A multi-factor integrated multi-agent energy system and a collaborative operation negotiation method

Abstract

The present invention discloses a multi-factor integrated multi-agent energy system and a collaborative operation negotiation method, which relates to the technical field of multi-agent energy systems. The collaborative operation negotiation method is as follows: construct an MAES-WHB model, based on the NB theory, construct an NB model of the MAES-WHB model, decompose the NB model into CPS and PAS, substitute the traded electricity obtained by solving CPS into the NB model, introduce the contribution degree as the bargaining power, and then construct an asymmetric PAS model. Based on the ADMM theory, decompose CPS and PAS, construct distributed models of each agent, improve the standard ADMM, and propose the TPCA-ADMM algorithm for solving CPS and PAS. The present invention comprehensively considers the electrical characteristics and thermal characteristics, and realizes the function of bidirectional interaction between the wind farm and the hydrogen energy system as well as between the wind farm and the building by setting the MAES-WHB model.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-agent energy systems, and in particular to a multi-factor integrated multi-agent energy system and a collaborative operation negotiation method. Background Art

[0002] Under the dual challenges of global climate change and resource depletion, the consumption of traditional fossil energy has risen sharply, leading to increased greenhouse gas emissions and increasingly severe climate warming problems. At the same time, the limited and non-renewable nature of fossil energy has also increased the pressure on energy supply. With the demand for global energy transformation and the rapid development of artificial intelligence technology, as well as the pursuit of intelligent, networked and digital management of energy systems, the traditional energy system can no longer meet the needs of sustainable development. Therefore, energy transformation has become a major challenge facing the world today, that is, the transformation from a traditional energy system with high pollution and high emissions to a new energy system that is clean, low-carbon and sustainable. In MAES, various entities are interconnected through information technology, communication technology and automation technology to form an intelligent energy network.

[0003] Due to the variability and uncertainty of wind power, large-scale wind power grid connection will inevitably lead to a match between wind power output and power load, increasing the difficulty of system scheduling. Therefore, the problem of wind power absorption needs to be solved urgently. Due to the different stakeholders of wind power, hydrogen energy systems and buildings, the specific interest relationship of the subjects in MAES is usually complex, and it is difficult to reasonably divide the interests. At the same time, MAES lacks operational flexibility, and the use of the standard ADMM method for privacy protection leads to slow convergence of distributed optimization, which cannot take into account both privacy protection and high solution efficiency.

[0004] 1. Patent document CN113591375B discloses an optimal coordinated operation method for a multi-energy system based on an intelligent agent. The above patent overcomes the problem that traditional mathematical algorithms require complex modeling of the coupling relationship between physical models, and expands the action space of general machine learning algorithms, so that decisions can be closer to the optimal decision.

[0005] 2. Patent document CN113467398B discloses a distributed control method and system for an integrated energy system based on a consistency algorithm. The above patent realizes the construction of a networked integrated energy system, analyzes the dynamics of each intelligent agent, and uses the consistency control algorithm of the multi-agent system to design a distributed control system for the integrated energy system, so that the electric power load rate of the gas generators of all CCHP units in the integrated energy system and the indoor temperature of the cooling load tend to be consistent.

[0006] 3. The patent document CN113098036B discloses an operation method for an integrated energy system based on a hydrogen fuel cell. The above patent realizes flexible adjustment of energy supply, energy consumption, and energy storage, thereby achieving flexible interaction of integrated energy and vertical integration of supply, demand, and storage.

[0007] 4. The patent document CN107665384B discloses a dispatching method for an integrated power-thermal energy system with multi-region energy stations. The above patent realizes, through the coordination of different types of regional energy stations, full exploration of the potential for complementary advantages of multiple energies to reduce the network loss of the distribution network, improve the peak load phenomenon of the distribution network, increase the consumption rate of renewable energy, and at the same time effectively reduce the operation cost of the regional power-thermal system.

[0008] In summary, inspired by the above patents, the present application has developed functions that can achieve two-way interaction between a wind farm and a hydrogen energy system and between a wind farm and a building, can achieve the function of asymmetric profit distribution, can achieve the function of improving the convergence speed and solution efficiency of distributed optimization, and can achieve the function of distributed energy management for each entity, so as to solve the problems of low energy utilization efficiency, unstable energy supply, low system flexibility, inability to reasonably distribute profits in the coordinated operation of MAES, slow system response speed, high consumption of computing resources, low system reliability, low efficiency, and poor user experience existing in the existing energy system;

[0009] Therefore, the present application proposes a multi-factor integrated multi-agent energy system and a coordinated operation negotiation method that can achieve the functions of two-way interaction between a wind farm and a hydrogen energy system and between a wind farm and a building, can achieve the function of asymmetric profit distribution, can achieve the function of improving the convergence speed and solution efficiency of distributed optimization, and can achieve the function of distributed energy management for each entity. Summary of the Invention

[0010] The purpose of the present invention is to provide a multi-factor integrated multi-agent energy system and a coordinated operation negotiation method to solve the technical problems of low energy utilization efficiency, unstable energy supply, low system flexibility, inability to reasonably distribute profits in the coordinated operation of MAES, slow system response speed, high consumption of computing resources, low system reliability, low efficiency, and poor user experience mentioned in the above background technology.

