Building flexible resource load optimization scheduling method and system based on game mechanism

Abstract

The invention belongs to the technical field of power system resource scheduling, and provides a building flexible resource load optimal scheduling method and system based on a game mechanism, and the method comprises the steps: obtaining a distributed architecture of a building flexible resource load; performing department-level and building-level layered modeling on the acquired distributed architecture to obtain a department-level model and a building-level model; carrying out distributed optimization solution on the constructed department-level model and building-level model by adopting an alternating direction multiplier method to obtain a distributed collaborative model; considering an aggregator, and constructing a single-leader-multi-follower game model which takes the transaction center as a dominant and takes the aggregator and a power generator as followers; and solving the constructed game model, and completing optimal scheduling of the building flexible resource load based on the game mechanism.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power system resource scheduling, and particularly relates to a building flexible resource load optimization scheduling method and system based on a game mechanism. BACKGROUND

[0002] The statements in this section merely provide background information related to the application and do not necessarily constitute prior art.

[0003] With the transformation of global energy, the energy management of self-owned buildings is facing a deep transformation from single energy saving to intelligence and flexible scheduling. In particular, in large-scale scenarios, there are a large number of adjustable power resources in buildings, such as air conditioning systems, lighting equipment, energy storage units and distributed renewable energy sources, etc. Among them, the air conditioning system as a typical thermostatically controlled load (TCL) not only has a high proportion of power consumption, but also has natural load adjustability, which has become an important entry point for realizing building flexible load management.

[0004] With the rapid development of new power systems and the continuous improvement of new energy penetration rate, the traditional power system with centralized supply as the main operation mode is facing increasingly severe regulation pressure. Aggregators, as an important bridge between power users and the market, gradually become a key carrier to stimulate user-side regulation potential and improve demand response capability. Through aggregation of adjustable loads and sharing of distributed resources such as energy storage, they realize large-scale management and have the ability to participate in unified bidding and centralized bidding in the spot market.

[0005] Currently, to improve the energy efficiency level and renewable energy consumption capacity of buildings, the aggregation and scheduling of flexible resources have gradually attracted attention. However, the existing scheduling technology still has the following shortcomings when facing building scenarios: (1) The traditional centralized scheduling method needs to collect a large amount of operation data of each floor and each subsystem, such as the on-off state of air conditioning equipment, temperature set value and actual load, etc. It faces the dual challenges of data privacy leakage and communication pressure, and the algorithm complexity increases rapidly with the increase of equipment quantity, making it difficult to realize real-time optimal regulation and control of large-scale buildings; (2) The existing distributed optimization method for buildings or microgrids often does not fully consider the operation law and thermal inertia characteristics of air conditioning loads in office buildings. For example, office area air conditioning loads have obvious time periods (such as load fluctuations during morning peak and lunch break), and there are user comfort constraints (such as temperature upper and lower limits and maximum change rate). However, the traditional model usually assumes that the load behavior is predictable and linearly responsive, which cannot truly reflect the dynamic feedback characteristics of TCLs in actual operation.

[0006] Meanwhile, the output of renewable energy equipment such as photovoltaic systems and energy storage batteries deployed in the building has obvious volatility and uncertainty; when the external electricity price changes (such as time-of-use electricity price) and the photovoltaic output fluctuation are superimposed, how to coordinate the response behavior of air conditioning load and distributed resources under the premise of ensuring comfort becomes a key problem faced by current flexible resource scheduling. The traditional static strategy or control method based on experience rules is difficult to adapt to such dynamic changes, often leading to failure of scheduling plan or unstable operation of equipment, limiting the economic and regulation potential of flexible load.

[0007] In the prior art, the aggregator can flexibly adjust the aggregated adjustable resources according to market demand: on the one hand, the aggregator can participate in demand response, intelligently schedule controllable load and energy storage resources, and realize load peak clipping and valley filling and peak regulation coordination according to electricity price fluctuation; on the other hand, in the electricity spot market, the aggregator can sell electricity during the peak period of electricity demand and buy electricity during the valley period to realize electricity arbitrage and obtain economic benefits.

[0008] Therefore, how to construct a flexible resource aggregation and scheduling framework suitable for the characteristics of the own building, while protecting the privacy of each department load, fully tap the regulation capacity of air conditioning and other loads, and establish a high-precision and low-complexity collaborative modeling and optimization mechanism, is of great significance for the aggregator to construct a set of aggregation and scheduling optimization methods for the aggregator bidding strategy in the spot market, realize coordinated control of the energy system in an uncertain environment, balance the game relationship between the supply and demand sides, reduce the electricity purchase cost under the premise of ensuring electricity acquisition, and improve the market adaptability and economy of the aggregator. SUMMARY

[0009] To solve the above problems, the present application provides a building flexible resource load optimization scheduling method and system based on a game mechanism, which optimizes building flexible resource load scheduling based on a game mechanism, builds a "BEMS-LEMS" distributed architecture, combines an ADMM distributed optimization algorithm, and realizes privacy-friendly scheduling of temperature-controlled loads such as air conditioners; introduces shared energy storage resources and a Stackelberg game mechanism, builds a multi-agent two-stage decision-making model in power trading, and effectively improves the market response ability and bidding strategy optimization level of the load aggregator.

[0010] According to some embodiments, the first aspect of the present application provides a building flexible resource load optimization scheduling method based on a game mechanism, which adopts the following technical scheme: A building flexible resource load optimization scheduling method based on a game mechanism, comprising: obtaining a distributed architecture of building flexible resource load; performing hierarchical modeling of the obtained distributed architecture at the department level and the building level to obtain a department-level model and a building-level model; The constructed department-level model and building-level model are solved by using an alternating direction multiplier method to obtain a distributed collaborative model. A single-leader-multiple-follower game model is constructed by considering aggregators, in which a transaction center is the leader and aggregators and power suppliers are followers. The constructed game model is solved to complete the optimization scheduling of building flexible resource load based on a game mechanism.

[0011] As a further technical limitation, the constructed game model is solved by using a genetic algorithm and a CPLEX solver, specifically: the parameters of the transaction center, the shared energy storage service provider and the aggregator are initialized; the acceptable purchase price of the transaction center is randomly generated by using the genetic algorithm, and the generated acceptable purchase price is passed to the aggregator as a population and iterated; the aggregator receives the market clearing price of all transaction centers, solves the load distribution and the shared energy storage service provider by using the CPLEX solver, calculates the current price income, and returns the purchase information of each period to the transaction center; the power supplier receives the market clearing price of all transaction centers, solves the optimal offer by using the CPLEX solver, calculates the current income, and returns the electricity sale information of each period to the transaction center; the transaction center performs market clearing according to the purchase information, the power supplier information and the offer scenario generated by the Latin hypercube sampling, obtains the market clearing price, calculates and retains the current external cost; the population is selected and mutated based on the genetic algorithm to generate a new generation of market clearing price, and the iteration is solved to calculate the transaction center income and the aggregator income; the transaction center income and the aggregator income obtained are compared, and the iteration is ended when the number of iterations reaches a certain value, and the solution of the game model is completed.

