of the changing trend of the degree of error in the historical forecast power load dispatching of users in the same month of a historical year; to The value of includes any one of positive numbers, negative numbers, and zero; among which to Positive values ​​are represented above The same month in historical years The error rate of historical power load dispatching operations by users has increased during the same months of historical years. to Negative values ​​indicate the above The same month in historical years The error rate of historical power load dispatching operations for the same month in historical years has decreased. to The value of zero indicates that the upper The same month in historical years Historically, there have been no errors in the user's predicted power load dispatching operations for the same month in previous years. to The sign of the value is the same.

[0036] Preferably, the steps for metering and processing the theoretical and actual power dispatch load prediction results for the user-end power load prediction time based on the user's historical predicted power dispatch load data, the user's historical actual power dispatch load data, and the user's historical predicted power dispatch load error change weight data, and generating the user's theoretical and actual power dispatch load prediction data are as follows:

[0037] S61. Obtain the user's historical predicted power dispatch load data. The user's historical actual power dispatch load data The user's historical predicted power dispatch load error change weighting data The user's historical predicted power dispatch load error change weighting data ;

[0038] S62, The user's historical predicted power dispatch load error ratio data The user's historical actual power dispatch load data Each of these data is compared with the weighted data of the user's historical predicted power dispatch load error change. The user's historical predicted power dispatch load error change weighting data Numerical metering processing of the theoretical and actual dispatch load forecast results for user-end power load forecasting time is performed to generate user power theoretical and actual dispatch load forecast data. ,in The measurement formula is as follows: The user power theoretical actual dispatch load prediction data mentioned above This represents the user power load forecast time data. The corresponding theoretical and actual power dispatch load forecast parameters for the specific month and time of the current year. This represents the user power load forecast time data. The corresponding theoretical forecast power dispatch load analysis parameters for the specific month and time of the current year. and All units are kilowatts.

[0039] Preferably, the steps for constructing user power load dispatching prediction result data and executing user power load dispatching operations are as follows:

[0040] S71, The user power load prediction time data The user's theoretical actual dispatch load prediction data User power load dispatch prediction data was constructed using data identification. ,in ;

[0041] S72. The power management platform uses the user's power load dispatching and prediction results data. The user power load forecasting time and user power dispatching load forecasting parameters are used to execute user power load dispatching operations.

[0042] The power load intelligent dispatching system based on user load profiles is used to implement the power load intelligent dispatching method based on user load profiles. The system includes a user load profile construction module, a user power load prediction module, and a user power load dispatching module.

[0043] The user load profile construction module includes a user historical power load scheduling feature information collection unit, a user load profile construction keyword storage unit, and a user load profile feature library establishment unit.

[0044] The user historical power load scheduling feature information collection unit collects user historical power load scheduling feature text data through the power management platform; the user load profile construction keyword storage unit is used to store user load profile construction keywords; the user load profile feature library establishment unit performs user-side power load profile feature library establishment processing based on the user historical power load scheduling feature text data and user load profile construction keywords, and generates the user load profile feature library.

[0045] The user power load forecasting module includes a user power load forecasting time acquisition unit, a user historical forecast power dispatching load information search unit, a user historical actual power dispatching load information search unit, a user historical forecast power dispatching load error ratio search unit, a user historical forecast power dispatching load error change weight parameter measurement unit, and a user power dispatching load forecasting unit.

[0046] The user power load prediction time acquisition unit collects user power load prediction time data through the power management platform; the user historical predicted power dispatch load information search unit performs historical predicted power dispatch load search processing based on the user power load prediction time data and the user load profile feature library, and generates user historical predicted power dispatch load data; the user historical actual power dispatch load information search unit performs historical actual power dispatch load search processing based on the user power load prediction time data and the user load profile feature library, and generates user historical actual power dispatch load data; the user historical predicted power dispatch load error ratio search unit performs search processing based on the user power load prediction time data and the user load profile feature library. The system searches and processes historical predicted power dispatch load error ratio information for the user's power load forecast time, and generates user historical predicted power dispatch load error ratio data. The user historical predicted power dispatch load error change weight parameter measurement unit performs numerical analysis of the historical predicted power dispatch load error ratio change trend based on the user historical predicted power dispatch load error ratio data, and generates user historical predicted power dispatch load error change weight data. The user power dispatch load forecasting unit performs measurement processing of the user's theoretical and actual dispatch load forecasting results for the user's power load forecast time based on the user historical predicted power dispatch load data, the user's historical actual power dispatch load data, and the user historical predicted power load error change weight data, and generates user theoretical and actual dispatch load forecasting data.

