Multi-agent joint transaction strategy optimization method and system for regional integrated energy system, and storage medium

Abstract

The invention relates to the technical field of regional integrated energy system optimization and power demand response transaction, in particular to a regional integrated energy system multi-subject joint transaction strategy optimization method, which comprises the following steps: establishing a multi-subject multi-source measurement and machine account acquisition system, and calculating the schedulable capacity of each subject in a time period t; constructing subitem cost functions including fuel and emission, user comfort / production loss, equipment life loss and rebound cost; generating a segmented incremental quotation curve conforming to energy and network security constraints; constructing a non-cooperative game model of the aggregator-response subject; and system transaction earnings are calculated, the overall cost is saved, the earnings are fairly distributed according to marginal contribution by using a Shapley value algorithm, and optimization of a multi-agent joint transaction strategy is realized. According to the method, through refined capacity modeling and multi-agent game clearing, operation boundary, cost and quotation are integrally linked, flexible resource quantification and transaction efficiency are improved, and source-network-load-storage collaborative optimization and safe and economical operation of a power grid are effectively supported.

Description

Technical Field

[0001] This invention relates to the field of demand response (DR) and market clearing technology for integrated energy systems (CIES), specifically to a method, system, and storage medium for optimizing multi-entity joint trading strategies in regional integrated energy systems. It covers a complete set of methods and systems for multi-source measurement and capacity characterization, component cost modeling and marginal cost curve generation, game clearing of aggregators and responders, and settlement and fair allocation of Shapley values. Background Technology

[0002] Regional integrated energy systems integrate combined heat and power (CHP), energy storage (ES), electric vehicles (EV), building air conditioning (HVAC), and electricity / heat / gas network coupling devices, exhibiting characteristics such as multi-energy coupling, time-varying load and price, and limited equipment lifespan. Existing demand response and trading mechanisms primarily focus on the single electricity side, often approximating terminal flexibility by reducing it proportionally to rated power. Cost modeling often neglects integrated pricing for comfort / output losses, lifespan reduction, rebound effects, and fuel / emissions, leading to uninterpretable bids, unstable clearing, and unfair settlement. Simultaneously, traditional centralized dispatch ignores differences in strategic behavior and risk preferences among multiple stakeholders, making it difficult to balance system security, economics, and user experience. Therefore, a joint trading strategy based on data-mechanism fusion, connecting the entire "capacity-cost-bid-clearing-allocation" chain, is urgently needed to improve executability and fairness. Summary of the Invention

[0003] This invention proposes an optimization method for multi-entity joint trading strategies in regional integrated energy systems. Based on multi-source operational data, it establishes an adjustable capacity library considering equipment physical constraints and multi-energy coupling; constructs marginal cost curves using cost functions for comfort / output loss, lifespan depreciation, fuel and emissions, and rebound; designs segmented incremental bidding curves that comply with energy and network security constraints; completes clearing of the unified time-period incentive price and optimal demand response configuration within a non-cooperative game framework between aggregators and responders; settles based on measured load shedding and uses Shapley values ​​for fair allocation, outputting the settlement and revenue distribution results. The technical solution of this invention to solve the above problems is as follows:

[0004] Establish a multi-entity, multi-source measurement and record-keeping system to calculate the dispatchable capacity of each entity in time period t; wherein the entities include HVAC groups, energy storage systems, electric vehicle clusters, and gas equipment;

[0005] Based on the schedulable capacity, a component cost function is constructed; wherein the component cost function includes fuel and emissions, user comfort / output loss, equipment life depreciation, and rebound cost.

[0006] Based on the itemized cost function, a piecewise increasing price curve that meets energy and network security constraints is generated;

[0007] Based on the quoted price curve, a non-cooperative game model of aggregator-respondent is constructed to determine the incentive price and optimal demand response configuration for a unified time period.

[0008] Based on a unified time-period incentive price and optimal demand response configuration, the system's transaction revenue and overall cost savings are calculated. The Shapley value algorithm is then used to fairly distribute the revenue according to marginal contribution, thereby optimizing the multi-entity joint transaction strategy.

[0009] The specific process for calculating the schedulable capacity of multi-entity, multi-source measurements is as follows:

[0010] By establishing a multi-source measurement system spanning electricity, heat, gas, and the terminal side, operational status and boundary condition data of flexible resources such as air conditioning (HVAC), electric vehicles (EV), energy storage systems (ES), and gas equipment in the integrated energy system are collected, and based on this, a dispatchable capacity database for each entity in time period t is constructed. Specifically:

[0011] Multi-source measurement data were collected through power monitoring, terminal control, and environmental sensing equipment, including:

[0012] Power side: via Collect household / building level active power Baseline power Node voltage The time resolution is 5 minutes;

[0013] Heating, ventilation and air conditioning (HVAC): Indoor temperature is collected through building temperature control terminals. Set temperature Permissible temperature difference Rated power Thermal inertia time constant Switch status .in Usually taken , Typical – min;

[0014] Electric Vehicles (EVs): On-site connection data is collected through the pile-side / BMS / station control system. Average state of charge Single vehicle battery capacity Health Station-level grid-connected power limit The time resolution is 5 minutes;

[0015] Energy storage system (ES): through Collect state of charge Health Maximum charge and discharge power , With rated energy capacity The time resolution is 5 minutes;

[0016] Hot air side: Regional heat load is collected through flow meters at heat exchange stations or gate stations. Maximum thermal power of gas boiler and CHP coupling coefficient ;in For constant values: back-pressure steam turbine , Gas-fired internal combustion engine , ;

[0017] Based on the collected multi-source measurement data and the physical parameters of various types of equipment, the schedulable capacity of each type of entity is calculated, including:

[0018] 1) Peak-shaving capacity of HVAC systems:

[0019] ;

[0020] in The control interval is set at 5 minutes. This indicates the air conditioner's on / off status.

[0021] 2) Dispatchable capacity of electric vehicle aggregation:

[0022] ;

[0023] in This represents the participation rate (typically 0.7%). The target upper limit is 0.8.

[0024] 3) Available load shedding capacity of the energy storage system:

[0025] ;

[0026] in The minimum permissible state of charge (fixed value 0.2).

[0027] 4) The adjustable capacity of the gas-fired boiler equipment is:

[0028] ;

[0029] in, The adjustable capacity of the gas-fired equipment at t; This represents the boiler's maximum thermal power. This represents the current electrical power output of the CHP. , The electrothermal coupling coefficient; This refers to the calorific value of the fuel. For boiler efficiency.

