An energy scheduling optimization method and system based on big data

Abstract

The application discloses an energy scheduling optimization method and system based on big data, relates to the technical field of community intelligent energy scheduling, is suitable for a community energy infrastructure inclusive access scene, collects operation data of community distributed photovoltaic, distributed wind power, energy storage devices, charging piles and resident flexible load and resident inclusive attribute data, completes multi-source data space-time fusion, automatically marks a rigid load exemption label for inclusive guarantee users, constructs an energy system panoramic situation map with an inclusive access adaptation layer, clusters and extracts user energy use preferences and response potential characteristics to form a three-dimensional user portrait, establishes a multi-objective collaborative scheduling optimization model to generate a scheduling strategy, and supports an inclusive tilt benefit sharing distribution mechanism. The application can effectively reduce community comprehensive energy consumption cost, improve local renewable energy consumption level, guarantee basic energy use rights and interests of special inclusive groups in the whole process, and balance energy utilization efficiency and public service fairness.

Description

Technical Field

[0001] This invention relates to the field of intelligent scheduling and optimization technology for community energy systems, and in particular to an energy scheduling and optimization method and system based on big data. Background Technology

[0002] With the popularization and promotion of distributed energy technology, more and more communities are beginning to connect to small energy units such as rooftop distributed photovoltaic power generation, distributed wind power generation, user-side energy storage, and electric vehicle charging facilities. The community energy system is gradually transforming from a traditional one-way power receiving network to a local energy system with multi-source interaction between power generation, grid, load, and storage. The traditional unified dispatch mode relying on the large power grid is mainly designed for large-scale energy facilities at the transmission and distribution level, which is difficult to adapt to the dispatch needs of a large number of small energy units with dispersed ownership and fluctuating operating status at the community level. Most existing community energy dispatch schemes take system operating efficiency and renewable energy consumption as the core dispatch objectives, and do not adequately consider the differences in energy security needs of different groups within the community. For low-income groups, people with disabilities, and other groups with special energy needs, their basic living loads are often indiscriminately included in the adjustable load range during demand response, which poses a risk of adjusting their basic electricity needs. Currently, energy data in most communities is managed by different operating entities. The data collection standards for electricity meters, photovoltaic systems, energy storage systems, and charging piles are not uniform, and there are differences in time calibration benchmarks and spatial positioning accuracy. In addition, there are compliance risks such as unclear collection scope and inadequate user authorization during the data collection process. It is difficult to effectively integrate multi-source data to form a globally consistent operational view, which can easily lead to deviations between dispatch instructions and the actual on-site operating status.

[0003] Existing user modeling processes for community energy dispatch mostly treat residential users as homogeneous dispatch objects. User feature extraction mainly focuses on dimensions such as adjustable load capacity and historical response frequency, without incorporating the user's inclusive security attributes as core features into the user profile system. Nor are dedicated rigid load exemption boundaries set in dispatch constraints. During the dispatch model solution process, it is impossible to provide rigid guarantees for the basic energy needs of inclusive users. In the event of local energy supply and demand tensions, dispatch algorithms often prioritize the adjustment of small user loads with fast response speeds and low adjustment thresholds. A considerable proportion of these loads belong to the residential electricity consumption of inclusive security groups, which can easily squeeze the energy rights of vulnerable groups.

[0004] Existing community energy dispatch contribution feedback rules mostly calculate contribution shares based on the absolute value of electricity adjustments made by users, lacking a reasonable tilt mechanism for the inclusive energy security group. The dispatch process is largely a one-way instruction from the management end, lacking a two-way interactive feedback channel with residential users. Users' willingness to use energy independently cannot be smoothly transmitted to the dispatch decision-making stage, which can easily lead to a decrease in residents' willingness to participate in the long run. Most community energy system resilience assessments mainly focus on the overall power supply capacity and the operating status of key equipment, failing to prioritize the power supply status of critical loads for inclusive energy security users. This makes it impossible to identify power supply gaps for inclusive energy security users in advance under special scenarios such as extreme weather events and external power grid failures.

[0005] Existing community energy user participation incentive mechanisms mostly adopt uniform point calculation rules, failing to set appropriate incentive standards for the participation capabilities of inclusive energy users. The value of points redemption is often fixed, making it prone to fluctuations with changes in point reserves and difficult to maintain a stable long-term incentive effect. As the inclusive access to community energy infrastructure continues to advance, community energy dispatch needs to simultaneously meet multiple objectives, including system safety operation, full renewable energy consumption, and protection of the rights of special groups. Existing dispatch methods, in terms of data fusion, user modeling, constraint settings, allocation mechanisms, and interaction processes, are no longer adequate to meet the construction requirements of balancing operational efficiency and the fairness of public services. Furthermore, some dispatch solutions utilizing big data and artificial intelligence technologies suffer from poor model-scenario integration and non-standard feature selection, easily introducing unnecessary user-sensitive data collection processes and even algorithmic bias due to inappropriate feature dimension selection, deviating from the development direction of equalizing community public services. Summary of the Invention

[0006] This invention proposes an energy dispatch optimization method and system based on big data to solve the problems mentioned in the prior art.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: an energy dispatch optimization method based on big data, comprising the following steps: S1. Construct a digital twin model of the community energy infrastructure, collect real-time and historical operation data within the community, collect community energy operation data and resident user attribute data, perform multi-source data fusion, identify inclusive security users and mark rigid load exemption attributes; S2. Perform cluster analysis on residential users' energy consumption data, extract energy consumption preferences, demand response potential and inclusive attributes, and establish a user profile database; S3. Construct a community energy collaborative dispatch optimization model with the joint objectives of minimizing community energy costs, maximizing renewable energy consumption rate, and maximizing the protection of the rights and interests of inclusive users. Under the constraints of distribution network security, rigid load exemption, and user energy preference, solve the optimal output plan of distributed energy and the user adjustable load adjustment scheme. S4. Establish a community energy benefit-sharing and distribution mechanism with inclusive incentives. Calculate the basic contribution value based on the user's contribution to the community's energy supply and demand balance. Multiply the basic contribution value of inclusive users by the inclusive incentive coefficient to obtain the final contribution value. S5. Push the scheduling strategy and allocation results to the user terminal, receive user feedback and iteratively optimize. When user participation is below the threshold, increase the inclusive tilt coefficient. After the participation reaches the target, lock the strategy and execute it.

[0008] Preferably, it also includes: constructing a community energy resilience assessment mechanism to quantitatively assess the community's independent operation capability under extreme weather or external power grid failure scenarios. The assessment process prioritizes simulating the power supply guarantee situation of critical loads of inclusive users, including simulating the maximum sustainable power supply duration of the community energy system in islanded operation mode, the critical load guarantee rate of inclusive users, and the evolution trend of the state of charge of the energy storage system.

[0009] Preferably, it also includes: establishing a community user participation incentive mechanism, issuing community energy points to users who participate in demand response and green electricity consumption, with the accumulated points being proportional to the carbon emission reduction during the dispatch cycle, and the accumulated points of universally benefited users being multiplied by an additional universal benefit bonus coefficient of 1.2 to 1.5 times; users can use community energy points to redeem electricity bill discounts, priority access to charging piles, or the right to lease community public energy storage capacity in the community energy trading market, with a dynamic pricing mechanism for point redemption, adjusting the redemption value in real time according to the supply and demand ratio of points.

[0010] Preferably, in step S2, the energy consumption preference pattern vector of residential users is extracted by analyzing the distribution pattern of hourly electricity consumption data of residential users in different seasons, different workday types, and different weather conditions. Specifically, the hourly electricity consumption data of a year is divided into twenty-four analysis dimensions according to workdays and rest days, spring, summer, autumn and winter, and three types of weather (sunny, cloudy and rainy). The average electricity consumption and standard deviation of each time period are statistically analyzed, and after normalization, the universal benefit attribute 0 / 1 identifier is concatenated to generate the energy consumption preference pattern vector. The demand response potential level is divided into three levels: high, medium and low, according to the ratio of the user's adjustable load capacity to the peak load.