[0047] The user power load scheduling module includes a user power load scheduling prediction result generation unit and a user power load scheduling execution unit.

[0048] The user power load scheduling prediction result generation unit constructs user power load scheduling prediction result data based on user power load prediction time information and user power theoretical actual scheduling load prediction information; the user power load scheduling execution unit executes user power load scheduling operations based on the user power load scheduling prediction result data.

[0049] (III) Beneficial Effects

[0050] This invention provides an intelligent power load dispatching system and method based on user load profiles. It has the following beneficial effects:

[0051] I. Accurately obtain historical power load dispatch information of users through the power management platform to provide real data support for the refined construction of user-end power load profile feature library; based on the historical power load dispatch information of users, combined with intelligent search algorithms and scientifically preset user load profile keywords, efficiently and intelligently establish user-end power load profile feature library, realize intelligent and customized establishment of user load profile feature library, and improve the scientific nature of power load dispatch.

[0052] Second, by dynamically collecting user power load forecast time information through the power management platform, reliable data support is provided for the accurate statistical calculation of power load dispatch parameters for forecast time. Based on user power load forecast time data, combined with intelligent recognition algorithms and a user load profile feature library, historical predicted power dispatch load information, historical actual power dispatch load information, and historical predicted power dispatch load error ratio information for user-end power load forecast time are accurately classified and searched, improving the efficiency and accuracy of power load dispatch. Based on the user's historical predicted power dispatch load error ratio parameter, numerical processing is used to statistically analyze the changing trend of the user's historical predicted power dispatch load error ratio, achieving accurate digital prediction of power load dispatch parameters based on the changing trend of the user's historical predicted power dispatch load error. Based on the user's historical predicted power dispatch load parameters, user's historical actual power dispatch load parameters, and user's historical predicted power load error change weight parameters, intelligent and accurate prediction of the theoretical and actual dispatch load forecast results for user-end power load forecast time is performed, achieving scientific statistical calculation of power load dispatch parameters based on the changing trend of user's power dispatch load error, improving the accuracy and reliability of power load dispatch.

[0053] Third, by combining user power load forecasting time information and user power theoretical actual dispatch load forecasting information with data processing, user power load dispatching forecasting result parameters are scientifically constructed, enabling timely and efficient collection of user power load dispatching forecasting results; the power management platform autonomously and accurately executes user power load dispatching operations based on user power load dispatching forecasting result information, improving the quality and applicability of power load dispatching. Attached Figure Description

[0054] Figure 1 A schematic diagram of the modules of the intelligent power load dispatching system based on user load profiles provided by the present invention;

[0055] Figure 2 The flowchart shows the intelligent power load scheduling method based on user load profile provided by the present invention. Detailed Implementation

[0056] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] An embodiment of the intelligent power load dispatching system and method based on user load profiles is as follows: Example

[0058] Please see Figure 1 - Figure 2 A method for intelligent power load dispatching based on user load profiles, comprising the following steps:

[0059] S1. Collect historical power load dispatch feature text data from users;

[0060] S2. Based on the user's historical power load dispatch feature text data and user load profile, keywords are constructed to establish a user-end power load profile feature library and generate the user load profile feature library.

[0061] S3. Collect user power load forecast time data;

[0062] S4. Based on the user power load prediction time data and the user load profile feature library, search and process the historical predicted power dispatch load information, historical actual power dispatch load information and historical predicted power dispatch load error ratio information of the user power load prediction time, and generate user historical predicted power dispatch load data, user historical actual power dispatch load data and user historical predicted power dispatch load error ratio data respectively.

[0063] S5. Based on the historical predicted power dispatch load error ratio data of users, perform numerical analysis and processing on the changing trend of the historical predicted power dispatch load error ratio of user-end power load prediction time, and generate weighted data of the historical predicted power dispatch load error change of users.

