Multi-agent collaborative source-load interaction and new energy management method and system

Abstract

This invention proposes a multi-agent collaborative source-load interaction and renewable energy management method and system. The method includes: acquiring and converting data from the electricity sales side, renewable energy aggregation side, user side, and distribution network side to generate a set of collaborative management scenario constraints; generating electricity demand response actions, renewable energy participation actions, and user response actions based on this set of constraints, and combining and filtering them to generate a set of feasible multi-agent strategies; scoring the strategy set using a collaborative utility function that includes a consistency gain term and a deviation correction term, and selecting the optimal strategy as the collaborative decision result; finally, mapping the strategies to specific execution instructions and performing consistency verification. The system includes a scenario constraint construction module, a strategy generation module, a decision solving module, and an execution transformation module. This solution effectively balances the interests of multiple parties through unified scenario expression and multi-agent game theory, achieving collaborative optimization of grid security, renewable energy consumption, and marketing efficiency.

Description

Technical Field

[0001] This invention belongs to the field of resource scheduling, and in particular relates to a method and system for multi-entity collaborative source-load interaction and new energy management. Background Technology

[0002] Traditional power system management architecture exhibits a clear "source-load" separation characteristic, and its technical bottlenecks mainly lie in the vertical independence of the business architecture and the local limitations of the optimization mechanism.

[0003] From the perspective of existing technical architecture, business management on the electricity sales side and operation management on the generation side are usually under different business systems. On the one hand, electricity marketing management mainly focuses on pricing strategy formulation, user load curve analysis, and market expansion. Its decision-making relies heavily on historical electricity data and static electricity price elasticity models, lacking the dynamic perception capability of the real-time operating status of the distribution network and the fluctuation of distributed renewable energy output. On the other hand, distributed renewable energy management focuses on single-point generation forecasting, power control, and grid connection protection, often neglecting the user-side electricity demand elasticity and commercial incentive strategies. This "marketing-operation" dual-track parallel model makes it difficult for user-side response behavior to effectively match the renewable energy output periods when implementing demand response or price incentives. This easily leads to situations where the total response load exceeds the distribution network's carrying capacity, thereby triggering technical contradictions such as grid overload or renewable energy curtailment.

[0004] Regarding existing collaborative control technologies, although some attempts have been made to introduce centralized optimization or master-slave game models to coordinate source-load relationships, these solutions still have significant technical shortcomings in practical applications. First, existing technologies often treat users as passive control objects, employing a "one-size-fits-all" approach to control commands. This fails to fully consider the response willingness boundaries and comfort constraints of user groups under different incentive levels, resulting in low user participation and significantly reduced strategy execution rates. Second, existing optimization models often neglect the commercial interest game among multiple stakeholders, such as renewable energy aggregators, electricity retailers, and grid companies, lacking quantitative modeling of the interest boundaries of each stakeholder. This leads to feasibility risks in the actual implementation of the generated strategies. Furthermore, due to the lack of a hard constraint verification mechanism for the carrying capacity boundaries of regional distribution networks, existing technologies pose a potential risk of local grid security when executing equipment-level commands.

[0005] In summary, existing technical solutions generally suffer from high data barriers, a lack of interest coordination mechanisms, and uncontrollable execution risks, failing to achieve multi-party synergy in ensuring safe grid operation, efficient renewable energy consumption, and economic benefits for market participants. Therefore, there is an urgent need to construct a source-load interaction management method that integrates multi-source data sensing, multi-stakeholder interest balancing, and dynamic boundary verification to address the strategic conflicts and low execution feasibility issues present in existing technologies. Summary of the Invention

[0006] This invention discloses a multi-entity collaborative source-load interaction and new energy management method and system to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the first aspect of the present invention provides a method for multi-entity collaborative source-load interaction and new energy management, the method comprising: Acquire demand response configuration data from the electricity sales side, aggregated operation data from the distributed renewable energy aggregation side, user group response data from the user side, and regional operation data from the distribution network operation side; perform interval-based processing on the demand response configuration data to convert it into a set of control constraint intervals, perform regional-level aggregation on the aggregated operation data to convert it into regional-level distributed renewable energy participation capability, perform boundary mapping on the user group response data to convert it into user group response boundaries, perform boundary mapping on the regional operation data to convert it into regional carrying capacity boundaries, and aggregate the set of control constraint intervals, regional-level distributed renewable energy participation capability, user group response boundaries, and regional carrying capacity boundaries to generate a set of constraints for collaborative management scenarios; Using the set of constraints in the collaborative management scenario as input, demand response actions are generated by combining user group response boundaries, distributed new energy participation actions are generated by combining regional carrying capacity boundaries, and user group response actions are generated by combining demand response actions. Demand response actions, distributed new energy participation actions, and user group response actions that satisfy regional coupling constraints are combined and filtered to generate a set of feasible strategies for multiple subjects. For each candidate item in the multi-agent feasible strategy set, calculate the collaborative utility value, which includes the matching degree between demand response actions and user group response boundaries, and the matching degree between distributed new energy participation actions and user group response actions; select the candidate item with the largest collaborative utility value as the collaborative decision result. The demand response actions in the collaborative decision-making results are mapped to actual execution prices and incentive parameters, distributed renewable energy participation actions are allocated as equipment-level scheduling instructions, and user group response actions are mapped to response instructions of the target user set. The mapped execution content is then checked against the regional carrying capacity boundary, and the checked equipment-level scheduling instructions and response instructions are output to the corresponding distributed renewable energy power generation equipment and user load control terminals to adjust the active power output of the distributed renewable energy power generation equipment and the power consumption of the user load control terminals.

[0008] Furthermore, the step of converting the demand response configuration data into a set of control constraint intervals includes: Extract the set of price tiers, incentive tiers, and green electricity consumption guidance parameters corresponding to the current decision-making period; Based on the user groups that can be reached during the current decision-making period, the set of price tiers, the set of incentive tiers, and the set of green electricity consumption guidance parameters are compressed into corresponding interval expressions to form the set of regulatory constraint intervals.

[0009] Furthermore, the process of performing regional-level aggregation on the aggregated runtime data includes: Read real-time output data and short-term forecast results from the same region in the distributed new energy aggregation platform; overlay the available capacity currently connected to the collaborative management scope to form a regional distributed new energy participation capability that represents the total participation capability that the region can form in the current time period.