[0011] To achieve the above purpose, the present invention provides the following technical solution: A coordinated operation negotiation method for a multi-factor integrated multi-agent energy system, and the coordinated operation negotiation method is as follows:

[0012] S1: Construct a MAES-WHB model;

[0013] S2: Based on the NB theory, construct an NB model of the MAES-WHB model;

[0014] S3: Decompose the NB model into CPS and PAS, substitute the transaction power obtained by solving CPS into the NB model, introduce the contribution degree as the bargaining power, and then construct an asymmetric PAS model;

[0015] S4: Based on the ADMM theory, decompose CPS and PAS to construct a distributed model for each subject;

[0016] S5: Improve the standard ADMM, and propose the TPCA-ADMM algorithm for solving CPS and PAS;

[0017] Among them, MAES-WHB is a multi-agent energy system integrating a wind farm, a hydrogen energy system, and a building, MAES is a multi-agent energy system, NB is Nash bargaining, ADMM is the alternating direction method of multipliers, and the TPCA-ADMM algorithm is a three-stage prediction-correction accelerated alternating direction method of multipliers algorithm improved based on the ADMM algorithm;

[0018] The specific content of S1 is as follows:

[0019] S11: Calculate the revenue and cost of the wind farm to construct the objective function of the wind farm;

[0020] S12: Calculate the cost of the hydrogen energy system and consider the power balance constraint to construct the objective function of the hydrogen energy system;

[0021] S13: Calculate the cost of the building and consider the power balance constraint to construct the objective function of the building;

[0022] S14: Construct the MAES-WHB model based on the wind farm model, the hydrogen energy system model, and the building model.

[0023] Preferably, the objective function of the wind farm is:

[0024] ;

[0025] Among them, is the overall profit of the wind farm, is the revenue obtained by the wind farm from selling electricity to the hydrogen energy system, is the revenue obtained by the wind farm from selling electricity to the building, is the revenue obtained by the wind farm from selling electricity to the PG, is the maintenance cost of the wind farm, is the transmission fee of the wind farm;

[0026] The objective function of the hydrogen energy system is:

[0027] ;

[0028] Among them, is the total cost of the hydrogen energy system, is the electricity cost paid by the hydrogen energy system to the PG, is the maintenance cost of the hydrogen energy system;

[0029] The objective function of the building is:

[0030] ;

[0031] Among them, is the total cost of the building, is the electricity cost paid by the building to the PG, is the maintenance cost of the building.

[0032] Preferably, the NB model is:

[0033] ;

[0034] ;

[0035] Among them, is the profit of the wind farm when cooperating, is the cost of the hydrogen energy system when not cooperating, is the cost of the building when not cooperating;

[0036] The CPS of the NB model is:

[0037] .

[0038] Preferably, the specific S3 is:

[0039] S31: Calculate the overall transaction electricity quantity in the operation period;

[0040] S32: Construct a mapping function to quantify the contribution degree of each subject to the transaction;

[0041] S33: Solve the CPS of the NB model to obtain the transaction electricity quantity;

[0042] S34: Substitute the transaction electricity quantity into the NB model to construct an asymmetric PAS model;

[0043] The overall transaction electricity quantity in the operation period is:

[0044] ;

[0045] ;

[0046] Among them, is the operation period, t is the moment, is the electricity quantity sold by the wind farm to the hydrogen energy system during the operation period, The electricity quantity sold by the wind farm to the hydrogen energy system at time t, The electricity quantity sold by the wind farm to the building during the operation period, The electricity quantity sold by the wind farm to the building at time t;

[0047] The method for constructing a mapping function to quantify the contribution degree of each entity to the transaction is as follows:

[0048] ;

[0049] ;

[0050] ;

[0051] Among them, is the contribution degree of the wind farm, is the contribution degree of the hydrogen energy system, is the contribution degree of the building, is a constant used to improve the contribution degree of the wind farm;

[0052] The asymmetric PAS model is:

[0053] .

[0054] Preferably, the specific content of S4 is as follows:

[0055] S41: Introduce new variables , , and , and convert the CPS into a minimization problem;

[0056] S42: Based on ADMM, decompose the augmented Lagrangian function of the CPS into distributed optimization models of the wind farm, hydrogen energy system, and building;

[0057] S43: Introduce new variables , , and , take the logarithm of the asymmetric PAS model, and convert it into a minimization problem;

[0058] S44: Based on ADMM, decompose the augmented Lagrangian function of the asymmetric PAS model into distributed optimization models of the wind farm, hydrogen energy system, and building;

[0059] Among them, represents the electricity quantity that the hydrogen energy system hopes to purchase from the wind farm, represents the electricity quantity that the building hopes to purchase from the wind farm, represents the electricity quantity that the wind farm hopes to sell to the hydrogen energy system, Represents the amount of electricity that the wind farm hopes to sell to the building, Represents the electricity price that the hydrogen energy system hopes to pay to the wind farm, Represents the electricity price that the building hopes to pay to the wind farm, Represents the electricity price that the wind farm hopes to charge from the hydrogen energy system, Represents the electricity price that the wind farm hopes to charge from the building, Is the electricity price between the wind farm and the hydrogen energy system at time t, Is the electricity price between the wind farm and the building at time t.

[0060] Preferably, the augmented Lagrangian function of the CPS is:

[0061] ;

[0062] Among them, and Are Lagrange multipliers, and Are penalty factors;

[0063] The augmented Lagrangian function of the asymmetric PAS is:

[0064]

[0065] Among them, and Are Lagrange multipliers, and Are penalty factors.

[0066] Preferably, the specific S5 is:

[0067] S51: Calculate the primal residual, dual residual and the difference between the primal residual and the dual residual in the first stage;

[0068] S52: Compare the difference with the set value. If the difference is less than the set value, the penalty factor remains unchanged. If the difference is not less than the set value, enter the second stage to update the penalty factor;

[0069] S53: Introduce a prediction-correction mechanism to update the Lagrange multiplier in the third stage;

[0070] S54: Meet the iteration termination condition and stop the solution.