[0011] Preferably, the energy consumption preference pattern vector of resident user i in step S2 is constructed by the following formula: ; in Let i be the average power consumption of residential user i in the k-th analysis period. Let i be the maximum average power consumption of residential user i across all analysis periods. Let be the standard deviation of the power consumption of residential user i in the k-th analysis period. Let M be the maximum standard deviation of electricity consumption of residential user i across all analysis periods, and M be the total number of analysis periods, taking a positive integer value. Inclusive attribute identifier for residential users Preferably, in step S3, the community energy collaborative dispatch optimization model simultaneously introduces a voltage deviation penalty term for community distribution network nodes, a charge-discharge cycle life loss penalty term for energy storage devices, and a load adjustment penalty term for inclusive users. This transforms the voltage amplitude constraints of each node in the community distribution network into soft constraints, and penalizes the degree of voltage exceeding the limit in the optimization objective function using a weighted sum of squares. The charge-discharge cycle life loss of the energy storage device is calculated by counting the equivalent charge-discharge cycles within the dispatch cycle using the rainflow counting method. The equivalent charge-discharge cycle count is multiplied by the equivalent capacity decay cost per cycle and then included in the energy storage operation cost term in the optimization objective function.

[0012] Preferably, in step S4, the actual contribution of residential users to the supply and demand balance of the community energy system is calculated using a multi-dimensional weighted contribution evaluation model. First, the net load adjustment amount per hour for each residential user during the scheduling cycle is calculated. The net load adjustment amount is the difference between the actual power consumption and the benchmark power consumption of the residential user. Second, the cumulative renewable energy power fed back to the community power grid by each residential user during the scheduling cycle is calculated and converted into an equivalent economic contribution value by weighting according to the time-of-use electricity price. Then, the net load adjustment amount and the equivalent economic contribution value are substituted into the preset contribution scoring function to obtain the load adjustment contribution score and the green energy contribution score. The two contribution scores are weighted and summed to obtain the basic contribution evaluation value of the residential user. The basic contribution evaluation value is multiplied by the corresponding inclusive tilt coefficient to obtain the final contribution evaluation value, which serves as the weighting coefficient for benefit distribution.

[0013] Preferably, the final contribution evaluation value of residential users to the supply and demand balance of the community energy system in step S4 is calculated by the following formula: ;in The final contribution evaluation value for resident user j; This is the universal benefit tilt coefficient; Contribute dimensional weighting factors to load regulation; The contribution dimension weighting factor for green energy is the absolute value of the cumulative net load adjustment of residential user j within the scheduling cycle; This represents the sum of the absolute values ​​of the cumulative net load adjustment for all residents within the community during the scheduling cycle. This refers to the renewable energy power fed back to the community power grid by residential user j during the dispatch cycle. The total renewable energy electricity fed back to the community grid by all residents within the community during the dispatch cycle is the total amount of renewable energy electricity. The benefit-sharing allocation for resident user j is equal to the total amount of the community benefit pool multiplied by the contribution evaluation value. The normalized percentage.

[0014] Preferably, in step S3, the community energy collaborative scheduling optimization model is solved using a mixed-integer nonlinear programming method. A constraint decoupling strategy based on Lagrange relaxation is introduced, decomposing the distribution network security constraints and rigid load exemption constraints into independent subproblems for each node, which are then solved in parallel. The coupling relationship between nodes is updated using Lagrange multipliers, and the global optimal solution is obtained after convergence. The optimization model uses a 24-hour rolling window and a 15-minute scheduling time resolution. At the start of each scheduling cycle, the parameters are refreshed based on the latest load forecast and renewable energy output forecast, and the solution is recalculated. Preferably, an energy dispatch optimization system based on big data includes: The community energy digital twin module is equipped with a multi-source data access interface, a universal user identification unit, and a spatiotemporal alignment and fusion unit, which are used to collect operation data of various energy equipment in the community and user universal attribute data. The user profile building module is equipped with user behavior clustering units, demand response potential assessment units, and inclusive attribute labeling units, which are used to extract multi-dimensional energy consumption characteristics of users and generate a three-dimensional user profile containing energy consumption preferences, response potential, and inclusive attributes. The collaborative scheduling optimization module is equipped with a multi-objective optimization solution unit, a rigid load exemption constraint unit, and a user preference matching unit, which are used to generate community energy collaborative scheduling strategies that meet the dual objectives of efficiency and fairness. The profit distribution module is equipped with a contribution calculation unit, a universal benefit adjustment unit, and a revenue calculation unit, which are used to distribute revenue to users according to the contribution ratio with universal benefit bonus. The interactive execution module is configured with a user mobile terminal interaction unit, a participation dynamic adjustment unit, and an instruction issuance unit, which are used to receive user feedback, iteratively optimize the scheduling scheme, and issue instructions for execution.

[0015] Compared with existing technologies, the beneficial effects of this invention are: The energy dispatch optimization method and system proposed in this technology can be adapted to the application scenario of universal access to community energy infrastructure. At the level of community energy data fusion, through the multi-source data mapping and fusion process under a unified spatiotemporal coordinate system, the data barriers of energy devices from different sources are broken down, forming a panoramic situation map of the community energy system covering all elements of source, grid, load and storage. The data collection process is strictly limited to the necessary scope of energy dispatch and strictly implements the user informed consent requirements, which can reduce unnecessary collection of sensitive user information and avoid data compliance risks.

[0016] The solution establishes an inclusive access adaptation layer at the data layer, which can automatically identify inclusive users and mark their basic living loads with rigid exemption tags, providing support for the protection of inclusive users' energy rights from the data layer. The accompanying community energy resilience assessment mechanism prioritizes the power supply status of key loads of inclusive users, simulating the system power supply status and energy storage evolution trends under islanded operation scenarios. It promptly pushes optimization suggestions when resilience indicators are insufficient, changing the previous approach of focusing only on overall system indicators and helping to improve the power supply security capabilities for special groups in extreme scenarios.

[0017] At the user modeling level, the solution constructs a three-dimensional community user profile system that incorporates inclusive attributes as a core dimension into the user feature set. The multi-dimensional clustering analysis process selects only feature dimensions directly related to energy consumption behavior and response capabilities, avoiding the introduction of user attributes unrelated to scheduling decisions, thus preventing algorithmic bias. Based on the energy consumption preference pattern vector formed by multi-dimensional feature extraction, it is possible to achieve group management of users with similar energy consumption habits, reducing the solution complexity of subsequent scheduling models. At the same time, it provides data support for user energy consumption preference matching constraints, making load adjustment schemes more in line with users' daily energy consumption habits and reducing unnecessary energy consumption interference.

[0018] At the scheduling model construction level, the proposed multi-objective collaborative scheduling optimization model takes the protection of the rights and interests of inclusive users as an independent optimization objective, and sets rigid load exemption constraints, user energy preference matching constraints, and distribution network safety operation constraints to ensure that the basic living loads of inclusive users are not included in the regulation scope from a rule-based perspective. The voltage deviation penalty, energy storage lifetime loss penalty, and inclusive user load regulation penalty terms introduced in the model can take into account the stability of distribution network operation and the long-term service life of energy storage devices during the solution process, while minimizing the adjustment magnitude of energy consumption for inclusive users.

[0019] The solution process employs a constraint-decoupled parallel computing framework, which is adaptable to the characteristics of dispersed community equipment and a large number of nodes, ensuring the efficiency of the scheduling scheme solution. The rolling optimization scheduling window setting can dynamically adjust the scheme according to the latest load and renewable energy output forecasts, better adapting to the fluctuation characteristics of renewable energy output and improving the local consumption level of renewable energy.

[0020] At the level of rights allocation and interaction, the solution establishes a community energy contribution feedback mechanism with a focus on universal benefit. Based on users' actual contributions in areas such as load regulation, green electricity feedback, and demand response execution, a basic contribution value is calculated. Then, a reasonable contribution value tilt coefficient is set for users receiving universal benefit, ensuring that users with different participation capabilities receive commensurate rights feedback. The two-way interactive iterative process can accept users' self-adjusted parameters and dynamically adjust the tilt coefficient based on the overall community participation, changing the traditional one-way dispatch command issuance model and allowing users to effectively participate in the dispatch decision-making process.