[0010] Furthermore, the action of generating distributed new energy sources includes: Read the regional-level distributed new energy participation capacity by region and determine the carrying capacity status of the regional carrying capacity boundary; When the regional carrying capacity boundary is in a high carrying capacity state, the regional distributed new energy participation capability is divided into distributed new energy participation actions including the basic participation layer, the enhanced participation layer and the high consumption participation layer. When the regional carrying capacity boundary is under tight constraints, the regional-level distributed new energy participation capability is divided into distributed new energy participation actions that include a basic participation layer and a conservative participation layer.

[0011] Furthermore, the generation of a multi-agent feasible strategy set specifically includes: The demand response action, the distributed new energy participation action, and the user group response action are combined to calculate the combined regional coupling constraint value. The regional coupling constraint value includes the superposition value of demand response action and user group response action, as well as the deviation occupancy value of distributed new energy participation action and user group response action. When the regional coupling constraint value is less than the regional carrying capacity boundary, the combination is retained in the multi-agent feasible strategy set.

[0012] Furthermore, the calculation of the synergistic utility value specifically includes: Calculate the positive collaborative product term between the demand response action and the user group response boundary; Calculate the positive collaborative product term between distributed new energy participation actions and user group response actions; Calculate the deviation penalty term between distributed new energy participation actions and user group response actions, as well as the deviation penalty term between demand response actions and user group response boundaries; The synergistic utility value is obtained by weighted summation of the four calculation results.

[0013] Furthermore, the selection of the collaborative decision-making result specifically includes: The collaborative utility value of each candidate item in the multi-agent feasible strategy set is compared; the candidate item with the largest collaborative utility value that simultaneously satisfies the demand response action constraint, the degree of local consumption of new energy, and the regional carrying capacity boundary is selected as the collaborative decision result.

[0014] Furthermore, the allocation of the participation actions of distributed new energy sources specifically includes: Read the mapping table between regions and devices in the aggregation platform, calculate the current available generation capacity ratio of each photovoltaic unit and energy storage unit in the region; allocate the distributed new energy participation actions as at least one device-level scheduling instruction according to the capacity ratio, and maintain the participation ratio relationship determined in the collaborative decision result.

[0015] Furthermore, the process of verifying the region carrying boundaries of the mapped execution content specifically includes: Calculate the sum of the total amount of new energy participation and the total amount of user group response within the region after the mapping is executed, as well as the mismatch deviation between the two; when the sum of the sum and the deviation exceeds the regional carrying capacity boundary, the execution content is synchronously scaled according to the ratio between the total amount of new energy participation and the total amount of user group response until the regional carrying capacity boundary is met.

[0016] A second aspect of the present invention provides a multi-entity collaborative source-load interaction and new energy management system, the system comprising: The scenario constraint construction module is used to acquire demand response configuration data from the electricity sales side, aggregated operation data from the distributed renewable energy aggregation side, user group response data from the user side, and regional operation data from the distribution network operation side. It performs interval-based processing on the demand response configuration data to convert it into a set of control constraint intervals, performs regional-level aggregation on the aggregated operation data to convert it into regional-level distributed renewable energy participation capabilities, performs boundary mapping on the user group response data to convert it into user group response boundaries, and performs boundary mapping on the regional operation data to convert it into regional carrying capacity boundaries. Finally, it aggregates the set of control constraint intervals, regional-level distributed renewable energy participation capabilities, user group response boundaries, and regional carrying capacity boundaries to generate a collaborative management scenario constraint set. The strategy generation module is used to take the set of constraints of the collaborative management scenario as input, generate demand response actions by combining user group response boundaries, generate distributed new energy participation actions by combining regional carrying capacity boundaries, and generate user group response actions by combining demand response actions; and combine and filter demand response actions, distributed new energy participation actions, and user group response actions that satisfy regional coupling constraints to generate a set of feasible multi-subject strategies. The decision-solving module is used to calculate the collaborative utility value for each candidate item in the set of feasible strategies for multiple subjects, including the matching degree between demand response actions and user group response boundaries, and the matching degree between distributed new energy participation actions and user group response actions; and select the candidate item with the largest collaborative utility value as the collaborative decision result. The execution conversion module is used to map the demand response actions in the collaborative decision-making results to actual execution prices and incentive parameters, allocate distributed renewable energy participation actions to equipment-level scheduling instructions, and map user group response actions to response instructions for the target user set; it performs regional carrying capacity boundary verification on the mapped execution content, and outputs the verified equipment-level scheduling instructions and response instructions to the corresponding distributed renewable energy power generation equipment and user load control terminals to adjust the active power output of the distributed renewable energy power generation equipment and the power consumption of the user load control terminals.

[0017] The beneficial technical effects of the present invention are at least as follows: To address the aforementioned issues, this invention provides a multi-entity collaborative source-load interaction and new energy management method and system. By constructing a multi-entity collaborative management framework, heterogeneous data from multiple sources, such as electricity sales configuration, distributed new energy aggregation operation, user-side group response, and distribution network regional operation, are uniformly mapped into control constraint ranges, new energy participation capabilities, user response boundaries, and regional carrying capacity boundaries. This breaks down data barriers between electricity marketing, new energy management, and distribution network operation, achieving a unified scenario expression across business scenarios. Furthermore, the solution utilizes regional coupling constraints to jointly screen demand response, new energy, and user response actions. Based on a collaborative utility function that includes consistency gain and deviation correction, it performs multi-entity game theory solving, effectively overcoming the strategy conflicts and local optima problems caused by traditional unilateral independent decision-making. This achieves dynamic equilibrium and global collaborative optimization of multi-party interests. Simultaneously, by implementing consistency verification and synchronous scaling mechanisms, the solution ensures that the final issued price parameters, generation dispatch, and load adjustment instructions are strictly within the regional carrying capacity boundaries. This significantly enhances the safety and stability of distribution network operation while improving new energy absorption rate and power allocation efficiency. Attached Figure Description

[0018] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0019] Figure 1 This is a flowchart of the multi-entity collaborative source-load interaction and new energy management method of the present invention.

[0020] Figure 2 This is a framework diagram of the multi-entity collaborative source-load interaction and new energy management system of the present invention. Detailed Implementation

[0021] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0022] In one or more embodiments, such as Figure 1 As shown, a multi-entity collaborative source-load interaction and new energy management method is disclosed, which includes the following: S1: Obtain demand response configuration data from the electricity sales side, aggregated operation data from the distributed renewable energy aggregation side, user group response data from the user side, and regional operation data from the distribution network operation side; perform interval-based processing on the demand response configuration data to convert it into a set of control constraint intervals; perform regional-level aggregation on the aggregated operation data to convert it into regional-level distributed renewable energy participation capability; perform boundary mapping on the user group response data to convert it into user group response boundary; perform boundary mapping on the regional operation data to convert it into regional carrying capacity boundary; and aggregate the set of control constraint intervals, regional-level distributed renewable energy participation capability, user group response boundary, and regional carrying capacity boundary to generate a set of collaborative management scenario constraints.