[0064] S6. Based on the user's historical predicted power dispatch load data, the user's historical actual power dispatch load data, and the user's historical predicted power dispatch load error change weight data, the user's theoretical and actual power dispatch load prediction results are metered and processed, and user's theoretical and actual power dispatch load prediction data are generated.

[0065] S7. Construct user power load scheduling prediction results data and execute user power load scheduling operations.

[0066] For further details, please refer to Figure 1 - Figure 2The steps for collecting historical power load dispatch feature text data from users are as follows:

[0067] S11. Collect historical power load dispatching characteristic information of target users online through the power management platform, and generate text data of users' historical power load dispatching characteristics. The text data on user historical power load dispatch characteristics includes the target user's historical annual monthly power forecast dispatch load information, historical annual monthly actual power dispatch load information, historical annual monthly power forecast dispatch load error ratio, and user identity characteristic information;

[0068] The steps for establishing and generating a user load profile feature library based on historical power load dispatch feature text data and user load profile keywords are as follows:

[0069] S21. Establish user load profiles and construct keyword sets. , ;in Indicates the first User load profile construction keywords This represents the maximum number of keywords used to construct a user load profile. Keywords for constructing a user load profile include keywords for historical year and month-on-month power forecasting and dispatching load, keywords for historical year and month-on-month power actual dispatching load, keywords for historical year and month-on-month power forecasting and dispatching load error ratio, and keywords for user identity feature information.

[0070] S22. Use the KD-tree nearest neighbor search algorithm to extract user historical power load scheduling feature text data. Build a keyword set based on user load profiles Keyword construction for user load profile Historical power load dispatching feature information of users is searched, classified, and processed to construct a user load profile feature library. ,in Keywords for building user load profiles The corresponding user load profile feature data represents keywords constructed based on the defined user load profile. We categorize and organize characteristic information of personal historical power load scheduling for each user.

[0071] By collecting historical power load dispatching characteristic information of users, the power management platform accurately obtains historical power load dispatching information of users, providing real data support for the refined construction of user-end power load profile feature library; the user load profile feature library establishment unit, based on the historical power load dispatching information of users, combined with intelligent search algorithms and scientifically preset user load profile construction keywords, efficiently and intelligently establishes the user-end power load profile feature library, realizing intelligent and customized establishment of user load profile feature library, and improving the scientific nature of power load dispatching.

[0072] For further details, please refer to Figure 1 - Figure 2 The steps for collecting user power load forecast time data are as follows:

[0073] S31. Collect the specific monthly time characteristics information of the current year's power load dispatch forecast from the target user terminal online through the power management platform, and generate user power load forecast time data. ,in The units consist of years and months.

[0074] The steps for searching and processing historical predicted power dispatch load information, historical actual power dispatch load information, and historical predicted power dispatch load error ratio information based on user power load forecast time data and user load profile feature database, and generating user historical predicted power dispatch load data, user historical actual power dispatch load data, and user historical predicted power dispatch load error ratio data are as follows:

[0075] S41. Transfer user power load forecast time data User load profile feature library User load profile feature data Based on time-related character matching, the system retrieves the predicted power dispatch load information, actual power dispatch load information, and predicted power dispatch load error ratio information for each month of the historical year corresponding to the user's power load prediction time. These data are then used to generate a user's historical predicted power dispatch load dataset after data identification. User historical actual power dispatch load data set Data set comparing the error ratio of historical power dispatching forecasts with user data Execute to generate user historical predicted power dispatch load data set User historical actual power dispatch load data set Data set comparing the error ratio of historical power dispatching forecasts with user data The specific operating steps are as follows:

[0076] S411. Initialize algorithm parameters: population size N, maximum number of iterations T;

[0077] S412. Initialize the population, calculate the fitness value, and determine the power load dispatching information search pathfinders and power load dispatching information search followers;

[0078] S413, According to the position formula In the user load profile feature library The search space is updated with power load dispatch information to search for the pathfinder's location, where t represents the current iteration number of the algorithm; This represents the pathfinder for searching power load dispatch information after the t-th iteration. In the user load profile feature library The location in the search space, This represents the pathfinder for searching power load dispatch information after the (t-1)th iteration. In the user load profile feature library The location in the search space, This represents the pathfinder for searching power load dispatch information after the (t+1)th iteration. In the user load profile feature library The location in the search space, This represents the step size factor for the pathfinder movement in the power load dispatch information search. It follows a uniform distribution in [0,1].

