A smart management system for adjustable loads in power systems

Abstract

This invention discloses an intelligent management system for adjustable loads in a power system, belonging to the field of power management technology. It acquires user load data and identifies user load devices in real time based on the load data and user information; acquires device characteristic data of the load devices and predicts their usage based on this data to obtain usage prediction data for each load device; summarizes the usage prediction data of each load device to obtain load prediction data; determines the load adjustment area corresponding to each user and obtains the area target for each load adjustment area in real time; evaluates whether the load adjustment area meets the area target based on the collected load data and load prediction data in real time to obtain the load evaluation result for each load adjustment area; performs load adjustment analysis on load adjustment areas whose load evaluation results do not meet the area target to obtain a load adjustment plan for the load adjustment area, and performs load adjustment according to the plan.

Description

Technical Field

[0001] This invention belongs to the field of power management technology, specifically a power system adjustable load intelligent management system. Background Technology

[0002] In power system operation, with the large-scale integration of renewable energy and the diversification of electricity demand, the traditional grid's single "source follows load" regulation mode is no longer adequate to adapt to the volatility and intermittent characteristics of high-proportion renewable energy sources and the flexible interaction needs of users. Current load management mainly relies on manual dispatch or simple threshold control, which suffers from problems such as regulation lag, insufficient resource integration, and low response accuracy. On the demand side, users' requirements for the quality of electricity services are constantly increasing. They not only expect a stable and reliable power supply but also hope to be better satisfied in terms of electricity costs, convenience, and personalized services. Traditional power load management methods mainly rely on administrative means and simple pricing mechanisms, such as time-of-use pricing and peak-valley pricing. These methods can guide users to adjust their electricity consumption behavior to some extent, but they lack real-time and precision, making it difficult to adapt to the complex and ever-changing operational needs of modern power systems. For example, while time-of-use pricing can guide users to consume electricity during off-peak hours, it lacks effective means for optimizing the electricity consumption of specific equipment within users.

[0003] Based on this, in order to achieve intelligent management of adjustable loads, the present invention provides an intelligent management system for adjustable loads in power systems. Summary of the Invention

[0004] To address the problems of the above solutions, this invention provides an intelligent management system for adjustable loads in power systems.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A smart management system for adjustable loads in a power system includes a platform and a user terminal;

[0007] The platform includes a data acquisition module and a platform analysis module;

[0008] The acquisition module is used to acquire load data in the power system in real time; classify and label the load data according to each user to obtain the user load data corresponding to each user; send the load data to the platform analysis module and send the user load data to the user analysis module at the user end.

[0009] The platform analysis module is used to perform load adjustment analysis, determine the load adjustment area corresponding to each user, and obtain the area target of each load adjustment area in real time.

[0010] Receive load acquisition data and load prediction data of each user, and evaluate in real time whether the load adjustment area meets the regional target based on the load acquisition data and load prediction data, and obtain the load assessment results of each load adjustment area;

[0011] For load adjustment areas where the load assessment results do not meet the regional targets, load adjustment analysis is performed to obtain the load adjustment plan for the load adjustment area, and load adjustment is carried out according to the load adjustment plan.

[0012] Furthermore, based on load acquisition data and load forecast data, the system assesses in real time whether the load adjustment area meets the regional targets, including:

[0013] Establish a target calibration model, the expression of which is:

[0014] ;

[0015] In the formula: (x, MB) are the input data, where x represents the load collection data and load prediction data corresponding to the load adjustment area, MB is the area target, and x is abnormal data, indicating that x does not meet the area target; the output data is the area evaluation value MP(x, MB), and the area evaluation value is 1 or 0.

[0016] By analyzing the load collection data, load prediction data, and regional targets of the corresponding load adjustment area through the target calibration model, the regional assessment value of the load adjustment area is obtained.

[0017] When the regional assessment value is 1, the load assessment result indicates that the regional objectives are not met.

[0018] When the regional assessment value is 0, the load assessment result is that the regional target is met.

[0019] Furthermore, the platform also includes a load optimization module, which is used to perform load optimization analysis on each user, evaluate the load of each user, and obtain the load management level of each user. The load management level includes excellent load management, normal load management, and weak load management.

[0020] Users with weak load management are marked as users to be optimized. Optimization reference users are determined for these users, and load optimization management is carried out on the users to be optimized based on the optimization reference users.

