A rolling optimization dispatching method for hydrogen energy microgrid considering the risk of unplanned off-grid disconnection

Abstract

The present invention relates to a microgrid optimization scheduling method. In order to propose an energy scheduling scheme for optimizing microgrids in the grid-connected and off-grid stages from a long-term perspective and improve the resilience of microgrids in dealing with unplanned off-grid, the present invention provides a rolling optimization scheduling method for hydrogen energy microgrids considering the risk of unplanned off-grid. First, the estimated start time and duration of microgrid off-grid are obtained through unplanned off-grid risk assessment; then, a long-term source-load energy matching assessment and optimization are performed: based on the energy prediction information of renewable energy and load in a long-term range, a plan for interaction with the grid, energy storage of the energy storage system, and load transfer is formulated in a long-term range, while maximizing the probability of reliable power supply, minimizing the interaction with the grid and the load transfer; finally, a random optimization scheduling based on conditional risk value is performed on the day-ahead: an optimal day-ahead scheduling plan is formulated. The present invention is mainly applied to microgrid optimization scheduling occasions.

Description

Technical Field

[0001] The present invention relates to a microgrid optimization dispatching method, and in particular to a hydrogen energy microgrid rolling optimization dispatching method taking into account the risk of unplanned off-grid operation. Background Art

[0002] In recent years, hydrogen-electricity coupled microgrids that produce hydrogen from distributed renewable energy have become a new model to promote the consumption of new energy and the development and utilization of hydrogen energy. Especially in the densely industrial and commercial areas in the southeastern coastal areas of China, there is a strong demand for nearby hydrogen, and hydrogen energy microgrids have good application prospects. However, typhoons frequently occur in coastal areas, and the distribution network connected to the microgrid is prone to line and tower failures during typhoons, resulting in unplanned islanding of the microgrid, which seriously threatens the reliability of the microgrid's power supply.

[0003] In order to improve the economy and reliability of microgrid operation, many scholars at home and abroad have conducted a lot of research on microgrid energy management strategies, mainly including rule-based methods and optimization-based methods. Based on the above two methods, domestic and foreign scholars have achieved rich research results in the study of microgrid energy management, including some that consider the scheduling problem of microgrids in the long term, but most of them are based on the establishment of optimization scheduling models based on typical days of each month or season, which means that they still do not consider the unplanned islanding risks that may occur in the future in actual operation. In addition, existing research can extend the operation time of microgrids through optimized scheduling after the microgrid is off-grid, but it only considers the day-ahead optimization problem of microgrids, and cannot optimize the source-load matching and energy reserve of microgrids over a longer time range. Moreover, since the risk of islanding is not considered, it is impossible to formulate a response plan in advance during the grid-connected stage to improve the operational reliability of the microgrid off-grid stage. Summary of the invention

[0004] In order to overcome the shortcomings of the prior art, the present invention aims to optimize the energy scheduling scheme of microgrids in the grid-connected and off-grid stages from a long-term perspective, and improve the resilience of microgrids to unplanned off-grid. To this end, the technical solution adopted by the present invention is a rolling optimization scheduling method for hydrogen energy microgrids considering the risk of unplanned off-grid. First, the estimated start time and duration of microgrid off-grid are obtained through unplanned off-grid risk assessment; then a long-term source-load energy matching assessment and optimization is performed: based on the energy forecast information of renewable energy and load in a long-term range, a plan for the amount of interaction with the grid, energy storage of the energy storage system, and load transfer in a long-term range is formulated, while maximizing the probability of reliable power supply, minimizing the amount of interaction with the grid and the amount of load transfer; finally, a random optimization scheduling based on conditional risk value is performed on the day-ahead: in the day-ahead scheduling stage, the energy reserve constraint after the end of the day-ahead scheduling is considered to formulate the optimal day-ahead scheduling plan.