[0021] The accompanying long-term points-based incentive mechanism sets up points-based bonus rules adapted to all users, and the dynamically priced points redemption market can maintain the relative stability of points value, supporting long-term user participation and ensuring that the operational results of the community energy system can cover all residents, meeting the construction requirements of equal access to community public services. The entire algorithm rule design revolves around the requirements of equal access to public services. All scheduling logic, allocation rules, and incentive standards are designed for the common energy needs of all residents, avoiding algorithmic discrimination against specific groups, and providing feasible technical support for inclusive energy access in different types of communities. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating the overall process of the big data-based energy dispatch optimization method proposed in this invention. Figure 2 This is a flowchart of the user energy consumption profile construction and resilience assessment process of the present invention; Figure 3 This is a flowchart illustrating the optimization and solution process for community energy collaborative scheduling in this invention. Figure 4 This is a flowchart illustrating the benefit-sharing allocation and incentive process with a universal benefit bias in this invention. Figure 5 This is a diagram of the energy dispatch optimization system based on big data according to the present invention. Detailed Implementation

[0023] 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.

[0024] Reference Figures 1 to 5 This invention discloses an energy dispatch optimization method based on big data, comprising the following steps: S1. Construct a digital twin model of the community energy infrastructure, collect real-time / historical operation data of distributed photovoltaic power generation units, distributed wind power generation units, energy storage devices, electric vehicle charging piles, and residents' flexible loads within the community, and collect data on the inclusive security attributes of residents. Map all data to a digital twin under a unified spatiotemporal coordinate system for multi-source heterogeneous data fusion, automatically mark rigid load exemption tags for inclusive security users, and generate a panoramic view of the community energy system operation status. The inclusive attributes and rigid load range of each user are marked in the figure simultaneously. S2. Based on the panoramic view of the community energy system, multi-dimensional cluster analysis is performed on the historical energy consumption behavior data, historical demand response execution data, and inclusive attribute data of each resident user in the community. The energy consumption preference pattern vector, demand response potential level, historical willingness fulfillment rate, and inclusive attribute label of resident users are extracted to establish a community user profile database. The user profile of each resident user includes five dimensions of attributes: energy consumption time period distribution characteristics, adjustable load capacity, historical participation frequency in demand response, green electricity consumption willingness intensity, and rigid load exemption mark. S3. Construct a community energy collaborative dispatch optimization model. With the joint optimization objectives of minimizing the overall energy cost of the community, maximizing the local consumption rate of renewable energy, and maximizing the protection of the rights and interests of inclusive users, and under the premise of satisfying the constraints of safe operation of the community distribution network, rigid load exemption constraints, and user energy preference matching constraints, solve the optimal output plan of each distributed energy unit and the optimal adjustment scheme of each residential user's adjustable load in each dispatch period, and generate a community energy collaborative dispatch strategy. The rigid load exemption constraint means that the basic living load of inclusive users shall not be included in the adjustable load range, and the user energy preference matching constraint means that the time deviation of the adjustment scheme shall not exceed the preset threshold of the user's historical energy consumption period. S4. Establish a community energy benefit-sharing and distribution mechanism with inclusive incentives. First, calculate the basic contribution value based on the actual contribution of each resident user to the supply and demand balance of the community energy system. Then, multiply the basic contribution value marked as inclusive guarantee user by the inclusive incentive coefficient to obtain the final contribution value. The actual contribution value comprehensively considers the adjustable load capacity provided by resident users during the dispatch cycle, the renewable energy power fed back to the community grid, the actual execution volume of demand response, and the value weight of the execution period. Calculate the share of economic benefits that each member should be allocated according to the proportion of the final contribution value. S5. Push the community energy collaborative scheduling strategy and benefit-sharing distribution results to each resident's mobile terminal through the community energy management cloud platform. Receive feedback from each resident on their willingness to adjust scheduling parameters. Iterate and update the community energy collaborative scheduling strategy and benefit distribution plan. If the overall community participation rate is lower than the preset threshold during the iteration process, automatically increase the benefit tilt coefficient for inclusive users. When the participation rate of residents in the community exceeds the preset target value, lock the final scheduling strategy and issue an execution command.

[0025] This invention also discloses a community energy resilience assessment mechanism introduced after constructing a panoramic situation map of the community energy system. This mechanism quantitatively assesses the independent operation capability of the community energy system under extreme weather events or external power grid failure scenarios. The assessment process prioritizes simulating the power supply guarantee situation of critical loads of inclusive users, including simulating the maximum sustainable power supply duration of the community energy system in islanded operation mode, the critical load guarantee rate of inclusive users, and the evolution trend of the state of charge of the energy storage system. When the resilience assessment index is lower than the preset safety threshold, the community energy system expansion warning is automatically triggered and a resilience improvement suggestion is pushed to the community energy management cloud platform. The resilience improvement suggestion includes at least one of the following measures: suggesting the addition of new distributed power source types and installed capacity, suggesting the addition of new energy storage capacity, and suggesting the optimization of the community distribution network topology.

[0026] This invention also discloses the establishment of a community energy points incentive mechanism, which issues community energy points to residential users who actively participate in demand response with an execution rate not lower than a preset threshold and who actively consume green electricity. The accumulated amount of community energy points is proportional to the amount of carbon emissions reduced by residential users during the dispatch cycle. The accumulated points of universally benefited users are multiplied by an additional universal benefit bonus coefficient of 1.2 to 1.5 times. Residential users can use community energy points to redeem electricity bill discounts, priority access to charging piles, or the right to lease community public energy storage capacity in the community energy trading market. The community energy points redemption market adopts a dynamic pricing mechanism, with the following specific rules: a benchmark redemption value V0 is set, and the total supply S and total demand D of points in the community are statistically analyzed in real time to calculate the supply-demand ratio k = S / D; when k > 1, the redemption value V = V0 / k; when k < 1, the redemption value V = V0 × (1 / k); when k = 1, V = V0; the fluctuation range of the redemption value does not exceed ±50% of the benchmark value, i.e., V ∈ [0.5V0, ... [1.5V0], maintaining the incentive effectiveness of the points market through supply and demand linkage.

[0027] In this invention, the energy preference pattern vector of residential users in step S2 is also disclosed. This vector is extracted by analyzing the distribution pattern of hourly electricity consumption data of residential users in different seasons, different workday types, and different weather conditions. Specifically, it includes dividing the hourly electricity consumption data of residential users for a year into twenty-four analysis dimensions according to workdays and rest days, the four seasons of spring, summer, autumn and winter, and three typical weather conditions: sunny, cloudy and rainy days. Under each analysis dimension, the average electricity consumption and standard deviation of electricity consumption of residential users in each time period are statistically calculated. After normalizing the average electricity consumption and standard deviation of electricity consumption, the 0 / 1 identifier of the inclusive attribute label is concatenated as a supplementary dimension to generate the energy preference pattern vector. The demand response potential level is divided into three levels: high potential users, medium potential users and low potential users based on the ratio of the adjustable load capacity of residential users to their peak load. The potential level data and the energy preference pattern vector together constitute the complete user profile of the residential user.

[0028] This invention also discloses that the energy consumption preference pattern vector of resident user i in step S2 is constructed by the following formula: in Let be the average power consumption of residential user i in the kth analysis period, in kilowatts; This represents the maximum average power consumption of residential user i across all analysis periods, expressed in kilowatts. Let be the standard deviation of the power consumption of residential user i in the k-th analysis period, in kilowatts; Let M be the maximum standard deviation of power consumption of residential user i across all analysis periods, in kilowatts; M is the total number of analysis periods, taking a positive integer value. This is a universal access attribute identifier for residential users, with a value of 1 indicating a universal access user and a value of 0 indicating a regular user; energy preference pattern vector. This is used to group residential users according to energy consumption similarity in step S3, and apply a uniform type of load adjustment signal to residential users in the same group to improve the solution efficiency of the coordinated scheduling strategy.

