Multi-market coordination and risk measurement of electricity-hydrogen-green fuel virtual power plant coordinated bidding method and device

Abstract

This invention belongs to the field of power market technology and discloses a method and apparatus for collaborative bidding in a virtual power plant based on multi-market collaboration and risk measurement for electricity-hydrogen-green fuel. The method includes: acquiring multi-timescale operating parameters and historical operating data; obtaining an initial operating dataset based on the green fuel delivery demand curve; acquiring a set of equipment operating constraints that meet green fuel delivery requirements based on the initial operating dataset and typical scenarios; acquiring a set of market constraints for collaborative bidding; constructing a stochastic look-ahead collaborative bidding objective function based on conditional value at risk (VAT) measurement; obtaining a bidding model with nonlinear terms to be solved based on the set of equipment operating constraints and the set of market constraints for collaborative bidding; linearizing the nonlinear terms in the bidding model to obtain a mixed-integer linear programming model; and obtaining multi-market collaborative bidding strategies and internal production scheduling schemes for various scenarios, providing support for the virtual power plant.

Description

Technical Field

[0001] This invention belongs to the field of electricity market technology and relates to a collaborative bidding method and device for virtual power plants of electricity-hydrogen-green fuel with multi-market collaboration and risk measurement. Background Technology

[0002] With the construction of new power systems and the continuous advancement of power market reforms, virtual power plants, as important players in aggregating distributed resources to participate in the power market, are evolving from a single day-ahead market to a multi-market collaborative approach encompassing medium- and long-term markets, spot markets, and ancillary services. Meanwhile, the coupling of electricity, hydrogen, and green fuels, due to its adjustable load, long-cycle energy storage, and cross-seasonal regulation potential, is participating in multi-market collaborative bidding.

[0003] Using long-term contract electricity as a fixed exogenous boundary condition or simply dividing it into equal time periods, combined with the spot market's "volume and price quotation" strategy for joint optimization, and introducing methods such as stochastic optimization or robust optimization to handle the uncertainty of new energy output and market prices, this approach suffers from several drawbacks. Firstly, the physical boundary of the bidding process is insufficiently defined and disconnected from fuel delivery, ignoring the complex physical coupling of the electricity-hydrogen-green fuel system and failing to treat the "rigid demand for green fuel delivery" as a hard constraint. This leads to bidding strategies that easily violate fuel delivery contracts. Secondly, there is a lack of coordination between long-term and spot market bidding strategies. Long-term contracts are often treated in isolation, failing to deeply couple with spot volume and price curve reporting and strict deviation assessment mechanisms. This results in a disconnect between the decomposition scheme and spot reporting, easily leading to huge deviation assessment costs. Thirdly, the spot bidding rules have low compatibility, making it difficult to accurately extract and linearize the transaction volume under the premise of the monotonicity, number of segments, and prefix structure of the "monotonic non-decreasing segmented volume and price curve across zero point." Finally, traditional optimization only pursues the maximization of expected returns, ignoring the risk of huge losses in extremely unfavorable scenarios, leading to overly aggressive bidding strategies and extremely volatile actual settlement returns. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide a collaborative bidding method and apparatus for virtual power plants of electricity, hydrogen and green fuels with multi-market collaboration and risk measurement.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: This invention provides a collaborative bidding method for electric-hydrogen-green fuel virtual power plants based on market coordination and risk measurement, comprising the following steps: acquiring multi-timescale operating parameters and historical operating data of heterogeneous resources within the virtual power plant; obtaining an initial operating dataset based on the green fuel delivery demand curve; acquiring a set of equipment operating constraints that meet green fuel delivery requirements based on the initial operating dataset and typical scenarios; acquiring a set of market constraints for collaborative bidding; constructing a stochastic look-ahead collaborative bidding objective function, wherein the collaborative bidding objective function is based on the set of equipment operating constraints and the set of market constraints for collaborative bidding to obtain a bidding model with nonlinear terms to be solved; linearizing the nonlinear terms in the bidding model to obtain a mixed-integer linear programming model; and obtaining multi-market collaborative bidding strategies and internal production scheduling schemes under various scenarios based on the mixed-integer linear programming model.

[0006] Furthermore, the market constraint set of the collaborative bidding includes medium- and long-term market constraints and spot market constraints; medium- and long-term market constraints include medium- and long-term transaction constraints and medium- and long-term deviation assessment constraints; spot market constraints include day-ahead segmented volume and price curve reporting constraints, transaction interception constraints under day-ahead prices in various scenarios, and real-time operation constraints.

[0007] Furthermore, the objective function of the collaborative bidding is:

[0008] in, Let be the probability of scenario s occurring. The medium-to-long-term contract electricity volume (MW) for time period t on day d is positive for grid connection and negative for grid disconnection. The price represents a medium- to long-term contract price, which is a fixed value derived from exogenous input. For scenario s, the price before the long-term settlement date on day d of time period t. For scenario s, the long-term settlement reference electricity price for time period t on day d. For time step, This represents the net trading power of the day before yesterday. This represents the actual net power in real time. This is the subsidy coefficient for hydrogen production. Let t be the hydrogen production power of the electrolyzer on day d in scenario s. For carbon credit pricing, CO2 emission intensity of fossil fuels, CO2 emission intensity of synthetic fuels For scenario s, the fuel synthesis output power during time period t on day d. This represents the maintenance cost coefficient for the electrolytic cell. For scenario s, the power consumption of the electrolytic cell during time period t on day d. This represents the fuel cell operation and maintenance cost coefficient. For scenario s, the fuel cell power generation capacity during time period t on day d is... This represents the operation and maintenance cost coefficient for the fuel synthesis unit. For scenario s, the fuel synthesis output power during time period t on day d. The degradation cost coefficient for energy storage charging. For the energy storage charging power of the s-th scenario on day d, time period t, This is the energy storage discharge degradation cost coefficient. Let t be the energy storage discharge power of the s-th scenario on day d. The excess performance fee for hour h under scenario s. The shortfall assessment fee for hour h under scenario s. Represents the risk aversion coefficient. This indicates that no risks are considered. This indicates extreme risk aversion; Represents conditional risk value.

[0009] Furthermore, the set of equipment operation constraints includes the operation constraints of the electrolyzer, fuel cell, hydrogen storage tank, fuel synthesis system, and fuel storage tank; the operation constraints of the electrolyzer include upper and lower limits of operating power, ramp rate, and electro-hydrogen conversion efficiency; the operation constraints of the fuel cell include upper and lower limits of power generation, ramp rate, and hydrogen-electric conversion efficiency; the operation constraints of the hydrogen storage tank include hydrogen charging / discharging power and hydrogen storage capacity boundaries; the operation constraints of the fuel synthesis system include upper and lower limits of synthesis power and hydrogen-fuel conversion efficiency; and the operation constraints of the fuel storage tank include the upper limit of fuel storage tank charging / discharging power and fuel order delivery.