[0021] Furthermore, identify the reference users for optimization for the user to be optimized, including:

[0022] Identify the load devices of the users to be optimized, and match each load device with users whose load management level is excellent to obtain each candidate reference user;

[0023] Perform priority analysis on each candidate reference user, and mark the candidate reference user with the highest priority as the optimized reference user.

[0024] Furthermore, perform priority analysis on each candidate reference user, including:

[0025] Estimate the difficulty value and effect value corresponding to load optimization management of the user to be optimized according to the candidate reference user;

[0026] Calculate the priority value of each candidate reference user according to the difficulty value and effect value, and the priority value formula is:

[0027] YU = b1×XG - b2×DZ;

[0028] In the formula: YU is the priority value; b1 and b2 are both proportionality coefficients, and the value range is 0 < b1 ≤ 1, 0 < b2 ≤ 1; XG is the effect value; DZ is the difficulty value;

[0029] Perform priority sorting in descending order according to the priority value.

[0030] The user terminal includes a user analysis module;

[0031] The user analysis module is used to perform power consumption analysis on users, obtain user load data, and identify the load equipment of users in real time according to the user load data and user information;

[0032] Obtain the device feature data of the load equipment, perform usage prediction on the load equipment according to the device feature data, and obtain the usage prediction data of each load equipment;汇总 the usage prediction data of each load equipment to obtain load prediction data, and send the load prediction data to the platform analysis module of the platform side.

[0033] Furthermore, identifying the load equipment of the user in real time according to the user load data and user information includes:

[0034] The platform party establishes a load feature library at the platform side, and the load feature library is used to store the load features of various load equipment;

[0035] Identify user information, and configure a user load feature library for the user terminal according to the user information and the load feature library;

[0036] Establish a feature recognition model, and use the feature recognition model to identify the user load data according to the user load feature library to obtain the load equipment dynamic information corresponding to the load feature data;汇总 the load equipment dynamic information corresponding to each load equipment in real time to obtain the device feature data of the load equipment;

[0037] The use of load equipment is evaluated based on load characteristic data to obtain the use evaluation results of the load equipment, which include whether the equipment is used or not.

[0038] The evaluation results will be used to output the load equipment.

[0039] Furthermore, the expression for the feature recognition model is:

[0040] ;

[0041] In the formula: (s) i C) represents the input data, s i This represents the corresponding load feature in the user load feature database, where i is the subscript, i = 1, 2, ..., n, and n is the number of load features in the user load feature database; C represents the user load data; the output data is the feature identification value TS(s) of the corresponding load feature. i (C), the feature identification value is 1 or 0;

[0042] When the feature recognition value is 1, dynamic information of the load equipment is generated based on the load equipment information corresponding to the load feature and the recognition time.

[0043] When the feature recognition value is 0, no corresponding processing is performed.

[0044] Compared with the prior art, the beneficial effects of the present invention are:

[0045] The intelligent management system for adjustable loads in power systems proposed in this invention effectively overcomes the adjustment dilemma of traditional power grids when facing the volatility and intermittent characteristics of high proportion of new energy sources and the flexible interaction needs of users. It breaks through the limitations of current load management that relies on manual scheduling or simple threshold control, resulting in adjustment lag, lack of resource integration, and low response accuracy. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a system composition diagram of the present invention. Detailed Implementation

[0048] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0049] like Figure 1 As shown, an intelligent management system for adjustable loads in a power system includes a platform and user terminals; the platform and each user terminal are connected via communication.

[0050] The platform includes a data acquisition module, a platform analysis module, and a load optimization module;

[0051] The acquisition module is used to collect relevant data of various adjustable loads in the power system in real time, including but not limited to information such as load power, voltage, current, running time, and equipment status, and mark it as load acquisition data; the load acquisition data is classified and marked according to each user to obtain the user load data corresponding to each user, that is, the user load data of a user can be quickly identified from the load acquisition data based on the user tag; the load acquisition data is sent to the platform analysis module, and the user load data is sent to the user analysis module at the user end.

[0052] In one embodiment, data acquisition can be achieved in various ways, such as by installing smart sensors on load equipment to acquire data in real time, or by obtaining relevant data from the existing monitoring system of the power system.

[0053] In one embodiment, data transmission employs high-speed and stable communication technologies, such as wireless communication (e.g., 4G / 5G, Wi-Fi) or wired communication (e.g., Ethernet, fiber optic communication), to quickly and accurately transmit the collected data to the platform analysis module. Simultaneously, it features data encryption and secure transmission functions to ensure the security and integrity of the data during transmission.