[0029] In this invention, the community energy collaborative dispatch optimization model in step S3 also discloses a penalty term for voltage deviation at community distribution network nodes, a penalty term for charge-discharge cycle life loss of energy storage devices, and a penalty term for load adjustment of inclusive users. This transforms the voltage amplitude constraints of each node in the community distribution network into soft constraints, and penalizes the degree of voltage exceedance in the optimization objective function using a weighted sum of squares. The charge-discharge cycle life loss of energy storage devices is calculated by counting the equivalent charge-discharge cycles within the dispatch cycle using the rainflow counting method. The equivalent charge-discharge cycle count is multiplied by the equivalent capacity decay cost per cycle and then included in the energy storage operation cost term in the optimization objective function. The load adjustment penalty term for inclusive users sets a penalty weight three times higher than that for ordinary users for the adjustable load adjustment amount of inclusive users, so that the optimized dispatch result can reduce energy costs while taking into account the voltage quality of the distribution network, the long-term service life of energy storage devices, and the energy rights of inclusive users.

[0030] In this invention, the actual contribution of residential users to the supply and demand balance of the community energy system in step S4 is calculated using a multi-dimensional weighted contribution evaluation model. First, the net load adjustment amount per hour for each residential user during the scheduling cycle is calculated. The net load adjustment amount is the difference between the actual power consumption and the benchmark power consumption of the residential user. Second, the cumulative renewable energy power fed back to the community power grid by each residential user during the scheduling cycle is calculated and converted into an equivalent economic contribution value by weighting according to the time-of-use electricity price. Then, the net load adjustment amount and the equivalent economic contribution value are substituted into a preset contribution scoring function to obtain the load adjustment contribution score and the green energy contribution score. The two contribution scores are weighted and summed to obtain the basic contribution evaluation value of the residential user. The basic contribution evaluation value is multiplied by the corresponding inclusive tilt coefficient to obtain the final contribution evaluation value, which directly determines the allocation weight coefficient of the residential user in the benefit sharing distribution.

[0031] This invention also discloses that the final contribution evaluation value of residential users to the supply and demand balance of the community energy system in step S4 is calculated by the following formula: ; in The final contribution evaluation value for resident user j, with a value ranging from zero to one; The inclusive benefit tilt coefficient is set at 1.2 to 1.8 for inclusive benefit users and 1 for ordinary users; Contribute dimensional weighting factors to load regulation; Contribute dimensional weighting factors to green energy, and satisfy the following conditions: ; This represents the absolute value of the cumulative net load adjustment of residential user j within the scheduling cycle, expressed in kilowatt-hours. It represents the sum of the absolute values ​​of the cumulative net load adjustment of all resident users in the community during the scheduling cycle, expressed in kilowatt-hours. The renewable energy power fed back to the community power grid by residential user j during the dispatch cycle is expressed in kilowatt-hours. The total amount of renewable energy electricity fed back to the community power grid by all residents within the community during the dispatch cycle is expressed in kilowatt-hours; the benefit-sharing allocation amount for resident user j is equal to the total amount of the community benefit pool multiplied by the contribution evaluation value. The normalized percentage.

[0032] In this invention, the community energy collaborative scheduling optimization model in step S3 is also disclosed to be solved using a mixed integer nonlinear programming method. During the solution process, a constraint decoupling strategy based on Lagrange relaxation is introduced to decompose the community distribution network safety operation constraints and rigid load exemption constraints into independent sub-problems for each node. After each sub-problem is solved in parallel, the coupling relationship between each node is updated and coordinated through Lagrange multipliers. After obtaining the global optimal solution after a preset number of convergence iterations, the optimization result is transformed into the output plan instructions of each distributed energy unit in each scheduling period and the adjustment signals of the adjustable load of each residential user. The optimization model uses a 24-hour rolling optimization window and a 15-minute scheduling period resolution. At the beginning of each scheduling cycle, the optimization model parameters are refreshed and the solution is recalculated based on the latest load forecast and renewable energy output forecast.

[0033] This invention also discloses an energy dispatch optimization system based on big data, comprising: The community energy digital twin module is equipped with multi-source data access interfaces, inclusive user identification units, and spatiotemporal alignment and fusion units. It is used to collect operation data of various energy equipment in the community and user inclusive attribute data to build a panoramic situation map of the community energy system with an inclusive access adaptation layer. The user profile building module is equipped with user behavior clustering units, demand response potential assessment units, and inclusive attribute labeling units, which are used to extract multi-dimensional energy consumption characteristics of users and generate a three-dimensional user profile containing energy consumption preferences, response potential, and inclusive attributes. The collaborative scheduling optimization module is equipped with a multi-objective optimization solution unit, a rigid load exemption constraint unit, and a user preference matching unit, which are used to generate community energy collaborative scheduling strategies that meet the dual objectives of efficiency and fairness. The profit distribution module is equipped with a contribution calculation unit, a universal benefit adjustment unit, and a revenue calculation unit, which are used to distribute revenue to users according to the contribution ratio with universal benefit bonus. The interactive execution module is configured with a user mobile terminal interaction unit, a participation dynamic adjustment unit, and an instruction issuance unit, which are used to receive user feedback, iteratively optimize the scheduling scheme, and issue instructions for execution.

[0034] The technical solution of the present invention will be further described in detail below with reference to specific application scenarios. The following embodiments are only used to explain the present invention and are not intended to limit the scope of protection of the present invention.

[0035] The first embodiment selects a resettlement community in the core area of ​​a provincial capital city in northern China as the application object. The community has more than 1,200 households, of which about 12% are inclusive security users recognized by the civil affairs department. The community has built rooftop distributed photovoltaic power generation units, small distributed wind power generation units in public areas, centralized energy storage devices, and public and private electric vehicle charging piles. Residents' homes are generally equipped with electric heating, electric water heaters and other callable electrical equipment.

[0036] After system startup, basic modeling is first completed through the community energy digital twin module. This module is configured with standardized multi-source data access interfaces to achieve unified access to operational data from various energy devices and user attribute data. Real-time operational data from various energy devices is collected every minute, and historical operational data is stored for at least three years, covering equipment operational status under different seasons and weather conditions. Simultaneously, through compliant integration with community civil affairs service portals, inclusive security attribute data for residents is collected, specifically including low-income user identification, disabled user identification, and infrastructure access subsidy application status. All data collection processes strictly adhere to relevant laws and regulations on personal information protection, obtaining explicit informed consent from residents in advance, and avoiding the collection of sensitive personal information unrelated to energy dispatch.

[0037] The module's built-in spatiotemporal alignment and fusion unit maps all collected data to a unified latitude and longitude grid coordinate system for the community, and uniformly calibrates the time dimension to Beijing time millisecond-level timestamps. After a data cleaning process involving outlier removal and missing value completion, multi-source heterogeneous data fusion is completed. The module's built-in inclusive user identification unit automatically matches the list of inclusive security users, marks the basic living loads of these users with rigid load exemption tags, and finally generates a panoramic situation map of the community energy system with an inclusive access adaptation layer. The inclusive access adaptation layer independently sets the load display priority, placing the security-related loads of inclusive users at the highest display level of situation awareness.

[0038] After completing the panoramic situational awareness map, the system first introduces a community energy resilience assessment mechanism to quantitatively evaluate the independent operation capability of the community energy system under scenarios where external power grid failures are caused by extreme blizzards and cold waves. The assessment process prioritizes simulating the power supply guarantee situation for critical loads of inclusive users, specifically simulating the maximum sustainable power supply duration of the community energy system in islanded operation mode, the critical load guarantee rate of inclusive users, and the state of charge evolution trend of the energy storage system. When the resilience assessment index is lower than the preset safety threshold, the system automatically triggers a community energy system capacity expansion warning and pushes resilience improvement suggestions to the community energy management cloud platform. The suggestions cover one or more of the following three categories of measures: suggesting the addition of new distributed power sources and installed capacity, suggesting the addition of new energy storage capacity, and suggesting the optimization of the community distribution network topology.

