Method for virtual power plant to participate in joint market bidding based on master-slave non-cooperative game

Abstract

The invention provides a bidding method for virtual power plants participating in a joint market based on a master-slave non-cooperative game. The bidding method comprises the following steps: constructing a double-layer game framework in which multiple virtual power plants participate in a day-ahead joint market; carrying out Wasserstein distance uncertainty modeling on the wind power, the photovoltaic power and the electric load power; based on the multi-source load power uncertainty set, an upper-layer multi-virtual power plant two-stage distribution robust bidding model is constructed, and the bidding model comprises a first objective function with the minimum sum of pre-scheduling cost for making a bidding strategy in a day-ahead stage and rescheduling cost expectation in an intra-day stage and a corresponding first class of constraint conditions; constructing a lower-layer joint market clearing model; the clearing model comprises a second objective function with the minimum day-ahead electricity-peak regulation-hydrogen combined market energy purchase cost and a corresponding second class of constraint conditions; performing equivalent conversion on the bidding model and the clearing model; and solving the converted bidding model and the converted clearing model to obtain a bidding strategy.

Description

Technical Field

[0001] The present invention relates to the field of power grid technology, and in particular to, but is not limited to, a method for a virtual power plant to participate in a joint market bidding based on a master-slave non-cooperative game. Background Art

[0002] As market reforms gradually deepen, a large amount of social capital has poured into the energy market to engage in fierce competition. Market participants are becoming increasingly diversified and belong to different interest groups. Each of them integrates diversified resources to optimize its own operating goals. In this context, the decision-making process needs to fully consider the mutual influence between the various subjects. Virtual power plants (VPPs) can get rid of the constraints of geographical space and power grid topology, aggregate widely distributed distributed resources based on advanced communication technology for centralized management, and optimize the scheduling of distributed resources with the help of energy market transactions. It has become a research hotspot. However, with the continuous diversification of VPP aggregation subjects, how to effectively coordinate the complementarity of these multi-source operating subjects, deeply explore and make full use of the flexible potential of massive distributed resources, and thus improve their market competitiveness has become an important problem that needs to be solved urgently.

[0003] At present, a lot of research has been conducted on the participation of multiple VPPs in market transactions. The main trading model is to use operators as leaders to guide energy transactions through price optimization. Each VPP acts as a follower and optimizes the internal aggregated units and loads according to the published prices. However, this model has a significant limitation, that is, it does not consider the impact of virtual power plants on market clearing results, resulting in VPPs participating in market operations only as price takers, which is essentially still VPP internal optimization.

[0004] With the development and application of power-to-hydrogen technology, hydrogen energy storage, as a new energy storage method, plays an important role in multi-energy virtual power plants. Hydrogen energy storage uses surplus electricity to convert it into hydrogen through water electrolysis for storage, and consumes hydrogen to generate electricity through hydrogen fuel cells when electricity is insufficient. Compared with other energy storage methods such as battery energy storage and pumped storage, hydrogen energy storage has more advantages in large-scale and long-term energy storage. However, current research fails to take into account the economic benefits of seasonal hydrogen energy storage participating in electricity market bidding. Summary of the invention

[0005] Based on the problems in the related art, an embodiment of the present invention provides a virtual power plant participating in joint market bidding based on a master-slave non-cooperative game.

[0006] The technical solution of the embodiment of the present invention is achieved as follows:

[0007] In a first aspect, an embodiment of the present invention provides a method for a virtual power plant to participate in a joint market bidding based on a master-slave non-cooperative game, the method comprising:

[0008] Construct a two-tier game framework for multiple virtual power plants to participate in the day-ahead joint market;

[0009] The uncertainty modeling of wind power, photovoltaic power and electric load power is carried out by using Wasserstein distance to obtain the uncertainty set of multi-source load power.

[0010] Based on the multi-source load power uncertainty set, a two-stage distributed robust bidding model for upper-layer multi-virtual power plants is constructed; the two-stage robust distributed bidding model includes a first objective function of minimizing the sum of the pre-dispatch cost of formulating the bidding strategy in the day-ahead stage and the expected re-dispatch cost in the intra-day stage; and a first type of constraint condition of the first objective function is constructed accordingly;

[0011] Constructing a lower-level joint market clearing model; the clearing model includes a second objective function of minimizing the energy purchase cost of the day-ahead electricity-peak-shaving-hydrogen joint market; and correspondingly constructing a second type of constraint condition of the second objective function;

[0012] The two-stage blue-stick bidding model and the lower-layer joint clearing model are respectively equivalently transformed to obtain a transformed bidding model and a transformed clearing model; the complementary relaxation conditions in the transformed bidding model are linearized to obtain processed complementary relaxation conditions;

[0013] Based on the processed complementary relaxation conditions, the first type of constraints and the second type of constraints, the converted bidding model and the converted clearing model are solved to obtain a bidding strategy.

[0014] In a second aspect, an embodiment of the present invention provides a virtual power plant participating in a joint market bidding device based on a master-slave non-cooperative game, the device comprising:

[0015] A building module for constructing a two-layer game framework for multiple virtual power plants to participate in the day-ahead joint market;

[0016] Modeling module, used to perform Wasserstein distance uncertainty modeling on wind power, photovoltaic power and electric load power, and obtain the multi-source load power uncertainty set;

[0017] The construction module is further used to construct an upper-layer multi-virtual power plant two-stage distributed robust bidding model based on the multi-source load power uncertainty set; the two-stage robust distributed bidding model includes a first objective function of minimizing the sum of the pre-dispatch cost of formulating the bidding strategy in the day-ahead stage and the expected re-dispatch cost in the intra-day stage; and correspondingly constructs a first type of constraint condition of the first objective function;

[0018] The construction module is also used to construct a lower-level joint market clearing model; the clearing model includes a second objective function of minimizing the energy purchase cost of the day-ahead electricity-peak-shaving-hydrogen joint market; and a second type of constraint condition for constructing the second objective function accordingly;

[0019] A transformation module, used to perform equivalent transformation on the two-stage robust distribution bidding model and the lower-layer joint clearing model, respectively, to obtain a transformed bidding model and a transformed clearing model; linearize the complementary relaxation conditions in the transformed bidding model to obtain processed complementary relaxation conditions;

[0020] A solution module is used to solve the converted bidding model and the converted clearing model based on the processed complementary relaxation conditions, the first type of constraints and the second type of constraints to obtain a bidding strategy.

[0021] In some embodiments, the modeling module is also used to determine the actual daily output of wind power and photovoltaic power based on the day-ahead predicted output of wind power and photovoltaic power and the power error vector of wind power and photovoltaic power source; determine the actual daily output of electric load based on the day-ahead predicted output of electric load and the power error vector of electric load source; calculate the first Wasserstein distance between the empirical distribution and the actual distribution of wind power output, the second Wasserstein distance between the empirical distribution and the actual distribution of photovoltaic output, and the third Wasserstein distance between the empirical distribution and the actual distribution of electric load respectively; and determine the wind power uncertainty set, the photovoltaic power uncertainty set and the electric load uncertainty set based on the first Wasserstein distance, the second Wasserstein distance and the third Wasserstein distance respectively.

[0022] In some embodiments, the building block is also used for expressing the first objective function as:

[0023]

[0024] Where: U i is the net operating cost of VPPi participating in various energy markets during the day-ahead bidding phase; f(x i ,δ i ) is the redispatch cost of VPPi during the intraday dispatch phase; To include uncertainty The function of the probability distribution P i WT / PV / el The expectation below; sup is the supremum; π E,i,t , π S,i,t and π H,i,t are the clearing prices of VPPi in the electricity market, peak load market and hydrogen market at time t respectively; PE,i,t , P S,i,t and P H,i,t are the winning bid powers of VPPi in the electricity market, peak load regulation market and hydrogen energy market at time t; C i is the operating cost of VPPi.