[0012] As a further technical limitation, the constructed game model at least includes a transaction center model, an aggregator model and a power supplier model; the aggregator participates in the spot market game, which includes a pricing decision stage and a quantitative decision stage; in the pricing decision stage, the transaction center as an upper subject performs clearing according to the supply-demand relationship and market information to minimize the external cost of the system, and completes the offer to the aggregator in the process of clearing the price; in the quantitative decision stage, the supply side and the user side as lower subjects respectively determine the optimal power generation and load demand according to the price signal of the transaction center, the optimal decision of the lower subject is a function of the upper decision variable, and the aggregator obtains the offered power according to the optimized load demand.

[0013] As a further technical limitation, in the process of distributed optimization solution, the alternating direction multiplier method (ADMM) is adopted to realize the distributed collaboration of the department-level model and the building-level model. Specifically, the building-level model initializes the Lagrange multiplier and the penalty factor, the department-level model optimizes the start-stop plan of the load based on the local model and the coordination signal of the building-level model, and feeds back the total load of the department to the building-level model; the building-level model updates the multiplier and the penalty factor according to the global constraint, and iterates cyclically until the original residual and the dual residual meet the convergence threshold, and the department privacy is protected through the interaction of aggregated load data; the convex relaxation technology is adopted to relax the binary variable to the continuous variable [0, 1], and the physical feasibility of the solution is enhanced through the covering inequality; after the relaxation solution converges, the iterative threshold rounding technology is used to restore the continuous variable to the binary value step by step, and the minimum start-stop time constraint is met through iterative adjustment, and the distributed optimization solution is completed.

[0014] According to some embodiments, the second aspect of the present application provides a building flexible resource load optimization scheduling system based on a game mechanism, which adopts the following technical scheme: A building flexible resource load optimization scheduling system based on a game mechanism, comprising: An acquisition module configured to acquire a distributed architecture of building flexible resource load; A hierarchical modeling module configured to perform hierarchical modeling of the acquired distributed architecture at the department level and the building level to obtain a department-level model and a building-level model; An optimization module configured to perform distributed optimization solution on the constructed department-level model and building-level model using the alternating direction multiplier method to obtain a distributed collaboration model; A game module configured to consider an aggregator, and construct a single-leader-multiple-follower game model with the transaction center as the leader and the aggregator and the power generator as the followers; A scheduling module configured to solve the constructed game model to complete the optimization scheduling of the building flexible resource load based on the game mechanism.

[0015] As a further technical limitation, in the scheduling model, the constructed game model is solved by a hybrid method of genetic algorithm and CPLEX solver, specifically: the parameters of the transaction center, shared energy storage service provider and aggregator are initialized; the acceptable purchase price of the transaction center is randomly generated by the genetic algorithm, and the generated acceptable purchase price is passed to the aggregator as a population and iterated; the aggregator receives the market clearing price of all transaction centers, solves the load distribution and shared energy storage service provider using the CPLEX solver, calculates the current price revenue, and returns the purchase information of each period to the transaction center; the power generator receives the market clearing price of all transaction centers, solves the optimal offer using the CPLEX solver, calculates the current revenue, and returns the electricity sale information of each period to the transaction center; the transaction center performs market clearing based on the purchase information, power generator information and offer scenarios generated by Latin hypercube sampling, obtains the market clearing price, calculates and retains the current external cost; the population is selected and mutated based on the genetic algorithm to generate a new generation of market clearing price, iteratively solve, and calculate the transaction center revenue and aggregator revenue; the transaction center revenue and aggregator revenue obtained are compared, and when the number of iterations reaches a certain value, the iteration is ended, and the solution of the game model is completed.

[0016] As a further technical limitation, in the game model, the constructed game model at least includes a transaction center model, an aggregator model and a power generator model; the aggregator participating in the spot market game includes a pricing decision stage and a quantitative decision stage; in the pricing decision stage, the transaction center as the upper subject performs clearing based on the supply and demand relationship and market information to minimize the system external cost, and completes the offer to the aggregator in the process of clearing the electricity price; in the quantitative decision stage, the supply side and the user side as the lower subject respectively determine the optimal power generation and load demand according to the price signal of the transaction center, the optimal decision of the lower subject is a function of the upper decision variable, and the aggregator refers to the optimized load demand to obtain the offered power.

[0017] As a further technical limitation, in the optimization module, in the process of distributed optimization solving, the alternating direction multiplier method (ADMM) is used to realize the distributed collaboration of department-level model and building-level model, specifically: the building-level model initializes the Lagrange multiplier and the penalty factor, the department-level model optimizes the start-stop plan of the load based on the local model and the coordination signal of the building-level model, and feeds back the total load of the department to the building-level model; the building-level model updates the multiplier and the penalty factor according to the global constraint, and iterates cyclically until the original residual and the dual residual meet the convergence threshold, the load data is interacted by aggregation to protect the privacy of the department; the convex relaxation technology is used to relax the binary variable to a continuous variable [0, 1], and the physical feasibility of the solution is enhanced through covering inequality; after the relaxation solution converges, the iterative threshold rounding technology is used to restore the continuous variable to a binary value step by step, and the minimum start-stop time constraint is met through iterative adjustment, and the distributed optimization solving is completed.

[0018] According to some embodiments, a third aspect of the present application provides a computer readable storage medium, adopting the technical scheme as follows: A computer readable storage medium, having a program stored thereon, which, when executed by a processor, implements the steps in the building flexible resource load optimization scheduling method based on a game mechanism according to the first aspect of the present application.

[0019] According to some embodiments, a fourth aspect of the present application provides an electronic device, adopting the technical scheme as follows: An electronic device, comprising a memory, a processor, and a program stored in the memory and running on the processor, wherein the processor implements the steps in the building flexible resource load optimization scheduling method based on a game mechanism according to the first aspect of the present application when executing the program.

[0020] According to some embodiments, a fifth aspect of the present application provides a computer program product, adopting the technical scheme as follows: A computer program product, comprising software code, wherein the program in the software code implements the steps in the building flexible resource load optimization scheduling method based on a game mechanism according to the first aspect of the present application.

