Hydrogen energy storage based robust planning method for islanded microgrid with distributed thermal

Abstract

The application discloses an independent micro-grid (IGDT) robust planning method containing hydrogen energy storage, belongs to the technical field of power supply capacity optimization planning in micro-grid planning, and has the advantages of scientific rationality, strong applicability and good effect. Firstly, a basic structure of the independent micro-grid containing hydrogen energy storage is constructed, the basic working principle of the hydrogen energy storage system is described based on the structure, and a simplified mathematical model of the hydrogen energy storage system is established. Secondly, an uncertain scenario set is established based on the IGDT robust optimization idea. Then, a double-layer robust optimization planning model of the independent micro-grid considering incentive DR is established, wherein the outer-layer model takes the minimum annual value of the comprehensive cost of the micro-grid as the optimization target, and the inner-layer model takes the minimum annual operation cost of the micro-grid as the optimization target. Finally, a genetic algorithm with an elite reservation strategy is adopted to solve the double-layer robust optimization planning model, so as to provide an optimization planning scheme for micro-grid decision makers. The method has the advantages of scientific rationality, strong applicability and good effect.

Description

Technical Field

[0001] This invention belongs to the field of power capacity optimization planning technology in microgrid planning, and in particular relates to a robust planning method for independent microgrids that comprehensively considers the uncertainties of hydrogen energy storage, wind and solar energy processing, and incentive-driven demand response. Background Technology

[0002] With rapid economic development, energy crises and environmental pollution have become increasingly prominent. Improving energy efficiency, exploring renewable energy sources, and promoting renewable energy production are critical issues that urgently need to be addressed. Microgrid technology, built using renewable energy and energy storage technologies, has become an important means to promote the integration and consumption of renewable energy. However, current microgrid energy storage mainly relies on batteries, which have short lifespans, are expensive, and cause environmental pollution, thus limiting the development of microgrids. Hydrogen energy storage, as a new type of energy storage, has advantages such as being clean, efficient, having high energy density, and large capacity. In the planning of independent microgrids with a high proportion of renewable energy penetration, the uncertainty of wind and solar power output makes it difficult to achieve lean planning goals. Considering the uncertainty of wind and solar power output can lead to more refined planning results. Furthermore, independent microgrids are not connected to the main power grid and are mainly used for power supply in remote areas, resulting in high construction costs. By introducing incentive-based demand response (DR), the electricity consumption time of users' transferable loads can be adjusted, making the renewable energy output curve and load curve more closely aligned in time during the operating cycle. This can promote the consumption of renewable energy and improve the economics of microgrids. Summary of the Invention

[0003] The technical problem to be solved by this invention is to provide a robust planning method for independent microgrids with hydrogen energy storage using IGDT (Information Gap Decision Theory). Using hydrogen energy storage as the energy storage system, an independent microgrid model is established, comprising a wind turbine (WT), a photovoltaic cell (PV), a diesel generator (DEG), an electrolyzer, a fuel cell, and a hydrogen storage tank. Based on the robust optimization concept of information gap decision theory (IGDT), uncertain operating scenarios in the microgrid are constructed. A two-layer robust optimization planning model for the independent microgrid, comprehensively considering IGDT and incentive-based DR (Devices Retention), is constructed. The outer layer model aims to minimize the annual value of the microgrid's overall cost, while the inner layer model aims to minimize the annual operating cost of the microgrid. A genetic algorithm with an elite retention strategy is used to solve the model, providing microgrid decision-makers with a more reasonable planning scheme.

[0004] A robust planning method for an independent microgrid IGDT containing hydrogen energy storage includes the following steps, which are performed sequentially.

[0005] Step 1: Establish an independent microgrid model of the hydrogen energy storage system, which includes an electrolyzer model, a fuel cell model, and a hydrogen storage tank model.