[0010] Furthermore, the linearization of the nonlinear terms in the bidding model is achieved by setting a constant M to transform the nonlinear product into a linear inequality constraint.

[0011] Furthermore, the conditional value of risk for:

[0012] in, Value at risk, For confidence level, For scene indexing, For the scene The probability of occurrence For the scene The auxiliary variable for excess loss.

[0013] This invention also provides a multi-market collaboration and risk measurement-based collaborative bidding device for an electric-hydrogen-green fuel virtual power plant, comprising: a data unit for acquiring multi-timescale operating parameters and historical operating data of heterogeneous resources within the virtual power plant, and obtaining an initial operating dataset based on the green fuel delivery demand curve; an equipment operation constraint unit for acquiring a set of equipment operation constraints that meet green fuel delivery requirements based on the initial operating dataset and typical scenarios; a collaborative bidding constraint unit for acquiring a set of market constraints for collaborative bidding; a model unit for constructing a stochastic look-ahead collaborative bidding objective function, wherein the collaborative bidding objective function is based on the set of equipment operation constraints and the set of market constraints for collaborative bidding to obtain a bidding model with nonlinear terms to be solved; and a solution unit for linearizing the nonlinear terms in the bidding model to obtain a mixed-integer linear programming model, and obtaining multi-market collaborative bidding strategies and internal production scheduling schemes under various scenarios based on the mixed-integer linear programming model.

[0014] The present invention also provides a computer device, comprising: a processor and a computer-readable storage medium; the processor being adapted to execute a computer program; the computer-readable storage medium storing a computer program, wherein when the computer program is executed by the processor, it implements the multi-market collaboration and risk measurement collaborative bidding method for electric-hydrogen-green fuel virtual power plants as described in any of the first aspects above.

[0015] The present invention also provides a computer-readable storage medium storing a computer program adapted to be loaded by a processor and executed as described in any of the first aspects above: a collaborative bidding method for electric-hydrogen-green fuel virtual power plants with multi-market collaboration and risk measurement.

[0016] The present invention also provides a computer program product comprising a computer program that, when executed by a processor, implements the multi-market collaboration and risk measurement collaborative bidding method for electric-hydrogen-green fuel virtual power plants as described in any of the first aspects above.

[0017] Compared with the prior art, the present invention has the following beneficial technical effects: This invention presents a collaborative bidding method for a virtual power plant integrating electricity, hydrogen, and green fuel, with multi-market collaboration and risk measurement. It uses the rigid delivery demand for green fuel as a constraint, combining multi-timescale operating parameters and equipment constraints of the electricity-hydrogen-green fuel system within the virtual power plant to determine the operating boundary that meets the green fuel delivery requirements, thus avoiding violations of fuel delivery contracts. By constructing collaborative bidding constraints, it deeply couples the decomposition of medium- and long-term contracts with the spot market's "volume and price quotation" curve reporting and deviation assessment mechanism. This aligns with the monotonicity, number of segments, and prefix structure rules of the spot market's "monotonically non-decreasing segmented volume and price curve across zero point." Linearization of nonlinear terms transforms the bidding model into a mixed-integer linear programming model, extracting and solving for transaction volume. Based on conditional value of risk, a stochastic forward-looking collaborative bidding objective function is constructed to optimize expected returns while controlling tail risk, maintaining stability even in extremely unfavorable scenarios. This provides support for virtual power plants.

[0018] By constructing a collaborative bidding objective function, the bidding model is upgraded from simply pursuing the maximization of expected returns to risk-averse optimization. Decision-makers can flexibly balance average profits and guaranteed returns in extreme scenarios by adjusting the risk aversion coefficient, thus addressing the uncertainties of spot prices and new energy output.

[0019] By setting a constant M and introducing auxiliary continuous variables for linearization, the complex nonlinear problem is transformed into a standard mixed-integer linear programming (MILP) model, which can be efficiently solved by commercial solvers such as Gurobi, thus meeting the timeliness requirements of multi-timescale rolling scheduling. Attached Figure Description

[0020] Figure 1 This is a flowchart of a collaborative bidding method for virtual power plants of electricity, hydrogen, and green fuel based on multi-market collaboration and risk measurement, according to the present invention. Figure 2 This is a schematic diagram of the multi-energy flow coupling and multi-market collaborative bidding interaction architecture of electricity-hydrogen-green fuel in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the principle of segmenting the monotonically non-decreasing price-volume curve across zero points in the spot market in an embodiment of the present invention. Figure 4 This is a block diagram of the internal structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0022] Time period index (15-minute step), ) Day Index ( , (Number of days in the month) Index of all time periods in the entire month (total) indivual) Curve segment index (maximum) part) : Curve breakpoint index (number of breakpoints = number of segments + 1) Hourly Index

[0023] : A collection of 4 15-minute time slots within an hour Scene Index

[0024] This invention discloses a collaborative bidding method for an electric-hydrogen-green fuel virtual power plant based on multi-market collaboration and risk measurement, comprising the following steps: acquiring multi-timescale operating parameters and historical operating data of heterogeneous resources within the virtual power plant; obtaining an initial operating dataset based on the green fuel delivery demand curve; acquiring a set of equipment operating constraints that meet green fuel delivery requirements based on the initial operating dataset and typical scenarios; acquiring a set of market constraints for collaborative bidding; constructing a stochastic look-ahead collaborative bidding objective function, wherein the collaborative bidding objective function is based on the set of equipment operating constraints and the set of market constraints for collaborative bidding to obtain a bidding model with nonlinear terms to be solved; linearizing the nonlinear terms in the bidding model to obtain a mixed-integer linear programming model; and obtaining multi-market collaborative bidding strategies and internal production scheduling schemes under various scenarios based on the mixed-integer linear programming model, such as... Figure 1 As shown.

[0025] The initial operating dataset is obtained based on the multi-timescale operating parameters of heterogeneous resources within the virtual power plant, multi-scenario market forecast data, and green fuel delivery demand curves. The time step of the initial operating dataset is set to 15 minutes.

[0026] like Figure 2 As shown, the multi-timescale operating parameters include the rated capacity, conversion efficiency (such as electro-hydrogen conversion efficiency and hydrogen-fuel synthesis efficiency, both calculated based on lower heating value), and ramp rate of the electrolyzer, fuel cell, hydrogen storage tank, fuel synthesis system, and fuel storage tank. Scenario market forecast data includes: day-ahead forecast price, real-time price, medium- to long-term settlement reference point spot price, upper limit of available photovoltaic power, and baseline load. The green fuel delivery demand curve is a hard constraint from an exogenous input, determining the lower limit of green fuel (such as green ammonia and green methanol) power that must be delivered at each time step.