[0039] After completing the resilience assessment, the 3D user profile construction module initiates user feature extraction. The user behavior clustering unit, demand response potential assessment unit, and inclusive attribute labeling unit configured in the module work together to conduct multi-dimensional clustering analysis on the historical energy consumption behavior data, historical demand response execution data, and inclusive attribute data of each resident user in the community. The clustering process adopts unsupervised clustering algorithms commonly used in this field, and only selects dimensions related to energy consumption behavior, response capability, and inclusive guarantee for analysis. User attributes such as gender, age, and ethnicity that are irrelevant to scheduling decisions are not introduced to avoid algorithmic bias.

[0040] During the clustering process, energy consumption preference pattern vectors, demand response potential levels, historical willingness fulfillment rates, and inclusive attribute tags of residential users are extracted to establish a three-dimensional community user profile database. Each residential user's profile covers five dimensions of attributes: energy consumption time period distribution characteristics, adjustable load capacity, historical frequency of participating in demand response, green electricity consumption willingness intensity, and rigid load exemption mark.

[0041] The extraction of the energy preference pattern vector covers twenty-four analytical dimensions. Specifically, the hourly electricity consumption data of residential users over a continuous year is cross-divided into two time attributes: weekdays and rest days; four seasonal attributes: spring, summer, autumn, and winter; and three typical weather attributes: sunny, cloudy, and rainy days. Under each analytical dimension, the average power consumption and power fluctuation level of residential users in each time period are statistically analyzed. After normalizing the two types of statistical values, a binary identifier of the inclusive attribute is concatenated as a supplementary dimension to form a complete energy preference pattern vector.

[0042] Demand response potential is divided into three levels based on the ratio of a residential user's adjustable load capacity to its peak load: users with a ratio of more than 30% are considered high-potential users, users with a ratio of 15% to 30% are considered medium-potential users, and users with a ratio of less than 15% are considered low-potential users. The potential level data and the energy consumption preference pattern vector together constitute a complete profile of residential users.

[0043] The generated energy preference pattern vector can group residential users according to the similarity of their energy consumption, and apply a uniform type of load adjustment signal to residential users in the same group, which can effectively improve the solution efficiency of subsequent coordinated scheduling strategies.

[0044] Subsequently, the collaborative scheduling optimization module initiates the construction of the community energy collaborative scheduling optimization model. The model has a built-in multi-objective optimization solution unit, which takes minimizing the overall energy cost of the community, maximizing the local consumption rate of renewable energy, and maximizing the protection of the rights and interests of inclusive users as joint optimization objectives. It combines the boundary conditions set by the rigid load exemption constraint unit and the user preference matching unit to carry out the solution. The constraints include the safe operation constraints of the community distribution network, the rigid load exemption constraints, and the user energy consumption preference matching constraints.

[0045] Among them, the rigid load exemption constraint clearly states that the basic living load of universally guaranteed users shall not be included in the adjustable load range, and the user energy preference matching constraint clearly states that the deviation between the time period of the load adjustment scheme and the user's historical energy consumption time period shall not exceed a preset threshold of 30 minutes.

[0046] During the model construction process, three types of penalty terms are introduced simultaneously: voltage deviation penalty term for community distribution network nodes, charge-discharge cycle life loss penalty term for energy storage devices, and load regulation penalty term for inclusive users. The voltage amplitude constraints of each node in the community distribution network are transformed into soft constraints, and the degree of voltage exceeding the limit is penalized in the optimization objective function by weighted sum of squares.

[0047] The charge-discharge cycle life loss of energy storage devices is calculated by counting the equivalent number of charge-discharge cycles within the scheduling cycle using the rainflow counting method. The equivalent number of charge-discharge cycles is then multiplied by the equivalent capacity decay cost per cycle and included in the energy storage operation cost term in the optimization objective function.

[0048] The load adjustment penalty term for inclusive users sets a penalty weight three times higher than that for ordinary users for the adjustable load adjustment amount of inclusive users, in order to minimize interference with the normal energy use of inclusive users. The model is solved using a mixed-integer nonlinear programming method. In the solution process, a constraint decoupling strategy based on Lagrange relaxation is introduced. The constraints of safe operation of community distribution network and rigid load exemption constraints are decomposed into independent subproblems for each node. After each subproblem is solved in parallel, the coupling relationship between nodes is coordinated by updating the Lagrange multipliers. The global optimal solution is obtained after a preset number of convergence iterations.

[0049] The model uses a 24-hour rolling optimization window and a 15-minute scheduling time resolution. At the beginning of each scheduling cycle, the optimization model parameters are refreshed and resolved based on the latest load forecast and renewable energy output forecast. Finally, the optimal output plan of each distributed energy unit and the optimal adjustment scheme of each residential user's adjustable load in each scheduling time period are obtained, forming a community energy collaborative scheduling strategy.

[0050] After the dispatch strategy is generated, the inclusive benefit distribution module starts the calculation of the community energy benefit sharing distribution mechanism with inclusive bias. The contribution calculation unit, inclusive bias adjustment unit and benefit accounting unit configured in the module work together. First, the actual contribution of each residential user to the supply and demand balance of the community energy system is calculated through a multi-dimensional weighted contribution evaluation model. The calculation process first counts the net load adjustment of each residential user per hour during the dispatch cycle, that is, the difference between the actual power consumption of the residential user and the benchmark power consumption. Then, the cumulative amount of renewable energy electricity fed back to the community grid by each residential user during the dispatch cycle is counted and converted into equivalent economic contribution value according to the time-of-use electricity price. The net load adjustment and equivalent economic contribution values ​​are substituted into a preset contribution scoring function to obtain load adjustment contribution scores and green energy contribution scores. The two scores are then weighted and summed to obtain the basic contribution evaluation value for residential users. The scoring process also considers the actual execution volume and execution time value weight of residential users' participation in demand response, with higher value weights assigned to response behavior during peak electricity consumption periods. The inclusive preferential adjustment unit multiplies the basic contribution evaluation value marked as inclusive guarantee users by a preferential preferential coefficient of 1.5 to obtain the final contribution evaluation value. The coefficient is set within the range of 1.2 to 1.8, while the preferential coefficient for ordinary users is set to 1. The sum of the two weighting factors for load adjustment and green energy contribution is 1. The revenue calculation unit calculates the economic benefit share to be allocated to each member based on the normalized proportion of each user's final contribution evaluation value. The allocated amount is the total amount of the community benefit pool multiplied by the corresponding proportion value.

[0051] After the allocation plan calculation is completed, the interactive iterative execution module carries out plan interaction and execution through the built-in user mobile terminal interaction unit, participation dynamic adjustment unit and instruction issuance unit. The generated community energy collaborative scheduling strategy and benefit sharing allocation results are pushed to each resident user's mobile terminal through the community energy management cloud platform. The module receives feedback from each resident user on the autonomous scheduling willingness adjustment parameters and iteratively updates the community energy collaborative scheduling strategy and benefit allocation plan.

[0052] During the iteration process, the participation dynamic adjustment unit continuously monitors the overall community participation in real time. If the participation rate falls below the preset threshold of 60%, the benefit tilt coefficient for inclusive users is automatically increased by 0.1 each time. When the participation rate of resident users in the community exceeds the preset target value of 85%, the final scheduling strategy is locked, and execution instructions are issued to each energy device and user terminal through the instruction issuance unit. The system also establishes a community user participation incentive mechanism, issuing community energy points to resident users who actively participate in demand response and green electricity consumption. The accumulated amount of community energy points is proportional to the amount of carbon emissions reduced by resident users during the scheduling cycle. The accumulated points of inclusive users are additionally multiplied by an inclusive bonus coefficient of 1.3, with the bonus coefficient falling within the set range of 1.2 to 1.5.

[0053] Residential users can use community energy points to redeem electricity bill credits, priority access to charging stations, or lease rights to community public energy storage capacity in the community energy trading market. The community energy points redemption market adopts a dynamic pricing mechanism, adjusting the redemption value of points in real time according to the supply and demand ratio of the points market to avoid large fluctuations in the value of points and maintain the incentive effectiveness of the points market.