[0005] The specific steps for unplanned off-grid risk assessment are as follows:

[0006] According to the Batts wind field model, the attenuation and duration of a typhoon are related to the pressure difference at the center of the typhoon. The typhoon duration T is shown in formula (1), where ΔH0(t) is the pressure difference at the center of the typhoon at time t0, in hPa; ξ and θ represent the clockwise angle between the coastline and the typhoon path and the north direction, respectively. T max is the upper limit of the typhoon duration; the number of days a typhoon lasts is d T Day, then

[0007]

[0008] First, the failure probability of the towers and lines is calculated according to the wind speed they are subjected to. Secondly, the cumulative failure probability of the overhead line components during the typhoon is calculated using the infinitesimal method. The towers and lines on the same overhead line form a series system, and then the cumulative failure probability of the overhead line is calculated:

[0009]

[0010] For the off-grid time, we can use formula (2) to determine the failure probability of the overhead line: g When it is greater than 80%, it is considered that the microgrid is off-grid unplanned. begin The time to start going off-grid.

[0011] The long-term source-load energy matching evaluation and optimization method seeks a balance between the maximum probability of reliable power supply from day i to day i+n, the minimum grid interaction cost, and the minimum load transfer energy, while considering a set of constraints related to operation. The mathematical formula is:

[0012] (1) Objective function

[0013]

[0014] In the formula, the probability of reliable power supply P i,d The calculation of includes two parts: the grid-connected stage and the off-grid stage, as shown in formula (4). The right side of the equation represents the probability that energy supply is greater than energy demand, reflecting the source-load energy matching characteristics of the day. and They are the predicted values / prediction errors of photovoltaic, wind power and load energy, respectively. They are all random variables, and their cumulative probability distribution can be obtained by fitting based on historical data; and are the net increase and decrease of energy of the battery energy storage system on day d, respectively; is the net load transfer before the i-th optimization; and are load transfer-in and load transfer-out respectively; and They are the electricity purchased and sold, both of which are 0 during the off-grid period; and are the discharge of the fuel cell and the power consumption of the water electrolysis hydrogen production system respectively;

[0015] (2) Constraints

[0016] 1) Energy storage system constraints

[0017] The energy loss of the energy storage system during the charging and discharging process is not considered in the long-term energy matching evaluation and optimization model. The energy storage system constraints include:

[0018] ① Timing constraints of the state of charge (SOC) of the energy storage system;

[0019] ② Maximum and minimum constraints of SOC;

[0020] ③Operation status constraints of the energy storage system;

[0021] 2) Fuel cell constraints

[0022] The fuel cell constraint is the constraint between the amount of hydrogen consumed by the fuel cell and the amount of electricity generated by the fuel cell;

[0023] 3) Electrolyzer Constraints

[0024] The electrolyzer constraint is the constraint between the amount of hydrogen produced by the electrolyzer and the amount of electricity used by the electrolyzer;

[0025] 4) Hydrogen storage tank constraints

[0026] Hydrogen tank constraints include:

[0027] ① Timing constraints of the hydrogen pressure level LOH of the hydrogen storage tank;

[0028] ② Maximum and minimum constraints of LOH;

[0029] 5) Load energy transfer constraints

[0030] Affected by load characteristics, non-transferable loads do not participate in the formulation of load transfer plans, so the daily load transfer amount does not exceed a certain range of the load energy forecast value of the day. Load energy transfer constraints include:

[0031] ①Energy balance constraints before and after load transfer;

[0032] ② Upper and lower limit constraints on load transfer.

[0033] 6) Constraints on interaction with the power grid

[0034] The interaction constraints with the grid are the maximum and minimum constraints on the microgrid’s purchase / sale of electricity from the distribution network.

[0035] For the optimization problem shown in formula (3), a multi-objective programming algorithm is used to solve it.

[0036] The specific steps of day-ahead stochastic optimization scheduling based on conditional value at risk are as follows:

[0037] A scenario-based day-ahead stochastic optimization model is established to deal with the uncertainty of photovoltaic power and load. The conditional value at risk (CVaR) is used to measure the risk cost caused by the uncertainty of photovoltaic power and load. The Latin hypercube sampling method is used to generate a large number of day-ahead photovoltaic, wind power and load forecast error scenarios, and the backward reduction method is used to reduce the scenarios to obtain typical scenarios of photovoltaic power, wind power and load:

[0038] (1) Objective function

[0039] The objective function of the day-ahead stochastic optimization is the comprehensive dispatch cost, which includes the dispatch cost and risk cost. The formula is as follows:

[0040] f=minC E +σC CVaR (5)

[0041] In formula (5), the scheduling cost C E Including electrolyzer operation cost, fuel cell operation cost, hydrogen storage tank operation cost, battery energy storage operation cost, electricity purchase and sale cost, solar abandonment penalty cost, wind abandonment penalty cost and load loss penalty cost, dispatch risk cost C CVaR It is expressed as:

[0042]

[0043] In the formula, ζ is an auxiliary variable, and its optimal value is Value at Risk, abbreviated as VaR; α is the confidence level of CVaR; C s is the scheduling cost under scenario s; [x] + =max{x,0};

[0044] (2) Constraints

[0045] 1) Electric power balance constraints

[0046] 2) Energy storage operation constraints

[0047] ① Timing constraints of the state of charge (SOC) of the energy storage system considering the charging / discharging efficiency;

[0048] ② Maximum and minimum constraints of SOC;

[0049] ③ Maximum and minimum constraints of charge / discharge power;

[0050] 3) Hydrogen energy subsystem operation constraints

[0051] ④ Constraints between fuel cell power, electrolyzer power and the amount of hydrogen;

[0052] ① Maximum power constraints of fuel cells and electrolyzers;

[0053] ② LOH timing constraints of hydrogen storage tanks;

[0054] ③ Maximum and minimum constraints of LOH;

[0055] 4) Constraints on interaction with the power grid

[0056] The interaction constraints with the grid are the maximum and minimum constraints on the microgrid’s purchase / sale of electricity from the distribution network;

[0057] 5) Reserve power demand constraints

[0058] In order to ensure that the microgrid has sufficient resilience to deal with the risk of grid disconnection, the dispatch strategy formulated by random optimization a few days ago should have a reserve energy of no less than the reserve energy of the hydrogen energy system and the battery energy storage system at the end of the dispatch on the i-th day, including the reserve energy of the hydrogen energy system and the battery energy storage system during the i-th long-term source-load matching evaluation and optimization on the i+1 to i+n days;

[0059] The day-ahead stochastic optimization model is a mixed integer linear programming problem, which is solved using a commercial solver.

[0060] The characteristics and beneficial effects of the present invention are:

[0061] (1) A microgrid island risk assessment model was established during the dynamic process of typhoon passage. Based on the off-grid probability risk and off-grid start and end time assessment, off-grid risk warning and preventive control were achieved.

[0062] (2) Based on the energy forecast of future distributed renewable energy generation and load demand, a rolling evaluation and optimization model for microgrid source-load matching during the grid-connected and off-grid stages is established to formulate energy reserve plans in advance for potential off-grid risks, which is used to guide the day-ahead optimization of microgrids, thereby improving the resilience of day-ahead decision-making plans to cope with future off-grid risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is the long-term optimization framework of the microgrid in the present invention.

[0064] Figure 2 It is a structural topology diagram of the microgrid in the present invention.

[0065] Figure 3 Schematic diagram of the impact of a typhoon on a distribution network in an embodiment of the present invention.

[0066] Figure 4 is the optimization result of the first and second source-load matching model in the embodiment of the present invention.

[0067] (a) The first source-charge matching optimization result;

[0068] (b) The second source-charge matching optimization result.

[0069] Figure 5 It is the scheduling result of 5 consecutive days of the day-ahead optimization in the embodiment of the present invention. DETAILED DESCRIPTION

[0070] In view of the problems existing in the prior art, the purpose of the present invention is to optimize the energy scheduling scheme of microgrids in the grid-connected and off-grid stages from a long-term perspective, and to improve the resilience of microgrids to unplanned off-grid. First, the estimated start time and duration of microgrid off-grid are obtained through the unplanned off-grid risk assessment method; then, based on the energy forecast information of renewable energy and load over a long period of time, the interaction with the grid, energy storage of the energy storage system and load transfer plan over a long period of time are formulated through the optimization method, so as to minimize the interaction with the grid and the load transfer while maximizing the probability of reliable power supply; finally, in the day-ahead scheduling stage, the energy reserve constraint after the end of the day-ahead scheduling is considered to formulate the optimal day-ahead scheduling plan.