[0071] Preferably, the formula for calculating the primal residual and the dual residual is:

[0072] ;

[0073] Among them, Is the primal residual, is the amount of electricity that the hydrogen energy system / building hopes to purchase from the wind farm, is the amount of electricity that the wind farm hopes to sell to the hydrogen energy system / building;

[0074] ;

[0075] Among them, is the dual residual, is the amount of electricity that the hydrogen energy system / building hopes to purchase from the wind farm;

[0076] The formula for calculating the difference between the original residual and the dual residual is:

[0077] ;

[0078] Among them, is the original residual;

[0079] ;

[0080] Among them, is the dual residual.

[0081] Preferably, the updated penalty factor of S52 is expressed as:

[0082] ;

[0083] Among them, and are penalty factors, is a constant;

[0084] The updated Lagrange multiplier of S53 is expressed as:

[0085] ;

[0086] Among them, is an intermediate parameter in the Lagrange multiplier update process, is the penalty factor, is the amount of electricity that the hydrogen energy system / building hopes to purchase from the wind farm, is the Lagrange multiplier, is the amount of electricity that the wind farm hopes to sell to the hydrogen energy system / building;

[0087] ;

[0088] Among them, is a constant;

[0089] ,

[0090] Among them, is the prediction correction factor, is a given constant.

[0091] Preferably, the iteration termination condition is:

[0092] ;

[0093] where is the number of iterations, is the maximum number of iterations.

[0094] Compared with the prior art, the beneficial effects of the present invention are:

[0095] 1. By setting up the MAES-WHB model, the present invention realizes the function of two-way interaction between the wind farm and the hydrogen energy system and between the wind farm and the building, solves the problems of low energy utilization efficiency, unstable energy supply and low system flexibility, comprehensively considers the electrical characteristics and thermal characteristics, enhances the flexibility of system operation, improves the user's living experience, improves the energy conservation and emission reduction effect, and improves the energy utilization rate;

[0096] 2. By setting up the distribution strategy with MAES collaborative operation, the present invention realizes the function of asymmetric profit distribution, solves the problem of reasonable profit distribution in MAES collaborative operation, can determine the respective bargaining prices through the contribution degrees of the wind farm, the hydrogen energy system and the building, takes into account both individual rationality and coalition rationality, promotes technological innovation and management innovation, and improves the economic benefits of the system;

[0097] 3. By setting up the construction of the CPS distributed model and the PAS distributed model, the present invention realizes the function of improving the convergence speed and solving efficiency of distributed optimization, solves the problems of slow system response speed, high computational resource consumption, low system reliability and poor user experience, can reduce the resource consumption during system operation, prolongs the service life of the equipment, improves the reliability, stability and accuracy of system calculation, and improves the adaptability of the system to the environment and demand;

[0098] 4. By setting up the TPCA-ADMM algorithm, the present invention realizes the function of distributed energy management of each subject, solves the problems of low system processing efficiency and low flexibility, can protect the information privacy of the wind farm, the hydrogen energy system and the building, improves the security and reliability of the energy system, and promotes the sustainable development of energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] Figure 1 is a schematic diagram of the energy system of the present invention;

[0100] Figure 2 is a schematic diagram of the CPS Lagrangian function decomposition of the present invention;

[0101] Figure 3Schematic diagram of Lagrangian function decomposition of the asymmetric PAS model of the present invention;

[0102] Figure 4 Schematic diagram of the distributed solution process of the present invention. Specific implementation manner

[0103] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0104] Embodiment 1: Please refer to Figure 1 , a collaborative operation negotiation method for a multi-factor integrated multi-agent energy system, and the collaborative operation negotiation method is as follows:

[0105] S1: Construct the MAES-WHB model;

[0106] S2: Based on the NB theory, construct the NB model of the MAES-WHB model;

[0107] S3: Decompose the NB model into CPS and PAS, substitute the transaction power obtained by solving CPS into the NB model, introduce the contribution degree as the bargaining power, and then construct an asymmetric PAS model;

[0108] S4: Based on the ADMM theory, decompose CPS and PAS to construct a distributed model for each agent;

[0109] S5: Improve the standard ADMM, and propose the TPCA-ADMM algorithm for solving CPS and PAS;

[0110] Wherein MAES-WHB is a multi-agent energy system integrating a wind farm, a hydrogen energy system, and a building, MAES is a multi-agent energy system, NB is Nash negotiation, ADMM is the alternating direction multiplier method, and the TPCA-ADMM algorithm is a three-stage prediction-correction accelerated alternating direction multiplier method algorithm improved based on the ADMM algorithm;

[0111] The specific content of S1 is as follows:

[0112] S11: Calculate the revenue and cost of the wind farm to construct the objective function of the wind farm;

[0113] S12: Calculate the cost of the hydrogen energy system and consider the power balance constraint to construct the objective function of the hydrogen energy system;

[0114] S13: Calculate the cost of the building and consider the power balance constraint to construct the objective function of the building;

[0115] S14: Construct the MAES-WHB model based on the wind farm model, the hydrogen energy system model, and the building model;

[0116] The objective function of the wind farm is:

[0117] ;

[0118] Where, is the overall profit of the wind farm, is the income obtained by the wind farm from selling electricity to the hydrogen energy system, is the income obtained by the wind farm from selling electricity to the building, is the income obtained by the wind farm from selling electricity to the PG, is the maintenance cost of the wind farm, is the transmission fee of the wind farm;

[0119] The objective function of the hydrogen energy system is:

[0120] ;

[0121] Where, is the total cost of the hydrogen energy system, is the electricity fee paid by the hydrogen energy system to the PG, is the maintenance cost of the hydrogen energy system;

[0122] The objective function of the building is:

[0123] ;

[0124] Where, is the total cost of the building, is the electricity fee paid by the building to the PG, is the maintenance cost of the building;