[0030] Based on the adjustable capacity of the above entities, a comprehensive schedulable capacity pool is formed;

[0031] ;

[0032] This capacity library As input for subsequent cost modeling, i represents the subject index, and T represents the transaction time period. The maximum demand response schedulable capacity (MW) that entity i can provide in time period t provides the physical basis for subsequent marginal cost and quotation generation.

[0033] Alternatively, the process of itemized cost modeling and marginal cost calculation is as follows:

[0034] Based on the acquired multi-entity schedulable capacity library We construct component cost functions that include fuel and emissions, user comfort / output loss, equipment life depreciation, and rebound costs, in order to establish the cost structure for each entity over a given period. The total cost model is used to calculate the corresponding marginal cost curve.

[0035] Read the schedulable capacity of each entity State variables (including indoor temperature) Set temperature Permissible temperature difference State of charge Health Equipment and price parameters (rated power) Rated energy storage capacity Gas prices Emission coefficient Heat source efficiency ).set up As the main body in time period The amount of power reduction.

[0036] Constructing the total cost function, main body The total cost function is defined as:

[0037] ;

[0038] ;

[0039] ;

[0040] ;

[0041] ;

[0042] in Temperature deviation weighting; Indoor temperature; Set the temperature; To allow for temperature difference (typical value) ); Power adjustment secondary weighting; To reduce power; The depreciation factor per unit energy flux is (e.g., 30 yuan / MWh for ES and 50 yuan / MWh for EV). Energy flux (MWh) for this period; Set as the control / settlement interval (fixed value 15 minutes). Health score (typically 0.92); This refers to the unit price of gas; This represents the equivalent heat demand or fuel flux for this period, used to quantify the additional fuel required for heat-side compensation due to power-side load shedding. For heat source or unit efficiency, the value ranges from 0 to 1; Carbon price (yuan / ton) ). Emission factor (tons) ); This is for a rebound with secondary weighting.

[0043] right Taking the partial derivative, we obtain the main body. The marginal cost curve.

[0044] ;

[0045] And require The upper bound remains monotonically constant to ensure economic rationality and the non-decreasing nature of the subsequent price curve. If this condition is not met, the lower bound of the secondary weight is increased (e.g., by increasing the weighting factor). or (and / or smooth the temperature deviation term) to convexity .

[0046] Output the marginal cost curves of each entity. With capacity constraints (From S1), and will The curve slope information is used as input to the quotation function to determine the starting price. With slope parameter .

[0047] Alternatively, the specific process of the segmented incremental price curve algorithm is as follows:

[0048] Based on the marginal cost curves of each entity obtained above Furthermore, by combining its capacity constraints and risk preference parameters, a piecewise increasing bid function that satisfies energy and network constraints is constructed. Specifically:

[0049] Extract the subject from the cost function modeling results. During the period Marginal cost starting point value and the upper limit of schedulable capacity Simultaneously read risk preference parameters. (Reflecting the degree of risk aversion) and price stability parameters (Reflecting price deviation tolerance), these parameters can be determined by contractual agreement.

[0050] Based on the obtained marginal cost starting point and capacity limit, construct the quantity-price linear quotation function for entity i:

[0051] ;

[0052] in, Price per unit of unloaded load (RMB / MWh); The starting price for the bid is defined as:

[0053] ;

[0054] in, From the starting point of the marginal cost curve mentioned above, This is a risk surcharge item used to reflect the entity's expected compensation for risk; The slope of the quote is based on the weighting parameters of the quadratic term mentioned above ( , ) and risk buffer The calculation yields a value used to describe the increase in price as the load reduction increases. Among these,

[0055] ;

[0056] For risk preference parameters; A higher value indicates a higher degree of risk aversion. The standard deviation of its expected revenue from load shedding services (RMB / MWh); low volatility ≈ 200 - 800 yuan / MWh corresponds to demand response projects with fixed-price compensation or long-term contracts, offering very stable and predictable returns; with moderate volatility. The price of approximately 800-1800 yuan / MWh corresponds to the day-ahead market or peak-shaving ancillary service market, and the price fluctuates to some extent depending on the supply and demand of the system; high volatility. ≈1800 - 3000+ yuan / MWh corresponds to a real-time balancing market or emergency demand response, where prices may fluctuate drastically in a very short period of time, or even reach extremely high prices;

[0057] The pricing constraints for the quantity-price linear pricing function are set as follows:

[0058] ;

[0059] in, as the main body The average starting price in historical transactions; The standard deviation of historical quotes; This represents the risk tolerance coefficient (typical value 1.64). This is the maximum allowable price slope (typically set at 120 yuan / MW²·h).

[0060] After obtaining a linear pricing function that satisfies the constraints, when the entity has graded response capabilities or different adjustable ranges, the marginal cost function can be used as a basis. For the adjustment range Perform piecewise fitting to generate piecewise linear price segments:

[0061] ;

[0062] in The price quote will be divided into segments, with the segmentation points selected based on the cost change rate. Each segment must meet the non-decreasing condition. .

[0063] If it appears If the non-convexity or the monotonicity of the price is violated, then the quadratic weighting coefficient ( , Perform convexity adjustment to make The constant validity of this rule guarantees a monotonically increasing price curve.

[0064] After constraint verification and segmented pricing, the feasible pricing points of each entity are compared. By summarizing, we obtain the set of price curves for time period t:

[0065] ;

[0066] in, The amount of regulating power (MW) that subject i can provide in time period t. To adjust the unit price (RMB / MWh) for this power level, point-to-point This represents the pricing intentions of the subjects at different levels of adjustment, and the set The bid collection comprises all resource entities participating in the bidding. This forms the complete market quote input for time period t.

[0067] This curve serves as the input to a non-cooperative game model, enabling market clearing calculations between aggregators and responders.

[0068] Alternatively, the non-cooperative game model of the aggregator-responder entity can be as follows:

[0069] The aggregator and the responders constitute a typical price-volume game model, in which the aggregator is the dominant party, responsible for determining the uniform incentive price of the system. The responding entity is the follower, which makes the optimal load reduction decision based on the quoted parameters and its own adjustable capacity.

[0070] Given an incentive price Below, each responding entity The goal is to maximize its own benefits, that is, to achieve a balance between benefits and costs. Its optimization problem can be formulated as:

[0071] ;

[0072] Combined with quotation function Differentiating the above equation and satisfying the KKT conditions, the optimal response of the subject is obtained as follows:

[0073] ;

[0074] That is, when the incentive price Below the initial marginal price When the incentive price is in effect, the main body does not participate in the response; when the incentive price is in effect... and When the incentive price is between these values, the main body responds according to a linear function; when the incentive price is too high, it is subject to the capacity limit. limit.