[0023] Specifically, in the scenario of collaborative management of electricity marketing and distributed renewable energy, although the data from the electricity sales side, the distributed renewable energy aggregation side, the user side, and the distribution network operation side are all related to collaborative decision-making, the organization of the original data is inconsistent. Directly inputting this data into subsequent strategy generation will lead to unclear boundaries, inconsistent granularity, and difficulty in synchronously reflecting regional constraints. Therefore, the task of step one is to organize these existing business data into a unified scenario constraint result, so that subsequent steps deal not with scattered meter entries, equipment readings, or group statistical fragments, but with a collaborative management scenario constraint set that already has clear business meaning. This set of constraints simultaneously reflects the selectable range of marketing parameters, the participation capability of distributed new energy aggregation units, the responsiveness range of user groups, and the carrying capacity boundary of local areas, thereby unifying "what demand response actions can be given," "how many new energy sources can be arranged to participate," "to what extent can user groups respond," and "to what level can local areas bear the load" into the same set of constraints.

[0024] The inputs come from four types of existing business interfaces. The first type is demand response configuration data from the electricity sales system, specifically including the time-of-use pricing tiers, demand response incentive tiers, and green electricity consumption guidance parameters configured for the current period. This data is generally stored in the marketing business database and can be retrieved by time period and user group through a structured query interface. The second type is aggregated operation data from the distributed renewable energy aggregation platform, specifically including real-time output, short-term forecast results, and currently available capacity for coordination in each access area. This data is typically uploaded by inverter-side acquisition terminals, edge aggregation devices, or regional aggregation master stations and provided periodically through the platform interface. The third type is user group response data output from the load forecasting module, specifically including the baseline load results and adjustable range of user groups for the current period. This data is usually calculated jointly from historical load sequences and intraday operation information and output in group-level format. The fourth type is regional operation data from the distribution network operation system, specifically including regional load levels, node operating status, and local renewable energy consumption capacity. This data typically comes from distribution automation terminals, line monitoring devices, and master station-side operation systems, and is aggregated to a regional-level view via the dispatch interface.

[0025] In this stage, the demand response configuration data in the electricity sales system is first organized into intervals. In actual business operations, marketing parameters are often configured in the form of price tiers, such as several price tiers, several incentive tiers, and guidance parameters visible to different user groups for a certain time period. These tiers are suitable for manual configuration but not for subsequent unified solution, so they need to be converted into interval expressions for collaborative management. The specific processing method is as follows: extract the corresponding set of price tiers according to the current decision-making time period, and then compress the discrete tiers into a set of control constraint intervals according to the user groups that can be reached during that time period; the incentive tiers and green electricity guidance parameters are processed in the same way, so that the output of the marketing side is not a set of scattered tier numbers, but a set of marketing boundaries that can be directly called by subsequent steps. For example, if a certain area opens three price tiers and two incentive tiers to commercial user groups during the evening peak period, after processing, a set of control constraint intervals for that group during that time period will be formed. Subsequently, instead of looking up tables to select tiers, feasible strategy candidates will be formed within this interval.

[0026] Subsequently, the operational data in the distributed renewable energy aggregation platform is aggregated at the regional level. Since raw data is often uploaded in the form of single devices, single nodes, or single branches, directly retaining it for subsequent steps would lead to overly granular decision-making, making it difficult to align with marketing and user group data. Therefore, this approach uses the actual execution granularity of collaborative management as the standard, grouping distributed renewable energy units within the same transformer area or region into regional aggregation units and extracting the region's participation capacity for the current time period. Specifically, real-time output and short-term forecast results are first read, then the currently accessible capacity within the collaborative management scope is overlaid to form a regional-level renewable energy participation boundary. For example, if a region includes rooftop photovoltaics, industrial and commercial photovoltaics, and some supporting energy storage, the aggregation platform initially uploads device-by-device readings, but the processed data represents the "total participation capacity that can be formed in the region during the current time period." This ensures that the generated result maintains consistency in granularity with subsequent regional carrying capacity boundaries and user group response intervals.

[0027] Next, the user group response data provided by the load forecasting module is processed. The key here is to organize the baseline load and adjustable range into user response boundaries that can be directly used to construct constraints. The load forecasting module has generally already grouped users by residential, commercial, industrial, or by contract type, and provides the forecast load and adjustable ratio for the current time period. After reading these group-level results in step one, and combining them with the marketing reachable targets for the current time period, the responsive range of each user group in that time period is represented as an interval value, which is used to determine how much response space the group has when facing different demand response actions. To give a more engineering-like example, commercial user groups may have significant transferable load in the afternoon, while industrial user groups have a smaller adjustable range when production schedules are stable. After this organization, the subsequent steps no longer obtain the original forecast curve, but rather the response boundary results of different user groups in the current time period.

[0028] Finally, the regional carrying capacity information in the distribution network operation system is processed and transformed into regional boundaries required for collaborative management. Distribution network operation data is originally distributed across multiple objects such as node load, line status, and local absorption capacity. If each of these is involved in subsequent calculations, the decision-making chain would be excessively long. Therefore, a regional mapping approach is adopted to group the operating status within the same collaborative management area into carrying capacity boundaries. Specifically, the load level and local renewable energy absorption capacity for the current time period are read by region, and then combined with the actual operating status of the region to form a unified carrying capacity expression. This expression describes the extent to which the region can accommodate marketing guidance, renewable energy participation, and user response during the current time period. To ensure that marketing intervals, renewable energy participation capacity, user group response boundaries, and regional carrying capacity boundaries can be placed within the same framework, the results are then uniformly mapped to the same numerical space. During mapping, the effective range of each region in the current time period is used as a benchmark for scaling, resulting in a directly combinable set of collaborative management scenario constraints. Its expression is ; in, This represents the set of constraints for collaborative management scenarios, whose value is composed of four types of regional constraints after being organized within the current time period. This represents the set of control constraint intervals, whose values ​​are derived from the price tiers, incentive tiers, and green electricity guidance parameters read from the electricity sales system by time period, and are obtained after intervalization. It represents the regional-level distributed renewable energy participation capability, and its value comes from the aggregated results of real-time output, short-term forecast, and available capacity of each access unit in the same region in the aggregation platform; This represents the user group response boundary, and its value comes from the group-level baseline load results and adjustable range adjustment results output by the load forecasting module. This represents the regional carrying capacity boundary, and its value originates from the mapping result between the regional load level and the local renewable energy absorption capacity in the distribution network operation system. This is how it is formed. The four types of business information have been compressed into a unified constraint structure that can be directly accessed in subsequent steps. For example, in a region where the proportion of commercial load is high and the photovoltaic output is strong at midday, This will simultaneously reflect the response space of business groups during that period, the participation capability of regional photovoltaic aggregation units, the selectable range of marketing parameters, and the carrying capacity limit of the region itself. Subsequent strategy generation can then be directly based on this scenario. After this processing, key information from the power marketing side, distributed renewable energy aggregation side, user group side, and distribution network operation side is uniformly organized into a collaborative management scenario constraint set. The subsequent steps can be directly centered around Generate feasible strategies for multiple stakeholders without having to go back to the original business data layer for splitting and reorganization.