[0125] Furthermore, calculate the revenue of the wind farm through , and where, is the on-grid electricity price of wind power, is the electricity quantity sold by the wind farm to the PG at time t. Calculate the maintenance cost and transmission fee of the wind farm through and where is the maintenance cost coefficient of the wind farm, is the wind power generation, and are the transmission fee coefficients of the wind farm. The constraints that the wind farm output needs to meet are and , where is the maximum wind power generation, and a wind farm model is constructed by combining the objective function of the wind farm. When the electrolyzer satisfies and , where is the maximum operating power of the electrolyzer, is the ramp power. The hydrogen production during the operation of the electrolyzer is calculated by , where is the hydrogen production rate, is the power consumed by the electrolyzer at time t, is the time interval. The operating mode of the compressor is and , where is the specific heat capacity constant of hydrogen, is the hydrogen production at time t, is the compressor efficiency, is the isentropic exponent of hydrogen, / is the hydrogen compression ratio, is the power consumed by the compressor at time t, is the maximum operating power of the compressor. The formula for indicating the hydrogen storage capacity through the internal pressure of the hydrogen storage tank is , and , where is the internal pressure of the hydrogen storage tank, is the internal temperature of the hydrogen storage tank, is the volume of the hydrogen storage tank, is the molar mass of hydrogen, is the hydrogen load demand at time t, is the maximum pressure of the hydrogen storage tank, is the minimum pressure of the hydrogen storage tank. The operating model of the battery is: , , , and , where is the stored energy at time t, is the charging power at time t, is the charging efficiency, is the discharging power at time t, is the discharging efficiency, is the maximum energy storage capacity, is the minimum energy storage capacity, is a binary variable used to avoid simultaneous charging and discharging at time t. The cost of the hydrogen energy system includes the cost of purchasing electricity from the wind farm and the PG, and the maintenance cost for the electrolyzer and the BS. Through and Calculation, where is the current market price, is the electricity quantity purchased by the hydrogen energy system from PG, is the electrolyzer maintenance cost coefficient, is the electrical energy storage maintenance cost coefficient. The operation of the hydrogen energy system should satisfy the following power balance constraint as , and a hydrogen energy system model is constructed by combining the objective function of the hydrogen energy system. By calculate the heat balance constraint of wall ij, where i and j are nodes, is the heat capacity of the wall, is the wall temperature, is the adjacent node of the wall, is the temperature of node j, is the thermal resistance. Since the thermal resistance values of the wall without window side and the wall with window side are different, it is necessary to analyze according to the actual situation. If the wall is irradiated by sunlight, then take 1, otherwise take 0, is the wall heat absorption rate, is the wall surface area, is the light intensity in the corresponding direction of the wall. By calculate the heat balance of area n, where n is the room index, is the area heat capacity, is the indoor temperature, is the adjacent node of the area, is the heat source in the area. If there is a window in this area, then take 1, otherwise take 0, is the window transmittance, is the window surface area of the area, is the light intensity in the corresponding direction of the window of the area, is the energy efficiency ratio, is the operating power of the air conditioner in room n at time t, is the window thermal resistance. The operating power of the air conditioner needs to satisfy , where is the maximum AC working power. The temperature limit for user comfort is , where is the lowest comfort temperature, is the highest comfort temperature. The trading model includes the building cost, including the cost of purchasing electricity from the wind farm and PG and the maintenance cost for air conditioners and base loads. The calculation formula is , and , where is the electricity quantity purchased by the building from PG, is the building maintenance cost, is the air conditioner maintenance cost coefficient, The electricity consumption of the air conditioner is the basic load maintenance cost coefficient is the electricity consumption of the basic load is the number of buildings, and the power balance constraint that the buildings should meet during operation is , combined with the objective function of the building to construct a building model, and through combining the wind farm model, the hydrogen energy system model and the building model to construct the MAES-WHB model, the function of two-way interaction between the wind farm and the hydrogen energy system and between the wind farm and the building is realized, the problems of low energy utilization efficiency, unstable energy supply and low system flexibility are solved, the electrical characteristics and thermal characteristics are comprehensively considered, the flexibility of system operation is enhanced, the living experience of users is improved, the effect of energy conservation and emission reduction is improved, and the utilization rate of energy is improved.

[0126] Example 2: Please refer to Figure 1 , a collaborative operation negotiation method for a multi-factor integrated multi-agent energy system, and the NB model is:

[0127] ;

[0128] ;

[0129] Among them, is the profit of the wind farm when cooperating is the cost of the hydrogen energy system when not cooperating is the cost of the building when not cooperating;

[0130] The CPS of the NB model is:

[0131] ;

[0132] The specific content of S3 is:

[0133] S31: Calculate the overall transaction electricity quantity during the operation period;

[0134] S32: Construct a mapping function to quantify the contribution degree of each subject to the transaction;

[0135] S33: Solve the CPS of the NB model to obtain the transaction electricity quantity;

[0136] S34: Substitute the transaction electricity quantity into the NB model to construct an asymmetric PAS model;

[0137] The overall transaction electricity quantity during the operation period is:

[0138] ;

[0139] ;

[0140] Among them, is the operating cycle, and t is the time instant, is the electricity quantity sold by the wind farm to the hydrogen energy system during the operating cycle, is the electricity quantity sold by the wind farm to the hydrogen energy system at time t, is the electricity quantity sold by the wind farm to the building during the operating cycle, is the electricity quantity sold by the wind farm to the building at time t;

[0141] The method for constructing a mapping function to quantify the contribution degree of each entity to the transaction is as follows:

[0142] ;

[0143] ;

[0144] ;

[0145] Among them, is the contribution degree of the wind farm, is the contribution degree of the hydrogen energy system, is the contribution degree of the building, is a constant used to increase the contribution degree of the wind farm;

[0146] The asymmetric PAS model is:

[0147] ;