[0025] In some embodiments, the building module is also used to build 1) gas turbine constraints

[0026]

[0027] Where: and are the electrical output and thermal output of the gas turbine GT at time t; ηp ,t and ηq ,t are the electrical efficiency and thermal efficiency of GT at time t respectively; is the gas consumption power of GT at time t; H gas It is the low calorific value of natural gas; and They are the upper and lower limits of GT output respectively; and They are the upper and lower limits of GT climbing power respectively;

[0028] 2) Constraints of Hydrogen Production by Electricity

[0029]

[0030] Where: are the P2H power consumption and hydrogen production power at time t respectively; η P2H is the conversion efficiency of P2H; and They are the upper and lower limits of P2H power consumption respectively; and They are the upper and lower limits of the P2H power consumption ramp, respectively;

[0031] 3) Hydrogen fuel cell constraints

[0032]

[0033] Where: is the hydrogen consumption power of HFC at time t; and are the electrical power and thermal power of HFC at time t respectively; η HFC,p,t and η HFC,q,t are the electrical efficiency and thermal efficiency of HFC at time t respectively; and are the upper and lower limits of HFC electric power, respectively; They are the upper and lower limits of HFC electric power ramping, respectively;

[0034] 4) Carbon capture equipment constraints

[0035]

[0036] Where: is the capture energy consumption required for CCS to capture CO2 at time t; is the energy consumption maintained at time t, which is a fixed value that is independent of the operating state of the CCS; is the operating energy consumption at time t, which changes with the operating state of CCS and is positively correlated with the ability of CCS to capture CO2; c The energy consumption required to capture a unit of CO2 for CCS; They are the upper and lower limits of CCS power consumption respectively;

[0037] 5) Constraints for electric energy storage electric vehicles, thermal energy storage electric vehicles, and hydrogen fuel cell electric vehicles

[0038]

[0039] Where: superscript m represents the energy type, which includes ES, TS, EV and HEV; is the capacity of the energy storage device at time t; φ m is the self-loss coefficient of energy storage device m; and They are the charging efficiency and discharging efficiency of the energy storage device respectively; and are the charging power and discharging power of the energy storage device respectively; and are the maximum and minimum storage capacities of the energy storage device m, respectively; is a 0-1 variable, used to indicate whether the energy storage device m can be charged and discharged at the same time. Indicates that the energy storage device is charged. Indicates that the energy storage device releases energy; and They represent the upper limits of the charging power and discharging power of the energy storage device m respectively; is the initial capacity of the energy storage device m; The remaining capacity of the energy storage device m after running for one cycle;

[0040] 6) Constraints on the energy balance of hydrogen storage day

[0041]

[0042] 7) Hydrogen storage at the beginning of each season

[0043]

[0044] 8) Hydrogen charge status after annual storage balance

[0045]

[0046] 9) SHS initial and year-end capacity constraints

[0047]

[0048] 10) Energy state constraints of SHS

[0049]

[0050] 11) Hydrogen charging / release rate constraints

[0051]

[0052] Where: parameter Π s The number of energy cycles s representing the typical days in the season;

[0053] 12) Time-shiftable electrical load

[0054]

[0055] Where: is the load power that can be shifted at time t; α i,TSE is the ratio of time-shiftable load to total load;

[0056] 13) Interruptible power load

[0057]

[0058] Where: They are the upper and lower limits of GT’s bidding price in the electricity market;

[0059] 14) Power balance constraints

[0060]

[0061] In some embodiments, the building block is also used for the second objective function to be expressed as:

[0062]

[0063] Where: U MA R is the total cost of energy purchase by the joint market operator; E,i,t , R S,i,t and R H,i,t are the energy purchase costs of the market operator in the electricity market, peak load market and hydrogen energy market at time t respectively; and They are the bidding prices of GT, electric energy storage ES, electric vehicles EV and renewable energy in the electricity market in VPPi; is the bidding price of GT in the peak load market in VPPi; and They are the bidding prices of P2H, SHS and hydrogen fuel cell electric vehicles HEV in the hydrogen energy market in VPPi; and are the bidding powers of GT, ES, EV and RE in the electric energy market in VPPi respectively; is the bidding power of GT in the peak load market in VPPi; and They are the bidding powers of P2H, SHS and HEV in VPPi in the hydrogen energy market.

[0064] In some embodiments, the building module is also used to build 1) power balance constraint conditions

[0065]

[0066] Where: and are the energy demands of the electricity market, peak load regulation market and hydrogen energy market at time t respectively;

[0067] 2) VPP bidding power constraints

[0068]

[0069] Where: and These are the upper limits of the winning bids for GT, ES, EV and RE in the electric energy market; The power cap for GT's winning bid in the peak-shaving market; and These are the winning power caps for GT, SHS and HEV in the hydrogen energy market;

[0070] is the dual variable of the corresponding constraint.

[0071] In some embodiments, the conversion module is further used to adopt duality theory to convert the supremum problem of the worst expected distribution in the real-time stage of the bidding model into the infimum problem, and obtain the single-layer minimum target optimization:

[0072]

[0073] Introduce auxiliary variable w n It can be simplified to:

[0074]

[0075] Since the supremum of the constraint condition in the above formula can only be obtained on the boundary of the uncertain variable, the first objective function is finally transformed into a single-layer linear programming model, and the transformed bidding model is expressed as:

[0076]

[0077] Where: i is the dual factor; δ i , They represent the uncertain variable sample set {δ i,1 , δ i,2 , …, δ i,N}'s infimum and supremum;

[0078] The Lagrangian function of the clearing model is expressed as:

[0079]

[0080] Taking partial derivatives of the variables, we can obtain the Lagrangian equilibrium condition of the clearing model:

[0081]

[0082] The relaxation condition corresponding to the inequality in the clearing model is:

[0083]

[0084] In some embodiments, the solution module is also used in step 1: before the bidding begins, VPPi fully integrates internal resources based on the predicted day-ahead market information and the bidding plans of other VPPs, and determines the first round of bidding plans as follows, taking the electric energy market as an example:

[0085] Step 2: After the bidding begins, VPPi will select the bidding plan of the k-1th round according to the bidding plan of other VPPj (j=1, 2, ..., n; j≠i) Solve the upper bidding model to formulate its own bidding plan for round k (k ≥ 2) After all VPPs have given their bids for the kth round, go to step 3;

[0086] Step 3: Each VPP takes the minimum operating cost as the optimization goal, repeats step 2, and repeats the quotation to the market operator; after multiple rounds of game, if no VPP changes its bidding plan, go to step 4; if there are still VPPs trying to change their bidding strategies and submit a bidding plan for the k+1th round, go to step 5;

[0087] Step 4: If all VPPs no longer try to profit by unilaterally changing their bidding plans, a Nash equilibrium is formed in the market. After the bidding is over, the market operators will clear out the market in a unified manner according to the results of the market game.

[0088] Step 5: Determine whether the upper limit of the game rounds has been reached; if the upper limit of the game rounds has not been reached at this time, return to step 1; if it has been reached, the market operator will no longer accept the new round of bidding proposals from each VPP; after the bidding is over, the market operator will force clearing according to the agreed market trading rules.

[0089] In the third aspect, an embodiment of the present invention provides a device for a virtual power plant to participate in a joint market bidding based on a master-slave non-cooperative game, comprising: a memory for storing executable instructions; and a processor for executing the executable instructions stored in the memory to implement the above-mentioned method for a virtual power plant to participate in a joint market bidding based on a master-slave non-cooperative game.

[0090] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium storing executable instructions for causing a processor to execute the executable instructions to implement the above-mentioned method for virtual power plants to participate in joint market bidding based on the master-slave non-cooperative game.

[0091] The embodiment of the present invention provides a method for virtual power plants to participate in joint market bidding based on master-slave non-cooperative game, that is, a distributed low-carbon optimization method for multiple virtual power plants based on two-stage carbon trading; firstly, a master-slave game day-ahead market two-layer bidding model with multiple VPPs as leaders and market operators as followers is constructed; the upper-layer leader is a VPP operator participating in multiple markets, and with the goal of minimizing its operating costs, combined with the uncertainty set based on Wasserstein distance, a two-stage distributed robust optimization model of VPP day-ahead bidding-intraday scheduling is constructed; the lower layer clears the market operator with the goal of minimizing energy purchases in the electricity-peak-shaving-hydrogen joint market; secondly, by introducing seasonal hydrogen energy storage, the mismatch between renewable energy generation and energy demand in time scale can be alleviated, and long-term storage and utilization of hydrogen energy across seasons and electricity across energy forms can be realized, thereby reducing system operating costs and promoting local consumption of renewable energy; the model can provide economic and low-carbon bidding strategies in a multi-market environment, and provides ideas for improving VPP revenue. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] Figure 1 It is a flow chart of a method for a virtual power plant to participate in a joint market bidding based on a master-slave non-cooperative game provided by an embodiment of the present invention;

[0093] Figure 2 It is a structural schematic diagram of a double-layer game framework for multiple virtual power plants to participate in a day-ahead joint market provided by an embodiment of the present invention;

[0094] Figure 3 It is a schematic diagram of the iterative solution process of the double-layer non-cooperative game model provided by an embodiment of the present invention;

[0095] Figure 4 It is a schematic diagram of the clearing power of each VPP in the electric energy market provided by an embodiment of the present invention;

[0096] Figure 5 It is a schematic diagram of the clearing volume of each VPP in the peak load market provided by an embodiment of the present invention;

[0097] Figure 6 It is a schematic diagram of the clearing volume of each VPP in the hydrogen energy market provided by an embodiment of the present invention;

[0098] Figure 7 A schematic diagram of the composition structure of a virtual power plant participating in a joint market bidding device based on a master-slave non-cooperative game provided in an embodiment of the present invention;

[0099] Figure 8 A schematic diagram of the composition structure of a virtual power plant participating in a joint market bidding device based on a master-slave non-cooperative game provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0100] In order to make the purpose, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present invention. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present invention.