[0029] S2: Using the set of constraints in the collaborative management scenario as input, generate demand response actions by combining user group response boundaries, generate distributed new energy participation actions by combining regional carrying capacity boundaries, and generate user group response actions by combining demand response actions; combine and filter demand response actions, distributed new energy participation actions, and user group response actions that satisfy regional coupling constraints to generate a set of feasible multi-subject strategies.

[0030] Specifically, the set of constraints for collaborative management scenarios formed step by step. The four types of regional constraints directly related to the scenario of this solution within the current time period have been compiled into a unified set of control constraint intervals. Distributed new energy aggregation capability User group response boundaries and regional carrying capacity boundaries Step Two With these four results as the sole input, instead of reorganizing the data back to the original business data layer, a multi-agent feasible strategy set is generated directly around these four results. The "feasible" here does not refer to a simple stacking of candidate actions, but rather to the fact that demand response actions, new energy participation actions, and user response actions are already within the same time period, the same region, and the same carrying capacity boundary when they are generated. Therefore, what subsequently enters the multi-stakeholder game is not an isolated price action, power generation action, or response action, but a combination of actions that already carry scenario constraints.

[0031] In specific implementation, firstly according to and Generate demand response actions. The time-based price tiers, incentive tiers, and green energy guidance parameters configured in the electricity sales system have already been organized into a set of control constraint intervals in step one. At this point, the generation of demand response actions is not directly from... Instead of selecting points at equal intervals, it combines... The corresponding user group response boundaries are mapped hierarchically. This mapping process can be completed by the rules engine within the main marketing site. The rules engine reads three types of indexes: "region ID, time period ID, and user group ID," queries the response boundary position of each user group in the current time period, and then... Extract the corresponding interval subset. Taking a mixed industrial and commercial area as an example, if the midday commercial user group is in If the middle is in the high-response zone, the rule engine will start from... The selection process involves choosing a subset of intervals with higher green electricity guidance intensity and more significant price difference incentives. If the industrial user group in the same region is in the low to medium response range during the current period, then a subset of intervals with slower price fluctuations and moderate incentive intensity is extracted. The demand response actions generated in this way already have user-side constraint attributes, eliminating the need for subsequent rematching of demand response actions with user response actions.

[0032] Subsequently according to and This generates distributed new energy participation actions. The generation method used here is at the regional aggregation level, specifically completed by the regional strategy unit of the aggregation platform. The regional strategy unit reads data by region. This already includes the region's real-time power output, short-term forecasts, and available capacity consolidation values ​​for the current time period, combined with... The given regional carrying capacity boundary divides the participation of new energy sources within the region into several participation levels. Instead of using a fixed proportional division, the division follows a sequence of "first defining the carrying capacity boundary, then further subdividing the aggregation capacity": first, examine the region... Whether it is under high load or tight constraint, then decide. It is divided into several participation levels. For example, in areas with a high proportion of photovoltaic power and large midday carrying capacity, regional strategy units can... It is divided into a basic participation layer, an enhanced participation layer, and a high absorption participation layer; in areas with tight evening peak loads and small regional carrying capacity, it is further divided into... The participation is divided into a basic participation layer and a conservative participation layer. This results in new energy participation actions that are not isolated segments of power generation plans, but rather regional-level actions that have already taken into account regional carrying capacity.

[0033] User group response actions revolved around Generate and synchronize with demand response actions during the formation process. This is achieved by the load interaction module reading data group by group. The response boundary is defined, and then, based on the marketing interval subset extracted from the main marketing site, response actions at the same level as demand response actions are generated. For example, for commercial groups in the high response range, the load interaction module generates multiple response actions within the complete response range. These actions can be paired with strong green electricity guidance and medium-to-high incentive demand response actions. For industrial groups in the medium-to-low response range, only narrower response actions are generated, paired with relatively mild demand response actions. This generation logic is particularly suitable for the scenario corresponding to this solution, because the collaborative management of electricity marketing and distributed new energy is not just about "giving prices and seeing responses," but also about simultaneously considering whether users can respond, whether photovoltaic power can be consumed locally, and whether the region itself can bear the load within the same area. By synchronously mapping demand response actions and response actions during the generation stage, the subsequent step three, when conducting multi-stakeholder game theory, deals with naturally coupled action units.

[0034] After demand response actions, renewable energy participation actions, and user response actions are all formulated, the next stage is the combination construction phase. The basic formula used here is derived from the feasible region representation in mathematical programming. Classical capacity constraints are typically written as the total resource occupancy not exceeding the carrying capacity boundary, i.e., adding the occupancy corresponding to different actions and comparing it to the regional boundary. Since linear capacity constraints alone are insufficient to express the business fact that "the degree of matching between renewable energy participation actions and user response actions affects the regional carrying capacity effect," a symmetric deviation term is added to the classic capacity constraint to characterize the mismatch in occupancy between renewable energy participation actions and user response actions. When the two are relatively close, this term is small, indicating that renewable energy participation and load response in the region are relatively coordinated; when the gap widens, this term increases, indicating that the regional carrying capacity needs to reserve additional margin for this mismatch. Based on this derivation, a multi-agent feasible strategy set is developed. Construct it according to the following formula: ; in, This represents the set of feasible multi-agent strategies, whose value is the set of all strategy triples that satisfy the regional coupling constraints. This represents a single electricity demand response action, and its value comes from... The marketing interval subset extracted by user group response level; This represents the action of a single distributed renewable energy source, and its value comes from... The new energy participation levels are divided according to regional carrying capacity conditions in China; This represents the response action of a single user group, and its value comes from... The response interval after synchronous mapping of demand-based response actions; This represents the set of control constraint intervals output in step one, which is derived from the current time period's price tier, incentive tier, and green electricity guidance parameters compiled by the electricity sales system. The distributed new energy aggregation capability output in step one is derived from the real-time output, short-term forecast results, and aggregated capacity of each access unit in the same area of ​​the aggregation platform. This indicates the user group response boundary output in step one, which is derived from the results of the load forecasting module's processing of the group-level baseline load and adjustable range. The regional load-bearing boundary output in step one is derived from the mapping result of the distribution network operation system on the regional load level and local absorption capacity. In the formula... This is a scenario-based correction to the classic linear capacity constraint. The absolute value term originates from a mathematical deviation measurement method. The coefficient is set to half, indicating that the mismatch between new energy participation actions and user response actions will occupy the regional carrying space symmetrically, but its occupancy level is lower than that of directly superimposed main terms. Since step one has already... , and Mapped to the same numerical space, the addition and comparison of this expression can be directly established.