[0071] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0072] 1. Unplanned off-grid risk assessment method

[0073] According to the Batts wind field model, the attenuation and duration of a typhoon are related to the pressure difference at the center of the typhoon. The typhoon duration T is shown in formula (1), where ΔH0(t) is the pressure difference at the center of the typhoon at time t0, in hPa; ξ and θ represent the clockwise angle between the coastline and the typhoon's path and the north direction, respectively. max is the upper limit of the typhoon duration, which is set to 240h in the present invention; the number of days the typhoon lasts is d T Day, then

[0074]

[0075] Typhoons usually have difficulty in destroying facilities in buildings and components buried underground, so the present invention only considers the damage of overhead transmission and distribution lines under typhoon disasters. First, the failure probability of the towers and lines is calculated according to the wind speed they are subjected to; secondly, the cumulative failure probability of the overhead line components (including towers and lines) during the typhoon is calculated using the differential element method; the towers and lines on the same overhead line form a series system, and then the cumulative failure probability of the overhead line can be calculated.

[0076]

[0077] The off-grid time can be determined according to formula (2), which represents the failure probability p of the overhead line.g When it is greater than 80%, it is considered that the microgrid is off-grid unplanned. begin is the time to start going offline, corresponding to Figure 1 Taking into account the repair time of the faulty components, the off-grid duration caused by the typhoon is defined as the mth to m+dth T +1 day.

[0078] 2. Long-term source-load energy matching evaluation and optimization

[0079] for Figure 2 For the microgrid shown in the figure, the long-period source-load energy matching evaluation and optimization method proposed in the present invention seeks a balance between the maximum probability of reliable power supply from day i to day i+n, the minimum grid interaction cost, and the minimum load transfer energy, while considering a set of constraints related to operation. The mathematical formula is expressed as:

[0080] (1) Objective function

[0081]

[0082]

[0083] In the formula, the probability of reliable power supply P i,d The calculation of includes two parts: the grid-connected stage and the off-grid stage, as shown in equation (4). The right side of the equation represents the probability that energy supply is greater than energy demand, reflecting the source-load energy matching characteristics of that day. and They are the predicted values / prediction errors of photovoltaic, wind power and load energy, respectively. They are all random variables, and their cumulative probability distribution can be obtained by fitting based on historical data; and are the net increase and decrease of energy of the battery energy storage system on day d, respectively; is the net load transfer before the i-th optimization; and are load transfer-in and load transfer-out respectively; and They are the electricity purchased and sold, both of which are 0 during the off-grid period; and They are the discharge of the fuel cell and the power consumption of the water electrolysis hydrogen production system.

[0084] (2) Constraints

[0085] 1) Energy storage system constraints

[0086] The energy loss of the energy storage system during the charging and discharging process is not considered in the long-term energy matching evaluation and optimization model. The energy storage system constraints include:

[0087] ① Timing constraints of the state of charge (SOC) of the energy storage system;

[0088] ② Maximum and minimum constraints of SOC;

[0089] ③Operation status constraints of the energy storage system.

[0090] 2) Fuel cell constraints

[0091] The fuel cell constraint is the constraint between the amount of hydrogen consumed by the fuel cell and the amount of electricity generated by the fuel cell.

[0092] 3) Electrolyzer Constraints

[0093] The electrolyzer constraint is the constraint between the amount of hydrogen produced by the electrolyzer and the amount of electricity used by the electrolyzer.

[0094] 4) Hydrogen storage tank constraints

[0095] Hydrogen tank constraints include:

[0096] ① Timing constraints of the hydrogen pressure level (LOH) of the hydrogen storage tank;

[0097] ② Maximum and minimum constraints of LOH.

[0098] 5) Load energy transfer constraints

[0099] Affected by load characteristics, non-transferable loads do not participate in the formulation of load transfer plans, so the daily load transfer amount does not exceed a certain range of the load energy forecast value of the day. Load energy transfer constraints include:

[0100] ①Energy balance constraints before and after load transfer;

[0101] ② Upper and lower limit constraints on load transfer.

[0102] 6) Constraints on interaction with the power grid

[0103] The interaction constraints with the grid are the maximum and minimum constraints on the microgrid’s purchase / sale of electricity from the distribution network.