[0021] Compared with the prior art, the present application has the following beneficial effects: The present application performs building flexible resource load optimization scheduling based on a game mechanism, realizes privacy-friendly scheduling of temperature control loads such as air conditioners by constructing a "BEMS-LEMS" distributed architecture and combining an ADMM distributed optimization algorithm, introduces shared energy storage resources and a Stackelberg game mechanism to construct a multi-agent two-stage decision-making model in power trading, effectively improves the market response capability and pricing strategy optimization level of load aggregators, and significantly improves the schedulability and participation depth of temperature control loads (especially air conditioning systems) in self-owned buildings in demand response scenarios, realizing efficient aggregation and scheduling of flexible resources. Compared with traditional methods, the present application has stronger privacy protection capability, higher convergence efficiency, and better overall energy efficiency performance, is suitable for enterprise headquarters, data centers, office buildings in parks, and other types of commercial building scenarios, and has wide popularization value and technical application prospect. BRIEF DESCRIPTION OF DRAWINGS

[0022] The drawings constituting a part of this embodiment are used to provide a further understanding of the embodiment, and the schematic embodiment and its description are used to explain the embodiment and do not constitute an improper limitation on the embodiment.

[0023] Figure 1 Flowchart of the building flexible resource load optimization scheduling method based on a game mechanism in the first aspect of the present application; Figure 2 A distributed architecture diagram of a building energy management system (BEMS) - department controller (DC) in the embodiment one of the present application; Figure 3 An information exchange diagram of a building energy management system (BEMS) - department controller (DC) in the embodiment one of the present application; Figure 4 An architecture schematic diagram of a game mechanism in the embodiment one of the present application; Figure 5 A flowchart of a game model solution in the embodiment one of the present application; Figure 6 A structure block diagram of a building flexible resource load optimization scheduling system based on a game mechanism in the embodiment two of the present application. DETAILED DESCRIPTION

[0024] The present application will be further described below with reference to the accompanying drawings and embodiments.

[0025] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present application. Unless otherwise indicated, all technical and scientific terms used herein have the same meaning as would be understood by one of ordinary skill in the art to which the present application pertains.

[0026] It should be noted that the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit exemplary embodiments according to the present application. As used herein, the singular form is intended to include the plural form unless the context clearly indicates otherwise, and it should also be understood that when the terms "comprise" and / or "include" are used in the specification, there is a presence of a feature, step, operation, device, component and / or combinations thereof.

[0027] In the present application, the terms such as "upper", "lower", "left", "right", "front", "back", "vertical", "horizontal", "side", "bottom", etc. indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, which is only a relationship word determined for the convenience of describing the structural relationship of the components or elements of the present application, and cannot be understood as a limitation on the present application.

[0028] In the present application, the terms such as "fixedly connected", "connected", "connected" should be understood broadly, which means that it can be fixedly connected, integrally connected or detachably connected; it can be directly connected or indirectly connected through an intermediate medium. For relevant researchers or technicians in the art, the specific meaning of the above terms in the present application can be determined according to the specific circumstances, and cannot be understood as a limitation on the present application.

[0029] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0030] Example 1 Embodiment 1 of this invention introduces a method for optimizing the scheduling of flexible building resource loads based on a game theory mechanism.

[0031] like Figure 1 The method for optimizing the scheduling of flexible building resource loads based on a game theory mechanism, as shown, includes: A distributed architecture for acquiring building flexible resource load; The obtained distributed architecture is modeled in a hierarchical manner at the department and building levels to obtain department-level models and building-level models; The distributed optimization solution of the constructed department-level model and building-level model is obtained by using the alternating direction multiplier method. Considering aggregators, construct a single-leader-multiple-follower game model with the trading center as the leader and aggregators and generators as followers; Solve the constructed game model to achieve optimal scheduling of building flexible resource load based on the game mechanism.

[0032] The specific implementation process of the building flexible resource load optimization scheduling method based on game theory mechanism in this embodiment includes the following steps: Step S01: System architecture setup and data initialization, constructing such a system within the building. Figure 2 The diagram illustrates a distributed architecture for a Building Energy Management System (BEMS) – Department Controllers (DCs). BEMS acts as the central controller, responsible for overall energy dispatch and optimization. Each department's DC independently manages its local air conditioning equipment, reporting departmental aggregated load data to BEMS only through an encrypted network to protect device-level privacy. During initialization, the DC collects parameters such as air conditioning rated power, temperature comfort range, and minimum start-stop time. BEMS acquires data including photovoltaic predicted output, energy storage status, grid time-of-use pricing, and demand response (DR) periods. The dispatch cycle is divided into 24 hours, defining binary variables to represent air conditioning start-stop status and continuous variables to represent real-time temperature.

[0033] Step S02: Layered modeling at the department and building levels. The department-level model (DC end) aims to minimize the department's air conditioning electricity cost, with constraints including temperature comfort, minimum equipment start-up and shutdown time, and load limits during DR (Dedicated Response) periods. The building-level model (BEMS end) aims to minimize the overall electricity purchase cost and maximize DR incentive benefits, with constraints covering the power balance between departmental air conditioning load, public air conditioning load, photovoltaic, energy storage, and the power grid, as well as energy storage charging and discharging power and state capacity limits. Through layered modeling, a balance is achieved between local departmental optimization and global building coordination.

[0034] Step S03: Distributed optimization solution (ADMM algorithm), using the Alternating Direction Multiplier Method (ADMM) to achieve distributed collaboration between departmental and building-level models; such as Figure 3 As shown, after BEMS initializes the Lagrange multipliers and penalty factors, each DC optimizes the air conditioning start-up and shutdown plan based on the coordination signals between the local model and BEMS, and only feeds back the total department load to BEMS; BEMS updates the multipliers and penalty factors according to global constraints, and iterates until the original residual and the dual residual meet the convergence threshold; centralized data collection is avoided, and only the aggregated load data is exchanged to protect department privacy.

[0035] Meanwhile, to address the obstacle to ADMM convergence caused by the discreteness of air conditioner start-stop states, a convex relaxation technique is employed to relax binary variables into continuous variables [0,1]. Furthermore, the physical feasibility of the solution is enhanced through a covering inequality (forced shutdown of some equipment when load exceeds limits during the DR period). After the relaxed solution converges, an iterative threshold rounding technique is used to successively restore the continuous variables to binary values: values ​​below the threshold are set to 0, and values ​​above the threshold are set to 1. The remaining variables are iteratively adjusted to satisfy the minimum start-stop time constraint, ensuring that the scheduling scheme conforms to the equipment operation rules.