[0035] Substituting this formula into a specific region allows for direct implementation calculations. Taking a region with a high proportion of commercial load and strong midday solar power output as an example, step one provides the region's... The aggregation platform provides an action for participating in new energy. The load interaction module provides a user response action corresponding to the medium-to-high guidance demand response action. Then the load occupancy of this combination is The result is in Therefore, this set of three movements was retained until... In the same region, if a higher level of new energy participation is chosen... The user's response action is still... Then the load occupancy becomes The result exceeded This combination does not enter. In another area, during the evening rush hour... Smaller and Also lower, the area strategy unit generation The market is already in a relatively conservative range, so even if the demand response remains moderately incentivized, the corresponding... It will also fall into a narrower response range, and the final strategy combination will still shrink around the region boundary.

[0036] After completing the above combinations, the system continues to... The process involves regional consolidation and reorganization, executed by the strategy portfolio management unit. This unit categorizes existing operations by region and marketing level. The strategy uses triple clustering to group combinations of new energy participation actions that are adjacent, user response actions that are adjacent, and that belong to the same marketing level into the same candidate cluster. Then, within each candidate cluster, it retains the closest combinations. The combination of boundaries is used as a representative action. The reason for this approach is that subsequent step three requires... In solving multi-agent game theory, the existence of numerous approximately equivalent combinations within the same region increases the problem-solving burden. Retaining representative combinations near the region boundary better reflects the coupling relationship between demand response actions, renewable energy participation actions, and user response actions. This process yields the following results. The process of moving from the scenario constraint set in step one has been completed. In the transformation to the strategy space, each strategy element simultaneously includes demand response actions, new energy participation actions, and user response actions. These actions have already been shaped during the generation phase by regional carrying capacity boundary constraints and deviation occupancy terms. Therefore, step three can directly revolve around... A multi-party game is conducted to further generate the final collaborative decision-making result.

[0037] S3: For each candidate item in the set of feasible multi-agent strategies, calculate the collaborative utility value, which includes the matching degree between demand response actions and user group response boundaries, and the matching degree between distributed new energy participation actions and user group response actions; select the candidate item with the largest collaborative utility value as the collaborative decision result.

[0038] Specifically, step three directly uses the multi-agent feasible strategy set output in step two. In step two, each candidate item has been written as a ternary combination. ,in Corresponding to electricity demand response actions, Corresponding to the participation of distributed new energy sources, The corresponding user group response actions, and these items already satisfy the collaborative management scenario constraint set formed in step one. The given region boundary conditions. Therefore, step three no longer regenerates actions, but instead revolves around... The process of "from feasible to optimal" involves identifying the most suitable collaborative decision-making outcome for the current region and time period from all feasible strategy combinations. .

[0039] The core calculation in step three is based on the "utility maximization" concept in classical optimization theory. Traditional approaches typically establish revenue functions for a single entity, such as calculating only electricity sales revenue, only renewable energy consumption revenue, or only user response revenue, and then ranking them separately. This step uses a unified collaborative utility expression, simultaneously incorporating marketing incentives, local renewable energy consumption, and the degree of coordination between actions into a single evaluation formula. Its initial sources are twofold: the first is the utility function concept from economics and game theory, which describes the beneficial effects of candidate actions through positive revenue terms; the second is the absolute deviation measurement method in mathematics, which describes the degree of deviation between two related actions through absolute value terms. In the scheme, demand response actions... User response actions The degree of alignment between these factors determines whether marketing efforts can translate into real responses, and the participation of new energy entities... User response actions The degree of compatibility between these factors determines whether new energy sources can be absorbed locally. Based on this understanding, we will start with the two most basic positive synergistic products, namely... and .in Reflecting the consistent gain between marketing intensity and response intensity, This reflects the consistent gain between renewable energy participation and load response. Two deviation correction terms are then added: Used to describe the mismatch between new energy participation and user response. This describes the mismatch between marketing intensity and user response boundaries. Considering that the former directly affects the consistency of renewable energy consumption and load absorption within the region, while the latter mainly affects execution stability indirectly through the marketing chain, a higher correction weight is assigned to the former, and a lower correction weight to the latter. In step two, all candidate actions have been mapped to the same numerical space, therefore the multiplication, addition, and comparison operations in the following formula are all performed on the same scale. This yields the collaborative utility value of a single candidate item: ; in, Indicates candidate entries The collaborative utility value is calculated item by item by the collaborative decision engine; This represents a single electricity demand response action generated in step two, the value of which is derived from the set of control constraint intervals. Write after group mapping Marketing fields; This represents the action of a single distributed renewable energy source generated in step two, and its value originates from the distributed renewable energy aggregation capability. Write after regional participation layer segmentation The new energy field; This represents the single user group response action generated in step two, and its value is derived from the user group response boundary. Write after marketing synchronization mapping The response field. The first term in the formula... The positive benefit term from the utility function indicates that the closer the demand response action is to the actual response boundary of the user group, the stronger its ability to transform into an effective response; the second term... Also derived from the positive benefit term in the utility function, it indicates that the more renewable energy can be absorbed by local response loads, the higher the degree of local consumption; the third term Derived from the absolute deviation penalty method, it is used to mitigate the mismatch between new energy participation and user response; the fourth item Also derived from the absolute deviation penalty method, it is used to narrow the mismatch between demand response actions and user response boundaries. One-half and one-quarter are scenario-based characterizations of the impact of the two types of deviations, demonstrating that the impact of new energy participation and response mismatch on regional coordination is stronger than the impact of demand response action and response boundary mismatch on regional coordination. This formula yields the comprehensive evaluation value of a single candidate item; the subsequent processing in step three involves... This evaluation value is compared and selected in the middle.