[0006] Step 2: Based on the IGDT robust optimization method, establish an independent microgrid operation scenario model, and use the entropy weight method to evaluate the weights of the uncertain parameters wind power and photovoltaic output;

[0007] Step 3: By considering the outer and inner optimization models, a two-layer robust optimization planning model for independent microgrids that takes into account IGDT is established. A genetic algorithm with an elite retention strategy is used to solve the two-layer robust optimization planning model to obtain a planning scheme for independent microgrids with hydrogen energy storage.

[0008] The independent microgrid structure in step one includes a photovoltaic array, a diesel generator set, a wind turbine generator set, loads, and a hydrogen energy storage system consisting of a hydrogen storage tank, a fuel cell, and an electrolyzer. The hydrogen energy storage system, the wind turbine generator set, and the photovoltaic array are connected to the AC bus through different converters. The user load consists of time-shiftable loads and rigid loads.

[0009] In a hydrogen energy storage system, when the power generated by wind and solar power exceeds the load demand, the remaining electrical energy is used to electrolyze water using an electrolyzer, and the generated hydrogen is stored in a hydrogen storage tank. When the power generated by wind and solar power is insufficient to meet the load demand, the fuel cell uses the hydrogen in the hydrogen storage tank as fuel to generate electricity to meet the load demand.

[0010] ① Electrolytic cell model

[0011] An alkaline electrolyzer is used to electrolyze water to produce hydrogen and oxygen. The output power is shown in equation (1).

[0012] P el,out =η el P el,in (1)

[0013] In the formula, P el,in P is the input power of the electrolytic cell. el,out η is the output power of the electrolytic cell. el The efficiency of the electrolytic cell;

[0014] The maximum input power of the electrolyzer is limited by its capacity and the remaining energy storage capacity of the hydrogen storage tank. The maximum input power of the electrolyzer is shown in equations (2) and (3).

[0015]

[0016]

[0017] In the formula, C is the maximum input power of the electrolytic cell; el,N This refers to the rated capacity of the electrolytic cell; C represents the maximum energy storage capacity of the hydrogen storage tank. tank,N E represents the rated capacity of the hydrogen storage tank. tank,t Let be the energy stored in the hydrogen storage tank at time t; Δt be the time interval; analogous to the state of charge of a battery, the state of charge of the hydrogen storage tank is defined as S. SOC,tank , This represents the maximum state of charge of the hydrogen storage tank.

[0018] ② Fuel cell model

[0019] Fuel cells release energy by burning hydrogen and oxygen, and their output power is shown in equation (4).

[0020] P fc,out =η fc P fc,in (4)

[0021] In the formula, P fc,out P represents the output power of the fuel cell. fc,in η is the input power of the fuel cell. fc To improve the working efficiency of fuel cells;

[0022] The maximum output power of a fuel cell is also limited by its capacity and the remaining capacity of the hydrogen storage tank, as shown in equations (5) and (6).

[0023]

[0024]

[0025] In the formula, This represents the maximum output power of the fuel cell; This represents the minimum state of charge of the hydrogen storage tank. C represents the minimum energy storage capacity of the hydrogen storage tank. fc,N This refers to the rated capacity of the fuel cell;

[0026] ③ Hydrogen storage tank model

[0027] The hydrogen storage tank is used to store hydrogen produced by the electrolyzer and to supply hydrogen to the fuel cell, as shown in equation (7).

[0028]

[0029] In the formula, η tank To improve the working efficiency of the hydrogen storage tank.

[0030] The independent microgrid operation scenario model in step two is as follows:

[0031]

[0032] In the formula, α is the fluctuation range of the uncertain parameter; y is the decision variable; and Y is the uncertain parameter. These are the predicted values ​​of uncertain parameters; The fluctuation range of the uncertain parameter Y does not exceed μ is the risk aversion coefficient, representing the percentage increase in costs due to uncertainty that microgrid decision-makers can accept. A larger aversion coefficient means a larger budget for the planning scheme that the decision-maker can tolerate, and a stronger ability of the planning model to cope with uncertainty. H(Y,y) and G(Y,y) are the equality and inequality constraints, respectively; F represents the choice of Y under the deterministic model. The target cost at that time; F' is the pessimistic cost; (1+μ)F is the maximum pessimistic cost target value that the decision-maker can accept;

[0033] The uncertain parameters are the output power of both wind power and photovoltaic power, and their fluctuation range is shown in equations (9) and (10).