[0036] Step S04: The aggregator architecture, i.e., the market trading mechanism, is established. The aggregator integrates its own buildings, shared energy storage, and photovoltaic power generation. Users adjust flexible loads based on incentive prices to reduce peak loads or increase off-peak loads. In addition, the aggregator can also purchase electricity from the external electricity spot market to assist the energy storage system and photovoltaic output in meeting electricity load demands. When the photovoltaic power generation configured by the aggregator cannot fully cover the electricity load, the aggregator can not only purchase electricity from the spot market but also draw electricity from the shared energy storage system. When there is a surplus in photovoltaic output, the surplus electricity can be stored in the shared energy storage system.

[0037] This embodiment uses aggregators as the main body and the trading center as the connecting hub, by constructing a trading center model, an aggregator model, and a power generator model. The trading center clears the market based on the quoted electricity volume and price information submitted by both supply and demand sides; the supply-side entities quote prices with reference to the day-ahead electricity market settlement price, their own unit costs, and installed capacity; the demand-side entities optimize the allocation of the daily electricity load based on the market clearing price of the trading center, the cost of adjustable load, and the cost of energy storage.

[0038] Based on the above process, this embodiment can achieve the aggregation and coordination of flexible building resources, providing scientific guidance for constructing a more intelligent, efficient, and low-carbon building energy system. To facilitate understanding of the technical solution of this embodiment, further explanations and descriptions of the specific implementation methods are provided, including: In step S01, for the scenario of aggregated scheduling of flexible resources (air conditioning) in self-owned buildings, the system architecture construction and data initialization can be achieved through the following steps: Step S101: Construct a distributed control system and collect air conditioning equipment parameters and energy data.

[0039] During the deployment of the distributed control architecture, a Building Energy Management System (BEMS) is established as the central controller, responsible for overall energy scheduling and optimization; each department is equipped with a Department Controller (DC) to independently manage the air conditioning equipment in its area. BEMS and DC communicate via an encrypted local area network, transmitting only department-level aggregated load data to protect device-level privacy.

[0040] Step S102: In the equipment and environment data acquisition, set the air conditioning parameters: collect the rated power of the air conditioner for each DC. P rated,h,n Comfortable temperature range T min,h,n , T max,h,n Minimum start-stop time Gamma h,n Permissible operating hours (office hours 8:00-18:00) thermal resistance R th,h,n Heat capacity C th,h,n Efficiency factor Eta h,n Energy data: BEMS obtains predicted output of rooftop solar photovoltaic (SPS) and battery energy storage (BESS) capacity / state of charge (SOC). j Time-of-use electricity pricing Gamma elec,j Demand Response (DR) period [ Tau s,r , Tau e,r and incentive electricity prices Gamma inc,j Scheduling cycle setting: Divided into 24-hour periods. J Each time step (15 minutes / step, ( J =96), Definition: Binary variable Omega j,h,n :air conditioner n During the period j Start / stop status (1 = running, 0 = off); continuous variable T j,h,n :air conditioner n During the period j Real-time temperature; slack variable Omega : A continuous value between 0 and 1 after convex relaxation.

[0041] In step S02, the following process can be implemented: Step S201: Department-end (DC) load optimization model: (1) Objective function: Minimize the air conditioning electricity cost of the department: ; Where, Gamma j,h,n is a binary variable: air conditioning n is the start-stop state of the air conditioner in period j (1=running, 0=off). P rated,h,n is the rated power of the air conditioner n , kW; Gamma elec,j is the on-grid electricity price of period j, yuan / kWh.

[0042] (2) Constraint function ; ; ; Where, T j,h,n为 is the real-time temperature of the air conditioner n in period j , ℃;[ T min,h,n , T max,h,n ] is the temperature comfort interval set by the department h for the air conditioner n ; Gamma h,n is the minimum continuous running time step number of the air conditioner n ;[ Tau s,r , Tau e,r ] is the DR period; Chi j,h is a flag indicating whether the department h participates in DR in period j (1=participates, 0=does not participate); Delta ∈(0,1] is the load reduction ratio set by the department h ; P base,h is the baseline load of the department h in the DR period, i.e., the average load when not participating in DR; P max,h is the maximum allowed load of the department h, kW.

[0043] Step S202: Building-level (BEMS) Aggregation Scheduling Model (1) Objective Function Minimize the total building electricity purchase cost, maximize the DR incentive revenue, and achieve the global optimization of shared energy (photovoltaic, energy storage) and air-conditioning load; that is, the objective function is ; where, P grid,j is the power of electricity purchased from the external grid at time period j , kW; P base is the baseline electricity purchase power of the building during the DR period, that is, the average electricity purchase amount when not participating in DR, kW; Gamma inc,j is the DR incentive electricity price at time period j , yuan / kWh; ( P base -P grid,j ) + is the electricity purchase reduction amount during the DR period (only valid when (Pgrid,j < Pbase)).

[0044] (2) Constraint Function ; ; ; ; ; ; where, P pub,j is the air-conditioning power of the public area, kW; P SPS,j is the rooftop photovoltaic output at time period j , kW; P chg BESS,j / P disch BESS,j is the charging / discharging power of the energy storage system at time period j , kW; SOC j is the state of charge of the energy storage at time period j , %; SOC min / SOC max is the upper and lower limits of the energy storage capacity; Eta chg / Eta disch is the charging / discharging efficiency of the energy storage; P pub,min is the minimum operating power of the air-conditioning in the public area, kW.

[0045] In step S03, collaboration between departmental and building-level models is achieved through alternating iterations to protect departmental privacy. This can be implemented through the following process: Initialization parameters: BEMS sets the initial Lagrange multipliers. Lambda j,h,0 Penalty Factor Rho 0 = 0.001, convergence threshold Epsilon prim =10 -3 , Epsilon dual =10 -4 Relax all binary variables for air conditioner start / stop. Omega j,h,n For continuous variables ∈[0,1], eliminating the obstacle of discreteness to convergence.

[0046] Step S301: Departmental DC Local Optimization. Each department's DC independently optimizes its own air conditioning load based on local air conditioning constraints and BEMS coordination signals, only feeding back the aggregated total power to BEMS. Optimization objective: ; in, P j,h department h During the period j The total air conditioning power is calculated by aggregating the start / stop status of all air conditioners in the department. ; For BEMS in the i In the next iteration, the department h Recommended time period j Air conditioning load, kW.

[0047] DC only reports the total departmental load to BEMS. P h,i+1,j The start / stop status of individual air conditioners is not exposed. Omega j,h,n .

[0048] Step S302: Building-wide optimization (BEMS). Based on the aggregated load reported by each department, BEMS optimizes the public air conditioning load and shared energy (photovoltaics, energy storage), updates coordination parameters, and feeds them back to the departmental DC. The optimization objectives are detailed in step S202.