[0079] S414, According to the position formula Update power load dispatch information by searching followers in the user load profile feature database. The location in the search space, where This represents the follower in the power load dispatch information search after the t-th iteration. In the user load profile feature library The location in the search space, This represents the follower in the power load dispatching information search after the (t+1)th iteration. In the user load profile feature library The position in the search space; This indicates the follower in the search for other power load scheduling information after the t-th iteration. In the user load profile feature library The position in the search space, the position of the follower in the search for power load dispatch information. Mobile search for Pathfinder locations is not only related to power load dispatching information. It is related to, and influenced by, the location of other power load dispatch information search followers. The impact, This indicates the search for power load dispatch information by followers within the user load profile feature database. The location distance parameter in the search space, This indicates the relationship between power load dispatching information search pioneers and power load dispatching information search followers in the user load profile feature database. The location distance parameter in the search space, , ; This represents the interaction coefficient between followers of the power load dispatch information search. This represents the attraction coefficient between a pathfinder for power load dispatching information search and a follower for power load dispatching information search. , All are uniformly distributed in [1,2]. The step size factor for the movement of power load dispatch information search followers and other power load dispatch information search followers. The step size factor for the movement of power load dispatching information search followers and power load dispatching information search pathfinders. , All are random numbers within the range [0,1].

[0080] S415. Calculate the user load profile feature library. All user load profile feature data in the search space With user power load forecast time data The fitness value, and in the user load profile feature library Update the search space with the time data of the user's power load forecast. Most matching user load profile feature data Global optimum;

[0081] S416. When the maximum number of iterations is met, the time data searched in step S415 corresponding to the user's power load forecast will be output. Matching user load profile feature data The corresponding historical monthly forecast power dispatch load information, historical monthly actual power dispatch load information, and historical monthly forecast power dispatch load error ratio information are processed through data identification to generate user historical forecast power dispatch load datasets. User historical actual power dispatch load data set Data set comparing the error ratio of historical power dispatching forecasts with user data ;in , ,in This indicates the search for time-based data related to user electricity load forecasts. Corresponding to Historical forecast power dispatch load data for users in the same month of a historical year. This represents the maximum number in a given historical year. The unit is kilowatt; among which The larger the number, the more historical predicted power dispatch load data of the user is collected. Time data for user power load forecast The larger the corresponding time interval;

[0082] ,in This indicates the search for time-based data related to user electricity load forecasts. Corresponding to Historical actual power dispatch load data for users in the same month of a historical year. The unit is kilowatt;

[0083] ,in This indicates the search for time-based data related to user electricity load forecasts. Corresponding to Historical forecast power dispatch load error ratio data for the same month in historical years. ; The values ​​of include positive numbers, negative numbers, and zero; among them A positive value indicates that the user's historical predicted power dispatch load is greater than the user's historical actual power dispatch load. A negative value indicates that the user's historical predicted power dispatch load is less than the user's historical actual power dispatch load. A value of zero indicates that the user's historical predicted power dispatch load is equal to the user's historical actual power dispatch load.

[0084] The steps for performing numerical analysis on the historical predicted power dispatch load error ratio data of users, and generating weighted data on the historical predicted power dispatch load error changes, are as follows:

[0085] S51. Obtain the user's historical predicted power dispatch load error ratio data set. ;

[0086] S52. Use a linear regression algorithm to analyze the user's historical predicted power dispatch load error ratio data set. Historical forecast power dispatch load error ratio data for Chinese users to A linear regression numerical analysis and measurement process is performed on the trend of the historical predicted power dispatch load error ratio of users, and a weighted dataset of the historical predicted power dispatch load error change of users is generated. ,in Indicates the first The error ratio of the user's historical predicted power dispatch load data to The weighted data of the change in the historical predicted power dispatch load error ratio of users is generated by linear regression numerical analysis and metering processing. Indicates the first The error ratio of the user's historical predicted power dispatch load data to The system generates weighted data on the changes in the historical predicted power dispatch load error ratio by performing linear regression numerical analysis and metering processing. This weighted data represents the historical predicted power dispatch load error ratio data. to The corresponding historical year and the same month Numerical characteristic parameters of the changing trend of the degree of error in the historical forecast power load dispatching of users in the same month of a historical year; to The value of includes any one of positive numbers, negative numbers, and zero; among which to Positive values ​​are represented above The same month in historical years The error rate of historical power load dispatching operations by users has increased during the same months of historical years. to Negative values ​​indicate the above The same month in historical years The error rate of historical power load dispatching operations for the same month in historical years has decreased. to The value of zero indicates that the upper The same month in historical years Historically, there have been no errors in the user's predicted power load dispatching operations for the same month in previous years. to The sign of the value is the same.