[0034]

[0035]

[0036] In the formula, and These are the predicted output values ​​for wind turbines and photovoltaic arrays, respectively; P wt,s,t and P pv,s,t These are the actual output values ​​of the wind turbine and the photovoltaic array, respectively; α wt and α pv These represent the fluctuation ranges in the output of wind turbines and photovoltaic arrays, respectively, and their values ​​are directly related to the prediction error. The set of fluctuation ranges in wind turbine output; This represents the set of fluctuation ranges for photovoltaic arrays.

[0037] When the actual output of wind and solar power is much less than the predicted value, the system operating cost is the highest, and its value is shown in equations (11) and (12).

[0038]

[0039]

[0040] The entropy weight method is used to calculate the weight values ​​of wind power output and photovoltaic output, and the importance of their uncertainty factors is evaluated, as shown in equation (13).

[0041]

[0042] In the formula, V is the standardized value of the i-th indicator in the t-th alternative; i,tH represents the proportion of the i-th evaluation index of the t-th alternative; i w is the entropy value of the i-th index; i Let be the entropy weight of the i-th index;

[0043] Each uncertain variable is considered as an indicator of system uncertainty. The corresponding entropy weights are calculated using the predicted values ​​of wind and solar power. Based on the entropy weights, the proportion of the corresponding uncertain variable in the total uncertainty is calculated, as shown in equation (14).

[0044]

[0045] In the formula, w wt The proportion of uncertainty in wind turbine output; w pv The percentage of uncertainty in the output of the photovoltaic array.

[0046] The method for constructing the independent microgrid two-layer robust optimization planning model based on IGDT in step three is as follows:

[0047] ①Outer layer optimization model

[0048] The outer layer aims to minimize the annual comprehensive cost Z of the microgrid. The set of decision variables x for outer layer optimization represents the rated capacity of fuel cells, electrolyzers, hydrogen storage tanks, photovoltaic arrays, and wind turbines. In a deterministic scenario, the expected target cost function Z of the outer layer is:

[0049]

[0050] In the formula, Z inv Z is the annual value of equipment investment cost; ope The annual operating cost of the system;

[0051] When wind and solar power output develops in an unfavorable direction, the maximum pessimistic operating cost and total cost acceptable to microgrid decision-makers after the introduction of IGDT can be rewritten as equations (16) and (17).

[0052] Z' ope =(1+μ)Z ope (16)

[0053]

[0054] In the formula, Z' ope Z' represents the maximum pessimistic operating cost acceptable to the decision-maker; Z' is the annual value of the overall cost of the microgrid.

[0055] Z inv This includes the average annual investment cost and annual maintenance cost of each piece of equipment, calculated using formulas (18) and (19).

[0056]

[0057]

[0058] In the formula, k represents the equipment type, k = 1, 2, ..., 5, representing wind turbine, photovoltaic array, electrolyzer, fuel cell, and hydrogen storage tank, respectively; R is the annual value operator; L is the system operating life; Z f,k Initial investment cost per unit capacity of equipment of type k; Z m,k Annual maintenance cost of the equipment; ε k λ represents the number or capacity of the equipment to be installed; λ is the discount rate.

[0059] Due to the influence of the site construction area, the inequality constraints on the investment capacity of various equipment are shown in equation (20).

[0060]

[0061] In the formula, C de,N C wt,N and C pv,N This refers to the rated installed capacity of diesel generator sets, wind turbine generator sets, and photovoltaic arrays. and These represent the minimum and maximum installed capacity of the photovoltaic array, respectively. and These are the minimum and maximum values ​​of the wind turbine assembly capacity, respectively. and These are the minimum and maximum values ​​of the diesel engine assembly capacity, respectively. and These are the minimum and maximum values ​​of the electrolytic cell capacity, respectively. and These represent the minimum and maximum values ​​of the fuel cell capacity, respectively. and These are the minimum and maximum values ​​of the hydrogen storage tank capacity, respectively.