[0040] After obtaining the collaborative utility values ​​of all candidate items, the maximum value selection rule from classical optimization theory is used to select the best option. The candidate item with the highest collaborative utility value is selected as the collaborative decision result. The initial source of this rule is in mathematical optimization. The selection operator, in essence, chooses the element from a given set that maximizes the objective function. The change in this step lies not in the operator itself, but in the replacement of the traditional single-benefit function with the scenario-based synergistic utility function given in the above formula. Therefore, the selected item is not a one-sided optimal choice like "strongest marketing" or "highest renewable energy," but rather an item that achieves overall optimality across marketing incentives, renewable energy consumption, and regional synergy. Specifically, it is written as: ; in, This represents the collaborative decision-making result. In the system implementation, an optimal entry is obtained for each region and time period, and then aggregated into a set of decision results for the current period. This represents the multi-agent feasible strategy set output in step two, which includes all policy triples that satisfy the feasible boundary within the current region and current time period. This represents the synergistic utility value calculated using the previous formula. The logical relationship between the two formulas is very direct: first use the first formula to calculate... For each candidate item in the set, calculate the collaborative utility value, then use the second equation to perform maximum selection on the same set, finally obtaining... Therefore, the first formula is the evaluation basis for the second formula, and the second formula is the selection rule for the first formula. The two are connected to form a complete decision-making process.

[0041] By combining the calculation process of a specific region, this decision-making chain can be written more clearly. Suppose there is a region with a high proportion of commercial load and concentrated distributed photovoltaic power output at midday. In the current time period, step two outputs three candidate items. The first candidate item is... The second candidate entry is The third candidate entry is Substituting into the first equation, we first calculate the collaborative utility value of the first candidate item: , , , ,therefore Then calculate the second candidate entry: , , , ,therefore Finally, the third candidate entry is calculated: , , , ,therefore Next, we substitute these three results into the second equation for comparison, because... The largest, so the second candidate was selected. From a scenario perspective, this result corresponds to "a good match between demand response actions and user group response boundaries, while the participation of new energy sources can be well absorbed by the response load within the same region." Therefore, it outperforms the other two candidate items in overall evaluation. Although the first candidate item is relatively balanced overall, its marketing traction and new energy absorption intensity are slightly low; although the third candidate item has high demand response actions and new energy participation actions, there is a large deviation between it and the user group response boundaries, so its utility value decreases under the deviation correction term.

[0042] At the system implementation level, the collaborative decision-making engine executes the above process cyclically by region and time period. For each region, it first starts from... The algorithm reads all candidate entries for the current time period in the region; then substitutes each entry into the first equation to obtain the corresponding collaborative utility value; finally, it selects the optimal entry based on the second equation and writes it into the database. This process can be directly verified through debugging logs, such as saving "candidate entry number, ..." in the engine log. value, value, The system uses six fields: "value," "synergistic utility value," and "whether it was selected." For a complete intraday cycle, the log will generate a sequence of results arranged by time slices, facilitating subsequent comparison of execution effects. For simulation verification, three schemes can be run simultaneously on historical samples: "sorted only by marketing revenue," "sorted only by new energy participation," and "sorted by synergistic utility of this step." Typically, the results obtained in this step will be observed... The improvement achieves a more balanced approach in terms of local renewable energy utilization and response stability. This improvement is not achieved by adding extra modules, but rather by further incorporating the three-element coupling action already formed in step two into a unified synergistic utility expression. This ensures that the scenario boundary in step one, the feasible combinations in step two, and the decision-making process in step three complete a closed loop along the same logical chain. The final output... Retained , and Based on the correspondence of the three types of actions, step four can transform demand response actions into price and incentive execution content, new energy participation actions into regional power generation participation arrangements, and user response actions into response organization content for target groups, thereby forming a complete collaborative management execution chain.

[0043] S4: Map the demand response actions in the collaborative decision-making results to actual execution prices and incentive parameters, allocate distributed new energy participation actions to equipment-level scheduling instructions, and map user group response actions to response instructions of the target user set; perform regional carrying capacity boundary verification on the mapped execution content, and output the verified equipment-level scheduling instructions and response instructions to the corresponding distributed new energy power generation equipment and user load control terminals to adjust the active power output of the distributed new energy power generation equipment and the power consumption of the user load control terminals.

[0044] Specifically, step three has already generated collaborative decision-making results. Each of these entries is named after a specific character. The form exists, and the control constraint range has been satisfied simultaneously in the aforementioned steps. Distributed new energy aggregation capability User group response boundaries and regional carrying capacity boundaries The synergistic relationship between them. Step four, based on this, will... The abstract strategy combination is transformed into specific execution content that can be directly executed in the power marketing system, the distributed new energy aggregation system, and the user side. This transformation process follows the mapping logic of "strategy value to execution quantity," based on the proportional mapping method in control systems. This involves mapping normalized strategy variables to actual control quantities while maintaining the proportional relationship between strategies. In this solution, at the execution layer, it's necessary to ensure that the synergistic relationship between marketing intensity, power generation participation, and user response amplitude remains consistent at the physical execution level. Therefore, a uniform scaling method is used in the mapping process to ensure... The relative relationship formed in step three is maintained during execution.

[0045] When performing the mapping, first respond to the demand action. Parameter conversion is performed. The electricity sales system internally stores price templates and incentive templates divided by time period and user group. Each template specifies the allowed parameter range for that region and group within the current period. The execution module reads... Then, this is used as a range location parameter to map to specific prices and incentive levels. For example, when When the corresponding interval median is reached, it is mapped to the median price in the template; when When the price is biased towards the higher end of the range, it is mapped to a higher price and stronger incentive in the template. This mapping process is completed by querying the marketing template library. Each item in the template library is configured by the operations side and stored in the database. The execution module reads and calculates the actual execution value through an interface.

[0046] Actions related to distributed new energy The processing employs a region-based decomposition approach. The aggregation platform maintains a table mapping regions to devices, recording the current available capacity of each photovoltaic unit and energy storage unit within each region. The execution module then... The indicated regional participation level is allocated to each device according to its capacity proportion. For example, if there are multiple photovoltaic units in a certain area, each with a different current generating capacity, the execution module first calculates the capacity proportion of each unit, and then allocates the resources according to that proportion. The allocation is divided into multiple device-level scheduling instructions, thereby obtaining specific power generation participation arrangements. This process ensures that the regional-level strategy can be mapped to device-level execution, while maintaining the participation ratio relationship determined in step three.