[0054] The platform analysis module is used to perform load adjustment analysis, determine the load adjustment area corresponding to each user, and divide the load adjustment area according to the existing power area division method. It is used to carry out differentiated management based on each load adjustment area. For example, the adjustment unit is divided according to the power grid topology (such as power supply area, feeder branch), and differentiated strategies are implemented for key areas such as high load density areas and industrial parks; it is divided according to load type (such as industrial, commercial, residential) or power supply characteristics (such as new energy rich area), etc.; the regional targets of each load adjustment area are obtained in real time, that is, the load management targets that need to be achieved in the load adjustment area are obtained through the power management system or other channels, such as the settings of the management personnel.

[0055] It receives load acquisition data and load forecast data from each user in real time, and evaluates whether the corresponding load adjustment area meets the regional target based on the load acquisition data and load forecast data in real time, and obtains the load assessment results of each load adjustment area.

[0056] For load adjustment areas where the load assessment results show that the regional targets are not met, load adjustment analysis is performed to obtain the load adjustment plan for the load adjustment area, and load adjustment is carried out according to the load adjustment plan.

[0057] In one embodiment, evaluating in real time whether the load adjustment area meets the regional target based on load acquisition data and load forecast data includes:

[0058] A target calibration model is established based on the isolated random forest algorithm. It is trained using a labeled training set of relevant historical data. Data that does not meet the regional target is considered anomaly, meaning the input data corresponding to the load collection data and load prediction data are considered anomalous. The expression for the target calibration model is:

[0059] ;

[0060] In the formula: (x, MB) are the input data, where x represents the load collection data and load prediction data corresponding to the load adjustment area, MB is the area target, and x is abnormal data, indicating that x does not meet the area target; the output data is the area evaluation value MP(x, MB), and the area evaluation value is 1 or 0.

[0061] By analyzing the load acquisition data, load prediction data, and regional targets of the corresponding load adjustment area through the target calibration model, the corresponding regional assessment values ​​are obtained.

[0062] When the regional assessment value is 1, the load assessment result indicates that the regional objectives are not met.

[0063] When the regional assessment value is 0, the load assessment result is that the regional target is met.

[0064] In one embodiment, the load adjustment area is evaluated in real time based on load acquisition data and load forecast data to determine whether the area meets the regional target. When the current load acquisition data and regional target are clear, the evaluation can be carried out based on various existing technologies by combining the estimated load forecast data of each user.

[0065] In one embodiment, load adjustment analysis is performed on load adjustment areas where the load assessment results do not meet the regional targets. Based on the regional targets, the analysis examines how to adjust the load to meet the regional targets, identifies the timeliness requirements of the regional targets, estimates the load devices that each user can adjust within the timeliness requirements, performs adjustment simulations based on each adjustable load device, determines the combined adjustment schemes for each load device, marks the candidate adjustment schemes, i.e., each combined adjustment scheme can achieve the load assessment results meeting the regional targets; prioritizes the candidate adjustment schemes, and selects the highest priority candidate adjustment scheme as the load adjustment scheme.

[0066] Specific priority ranking can be based on existing priority algorithms, such as evaluation based on cost, implementation difficulty (whether users cooperate, etc.), success rate, and effectiveness.

[0067] In one embodiment, load adjustment analysis is performed on load adjustment areas where the load assessment results do not meet the regional targets, and load adjustment schemes can also be determined based on other analysis methods.

[0068] The load optimization module is used to perform load optimization analysis on each user, evaluate the load of each user, and obtain the load management level of each user. The load management level includes excellent load management, normal load management, and weak load management.

[0069] Users with weak load management are marked as users to be optimized. Optimization reference users are determined for these users, and load optimization management is carried out on the users to be optimized based on the optimization reference users.

[0070] In one embodiment, load assessment for each user can be performed by the platform setting reference standards for excellent load management, normal load management, and weak load management. Based on these reference standards, the platform can evaluate the historical load data of each user to determine the corresponding load management level. For example, the platform can assess the adverse impact on the regional load, classify the degree of impact corresponding to different load management levels, and then match them accordingly. Alternatively, users can be classified according to the load equipment, and the scope corresponding to different load management levels can be divided based on the impact range achievable by each user classification.

[0071] In one embodiment, load assessment can be performed on individual users, or it can be performed based on existing methods, such as manual assessment or AI assessment.