[0101] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict. Unless otherwise defined, all technical and scientific terms used in the embodiments of the present invention have the same meaning as those commonly understood by those skilled in the art to which the embodiments of the present invention pertain. The terms used in the embodiments of the present invention are only for the purpose of describing the embodiments of the present invention and are not intended to limit the present invention.

[0102] The embodiment of the present invention provides a method for virtual power plants to participate in joint market bidding based on master-slave non-cooperative game, see Figure 1 , Figure 1 is a flowchart of a method for virtual power plants to participate in joint market bidding based on master-slave non-cooperative game provided by an embodiment of the present invention, which will be combined with Figure 1 The steps shown are explained.

[0103] Step S110, constructing a two-layer game framework for multiple virtual power plants to participate in the day-ahead joint market.

[0104] Step S120, performing Wasserstein distance uncertainty modeling on wind power, photovoltaic power and electric load power to obtain a multi-source load power uncertainty set.

[0105] Step S130, based on the multi-source load power uncertainty set, construct an upper-level two-stage distributed robust bidding model for multiple virtual power plants; the two-stage robust distributed bidding model includes a first objective function of minimizing the sum of the pre-dispatching cost of formulating the bidding strategy in the day-ahead stage and the expected re-dispatching cost in the intra-day stage; and correspondingly constructs the first type of constraint condition of the first objective function.

[0106] Step S140, constructing a lower-level joint market clearing model; the clearing model includes a second objective function of minimizing the cost of purchasing energy in the day-ahead electricity-peak-shaving-hydrogen joint market; and correspondingly constructing a second type of constraint condition for the second objective function.

[0107] Step S150, respectively perform equivalent transformation on the two-stage blue-stick bidding model and the lower-layer joint clearing model to obtain a transformed bidding model and a transformed clearing model respectively; linearize the complementary relaxation conditions in the transformed bidding model to obtain processed complementary relaxation conditions.

[0108] Step S160, based on the processed complementary relaxation conditions, the first type of constraints and the second type of constraints, the converted bidding model and the converted clearing model are solved to obtain a bidding strategy.

[0109] The embodiment of the present invention provides a method for virtual power plants to participate in joint market bidding based on master-slave non-cooperative game, that is, a distributed low-carbon optimization method for multiple virtual power plants based on two-stage carbon trading; firstly, a master-slave game day-ahead market two-layer bidding model with multiple VPPs as leaders and market operators as followers is constructed; the upper-layer leader is a VPP operator participating in multiple markets, and with the goal of minimizing its operating costs, combined with the uncertainty set based on Wasserstein distance, a two-stage distributed robust optimization model of VPP day-ahead bidding-intraday scheduling is constructed; the lower layer clears the market operator with the goal of minimizing energy purchases in the electricity-peak-shaving-hydrogen joint market; secondly, by introducing seasonal hydrogen energy storage, the mismatch between renewable energy generation and energy demand in time scale can be alleviated, and long-term storage and utilization of hydrogen energy across seasons and electricity across energy forms can be realized, thereby reducing system operating costs and promoting local consumption of renewable energy; the model can provide economic and low-carbon bidding strategies in a multi-market environment, and provides ideas for improving VPP revenue.

[0110] In some embodiments, the above step S120 may be implemented by steps S121 to S123:

[0111] Step S121, based on the day-ahead predicted output of wind power and photovoltaic power and the power error vector of wind power and photovoltaic power source load, determine the actual output of wind power and photovoltaic power during the day; based on the day-ahead predicted output of electric load and the power error vector of electric load source load, determine the actual output of electric load during the day.

[0112] Step S122, respectively calculating a first Wasserstein distance between the empirical distribution and the actual distribution of wind power output, a second Wasserstein distance between the empirical distribution and the actual distribution of photovoltaic output, and a third Wasserstein distance between the empirical distribution and the actual distribution of electric load.

[0113] Step S123: determining the wind power uncertainty set, the photovoltaic power uncertainty set, and the electric load uncertainty set based on the first Wasserstein distance, the second Wasserstein distance, and the third Wasserstein distance, respectively.

[0114] In some embodiments, the first objective function in step S130 is expressed as:

[0115]

[0116] Where: U i is the net operating cost of VPPi participating in various energy markets during the day-ahead bidding phase; f(x i ,δ i ) is the redispatch cost of VPPi during the intraday dispatch phase; To include uncertainty The function of the probability distribution P i WT / PV / el The expectation below; sup is the supremum; π E,i,t , π S,i,t and π H,i,t are the clearing prices of VPPi in the electricity market, peak load market and hydrogen market at time t respectively; P E,i,t , P S,i,t and P H,i,t are the winning bid powers of VPPi in the electricity market, peak load regulation market and hydrogen energy market at time t; C i is the operating cost of VPPi.

[0117] In some embodiments, the first type of constraint condition of the first objective function is:

[0118] 1) Gas turbine constraints

[0119]

[0120] Where: and are the electrical output and thermal output of the gas turbine GT at time t; ηp ,t and ηq ,t are the electrical efficiency and thermal efficiency of GT at time t respectively; is the gas consumption power of GT at time t; H gas It is the low calorific value of natural gas; and They are the upper and lower limits of GT output respectively; and They are the upper and lower limits of GT climbing power respectively;

[0121] 2) Constraints of Hydrogen Production by Electricity

[0122]

[0123] Where: are the P2H power consumption and hydrogen production power at time t respectively; η P2H is the conversion efficiency of P2H; and They are the upper and lower limits of P2H power consumption respectively; and They are the upper and lower limits of the P2H power consumption ramp, respectively;

[0124] 3) Hydrogen fuel cell constraints

[0125]

[0126] Where: is the hydrogen consumption power of HFC at time t; and are the electrical power and thermal power of HFC at time t respectively; η HFC,p,t and η HFC,q,t are the electrical efficiency and thermal efficiency of HFC at time t respectively; and are the upper and lower limits of HFC electric power, respectively; They are the upper and lower limits of HFC electric power ramping, respectively;

[0127] 4) Carbon capture equipment constraints

[0128]

[0129] Where: is the capture energy consumption required for CCS to capture CO2 at time t; is the energy consumption maintained at time t, which is a fixed value that is independent of the operating state of the CCS; is the operating energy consumption at time t, which changes with the operating state of CCS and is positively correlated with the ability of CCS to capture CO2;c The energy consumption required to capture a unit of CO2 for CCS; They are the upper and lower limits of CCS power consumption respectively;

[0130] 5) Constraints for electric energy storage electric vehicles, thermal energy storage electric vehicles, and hydrogen fuel cell electric vehicles

[0131]

[0132] Where: superscript m represents the energy type, which includes ES, TS, EV and HEV; is the capacity of the energy storage device at time t; φ m is the self-loss coefficient of energy storage device m; and They are the charging efficiency and discharging efficiency of the energy storage device respectively; and are the charging power and discharging power of the energy storage device respectively; and are the maximum and minimum storage capacities of the energy storage device m, respectively; is a 0-1 variable, used to indicate whether the energy storage device m can be charged and discharged at the same time. Indicates that the energy storage device is charged. Indicates that the energy storage device releases energy; and They represent the upper limits of the charging power and discharging power of the energy storage device m respectively; is the initial capacity of the energy storage device m; The remaining capacity of the energy storage device m after running for one cycle;

[0133] 6) Constraints on the energy balance of hydrogen storage day

[0134]

[0135] 7) Hydrogen storage at the beginning of each season

[0136]

[0137] 8) Hydrogen charge status after annual storage balance

[0138]

[0139] 9) SHS initial and year-end capacity constraints

[0140]

[0141] 10) Energy state constraints of SHS

[0142]