[0075] Aggregators need to reduce the system load by a certain amount. Under constraints, select the optimal incentive price. To minimize system payment costs:

[0076] ;

[0077] Simultaneously satisfying system objective constraints:

[0078] ;

[0079] This constraint reflects the condition where the sum of the optimal responses of all agents exactly satisfies the system load reduction target. This is the clearing incentive price. Because each All about Since it is a monotonically non-decreasing function, the above equation has a unique monotonically increasing solution. The clearing price can be obtained through numerical iteration (such as the bisection method or Newton's iteration). .

[0080] When clearing price After obtaining the result, substitute it into the follower response function to obtain the optimal demand response configuration result for each subject:

[0081] ;

[0082] At this point, the sum of the load reductions of all responding entities satisfies the system objective:

[0083] ;

[0084] in, The incentive price is set at a uniform time period (RMB / MWh). as the main body During the period The optimal load reduction decision result (MW).

[0085] This clearing process simultaneously satisfies the goals of minimizing aggregator payment costs and reducing system load, while also ensuring the Nash equilibrium of the game and market stability.

[0086] Alternatively, the specific process of the profit distribution algorithm is as follows:

[0087] Based on the obtained uniform incentive price Optimal demand response results of each entity The transaction revenue and system cost savings of each entity are calculated, and the Shapley value algorithm is used for fair allocation. Specifically:

[0088] Calculate the time period for each entity based on the measurement data. Actual load reduction:

[0089] ;

[0090] in, The baseline power established in step S1, This is the actual measured power.

[0091] Combined with the clearing incentive price in step S4 Calculation subject Settlement and payment amount:

[0092] ;

[0093] in, For settlement interval, The performance coefficient determined by the accuracy of response execution ( ).

[0094] Calculate the system savings based on the obtained total system incentive cost and the baseline system operating cost:

[0095] ;

[0096] in, This represents the total system benefit with the participation of all stakeholders. This represents the system operating cost when there is no demand response.

[0097] definition as a subset of the main body The total benefit of the system upon participation. For any given entity. Its alliance The marginal contribution is:

[0098] ;

[0099] in, The calculations are the same as above, both based on S4. and Obtain it again.

[0100] Based on the marginal contribution of each entity, the Shapley value method is used for fair allocation:

[0101] ;

[0102] in, as the main body In the league The weight of fair benefit allocation in the process.

[0103] Each subject The final profit is:

[0104] ;

[0105] in, The apportionment weighting coefficient (take) ( ), used to balance immediate gains with long-term equitable distribution.

[0106] This algorithm drives revenue settlement through the output price-response results and combines the Shapley value method to quantify the contribution of each entity to the alliance's revenue, ensuring the fairness and compatibility of transaction incentives and achieving the optimal distribution of revenue from multi-entity joint transactions.

[0107] This invention also provides a multi-entity joint trading strategy optimization system for regional integrated energy systems, comprising the following core modules:

[0108] Measurement and data acquisition module: Establishes a multi-entity, multi-source measurement and data acquisition system to calculate the dispatchable capacity of each entity in time period t; wherein the entities include HVAC groups, energy storage systems, electric vehicle clusters, and gas equipment;

[0109] Cost modeling module: Based on the schedulable capacity, construct component cost functions; wherein the component cost functions include fuel and emissions, user comfort / output loss, equipment life depreciation, and rebound costs;

[0110] Quotation generation module: Based on the itemized cost function, generates a piecewise increasing quotation curve that meets energy and network security constraints;

[0111] Clearing and game theory module: Based on the quoted price curve, construct a non-cooperative game model between aggregators and responders to determine the incentive price and optimal demand response configuration for a unified time period;

[0112] Revenue sharing module: Based on the unified time-period incentive price and optimal demand response configuration, it calculates the system's transaction revenue and overall cost savings, and uses the Shapley value algorithm to fairly distribute revenue according to marginal contribution, thereby optimizing the multi-entity joint transaction strategy.

[0113] Optionally, the measurement and acquisition module can collect the operating status and boundary condition data of flexible resources such as air conditioning, electric vehicles, energy storage systems, and gas equipment in the integrated energy system by establishing a multi-source measurement system across electricity, heat, gas and the terminal side, and construct a schedulable capacity library for each entity in time period t based on this data.

[0114] Optionally, based on the acquired multi-entity schedulable capacity library, a component cost function is constructed that includes fuel and emissions, user comfort / output loss, equipment life depreciation, and rebound costs, in order to establish the cost structure for each entity during a given time period. The total cost model is used to calculate the corresponding marginal cost curve.

[0115] The present invention also provides a computer-readable storage medium having program instructions stored thereon, which, when executed by a processor, cause the processor to perform the method described above.

[0116] The present invention has the following beneficial effects:

[0117] This invention provides an innovative solution for integrated energy management in industrial parks / communities through refined modeling and game theory optimization. Within an electrothermal synergy framework, it quantifies the adjustable potential and control costs of key components such as air conditioning, energy storage, and electric vehicles, ensuring that demand response balances comfort and equipment safety. A Shapley value-based revenue distribution mechanism fairly reflects the contributions of various resources, significantly enhancing user participation. The technology can reduce energy costs, and its unified clearing method and collaborative control logic can be extended to diverse scenarios such as industry and commerce, guiding users to adjust energy consumption during off-peak hours, reducing energy costs and waste, and providing a standardized tool for distributed resource aggregation under new power systems. Attached Figure Description

[0118] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0119] The principles and features of the present invention will be described in detail below with reference to the accompanying drawings. The embodiments given are only for illustrating the technical solutions of the present invention and are not intended to limit the scope of protection.

[0120] This example uses a multi-entity joint transaction of an integrated energy system in a certain industrial park as an example. The park includes an office building (HVAC load shaving), a centralized EV charging station, an electrochemical energy storage (ES) system, and a CHP+ gas boiler. The heat side is used for measuring rebound and comfort assurance costs. The evaluation period is 18:00-18:15 (settlement interval). ), System DR Target .

[0121] S1: Establish a multi-entity, multi-source measurement and ledger collection system to calculate the dispatchable capacity of each entity in time period t; wherein the entities include HVAC groups, energy storage systems, electric vehicle clusters, and gas equipment.