[0104] In view of the optimization problem shown in formula (3), the present invention adopts a multi-objective programming algorithm to solve it.

[0105] 3. Day-ahead stochastic optimization scheduling based on conditional value at risk

[0106] The present invention establishes a scenario-based day-ahead stochastic optimization model to deal with the uncertainty of photovoltaic and load, and CVaR (Conditional Value at Risk) is used to measure the risk cost caused by the uncertainty of photovoltaic and load. The Latin hypercube sampling method is used to generate a large number of day-ahead photovoltaic, wind power and load forecast error scenarios, and the backward reduction method is used to reduce the scenarios to obtain typical scenarios of photovoltaic, wind power and load.

[0107] (1) Objective function

[0108] The objective function of the day-ahead stochastic optimization is the comprehensive dispatch cost, which includes the dispatch cost and risk cost. The formula is as follows:

[0109] f=minC E +σC CVaR (5)

[0110] In formula (5), the scheduling cost C E It includes the operation cost of electrolyzer, fuel cell, hydrogen storage tank, battery energy storage, electricity purchase and sale cost, solar power abandonment penalty cost, wind power abandonment penalty cost and load loss penalty cost. CVaR It can be expressed as:

[0111]

[0112] In the formula, ζ is an auxiliary variable, and its optimal value is Value at Risk (VaR); α is the confidence level of CVaR; C s is the scheduling cost under scenario s; [x] + =max{x,0}.

[0113] (2) Constraints

[0114] 1) Electric power balance constraints

[0115] 2) Energy storage operation constraints

[0116] ① Timing constraints of the state of charge (SOC) of the energy storage system considering the charging / discharging efficiency;

[0117] ② Maximum and minimum constraints of SOC;

[0118] ③ Maximum and minimum constraints on charge / discharge power.

[0119] 3) Hydrogen energy subsystem operation constraints

[0120] ① Constraints between fuel cell power, electrolyzer power and the amount of hydrogen;

[0121] ② Maximum power constraints of fuel cells and electrolyzers;

[0122] ③ LOH timing constraints of hydrogen storage tanks;

[0123] ④ Maximum and minimum constraints of LOH.

[0124] 4) Constraints on interaction with the power grid

[0125] The interaction constraints with the grid are the maximum and minimum constraints on the microgrid’s purchase / sale of electricity from the distribution network.

[0126] 5) Reserve power demand constraints

[0127] To ensure that the microgrid has sufficient resilience to cope with the risk of off-grid disconnection, the dispatch strategy formulated by random optimization a few days ago should have a reserve energy (including hydrogen energy system and battery energy storage system) at the end of the i-th day of dispatching, which should not be less than the reserve energy from i+1 to i+n days during the i-th long-cycle source-load matching evaluation and optimization.

[0128] The day-ahead stochastic optimization model is a mixed integer linear programming problem, which is solved by a mature commercial solver in the present invention.

[0129] 4. Examples

[0130] Schematic diagram of the impact of typhoon on the distribution network connected to the microgrid Figure 3 As shown in the figure, the microgrid is connected to the upper substation via towers 1, 4, 5, 8, 9, and 10. The typhoon landed on the coastline and gradually affected the connection line between the microgrid and the upper substation during its movement.

[0131] According to the aforementioned off-grid risk assessment model, the probability of distribution network failure on the first and second days after the typhoon landed is 0.9238 and 0.9248, which is greater than the set threshold of 0.8. Therefore, the typhoon lasts for 2 days, and considering the maintenance time is 1 day, the continuous off-grid time of the microgrid is 3 days. Assuming that the typhoon information is obtained from the meteorological department 2 days in advance, the grid-connected operation time of this example is 2 days, and the off-grid operation time is 3 days.

[0132] (1) Energy matching evaluation and optimization results analysis

[0133] In the five consecutive long-term source-load matching assessments and optimizations, the probability of reliable power supply every day was greater than 0.9. The results of the first and second matching and optimizations were as follows: Figure 4 As shown in the figure, it can be seen that during the off-grid period, wind and solar power generation is low, and part of the load is transferred to the grid-connected stage to ensure the maximum probability of reliable power supply. For example, Figure 4 Part of the load in (a) is transferred from day 3 to day 1. During the grid connection period, the surplus electricity from wind and solar power is stored through water electrolysis to produce hydrogen and battery energy storage charging, such as Figure 4As shown on days 1 and 2 in (a) and day 2 in (b), during the off-grid period when the wind and solar output is low, the fuel cell and battery energy storage system supplement the load power shortage, as shown on days 3, 4, and 5 in the figure.