[0143] 11) Hydrogen charging / release rate constraints

[0144]

[0145] Where: parameter Π s The number of energy cycles s representing the typical days in the season;

[0146] 12) Time-shiftable electrical load

[0147]

[0148] Where: is the load power that can be shifted at time t; α i,TSE is the ratio of time-shiftable load to total load;

[0149] 13) Interruptible power load

[0150]

[0151] Where: They are the upper and lower limits of GT’s bidding price in the electricity market;

[0152] 14) Power balance constraints

[0153]

[0154] In some embodiments, the second objective function in step S140 above is expressed as:

[0155]

[0156] Where: U MA is the total cost of energy purchase by the joint market operator; R E,i,t , R S,i,t and R H,i,t are the energy purchase costs of the market operator in the electricity market, peak load market and hydrogen energy market at time t respectively; They are the bidding prices of GT, electric energy storage ES, electric vehicles EV and renewable energy RE in the electricity market in VPPi; is the bidding price of GT in the peak load market in VPPi; They are the bidding prices of P2H, SHS and hydrogen fuel cell electric vehicles HEV in the hydrogen energy market in VPPi; and are the bidding powers of GT, ES, EV and RE in the electric energy market in VPPi respectively; is the bidding power of GT in the peak load market in VPPi; and They are the bidding powers of P2H, SHS and HEV in VPPi in the hydrogen energy market.

[0157] In some embodiments, the second type of constraint condition of the second objective function is:

[0158] 1) Power balance constraints

[0159]

[0160] Where: and are the energy demands of the electricity market, peak load regulation market and hydrogen energy market at time t respectively;

[0161] 2) VPP bidding power constraints

[0162]

[0163] Where: These are the upper limits of the winning bids for GT, ES, EV and RE in the electric energy market; The power cap for GT's winning bid in the peak-shaving market; These are the upper limits of the winning bids for GT, SHS and HEV in the hydrogen energy market;

[0164] is the dual variable of the corresponding constraint.

[0165] In some embodiments, the above step S150 can be implemented in the following manner:

[0166] Using the duality theory, the supremum problem of the worst expected distribution in the real-time stage of the bidding model is transformed into the infimum problem, and the single-layer minimum target optimization is obtained:

[0167]

[0168] Introduce auxiliary variable w n It can be simplified to:

[0169]

[0170] Since the supremum of the constraint condition in the above formula can only be obtained on the boundary of the uncertain variable, the first objective function is finally transformed into a single-layer linear programming model, and the transformed bidding model is expressed as:

[0171]

[0172] Where: i is the dual factor; δ i , They represent the uncertain variable sample set {δ i,1 , δ i,2 , …, δ i,N}'s infimum and supremum;

[0173] The Lagrangian function of the clearing model is expressed as:

[0174]

[0175] Taking partial derivatives of the variables, we can obtain the Lagrangian equilibrium condition of the clearing model:

[0176]

[0177] The relaxation condition corresponding to the inequality in the clearing model is:

[0178]

[0179] In some embodiments, the above step S160 can be implemented in the following manner:

[0180] Step 1: Before the bidding begins, VPPi fully integrates internal resources based on the predicted day-ahead market information and the bidding plans of other VPPs. Taking the electricity market as an example, the first round of bidding plans are determined as follows:

[0181] Step 2: After the bidding begins, VPPi will select the bidding plan of the k-1th round according to the bidding plan of other VPPj (j=1, 2, ..., n; j≠i) Solve the upper bidding model to formulate its own bidding plan for round k (k ≥ 2) After all VPPs have given their bids for the kth round, go to step 3;

[0182] Step 3: Each VPP takes the minimum operating cost as the optimization goal, repeats step 2, and repeats the quotation to the market operator; after multiple rounds of game, if no VPP changes its bidding plan, go to step 4; if there are still VPPs trying to change their bidding strategies and submit a bidding plan for the k+1th round, go to step 5;

[0183] Step 4: If all VPPs no longer try to profit by unilaterally changing their bidding plans, a Nash equilibrium is formed in the market. After the bidding is over, the market operators will clear out the market in a unified manner according to the results of the market game.

[0184] Step 5: Determine whether the upper limit of the game rounds has been reached; if the upper limit of the game rounds has not been reached at this time, return to step 1; if it has been reached, the market operator will no longer accept the new round of bidding proposals from each VPP; after the bidding is over, the market operator will force clearing according to the agreed market trading rules.

[0185] The following is an explanation of an exemplary application of the embodiments of the present application in a practical application scenario.

[0186] 1) Virtual power plant participation in joint market bidding model based on master-slave non-cooperative game

[0187] (1) Two-tier game framework of multiple virtual power plants (VPPs) participating in the day-ahead joint market

[0188] In the embodiment of the present invention, the day-ahead joint market includes the electricity market, the peak-shaving market and the hydrogen energy market. When the market operator releases the market demand information, each VPP submits a "quotation-volume" plan in each energy market in the form of an energy seller to participate in market transactions. Its trading behavior is based on the reasonable scheduling of internal resources, and bids are made after fully considering the bidding strategies of competitors and known market information. The present invention provides a non-cooperative two-layer game framework for multiple virtual power plants VPPs to participate in the day-ahead joint market, such as Figure 2 shown.

[0189] Depend on Figure 2 It can be seen that each VPP can effectively integrate various types of distributed energy, energy storage, load and other equipment, including wind turbines (WT), photovoltaic panels (PV), gas turbines (GT), gas boilers (GB), P2H, hydrogen fuel cells (HFC), carbon capture (CCS), electric energy storage (ES), thermal storage tanks (TS), SHS, electric vehicles (EV), hydrogen fuel cell electric vehicles (HEV), and user electrical loads and thermal loads.

[0190] (2) Uncertainty modeling based on Wasserstein distance

[0191] The source-load power error vectors are defined as Then the source load power containing uncertain variables is:

[0192]

[0193] Where: is the actual daily output of wind power and photovoltaic power at time t; The day-ahead predicted output of wind power and photovoltaic power at time t; is the actual value of the daily electric load at time t; is the day-ahead forecast value of the electric load at time t.

[0194] Taking wind power as an example, the empirical distribution of wind power forecast error is N WT Independent wind power error vector To determine, when a large number of error vectors δ are obtained, the empirical distribution will approach the true distribution The present invention uses Wasserstein distance to measure the distance between the probability distribution and the true distribution in a given variable space Ξ To measure the similarity between distributions, the expression is as follows:

[0195]

[0196] Where: They are respectively subject to the wind power experience distribution The actual distribution of wind power A random variable of is the wind power random variable The Euclidean norm between ; Experience distribution for wind power The actual distribution of wind power The joint probability distribution of ; similarly, the Wasserstein distance of photovoltaic and electric load can be expressed as:

[0197]

[0198] According to the above definition, the wind power uncertainty set based on Wasserstein distance Photovoltaic power uncertainty set Uncertain set of load power Can be expressed as:

[0199]

[0200] Where: and are the probability distribution sets of all random variable spaces Ξ for wind power, photovoltaic power and electric load respectively; taking wind power as an example, Indicates a As the center of the ball, The radius of the wind power Wasserstein ball is , and the size of the fuzzy set can be controlled by the radius to control the degree of conservatism; if the uncertainty set is set The confidence level is 1-β, so the radius of the wind power Wasserstein sphere is set to the following value to ensure that the confidence level is not less than 1-β:

[0201]

[0202] Where: D is the diameter of the random variable space Ξ, representing the maximum value of the Wasserstein distance between any pair of atoms in the space Ξ; 1-β represents the confidence of the radius; η is an auxiliary variable; is the mean of all samples; is the nth group of samples of wind power in VPPi; Representation sample With the mean The 1 norm between .

[0203] (3) Upper-level multi-virtual power plant two-stage decentralized bidding model

[0204] 3.1) First objective function

[0205] In the overall game bidding framework of the present invention for VPPs to participate in the joint market, the upper-layer VPP optimizes its bidding price according to the dynamic bidding model and passes it to the lower layer. The lower-layer joint market followers perform market clearing based on the bidding strategy of the market entities and pass the clearing price and clearing volume to the upper layer.

[0206] Based on the uncertainty characterization of source and load power, a two-stage DRO model of VPP day-ahead and day-intraday is constructed. The goal of this model is to calculate the source and load power prediction error δ i In the uncertain set Under the worst probability distribution, the expected sum of the pre-dispatching cost of formulating the bidding strategy in the day-ahead stage and the re-dispatching cost in the intra-day stage is the smallest, that is:

[0207]

[0208] Where: U i is the net operating cost of VPPi participating in various energy markets during the day-ahead bidding phase; f(x i ,δ i ) is the redispatch cost of VPPi during the intraday dispatch phase; To include uncertainty The function of the probability distribution P i WT / PV / el The expectation below is the expectation below; sup is the supremum.