[0062] ②Inner layer optimization model

[0063] The annual operating cost of the inner layer includes fuel cost and DR compensation cost. The decision variables of the inner layer are the charging and discharging power of wind turbines, photovoltaic arrays, diesel generator sets and hydrogen energy storage systems at different times of the year. The inner layer optimization model is shown in Equation (21).

[0064]

[0065] In the formula, Z dr Annual load transfer compensation cost; Z fuel The annual fuel cost for the diesel generator set;

[0066] The wind speed, solar irradiance, and load curves of typical days in summer, winter, spring, and autumn were obtained using the K-means clustering algorithm to replace the wind speed, solar irradiance, and load data for each day of the season, as shown in Equation (22).

[0067]

[0068] In the formula, S represents the number of typical days in a year, s = 1, 2, ..., S; n s Z represents the number of days in a typical year, specifically the s-th day; dr,s and Z fuel,s This is expressed as the load transfer compensation cost and diesel generator set fuel cost on the s-th typical day;

[0069] The fuel cost of diesel generator sets is shown in equations (23) and (24).

[0070]

[0071]

[0072] In the formula, β is the unit price of diesel fuel; Ω s It is a set of three typical days in spring, autumn, summer, and winter; ω s,t This is a binary variable; a value of 0 indicates that the diesel generator unit is shut down during the t-th typical day, and a value of 1 indicates that it is operating. s,t P represents the fuel consumption of the diesel generator set. de,s,t The output of the diesel generator set during s typical days; and , representing the maximum and minimum output of the diesel generator set, respectively; b represents the slope of the fuel consumption-power curve; c represents the intercept coefficient.

[0073] An incentive-based DR is introduced, with the minimum sum of the absolute values ​​of the differences between renewable energy generation and load demand at each time point as the optimization objective of the incentive-based DR. The incentive-based DR is shown in equation (25).

[0074]

[0075] In the formula, P pv,s,t P wt,s,t The output power of the photovoltaic array and the wind turbine at time t on the s-th typical day are respectively; P L0,s,t P L1,s,t These represent the load power at time t before and after the excitation-type DR; ΔP L,s,t The load transfer amount at time t on the s-th typical day is greater than 0 if it is load transfer in, and less than 0 if it is load transfer out.

[0076] After considering DR, the main cost is the compensation cost that needs to be given to users to adjust time-shiftable loads, as shown in equation (26).

[0077]

[0078] In the formula, C dr The cost of transferring and compensating for unit electricity load;

[0079] ③Establish a two-layer robust optimization planning model for independent microgrids that takes into account IGDT, as shown in equation (33).

[0080]

[0081] The inner optimization model satisfies the following constraints:

[0082] a. Power balance constraint

[0083] P el,in,s,t +P L1,s,t =P pv,s,t +P wt,s,t +P de,s,t +P fc,out,s,t (27)

[0084] In the formula, P el,in,s,t and P fc,out,s,t These represent the power outputs of the electrolyzer and the fuel cell at time t on the s-th typical day, respectively.

[0085] b. Output constraints of distributed power sources

[0086]

[0087] In the formula, P pv,N P wt,N and P de,N These are the rated outputs of the photovoltaic array, wind turbine, and diesel generator, respectively.