[0054] This embodiment addresses the characteristics of older resettlement communities in northern China, which have a high proportion of inclusive energy users, large winter heating loads, and high risks of external power grid failures. It deeply integrates the logic of inclusive energy security into the entire process of digital twin modeling, user profile construction, collaborative scheduling solution, benefit distribution calculation, and interactive iteration. Through multi-objective optimization, it achieves a reduction in community energy costs and an increase in renewable energy consumption rate. Furthermore, through mechanisms such as rigid load exemption, high penalty weighting, and preferential benefit distribution, it fully protects the basic energy rights of inclusive users such as those on minimum living allowances and people with disabilities. At the same time, it is equipped with resilience assessment and long-term credit mechanism, which effectively improves the community's energy system's ability to withstand extreme scenarios and residents' long-term participation, achieving a balance between energy dispatch efficiency and the fairness of public services.

[0055] The second embodiment selects a new mixed-use residential and commercial community in a coastal city in southern China as the application object. The community has more than 800 households and more than 70 supporting commercial stores. The proportion of users covered by the inclusive insurance is about 5%. The community has photovoltaic carports, building-integrated distributed photovoltaics, small distributed wind power generation units in buildings along the street, centralized and user-side combined energy storage systems, and high coverage electric vehicle charging piles. Residential households and commercial stores are generally equipped with adjustable load equipment such as air conditioners and refrigerators. The cooling load accounts for a high proportion in summer and is also greatly affected by extreme weather such as typhoons and rainstorms.

[0056] After system deployment, basic modeling is first conducted through the community energy digital twin module. Multi-source data access interfaces simultaneously cover various energy devices on both the residential and commercial sides. Real-time equipment operation data is collected every 30 seconds, and historical operation data is stored for at least three years, covering operation records under typical extreme scenarios such as typhoons, high temperatures, and heavy rainfall. The module connects to the compliant port of the community service center to collect data on the inclusive security attributes of resident users. All data collection is subject to explicit user authorization, and data permissions are strictly managed in accordance with personal information protection requirements, ensuring that user information is not collected beyond the permitted scope. The spatiotemporal alignment and fusion unit maps all data to a community latitude and longitude grid coordinate system with an accuracy of 1 meter. The time dimension uses millisecond-level timestamps calibrated by the Network Time Protocol. After data cleaning, multi-source heterogeneous data fusion is completed.

[0057] The inclusive user identification unit automatically identifies inclusive users and marks users' basic living loads, charging loads of assistive devices for disabled users, and electricity consumption of community barrier-free facilities as rigid load exemption tags. It generates a panoramic view of the community energy system with an inclusive access adaptation layer. The inclusive adaptation layer highlights the load locations of inclusive users in low-lying and high-rise areas to ensure rapid location in extreme scenarios.

[0058] After the panoramic situation map is generated, the system first activates the community energy resilience assessment mechanism. For typical risk scenarios such as typhoons, flooding, and external power grid outages in coastal communities, it quantitatively assesses the independent operation capability of the community energy system. The assessment process prioritizes simulating the power supply guarantee of critical loads for inclusive users, focusing on simulating the power consumption of drainage equipment for inclusive users in low-lying residential areas and the power supply guarantee level for elevators and basic living needs for inclusive users in high-rise buildings. Specifically, it calculates the maximum sustainable power supply duration under the island operation mode, the critical load guarantee rate of inclusive users, and the evolution trend of the charge state of the energy storage system.

[0059] When the resilience assessment index is lower than the preset safety threshold, an expansion warning is automatically triggered, and a resilience improvement plan is pushed out. This plan includes measures adapted to coastal scenarios, such as adding wind-resistant distributed power sources, increasing the installation height of energy storage equipment, adding energy storage capacity, and optimizing the structure of distribution network interconnection switches.

[0060] After completing the resilience assessment, the 3D user profile construction module extracts user features. The user behavior clustering unit covers both residential and commercial users. Multi-dimensional clustering analysis is carried out on the historical energy consumption behavior data, historical demand response execution data, and inclusive attribute data of the two types of users. The clustering process strictly follows the principle of fairness, selecting only the feature dimensions related to energy consumption adjustment, contribution accounting, and rights protection, and does not introduce any irrelevant attributes that may lead to discrimination.

[0061] During the clustering process, users' energy consumption preference pattern vectors, demand response potential levels, historical willingness fulfillment rates, and inclusive attribute tags are extracted to establish a three-dimensional community user profile database covering both residential and commercial users. Each user's profile includes five dimensions of attributes: energy consumption time period distribution characteristics, adjustable load capacity, historical frequency of participation in demand response, green electricity consumption willingness intensity, and rigid load exemption mark.

[0062] The extraction of the energy consumption preference pattern vector covers twenty-four analytical dimensions. Targeting the operational characteristics of commercial users, and based on the cross-division of weekdays and weekends, seasons, and three typical weather conditions, it focuses on statistically analyzing power variation characteristics during business hours. Under each analytical dimension, the average power consumption and power consumption fluctuation level for each time period are calculated. After normalization, a binary identifier representing the inclusive attribute is concatenated to form a complete energy consumption preference pattern vector. The demand response potential assessment unit classifies potential levels according to the ratio of adjustable load capacity to peak load: above 30% is high-potential users, 15% to 30% is medium-potential users, and below 15% is low-potential users. Commercial users, due to the higher proportion of adjustable loads such as refrigerators and lighting, have an overall higher potential level than ordinary residential users.

[0063] The generated energy preference pattern vector is used to group users according to energy consumption similarity. Users in the same group are given the same type of adjustment signal, which improves the scheduling solution efficiency.

[0064] Subsequently, the collaborative scheduling optimization module constructs a community energy collaborative scheduling optimization model. The multi-objective optimization solution unit takes minimizing the overall energy cost of the community, maximizing the local consumption rate of renewable energy, and maximizing the protection of the rights and interests of inclusive users as joint optimization objectives. It combines the boundary conditions set by the rigid load exemption constraint unit and the user preference matching unit to carry out the solution, ensuring that the rigid load of inclusive users is not included in the regulation range, the deviation between the load regulation period and the user's historical energy consumption period does not exceed the preset threshold of 30 minutes, and the load rate and node voltage of all lines are within the qualified range.

[0065] The model introduces three types of penalty terms, incorporating node voltage deviation, energy storage cycle life loss, and load adjustment for inclusive users into cost accounting. Voltage deviation is penalized using a weighted sum of squares soft constraint, energy storage life loss is calculated by counting equivalent cycles using rainflow counting and converted into operating costs, and the load adjustment for inclusive users is set with a penalty weight of 3 times that of ordinary users to minimize the impact on energy consumption of inclusive users.

[0066] The model is solved using a mixed-integer nonlinear programming method. The complex coupling constraints are decoupled into independent subproblems for each node through a Lagrange relaxation strategy. Considering the large number of nodes in a mixed commercial and residential community, multi-process parallel computing is used to improve the solution speed. The coupling relationship between nodes is coordinated by iteratively updating the Lagrange multipliers. The global optimal solution is obtained after the convergence condition is met.

[0067] The model adopts a 24-hour rolling optimization window and a 15-minute scheduling time resolution. Considering the rapid changes in coastal weather conditions, the load forecast, renewable energy output forecast, and weather warning data are refreshed every 5 minutes at the beginning of each scheduling cycle. The model parameters are dynamically updated and the solution is recalculated to ultimately generate the optimal output plan for each distributed energy unit and the optimal adjustment scheme for the user's adjustable load, thus forming a collaborative scheduling strategy.

[0068] After the dispatch strategy is generated, the inclusive benefit distribution module conducts benefit sharing accounting, and the contribution calculation unit establishes a multi-dimensional weighted contribution evaluation model. First, it calculates the hourly net load adjustment of users in each dispatch cycle, that is, the difference between the actual power consumption and the benchmark power consumption. Then, it calculates the cumulative amount of renewable energy fed back to the community grid by users and converts it into an equivalent economic contribution value according to the time-of-use price. Combining the actual execution volume of users participating in demand response and the value weight of the execution period, it calculates the load adjustment contribution score and the green energy contribution score respectively. The two scores are weighted and summed to obtain the basic contribution evaluation value. The weight factors of the two dimensions sum to 1.