[0122] The parameters collected based on the subject's multi-source measurement and ledger collection system are as follows:

[0123] HVAC: ; ; ; ; ; .

[0124] EV: ; ; ; ; ; .

[0125] ES: , , , .

[0126] Gas side: ; .

[0127] The schedulable capacity calculation process is as follows:

[0128] HVAC dispatchable capacity

[0129] ;

[0130] EV aggregate schedulable capacity

[0131] ;

[0132] ES schedulable capacity

[0133] ;

[0134] In summary, the dispatchable capacity on the power side during this period is as follows:

[0135] ;

[0136] Gas equipment The thermal capability can be calculated according to the formula in the claims, but this embodiment mainly uses it for... Fuel / emissions and rebound costs in Quantization, not directly included in the total capacity of the electrical-side DR.

[0137] S2: Based on the schedulable capacity, construct a component cost function; wherein the component cost function includes fuel and emissions, user comfort / output loss, equipment life depreciation, and rebound cost;

[0138] Based on the multi-entity schedulable capacity library obtained in step S1 We construct component cost functions that include fuel and emissions costs, user comfort / damage costs, equipment life depreciation costs, and rebound costs, in order to establish the cost structure for each entity over a given period. The total cost model is used, and the corresponding marginal cost curve is calculated to provide the basic input for the subsequent pricing function.

[0139] Read the following input variables:

[0140] Capacity parameters: the schedulable capacity of each entity ;

[0141] State variable: Indoor temperature Set temperature Permissible temperature difference State of charge Health ;

[0142] Equipment and price parameters:

[0143] Rated power Rated energy storage capacity Gas prices Emission coefficient Heat source efficiency carbon price Etc. Let. As the main body in time period If the power reduction (MW) is calculated, then this variable will be used as an input independent variable in each component cost function.

[0144] Example values: ; ; ; ; ; ; ; ; .

[0145] Construct the total cost function:

[0146] main body The total cost function is defined as:

[0147] ;

[0148] Comfort / Production Loss Costs:

[0149] ;

[0150] Characterizing the cost of reduced indoor temperature or decreased production comfort due to reduced air conditioning load, among which Temperature deviation weight (typically taken in HVAC) Yuan ); The power is adjusted with a second-order weight to maintain the convexity of the curve (typically taking...). Yuan ).

[0151] Cost of depreciation over life:

[0152] ;

[0153] in This refers to the depreciation factor per unit energy flux, for example, 30 yuan for energy storage (ES). EV is 50 yuan ; For control / settlement intervals, this embodiment uses 15 minutes (0.25 hours). This reflects the lifespan loss of the energy storage device due to charge-discharge cycles.

[0154] Fuel and emissions costs

[0155] ;

[0156] in This refers to the unit price of gas; This represents the equivalent fuel flux compensated on the thermal side due to the reduction in load on the electrical side during this period. For heat source or unit efficiency (take) ); Carbon price (yuan / ton) ); Emission factor (tons) ).

[0157] Rebound penalty cost

[0158] ;

[0159] in For rebound secondary weight (typical) Yuan / ), used to punish rebound behavior after a cut.

[0160] Power reduction Taking the partial derivative, we obtain the main body. Marginal cost curve:

[0161] ;

[0162] Right now:

[0163] ;

[0164] Require In the interval The upper bound is monotonically non-decreasing to ensure economic rationality and the non-decreasing nature of subsequent price functions. If a non-convex phenomenon occurs, it is necessary to increase the upper bound. or The lower bound of the value, or smoothing of the temperature deviation term to make it convex. .

[0165] Marginal cost calculation example (this embodiment)

[0166] Taking HVAC, ES, and EV as examples, and using typical parameters (Table 1), the marginal cost curves for the three types of entities are calculated:

[0167]

[0168] It is evident that HVAC costs are the lowest and rise the fastest, ES costs are moderate and relatively smooth, and EV costs are high and only participate when there is a high incentive price.

[0169]

[0170] Results output and connection to the next step

[0171] Output the marginal cost curves of each entity. Corresponding capacity constraints (From S1); Initial Marginal Cost ; curve slope These results serve as input to the S3 pricing function to determine the starting price for the bid. With slope parameter .

[0172] S3: Based on the component cost function of S2, generate a piecewise increasing price curve that meets the constraints of energy and network security.

[0173] Based on the marginal cost curves of each entity obtained in step S2 Furthermore, by combining capacity constraints and risk preference parameters, a piecewise increasing bidding function that satisfies energy and network constraints is constructed to realize the economic bidding modeling of each responding entity.

[0174] Extract the main body from the cost modeling results of step S2. During the period Marginal cost starting point value and the upper limit of schedulable capacity At the same time, a risk parameter is introduced. Risk aversion parameter. Reflects the sensitivity of the subject to price fluctuations; price stability parameters This reflects the tolerance for price deviations.

[0175] In this embodiment:

[0176] HVAC: ;

[0177] ES: ;

[0178] EV: .

[0179] Depend on Derivation of the marginal cost function:

[0180] ;

[0181] This is used as the benchmark for the pricing function, followed by the addition of a risk premium term. ,in This represents the standard deviation of historical quotes.

[0182] Construction Main Body Based on the capacity limit and the starting point of marginal cost, generate a quantity-price linear quotation function:

[0183] ;

[0184] in Price per unit of unloaded load (RMB / MWh); This is the starting price for the quote; This represents the slope of the quote.

[0185] in, From the starting point of the marginal cost curve mentioned above, This is a risk premium item, used to reflect the entity's expected risk compensation; slope The weight parameters of the quadratic term in S2 and risk buffer items Jointly determined:

[0186] ;

[0187] It reflects the increase in price as the power reduction changes.

[0188] in,

[0189] ;

[0190] For risk preference parameters; A higher value indicates a higher degree of risk aversion. The standard deviation of its expected revenue from load shedding services (RMB / MWh); low volatility ≈ 200 - 800 yuan / MWh corresponds to demand response projects with fixed-price compensation or long-term contracts, offering very stable and predictable returns; with moderate volatility. The price of approximately 800-1800 yuan / MWh corresponds to the day-ahead market or peak-shaving ancillary service market, and the price fluctuates to some extent depending on the supply and demand of the system; high volatility. ≈1800 - 3000+ yuan / MWh corresponds to a real-time balancing market or emergency demand response, where prices may fluctuate drastically in a very short period of time, or even reach extremely high prices;

[0191] In this embodiment, the linear pricing function parameters for the three types of entities are calculated as shown in Table 2. HVAC has the lowest starting price (6.25 yuan / MWh) but the highest slope (200 yuan / MW), reflecting its flexible response but sensitivity to adjustment costs and smallest capacity. ES has a moderate starting price (16.15 yuan / MWh) and a gentle slope (100 yuan / MW), possessing core adjustment capabilities (0.2575MW capacity). EV has an extremely high starting price (83.16 yuan / MWh), a high economic threshold, and only participates under high incentives (0.21MW capacity). ES, with its large capacity and cost advantages, becomes the main response unit.