[0148] Furthermore, NB is a cooperative game that can take into account both individual and collective interests. By applying the NB theory and combining the wind farm model, hydrogen energy system model, and building model, the NB model of the MAES-WHB model is obtained. As an alliance of the wind farm, hydrogen energy system, and building, the CPS of the MAES-WHB model is , in the CPS, the transaction amounts I W2H and I W2B can cancel each other out. By solving the CPS, the optimal transaction electricity quantity, as well as the operating powers of the electrolyzer in the hydrogen energy system, BS, and the air conditioner in the building, are obtained. By calculating the overall transaction electricity quantity during the operation cycle, a mapping function is constructed to quantify the contribution degree of each entity to the transaction. Each entity participating in the electricity transaction will obtain a contribution degree, and the more electricity the entity trades, the higher its corresponding contribution degree. When the transaction electricity quantities are the same, the contribution degree of the electricity selling entity is higher than that of the electricity buying entity. By introducing the contribution degree as the bargaining power and substituting the optimal transaction electricity quantity obtained from solving the CPS into the asymmetric NB model, the asymmetric PAS model is constructed , and the optimal transaction electricity price and are obtained by solving the PAS, and then the asymmetric distribution of profits is realized. In In this case, the profit distribution is symmetric. This strategy can determine the bargaining power of each party through the contribution of the wind farm, hydrogen energy system, and building, thereby solving the problem of reasonable profit distribution in the collaborative operation of MAES, taking into account both individual rationality and coalition rationality, promoting technological innovation and management innovation, and improving the economic efficiency of the system.

[0149] Example 3: Please refer to Figure 1 、 Figure 2 and Figure 3 , a collaborative operation negotiation method for a multi-factor integrated multi-agent energy system, where the NB model is:

[0150] ;

[0151] ;

[0152] Among them, is the profit of the wind farm when cooperating, is the cost of the hydrogen energy system when not cooperating, is the cost of the building when not cooperating;

[0153] The CPS of the NB model is:

[0154] ;

[0155] The specific content of S3 is:

[0156] S31: Calculate the overall transaction power during the operation period;

[0157] S32: Construct a mapping function to quantify the contribution of each agent to the transaction;

[0158] S33: Solve the CPS of the NB model to obtain the transaction power;

[0159] S34: Substitute the transaction power into the NB model to construct an asymmetric PAS model;

[0160] The overall transaction power during the operation period is:

[0161] ;

[0162] ;

[0163] Among them, is the operation period, t is the moment, is the power sold by the wind farm to the hydrogen energy system during the operation period, is the power sold by the wind farm to the hydrogen energy system at time t, is the power sold by the wind farm to the building during the operation period, The electricity quantity sold by the wind farm to the building at time t;

[0164] The method of constructing a mapping function to quantify the contribution degree of each entity to the transaction is as follows:

[0165] ;

[0166] ;

[0167] ;

[0168] Among them, is the contribution degree of the wind farm, is the contribution degree of the hydrogen energy system, is the contribution degree of the building, is a constant used to improve the contribution degree of the wind farm;

[0169] The asymmetric PAS model is:

[0170] ;

[0171] The specific content of S4 is:

[0172] S41: Introduce new variables , , and , and convert the CPS into a minimization problem;

[0173] S42: Based on ADMM, decompose the augmented Lagrangian function of the CPS into distributed optimization models of the wind farm, hydrogen energy system and building;

[0174] S43: Introduce new variables , , and , take the logarithm of the asymmetric PAS model, and convert it into a minimization problem;

[0175] S44: Based on ADMM, decompose the augmented Lagrangian function of the asymmetric PAS model into distributed optimization models of the wind farm, hydrogen energy system and building;

[0176] Among them, represents the electricity quantity that the hydrogen energy system hopes to purchase from the wind farm, represents the electricity quantity that the building hopes to purchase from the wind farm, represents the electricity quantity that the wind farm hopes to sell to the hydrogen energy system, represents the electricity quantity that the wind farm hopes to sell to the building, represents the electricity price that the hydrogen energy system hopes to pay to the wind farm, Represents the electricity price that the building hopes to pay to the wind farm, Represents the electricity price that the wind farm hopes to charge from the hydrogen energy system, Represents the electricity price that the wind farm hopes to charge from the building, Is the electricity price between the wind farm and the hydrogen energy system at time t, Is the electricity price between the wind farm and the building at time t;

[0177] The augmented Lagrangian function of the CPS is:

[0178] ;

[0179] Among them, and Are Lagrange multipliers, and Are penalty factors;

[0180] The augmented Lagrangian function of the asymmetric PAS is:

[0181]

[0182] Among them, and Are Lagrange multipliers, and Are penalty factors;

[0183] Furthermore, the ADMM algorithm efficiently handles constrained optimization problems in a distributed and parallel environment by alternately optimizing the primal variables and dual variables, combines the decomposability of the dual ascent method and the excellent convergence properties of the augmented Lagrangian multiplier method, decomposes the large global problem into multiple smaller and easier-to-solve local sub-problems, and obtains the solution of the large global problem by coordinating the solutions of the sub-problems. When and , the wind farm, hydrogen energy system, and building reach an agreement on the traded electricity volume, convert the CPS into a minimization problem, and decompose the augmented Lagrangian function of the CPS into distributed optimization models for the wind farm, hydrogen energy system, and building based on the ADMM algorithm. The distributed optimization model of the wind farm is , the distributed optimization model of the hydrogen energy system is , the distributed optimization model of the building is , when and , it is considered that each entity reaches an agreement on the electricity price, takes the logarithm of the asymmetric PAS model, converts it into a minimization problem, and decomposes the augmented Lagrangian function of the asymmetric PAS model into distributed optimization models for the wind farm, hydrogen energy system, and building based on the ADMM. The optimization model of the wind farm is , the distributed optimization model of the hydrogen energy system is , the distributed optimization model of the building is , which realizes the function of improving the convergence speed and solution efficiency of distributed optimization, solves the problems of slow system response speed, high computing resource consumption, low system reliability and poor user experience, can reduce the resource consumption during system operation, extend the service life of equipment, improve the reliability, stability and accuracy of system computing, and improve the adaptability of the system to the environment and requirements.