[0134] (2) Analysis of day-ahead random optimization results

[0135] The day-ahead random optimization scheduling results for each day within the optimization period are as follows: Figure 5 As shown. Figure 5 It can be seen that on the first day, the wind and solar output was higher, and part of the load in the off-grid stage was transferred in. At the same time, during the period of high wind and solar output, the excess wind and solar power generation in addition to the supply load is stored through the battery energy storage system and water electrolysis hydrogen production, and electricity is sold during the period of higher electricity prices (such as 18:00-22:00 on the first day and 18:00-20:00 on the second day), and electricity is purchased during the period of lower electricity prices (such as 23:00-00:00 on the first and second days) to obtain maximum benefits.

[0136] (3) Conclusion

[0137] The results of the embodiments show that the method proposed in the present invention can significantly enhance the risk resistance of the microgrid to cope with extreme natural disasters and improve the power supply reliability.

[0138] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed in the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for rolling optimization scheduling of hydrogen energy microgrids considering the risk of unplanned off-grid operation, characterized in that: Firstly, the estimated start time and duration of microgrid off-grid are obtained through unplanned off-grid risk assessment; Then, long-term source-load energy matching evaluation and optimization are carried out: based on the energy forecast information of renewable energy and load in the long term, a plan for interaction with the grid, energy storage of the energy storage system and load transfer in the long term is formulated, so as to minimize the interaction with the grid and load transfer while maximizing the probability of reliable power supply; finally, a day-ahead random optimization scheduling based on conditional risk value is carried out: in the day-ahead scheduling stage, the energy reserve constraint after the day-ahead scheduling is considered to formulate the optimal day-ahead scheduling plan; among them, the specific steps of unplanned off-grid risk assessment are as follows: According to the Batts wind field model, the attenuation and duration of a typhoon are related to the pressure difference at the center of the typhoon. The typhoon duration T is shown in formula (1), where ΔH0(t) is the pressure difference at the center of the typhoon at time t0, in hPa; ξ and θ represent the clockwise angle between the coastline and the typhoon path and the north direction, respectively. T max is the upper limit of the typhoon duration; the number of days a typhoon lasts is d T Day, then First, the failure probability of the towers and lines is calculated according to the wind speed they are subjected to. Secondly, the cumulative failure probability of the overhead line components during the typhoon is calculated using the infinitesimal method. The towers and lines on the same overhead line form a series system, and then the cumulative failure probability of the overhead line is calculated: For the off-grid time, we can use formula (2) to determine the failure probability of the overhead line: g When it is greater than 80%, it is considered that the microgrid is off-grid unplanned. begin The time to start going off-grid.