[0049] Lagrange multiplier update ; Penalty factor adjustment ; in, Mu >1 is used to accelerate convergence; Epsilon dec >1 is used to avoid excessive punishment.

[0050] Step S303: Convergence judgment and residual calculation. Determine whether the department and building reach an agreement through residual indicators, and terminate iteration to enter the solution processing stage: (1) Original residual (load deviation) ; (2) Dual residual (multiplier change) ; (3) Termination condition: when r i < Epsilon prim and s i < Gamma dual , stop iteration, otherwise return to step S3-1 to continue iteration.

[0051] Step S304: Solution refinement processing. Convex relaxation and iteration threshold rounding. Restore the continuous solution after relaxation to binary on-off state to ensure compliance with air conditioning physical constraints.

[0052] (1) Covering inequality enhances integrity Add constraints in the relaxation model to avoid DR period load overlimit. If the total power of the department , force part of the air conditioner to shut down: , C j,h is the subset of overloaded air conditioners, and C j,h is the subset size.

[0053] (2) Iteration threshold rounding technique Threshold filtering: , ; Iteration adjustment: for the undetermined , solve the local model by iteration, preferentially meet the minimum start-stop time constraint, and gradually correct to binary value. ∈(ϵ,1-ϵ)if air conditioner n needs to run continuously for Figure 4 h,n time steps, force it to remain in the running state after starting.

[0054] In step S04, check whether all air conditioner states meet the constraints (such as temperature, start-stop time), if not, re-adjust the non-converged variables until the solution is completely feasible; Specifically: Step S401: As shown in Lambda , the aggregator's game mechanism in the spot market includes two stages: pricing decision and quantitative decision, which have a sequential order and influence each other in the iteration process until equilibrium is reached: The first stage (pricing stage): the transaction center as the upper subject clears according to the supply-demand relationship and market information, and the target is to minimize the external cost of the system. The clearing price plays a guiding role in the bidding behavior of the aggregator in this process.

[0055] The second stage (quantitative stage): the supply side and the user side as the lower subject respectively determine the optimal generation capacity and load demand according to the price signal of the transaction center. The optimal decision of the lower subject can be regarded as a function of the upper decision variable, and the aggregator can refer to the optimized load demand to submit reasonable bidding capacity in the spot market.

[0056] Step S402: construction of the bidding price model based on Stackelberg game-the transaction center model, whose benefits are to maximize the aggregated transaction capacity and minimize the external cost: ; Among them, F iso represents the degree of deviation of the aggregator's bidding; pl b,t t represents the system marginal clearing price (System Marginal Clearing Price) of the day-ahead power market at time t Eu(e)l represents the power purchase price submitted by the aggregator at time t

[0057] Step S403: construction of the bidding price model based on Stackelberg game-the aggregator model: (1) Objective function The aggregator optimizes the load dispatching strategy and shared energy storage service use strategy in each period based on the price given by the power transaction center. The target is to maximize its own revenue: ; Among them, F l represents the net income of the aggregator; Cel is the power consumption utility function of the aggregator, reflecting the satisfaction obtained by the user in power consumption, which is usually a non-decreasing and convex function, and a quadratic function is used in this embodiment; CESS l is the power purchase cost of the aggregator in the day-ahead market; Wsel,j is the service fee paid by the aggregator when using the shared energy storage system.

[0058] The power purchase cost is represented as: ; The power consumption utility function is represented as: ;​​ The shared energy storage service cost is expressed as: ; wherein, represents the adjusted electricity load of the aggregator in the tth time period; t a , b , c is a preference coefficient in the aggregator's electricity consumption utility function, used to reflect the degree of preference and sensitivity of the user to electricity consumption; F ESS is the basic service cost to be paid for using the shared energy storage system.

[0059] (2) Constraint conditions 1) Adjustable load constraint Considering that there are flexible electricity load resources on the user side, the electricity load can be divided into rigid load (fixed load) and flexible load (adjustable load), and the electricity load of the aggregator in the tth time period can be expressed as: t ; wherein, represents the fixed load in the tth time period, which has a high reliability requirement and a determined electricity consumption time; represents the adjustable load, which can be adjusted according to the electricity price information to realize demand side response (DR).

[0060] The adjustable load should satisfy the following restriction conditions: ; ; wherein, is the upper limit value of the adjustable load in the tth time period; t is the total amount of adjustable load in the adjustment period, which means that the total amount of adjustable load before and after the response remains unchanged. Eta 2) Power balance constraint

[0061] Since the shared energy storage mechanism is introduced, the aggregator can perform energy storage or power taking operation in each time period. Under the premise of considering the user side load response, the aggregator needs to satisfy the following power balance constraint: ; wherein, represents the electricity purchase amount declared by the aggregator in the tth time period in the day-ahead market; is the photovoltaic power generation output in the time period; , respectively represents the charging power and discharging power in the tth time period.

[0062] ​​​​3) Shared energy storage constraints Assume the shared energy storage system is independently configured by an aggregator, primarily providing charging and discharging services to users. Divide a day into 24 time periods. The energy storage system in the [number]th period... t The energy storage state at time +1 can be represented as: ; in, , These represent the aggregator in the first... t Charging power and discharging power for each time period; Eta d c , EESS min These are the efficiency factors during the charging and discharging processes, respectively. To ensure the continuous and stable operation of the shared energy storage system, it is assumed that the total charging and discharging power is balanced throughout the 24 time periods of a day, i.e.: ; In addition, to prevent the charging and discharging power from exceeding the capacity limit of the shared energy storage system at any given time, the following capacity constraints must also be met: ; in, EESS max , It G,i These represent the minimum and maximum capacity limits of the shared energy storage system, respectively. During any given time period, the charging or discharging power of all users must not exceed the maximum power limit allowed by the shared energy storage system. t The following constraints must be met: ; in, , These represent the maximum allowable charging power and maximum discharging power of the shared energy storage system, respectively. When a user uses the shared energy storage system, they need to pay a corresponding service fee to the energy storage service provider. The total cost for shared energy storage that a user needs to pay per day is: ; in, Indicates the first t The rental price per unit of charging or discharging power for each time period. Meanwhile, the costs incurred by energy storage service providers to complete the charging and discharging operations are as follows: ; in, p c , p d These are the cost coefficients for unit charging power and unit discharging power, respectively.