[0069] The inclusive benefit adjustment unit sets an inclusive benefit tilt coefficient of 1.3 for inclusive benefit guarantee users, with the coefficient ranging from 1.2 to 1.8. The tilt coefficient for ordinary users and commercial users is 1. The final contribution evaluation value is obtained by multiplying the basic contribution by the tilt coefficient. The revenue calculation unit normalizes the final contribution of all users and calculates the economic benefit share that each user should receive according to the proportion. Funds are allocated from the common interest pool formed by community energy conservation and green electricity trading to complete the distribution.

[0070] The interactive iteration execution module then connects to both residential mobile terminals and commercial user energy management terminals. Through the user mobile terminal interaction unit, it pushes the scheduling strategy and estimated revenue to all users, receives user feedback on their willingness to adjust scheduling parameters, and iteratively updates the scheduling strategy and allocation plan. The participation dynamic adjustment unit monitors the overall community participation in real time. If the participation rate falls below a preset threshold of 70%, the inclusive tilt coefficient is increased by 0.1 each time. When the participation rate reaches a preset target value of 90%, the final plan is locked, and execution instructions are issued to each device and user terminal through the instruction issuance unit.

[0071] The system includes an incentive mechanism that awards energy points to users participating in demand response and green electricity consumption. The accumulated points are proportional to the user's carbon reduction, with points for general users multiplied by an additional 1.4 times the base rate, which is set between 1.2 and 1.5. Users can redeem points for electricity bill discounts, priority access to charging stations, and public energy storage capacity leasing rights. Commercial users can also redeem points for priority access to promotional resources in community public areas. The points market employs a dynamic pricing mechanism, adjusting the redemption value in real time based on supply and demand to maintain the incentive effectiveness of the points system.

[0072] This embodiment addresses the characteristics of mixed-use residential and commercial communities in southern coastal areas, such as high load density, a large proportion of commercial users, and diverse extreme weather risks. It constructs an inclusive energy dispatch system adapted to mixed-use scenarios. While achieving high-proportion renewable energy consumption and reducing the overall energy cost of the community, it extends the scope of inclusive protection to the electricity needs of special groups for barrier-free facilities and auxiliary equipment. Through differentiated resilience assessment and points incentive design, it not only stimulates the participation of commercial users and ordinary residents, but also fully protects the energy rights of inclusive users in daily dispatch and extreme scenarios, forming a win-win community energy governance model for all parties.

[0073] Reference Figure 1This flowchart illustrates the overall execution steps of a big data-based energy dispatch optimization method. First, the method constructs a digital twin model of the community energy infrastructure. By collecting multi-source real-time and historical operational data, as well as resident attribute data, it achieves multi-source data fusion and identifies users eligible for universal benefit, while also marking rigid load exemption attributes. Subsequently, the system performs cluster analysis on energy consumption data, extracting users' energy preferences and demand response potential to establish a multi-dimensional user profile database. Based on this, with the joint objectives of minimizing energy costs, maximizing renewable energy absorption rate, and maximizing the protection of universal benefit users' rights, a community energy collaborative dispatch optimization model is constructed to derive the optimal output plan for distributed energy and the adjustable load adjustment scheme for users. Furthermore, a community energy benefit-sharing allocation mechanism with a universal benefit bias is established, calculating the final contribution value of users. Finally, the system pushes the dispatch strategy and allocation results to user terminals, iteratively optimizing the strategy based on user feedback. When participation is below a threshold, the universal benefit bias coefficient is dynamically increased until the participation reaches the target, at which point the strategy is locked and finally executed, forming a complete intelligent energy dispatch closed loop.

[0074] Reference Figure 2 This flowchart details the process of processing residential user energy consumption data, constructing user profiles, and assessing community energy resilience. The system first acquires residential energy consumption data, dividing it into multiple analysis periods based on weekday type, seasonal characteristics, and weather conditions. By statistically analyzing the average power consumption and standard deviation for each period, and combining this with normalization processing and inclusive attribute identification, a precise energy preference pattern vector is generated. Simultaneously, based on the ratio of user adjustable load capacity to peak load, the system categorizes demand response potential into high, medium, and low levels, establishing a comprehensive user profile database based on these characteristics. Furthermore, the process includes constructing a community energy resilience assessment mechanism, specifically targeting extreme weather or external grid failure scenarios to quantitatively assess the independent operating capability of the community energy system. During the assessment, priority is given to simulating the power supply guarantee situation for critical loads of inclusive users, specifically including simulating the maximum sustainable power supply duration under islanded operation mode, the critical load guarantee rate for inclusive users, and the evolution trend of the energy storage system's state of charge.

[0075] Reference Figure 3This flowchart illustrates the construction and solution logic of the community energy collaborative scheduling optimization model. The system sets the minimum community energy cost, the maximum renewable energy absorption rate, and the maximum protection of the rights and interests of inclusive users as joint optimization objectives. The objective function incorporates penalties for distribution network node voltage deviation, energy storage device charge / discharge cycle life loss, and inclusive user load adjustment. Regarding constraints, it comprehensively covers distribution network security constraints, rigid load exemption constraints, and user energy preference constraints. To efficiently solve this mixed-integer nonlinear programming problem, the process introduces a constraint decoupling strategy based on Lagrange relaxation. The global distribution network security constraints and rigid load exemption constraints are decomposed into independent subproblems for each node and solved in parallel. The coupling relationship between nodes is updated and coordinated using Lagrange multipliers to obtain the global optimal solution. The entire solution process uses a fixed number of hours as the rolling window and a fixed number of minutes as the scheduling time resolution. At the beginning of each scheduling cycle, the parameters are refreshed based on the latest load forecast and renewable energy output forecast, and the solution is recalculated.

[0076] Reference Figure 4 This flowchart illustrates the implementation steps of the community energy benefit-sharing allocation mechanism with inclusive incentives and the user participation incentive mechanism. In the benefit-sharing allocation phase, the system calculates each user's net load adjustment and renewable energy feedback during the scheduling cycle using a multi-dimensional weighted contribution evaluation model. These two indicators are converted and substituted into a preset contribution scoring function to obtain load adjustment contribution scores and green energy contribution scores, which are then weighted and summed to obtain the basic contribution evaluation value. For users receiving inclusive benefits, the system automatically multiplies the basic contribution evaluation value by an inclusive incentive coefficient to generate the final contribution evaluation value, and allocates the total amount of the community benefit pool accordingly. Regarding user participation incentives, the system issues community energy points to users participating in demand response and green electricity consumption. The accumulated points are proportional to carbon emission reductions, and users receiving inclusive benefits enjoy an additional inclusive bonus coefficient. Users can use community energy points to redeem electricity bill discounts or priority access to charging stations in the community energy trading market. The point redemption adopts a dynamic pricing mechanism, adjusting the value in real time according to the supply-demand ratio.

[0077] Reference Figure 5This diagram illustrates the system logic architecture used to implement the aforementioned big data-based energy dispatch optimization method. The system consists of five highly integrated core modules. The community energy digital twin module is equipped with multi-source data access interfaces, a universal user identification unit, and a spatiotemporal alignment fusion unit, responsible for comprehensively collecting operational data from various energy equipment and user universal attribute data. The user profile construction module includes a user behavior clustering unit, a demand response potential assessment unit, and a universal attribute tagging unit, used to extract multi-dimensional energy consumption characteristics and generate a three-dimensional user profile. The collaborative dispatch optimization module integrates a multi-objective optimization solution unit, a rigid load exemption constraint unit, and a user preference matching unit, outputting a collaborative dispatch strategy that balances efficiency and fairness. The benefit distribution module includes a contribution calculation unit, a universal benefit tilt adjustment unit, and a revenue calculation unit, responsible for executing revenue distribution calculations with universal benefit bonuses. The interactive iterative execution module is equipped with a user mobile terminal interaction unit, a participation dynamic adjustment unit, and an instruction issuance unit, used to receive feedback, iterate strategies, and issue execution instructions. The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A big data-based energy dispatch optimization method, characterized in that, Includes the following steps: S1. Construct a digital twin model of the community energy infrastructure, collect real-time and historical operation data within the community, collect community energy operation data and resident user attribute data, perform multi-source data fusion, identify inclusive security users and mark rigid load exemption attributes; S2. Perform cluster analysis on residential users' energy consumption data, extract energy consumption preferences, demand response potential and inclusive attributes, and establish a user profile database; S3. Construct a community energy collaborative dispatch optimization model with the joint objectives of minimizing community energy costs, maximizing renewable energy consumption rate, and maximizing the protection of the rights and interests of inclusive users. Under the constraints of distribution network security, rigid load exemption, and user energy preference, solve the optimal output plan of distributed energy and the user adjustable load adjustment scheme. S4. Establish a community energy benefit-sharing and distribution mechanism with inclusive incentives. Calculate the basic contribution value based on the user's contribution to the community's energy supply and demand balance. Multiply the basic contribution value of inclusive users by the inclusive incentive coefficient to obtain the final contribution value. S5. Push the scheduling strategy and allocation results to the user terminal, receive user feedback and iteratively optimize. When user participation is below the threshold, increase the inclusive tilt coefficient. After the participation reaches the target, lock the strategy and execute it.