[0088] c. Constraints of hydrogen energy storage systems

[0089]

[0090] In the formula, E tank,s,t The energy storage capacity of the hydrogen storage tank during the typical t-period within a day; and These are the upper and lower limits of the energy storage capacity of hydrogen storage tanks; and This represents the maximum power of fuel cells and electrolyzers during a typical daytime period (t). and These are the upper and lower limits of the state of charge of the hydrogen storage tank; S SOC,tank,s,t The state of charge of a hydrogen storage tank during a typical daytime period t; and These represent the state of charge of the hydrogen storage tank at the beginning and end of an operating cycle; t0 and tN These represent the beginning and end periods of the operating cycle;

[0091] d. Constraints on the proportion of electricity generated by diesel generator sets

[0092]

[0093] e. Load transfer constraints

[0094]

[0095]

[0096] In the formula, The maximum load transfer amount at time t on a typical day s; t0 is the load transfer-in period; t1 is the load transfer-out period; T n This is the nth running cycle.

[0097] Through the above design scheme, the present invention can bring the following beneficial effects: a robust planning method for independent microgrids with hydrogen energy storage (IGDT). First, the basic structure of an independent microgrid with hydrogen energy storage is constructed, and the basic working principle of the hydrogen energy storage system is explained based on this structure, and its simplified mathematical model is established. Second, an uncertain scenario set is established based on the robust optimization idea of ​​IGDT. Then, a two-layer robust optimization planning model for independent microgrids considering incentive-based dynamic range (DR) is established, wherein the outer model takes the minimization of the annual value of the microgrid's comprehensive cost as the optimization objective, and the inner model takes the minimization of the microgrid's annual operating cost as the optimization objective. Finally, a genetic algorithm with an elite retention strategy is used to solve the two-layer robust optimization planning model, providing an optimized planning scheme for microgrid decision-makers. This method has the advantages of being scientifically sound, highly applicable, and effective. Attached Figure Description

[0098] The present invention will be further described below with reference to the accompanying drawings and specific embodiments:

[0099] Figure 1 This is a schematic diagram of the independent microgrid structure involved in the robust planning method of the independent microgrid IGDT containing hydrogen energy storage according to the present invention.

[0100] Figure 2 This is a schematic diagram of a typical daily wind speed curve for the robust planning method of an independent microgrid IGDT containing hydrogen energy storage according to the present invention.

[0101] Figure 3 This is a schematic diagram of a typical daily solar irradiance curve for the robust planning method of an independent microgrid IGDT containing hydrogen energy storage according to the present invention.

[0102] Figure 4 This is a schematic diagram of a typical daily load curve for the robust planning method of an independent microgrid IGDT containing hydrogen energy storage according to the present invention.

[0103] Figure 5 This invention presents a robust planning method for independent microgrids containing hydrogen energy storage (IGDT), and a flowchart of the solution process for the two-layer robust optimization planning of microgrids.

[0104] Figure 6 This is a schematic diagram of the operation results of a microgrid under a typical summer day in a deterministic scenario, which is a specific embodiment of the present invention 2.

[0105] Figure 7 This is a schematic diagram of the operation results of a microgrid under a typical summer day with uncertainty, which is a specific embodiment of the present invention.

[0106] Figure 8 This is a schematic diagram of the operation results of a typical microgrid in winter after introducing an excitation-type DR in specific implementation scheme 4 of the present invention.

[0107] Figure 9 This is a schematic diagram of the operation results of a typical microgrid in winter after introducing an excitation-type DR in specific implementation scheme 5 of the present invention.

[0108] Figure 10 This is a schematic diagram of the iterative convergence curve of specific embodiment 5 of the present invention.

[0109] Figure 11 This diagram illustrates the impact of different μ values ​​on α and the annual operating cost of the microgrid under specific embodiments of the present invention.

[0110] Figure 12 This diagram illustrates the annual operating cost of a microgrid under uncertain scenarios in specific embodiments of the present invention. Detailed Implementation

[0111] The following uses appendix Figures 1-12 The present invention will be further illustrated by the examples.