[0047] Response actions for user groups Execution relies on a load interaction system. This system maintains a mapping between user groups and specific users, and records the current participation status of each user. The execution module then... The target response strength is determined, and a set of users meeting the criteria is selected from the corresponding groups. The selection process is based on users' historical response records and current online status. Subsequently, response requests are sent to the target users via a communication interface, or control commands are directly issued. For users with automatic control capabilities, load adjustment signals can be directly issued; for ordinary users, price signals and incentive notifications guide their participation in the response.

[0048] After completing the above three types of mapping, the execution results need to be consistent to ensure consistency between the execution layer and the strategy layer. This consistency check method originates from classic capacity constraint judgment, which involves superimposing the new energy participation and user response quantities after execution and comparing them with the regional carrying capacity boundary. In step three, the collaborative decision-making results have already been filtered based on the utility function, but due to discrete mapping and equipment decomposition at the execution layer, slight deviations may occur, thus requiring another rapid consistency check. This consistency check is expressed as: ; in, This represents the total amount of new energy participating in the region after the mapping is executed, which is obtained by summarizing the instructions from each device after decomposition by the aggregation platform; This represents the total response of the user group after the mapping is executed, which is obtained by the load interaction system by statistically analyzing the response magnitude of the target user set; This represents the regional carrying capacity boundary obtained in step one, whose value is provided by the distribution network operation system. The absolute value term comes from the deviation measurement method introduced in steps two and three, used to characterize the degree of mismatch between renewable energy participation and user response. This expression is consistent with the feasibility constraints in step two, ensuring that the execution layer and the strategy generation layer use the same logic.

[0049] The specific calculation process can be illustrated using a regional example. Assume the collaborative decision-making result for a certain region before execution is... The execution module will The photovoltaic units were allocated to three photovoltaic modules, with their respective capacity proportions as follows: , and The corresponding allocation result is , and The summaries are still as follows At the same time Mapped to a business user group, a number of users are selected to form a response set, the total response magnitude of which is Substituting into the verification expression, we obtain... This result is at the regional carrying capacity boundary. If the result of the mapping exceeds the allowed range, the execution entry is deemed valid and enters the actual execution process. Then the execution module will be based on and The proportional relationship is synchronously scaled to re-satisfy the above conditions, thereby ensuring that the execution process always remains within the safe range of the region. After the above processing, the execution content obtained in step four is... It consists of execution items across multiple regions and time periods. Each item includes corresponding marketing execution parameters, renewable energy dispatch instructions, and user response instructions. These instructions maintain a consistent logical relationship with steps one through three throughout the generation and verification process. In this way, the collaborative decision-making results can be stably executed in the actual system, thereby achieving the overall goal of collaborative management of electricity marketing and distributed renewable energy.

[0050] In one or more embodiments, such as Figure 2 As shown, a multi-entity collaborative source-load interaction and new energy management system is disclosed, the system comprising: The scenario constraint construction module is used to acquire demand response configuration data from the electricity sales side, aggregated operation data from the distributed renewable energy aggregation side, user group response data from the user side, and regional operation data from the distribution network operation side. It performs interval-based processing on the demand response configuration data to convert it into a set of control constraint intervals, performs regional-level aggregation on the aggregated operation data to convert it into regional-level distributed renewable energy participation capabilities, performs boundary mapping on the user group response data to convert it into user group response boundaries, and performs boundary mapping on the regional operation data to convert it into regional carrying capacity boundaries. Finally, it aggregates the set of control constraint intervals, regional-level distributed renewable energy participation capabilities, user group response boundaries, and regional carrying capacity boundaries to generate a collaborative management scenario constraint set. The strategy generation module is used to take the set of constraints of the collaborative management scenario as input, generate demand response actions by combining user group response boundaries, generate distributed new energy participation actions by combining regional carrying capacity boundaries, and generate user group response actions by combining demand response actions; and combine and filter demand response actions, distributed new energy participation actions, and user group response actions that satisfy regional coupling constraints to generate a set of feasible multi-subject strategies. The decision-solving module is used to calculate the collaborative utility value for each candidate item in the set of feasible strategies for multiple subjects, including the matching degree between demand response actions and user group response boundaries, and the matching degree between distributed new energy participation actions and user group response actions; and select the candidate item with the largest collaborative utility value as the collaborative decision result. The execution conversion module is used to map the demand response actions in the collaborative decision-making results to actual execution prices and incentive parameters, allocate distributed renewable energy participation actions to equipment-level scheduling instructions, and map user group response actions to response instructions for the target user set; it performs regional carrying capacity boundary verification on the mapped execution content, and outputs the verified equipment-level scheduling instructions and response instructions to the corresponding distributed renewable energy power generation equipment and user load control terminals to adjust the active power output of the distributed renewable energy power generation equipment and the power consumption of the user load control terminals.

[0051] It is worth noting that the specific workflow of the multi-entity collaborative source-load interaction and new energy management system provided in this embodiment of the invention is the same as that of the multi-entity collaborative source-load interaction and new energy management method described in the above embodiments, and will not be repeated here.

[0052] This invention also provides a multi-entity collaborative source-load interaction and new energy management device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps described in the above embodiments of the multi-entity collaborative source-load interaction and new energy management method, for example... Figure 1 The steps S1 to S4 described above; or, when the processor executes the computer program, it implements the functions of each module in the above system embodiments.

[0053] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A multi-agent collaborative source-load interaction and new energy management method, characterized in that, The method includes: Acquire demand response configuration data from the electricity sales side, aggregated operation data from the distributed renewable energy aggregation side, user group response data from the user side, and regional operation data from the distribution network operation side; perform interval-based processing on the demand response configuration data to convert it into a set of control constraint intervals, perform regional-level aggregation on the aggregated operation data to convert it into regional-level distributed renewable energy participation capability, perform boundary mapping on the user group response data to convert it into user group response boundaries, perform boundary mapping on the regional operation data to convert it into regional carrying capacity boundaries, and aggregate the set of control constraint intervals, regional-level distributed renewable energy participation capability, user group response boundaries, and regional carrying capacity boundaries to generate a set of constraints for collaborative management scenarios; Using the set of constraints in the collaborative management scenario as input, demand response actions are generated by combining user group response boundaries, distributed new energy participation actions are generated by combining regional carrying capacity boundaries, and user group response actions are generated by combining demand response actions. Demand response actions, distributed new energy participation actions, and user group response actions that satisfy regional coupling constraints are combined and filtered to generate a set of feasible strategies for multiple subjects. For each candidate item in the multi-agent feasible strategy set, calculate the collaborative utility value, which includes the matching degree between demand response actions and user group response boundaries, and the matching degree between distributed new energy participation actions and user group response actions; select the candidate item with the largest collaborative utility value as the collaborative decision result. The demand response actions in the collaborative decision-making results are mapped to actual execution prices and incentive parameters, distributed renewable energy participation actions are allocated as equipment-level scheduling instructions, and user group response actions are mapped to response instructions of the target user set. The mapped execution content is then checked against the regional carrying capacity boundary, and the checked equipment-level scheduling instructions and response instructions are output to the corresponding distributed renewable energy power generation equipment and user load control terminals to adjust the active power output of the distributed renewable energy power generation equipment and the power consumption of the user load control terminals.