2. The energy dispatch optimization method based on big data according to claim 1, characterized in that, Also includes: Establish a community energy resilience assessment mechanism to quantitatively assess the community's independent operation capability under extreme weather or external power grid failure scenarios. The assessment process prioritizes simulating the power supply guarantee situation of critical loads of inclusive users, including simulating the maximum sustainable power supply duration of the community energy system in islanded operation mode, the critical load guarantee rate of inclusive users, and the evolution trend of the energy storage system's state of charge.

3. The energy dispatch optimization method based on big data according to claim 1, characterized in that, Also includes: Establish a community user participation incentive mechanism, issue community energy points to users who participate in demand response and green electricity consumption. The accumulated points are proportional to the carbon emission reduction during the dispatch cycle. The accumulated points of universally benefited users are multiplied by an additional universal benefit bonus coefficient of 1.2 to 1.5 times. Users can use community energy points to redeem electricity bill discounts, priority access to charging stations, or lease rights to community public energy storage capacity in the community energy trading market. The points redemption adopts a dynamic pricing mechanism, adjusting the redemption value in real time according to the supply and demand ratio of points.

4. The energy dispatch optimization method based on big data according to claim 1, characterized in that, In step S2, the energy consumption preference pattern vector of residential users is extracted by analyzing the distribution pattern of hourly electricity consumption data of residential users in different seasons, different workday types, and different weather conditions. Specifically, the hourly electricity consumption data of a year is divided into twenty-four analysis dimensions according to workdays and rest days, spring, summer, autumn and winter, and three types of weather (sunny, cloudy and rainy). The average electricity consumption and standard deviation of each time period are statistically analyzed, and after normalization, the inclusive attribute 0 / 1 identifier is concatenated to generate the energy consumption preference pattern vector. The demand response potential level is divided into three levels: high, medium and low, according to the ratio of the user's adjustable load capacity to the peak load.

5. The energy dispatch optimization method based on big data according to claim 4, characterized in that, In step S2, the energy consumption preference pattern vector of resident user i is constructed by the following formula: ； in Let i be the average power consumption of residential user i in the k-th analysis period. Let i be the maximum average power consumption of residential user i across all analysis periods. Let be the standard deviation of the power consumption of residential user i in the k-th analysis period. Let M be the maximum standard deviation of electricity consumption of residential user i across all analysis periods, and M be the total number of analysis periods, taking a positive integer value. This serves as an identifier for the universal benefit attribute of residential users.

6. The energy dispatch optimization method based on big data according to claim 1, characterized in that, In step S3, the community energy collaborative dispatch optimization model simultaneously introduces a voltage deviation penalty term for community distribution network nodes, a charge-discharge cycle life loss penalty term for energy storage devices, and a load adjustment penalty term for inclusive users. This transforms the voltage amplitude constraints of each node in the community distribution network into soft constraints, and penalizes the degree of voltage exceeding the limit in the optimization objective function using a weighted sum of squares. The charge-discharge cycle life loss of the energy storage device is calculated by counting the equivalent charge-discharge cycles within the dispatch cycle using the rainflow counting method. The equivalent charge-discharge cycle count is multiplied by the equivalent capacity decay cost of a single cycle and then included in the energy storage operation cost term in the optimization objective function.

7. The energy dispatch optimization method based on big data according to claim 1, characterized in that, In step S4, the actual contribution of residential users to the supply and demand balance of the community energy system is calculated using a multi-dimensional weighted contribution evaluation model. First, the net load adjustment amount per hour for each residential user during the scheduling cycle is calculated. The net load adjustment amount is the difference between the actual power consumption and the benchmark power consumption of the residential user. Second, the cumulative renewable energy power fed back to the community power grid by each residential user during the scheduling cycle is calculated and converted into an equivalent economic contribution value by weighting according to the time-of-use electricity price. Then, the net load adjustment amount and the equivalent economic contribution value are substituted into the preset contribution scoring function to obtain the load adjustment contribution score and the green energy contribution score. The two contribution scores are weighted and summed to obtain the basic contribution evaluation value of the residential user. The basic contribution evaluation value is multiplied by the corresponding inclusive tilt coefficient to obtain the final contribution evaluation value, which serves as the weighting coefficient for benefit distribution.

8. The energy dispatch optimization method based on big data according to claim 7, characterized in that, In step S4, the final contribution evaluation value of residential users to the supply and demand balance of the community energy system is calculated by the following formula: ;in The final contribution evaluation value for resident user j; This is the universal benefit tilt coefficient; Contribute dimensional weighting factors to load regulation; Contribute dimensional weighting factors to green energy Let J be the absolute value of the cumulative net load adjustment of residential user j during the scheduling cycle; This represents the sum of the absolute values ​​of the cumulative net load adjustment for all residents within the community during the scheduling cycle. This refers to the renewable energy power fed back to the community power grid by residential user j during the dispatch cycle. The total renewable energy electricity fed back to the community grid by all residents within the community during the dispatch cycle is the total amount of renewable energy electricity. The benefit-sharing allocation for resident user j is equal to the total amount of the community benefit pool multiplied by the contribution evaluation value. The normalized percentage.

9. The energy dispatch optimization method based on big data according to claim 6, characterized in that, In step S3, the community energy collaborative scheduling optimization model is solved using a mixed integer nonlinear programming method. A constraint decoupling strategy based on Lagrange relaxation is introduced to decompose the distribution network security constraints and rigid load exemption constraints into independent subproblems of each node and solve them in parallel. The coupling relationship between nodes is updated and coordinated through Lagrange multipliers, and the global optimal solution is obtained after convergence. The optimization model uses a 24-hour rolling window and a 15-minute scheduling time resolution. At the beginning of each scheduling cycle, the parameters are refreshed and the solution is recalculated based on the latest load forecast and renewable energy output forecast.

10. A big data-based energy dispatch optimization system, used to implement the energy dispatch optimization method according to any one of claims 1-9, characterized in that, include: The community energy digital twin module is equipped with a multi-source data access interface, a universal user identification unit, and a spatiotemporal alignment fusion unit, which are used to collect operation data of various energy equipment in the community and user universal attribute data. The user profile building module is equipped with user behavior clustering units, demand response potential assessment units, and inclusive attribute labeling units, which are used to extract multi-dimensional energy consumption characteristics of users and generate a three-dimensional user profile containing energy consumption preferences, response potential, and inclusive attributes. The collaborative scheduling optimization module is equipped with a multi-objective optimization solution unit, a rigid load exemption constraint unit, and a user preference matching unit, which are used to generate community energy collaborative scheduling strategies that meet the dual objectives of efficiency and fairness. The profit distribution module is equipped with a contribution calculation unit, a universal benefit adjustment unit, and a revenue calculation unit, which are used to distribute revenue to users according to the contribution ratio with universal benefit bonus. The interactive execution module is configured with a user mobile terminal interaction unit, a participation dynamic adjustment unit, and an instruction issuance unit, which are used to receive user feedback, iteratively optimize the scheduling scheme, and issue instructions for execution.

Images