[0027] Typical Scenario: Obtain 90 days of historical data for the corresponding months over the past three years. Each day's data includes five complete curves: photovoltaic output, electricity load, day-ahead electricity price, real-time electricity price, and settlement reference point electricity price. Each curve is further subdivided into 96 time periods. Set the initial total number of scenarios N=500. Each scenario corresponds to a combination of operating conditions within a 30-day optimization cycle. All initial scenarios are uniformly weighted at 1 / N. Generate 500 initial scenarios in batches. For each scenario, independently and randomly sample each day within the 30-day optimization cycle. Randomly extract a set of daily photovoltaic, load, and electricity price curves from the 90-day historical data. Complete the extraction and splicing of 30 days of data to obtain a time series sequence of 2880 data points for a single scenario (30 days × 96 time periods).

[0028] After the initial scenes are generated, typical scenes are selected and initial settings are performed. All 500 batch-generated initial scenes are included in the initial scene set as initial calculation samples for algorithm iteration. Then, an iterative reduction loop is entered, combining scene weights (scene occurrence probabilities) with the minimum distance between scenes to perform comprehensive calculations. Scenes with the smallest product of weight and minimum distance are selected as redundant scenes and deleted. The weight distribution of the remaining scenes is updated, and all the weights of the deleted scenes are added to the remaining scene closest to the deleted scene, ensuring that the sum of all scene weights is always 1. After each single scene deletion and weight update, it is checked whether the current number of scenes has reached the preset target number (target number is 8). If the current number of remaining scenes is greater than the target number, the scene selection, deletion of redundant scenes, and weight update are repeated; if the current number of scenes has decreased to the target number, the iteration process is terminated, and the scenes retained are typical scenes.

[0029] Based on the initial operational dataset and typical scenarios, a set of equipment operation constraints that meet the requirements for green fuel delivery is obtained. This set of constraints includes the operation constraints of electrolyzers, fuel cells, hydrogen storage tanks, fuel synthesis systems, and fuel storage tanks.

[0030] 1. The operating constraints of the electrolyzer include: upper and lower limits of operating power, ramp rate, and electro-hydrogen conversion efficiency.

[0031] Operating power limits:

[0032] in, Let t be the power consumption of the electrolytic cell on day d in scenario s. This is the maximum operating power of the electrolytic cell.

[0033] Climbing speed:

[0034] in, This represents the up / down ramp rate of the electrolytic cell. Let t be the power consumption of the electrolytic cell on day d in scenario s during time period t-1. For time step, .

[0035] Electro-hydrogen conversion efficiency:

[0036] in, The electro-hydrogen conversion efficiency of the electrolyzer. Let t be the hydrogen production power of scenario s on day d during time period t.

[0037] 2. The operating constraints of fuel cells include: upper and lower limits of power generation, ramp rate, and hydrogen-to-electricity conversion efficiency.

[0038] Upper and lower limits of power generation:

[0039] in, Let fetch the fuel cell power generation capacity for day d, time period t, in scenario s. This represents the maximum power output of the fuel cell.

[0040] Climbing speed:

[0041] in, , The ramp rate for the fuel cell. Let fetch the fuel cell power generation capacity for day d, time period t, in scenario s. Let fetch the fuel cell power generation capacity for day d, time period t-1 in scenario s. For time step, .

[0042] Hydrogen-to-electricity conversion efficiency:

[0043] in, For the hydrogen power consumption of fuel cells, This refers to the hydrogen-to-electricity conversion efficiency of a fuel cell.

[0044] 3. The operational constraints of hydrogen storage tanks include hydrogen charging and discharging power and hydrogen storage capacity limits.

[0045] Hydrogen charge / discharge power:

[0046] in, For scenario s, the hydrogen charging power of the hydrogen storage tank during time period t on day d. Let t be the hydrogen discharge power of the hydrogen storage tank on day d in scenario s. Let t be the hydrogen production power of scenario s on day d. For the hydrogen power consumption of fuel cells, Hydrogen consumption power for fuel synthesis.

[0047]

[0048] in, This represents the maximum charging / discharging power of the hydrogen storage tank. The hydrogen storage tank is in hydrogen filling mode. The hydrogen storage tank is in hydrogen release / idle mode.

[0049] The state of hydrogen storage changes over time:

[0050] in, For scenario s, the hydrogen storage energy state at the end of time period t on day d. For scenario s, the hydrogen storage energy state at the end of day d, time period t-1. The charging / discharging efficiency of the hydrogen storage tank, For scenario s, the hydrogen charging power of the hydrogen storage tank during time period t on day d. Let t be the hydrogen release power of the hydrogen storage tank on day d in scenario s.

[0051] Hydrogen storage capacity boundary:

[0052] in, These represent the upper and lower limits of hydrogen storage energy. This refers to the hydrogen storage capacity.

[0053] To ensure the continuity of hydrogen storage during the scheduling cycle, the energy status is continuously transferred between adjacent trading days without any interruption.

[0054]

[0055] in, The energy state of the hydrogen storage tank at the end of day 1 in scenario s. Initial hydrogen storage setpoint, Let's consider the energy state of the hydrogen storage tank at the end of time period T on day D in scenario s. The end-of-period hydrogen storage setpoint, The energy state of the hydrogen storage tank at the end of day d in scenario s. The energy state of the hydrogen storage tank at the end of day T in scenario s is the time period d-1.

[0056] 4. The operational constraints of the fuel synthesis system include the upper and lower limits of synthesis power and the hydrogen-fuel conversion efficiency.

[0057] upper and lower limits of synthetic power:

[0058] in, For scenario s, the fuel synthesis power is the fuel synthesis power during time period t on day d. for Maximum fuel synthesis power of the fuel synthesis system.

[0059] Hydrogen-fuel conversion efficiency:

[0060] in, Hydrogen power consumption for fuel synthesis The efficiency of hydrogen-to-fuel synthesis.

[0061] 5. Operational constraints on fuel storage tanks include the upper limit of fuel tank filling and discharging power and fuel order delivery.

[0062] Fuel tank charging / discharging capacity:

[0063] in, The fuel synthesis output power of scenario s on day d, time period t; For scenario s, the fuel sales power during time period t on day d; Power is supplied to the fuel storage tank during the t period on day d in scenario s. The power released by the fuel storage tank on day d in scenario s is given.

[0064] Maximum charging / discharging capacity of fuel storage tanks:

[0065] In the formula, This represents the maximum charging / discharging power of the fuel storage tank.

[0066] The energy state of fuel storage tanks changes over time:

[0067] in, For the energy state of the fuel storage tank at the end of time period t on day d in scenario s, For the energy state of the fuel storage tank at the end of time period t-1 on day d in scenario s, To improve the charging efficiency of fuel storage tanks, The energy release efficiency of the fuel storage tank.

[0068] Fuel storage tank capacity limits:

[0069] in, This is the upper limit of the fuel storage tank capacity. This is the lower limit for the capacity of fuel storage tanks.

[0070] Cycle start and end constraints and cross-day connections:

[0071] in, For the energy state of the fuel storage tank at the end of day 1 in scenario s, Initial fuel storage For the energy state of the fuel storage tank at the end of time period T on day D in scenario s, End-of-period fuel reserves For the energy state of the fuel storage tank at the end of day d in scenario s, The energy state of the fuel storage tank at the end of day T in the s-th scenario (d-1).