[0192]

[0193] Setting quotation constraints:

[0194] To ensure historical consistency and network security in quotations, the quotation parameters must meet the following requirements:

[0195] ;

[0196] in as the main body Average price in historical transactions; The standard deviation of the quoted price (8.0 yuan / MWh in this example); The maximum allowable price slope (typically 120 yuan / ). ).

[0197] In this embodiment, all entities satisfy the constraint condition, with ES pricing being relatively smooth and HVAC pricing being slightly more sensitive, but still within the range of... within limits.

[0198] When the entity has a graded response capability (such as segmented charging and discharging power of ES, and grouped control of HVAC groups), based on the monotonicity of the marginal cost curve of S2, in the interval The segmentation points are selected based on the rate of cost change. If the response of a single segment can be approximated as linear, then the number of segments is... If the cost changes significantly with the response rate, then take... In this embodiment, because the HVAC temperature buffer zone is relatively small, it is set... ;ES discharge power is relatively large, divided into two power ranges:

[0199]

[0200] All segments must satisfy the property of non-decreasing. If it appears In cases where the price is non-convex or the monotonicity of the price quote is violated, the secondary weighting coefficient in S2 needs to be reverted. ,make sure Heng was established.

[0201] After feasibility verification, all bids from various entities passed constraint checks, forming the final bid set.

[0202] After constraint testing and convexity adjustment, the main quotation set is formed:

[0203] ;

[0204] in:

[0205] ;

[0206] This curve serves as the input to the S4 aggregator-responder non-cooperative game model, enabling the determination of market clearing and incentive prices.

[0207] S4: Based on the price curve in S3, construct a non-cooperative game model between aggregators and responders to determine the incentive price and optimal demand response configuration for a unified time period.

[0208] The aggregator and the respondent constitute a typical price-follower game model, in which the aggregator, as the dominant party, is responsible for determining the uniform incentive price of the system. Each responding entity (HVAC, ES, EV) is a follower, making the optimal load reduction decision based on the quoted parameters and its own adjustable capacity.

[0209] (1) Optimization objective of the response subject

[0210] Given an incentive price Below, each responding entity The goal is to maximize its own gains, that is, to strike a balance between incentive revenue and response costs. Its optimization problem can be formulated as:

[0211] ;

[0212] (2) Solving for the optimal response by combining the pricing function

[0213] From the quote function in step S3:

[0214] ;

[0215] The optimal solution satisfies the first-order optimality (KKT) condition:

[0216] ;

[0217] Right now:

[0218] ;

[0219] The optimal response function of the subject is obtained:

[0220] ;

[0221] (3) Description of the characteristics of the optimal response behavior

[0222] According to the above relationship, when When the incentive price is lower than the initial marginal price, the entity does not participate in the response; The subject responds in a linear fashion; when The time response is limited by the upper limit of the adjustable capacity.

[0223] This reflects the dynamic change pattern of the response subject from not participating in a partial response to a full-capacity response.

[0224] (4) Aggregator optimization and clearing conditions

[0225] Aggregators need to select the optimal incentive price under the constraint of the system's target load reduction. To minimize the system's payment costs:

[0226] ;

[0227] Satisfy the system's total load reduction constraint:

[0228] ;

[0229] in Reduce the load to achieve system objectives.

[0230] Because each All about Since it is a monotonically non-decreasing function, the above equation has a unique monotonically increasing solution, which can be obtained by numerical iteration.

[0231] (5) Numerical substitution and calculation in the example

[0232] In this embodiment, step S3 has already obtained the quotation parameters and the system target reduction amount. The response functions of each subject are:

[0233] ;

[0234] It is evident that EV has the highest initial marginal price and only participates when the incentive price is high; ES has moderate cost and is the main target for reduction; HVAC has low cost but small capacity and is the auxiliary response target.

[0235] (6) Solve for the uniform incentive price

[0236] System constraints:

[0237] ;

[0238] Solve using binary iterative method:

[0239] when Total Response ;

[0240] when Total Response ;

[0241] when Total Response .

[0242] Therefore, the clearing incentive price is:

[0243] ;

[0244] (7) Optimal Demand Response Configuration Results

[0245] Substitution The optimal response values ​​of each subject are obtained.

[0246] HVAC: The limit is 0.0425.

[0247] ES: ;

[0248] EV: ;

[0249] ;

[0250] The system has met its target load reduction target and the clearing process is complete.

[0251] (8) Analysis of the characteristics of clearing results

[0252] The pricing result is a uniform incentive price. The response structure is dominated by ES. HVAC provides remaining EVs do not participate. Market characteristics include consistent marginal costs among responders, achieving a local Nash equilibrium; minimized system payment costs; and a clearing outcome satisfying monotonicity and stability. This clearing process simultaneously satisfies the minimization of aggregator payment costs and the system load reduction objective, ensuring the Nash equilibrium of the game and the dynamic stability of market operation.

[0253] S5: Based on the unified time-period incentive price and optimal demand response configuration derived from S4, calculate the system transaction revenue and overall cost savings, and use the Shapley value algorithm to fairly distribute the revenue according to the marginal contribution, thereby optimizing the multi-entity joint transaction strategy.

[0254] Based on the uniform incentive price obtained in step S4 Optimal demand response results of each entity The system calculates the transaction revenue and system cost savings of each entity, and uses the Shapley value algorithm for fair allocation to ensure the fairness of revenue distribution and incentive compatibility among multiple entities.

[0255] (1) Revenue settlement calculation

[0256] Calculate the time period for each entity based on the measurement data. Actual load reduction:

[0257] ;

[0258] in The baseline power (MW) established for step S1; The actual measured power (MW).

[0259] Combined with step S4 clearing incentive price Calculation subject Settlement and payment amount:

[0260] ;

[0261] in For the settlement interval, this embodiment takes... ; The performance coefficient determined by the response execution accuracy .