[0063] In summary, the revenue function of a shared energy storage service provider within a day can be expressed as: ; Step S404: Construction of a bidding price model based on a Stackelberg game - a power plant model: (1) Construction of a power plant price scene In order to more accurately depict the bidding behavior of the power plant, the Latin Hypercube Sampling (LHS) method is used in this embodiment to generate a set of power plant bidding random scene, and on this basis, the Kantorovich distance model is combined to reduce the scene, and finally 5 typical bidding scenes are selected; Specifically: 1) The cumulative distribution of each random variable is divided into N equal probability intervals; 2) A sample probability is randomly selected in the i th interval, which is in the form of: ; 3) The sample probability value is converted into a variable value by the inverse function F -1 (.) of the distribution function: ; 4) The N sample values obtained by each variable are randomly combined with equal probability to construct N power plant bidding scenes.

[0064] The scene reduction method based on Kantorovich distance is used to improve the calculation efficiency and representativeness, specifically: 1) Calculate the Kantorovich distance s i , s j between any two scenes D k ( s i , s j ); 2) Calculate the distance expectation value (i.e. distance multiplied by probability) of each scene relative to the target scene s , delete the scene with the smallest distance from s , and transfer its probability to s ; 3) Repeat the above steps until the number of scenes is reduced to the preset n .

[0065] (2) Objective function The goal of the power plant is to maximize its daily income in the day-ahead market, and the objective function is: ; wherein, F G,i the total revenue of the power producer; i the power producer Ct G,i the power selling revenue of the power producer in the time period t; i the power producer t the power selling revenue of the power producer in the time period t; pl b,t the power producer i the power generation cost of the power producer in the time period t. t the power selling revenue

[0066] ; the power generation cost ; wherein, is the quoted price of the power producer in the time period t; is the quoted power of the power producer in the time period t; ai, bi, ci are respectively the fuel cost coefficient, the power generation cost of the power producer. is a quadratic function.

[0067] (3) Constraint condition the power generation output of the power producer needs to meet the installed capacity limit ; wherein, g i,min , g i,max : the minimum and maximum output boundary of the power producer i reflecting the physical operation constraints.

[0068] In this embodiment, the transaction center acts as the leader of the game, and its goal is to aggregate the maximum transaction power and minimize the external cost after obtaining the quotes of each market subject, to clear the day-ahead electricity market, and to obtain the market clearing price of each time period; the aggregator acts as a follower, aiming to maximize the user surplus, and adjusts the use strategy of the shared energy storage service and the quoted power strategy of each time period according to the clearing price, user demand and shared energy storage rental fee; and the power producer acts as a passive party, aiming to maximize its own revenue, and optimizes the power generation plan according to the clearing price and its own cost function.

[0069] Therefore, the process of the aggregator participating in the spot market bidding can be regarded as a Stackelberg game, and its game model can be expressed as ; the game model includes three basic elements of the Stackelberg game: participants L , strategy set U and revenue function F . In the game process, the transaction center adjusts the market clearing price Lt sel as its strategyU G to regulate external costs F iso ; aggregators use adjustable loads qt G,i as a strategy U LA to adjust their revenues F l ; power producers use time-of-use pricing Figure 5 as a strategy U G,i to form the overall decision variable set U .

[0070] In this game process, when all followers make the optimal response according to the leader strategy, and the leader accepts the response, the Stackelberg equilibrium is reached. , the Stackelberg equilibrium solution in this embodiment, satisfies the following conditions: ; This embodiment uses a genetic algorithm and a CPLEX hybrid solution as shown in Fk l , and the specific process is as follows: (1) Initialize the parameters of the trading center, shared energy storage service providers and aggregators, let the iteration number k = 0, set the population size m = 40, the maximum iteration number is 200, the mutation rate is 5%, and the crossover probability is 70%.

[0071] (2) Randomly generate m groups of acceptable electricity purchase prices of the trading center using the genetic algorithm as the population, and pass them to the aggregators.

[0072] (3) Let k = k + 1.

[0073] (4) The aggregators receive m groups of market clearing prices, use the CPLEX solver to solve the flexible load distribution and shared energy storage service usage strategy, calculate the current revenue Fk G,i , and return the electricity purchase information of each period to the trading center.

[0074] (5) The power producers receive m groups of market clearing prices, use the CPLEX solver to solve the optimal pricing strategy, calculate the current revenue Fk is , and return the electricity sale information of each period to the trading center.

[0075] (6) The trading center performs market clearing based on the electricity purchase information of aggregators, the power generation information of generators, and the bidding scenario generated by Latin Hypercube sampling, obtains the market clearing price, and calculates and retains the current external costs. Fk iso .

[0076] (7) The genetic algorithm performs selection and mutation operations to generate a new generation of market-clearing electricity prices. Steps (4)–(6) are repeated to calculate the new trading center revenue. FK l and aggregator revenue F'k iso > Fk iso .

[0077] (8) If Fk+1 iso Then let Fk iso = Fk+1 l , Fk l = Fk+1 G,i , Fk G,i = Fk+1 iso Otherwise, F'k iso, Fk+1 l = Fk+1 G,i = F ′kl, Figure 6 = F ′k G,i; (9) If k If the value is 200, the program ends; otherwise, return to step (3).

[0078] This embodiment utilizes a game theory mechanism to optimize the scheduling of flexible building resource loads. By constructing a "BEMS-LEMS" distributed architecture and combining it with the ADMM distributed optimization algorithm, it achieves privacy-friendly scheduling of temperature-controlled loads such as air conditioning. It introduces shared energy storage resources and a Stackelberg game mechanism to construct a multi-agent, two-stage decision-making model for electricity trading, effectively improving the market responsiveness and pricing strategy optimization level of load aggregators. It significantly enhances the schedulability and participation depth of temperature-controlled loads (especially air conditioning systems) in demand response scenarios within self-owned buildings, achieving efficient aggregation and scheduling of flexible resources. Compared to traditional methods, this invention possesses stronger privacy protection capabilities, higher convergence efficiency, and superior overall energy efficiency. It is applicable to various commercial building scenarios such as corporate headquarters, data centers, and park office buildings, and has broad promotional value and promising technological applications.

[0079] Example 2 Embodiment 2 of the present invention introduces a building flexible resource load optimization scheduling system based on game theory mechanism.

[0080] like Figure 1 The illustrated building flexible resource load optimization scheduling system includes: The acquisition module is configured to acquire building flexible resource loads in a distributed architecture. a hierarchical module configured to perform hierarchical modeling of the acquired distributed architecture at the department level and the building level to obtain department-level models and building-level models; an optimization module configured to perform distributed optimization solution of the constructed department-level models and building-level models by using an alternating direction multiplier method to obtain a distributed collaborative model; a game module configured to consider aggregators and construct a single-leader-multiple-follower game model with a transaction center as the leader and the aggregators and power suppliers as the followers; a scheduling module configured to solve the constructed game model to complete optimization scheduling of building flexible resource loads based on a game mechanism.