[0072] Fuel order delivery:

[0073] in, This represents the fuel delivery demand for time period t on day d. The fuel sales power for each 15-minute time period is not lower than the corresponding value on the demand curve. When the synthesis system... Due to fluctuations in electricity prices / photovoltaic power, real-time support cannot be provided. When the tank is empty, it will automatically release energy to fill the gap; conversely, when... When needed, excess fuel is automatically stored in a storage tank.

[0074] Obtain the market constraint set for collaborative bidding. The market constraint set for collaborative bidding includes medium- and long-term market constraints and spot market constraints. Medium- and long-term market constraints include medium- and long-term transaction constraints and medium- and long-term deviation assessment constraints. Spot market constraints include day-ahead segmented volume and price curve reporting constraints, transaction interception constraints under day-ahead prices in various scenarios, and real-time operation constraints.

[0075] 1. Medium- and long-term transaction constraints include medium- and long-term contract electricity volume, monthly electricity purchase limit, medium- and long-term contract time period power, hourly electricity volume, and mutual exclusion of buying and selling of medium- and long-term hourly contract electricity volume.

[0076] 1) Medium- and long-term contract electricity volume:

[0077] in, For medium- and long-term contract electricity volume, For the medium- and long-term contract grid-connected power (MW) during time period t on day d; Let represent the grid connection power (MW) of the medium- and long-term contract for time period t on day d.

[0078] 2) Monthly electricity purchase limit:

[0079] in, , For maximum electrical load, This refers to the number of days in the execution cycle.

[0080] 3) Power during medium- and long-term contract periods:

[0081] in, To maximize download capacity, Power for medium- to long-term contract periods. For maximum internet access capability.

[0082] 4) Hourly electricity consumption: The electricity consumption of each segment (15 minutes) of the contract within the same hour must be equal.

[0083]

[0084] in, The electricity volume is allocated to the first contract segment within the same hour. To allocate electricity for the second contract segment within the same hour. The electricity volume is allocated to the third contract segment within the same hour. The electricity volume of the fourth contract segment within the same hour is allocated.

[0085] 5) Medium- and long-term hourly contract electricity trading is mutually exclusive:

[0086] in, For medium- and long-term contract electricity volume, the Big-M constant is used. For medium- to long-term buy / sell mutually exclusive binary variables, This refers to the grid connection power after the medium- and long-term contracts are split. This refers to the power output after the splitting of medium- and long-term contracts.

[0087] 2. Medium- and long-term deviation assessment constraints include electricity volume summary, price difference calculation, excess assessment, and shortfall assessment.

[0088] 1) The total electricity volume includes both medium- and long-term contracted electricity volume and actual transaction volume. Medium and long-term contract electricity volume:

[0089] in, This refers to the sum of the electricity consumption allocated to the same hour (h) within each day during the medium- to long-term contract period. This represents the net power of medium- and long-term contracts for the 4-hour period on day d (positive for online transactions, negative for offline transactions). Since the contract power allocation is equal for each segment (15 minutes) within the same hour, i.e. ,so This represents the breakdown of medium- to long-term contract electricity volume for hour h on day d.

[0090] Actual transaction volume:

[0091] in, This represents the sum of real-time transaction volumes for the same hour (h) each day within the medium- to long-term contract period under scenario s. The real-time actual net power of the s-th scenario on day d, 4h-3 (i.e., the real-time actual net power of the first segment within the same hour). This refers to the real-time actual net power of the second segment within the same hour for the s-th scenario on day d, 4h-2. This refers to the real-time actual net power of the s-th scenario on day d, time period 4h-1 (i.e., the real-time actual net power of the third segment within the same hour). This refers to the real-time actual net power of the s-th scenario on day d, 4 hours later (i.e., the real-time actual net power of the fourth segment within the same hour).

[0092] 2) Price difference calculation includes real-time average price, excess price difference, and shortfall price difference. Real-time average price:

[0093] in, This refers to the real-time average price of electricity for the same hour (h) each day within the medium- to long-term contract period under scenario s. The real-time price for the s-th scenario on day d, time period 4h-3 (i.e., the real-time price of the first segment within the same hour). This is the real-time price for the s-th scenario on day d, time period 4h-2 (i.e., the real-time price for the second segment within the same hour). The real-time price for the s-th scenario on day d, time period 4h-1 (i.e., the real-time price for the third segment within the same hour). This is the real-time price for the s-th scenario on the d-th day, 4 hours later (i.e., the real-time price for the fourth segment within the same hour).

[0094] Excess spread:

[0095] in, For the excess price difference of the same hour h within the same day during the medium-to-long-term contract period under scenario s; This refers to the real-time average price of electricity for the same hour (h) each day within the medium- to long-term contract period under scenario s. The price is the price of a medium- to long-term contract and is a fixed value of exogenous input.

[0096] Deficit price difference:

[0097] in, For the shortfall price difference in the same hour h within the medium to long term contract period under scenario s; This refers to the real-time average price of electricity for the same hour (h) each day within the medium- to long-term contract period under scenario s. The price is the price of a medium- to long-term contract and is a fixed value of exogenous input.

[0098] 3) Excess performance constraints include excess range coefficient, excess judgment, price judgment, and excess cost. Excess range coefficient :

[0099] in, This represents the actual volume of electricity traded.

[0100] Excess judgment:

[0101] in, For medium- and long-term contract electricity volume, This represents the actual transaction volume. To assess the excess binary variable, if the excess If not exceeded, It can be 0 or 1.

[0102] Price judgment:

[0103] in, For excess price difference, The Big-M constant is used to assess deviation. Let be a binary variable representing the excess price difference. If the excess price difference is positive, then... If the excess spread is 1, then... It can be 0 or 1.

[0104] Excess Fees:

[0105] in, For excess fees, For medium- and long-term contract electricity volume, This is the excess range coefficient. This represents the actual transaction volume. For excess price difference, The excess recovery coefficient is the assessment factor. For excess price spread, binary variable. The Big-M constant is used to assess deviation. The binary variable is used for over-performance assessment.

[0106] If an excess occurs, The value is 1.

[0107] The excess fee is:

[0108] If the excess spread is positive, then The value is 1.

[0109] The excess fee is:

[0110] The actual equivalent is:

[0111] If the excess spread is not positive, then It can be 0 or 1.

[0112] To maximize profits, , .

[0113] If no excess occurs, then To maximize profit, the value should be either 0 or 1. , .

[0114] The boundary for excess performance assessment fees is:

[0115] in, This is the upper limit for expenses exceeding the performance target.

[0116] 4) Deficit assessment constraints include deficit range coefficient, deficit judgment, price judgment, and deficit cost. Deficit range coefficient:

[0117] in, This represents the actual transaction volume. This is the coefficient for the scope of the shortfall.