[0184] Example 4: Please refer to Figure 4 , a collaborative operation negotiation method for a multi-factor integrated multi-agent energy system, where S5 is specifically:

[0185] S51: Calculate the original residual, the dual residual, and the difference between the original residual and the dual residual in the first stage;

[0186] S52: Compare the difference with the set value. If the difference is less than the set value, the penalty factor remains unchanged. If the difference is not less than the set value, enter the second stage to update the penalty factor;

[0187] S53: Introduce a prediction correction mechanism to update the Lagrange multiplier in the third stage;

[0188] S54: Meet the iteration termination condition and stop solving;

[0189] The formula for calculating the original residual and the dual residual is:

[0190] ;

[0191] Among them, is the original residual, is the electricity quantity that the hydrogen energy system / building hopes to purchase from the wind farm, is the electricity quantity that the wind farm hopes to sell to the hydrogen energy system / building;

[0192] ;

[0193] Among them, is the dual residual, the electricity quantity that the hydrogen energy system / building hopes to purchase from the wind farm;

[0194] The formula for calculating the difference between the original residual and the dual residual is:

[0195] ;

[0196] Among them, is the original residual;

[0197] ;

[0198] Among them, is the dual residual;

[0199] The updated penalty factor of S52 is expressed as:

[0200] ;

[0201] where and are penalty factors, is a constant;

[0202] The updated Lagrange multiplier of S53 is expressed as:

[0203] ;

[0204] where is an intermediate parameter in the Lagrange multiplier update process, is a penalty factor, is the amount of electricity that the hydrogen energy system / building hopes to purchase from the wind farm, is the Lagrange multiplier, is the amount of electricity that the wind farm hopes to sell to the hydrogen energy system / building;

[0205] ;

[0206] where is a constant;

[0207] ,

[0208] where is the prediction correction factor, is a given constant;

[0209] The iteration termination condition is:

[0210] ;

[0211] where is the number of iterations, is the maximum number of iterations;

[0212] Furthermore, it is solved by the TPCA-ADMM algorithm improved based on the ADMM algorithm. In the first stage, the original residual and the dual residual are calculated through and , and then the difference between and is calculated through and , and When it is less than the set value and the primal residual and dual residual change in a decreasing or acceptable direction, the penalty factor remains unchanged and enters the third stage. By introducing a correction mechanism to update the Lagrange multiplier, if and are greater than or equal to the set value, it enters the second stage to update the penalty factor to prevent the current penalty factor from deteriorating the convergence efficiency. In the second stage, the penalty factor is updated according to the current primal residual and dual residual values. After the penalty factor is updated, it enters the third stage to complete the update of the Lagrange multiplier. When the iteration termination condition is met, the update stops, realizing the function of distributed energy management for each entity, solving the problems of low system processing efficiency and low flexibility, being able to protect the information privacy of the wind farm, hydrogen energy system and building, improving the security and reliability of the energy system, and promoting the sustainable development of energy.

[0213] Example 5: Please refer to Figure 1 , a collaborative operation negotiation method for a multi-factor integrated multi-agent energy system, and the collaborative operation negotiation method is as follows:

[0214] S1: Construct the MAES-WHB model;

[0215] S2: Based on the NB theory, construct the NB model of the MAES-WHB model;

[0216] S3: Decompose the NB model into CPS and PAS, substitute the traded electricity obtained by solving CPS into the NB model, introduce the contribution degree as the bargaining power, and then construct an asymmetric PAS model;

[0217] S4: Based on the ADMM theory, decompose CPS and PAS to construct a distributed model for each entity;

[0218] S5: Improve the standard ADMM and propose the TPCA-ADMM algorithm for solving CPS and PAS;

[0219] Among them, MAES-WHB is a multi-agent energy system integrating a wind farm, a hydrogen energy system and a building, MAES is a multi-agent energy system, NB is Nash negotiation, ADMM is the alternating direction multiplier method, and the TPCA-ADMM algorithm is a three-stage prediction-correction accelerated alternating direction multiplier method algorithm improved based on the ADMM algorithm;

[0220] The specific content of S1 is as follows:

[0221] S11: Calculate the revenue and cost of the wind farm to construct the objective function of the wind farm;

[0222] S12: Calculate the cost of the hydrogen energy system and consider the power balance constraint to construct the objective function of the hydrogen energy system;

[0223] S13: Calculate the cost of the building and consider the power balance constraint to construct the objective function of the building;

[0224] S14: Construct the MAES-WHB model based on the wind farm model, the hydrogen energy system model, and the building model;

[0225] The objective function of the wind farm is:

[0226] ;

[0227] Where, is the overall profit of the wind farm, is the income obtained by the wind farm from selling electricity to the hydrogen energy system, is the income obtained by the wind farm from selling electricity to the building, is the income obtained by the wind farm from selling electricity to the PG, is the maintenance cost of the wind farm, is the transmission fee of the wind farm;

[0228] The objective function of the hydrogen energy system is:

[0229] ;

[0230] Where, is the total cost of the hydrogen energy system, is the electricity fee paid by the hydrogen energy system to the PG, is the maintenance cost of the hydrogen energy system;

[0231] The objective function of the building is:

[0232] ;

[0233] Where, is the total cost of the building, is the electricity fee paid by the building to the PG, is the maintenance cost of the building;