[0081] As one or more implementations, in the scheduling model, the constructed game model is solved by using a genetic algorithm and a CPLEX solver, specifically: parameters of the transaction center, the shared energy storage service provider, and the aggregator are initialized; a genetic algorithm is used to randomly generate a plurality of groups of acceptable electricity purchase prices of the transaction center, and the generated acceptable electricity purchase prices are passed to the aggregator as a population and iterated; the aggregator receives market clearing prices of all transaction centers, uses a CPLEX solver to solve load distribution and the shared energy storage service provider, calculates current electricity price revenue, and returns electricity purchase information of each period to the transaction center; the power supplier receives market clearing prices of all transaction centers, uses a CPLEX solver to solve optimal offers, calculates current revenue, and returns electricity quantity sale information of each period to the transaction center; the transaction center performs market clearing based on the electricity purchase information, the power supplier information, and the offer scenarios generated by Latin hypercube sampling, obtains market clearing prices, calculates and retains current external costs; the population is selected and mutated based on the genetic algorithm to generate a new generation of market clearing prices, and the iteration is solved to calculate the transaction center revenue and the aggregator revenue; the transaction center revenue and the aggregator revenue obtained are compared, and the iteration is ended when the number of iterations reaches a certain value, and the solution of the game model is completed.

[0082] As one or more implementations, in the game model, the constructed game model at least includes a transaction center model, an aggregator model, and a power supplier model; the participation of the aggregator in the spot market game includes a pricing decision stage and a quantitative decision stage; in the pricing decision stage, the transaction center as an upper subject performs clearing based on supply and demand relations and market information to minimize system external costs, and completes pricing of the aggregator in the clearing price process; in the quantitative decision stage, the supply side and the user side as lower subjects respectively determine optimal power generation and load demand based on the price signal of the transaction center, the optimal decision of the lower subject is a function of the upper decision variable, and the aggregator obtains the priced electricity quantity by referring to the optimized load demand.

[0083] As one or more implementations, in the optimization module, in the process of distributed optimization solving, an alternating direction multiplier method (ADMM) is used to realize distributed collaboration of the department-level model and the building-level model, specifically: the building-level model initializes a Lagrange multiplier and a penalty factor, the department-level model optimizes a start-stop plan of a load based on a local model and a coordination signal of the building-level model, and feeds back a total load of the department to the building-level model; the building-level model updates the multiplier and the penalty factor according to global constraints, and iterates cyclically until original residual error and dual residual error meet a convergence threshold, and through interaction of aggregated load data, privacy of the department is ensured; a convex relaxation technique is used to relax binary variables into continuous variables [0, 1], and through covering inequalities, physical feasibility of a solution is enhanced; after the relaxed solution converges, an iterative threshold rounding technique is used to restore the continuous variables into binary values successively, and through iterative adjustment, a minimum start-stop time constraint is met, and distributed optimization solving is completed.

[0084] The detailed steps are the same as those of the building flexible resource load optimization scheduling method based on the game mechanism provided in Embodiment One, and will not be described here again.

[0085] Embodiment Three The embodiment three of the present application provides a computer readable storage medium.

[0086] A computer readable storage medium has a program stored thereon, and the program is executed by a processor to implement the steps in the building flexible resource load optimization scheduling method based on the game mechanism according to the embodiment one of the present application.

[0087] The detailed steps are the same as those of the building flexible resource load optimization scheduling method based on the game mechanism provided in Embodiment One, and will not be described here again.

[0088] Embodiment Four The embodiment four of the present application provides an electronic device.

[0089] An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor, and the processor implements the steps in the building flexible resource load optimization scheduling method based on the game mechanism according to the embodiment one of the present application when executing the program.

[0090] The detailed steps are the same as those of the building flexible resource load optimization scheduling method based on the game mechanism provided in Embodiment One, and will not be described here again.

[0091] Embodiment Five The embodiment five of the present application provides a computer program product.

[0092] A computer program product includes software code, and a program in the software code executes the steps in the building flexible resource load optimization scheduling method based on the game mechanism according to the embodiment one of the present application.

[0093] The detailed steps are the same as the building flexible resource load optimization scheduling method based on the game mechanism provided in Embodiment 1, and will not be described herein.

[0094] Those skilled in the art will understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application 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-ROMs, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present application can be implemented in various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0095] The present application is described with reference to flowcharts and / or block diagrams of the method, device (system), and computer program product according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus generate a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 means for carrying out the functions specified in the flowchart

[0096] These computer program instructions can also be stored in a computer-readable memory capable of causing the computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including instruction means, which implements the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 means for carrying out the functions specified in the flowchart

[0097] These computer program instructions can also be loaded into a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable data processing apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable data processing apparatus provide a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks ​ means for carrying out the functions specified in the flowchart

[0098] While the preferred embodiments of the application have been described, additional variations and modifications can be made to the preferred embodiments without departing from the spirit and scope of the application. Therefore, it is intended that such additional variations and modifications be included within the scope of the application as defined by the following claims and their equivalents.

[0099] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.

[0100] The above description is merely illustrative of the preferred embodiments of the present application and is not intended to limit the scope of the application. Variations and modifications can be made to the preferred embodiments without departing from the spirit and scope of the present application. Therefore, the present application should not be limited to any single embodiment, but should be construed in the broadest scope possible with the appended claims and their equivalents.

Claims

1. A method for building flexible resource load optimization scheduling based on game mechanism, characterized in that, The application relates to a method for optimizing and scheduling building flexible resource load based on a game mechanism. The method comprises the following steps: a distributed architecture of building flexible resource load is acquired; department-level and building-level hierarchical modeling is performed on the acquired distributed architecture to obtain a department-level model and a building-level model; distributed optimization solving is performed on the constructed department-level model and building-level model by using an alternating direction multiplier method to obtain a distributed collaborative model; a single-leader-multiple-follower game model is constructed by taking a transaction center as a leader and taking aggregators and power suppliers as followers, and the aggregators are considered; 2. The building flexible resource load optimization scheduling method based on game mechanism according to claim 1, characterized in that, the game model is solved to complete optimization and scheduling of the building flexible resource load based on the game mechanism. The game model is solved by using a genetic algorithm and a CPLEX solver, specifically: parameters of the transaction center, the shared energy storage service provider and the aggregators are initialized; a genetic algorithm is used to randomly generate a plurality of groups of acceptable power purchase prices of the transaction center, the generated acceptable power purchase prices are taken as a population and are transmitted to the aggregators and iterated; the aggregators receive market clearing prices of all the transaction centers, solve load distribution and the shared energy storage service provider by using the CPLEX solver, calculate current power purchase information and return the current power purchase information to the transaction center; the power suppliers receive market clearing prices of all the transaction centers, solve optimal offers by using the CPLEX solver, calculate current revenue and return power sale information of each period to the transaction center; the transaction center performs market clearing according to the power purchase information, the power supplier information and offer scenarios generated by Latin hypercube sampling, obtains market clearing prices, calculates and retains current external costs; 3. The building flexible resource load optimization scheduling method based on game mechanism according to claim 1, characterized in that, selection and variation of the population are performed based on the genetic algorithm, a new generation of market clearing prices is generated, iterative solving is performed, transaction center revenue and aggregator revenue are calculated; the transaction center revenue and the aggregator revenue are compared, and when the number of iterations reaches a certain value, the iteration is ended, and solving of the game model is completed.