[0118] Missing Fund Assessment:

[0119] in, This represents the actual transaction volume. For medium- and long-term contract electricity volume, The Big-M constant is used to assess deviation. The binary variable is used for assessment of vacancies. If there is a vacancies, the value is 1; if no vacancies have occurred, the value is 0 or 1.

[0120] Price judgment:

[0121] in, For the price difference of the shortfall, The Big-M constant is used to assess deviation. Let be a binary variable representing the price difference in the shortfall. If the price difference in the shortfall is positive, then... If the shortfall is 1, then... It can be 0 or 1.

[0122] Missing fees:

[0123] in, For the missing fees, This is the coefficient for the range of shortfalls. This represents the actual transaction volume. For medium- and long-term contract electricity volume, For the price difference of the shortfall, The recovery coefficient for the shortfall assessment. The difference in price is a binary variable representing the price gap. The Big-M constant is used to assess deviation. The binary variable is for the assessment of vacancies.

[0124] If a vacancy occurs, then The value is 1.

[0125] The shortfall is:

[0126] If the price difference is positive, then The value is 1.

[0127] The shortfall fee is then:

[0128] The actual equivalent is:

[0129] If the shortfall price difference is not positive, then It can be 0 or 1.

[0130] To maximize profits, , .

[0131] If no vacancies occur, then To maximize profit, the value should be either 0 or 1. .

[0132] The boundary for shortfall assessment fees is:

[0133] in, This is the upper limit for the assessment fee for shortfalls.

[0134] 3. Constraints on segmented volume and price curve reporting in the previous day: The two ends of the application curve are fixed at the maximum offline capacity and the maximum online capacity to ensure that the net power curve crosses 0.

[0135]

[0136] in, The maximum grid connection capacity (not exceeding the maximum power load / application capacity). Maximum internet access capacity; The k-th breakpoint of the net power curve for the entire day of day d is defined as positive for online applications and negative for offline applications. The zeroth breakpoint of the net power curve for the entire day on day d represents the starting point; The full-day net power curve for day d. Each breakpoint represents the endpoint ( (The maximum number of segments allowed under the electricity market rules).

[0137] Breakpoint monotonicity: The power of the curve does not decrease with the segment number, such as... Figure 3 As shown.

[0138]

[0139] This constraint ensures that the breakpoints are arranged in ascending order of power.

[0140] Prefix structure for segment enablement: No gaps are allowed where “the middle is not enabled but the later part is enabled”.

[0141]

[0142] in, This is a flag indicating whether the power curve segment k on day d is enabled. This indicates that segment k is enabled on day d. This indicates that segment k is not enabled on day d; This is a flag indicating whether the power curve for segment k on day d is enabled.

[0143] Segment number constraint: The total number of curve segments throughout the day must meet the application requirements (3–10 segments).

[0144]

[0145] in, , The upper and lower limits of the number of segments are preset. The number of segments used for the entire day's reporting curve (i.e., the final number of reporting segments).

[0146] Power span bidirectional constraint:

[0147] in, The minimum segment length (MW) is defined, ranging from 1 to 5MW. When the k-th segment is enabled... The constraint becomes That is, the power span of the enabled segment must not be less than the minimum segment length to avoid excessive fragmentation; when the k-th segment is not enabled, The constraint becomes ,Right now This means that the power of the unused segment must be exactly the same as the previous segment.

[0148] Price range two-way constraint:

[0149]

[0150] in, It is a very small constant (yuan / MWh). This is a flag indicating whether the power curve segment k on day d is enabled. This indicates that segment k+1 is enabled on day d. This indicates that segment k+1 on day d is not enabled; The price quoted for the (k+1)th segment on day d is (in yuan / MWh). The price (in yuan / MWh) for the kth segment on day d. This is the Big-M constant for the day-to-day reported price. When the (k+1)th segment is activated, The constraint becomes That is, the declared price increases as the segment number increases (the rule is non-decreasing, but a strictly increasing treatment is chosen here to ensure a clear gradient in the price-volume curve). When the (k+1)th segment is not enabled, At this point, the constraint becomes ,Right now This means that the price of the unused segment is exactly the same as the previous segment.

[0151] 4. Constraints on transaction retrieval at the day-ahead price for each scenario: Prefix property of the accepted segment: If the (k+1)th segment is accepted, then the kth segment must also be accepted; and the accepted segment must be an enabled segment.

[0152]

[0153] in, This is a flag indicating whether the k-th segment is accepted in the s-th scenario during the d-th time period t. This is a flag indicating whether the (k+1)th segment is accepted in the (s)th scenario on the (d)th day, time period (t). This is a flag indicating whether the power curve for segment k on day d is enabled.

[0154] Price threshold constraint: If segment k is accepted in time period t on day d, then the price of that segment shall not be higher than the price before the forecast date (the clearing price before the forecast date).

[0155]

[0156] in, The price declared for segment k on day d. For the day-ahead predicted price of the s-th scenario on day d, time period t, The Big-M constant is the price reported so far. This is a flag indicating whether the k-th segment is accepted in the s-th scenario on day d, time period t. The price (in yuan / MWh) for the kth segment on day d. A value greater than 0 represents a very small constant used to strengthen the logic that "if the offer is not accepted, the quoted price must be higher than the previous day's price." This is a flag indicating whether the power curve segment k on day d is enabled. When segment k is enabled but not truncated, , The constraint becomes This means that the price quoted for the unreceived segment is higher than the day-ahead price; when the k-th segment is enabled and truncation occurs, Then the constraint becomes This means the price accepted for the segment is no higher than the price quoted earlier in the day. When segment k is not enabled, Forced If so, the constraint is relaxed as a whole.

[0157] Linearized variables : Represents the power length accepted in the k-th segment of the s-th scenario on the d-th day (linearized by multiplying the breakpoint difference by binary).

[0158]

[0159] in, Let be a linearized variable, representing the length of power received in the k-th segment of the time period t on the d-th day of the s-th scenario. Let k be the breakpoint of the net power curve for the entire day of day d. This represents the (k-1)th breakpoint on the net power curve of the entire day's declarations on day d. This refers to the Big-M constant for power recently declared. To determine whether the k-th segment is accepted in the s-th scenario on day d, time period t, when... hour, ;when hour, .

[0160] The net trading power of the day is obtained by summing the starting point of the curve and the length of the accepted segment.

[0161]

[0162] in, The net power (MW) traded before day t in the s-th scenario on day d. This is the 0th breakpoint (i.e., the starting point) of the net power curve for the entire day of day d. This represents the sum of the power lengths accepted in each segment during the time period t on day d in the s-th scenario.

[0163] The current net power trading is split and mutually exclusive: in each period, there is either net power supply (selling electricity) or net power outflow (buying electricity).

[0164]

[0165] in, Let t be the net daily trading power of the s-th scenario on day d. , These represent the power consumption for internet access and offline access on the day of the s-th scenario. This refers to the Big-M constant for power recently declared. This refers to the binary variable for mutual exclusion between buying and selling transactions.