[0087] The steps for metering and processing the theoretical and actual power dispatch load prediction results for the user's power load prediction time based on the user's historical predicted power dispatch load data, the user's historical actual power dispatch load data, and the user's historical predicted power dispatch load error change weight data, and generating the user's theoretical and actual power dispatch load prediction data are as follows:

[0088] S61. Obtain historical predicted power dispatch load data from users. Historical actual power dispatch load data of users User historical predicted power dispatch load error change weighting data User historical predicted power dispatch load error change weighting data ;

[0089] S62. Use the user's historical predicted power dispatch load error ratio data. Historical actual power dispatch load data of users Weighted data of changes in power dispatch error in user's historical forecasts. User historical predicted power dispatch load error change weighting data Numerical metering processing of the theoretical and actual dispatch load forecast results for user-end power load forecasting time is performed to generate user power theoretical and actual dispatch load forecast data. ,in The measurement formula is as follows: Among them, the user's theoretical actual dispatch load prediction data Represents user power load forecast time data The corresponding theoretical and actual power dispatch load forecast parameters for the specific month and time of the current year. Represents user power load forecast time data The corresponding theoretical forecast power dispatch load analysis parameters for the specific month and time of the current year. and All units are kilowatts.

[0090] The user power load forecasting time acquisition unit dynamically collects user power load forecasting time information through the power management platform, providing reliable data support for accurate statistical analysis of forecasting time power load dispatching parameters. The user historical forecasting power dispatching load information search unit, the user historical actual power dispatching load information search unit, and the user historical forecasting power dispatching load error ratio search unit work together to accurately classify and search for historical forecasting power dispatching load information, historical actual power dispatching load information, and historical forecasting load error ratio information based on user power load forecasting time data, combined with intelligent recognition algorithms and a user load profile feature library. This improves the efficiency and accuracy of power load dispatching. The user historical forecasting power dispatching load error ratio... The difference change weight parameter measurement unit, based on the user's historical predicted power dispatch load error ratio parameter and combined with numerical processing, performs numerical statistics on the historical predicted power dispatch load error ratio change trend of the user's power load prediction time, realizing the digital and accurate prediction of power load dispatch parameters based on the user's historical predicted power dispatch load error change trend; the user power dispatch load prediction unit, based on the user's historical predicted power dispatch load parameters, the user's historical actual power dispatch load parameters, and the user's historical predicted power load error change weight parameter, performs intelligent and accurate prediction of the user's power load prediction time based on the theoretical and actual dispatch load prediction results, realizing the scientific statistical analysis of power load dispatch parameters based on the user's power dispatch load error change trend, improving the accuracy and reliability of power load dispatch.

[0091] For further details, please refer to Figure 1 - Figure 2 The steps for constructing user power load dispatching prediction data and executing user power load dispatching operations are as follows:

[0092] S71, Use user power load forecast time data User power theoretical and actual dispatch load forecast data User power load dispatch prediction data was constructed using data identification. ,in ;

[0093] S72, The power management platform uses user power load dispatching and forecasting data. The user power load forecasting time and user power dispatching load forecasting parameters are used to execute user power load dispatching operations.

[0094] The user power load dispatching prediction result generation unit scientifically constructs user power load dispatching prediction result parameters based on user power load prediction time information, user power theoretical and actual dispatching load prediction information, and data processing, thereby enabling timely and efficient collection of user power load dispatching prediction results. The user power load dispatching execution unit enables the power management platform to autonomously and accurately execute user power load dispatching operations based on user power load dispatching prediction result information, thereby improving the quality and applicability of power load dispatching.