[0209] In the day-ahead stage, VPP absorbs as much renewable energy as possible according to the predicted power generation, pre-dispatches the output of each unit, and participates in multi-market bidding in the form of non-cooperative game. The optimization goal of the VPPi operator in this stage is to minimize its net operating cost in the day-ahead market. As shown in formula (11), the net operating cost of each VPP consists of the following two parts: the first part is the operating cost of the VPPi operator; the second part is the income obtained by the VPPi operator in the electricity market, peak load market, and hydrogen market.

[0210]

[0211] Where: π E,i,t , π S,i,t , π H,i,t are the clearing prices of VPPi in the electricity market, peak load market, and hydrogen market at time t; P E,i,t , PS,i,t , P H,i,t are the winning bid powers of VPPi in the electricity market, peak load regulation market, and hydrogen energy market at time t; C i is the operating cost of VPPi, specifically expressed as:

[0212] C i =C i,grid +C i,car +C i,idr +C i,op +C i,cu +C i,re (12)

[0213]

[0214] Where: C i,grid is the energy purchase cost, C i,car is the carbon trading cost, C i,idr is the demand response compensation cost, C i,op is the equipment operation and maintenance cost, C i,cu is the cost of renewable energy abandonment, C i,re is the spare capacity cost. is the step-type carbon trading cost at time t; is the carbon trading base price; ξ is the price growth rate; r IE is the unit compensation price of interruptible load; ω op,x is the unit operation and maintenance cost of unit x, ω op,m is the unit operation and maintenance cost of energy storage m; unit x includes GT, P2H, HFC and CCS; r loss Penalize prices for renewable energy units that are abandoned, curtailing power for renewable energy; are the spare capacities provided by GT, P2H and HFC in VPPi at time t; ω re,GT ,ω re,P2H ,ω re,HFC They are the unit standby costs of GT, P2H and HFC respectively.

[0215] 3.2) Constraints

[0216] ①Power balance

[0217]

[0218] ② Gas turbine

[0219]

[0220] Where: is the electrical output and thermal output of GT at time t; η p,t , η q,tis the electrical efficiency and thermal efficiency of GT at time t; is the gas consumption power of GT at time t; H gas It is the low calorific value of natural gas; The upper and lower limits of GT output; It is the upper and lower limits of GT climbing power.

[0221] ③ Hydrogen production by electricity

[0222]

[0223] Where: are the P2H power consumption and hydrogen production power at time t respectively; η P2H is the conversion efficiency of P2H; The upper and lower limits of P2H power consumption; It is the upper and lower limits of P2H power consumption climbing.

[0224] ④ Hydrogen fuel cell

[0225]

[0226] Where: is the hydrogen consumption power of HFC at time t; are the electrical power and thermal power of HFC at time t respectively; η HFC,p,t , η HFC,q,t are the electrical efficiency and thermal efficiency of HFC at time t respectively; They are the upper and lower limits of HFC electric power respectively; They are the upper and lower limits of HFC power ramping respectively.

[0227] ⑤Carbon capture equipment

[0228]

[0229] Where: is the capture energy consumption required for CCS to capture CO2 at time t; is the energy consumption maintained at time t, which is a fixed value that is independent of the operating state of the CCS; is the operating energy consumption at time t, which changes with the operating state of CCS and is positively correlated with the ability of CCS to capture CO2; c The energy consumption required to capture a unit of CO2 for CCS; They are the upper and lower limits of CCS power consumption respectively.

[0230] ⑥Electric energy storage electric vehicles, thermal energy storage electric vehicles, hydrogen fuel cell electric vehicles

[0231]

[0232] The superscript m indicates the energy type, which includes ES, TS, EV, and HEV; is the capacity of the energy storage device at time t; φ m is the self-loss coefficient of energy storage device m; are the charging and discharging efficiencies of the energy storage device, respectively; are the charging and discharging power of the energy storage device respectively; and are the maximum and minimum storage capacities of the energy storage device m, respectively; is a 0-1 variable, used to indicate whether the energy storage device m can be charged and discharged at the same time. Indicates that the energy storage device is charged. Indicates that the energy storage device releases energy; They represent the upper limits of the charging and discharging power of the energy storage device m respectively; is the initial capacity of the energy storage device m; It is the remaining capacity of energy storage device m after running one cycle.

[0233] ⑦Seasonal hydrogen energy storage

[0234] In a single typical day, SHS can only charge hydrogen or only discharge hydrogen, so that SHS works in different typical days and realizes hydrogen energy interaction between days. The energy self-dissipation rate of SHS is almost close to 0, and the accumulation of hydrogen charging and discharging in the previous typical day scene needs to be considered at the initial moment of the next typical day scene. Therefore, the mathematical model of SHS can be expressed as:

[0235]

[0236] Where: Equation (26) is the energy balance of hydrogen storage day; Equation (27) is the hydrogen storage at the beginning of each season, where the parameter Π s represents the number of energy cycles s for typical days in a season; Equation (28) is the hydrogen charge state after annual storage balance; Equation (29) is the capacity constraint of SHS at the beginning and end of the year; Equations (30) and (31) are the energy state and hydrogen charging / release rate constraints of SHS, respectively.

[0237] ⑧ Time-shiftable electrical load

[0238]

[0239] Where: is the load power that can be shifted at time t; α i,TSE It is the ratio of time-shiftable load to total load.

[0240] ⑨Can interrupt electrical load

[0241]

[0242] Where: is the interruptable load power at time t; α i,IE It is the ratio of interruptible load to total load.

[0243] ⑩ Bidding price constraints

[0244]

[0245] Where: are the upper and lower limits of GT's bidding price in the electricity market, and the other symbols are similar.

[0246] (4) Lower-level joint market clearing model

[0247] 4.1) Second objective function

[0248] In the lower model, the market operator jointly clears the day-ahead electricity market, peak load market and hydrogen market based on the bid prices submitted by each VPP, with the goal of minimizing the day-ahead joint market energy purchase cost, and feeds back the clearing results to each market participant; the second objective function is as follows:

[0249]

[0250] Where: U MA R is the total cost of energy purchase by the joint market operator; E,i,t , R S,i,t , R H,i,t are the energy purchase costs of the market operator in the electricity market, peak load market, and hydrogen market at time t respectively; They are the bidding prices of GT, ES, EV and RE in the electricity market in VPPi; is the bidding price of GT in the peak load market in VPPi; They are the bidding prices of P2H, SHS and HEV in the hydrogen energy market in VPPi; They are the bidding powers of GT, ES, EV and RE in the electric energy market in VPPi; is the bidding power of GT in the peak load market in VPPi; They are the bidding powers of P2H, SHS and HEV in VPPi in the hydrogen energy market.

[0251] 4.2) Constraints

[0252] ①Power balance

[0253]

[0254] Where: They are the energy demands of the electricity market, peak load regulation market and hydrogen energy market at time t respectively.

[0255] ②VPP bidding power constraints

[0256]

[0257] Where: They are the upper and lower limits of GT’s winning bid power in the electric energy market; is the dual variable of the corresponding constraint.

[0258] 2) Two-layer model conversion and solution

[0259] (1) Equivalent transformation of the two-stage distributed robust model

[0260] The present invention adopts duality theory to transform the inner layer real-time stage worst expected distribution supremum problem into the infimum problem, thereby obtaining the single-layer minimum target optimization.

[0261]

[0262] Where: i is the dual factor; δ i , They represent the uncertain variable sample set {δ i,1 , δ i,2 , …, δ i,N}. Introduce auxiliary variable w n It can be simplified to:

[0263]

[0264] Since the supremum of the constraint condition in formula (40) can only be obtained on the boundary of the uncertain variable, the objective function is finally transformed into a single-layer linear programming model as follows:

[0265]

[0266] (2) Transformation of the lower joint clearing model of the two-layer game bidding model

[0267] 2.1) Transformation of the two-layer model

[0268] The relationship between VPP and market operators is a two-layer master-slave non-cooperative game. When the followers make the best response according to the strategies of all leaders and the leaders also accept this response, the game reaches equilibrium. At this time, it is impossible for any participant to gain greater benefits by unilaterally changing the strategy. The lower model in the two-layer model proposed in this paper is non-convex and nonlinear. The KKT (Karush Kuhn-Tucker) theorem is used to transform the lower-layer clearing model into KKT constraints, thereby transforming the equilibrium constraint programming problem into a mixed integer linear programming problem.