[0166] Electricity purchase constraints: The maximum net off-grid capacity per time period shall not exceed the maximum power load / application capacity of the associated load category resources.

[0167]

[0168] in, The maximum power load / application capacity of the associated load resources of the transaction unit is limited, and the upper limit of the power purchase capacity is restricted for each time period.

[0169] 5. Real-time operational constraints Photovoltaic utilization and curtailment balance: The sum of actual utilization and curtailment equals the upper limit of availability.

[0170]

[0171] in, The predicted available photovoltaic power for the s-th scenario on day d, time period t. Let t be the amount of solar power curtailed on day d in scenario s. This represents the upper limit of the predicted photovoltaic available power for the s-th scenario on day d, time period t.

[0172] Energy storage charging and discharging mutual exclusion and power limit.

[0173]

[0174] in, For the energy storage charging power of the s-th scenario on day d, time period t, This is the maximum charging power for energy storage. Let t be the energy storage discharge power of the s-th scenario on day d. To store the maximum discharge power, For energy storage, use mutually exclusive binary variables for charging and discharging. Indicates the charging mode. Indicates discharge / idle mode.

[0175] State of charge energy recursion (15-minute step).

[0176]

[0177] in, The SOC energy (MWh) at the end of day d in the s-th scenario is the energy at the end of time period t. Let SOC energy be the energy at the end of day d-1 in the s-th scenario. For energy storage charging efficiency, For the energy storage charging power of the s-th scenario on day d, time period t, For energy storage and discharge efficiency, Let t be the energy storage discharge power of the s-th scenario on day d during time period t.

[0178] Upper and lower bounds of the charged state and the initial and final targets.

[0179]

[0180] in, These are the lower and upper limits of the energy storage SOC, respectively. The SOC energy (MWh) at the end of day d in the s-th scenario is the energy at the end of time period t. The SOC energy (MWh) at the end of day 1 in the s-th scenario. The initial SOC energy, For the s-th scene on day D, the time period is... Final SOC energy (MWh) This refers to the end-of-term SOC energy.

[0181] Real-time net power (exchanged with the grid): PV utilization, energy storage discharge and fuel cell power generation to offset loads, energy storage charging and electrolyzer power consumption.

[0182]

[0183] in, For the s-th scenario, the real-time actual net power during the d-th day, t. The predicted available photovoltaic power for the s-th scenario on day d, time period t. Let t be the energy storage discharge power of the s-th scenario on day d. Let t be the fuel cell power generation capacity during the d-th day of the s-th scenario. The baseline load for scenario s on day d, time period t. For the energy storage charging power of the s-th scenario on day d, time period t, Let t represent the power consumption of the electrolytic cell during the d-th day of the s-th scenario. This represents the net internet usage in the s-th scenario; This indicates the net offline rate in the s-th scenario.

[0184] Grid connection capacity constraint: The actual net power in each time period is limited by the maximum grid connection / disconnection capacity.

[0185]

[0186] in, To maximize download capacity, For maximum internet access capability, The actual net power in real time for the s-th scenario on day d is t.

[0187] Real-time buy / sell splitting and mutual exclusion (used for settlement output).

[0188]

[0189] in, For the s-th scenario, the real-time actual net power during the d-th day, t. , These are the real-time online and offline power consumption, respectively. The real-time power Big-M constant, For real-time buying and selling of mutually exclusive binary variables.

[0190] A stochastic prospective collaborative bidding objective function is constructed, and a bidding model with nonlinear terms is obtained based on the set of equipment operation constraints and the set of market constraints for collaborative bidding.

[0191] Specifically, the total bidding revenue for each scenario is defined, including medium- to long-term fixed revenue, day-ahead / real-time deviation settlement revenue, and policy subsidy revenue, minus equipment operation and maintenance costs, energy storage degradation costs, and deviation assessment fees. An excess loss auxiliary variable is defined for each scenario, representing the difference between the total bidding revenue and the Value at Risk (VaR) threshold. A collaborative bidding objective function is constructed, incorporating both expected total bidding revenue and a conditional Value at Risk (CVaR) penalty term.

[0192] By introducing a risk aversion coefficient, the degree of avoidance of extremely unfavorable scenarios is adjusted. This risk aversion coefficient represents the number of dollars of average expected return that a decision-maker is willing to sacrifice to increase the guaranteed return of 1 dollar in the worst-case scenario. This achieves a balance between maximizing expected bidding returns and minimizing tail risk, effectively avoiding drastic fluctuations in actual settlement returns caused by overly aggressive bidding strategies.

[0193] Specifically, the constraints for CVaR risk measurement are as follows: CVaR represents the worst-case scenario. Average returns in proportional scenarios are an important indicator for measuring tail risk. As a collaborative bidding model, risk is a crucial dimension for evaluation. CVaR risk measurement constraints upgrade the model from simply pursuing the maximization of expected returns to risk-averse optimization that considers extremely unfavorable scenarios.

[0194] Benefits of scenario s

[0195] in, The medium-to-long-term contract electricity volume (MW) for time period t on day d is positive for grid connection and negative for grid disconnection. The price represents a medium- to long-term contract price, which is a fixed value derived from exogenous input. For scenario s, the price before the long-term settlement date on day d of time period t. For scenario s, the long-term settlement reference electricity price for time period t on day d. For time step, This represents the net trading power of the day before yesterday. This represents the actual net power in real time. This is the subsidy coefficient for hydrogen production. Let t be the hydrogen production power of the electrolyzer on day d in scenario s. For carbon credit pricing, CO2 emission intensity of fossil fuels, CO2 emission intensity of synthetic fuels For scenario s, the fuel synthesis output power during time period t on day d. This represents the maintenance cost coefficient for the electrolytic cell. For scenario s, the power consumption of the electrolytic cell during time period t on day d. This represents the fuel cell operation and maintenance cost coefficient. For scenario s, the fuel cell power generation capacity during time period t on day d is... This represents the operation and maintenance cost coefficient for the fuel synthesis unit. For scenario s, the fuel synthesis output power during time period t on day d. The degradation cost coefficient for energy storage charging. For the energy storage charging power of the s-th scenario on day d, time period t, This is the energy storage discharge degradation cost coefficient. Let t be the energy storage discharge power of the s-th scenario on day d. The excess performance fee for hour h under scenario s. The shortfall assessment fee for hour h under scenario s. It represents the total revenue of scenario s.

[0196] Definition of auxiliary variable for excess loss:

[0197] in, Let this be the auxiliary variable for the excess loss in scenario s. VaR, i.e. The quantile-based payoff threshold It represents the total revenue of scenario s.

[0198] like ,but This indicates the difference between the current scenario and VaR; if ,but This indicates that there was no excess loss.

[0199] Conditional Value at Risk (VaR) Calculation:

[0200] in, Value at risk, For confidence level, For scene indexing, For the scene The probability of occurrence For the scene The auxiliary variable for excess loss.