[0095] Please see Figure 1 - Figure 2 A power load intelligent dispatching system based on user load profiles is used to implement a power load intelligent dispatching method based on user load profiles. The system includes a user load profile construction module, a user power load prediction module, and a user power load dispatching module.

[0096] The user load profile construction module includes a user historical power load dispatch feature information collection unit, a user load profile construction keyword storage unit, and a user load profile feature library establishment unit.

[0097] The system includes: a user historical power load dispatch feature information collection unit, which collects user historical power load dispatch feature text data through the power management platform; a user load profile construction keyword storage unit, which stores user load profile construction keywords; and a user load profile feature library establishment unit, which processes the user's power load profile feature library establishment based on the user historical power load dispatch feature text data and user load profile construction keywords, and generates the user load profile feature library.

[0098] The user power load forecasting module includes a user power load forecasting time acquisition unit, a user historical forecast power dispatching load information search unit, a user historical actual power dispatching load information search unit, a user historical forecast power dispatching load error ratio search unit, a user historical forecast power dispatching load error change weight parameter measurement unit, and a user power dispatching load forecasting unit.

[0099] The user power load forecasting time acquisition unit collects user power load forecasting time data through the power management platform; the user historical forecasted power dispatching load information search unit searches for historical forecasted power dispatching load data based on the user power load forecasting time data and the user load profile feature library, and generates user historical forecasted power dispatching load data; the user historical actual power dispatching load information search unit searches for historical actual power dispatching load data based on the user power load forecasting time data and the user load profile feature library, and generates user historical actual power dispatching load data; the user historical forecasted power dispatching load error ratio search unit searches for user power load forecasting time error ratio data based on the user power load forecasting time data and the user load profile feature library. The system performs a search and processing of historical predicted power dispatch load error ratio information for the load forecasting time, and generates historical predicted power dispatch load error ratio data for users. A user historical predicted power dispatch load error change weight parameter measurement unit performs numerical analysis of the historical predicted power dispatch load error ratio change trend based on the user historical predicted power dispatch load error ratio data, and generates user historical predicted power dispatch load error change weight data. A user power dispatch load forecasting unit performs measurement processing of the user's theoretical and actual dispatch load forecasting results for the user's power load forecasting time based on the user historical predicted power dispatch load data, user historical actual power dispatch load data, and user historical predicted power dispatch load error change weight data, and generates user theoretical and actual dispatch load forecasting data.

[0100] The user power load dispatching module includes a user power load dispatching prediction result generation unit and a user power load dispatching execution unit.

[0101] The user power load dispatching prediction result generation unit constructs user power load dispatching prediction result data based on user power load prediction time information and user power theoretical actual dispatching load prediction information; the user power load dispatching execution unit executes user power load dispatching operations based on the user power load dispatching prediction result data of the power management platform.