[0262] The revenue settlement for the three main entities was calculated using the example, as shown in Table 3. The HVAC system precisely completed a 0.0425MW load reduction, generating a revenue of 0.4452 yuan; the ES system, as the main contributor, achieved 85.8% load reduction, generating a revenue of 2.6973 yuan; the EV system did not participate due to price barriers. The total expenditure of 3.1425 yuan achieved the 100% system response target. The ES system's 85.8% contribution validates its core scheduling value, demonstrating optimal cost-effectiveness.

[0263]

[0264] (2) Quantification of system savings

[0265] Based on the total system incentive cost obtained in step S4 and the baseline system operating cost, calculate the system savings:

[0266] ;

[0267] in The total system benefit with the participation of all stakeholders; This represents the baseline operating cost when there is no demand response.

[0268] Example calculation of electricity purchase cost without DR:

[0269] ;

[0270] ;

[0271] That is, the system saved 1.3575 yuan during this period.

[0272] (3) Calculation of Alliance Marginal Contribution

[0273] definition as a subset of the main body The total benefit of the system upon participation. For any given entity. Its alliance The marginal contribution is:

[0274] ;

[0275] This embodiment only considers HVAC and ES as the main response alliance, and EV does not participate.

[0276] When HVAC is involved alone ( ): The capacity is only 0.0425MW, and the target of 0.3MW cannot be achieved, so the revenue is recorded as 0.

[0277] When ES participates alone : Reduced by 0.2575MW, resulting in savings in electricity purchase costs .

[0278] When all participate Save ¥1.3575.

[0279] ;

[0280] ;

[0281] ;

[0282] ;

[0283] (4) Calculation of fair distribution of income

[0284] Based on the marginal contribution of each entity, the Shapley value method is used for fair allocation:

[0285] ;

[0286] in as the main body In the league The weight of fair benefit allocation in the process.

[0287] Example calculation:

[0288] ;

[0289] ;

[0290] ;

[0291] check: .

[0292] (5) Final Profit Output

[0293] Each subject The final profit is:

[0294] ;

[0295] in The allocation weighting coefficient (set to 1.0) is used to balance immediate revenue with long-term fair distribution. The calculation results after substituting the examples are shown in Table 4. ES receives 3.8954 yuan in revenue, accounting for 78.4% of the total revenue, reflecting its core contribution value; HVAC receives 0.6046 yuan, reflecting its auxiliary role; EV receives zero revenue due to not participating. The total revenue of 1.3575 yuan allocated by Shapley perfectly matches the system savings, verifying the fairness and economic rationality of the allocation mechanism and effectively incentivizing all stakeholders to actively participate in demand response.

[0296]

[0297] ES receives approximately 86% of the revenue due to undertaking the main reduction; HVAC receives approximately 14% due to its auxiliary response; EV does not participate. Shapley allocation is dynamically adjusted based on marginal contribution, ensuring that even smaller HVAC units receive positive revenue, avoiding a "zero incentive" situation. This guarantees both immediate economic incentives and long-term cooperative stability, preventing entities from leaving the alliance due to a single low return. The algorithm drives revenue settlement through the price-response result output by S4 and combines it with the Shapley value method to quantify each entity's contribution to the alliance's revenue, ensuring fairness and compatibility of transaction incentives and achieving optimal revenue allocation for multi-entity joint transactions.

[0298] In summary, this invention proposes a multi-stakeholder joint trading strategy optimization method for regional integrated energy systems. By constructing a multi-source data acquisition system spanning electricity, heat, and gas, implementing component cost modeling, quantity-price segmented bidding, non-cooperative game clearing, and a Shapley value-based fair revenue sharing mechanism, it achieves collaborative optimization among multiple stakeholders within the same trading framework. This method significantly improves the flexible adjustment capability and market response efficiency of integrated energy systems while ensuring system energy and security constraints. It enables unified incentive price determination, optimal demand response configuration, and fair revenue distribution under various scenarios, including load fluctuations, uncertain renewable energy output, and dynamic price signal changes.

[0299] Compared to traditional demand response mechanisms based on single entities, fixed incentives, or static pricing, this invention, grounded in game theory and marginal cost theory, balances economic efficiency and fairness. It establishes a complete chain from physical-layer adjustable capacity calculation to market-layer revenue sharing, providing a replicable strategy model for rolling optimization scheduling, two-way market clearing, and multi-energy complementary operation of regional integrated energy systems. This method effectively promotes local consumption of renewable energy and efficient aggregation of flexible resources, exhibiting good scalability and promising engineering applications.

[0300] This invention designs a multi-entity joint trading strategy optimization system for regional integrated energy systems, comprising the following core modules:

[0301] Measurement and data acquisition module: Establishes a multi-entity, multi-source measurement and data acquisition system to calculate the dispatchable capacity of each entity in time period t; wherein the entities include HVAC groups, energy storage systems, electric vehicle clusters, and gas equipment;

[0302] Cost modeling module: Based on the schedulable capacity, construct component cost functions; wherein the component cost functions include fuel and emissions, user comfort / output loss, equipment life depreciation, and rebound costs;

[0303] Quotation generation module: Based on the itemized cost function, generates a piecewise increasing quotation curve that meets energy and network security constraints;

[0304] Clearing and game theory module: Based on the quoted price curve, construct a non-cooperative game model between aggregators and responders to determine the incentive price and optimal demand response configuration for a unified time period;

[0305] Revenue sharing module: Based on the unified time-period incentive price and optimal demand response configuration, it calculates the system's transaction revenue and overall cost savings, and uses the Shapley value algorithm to fairly distribute revenue according to marginal contribution, thereby optimizing the multi-entity joint transaction strategy.

[0306] As an improvement to the solution, the measurement and acquisition module establishes a multi-source measurement system spanning electricity, heat, gas, and the terminal side to collect data on the operating status and boundary conditions of flexible resources such as air conditioning, electric vehicles, energy storage systems, and gas equipment in the integrated energy system, and constructs a schedulable capacity database for each entity in time period t based on this data.

[0307] As an improvement to the solution, the cost modeling module, based on the acquired multi-entity schedulable capacity library, constructs a component cost function that includes fuel and emissions, user comfort / output loss, equipment lifespan depreciation, and rebound costs, in order to establish the cost functions for each entity during different time periods. The total cost model is used to calculate the corresponding marginal cost curve.