[0269] First, write the Lagrangian function of the lower-level joint clearing model, as shown in Equation (42).

[0270]

[0271] Taking partial derivatives of the variables, we can get the Lagrangian equilibrium condition of the underlying problem:

[0272]

[0273] The relaxation condition corresponding to the inequality in the lower problem is:

[0274]

[0275] 4.2) Linearization

[0276] The transformed model still has nonlinear terms: one is the complementary relaxation condition; the other is the product of the double decision variables in the objective function. The linear transformation process is as follows:

[0277] ① Linearization of objective function

[0278] The nonlinearity of the above-mentioned model is mainly caused by the product of two continuous variables in the objective function (11), which makes the model impossible to solve directly.

[0279] Taking the electricity market as an example, when the market bidding entities adopt the bidding method of "quoting price-quoting quantity", the bidding electricity price of each entity in VPPi is Bidding power Clearing electricity price π E,i,t , Clearing Power P E,i,t The following relationship exists:

[0280]

[0281] There are still two products of decision variables in formula (43), which makes the VPPi bidding model a nonlinear and non-convex optimization problem, but it can be processed by binary expansion method. and For example, the processing process is:

[0282]

[0283] Where: N π 、N P is the number of binary expansion variables; are newly introduced auxiliary variables, and their calculation method is shown in formula (47). The calculation method of is shown in formula (48), and the relaxation method of the nonlinear part in formula (46) is shown in formula (49):

[0284]

[0285] Where: All are Boolean variables; M π 、M P are all sufficiently large positive integers.

[0286] ② Linearization of constraints

[0287] The transformed model still has nonlinear terms. For the complementary constraints in formula (44), the big M method is used to introduce Boolean variables to linearize the product of the dual variables and decision variables in the above complementary constraints. Taking the complementary constraint of the upper power limit of GT power sales as an example, the constraint can be used to correspond to the auxiliary Boolean variable and a sufficiently large positive number M can be reformulated as:

[0288]

[0289] The linearization method for the remaining complementary relaxation constraints in equation (44) is similar to equation (50).

[0290] (3) Solution process

[0291] Step 1: Before the bidding begins, VPPi fully integrates internal resources based on the predicted day-ahead market information and the bidding plans of other VPPs. Taking the electricity market as an example, the first round of bidding plans are determined as follows:

[0292] Step 2: After the bidding begins, VPPi will select the bidding plan of the k-1th round according to the bidding plan of other VPPj (j=1, 2, ..., n; j≠i) Solve the upper bidding model to formulate its own bidding plan for round k (k ≥ 2) After all VPPs have given their bids for the kth round, go to step 3;

[0293] Step 3: Each VPP takes the minimum operating cost as the optimization goal, repeats step 2, and repeats the quotation to the market operator; after multiple rounds of game, if no VPP changes its bidding plan, go to step 4; if there are still VPPs trying to change their bidding strategies and submit a bidding plan for the k+1th round, go to step 5;

[0294] Step 4: If all VPPs no longer try to profit by unilaterally changing their bidding plans, a Nash equilibrium is formed in the market. After the bidding is over, the market operators will clear out the market in a unified manner according to the results of the market game.

[0295] Step 5: Determine whether the upper limit of the game rounds has been reached; if the upper limit of the game rounds has not been reached at this time, return to step 1; if it has been reached, the market operator will no longer accept the new round of bidding proposals from each VPP; after the bidding is over, the market operator will force clearing according to the agreed market trading rules.

[0296] This embodiment 1

[0297] Functional test of the optimization model of cooperative operation of multiple virtual power plants under the carbon trading mechanism

[0298] (1) Parameter settings

[0299] To verify the effectiveness of the strategy proposed in this paper, a simulation is conducted using the energy market in Xi'an, Shaanxi Province, China as an example. In this day-ahead market, there are three VPPs with different structures participating in the bidding. The internal structure of each VPP is as follows: Figure 1 As shown. Taking the typical days of the four seasons as the research objects, VPP1 and VPP3 are located in areas with low building density and are mainly equipped with wind power; VPP2 is located in areas with high building density and is mainly equipped with photovoltaics. Among them, VPP3 contains 20 charging piles and 10 hydrogen charging piles. The basic parameters of the internal equipment of each VPP are shown in Table 1. The present invention calls the Gurobi solver for solution in the MATLAB2021a environment.

[0300] Table 1 Basic parameters of each VPP internal equipment

[0301]

[0302] (2) Analysis of VPP’s participation in multi-market clearing results

[0303] 1. Convergence analysis

[0304] Figure 3 The iterative solution process of the two-level non-cooperative game model is demonstrated. Figure 3It can be seen that after 26 rounds of non-cooperative games, the operating costs of each VPP in the joint market tend to be stable, and no operator can obtain additional benefits by unilaterally changing its bidding plan. The bidding strategy at this time is the Nash equilibrium solution that optimizes the benefits of each subject, among which the equilibrium cost of the joint market is 24755.04￥. The equilibrium costs corresponding to VPP1, VPP2, and VPP3 are -665.64￥, -2740.10￥, and -5901.12￥, respectively. In the solution process, the operating costs of each subject show a law of continuous fluctuations and increases and decreases. This is because the dynamic bidding strategy proposed in the present invention is closely related to factors such as carbon emissions and load demand response. During each iteration, these factors fluctuate, resulting in prices fluctuating within a certain range. It can be seen that the bidding game method proposed in the present invention can effectively solve the bidding decision-making schemes for multiple VPPs to participate in the electricity market, peak-shaving market, and hydrogen energy market at the same time, and realize the optimal allocation of the benefits of each subject in a complex game scenario.

[0305] 2. Analysis of the results of VPP participation in the electricity market clearing

[0306] The clearing power of each VPP in the electricity market is as follows: Figure 4 As shown. Figure 4 It can be seen that the dynamic bidding strategy makes the winning bids of each VPP different. The winning bids of VPP are positively correlated with the wind and solar output, that is, VPP will obtain more market winning bids when the renewable energy power generation is high. Among them, the winning bids of VPP3 are large in most periods, and it has an advantage in the bidding game. The reason is that the wind power in VPP3 has output at all times, and its lower operating cost makes VPP3 have a lower bid level in the game bidding. The winning bid periods of VPP2 are mainly concentrated in 8:00-18:00, 32:00-42:00, 56:00-66:00, and 80:00-90:00. The photovoltaic output in the above periods is sufficient, so it can provide a lower bid in the electricity market. When the winning bids of VPP2 and VPP3 are difficult to meet the market electricity demand, the surplus renewable energy output in VPP1 wins the bid for the market power shortage at a lower electricity price and obtains benefits.

[0307] 3. Analysis of the results of VPP participating in the peak load-shaving market clearing

[0308] The clearing volume of each VPP in the peak-shaving market, such as Figure 5 As shown. Figure 5It can be seen that during the peak shaving period, VPP1 and VPP2 have a relatively large proportion of bidding capacity invested in the peak shaving market. The reason is that the gas turbines in VPP1 and VPP2 are at a disadvantage in the bidding in the electricity market, and will give priority to participating in the peak shaving market to maximize their own profits. This is especially important for VPP1 with a larger gas turbine power generation capacity. In addition, because gas turbines are more stable and efficient than renewable energy output and have high peak shaving quality, VPP operators can compensate for their economic losses caused by failed bidding in the electricity market by reserving more bidding capacity in the peak shaving market.

[0309] 4. Analysis of the results of VPP’s participation in hydrogen market clearing

[0310] The amount of each VPP cleared in the hydrogen market, such as Figure 6 As shown. Figure 6 It can be seen that in most periods, the hydrogen energy winning bids of VPP1 and VPP3 are relatively large, with obvious advantages, while VPP2 is in a relatively disadvantaged position. The reason is that VPP3 has lower operating costs throughout the entire period, and hydrogen fuel cell electric vehicles make its bidding scheme more economical and flexible, which can help VPP3 obtain more market share with a lower bid in the game. In addition to participating in electricity supply through hydrogen fuel cells, the hydrogen energy produced in VPP1 also participates in market transactions. The power market side adheres to the principle of the lowest energy purchase cost and will also allocate more market share to VPP1 operators. The renewable energy output in VPP2 mainly participates in the clearing of the electricity market. The hydrogen energy produced by electricity-to-hydrogen production is relatively small, and VPP2 is not equipped with seasonal hydrogen energy storage, which makes it occupy a relatively small market share.