[0102] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A power load intelligent scheduling method based on user load profiling, characterized in that, The method comprises the following steps: S1, collecting user historical power load scheduling feature text data; The S1 comprises the following steps: S11, collecting historical power load scheduling characteristic information of the target user terminal online through the power management platform, and generating user historical power load scheduling characteristic text data ; S2, performing user end power load image feature library establishment processing, and generating a user load image feature library; The S2 comprises the following steps: S21, establish a user load profile construction keyword set , ; wherein denotes the th user load profile construction keyword, denotes the maximum value of the number of user load profile construction keywords S22. The KD-tree nearest neighbor search algorithm is used to... According to the above The above Historical power load dispatching feature information of users is searched, classified, and processed to construct a user load profile feature library. ,in Indicates the The corresponding user load profile feature data, wherein the user load profile feature data represents the data based on the set... Personalized historical power load scheduling feature information is categorized and organized for each user. S3, collecting user power load prediction time data; S4, performing user end power load prediction time historical prediction power scheduling load information, historical actual power scheduling load information and historical prediction power scheduling load error ratio information search processing, and respectively generating user historical prediction power scheduling load data, user historical actual power scheduling load data and user historical prediction power scheduling load error ratio data; S5, performing user end power load prediction time historical prediction power scheduling load error ratio change trend numerical analysis processing, and generating user historical prediction power scheduling load error change weight data; S6, performing user end power load prediction time power theoretical actual scheduling load prediction result measurement processing, and generating user power theoretical actual scheduling load prediction data; S7, constructing user power load scheduling prediction result data and executing user power load scheduling operation. 2.The method of claim 1, wherein: The S3 comprises the following steps: S31, collecting the current annual specific month time characteristic information of the power load scheduling prediction of the target user terminal through the power management platform, and generating user power load prediction time data wherein The unit is composed of years and months. 3.The method of claim 2, wherein: The S4 comprises the following steps: S41, the above With the The above Based on time-related character matching, the system retrieves the predicted power dispatch load information, actual power dispatch load information, and predicted power dispatch load error ratio information for each month of the historical year corresponding to the user's power load prediction time. These data are then used to generate a user's historical predicted power dispatch load dataset after data identification. User historical actual power dispatch load data set Data set comparing the error ratio of historical power dispatching forecasts with user data The process generates the user's historical predicted power dispatch load data set. The user's historical actual power dispatch load data set and the user's historical predicted power dispatch load error ratio data set The specific operating steps are as follows: S411, initializing algorithm parameters, population number N, and maximum iteration number T; S412, initializing population, calculating fitness value, and determining power load scheduling information search pathfinder and power load scheduling information search follower; S413, updating the power load dispatching information search explorer position in the search space according to the position formula ; S414, updating the position of the search follower in the search space of the position formula according to the power load scheduling information ​ S415, Calculate the above All of the above in the search space With the The fitness value, and in the Update the search space to find results related to the above. The most matching Global optimum; S416. When the maximum number of iterations is satisfied, output the results from step S415, which searches for results matching the given condition. The matching The corresponding historical monthly forecast power dispatch load information, historical monthly actual power dispatch load information, and historical monthly forecast power dispatch load error ratio information are processed through data identification to generate user historical forecast power dispatch load datasets. User historical actual power dispatch load data set Data set comparing the error ratio of historical power dispatching forecasts with user data ;in , ,in Indicates the search and the Corresponding to Historical forecast power dispatch load data for users in the same month of a historical year. This represents the maximum number in a given historical year. The unit is kilowatt; among which The larger the number value, the more data has been collected. Distance The larger the corresponding time interval; ,in Indicates the search and the Corresponding to Historical actual power dispatch load data for users in the same month of a historical year. The unit is kilowatt; ,in This indicates the search for time data related to the user's electricity load forecast. Corresponding to Historical forecast power dispatch load error ratio data for users in the same month of a historical year. 4.The method of claim 3, wherein: The S5 comprises the following steps: S51, obtaining the ; S52, The linear regression algorithm is used to analyze the... The above to A linear regression numerical analysis and measurement process is performed on the trend of the historical predicted power dispatch load error ratio of users, and a weighted dataset of the historical predicted power dispatch load error change of users is generated. ,in Indicates the first The next time on the to The weighted data of the change in the historical predicted power dispatch load error ratio of users is generated by linear regression numerical analysis and metering processing. Indicates the first The next time on the to The weighted data of the change in the historical predicted power dispatch load error ratio of users is generated through linear regression numerical analysis and metering processing. 5.The method of claim 4, wherein: The S6 comprises the following steps: S61、acquire the , the , the , the ; S62, the above The above respectively with the above The above Numerical metering processing of the theoretical and actual dispatch load forecast results for user-end power load forecasting time is performed to generate user power theoretical and actual dispatch load forecast data. ,in The measurement formula is as follows: , wherein Indicates the The corresponding theoretical and actual power dispatch load forecast parameters for the specific month and time of the current year. Indicates the The corresponding theoretical forecast power dispatch load analysis parameters for the specific month and time of the current year. and All units are kilowatts. 6.The method of claim 5, wherein: The S7 comprises the following steps: S71、constructing the user power load scheduling prediction result data through the data identification , the data identification constructing the user power load scheduling prediction result data through the data identification ; S72, the power management platform performs a user power load dispatching operation according to the user power load prediction time and the user power dispatching load prediction parameter. S72, the power management platform performs a user power load dispatching operation according to the user power load prediction time and the user power dispatching load prediction parameter.

7. A power load intelligent scheduling system based on user load portrait, used for implementing the power load intelligent scheduling method based on user load portrait in any one of claims 1-6, characterized in that: The system comprises a user load image construction module, a user power load prediction module and a user power load scheduling module.

Images