[0072] In one embodiment, determining the optimization reference user for the user to be optimized includes:

[0073] Identify each load device of the user to be optimized, match each load device with users with excellent load management according to the load management level, and mark users with excellent load management having the same or differences within the allowable range as candidate reference users; differences within the allowable range mean that the different load devices have little impact and can be matched based on similarity algorithms, such as selecting weighted cosine similarity, weighted Euclidean distance, etc.;

[0074] Preset the weight coefficients of each load device; for example, set them according to the following weight setting rules;

[0075] By load regulation ability: the stronger the adjustability → the higher the weight;

[0076] By influence range: the greater the impact on the overall load curve → the higher the weight;

[0077] By management difficulty: the more difficult to control and the more critical → the higher the weight;

[0078] Specifically set the weight coefficients according to user needs.

[0079] Perform a priority analysis on each candidate reference user, and mark the candidate reference user with the highest priority as the optimization reference user.

[0080] In one embodiment, performing a priority analysis on each candidate reference user includes:

[0081] Estimate the difficulty value and effect value corresponding to the load optimization management of the user to be optimized according to the candidate reference user. The difficulty value is set according to the difficulty of user conversion. The difficulty value can be evaluated from the conversion ratio of the user. For example, the conversion ratio is inversely proportional to the difficulty value. When the conversion ratio is 100%, the difficulty value is 0. When the conversion ratio is 0, the difficulty value is 100. It can also be that when the conversion ratio is greater than the preset value, the difficulty value is 0, and the difficulty value of 100 is equally divided from 0 to this preset value. Subsequently, the difficulty value is calculated according to the conversion ratio. For example, the preset value is 50, that is, when the conversion ratio is greater than 50, the difficulty value is 0, and 0 - 50 corresponds to the difficulty value of 100 - 0; the effect value is the degree of beneficial effect from the user's perspective, such as changes in economic benefits. Using the difference between the two can estimate the corresponding economic difference, such as electricity bill savings and rewards, etc.;

[0082] Calculate the priority value of each candidate reference user according to the difficulty value and the effect value. The priority value formula is:

[0083] YU = b1×XG - b2×DZ;

[0084] In the formula: YU is the priority value; b1 and b2 are both proportionality coefficients, and the value range is 0 < b1 ≤ 1, 0 < b2 ≤ 1; XG is the effect value; DZ is the difficulty value;

[0085] Sort by priority value from largest to smallest.

[0086] In one embodiment, the difficulty and effectiveness values ​​of load optimization management for the user to be optimized according to the candidate reference user can be estimated. The difficulty and effectiveness values ​​can be evaluated in the existing way, for example, by building an intelligent evaluation model based on machine learning, deep learning algorithms, etc., and training the corresponding training set by manually labeling it. The training set includes input data and output data. The input data is the load information of the user to be optimized and the load information of the candidate reference user; the output data is the difficulty value and effectiveness value.

[0087] In one embodiment, because different users have different ratios for difficulty and effect values, there will be a certain deviation if calculated according to a uniform ratio. Therefore, in this embodiment, the corresponding ratio is dynamically evaluated based on user information and load optimization management records. An initial ratio is first matched based on user information, and then dynamically adjusted based on load adjustment and reception records to understand the user's acceptance of the difficulty value. It can also be displayed to the user so that the user can adjust the ratio themselves.

[0088] In one embodiment, load optimization management is performed on the user to be optimized based on the optimization reference user. The load equipment usage of the optimization reference user is unhidden and displayed to the user to be optimized, and the user to be optimized is advised or urged to make adjustments.

[0089] In one embodiment, load optimization management is performed on the user to be optimized based on the optimization reference user; alternatively, optimization management can be performed based on other existing methods.

[0090] The user terminal is used by various electricity users, such as factories, residents, shopping malls, etc.; it includes a user analysis module.

[0091] In one embodiment, the client also includes other functional modules that users need to use, such as login, display, and fee inquiry modules.

[0092] The user analysis module is used to analyze user electricity consumption, obtain user load data, and estimate the load devices owned by users in real time based on user load data and user information; obtain equipment characteristic data of load devices, predict the usage of load devices based on equipment characteristic data, obtain usage prediction data for each load device; summarize the usage prediction data of each load device to obtain load prediction data, and send the load prediction data to the platform.

[0093] In one embodiment, the load devices owned by a user are estimated in real time based on user load data and user information. The main load devices used by the user are determined based on continuously accumulated historical user load data and real-time user load data. This is because different load devices have unique characteristics such as current, voltage, and power factor when starting up, running, and in standby mode. For example, air conditioners have a sudden increase in power when starting up, small power fluctuations during operation, and are related to outdoor temperature; refrigerators have periodic start-stop cycles, with higher power consumption when the compressor is running and extremely low power consumption when in standby mode; electric water heaters have stable power consumption when heating and a sharp drop in power consumption when maintaining temperature; lighting equipment has low and stable power consumption, but may vary with day and night or usage habits.