2. The method for rolling optimization scheduling of hydrogen energy microgrids considering the risk of unplanned off-grid operation as claimed in claim 1 is characterized in that: The long-term source-load energy matching evaluation and optimization method seeks a balance between the maximum probability of reliable power supply, the minimum grid interaction cost, and the minimum load transfer energy between the i-th day and the i+n-th day, while considering a set of constraints related to operation. The mathematical formula is expressed as: (1) Objective function Where, the probability of reliable power supply P i,d The calculation of includes two parts: the grid-connected stage and the off-grid stage, as shown in formula (4). The right side of the equation represents the probability that energy supply is greater than energy demand, reflecting the source-load energy matching characteristics of the day. and They are the predicted values / prediction errors of photovoltaic, wind power and load energy, respectively. They are all random variables, and their cumulative probability distribution can be obtained by fitting the historical data; and are the net increase and decrease of energy of the battery energy storage system on day d, respectively; is the net load transfer before the i-th optimization; and are load transfer-in and load transfer-out respectively; and They are the electricity purchased and sold, both of which are 0 during off-grid periods; and are the discharge of the fuel cell and the power consumption of the water electrolysis hydrogen production system respectively; (2) Constraints 1) Energy storage system constraints The energy loss of the energy storage system during the charging and discharging process is not considered in the long-term energy matching evaluation and optimization model. The energy storage system constraints include: ① Timing constraints of the state of charge (SOC) of the energy storage system; ② Maximum and minimum constraints of SOC; ③Operation status constraints of the energy storage system; 2) Fuel cell constraints The fuel cell constraint is the constraint between the amount of hydrogen consumed by the fuel cell and the amount of electricity generated by the fuel cell; 3) Electrolyzer Constraints The electrolyzer constraint is the constraint between the amount of hydrogen produced by the electrolyzer and the amount of electricity used by the electrolyzer; 4) Hydrogen storage tank constraints Hydrogen tank constraints include: ① Timing constraints of the hydrogen pressure level LOH of the hydrogen storage tank; ② Maximum and minimum constraints of LOH; 5) Load energy transfer constraints Affected by load characteristics, non-transferable loads do not participate in the formulation of load transfer plans. Therefore, the daily load transfer amount does not exceed a certain range of the load energy forecast value of the day. The load energy transfer constraints include: ①Energy balance constraints before and after load transfer; ② Upper and lower limit constraints on load transfer; 6) Constraints on interaction with the power grid The interaction constraints with the grid are the maximum and minimum constraints on the microgrid’s purchase / sale of electricity from the distribution network.

3. The method for rolling optimization scheduling of hydrogen energy microgrids considering the risk of unplanned off-grid operation as claimed in claim 2 is characterized in that: For the optimization problem shown in formula (3), a multi-objective programming algorithm is used to solve it.

4. The method for rolling optimization scheduling of hydrogen energy microgrids considering the risk of unplanned off-grid operation as claimed in claim 3 is characterized in that: The specific steps of day-ahead stochastic optimization scheduling based on conditional value at risk are as follows: A scenario-based day-ahead stochastic optimization model is established to deal with the uncertainty of photovoltaic power and load. The conditional value at risk (CVaR) is used to measure the risk cost caused by the uncertainty of photovoltaic power and load. The Latin hypercube sampling method is used to generate a large number of day-ahead photovoltaic, wind power and load forecast error scenarios, and the backward reduction method is used to reduce the scenarios to obtain typical scenarios of photovoltaic power, wind power and load: (1) Objective function The objective function of the day-ahead stochastic optimization is the comprehensive dispatch cost, which includes the dispatch cost and risk cost. The formula is as follows: f=minC E +σC CVaR (5) In formula (5), the scheduling cost C E Including electrolyzer operation cost, fuel cell operation cost, hydrogen storage tank operation cost, battery energy storage operation cost, electricity purchase and sale cost, solar abandonment penalty cost, wind abandonment penalty cost and load loss penalty cost, dispatch risk cost C CVaR It is expressed as: Where ζ is an auxiliary variable, whose optimal value is Value at Risk, or VaR; α is the confidence level of CVaR; Cs is the dispatch cost under scenario s; [x] + =max{x,0}; (2) Constraints 1) Electric power balance constraints 2) Energy storage operation constraints ① Timing constraints of the state of charge (SOC) of the energy storage system considering the charging / discharging efficiency; ② Maximum and minimum constraints of SOC; ③ Maximum and minimum constraints of charge / discharge power; 3) Operation constraints of hydrogen energy subsystem ① Constraints between fuel cell power, electrolyzer power and the amount of hydrogen; ② Maximum power constraints of fuel cells and electrolyzers; ③ LOH timing constraints of hydrogen storage tanks; ④ Maximum and minimum constraints of LOH; 4) Constraints on interaction with the power grid The interaction constraints with the grid are the maximum and minimum constraints on the microgrid’s purchase / sale of electricity from the distribution network; 5) Reserve power demand constraints To ensure that the microgrid has sufficient resilience to cope with the risk of off-grid disconnection, the dispatch strategy formulated by random optimization a few days ago should ensure that the reserve energy of the hydrogen energy system and the battery energy storage system at the end of the i-th day of dispatch should not be less than the reserve energy of the i+1 to i+n days during the i-th long-cycle source-load matching evaluation and optimization.

Images