[0201] As the conditional value at risk, the smaller (more negative) the CVaR, the greater the risk.

[0202] for Explanation: It is not the "weight ratio" within the [0,1] interval, but the risk aversion coefficient. It represents the number of "average expected returns" that decision-makers are willing to sacrifice to "increase the guaranteed return of 1 yuan in the worst-case 5% scenario".

[0203] Larger size → More conservative: It's better to earn less on average than to be extremely risky and prone to losses. Smaller → More aggressive: Pursues maximum average profit, tolerates large tail fluctuations Value retrieval logic: The value of is no longer restricted to [0,1], and its reasonable value depends on the expected return and the order of magnitude of CVaR.

[0204] Based on this, the model objective is to maximize the sum of medium- to long-term returns, day-ahead deviation settlement returns, and real-time deviation settlement returns, while including policy subsidies for hydrogen production and synthetic green fuels, and deducting equipment operation and maintenance costs, energy storage degradation costs, and deviation assessment fees. Under the settlement caliber of "medium- to long-term returns + day-ahead deviation + real-time deviation," the collaborative bidding objective function is:

[0205] in, Let be the probability of scenario s occurring. The medium-to-long-term contract electricity volume (MW) for time period t on day d is positive for grid connection and negative for grid disconnection. The price represents a medium- to long-term contract price, which is a fixed value derived from exogenous input. For scenario s, the price before the long-term settlement date on day d of time period t. For scenario s, the long-term settlement reference electricity price for time period t on day d. For time step, This represents the net trading power of the day before yesterday. This represents the actual net power in real time. This is the subsidy coefficient for hydrogen production. Let t be the hydrogen production power of the electrolyzer on day d in scenario s. For carbon credit pricing, CO2 emission intensity of fossil fuels, CO2 emission intensity of synthetic fuels For scenario s, the fuel synthesis output power during time period t on day d. This represents the maintenance cost coefficient for the electrolytic cell. For scenario s, the power consumption of the electrolytic cell during time period t on day d. This represents the fuel cell operation and maintenance cost coefficient. For scenario s, the fuel cell power generation capacity during time period t on day d is... This represents the operation and maintenance cost coefficient for the fuel synthesis unit. For scenario s, the fuel synthesis output power during time period t on day d. The degradation cost coefficient for energy storage charging. For the energy storage charging power of the s-th scenario on day d, time period t, This is the energy storage discharge degradation cost coefficient. Let t be the energy storage discharge power of the s-th scenario on day d. The excess performance fee for hour h under scenario s. The shortfall assessment fee for hour h under scenario s. Represents the risk aversion coefficient. This indicates that no risks are considered. This indicates extreme risk aversion; Represents conditional risk value.

[0206] Linearization is performed on the nonlinear terms in the bidding model to obtain a mixed-integer linear programming model, which yields a multi-market collaborative bidding strategy and internal production scheduling schemes for various scenarios.

[0207] For the nonlinear terms of the product of continuous and binary variables generated during the calculation of deviation assessment costs and the extraction of quantity-price curves, the Big-M method is used to introduce auxiliary continuous variables. By setting a constant M as an auxiliary quantity, linearization is achieved, and the original nonlinear product relationship is transformed into a series of linear inequality constraints, so that the overall model meets the solution requirements of mixed integer linear programming (MILP).

[0208] Linearization includes: The formulas for calculating excess and deficit costs involve the product of continuous and binary variables, which can lead to nonlinearity. The Big M method is used to address this nonlinearity. The processing methods for excess and deficit cost calculations are completely consistent; the excess cost calculation will be used as an example for illustration.

[0209] Introducing auxiliary variables ,make:

[0210] in, As an auxiliary variable for linearizing the excess cost calculation formula, For medium- and long-term contract electricity volume, This is the excess range coefficient. This refers to the actual transaction volume. For excess price difference, The excess recovery coefficient is the assessment factor. This is a binary variable representing the excess price difference.

[0211] Linearize this product using the Big M method: ( (for constants)

[0212]

[0213]

[0214]

[0215] The original formula for calculating excess fees is replaced with:

[0216] in, For excess fees, As an auxiliary variable for linearizing the excess cost calculation formula, For excess price difference, The excess recovery coefficient is the assessment factor. The Big-M constant is used to assess deviation. The binary variable is used for over-performance assessment.

[0217] Through the above linearization process, the original mixed-integer nonlinear programming (MINLP) model is transformed into a mixed-integer linear programming (MILP) model, which can be solved directly using commercial solvers such as Gurobi and CPLEX.

[0218] The solution is obtained by inputting commercial solvers such as Gurobi. Since the collaborative bidding model is a stochastic look-ahead optimization model, its output scheduling plan represents the internal production scheduling scheme for each scenario, used to guide deviation adjustments and equipment control in the real-time phase. The deterministic "first-order decision" that maintains consistency across scenarios—namely, the medium-to-long-term contract decomposition curve and the day-ahead declared quantity and price curve—serves as the final multi-market collaborative bidding strategy, directly used for market declarations to the power trading center. This output format perfectly meets the deterministic declaration requirements of multi-timescale market transactions.

[0219] Example 2 This invention discloses a collaborative bidding device for a virtual power plant of electricity, hydrogen and green fuel with multi-market collaboration and risk measurement, comprising a data unit, an equipment operation constraint unit, a collaborative bidding constraint unit, a model unit, and a solution unit.

[0220] The data unit is used to acquire multi-timescale operating parameters and historical operating data of heterogeneous resources within the virtual power plant, and to obtain an initial operating dataset based on the green fuel delivery demand curve. The equipment operation constraint unit is used to acquire a set of equipment operation constraints that meet the green fuel delivery requirements based on the initial operating dataset and typical scenarios. The collaborative bidding constraint unit is used to acquire a set of market constraints for collaborative bidding. The model unit is used to construct a stochastic look-ahead collaborative bidding objective function, which, based on the set of equipment operation constraints and the set of market constraints for collaborative bidding, yields a bidding model with nonlinear terms to be solved. The solution unit linearizes the nonlinear terms in the bidding model to obtain a mixed-integer linear programming model, and uses this model to obtain multi-market collaborative bidding strategies and internal production scheduling schemes for each scenario.

[0221] For specific limitations regarding the device, please refer to the limitations of the method described above; the corresponding technical effects can be obtained equivalently and will not be repeated here. Each module in the above system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0222] Example 3 Computer equipment can specifically be a terminal or a server. For example... Figure 4 As shown, the computer device includes a processor, memory, network interface, display, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the aforementioned collaborative bidding method for electric-hydrogen-green fuel virtual power plants involving multi-market collaboration and risk measurement. The display screen can be an LCD screen or an e-ink display screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device casing, or an external keyboard, touchpad, or mouse.

[0223] As will be understood by those skilled in the art, computer equipment Figure 4The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computing devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0224] Example 4 A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method.