[0308] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of this application can be implemented in various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0309] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these changes and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for optimizing multi-entity joint trading strategies in a regional integrated energy system, characterized in that, include: Establish a multi-entity, multi-source measurement and record-keeping system to calculate the schedulable capacity of each entity in time period t; The main components include HVAC systems, energy storage systems, electric vehicle clusters, and gas equipment; Based on the schedulable capacity, a component cost function is constructed, wherein the component cost function includes fuel and emissions, user comfort / output loss, equipment life depreciation, and rebound cost; Based on the itemized cost function, a piecewise increasing price curve that meets energy and network security constraints is generated; Based on the quoted price curve, a non-cooperative game model of aggregator-responder is constructed to determine the incentive price and optimal demand response configuration for a unified time period. Based on the derived unified time-period incentive price and optimal demand response configuration, the system's transaction revenue and overall cost savings are calculated. The Shapley value algorithm is then used to fairly distribute the revenue according to marginal contribution, thereby optimizing the multi-entity joint transaction strategy.

2. The method according to claim 1, characterized in that, The specific process for calculating the schedulable capacity of multi-entity, multi-source measurements is as follows: By establishing a multi-source measurement system spanning electricity, heat, gas, and the terminal side, operational status and boundary condition data of air conditioning (HVAC), electric vehicles (EV), energy storage systems (ES), and flexible resources of gas equipment in the integrated energy system are collected, and based on this, a database of the dispatchable capacity of each entity in time period t is constructed. Specifically: Multi-source measurement data are collected through power monitoring, terminal control, and environmental sensing equipment; among which, Power side: via Collect household / building level active power Baseline power Node voltage The time resolution is 5 minutes; Heating, ventilation and air conditioning (HVAC): Indoor temperature is collected through building temperature control terminals. Set temperature Permissible temperature difference Rated power Thermal inertia time constant Switch status ;in Usually taken , Typical – min; Electric Vehicles (EVs): On-site connection data is collected through the pile-side / BMS / station control system. Average state of charge Single vehicle battery capacity Health Station-level grid-connected power limit The time resolution is 5 minutes; Energy storage system (ES): through Collect state of charge Health Maximum charge and discharge power , With rated energy capacity The time resolution is 5 minutes; Hot air side: Regional heat load is collected through flow meters at heat exchange stations or gate stations. Maximum thermal power of gas boiler and CHP coupling coefficient ;in For constant values: back-pressure steam turbine , Gas-fired internal combustion engine , ; Based on the collected multi-source measurement data and the physical parameters of various devices, the schedulable capacity of each type of entity is calculated: among which, 1) Peak-shaving capacity of HVAC systems: ； in The control interval is set at 5 minutes. Indicates the air conditioner's on / off status; 2) Dispatchable capacity of electric vehicle aggregation: ； in This represents the participation rate (typically 0.7%). The target upper limit is 0.8; 3) Available load shedding capacity of the energy storage system: ； in The minimum permissible state of charge is 0; 4) The adjustable capacity of the gas-fired boiler equipment is: ； in, The adjustable capacity of the gas-fired equipment at t; This represents the boiler's maximum thermal power. This represents the current electrical power output of the CHP. , The electrothermal coupling coefficient; This refers to the calorific value of the fuel. For boiler efficiency; Based on the adjustable capacity of the above entities, a comprehensive schedulable capacity pool is formed; ； This capacity library As input for subsequent cost modeling, i represents the subject index, and T represents the transaction time period. The maximum demand response schedulable capacity (MW) that entity i can provide in time period t provides the physical basis for subsequent marginal cost and quotation generation.

3. The method according to claim 1, characterized in that, The specific process for calculating the itemized cost function is as follows: Based on the acquired multi-entity schedulable capacity library We construct component cost functions that include fuel and emissions, user comfort / output loss, equipment life depreciation, and rebound costs, in order to establish the cost structure for each entity over a given period. The total cost model is used to calculate the corresponding marginal cost curve, specifically: Read the schedulable capacity of each entity Indoor temperature Set temperature Permissible temperature difference State of charge Health ; Equipment and pricing parameters, including rated power Rated energy storage capacity Gas prices Emission coefficient Heat source efficiency ; set up As the main body in time period The amount of power reduction; Construct the total cost function: where, main body The total cost function is defined as: ； ； ； ； ； in Temperature deviation weighting; Indoor temperature; Set the temperature; To allow for temperature difference (typical value) ); It is a power adjustment secondary weight; To reduce power; The depreciation factor per unit energy flux; Energy flux (MWh) for this period; To control / settlement interval (set to 15 minutes); Health score (typically 0.92); This refers to the unit price of gas; This is the equivalent heat demand or fuel flux for this period, used to quantify the additional fuel required for heat-side compensation due to power-side load shedding. For heat source or unit efficiency, the value ranges from 0 to 1; For carbon price; Emission factor (tons) ); To rebound the secondary weighting; right Taking the partial derivative, we obtain the main body. Marginal cost curve: ； And require The upper bound remains monotonically constant to ensure economic rationality and the non-decreasing nature of the subsequent price curve; if this is not satisfied, the lower bound of the secondary weight is increased (e.g., by increasing the weighting). or (and / or smooth the temperature deviation term) to convexity ; Output the marginal cost curves of each entity. With capacity constraints and will The curve slope information is used as input to the quotation function to determine the starting price. With slope parameter .