[0234] Furthermore, by constructing a wind farm model, a hydrogen energy system model, and a building model, the integration of wind power, hydrogen power, and buildings is achieved. The wind farm model directly sells electricity to the hydrogen energy system and buildings, and sells the excess electricity to the PG. The construction of the hydrogen energy system model calculates the consumption of the main equipment in the hydrogen production system, including the hydrogen production volume of the electrolyzer during stable operation, the constraint conditions that the electrolyzer operation needs to meet, the operation mode of the compressor when compressing hydrogen into high-pressure hydrogen, the hydrogen storage volume calculated through the internal air pressure of the hydrogen storage tank, and the operation mode of the BS. The building model analyzes data on the heat balance constraint of wall ij, the heat balance of area n, the operation power of the air conditioner, and the comfort temperature limit of users, realizing the function of detecting data in each link during the operation of the wind farm, hydrogen energy system, and building, solving the problems of energy loss and interest disputes caused by untimely and inaccurate information collection of each part of the system, being able to accurately collect real-time information of each part of the system, improving the flexibility and accuracy of the system, enhancing the working efficiency of the system, and reducing the contradictions among various entities.

[0235] The working principle is to integrate wind power, hydrogen power, and buildings by constructing a wind farm model, a hydrogen energy system model, and a building model. The wind farm model directly sells electricity to the hydrogen energy system and buildings, and sells the excess electricity to the PG. By calculating and solving the revenue of the wind farm, the maintenance cost and transmission fees of the wind farm, combined with the constraints that the wind farm output needs to meet, with the goal of improving the overall profit of the wind farm, the objective function of the wind farm is obtained as , The construction of the hydrogen energy system model includes the hydrogen production volume of the electrolyzer in a stable operation state, the operation mode of the compressor, the hydrogen storage volume indicated by the internal air pressure of the hydrogen tank, and the operation mode of the BS. The cost of the hydrogen energy system model includes the cost of purchasing electricity from the wind farm and the PG, and the maintenance cost mainly for the electrolyzer and the BS. Analyzing the power balance constraint that the hydrogen energy system operation needs to meet, with the goal of reducing the overall cost, the objective function of the hydrogen energy system is obtained as , The building model analyzes the heat balance constraint of wall ij, the heat balance of area n, the operation power of the air conditioner, and the comfort temperature limit of users, and the cost of the building model includes the cost of purchasing electricity from the wind farm and the PG, and the maintenance cost mainly for the air conditioner and the base load. Combining the power balance constraint that the building operation needs to meet, with the goal of reducing the overall cost, the objective function of the building is obtained as , The wind farm model, hydrogen energy system model, and building model are constructed;

[0236] Through the application of the NB theory, the NB model of the MAES-WHB model is obtained by combining the wind farm model, the hydrogen energy system model, and the building model. The MAES-WHB model is an alliance of the wind farm, the hydrogen energy system, and the building, and its CPS is , in the CPS, the transaction amounts I W2H and I W2B can offset each other. By solving the CPS, the optimal transaction power and the operating powers of the electrolyzer in the hydrogen energy system, the BS, and the air conditioner in the building are obtained. By calculating the total transaction power within the operating cycle, a mapping function that can quantify the contribution degree of each entity to the transaction is constructed. Each entity participating in the power transaction will obtain a contribution degree. The more power an entity trades, the higher its corresponding contribution degree. When the transaction powers are the same, the contribution degree of the power-selling entity is higher than that of the power-buying entity. By introducing the contribution degree as the bargaining power and substituting the optimal transaction power obtained from solving the CPS into the NB model, an asymmetric PAS model is constructed , and the optimal transaction electricity price is obtained by solving the PAS and In 's case, the profit distribution is symmetric;

[0237] Convert the CPS into a minimization problem. Based on ADMM, decompose the augmented Lagrangian function of the CPS into distributed optimization models of the wind farm, the hydrogen energy system, and the building. When and , it is considered that the wind farm, the hydrogen energy system, and the building reach an agreement on the transaction power. Take the logarithm of the asymmetric PAS model and convert it into a minimization problem. Based on ADMM, decompose the augmented Lagrangian function of the asymmetric PAS model into distributed optimization models of the wind farm, the hydrogen energy system, and the building. When and , it is considered that each entity reaches an agreement on the electricity price;

[0238] Solve through the TPCA-ADMM algorithm improved based on the ADMM algorithm. In the first stage, calculate the original residual and the dual residual through and , and then calculate and through and . When and are less than the set value and the original residual and the dual residual change in a decreasing or acceptable direction, the penalty factor remains unchanged and enters the third stage. Update the Lagrange multiplier by introducing a correction mechanism. If and If it is greater than or equal to the set value, enter the second stage to update the penalty factor to prevent the current penalty factor from deteriorating the convergence efficiency. In the second stage, update the penalty factor according to the current original residual and dual residual values. After the penalty factor is updated, enter the third stage to complete the update of the Lagrange multiplier, and stop the update when the iteration termination condition is met.

[0239] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be construed as limiting the claimed rights.