[0225] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0226] The various embodiments in this specification are described in a progressive manner. For directly identical or similar parts among the embodiments, refer to each other. Each embodiment focuses on its differences from other embodiments. In particular, the device embodiments are basically similar to the method embodiments, so the description is relatively simple; refer to the description of the method embodiments for relevant details. It should be noted that the technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0227] The above-described embodiments are merely preferred embodiments of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various improvements and substitutions without departing from the principles of the present invention, and these improvements and substitutions should also be considered within the scope of protection of the present invention. Therefore, the scope of protection of this invention should be determined by the scope of the claims.

[0228] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

Claims

1. A collaborative bidding method for virtual power plants using electricity, hydrogen, and green fuels, involving multi-market collaboration and risk measurement, characterized in that... Includes the following steps: Obtain multi-timescale operating parameters and historical operating data of heterogeneous resources within the virtual power plant, and obtain the initial operating dataset based on the green fuel delivery demand curve; Based on the initial operational dataset and typical scenarios, obtain a set of equipment operation constraints that meet the requirements for green fuel delivery; Obtain the set of market constraints for collaborative bidding; A stochastic prospective collaborative bidding objective function is constructed, which is a bidding model with nonlinear terms to be solved based on the set of equipment operation constraints and the set of market constraints for collaborative bidding. The nonlinear terms in the bidding model are linearized to obtain a mixed-integer linear programming model. Based on the mixed-integer linear programming model, a multi-market collaborative bidding strategy and an internal production scheduling scheme for each scenario are obtained.

2. The collaborative bidding method for virtual power plants of electricity, hydrogen, and green fuel based on multi-market collaboration and risk measurement as described in claim 1, characterized in that: The set of market constraints for the coordinated bidding includes medium- and long-term market constraints and spot market constraints. Medium- and long-term market constraints include medium- and long-term transaction constraints and medium- and long-term deviation assessment constraints; The constraints in the spot market include the day-ahead segmented volume and price curve reporting constraints, the transaction interception constraints under the day-ahead price in various scenarios, and the real-time operation constraints.

3. The collaborative bidding method for electric-hydrogen-green fuel virtual power plants based on multi-market collaboration and risk measurement as described in claim 2, characterized in that: The objective function for the collaborative bidding is: in, Let be the probability of scenario s occurring. The medium-to-long-term contract electricity volume (MW) for time period t on day d is positive for grid connection and negative for grid disconnection. The price represents a medium- to long-term contract price, which is a fixed value derived from exogenous input. For scenario s, the price before the long-term settlement date on day d of time period t. For scenario s, the long-term settlement reference electricity price for time period t on day d. For time step, This represents the net trading volume of the day before yesterday. This represents the actual net power in real time. This is the subsidy coefficient for hydrogen production. Let t be the hydrogen production power of the electrolyzer on day d in scenario s. For carbon credit pricing, CO2 emission intensity of fossil fuels, CO2 emission intensity of synthetic fuels For scenario s, the fuel synthesis output power during time period t on day d. This represents the maintenance cost coefficient for the electrolytic cell. For scenario s, the power consumption of the electrolytic cell during time period t on day d. This represents the fuel cell operation and maintenance cost coefficient. For scenario s, the fuel cell power generation capacity during time period t on day d is... This represents the operation and maintenance cost coefficient for the fuel synthesis unit. For scenario s, the fuel synthesis output power during time period t on day d. The degradation cost coefficient for energy storage charging. For the energy storage charging power of the s-th scenario on day d, time period t, This is the energy storage discharge degradation cost coefficient. Let t be the energy storage discharge power of the s-th scenario on day d, The excess performance fee for hour h under scenario s. The shortfall assessment fee for hour h under scenario s. Represents the risk aversion coefficient. This indicates that no risks are considered. This indicates extreme risk aversion; Represents conditional risk value.

4. The collaborative bidding method for virtual power plants of electricity, hydrogen, and green fuel based on multi-market collaboration and risk measurement as described in claim 3, characterized in that: The set of equipment operation constraints includes the operation constraints of electrolyzers, fuel cells, hydrogen storage tanks, fuel synthesis systems, and fuel storage tanks; The operating constraints of the electrolyzer include upper and lower limits of operating power, ramp rate, and electro-hydrogen conversion efficiency; The operating constraints of the fuel cell include upper and lower limits of power generation, ramp rate, and hydrogen-to-electricity conversion efficiency. The operational constraints of the hydrogen storage tank include the hydrogen charging and discharging power and the hydrogen storage capacity boundary. The operational constraints of the fuel synthesis system include the upper and lower limits of synthesis power and the hydrogen-fuel conversion efficiency. The operational constraints of the fuel storage tank include the upper limit of the fuel storage tank's charging and discharging power and fuel order delivery.

5. The collaborative bidding method for virtual power plants of electricity-hydrogen-green fuel based on multi-market collaboration and risk measurement according to claim 4, characterized in that: The linearization of nonlinear terms in the bidding model is achieved by setting a constant M to transform the nonlinear product into a linear inequality constraint.

6. The collaborative bidding method for electric-hydrogen-green fuel virtual power plants based on multi-market collaboration and risk measurement according to claim 3, characterized in that: The conditional risk value for: in, Value at risk, For confidence level, For scene indexing, For the scene The probability of occurrence For the scene The auxiliary variable for excess loss.

7. A collaborative bidding device for a virtual power plant of electricity, hydrogen, and green fuel with multi-market collaboration and risk measurement, characterized in that, include: Data Unit: Used to acquire multi-timescale operating parameters and historical operating data of heterogeneous resources within the virtual power plant, and to obtain the initial operating dataset based on the green fuel delivery demand curve; Equipment operation constraint unit: used to obtain a set of equipment operation constraints that meet the requirements for green fuel delivery based on the initial operation dataset and typical scenarios; Collaborative bidding constraint unit: used to obtain the set of market constraints for collaborative bidding; Model Unit: Used to construct a stochastic prospective collaborative bidding objective function, which is a bidding model with nonlinear terms to be solved based on the set of equipment operation constraints and the set of market constraints for collaborative bidding. Solving Unit: Linearizes the nonlinear terms in the bidding model to obtain a mixed-integer linear programming model. Based on the mixed-integer linear programming model, it obtains multi-market collaborative bidding strategies and internal production scheduling schemes for various scenarios.

8. A computer device, characterized in that, include: Processor and computer-readable storage media; The processor is adapted to execute computer programs; The computer-readable storage medium stores a computer program, which, when executed by the processor, implements the multi-market collaboration and risk measurement collaborative bidding method for electric-hydrogen-green fuel virtual power plants as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program adapted to be loaded by a processor and executed as described in any one of claims 1 to 6: a collaborative bidding method for electric-hydrogen-green fuel virtual power plants with multi-market collaboration and risk measurement.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the collaborative bidding method for electric-hydrogen-green fuel virtual power plants with multi-market collaboration and risk measurement as described in any one of claims 1 to 6.

Images