[0311] Figure 7 is a schematic diagram of the composition structure of a virtual power plant participating in a joint market bidding device based on a master-slave non-cooperative game provided by an embodiment of the present invention, such as Figure 7As shown, a virtual power plant participating in a joint market bidding device 700 based on a master-slave non-cooperative game includes: a construction module 701, which is used to construct a two-layer game framework for multiple virtual power plants to participate in a day-ahead joint market; a modeling module 702, which is used to perform Wasserstein distance uncertainty modeling on wind power, photovoltaic power and electric load power to obtain a multi-source load power uncertainty set; the construction module 701 is also used to construct an upper-layer two-stage distributed robust bidding model for multiple virtual power plants based on the multi-source load power uncertainty set; the two-stage robust distributed bidding model includes a first objective function of minimizing the sum of the pre-dispatch cost of formulating a bidding strategy in the day-ahead stage and the expected sum of the re-dispatch cost in the intra-day stage; and correspondingly constructs a first type of constraint condition for the first objective function; the construction module 701 01, and is also used to construct a lower-level joint market clearing model; the clearing model includes a second objective function of minimizing the cost of purchasing energy in the day-ahead electricity-peak-shaving-hydrogen joint market; and a second type of constraint condition for constructing the second objective function accordingly; a conversion module 703 is used to perform equivalent conversions on the two-stage robust distribution bidding model and the lower-level joint clearing model respectively, and obtain a converted bidding model and a converted clearing model accordingly; the complementary relaxation conditions in the converted bidding model are linearized to obtain processed complementary relaxation conditions; a solution module 704 is used to solve the converted bidding model and the converted clearing model based on the processed complementary relaxation conditions, the first type of constraint conditions and the second type of constraint conditions to obtain a bidding strategy.

[0312] It should be noted that the description of the device of the embodiment of the present invention is similar to the description of the above method embodiment, and has similar beneficial effects as the method embodiment, so it is not repeated. For technical details not disclosed in the device embodiment, please refer to the description of the method embodiment of the present invention for understanding.

[0313] It should be noted that in the embodiments of the present invention, if the above-mentioned method for virtual power plants to participate in joint market bidding based on master-slave non-cooperative game is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions for a terminal to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM), a magnetic disk or an optical disk. In this way, the embodiments of the present invention are not limited to any specific combination of hardware and software.

[0314] Correspondingly, an embodiment of the present invention provides a virtual power plant participating in a joint market bidding device based on a master-slave non-cooperative game. Figure 8 is a schematic diagram of the composition structure of a virtual power plant participating in a joint market bidding device based on a master-slave non-cooperative game provided by an embodiment of the present invention, such as Figure 8 As shown, the virtual power plant participating in the joint market bidding device 800 based on the master-slave non-cooperative game at least includes: a processor 801 and a computer-readable storage medium 802 configured to store executable instructions, wherein the processor 801 generally controls the overall operation of the virtual power plant participating in the joint market bidding device 800 based on the master-slave non-cooperative game. The computer-readable storage medium 802 is configured to store instructions and applications executable by the processor 801, and can also cache data to be processed or processed by each module in the processor 801 and the virtual power plant participating in the joint market bidding device 800 based on the master-slave non-cooperative game, which can be implemented through a flash memory (FLASH) or a random access memory (Random Access Memory, RAM).

[0315] An embodiment of the present invention provides a storage medium storing executable instructions, wherein the executable instructions are stored. When the executable instructions are executed by a processor, the processor will be caused to execute the method provided by the embodiment of the present invention, for example, Figure 1 The method shown.

[0316] In some embodiments, the storage medium can be a computer-readable storage medium, for example, a ferroelectric random access memory (FRAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a flash memory, a magnetic surface memory, an optical disk, or a compact disk read-only memory (CD-ROM), etc.; it can also be various devices including one or any combination of the above memories.

[0317] In some embodiments, executable instructions may be in the form of a program, software, software module, script or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine or other unit suitable for use in a computing environment.

[0318] As an example, executable instructions may, but need not necessarily, correspond to a file in a file system, may be stored as part of a file storing other programs or data, for example, in one or more scripts in a Hyper Text Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files storing one or more modules, subroutines, or code portions). As an example, executable instructions may be deployed to be executed on one electronic device, or on multiple electronic devices located at one location, or on multiple electronic devices distributed at multiple locations and interconnected by a communication network.

[0319] The above description is only an embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement and improvement made within the spirit and scope of the present invention are included in the protection scope of the present invention.

[0320] It should be understood that "one embodiment" or "an embodiment" mentioned throughout the specification means that specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present invention. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present invention, the size of the serial number of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention. The serial numbers of the above-mentioned embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments.

[0321] It should be noted that, in this article, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method or device. In the absence of further restrictions, an element defined by the statement "comprises one..." does not exclude the presence of other identical elements in the process, method, article or device including the element. In several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed.

[0322] The above is only an embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for virtual power plants to participate in joint market bidding based on master-slave non-cooperative game, characterized in that: The method comprises: Construct a two-tier game framework for multiple virtual power plants to participate in the day-ahead joint market; The uncertainty modeling of wind power, photovoltaic power and electric load power is carried out by using Wasserstein distance to obtain the uncertainty set of multi-source load power. Based on the multi-source load power uncertainty set, a two-stage distributed robust bidding model for upper-layer multi-virtual power plants is constructed; the two-stage robust distributed bidding model includes a first objective function of minimizing the sum of the pre-dispatch cost of formulating the bidding strategy in the day-ahead stage and the expected re-dispatch cost in the intra-day stage; and a first type of constraint condition of the first objective function is constructed accordingly; Constructing a lower-level joint market clearing model; the clearing model includes a second objective function of minimizing the energy purchase cost of the day-ahead electricity-peak-shaving-hydrogen joint market; and correspondingly constructing a second type of constraint condition of the second objective function; The two-stage blue-stick bidding model and the lower-layer joint clearing model are respectively equivalently transformed to obtain a transformed bidding model and a transformed clearing model; the complementary relaxation conditions in the transformed bidding model are linearized to obtain processed complementary relaxation conditions; Based on the processed complementary relaxation conditions, the first type of constraints and the second type of constraints, the converted bidding model and the converted clearing model are solved to obtain a bidding strategy.

2. The method according to claim 1, characterized in that The uncertainty modeling of the Wasserstein distance for wind power, photovoltaic power and electric load power is performed to obtain a multi-source load power uncertainty set, including: Based on the day-ahead forecast output of wind power and photovoltaic power and the error vector of wind power and photovoltaic power source load, the actual output of wind power and photovoltaic power within the day is determined; based on the day-ahead forecast output of electric load and the error vector of electric load source load power, the actual output of electric load within the day is determined; The first Wasserstein distance between the empirical distribution and the real distribution of wind power output, the second Wasserstein distance between the empirical distribution and the real distribution of photovoltaic output, and the third Wasserstein distance between the empirical distribution and the real distribution of electric load are calculated respectively; The wind power uncertainty set, the photovoltaic power uncertainty set and the electric load uncertainty set are correspondingly determined based on the first Wasserstein distance, the second Wasserstein distance and the third Wasserstein distance respectively.

3. The method according to claim 1, characterized in that Based on the multi-source load power uncertainty set, an upper-layer multi-virtual power plant two-stage distributed robust bidding model is constructed; the two-stage robust distributed bidding model includes a first objective function of minimizing the sum of the pre-dispatch cost of formulating the bidding strategy in the day-ahead stage and the expected re-dispatch cost in the intra-day stage, including: The first objective function is expressed as: Where: U i is the net operating cost of VPPi participating in various energy markets during the day-ahead bidding phase; f(x i ,δ i ) is the redispatch cost of VPPi during the intraday dispatch phase; To include uncertainty The function of the probability distribution The expectation below; sup is the supremum; π E,i,t , π S,i,t and π H,i,t are the clearing prices of VPPi in the electricity market, peak load market and hydrogen market at time t respectively; P E,i,t , P S,i,t and P H,i,t are the winning bid powers of VPPi in the electricity market, peak load regulation market and hydrogen energy market at time t; C i is the operating cost of VPPi.