4. The method according to claim 3, characterized in that, The specific process of the segmented incremental price curve algorithm is as follows: Based on the marginal cost curves of each entity obtained above Furthermore, by combining its capacity constraints and risk preference parameters, a piecewise increasing bid function that satisfies both energy and network constraints is constructed; specifically: Extract the subject from the cost function modeling results. During the period Marginal cost starting point value and the upper limit of schedulable capacity Simultaneously read risk preference parameters With price stability parameters , Reflects the degree of risk aversion; It reflects the tolerance for price deviation and is typically used to control the deviation between quoted prices and historical prices; These parameters can be determined through contractual agreement or historical statistics; Based on the obtained marginal cost starting point and capacity limit, construct the quantity-price linear quotation function for entity i: ； in, Price per unit of unloaded load (RMB / MWh); The starting price for the bid is defined as: ； in, From the starting point of the marginal cost curve mentioned above, This is a risk surcharge item used to reflect the entity's expected compensation for risk; The slope of the quote is based on the weighting parameters of the quadratic term mentioned above ( , ) and risk buffer The calculation yields a value used to describe the increase in price as the load reduction increases; where, ； For risk preference parameters; A higher value indicates a higher degree of risk aversion. The standard deviation of its expected revenue from load shedding services (RMB / MWh); low volatility ≈ 200 - 800 yuan / MWh corresponds to demand response projects with fixed-price compensation or long-term contracts, offering very stable and predictable returns; with moderate volatility. The price ranges from approximately 800 to 1800 yuan / MWh, corresponding to the day-ahead market or peak-shaving ancillary service market. Prices fluctuate to some extent depending on system supply and demand; high volatility. ≈ 1800 -3000+ yuan / MWh corresponds to a real-time balancing market or emergency demand response, where prices may fluctuate drastically in a very short period of time, or even reach extremely high prices; The pricing constraints for the quantity-price linear pricing function are set as follows: ； in, as the main body The average starting price in historical transactions; The standard deviation of historical quotes; This refers to the risk tolerance coefficient. The maximum allowed bid slope; After obtaining a linear pricing function that satisfies the constraints, when the entity has graded response capabilities or different adjustable ranges, the marginal cost function can be used as a basis. For the adjustment range Perform piecewise fitting to generate piecewise linear price segments: ； in The price quote is divided into segments, with the segment points selected based on the cost change rate; each segment must meet the non-decreasing condition. ; If it appears If the non-convexity or the monotonicity of the price is violated, then the quadratic weighting coefficient of the pair ( , Perform convexity adjustment to make The constant validity ensures that the price curve monotonically increases; After constraint verification and segmented pricing, the feasible pricing points of each entity are compared. By summarizing, we obtain the set of price curves for time period t: ； in, The amount of regulating power (MW) that subject i can provide in time period t. To adjust the unit price (RMB / MWh) for this power level, point-to-point This represents the pricing intentions of the subjects at different levels of adjustment, and the set The bid collection comprises all resource entities participating in the bidding. This forms the complete market quote input for time period t.

5. The method according to claim 4, characterized in that, The non-cooperative game model of the aggregator-responder entity is as follows: The aggregator and the responders constitute a typical price-volume game model, in which the aggregator is the dominant party, responsible for determining the uniform incentive price of the system. The responding entity is the follower, making the optimal load reduction decision based on the quoted parameters and its own adjustable capacity; Given an incentive price Below, each responding entity The goal is to maximize its own benefits, that is, to achieve a balance between benefits and costs; its optimization problem can be expressed as: ; Combined with quotation function Differentiating the above equation and satisfying the KKT conditions, the optimal response of the subject is obtained as follows: ; That is, when the incentive price Below the initial marginal price When the incentive price is in effect, the main body does not participate in the response; when the incentive price is in effect... and When the incentive price is between these values, the main body responds according to a linear function; when the incentive price is too high, it is subject to the capacity limit. limit; Aggregators need to reduce the system load by a certain amount. Under constraints, select the optimal incentive price. To minimize system payment costs: ; Simultaneously satisfying system objective constraints: ; This constraint reflects the condition where the sum of the optimal responses of all agents exactly satisfies the system load reduction target. This is the clearing incentive price; because each All about Since it is a monotonically non-decreasing function, the above equation has a unique monotonically increasing solution; the clearing price can be obtained through numerical iteration (such as the bisection method or Newton's iteration). ; When clearing price After obtaining the result, substitute it into the follower response function to obtain the optimal demand response configuration result for each subject: ； At this point, the sum of the load reductions of all responding entities satisfies the system objective: ； in, The incentive price is set at a uniform time period (RMB / MWh). as the main body During the period The optimal load reduction decision result (MW); This clearing process simultaneously satisfies the goals of minimizing aggregator payment costs and reducing system load, while also ensuring the Nash equilibrium of the game and market stability.

6. The method according to claim 1, characterized in that, The specific process of the profit distribution algorithm is as follows: Based on the obtained uniform incentive price... Optimal demand response results of each entity The transaction revenue and system cost savings of each entity are calculated, and the Shapley value algorithm is used for fair allocation. Specifically: Calculate the time period for each subject based on the measurement data. Actual load reduction: ； in, Baseline power, For actual measured power; Combining the above-mentioned clearing incentive price Calculation subject Settlement and payment amount: ； in, For settlement interval, The performance coefficient determined by the accuracy of response execution ( ); Calculate the system savings based on the obtained total system incentive cost and the baseline system operating cost: ; in, This represents the total system benefit with the participation of all stakeholders. This refers to the system operating cost when there is no demand response. definition as a subset of the main body The total benefit of the system upon participation; for any subject Its alliance The marginal contribution is: ; in, The calculations are the same as above, both based on S4. and Obtained again; Based on the marginal contribution of each entity, the Shapley value method is used for fair allocation: ； in, as the main body In the league The weight of fair benefit allocation in the process; Each subject The final profit is: ； in, The apportionment weighting coefficient (take) ( ), used to balance immediate gains with long-term equitable distribution.

7. A multi-entity joint trading strategy optimization system for a regional integrated energy system, characterized in that, include: Measurement and data acquisition module: Establish a multi-entity, multi-source measurement and data acquisition system, and calculate the schedulable capacity of each entity in time period t; The main components include HVAC systems, energy storage systems, electric vehicle clusters, and gas equipment; Cost modeling module: Based on the schedulable capacity, construct component cost functions; wherein the component cost functions include fuel and emissions, user comfort / output loss, equipment life depreciation, and rebound costs; Quotation generation module: Based on the itemized cost function, generates a piecewise increasing quotation curve that meets energy and network security constraints; Clearing and game theory module: Based on the quoted price curve, construct a non-cooperative game model between aggregators and responders to determine the incentive price and optimal demand response configuration for a unified time period; Revenue sharing module: Based on the unified time-period incentive price and optimal demand response configuration, it calculates the system's transaction revenue and overall cost savings, and uses the Shapley value algorithm to fairly distribute revenue according to marginal contribution, thereby optimizing the multi-entity joint transaction strategy.

8. The system as described in claim 7, characterized in that: The measurement and acquisition module establishes a multi-source measurement system spanning electricity, heat, gas, and the terminal side to collect data on the operating status and boundary conditions of flexible resources such as air conditioning, electric vehicles, energy storage systems, and gas equipment in the integrated energy system, and constructs a schedulable capacity database for each entity in time period t based on this data.

9. The system as described in claim 7, characterized in that: The cost modeling module, based on the acquired multi-entity schedulable capacity library, constructs component cost functions including fuel and emissions, user comfort / output loss, equipment lifespan depreciation, and rebound costs, to establish the cost structure for each entity within a given time period. The total cost model is used to calculate the corresponding marginal cost curve.

10. A computer-readable storage medium having stored thereon program instructions that, when executed by a processor, cause the processor to perform the method according to any one of claims 1-6.

Images