2. The multi-agent coordinated source-load interaction and new energy management method according to claim 1, characterized in that, The step of converting the demand response configuration data into a set of control constraint intervals includes: Extract the set of price tiers, incentive tiers, and green electricity consumption guidance parameters corresponding to the current decision-making period; Based on the user groups that can be reached during the current decision-making period, the set of price tiers, the set of incentive tiers, and the set of green electricity consumption guidance parameters are compressed into corresponding interval expressions to form the set of regulatory constraint intervals.

3. The multi-agent coordinated source-load interaction and new energy management method according to claim 1, characterized in that, The process of performing regional-level aggregation on the aggregated runtime data includes: Read real-time output data and short-term forecast results from the same region in the distributed new energy aggregation platform; overlay the available capacity currently connected to the collaborative management scope to form a regional distributed new energy participation capability that represents the total participation capability that the region can form in the current time period.

4. The multi-agent coordinated source-load interaction and new energy management method according to claim 1, characterized in that, The actions for generating distributed new energy sources include: Read the regional-level distributed new energy participation capacity by region and determine the carrying capacity status of the regional carrying capacity boundary; When the regional carrying capacity boundary is in a high carrying capacity state, the regional distributed new energy participation capability is divided into distributed new energy participation actions including the basic participation layer, the enhanced participation layer and the high consumption participation layer. When the regional carrying capacity boundary is under tight constraints, the regional-level distributed new energy participation capability is divided into distributed new energy participation actions that include a basic participation layer and a conservative participation layer.

5. The multi-agent coordinated source-load interaction and new energy management method according to claim 1, characterized in that, The generation of the multi-agent feasible strategy set specifically includes: The demand response action, the distributed new energy participation action, and the user group response action are combined to calculate the combined regional coupling constraint value. The regional coupling constraint value includes the superposition value of demand response action and user group response action, as well as the deviation occupancy value of distributed new energy participation action and user group response action. When the regional coupling constraint value is less than the regional carrying capacity boundary, the combination is retained in the multi-agent feasible strategy set.

6. The multi-agent coordinated source-load interaction and new energy management method according to claim 1, characterized in that, The calculation of the synergistic utility value specifically includes: Calculate the positive collaborative product term between the demand response action and the user group response boundary; Calculate the positive collaborative product term between distributed new energy participation actions and user group response actions; Calculate the deviation penalty term between distributed new energy participation actions and user group response actions, as well as the deviation penalty term between demand response actions and user group response boundaries; The synergistic utility value is obtained by weighted summation of the four calculation results.

7. The multi-entity collaborative source-load interaction and new energy management method according to claim 1, characterized in that, The selection of the collaborative decision-making results specifically includes: The collaborative utility value of each candidate item in the multi-agent feasible strategy set is compared; the candidate item with the largest collaborative utility value that simultaneously satisfies the demand response action constraint, the degree of local consumption of new energy, and the regional carrying capacity boundary is selected as the collaborative decision result.

8. The multi-entity collaborative source-load interaction and new energy management method according to claim 1, characterized in that, The allocation of actions for participation in distributed new energy sources specifically includes: Read the mapping table between regions and devices in the aggregation platform, calculate the current available generation capacity ratio of each photovoltaic unit and energy storage unit in the region; allocate the distributed new energy participation actions as at least one device-level scheduling instruction according to the capacity ratio, and maintain the participation ratio relationship determined in the collaborative decision result.

9. The multi-entity collaborative source-load interaction and new energy management method according to claim 1, characterized in that, The specific steps for performing region-bearing boundary verification on the mapped execution content include: Calculate the sum of the total amount of new energy participation and the total amount of user group response within the region after the mapping is executed, as well as the mismatch deviation between the two; when the sum of the sum and the deviation exceeds the regional carrying capacity boundary, the execution content is synchronously scaled according to the ratio between the total amount of new energy participation and the total amount of user group response until the regional carrying capacity boundary is met.

10. A multi-entity collaborative source-load interaction and new energy management system, characterized in that, include: The scenario constraint construction module is used to acquire demand response configuration data from the electricity sales side, aggregated operation data from the distributed renewable energy aggregation side, user group response data from the user side, and regional operation data from the distribution network operation side. It performs interval-based processing on the demand response configuration data to convert it into a set of control constraint intervals, performs regional-level aggregation on the aggregated operation data to convert it into regional-level distributed renewable energy participation capabilities, performs boundary mapping on the user group response data to convert it into user group response boundaries, and performs boundary mapping on the regional operation data to convert it into regional carrying capacity boundaries. Finally, it aggregates the set of control constraint intervals, regional-level distributed renewable energy participation capabilities, user group response boundaries, and regional carrying capacity boundaries to generate a collaborative management scenario constraint set. The strategy generation module is used to take the set of constraints of the collaborative management scenario as input, generate demand response actions by combining user group response boundaries, generate distributed new energy participation actions by combining regional carrying capacity boundaries, and generate user group response actions by combining demand response actions; and combine and filter demand response actions, distributed new energy participation actions, and user group response actions that satisfy regional coupling constraints to generate a set of feasible multi-subject strategies. The decision-solving module is used to calculate the collaborative utility value for each candidate item in the set of feasible multi-agent strategies, including the matching degree between demand response actions and user group response boundaries, and the matching degree between distributed new energy participation actions and user group response actions. The candidate item with the highest collaborative utility value is selected as the collaborative decision result. The execution conversion module is used to map the demand response actions in the collaborative decision-making results to actual execution prices and incentive parameters, allocate distributed renewable energy participation actions to equipment-level scheduling instructions, and map user group response actions to response instructions for the target user set; it performs regional carrying capacity boundary verification on the mapped execution content, and outputs the verified equipment-level scheduling instructions and response instructions to the corresponding distributed renewable energy power generation equipment and user load control terminals to adjust the active power output of the distributed renewable energy power generation equipment and the power consumption of the user load control terminals.

Images