[0094] Then, the user information is combined to make a determination; specifically, identification can be based on existing technologies.

[0095] In one embodiment, real-time estimation of the load devices owned by a user based on user load data and user information includes:

[0096] The platform provider establishes a load characteristic database on the platform side. The load characteristic database is used to store the load characteristics of various load devices that various users may have. For example, it can pre-statistically count the various load devices that various users may have, such as different production plants, shopping malls, residential buildings, etc. The platform provider can then use the historical relevant data of various load devices accumulated by the platform provider to set the corresponding load characteristics.

[0097] Identify user information and configure the user load feature library for the user terminal based on the user information and the load feature library, that is, only store the load features of each load device related to the user;

[0098] A feature recognition model is established to identify user load data based on a user load feature database, thereby obtaining dynamic information of load equipment corresponding to the load feature data, including load equipment type, identification time, and other relevant information, which facilitates subsequent analysis of the usage characteristics of the load equipment; the dynamic information of the load equipment corresponding to each load equipment is summarized in real time to obtain the equipment feature data of the load equipment;

[0099] Determine whether a user still uses the load equipment based on load characteristic data. That is, evaluate the use of the load equipment based on load characteristic data and obtain the usage evaluation results of the load equipment. The usage evaluation results include whether the user uses the equipment or not.

[0100] The evaluation results will be used to output the load equipment.

[0101] In one embodiment, the feature recognition model can be built based on existing intelligent technologies to identify user load data according to various load features in the user load feature database, determine whether it has the load features of each load device, and if so, form load device dynamic information based on the load device and the corresponding identification time; for example, the feature recognition model can be built based on machine learning, deep learning algorithms, etc.

[0102] In one embodiment, the expression for the feature recognition model is:

[0103] ;

[0104] In the formula: (s) i C) represents the input data, s i This represents the corresponding load feature in the user load feature database, where i is the subscript, i = 1, 2, ..., n, and n is the number of load features in the user load feature database; C represents the user load data; s i →C indicates that the user load data contains corresponding load characteristics from the user load feature library. For example, feature parameters corresponding to the load characteristics in the load feature library are extracted from each user load data (i.e., features such as power range, operating period, and power-time curve are extracted from the user load data). Using a "one-to-one feature matching" mode, the extracted user load feature parameters are compared with the corresponding load characteristics, for example:

[0105] Power feature matching: The power value and power fluctuation range in the user load data are compared with the rated power range and fluctuation threshold of the corresponding load equipment in the feature library to determine whether the user load power falls within the power feature range of the equipment.

[0106] Curve feature matching: Using curve similarity comparison methods (such as Dynamic Time Warping algorithm DTW), the power-time curve of user load is compared with the standard curve corresponding to the load feature in the feature library to calculate the similarity and determine the consistency of the curve shape; by combining the matching results of multiple features, it is determined whether the user load data has the corresponding load feature in the user load feature library.

[0107] The output data is the feature identification value TS(s) of the corresponding load characteristics. i (C), the feature identification value is 1 or 0; the corresponding training set is labeled with the corresponding historical load data for training.

[0108] When the feature recognition value is 1, dynamic information of the load equipment is generated based on the load equipment information corresponding to the load feature and the recognition time.

[0109] When the feature recognition value is 0, no corresponding processing is performed.

[0110] In one embodiment, the use of load equipment can be evaluated based on load characteristic data. This evaluation can be based on the type of load equipment, the usage time interval, the current time, etc. For example, some load equipment is used seasonally and is used in a specific season, so the evaluation can be based on the unused interval and the current time. Alternatively, the evaluation can be directly based on the unused interval. Specifically, the evaluation can be based on existing technologies, such as using intelligent models built based on machine learning and deep learning algorithms.

[0111] In one embodiment, the usage prediction of load equipment is made based on equipment characteristic data. This prediction can be made using existing machine learning, deep learning algorithms, or other prediction techniques based on equipment characteristic data.

[0112] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of an intelligent power system adjustable load management system as described in the above embodiments.

[0113] The above formulas are all numerical calculations after removing dimensions. The formulas are obtained by software simulation based on a large amount of data and are closest to the real situation. The preset parameters and preset thresholds in the formulas are set by those skilled in the art according to the actual situation or obtained by simulation based on a large amount of data.