Claims

1. A collaborative operation negotiation method for a multi-factor integrated multi-agent energy system, characterized in that: The collaborative operation negotiation method is as follows: S1: Construct the MAES-WHB model; S2: Based on the NB theory, construct the NB model of the MAES-WHB model; S3: Decompose the NB model into CPS and PAS, substitute the transaction power obtained by solving CPS into the NB model, introduce the contribution degree as the bargaining power, and then construct an asymmetric PAS model; S4: Based on the ADMM theory, decompose CPS and PAS to construct the distributed models of each subject; S5: Improve the standard ADMM, and propose the TPCA-ADMM algorithm for solving CPS and PAS; Among them, MAES-WHB is a multi-agent energy system integrating a wind farm, a hydrogen energy system, and a building. MAES is a multi-agent energy system, NB is Nash negotiation, ADMM is the alternating direction multiplier method, and the TPCA-ADMM algorithm is a three-stage prediction-correction accelerated alternating direction multiplier method algorithm improved based on the ADMM algorithm; The specific content of S1 is as follows: S11: Calculate the revenue and cost of the wind farm and construct the objective function of the wind farm; S12: Calculate the cost of the hydrogen energy system and consider the power balance constraint to construct the objective function of the hydrogen energy system; S13: Calculate the cost of the building and consider the power balance constraint to construct the objective function of the building; S14: Based on the wind farm model, hydrogen energy system model, and building model, construct the MAES-WHB model; The NB model is: Among them, I W is the overall profit of the wind farm, is the profit of the wind farm before cooperation, C B is the total cost of the building, is the cost of the hydrogen energy system before cooperation, C H is the total cost of the hydrogen energy system, is the cost of the building before cooperation; The CPS of the NB model is: max(I W +C H +C B ); The specific content of S3 is as follows: S31: Calculate the overall transaction power of the operation period; S32: Construct a mapping function to quantify the contribution degree of each subject to the transaction; S33: Solve the CPS of the NB model to obtain the transaction power; S34: Substitute the transaction power into the NB model to construct an asymmetric PAS model; The overall transaction power of the operation period is: Among them, T is the operating cycle, t is the moment, and P W2H is the electricity quantity sold by the wind farm to the hydrogen energy system within the operating cycle, is the electricity quantity sold by the wind farm to the hydrogen energy system at time t, and P W2B is the electricity quantity sold by the wind farm to the building within the operating cycle, is the electricity quantity sold by the wind farm to the building at time t; The method for constructing a mapping function to quantify the contribution degree of each subject to the transaction is: Among them, d W is the contribution degree of the wind farm, d H is the contribution degree of the hydrogen energy system, d B is the contribution degree of the building, is a constant used to increase the contribution degree of the wind farm; The asymmetric PAS model is: The specific content of S4 is as follows: S41: Introduce a new variable and Convert the CPS into a minimization problem; S42: Based on ADMM, decompose the augmented Lagrangian function of CPS into the distributed optimization models of the wind farm, hydrogen energy system, and building; S43: Introduce a new variable and Take the logarithm of the asymmetric PAS model and transform it into a minimization problem; S44: Based on ADMM, decompose the augmented Lagrangian function of the asymmetric PAS model into the distributed optimization models of the wind farm, hydrogen energy system, and building; Among them, represents the amount of electricity that the hydrogen energy system hopes to purchase from the wind farm, represents the amount of electricity that the building hopes to purchase from the wind farm, represents the electricity price that the hydrogen energy system hopes to pay to the wind farm, represents the electricity price that the building hopes to pay to the wind farm, is the electricity price between the wind farm and the hydrogen energy system at time t, is the electricity price between the wind farm and the building at time t; The augmented Lagrangian function of CPS is: Among them, I W2G is the revenue obtained by the wind farm from selling electricity to the PG, C WM is the maintenance cost of the wind farm, C WW is the transmission fee of the wind farm, C G2H is the electricity fee paid by the hydrogen energy system to the PG, C HM is the maintenance cost of the hydrogen energy system, C G2B is the electricity fee paid by the building to the PG, C BM is the maintenance cost of the building, and are the Lagrange multipliers, ρ W2H and ρ W2B are the penalty factors; The augmented Lagrangian function of the asymmetric PAS is: wherein, and are Lagrange multipliers, and ψ W2H and ψ W2B are penalty factors.

2. The collaborative operation negotiation method for a multi-factor integrated multi-agent energy system according to claim 1, characterized in that: The objective function of the wind farm is: max I W = I W2H + I W2B + I W2G - C WM - C WW ; Among them, I W2H is the income obtained by the wind farm from selling electricity to the hydrogen energy system, and I W2B is the income obtained by the wind farm from selling electricity to buildings; The objective function of the hydrogen energy system is: max C H = -(I W2H + C G2H + C HM ); The objective function of the building is: maxC B = -(I W2B + C G2B + C BM ); 3. The collaborative operation negotiation method of a multi-factor integrated multi-agent energy system according to claim 1, characterized in that: The specific content of S5 is as follows: S51: In the first stage, calculate the primal residual, dual residual, and the difference between the primal residual and the dual residual; S52: Compare the difference with the set value. If the difference is less than the set value, the penalty factor remains unchanged. If the difference is not less than the set value, enter the second stage to update the penalty factor; S53: In the third stage, introduce a prediction-correction mechanism to update the Lagrange multiplier; S54: When the iteration termination condition is met, stop the solution.

4. The collaborative operation negotiation method of a multi-factor integrated multi-agent energy system according to claim 3, characterized in that: The formula for calculating the primal residual and the dual residual is: Among them, is the original residual, is the electricity quantity that the hydrogen energy system / building hopes to purchase from the wind farm, is the electricity quantity that the wind farm hopes to sell to the hydrogen energy system / building; Among them, is the dual residual, the amount of electricity that the hydrogen energy system / building hopes to purchase from the wind farm; The formula for calculating the difference between the primal residual and the dual residual is: Among them, is the original residual; Among them, is the dual residual.

5. The collaborative operation negotiation method of a multi-factor integrated multi-agent energy system according to claim 3, characterized in that: The update of the penalty factor in S52 is expressed as: Among them, and are penalty factors, and θ is a constant; The updated Lagrange multiplier of S53 is expressed as: Among them, is an intermediate parameter in the Lagrange multiplier update process, is the amount of electricity that the hydrogen energy system / building hopes to purchase from the wind farm, is the Lagrange multiplier, is the amount of electricity that the wind farm hopes to sell to the hydrogen energy system / building; where ζ is a constant; where χ is a predictor-corrector factor.

6. The collaborative operation negotiation method of a multi-factor integrated multi-agent energy system according to claim 3, characterized in that: The iteration termination condition is: where k is the number of iterations, and k max is the maximum number of iterations.

Images