4. The method according to claim 1, characterized in that: The first type of constraint conditions corresponding to constructing the first objective function include: 1) Gas turbine constraints Where: and are the electrical output and thermal output of the gas turbine GT at time t; ηp ,t and ηq ,t are the electrical efficiency and thermal efficiency of GT at time t respectively; is the gas consumption power of GT at time t; H gas It is the low calorific value of natural gas; and They are the upper and lower limits of GT output respectively; and They are the upper and lower limits of GT climbing power respectively; 2) Constraints of Hydrogen Production by Electricity Where: are the P2H power consumption and hydrogen production power at time t respectively; η P2H is the conversion efficiency of P2H; and They are the upper and lower limits of P2H power consumption respectively; and They are the upper and lower limits of the P2H power consumption ramp, respectively; 3) Hydrogen fuel cell constraints Where: is the hydrogen consumption power of HFC at time t; and are the electrical power and thermal power of HFC at time t respectively; η HFC,p,t and η HFC,q,t are the electrical efficiency and thermal efficiency of HFC at time t respectively; and are the upper and lower limits of HFC electric power, respectively; They are the upper and lower limits of HFC electric power ramping, respectively; 4) Carbon capture equipment constraints Where: is the capture energy consumption required for CCS to capture CO2 at time t; is the energy consumption maintained at time t, which is a fixed value that is independent of the operating state of the CCS; is the operating energy consumption at time t, which changes with the operating state of CCS and is positively correlated with the ability of CCS to capture CO2; c The energy consumption required to capture a unit of CO2 for CCS; They are the upper and lower limits of CCS power consumption respectively; 5) Constraints for electric energy storage electric vehicles, thermal energy storage electric vehicles, and hydrogen fuel cell electric vehicles Where: superscript m indicates energy type, which includes ES, TS, EV and HEV; is the capacity of the energy storage device at time t; φ m is the self-loss coefficient of energy storage device m; and They are the charging efficiency and discharging efficiency of the energy storage device respectively; and are the charging power and discharging power of the energy storage device respectively; and are the maximum and minimum storage capacities of the energy storage device m, respectively; is a 0-1 variable, used to indicate whether the energy storage device m can be charged and discharged at the same time. Indicates that the energy storage device is charged. Indicates that the energy storage device releases energy; and They represent the upper limits of the charging power and discharging power of the energy storage device m respectively; is the initial capacity of the energy storage device m; The remaining capacity of the energy storage device m after running for one cycle; 6) Constraints on the energy balance of hydrogen storage day 7) Hydrogen storage at the beginning of each season 8) Hydrogen charge status after annual storage balance 9) SHS initial and year-end capacity constraints 10) Energy state constraints of SHS 11) Hydrogen charging / release rate constraints Where: parameter Π s The number of energy cycles s representing the typical days in the season; 12) Time-shiftable electrical load Where: is the load power that can be shifted at time t; α i,TSE is the ratio of time-shiftable load to total load; 13) Interruptible power load Where: They are the upper and lower limits of GT’s bidding price in the electricity market; 14) Power balance constraints 5. The method according to claim 1, characterized in that The lower-level joint market clearing model is constructed; the clearing model includes a second objective function of minimizing the energy purchase cost of the day-ahead electricity-peak-shaving-hydrogen joint market, including: The second objective function is expressed as: Where: U MA is the total cost of energy purchase by the joint market operator; R E,i,t , R S,i,t and R H,i,t are the energy purchase costs of the market operator in the electricity market, peak load market and hydrogen energy market at time t respectively; and They are the bidding prices of GT, electric energy storage ES, electric vehicles EV and renewable energy RE in the electricity market in VPPi; is the bidding price of GT in the peak load market in VPPi; and They are the bidding prices of P2H, SHS and hydrogen fuel cell electric vehicles HEV in the hydrogen energy market in VPPi; and are the bidding powers of GT, ES, EV and RE in the electric energy market in VPPi respectively; is the bidding power of GT in the peak load market in VPPi; and They are the bidding powers of P2H, SHS and HEV in VPPi in the hydrogen energy market.

6. The method according to claim 1, characterized in that The second type of constraint conditions corresponding to constructing the second objective function include: 1) Power balance constraints Where: and are the energy demands of the electricity market, peak load regulation market and hydrogen energy market at time t respectively; 2) VPP bidding power constraints Where: and These are the upper limits of the winning bids for GT, ES, EV and RE in the electric energy market; The power cap for GT's winning bid in the peak-shaving market; They are the upper limits of the winning bids for GT, SHS and HE V in the hydrogen energy market; is the dual variable of the corresponding constraint.

7. The method according to claim 1, characterized in that The two-stage blue-stick bidding model and the lower-layer joint clearing model are respectively converted equivalently to obtain a converted bidding model and a converted clearing model, including: Using the duality theory, the supremum problem of the worst expected distribution in the real-time stage of the bidding model is transformed into the infimum problem, and the single-layer minimum target optimization is obtained: Introduce auxiliary variable w n It can be simplified to: Since the supremum of the constraint condition in the above formula can only be obtained on the boundary of the uncertain variable, the first objective function is finally transformed into a single-layer linear programming model, and the transformed bidding model is expressed as: Where: i is the dual factor; δ i , They represent the uncertain variable sample sets {δ i,1 , δ i,2 , …, δ i,N }'s infimum and supremum; The Lagrangian function of the clearing model is expressed as: Taking partial derivatives of the variables, we can obtain the Lagrangian equilibrium condition of the clearing model: The relaxation condition corresponding to the inequality in the clearing model is:

8. The method according to claim 1, characterized in that Solving the converted bidding model and the converted clearing model includes: Step 1: Before the bidding begins, VPPi fully integrates internal resources based on the predicted day-ahead market information and the bidding plans of other VPPs. Taking the electricity market as an example, the first round of bidding plans are determined as follows: Step 2: After the bidding begins, VPPi will select the bidding plan of the k-1th round according to the bidding plan of other VPPj (j=1, 2, ..., n; j≠i) Solve the upper bidding model to formulate its own bidding plan for round k (k ≥ 2) After all VPPs have given their bids for the kth round, go to step 3; Step 3: Each VPP takes the minimum operating cost as the optimization goal, repeats step 2, and repeats the quotation to the market operator; after multiple rounds of game, if no VPP changes its bidding plan, go to step 4; if there are still VPPs trying to change their bidding strategies and submit a bidding plan for the k+1th round, go to step 5; Step 4: If all VPPs no longer try to profit by unilaterally changing their bidding plans, a Nash equilibrium is formed in the market. After the bidding is over, the market operators will clear out the market in a unified manner according to the results of the market game. Step 5: Determine whether the upper limit of the game rounds has been reached; if the upper limit of the game rounds has not been reached at this time, return to step 1; if it has been reached, the market operator will no longer accept the new round of bidding proposals from each VPP; after the bidding is over, the market operator will force clearing according to the agreed market trading rules.

9. A virtual power plant participating in a joint market bidding device based on a master-slave non-cooperative game, characterized in that: The device comprises: A building module for constructing a two-layer game framework for multiple virtual power plants to participate in the day-ahead joint market; Modeling module, used to perform Wasserstein distance uncertainty modeling on wind power, photovoltaic power and electric load power, and obtain the multi-source load power uncertainty set; The construction module is further used to construct an upper-layer multi-virtual power plant two-stage distributed robust bidding model based on the multi-source load power uncertainty set; the two-stage robust distributed bidding model includes a first objective function of minimizing the sum of the pre-dispatch cost of formulating the bidding strategy in the day-ahead stage and the expected re-dispatch cost in the intra-day stage; and correspondingly constructs a first type of constraint condition of the first objective function; The construction module is also used to construct a lower-level joint market clearing model; the clearing model includes a second objective function of minimizing the energy purchase cost of the day-ahead electricity-peak-shaving-hydrogen joint market; and a second type of constraint condition for constructing the second objective function accordingly; A transformation module, used to perform equivalent transformation on the two-stage robust distribution bidding model and the lower-layer joint clearing model, respectively, to obtain a transformed bidding model and a transformed clearing model; linearize the complementary relaxation conditions in the transformed bidding model to obtain processed complementary relaxation conditions; A solution module is used to solve the converted bidding model and the converted clearing model based on the processed complementary relaxation conditions, the first type of constraints and the second type of constraints to obtain a bidding strategy.

10. A virtual power plant participating in a joint market bidding device based on a master-slave non-cooperative game, characterized in that: include: A memory for storing executable instructions; A processor is used to implement the method for virtual power plants to participate in joint market bidding based on master-slave non-cooperative game as described in any one of claims 1 to 8 when executing executable instructions stored in the memory.

Images