4. The building flexible resource load optimization scheduling method based on game mechanism according to claim 1, characterized in that, The constructed game model at least comprises a transaction center model, an aggregator model and a power supplier model; in the game of the aggregator in the spot market, a pricing decision stage and a quantitative decision stage are included; in the pricing decision stage, the transaction center as an upper subject performs clearing according to a supply-demand relationship and market information, takes minimizing system external costs as a target, and completes pricing of the aggregators in the process of clearing the prices; in the quantitative decision stage, the supply side and the user side as lower subjects respectively determine optimal power generation and load demand according to the price signal of the transaction center, the optimal decision of the lower subject is a function of the upper decision variable, and the aggregators refer to the optimized load demand to obtain priced power. In the process of distributed optimization solving, the alternating direction multiplier method (ADMM) is used to realize distributed collaboration of the department-level model and the building-level model, specifically: the building-level model initializes Lagrange multipliers and penalty factors, the department-level model optimizes start-stop plans of the load based on local models and coordination signals of the building-level model, and feeds back department total load to the building-level model; The building-level model is updated according to global constraints, and the multipliers and penalty factors are updated, and the iteration is circular until the original residual error and the dual residual error meet the convergence threshold, and the privacy of the department is ensured through the interaction of aggregated load data; the convex relaxation technology is used to relax the binary variable into a continuous variable [0, 1], and the physical feasibility of the solution is enhanced through covering inequalities; after the relaxation solution converges, the continuous variable is restored to a binary value through the iterative threshold rounding technology, and the minimum start-stop time constraint is met through iterative adjustment, and the distributed optimization solution is completed.

5. A building flexible resource load optimization scheduling system based on a game mechanism, characterized in that, Comprise: An acquisition module configured to acquire a distributed architecture of building flexible resource load; A hierarchical modeling module configured to hierarchically model the acquired distributed architecture at the department level and the building level to obtain a department-level model and a building-level model; An optimization module configured to use an alternating direction multiplier method to distribute the department-level model and the building-level model to obtain a distributed collaborative model; A game module configured to consider aggregators, and construct a single-leader-multiple-follower game model with a transaction center as the leader and aggregators and power suppliers as the followers; A scheduling module configured to solve the constructed game model to complete the optimization scheduling of building flexible resource load based on the game mechanism. 6.The building flexible resource load optimization scheduling system based on game mechanism of claim 5, wherein, In the scheduling model, the genetic algorithm and the CPLEX solver are used to solve the constructed game model, specifically: the parameters of the transaction center, the shared energy storage service provider and the aggregator are initialized; the genetic algorithm is used to randomly generate a plurality of groups of acceptable electricity purchase prices of the transaction center, and the generated acceptable electricity purchase prices are transmitted to the aggregator as a population and iterated; the aggregator receives the market clearing price of all transaction centers, uses the CPLEX solver to solve the load distribution and the shared energy storage service provider, calculates the current electricity price income and returns the electricity purchase information of each period to the transaction center; the power supplier receives the market clearing price of all transaction centers, uses the CPLEX solver to solve the optimal offer, calculates the current income, and returns the electricity sale information of each period to the transaction center; the transaction center performs market clearing according to the electricity purchase information, the power supplier information and the offer scenario generated by the Latin hypercube sampling to obtain the market clearing price, calculate and retain the current external cost; The genetic algorithm is used to select and mutate the population to generate a new generation of market clearing prices, and the iteration is solved to calculate the transaction center income and the aggregator income; the transaction center income and the aggregator income obtained are compared, and the iteration is ended when the number of iterations reaches a certain value, and the solution of the game model is completed. 7.The building flexible resource load optimization scheduling system based on game mechanism of claim 5, wherein, In the game model, the game model constructed at least includes a transaction center model, a wholesaler model and a power generator model; in the wholesaler participating in the spot market game, there are a pricing decision stage and a quantitative decision stage; in the pricing decision stage, the transaction center as an upper subject clears according to the supply-demand relationship and market information, to minimize the external cost of the system, and completes the pricing for the wholesaler in the process of clearing the electricity price; in the quantitative decision stage, the supply side and the user side as lower subjects respectively determine the optimal power generation and load demand according to the price signal of the transaction center, the optimal decision of the lower subject is a function of the upper decision variable, and the wholesaler obtains the quoted power according to the optimized load demand. 8.The building flexible resource load optimization scheduling system based on game mechanism of claim 5, wherein, In the optimization module, in the process of distributed optimization solving, the alternating direction multiplier method (ADMM) is adopted to realize the distributed collaboration of the department-level model and the building-level model, specifically: the building-level model initializes the Lagrange multiplier and the penalty factor, the department-level model optimizes the start-stop plan of the load based on the local model and the coordination signal of the building-level model, and feeds back the total load of the department to the building-level model; The building-level model updates the multiplier and the penalty factor according to the global constraint, and iterates cyclically until the original residual and the dual residual meet the convergence threshold, the load data is aggregated to interact, the privacy of the department is ensured, the convex relaxation technology is adopted to relax the binary variable into a continuous variable [0, 1], the physical feasibility of the solution is enhanced through covering inequalities, and after the relaxation solution converges, the iterative threshold rounding technology is used to restore the continuous variable to a binary value, the minimum start-stop time constraint is met through iterative adjustment, and the distributed optimization solving is completed.

9. A computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to realize the steps of the building flexible resource load optimization scheduling method based on the game mechanism in any one of claims 1-4.

10. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor executes the program to realize the steps of the building flexible resource load optimization scheduling method based on the game mechanism in any one of claims 1-4.

11. A computer program product comprising software code, characterized in that, The program in the software code executes the steps of the building flexible resource load optimization scheduling method based on the game mechanism in any one of claims 1-4.

Images