[0114] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A power system adjustable load intelligent management system, characterized in that, It includes a platform side and a user side; The platform side includes a collection module and a platform analysis module; the user side includes a user analysis module; The collection module is used to collect load collection data in the power system in real time; classify and label the load collection data according to each user to obtain user load data corresponding to each user; send the load collection data to the platform analysis module and send the user load data to the user analysis module on the user side; The platform analysis module is used to conduct load adjustment analysis, determine the load adjustment area corresponding to each user, and obtain the area target of each load adjustment area in real time; Receive the load collection data and the load prediction data of each user sent by the user side, evaluate in real time whether the load adjustment area meets the area target according to the load collection data and the load prediction data, and obtain the load evaluation result of each load adjustment area; identify the load adjustment area whose load evaluation result does not meet the area target, conduct load adjustment analysis on the load adjustment area, obtain the load adjustment plan of the load adjustment area, and conduct load adjustment according to the load adjustment plan; The user analysis module is used to conduct power consumption analysis on users, obtain user load data, and identify the load equipment of users in real time according to the user load data and user information; obtain the equipment characteristic data of the load equipment, conduct usage prediction on the load equipment according to the equipment characteristic data, and obtain the usage prediction data of each load equipment; summarize the usage prediction data of each load equipment to obtain the load prediction data, and send the load prediction data to the platform analysis module on the platform side.

2. The intelligent management system for adjustable loads in a power system according to claim 1, characterized in that, The platform side further includes a load optimization module, which is used to conduct load optimization analysis on each user, evaluate the load of each user, and obtain the load management level of each user. The load management level includes excellent load management, normal load management, and weak load management; Mark the users with weak load management level as users to be optimized, determine the optimization reference users of the users to be optimized, and conduct load optimization management on the users to be optimized according to the optimization reference users.

3. The intelligent management system for adjustable loads in a power system according to claim 2, characterized in that, Determining the optimization reference users of the users to be optimized includes: Identify each load equipment of the user to be optimized, match each load equipment with the users with excellent load management level, and obtain each candidate reference user; Conduct priority analysis on each candidate reference user, and mark the candidate reference user with the highest priority as the optimization reference user.

4. The intelligent management system for adjustable loads in a power system according to claim 3, characterized in that, Conducting priority analysis on each candidate reference user includes: Estimate the difficulty value and effect value corresponding to the load optimization management of the user to be optimized according to the candidate reference user; Calculate the priority value of each candidate reference user according to the difficulty value and the effect value; Conduct priority ranking in descending order of the priority value.

5. The intelligent management system for adjustable loads in a power system according to claim 1, characterized in that, The priority value formula is: YU = b1×XG - b2×DZ; In the formula: YU is the priority value; b1 and b2 are both proportionality coefficients, and the value range is 0 < b1 ≤ 1, 0 < b2 ≤ 1; XG is the effect value; DZ is the difficulty value.

6. The intelligent management system for adjustable loads in a power system according to claim 1, characterized in that, Identifying the load equipment of users in real time according to the user load data and user information includes: The platform provider establishes a load characteristic database on the platform side, which is used to store the load characteristics of various load devices; Identify user information and configure the user load characteristic database for the user terminal based on the user information and the load characteristic database; Establish a feature recognition model, and use the feature recognition model to identify user load data based on the user load feature database to obtain the dynamic information of load equipment corresponding to the load feature data; summarize the dynamic information of load equipment corresponding to each load equipment in real time to obtain the equipment feature data of the load equipment; The use of load equipment is evaluated based on load characteristic data to obtain the use evaluation results of the load equipment, which include whether the equipment is used or not. The evaluation results will be used to output the load equipment.

7. The intelligent management system for adjustable loads in a power system according to claim 6, characterized in that, The expression for the feature recognition model is: ； In the formula: (s) i C) represents the input data, s i This represents the corresponding load feature in the user load feature database, where i is the subscript, i = 1, 2, ..., n, and n is the number of load features in the user load feature database; C represents the user load data; the output data is the feature identification value TS(s) of the corresponding load feature. i (C), the feature identification value is 1 or 0; When the feature recognition value is 1, dynamic information of the load equipment is generated based on the load equipment information corresponding to the load feature and the recognition time. When the feature recognition value is 0, no corresponding processing is performed.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements a power system adjustable load intelligent management system as described in any one of claims 1 to 7.

Images