[0112] Specific example: The robust planning method of IGDT for independent microgrids containing hydrogen energy storage according to the present invention includes: establishing an independent microgrid model containing hydrogen energy storage, constructing an independent microgrid operation scenario set based on the IGDT robust optimization idea, and establishing a two-layer robust optimization planning model for independent microgrids that comprehensively considers wind and solar uncertainties and incentive-driven demand response. The specific content is as follows:

[0113] 1) Example Background and Parameter Settings

[0114] Figure 1This diagram illustrates an independent microgrid structure, including a photovoltaic array, diesel generators, wind turbines, loads, and a hydrogen energy storage system comprised of a hydrogen storage tank, fuel cells, and an electrolyzer. The hydrogen energy storage system, wind turbines, and photovoltaic array are connected to the AC bus via different converters. User loads consist of time-shiftable loads and rigid loads. In the hydrogen energy storage system, when wind and solar power generation exceeds load demand, the electrolyzer uses the surplus electricity to electrolyze water, storing the generated hydrogen in the hydrogen storage tank. When wind and solar power generation is insufficient to meet load demand, the fuel cells use the hydrogen in the storage tank as fuel to generate electricity to meet load demand. Distributed power generation parameters and other parameters in the microgrid are shown in Tables 1 and 2.

[0115] Table 1 Distributed Power Generation Parameters

[0116]

[0117] Table 2 Other parameters within the microgrid

[0118]

[0119] Figure 2 , Figure 3 and Figure 4 To generate curves for typical daytime wind speed, solar irradiance, and load, the annual meteorological and load data are divided into three categories: 60 days in summer, 120 days in winter, and 185 days in spring and autumn. Each category can be represented by a typical day. The K-means clustering algorithm is used to obtain the wind speed, solar irradiance, and load curves for typical days in summer, winter, and spring and autumn to replace the wind speed, solar irradiance, and load data for each day in that season.

[0120] 2) Solve the two-layer robust optimization programming model for the microgrid.

[0121] Figure 5 The flowchart for solving the two-layer robust optimization planning of independent microgrids is as follows: the outer layer uses a genetic algorithm with an elite retention strategy to solve the problem, while the inner layer calls the solver Cplex to solve the problem. The capacity configuration calculated by the outer layer is substituted into the inner layer, and the inner layer calculates the optimal operating result based on the known capacity configuration and then substitutes it into the outer layer. The inner and outer layers iterate to obtain the optimal planning scheme.

[0122] Five schemes were set up for comparative analysis, and the schemes are set as follows:

[0123] Option 1: Ignoring uncertainties in wind and solar power output and incentive-based DR;

[0124] Option 2: Considering the uncertainty of wind and solar power output, stimulating DR is not considered;

[0125] Option 3 is an application of the capacity configuration results obtained from Option 2 in a scenario with uncertain wind and solar power output.

[0126] Option 4: Based on the capacity configuration of Option 2, consider the calculation of the optimal operating cost using an incentive-based DR.

[0127] Option 5: Simultaneously consider the uncertainty of wind and solar power output and the optimal capacity configuration and operating costs calculated by incentive-based DR;

[0128] The calculation results for the five schemes are shown in Table 3.

[0129] Table 3. Capacity configuration and cost calculation results for different schemes.

[0130]

[0131] a. The impact of wind and solar uncertainties on capacity allocation and costs

[0132] Comparing Schemes 1 and 2, it can be found that the maximum pessimistic cost of Scheme 2, considering uncertainty, is 294,800 yuan higher than that of Scheme 1. This is because in the deterministic scenario of Scheme 1, the microgrid is planned and invested based on wind and solar forecasts, meaning that both the risk aversion coefficient and the uncertainty parameter are 0. After introducing IGDT to simulate wind and solar uncertainty in Scheme 2, the calculated uncertainty parameter α is 9.43%, which can be interpreted as the maximum range of wind and solar fluctuations that the generated planning scheme can withstand within the planning period being 9.43%, i.e., the system's robustness level to wind and solar uncertainty is 9.43%. In this case, the microgrid decision-maker must bear an investment budget with a risk aversion coefficient of 10%.

[0133] Based on the calculation results of Scheme 2, Scheme 3 takes α as 4%. As shown in Table 3, under the same capacity configuration, on the one hand, the total cost Z of Scheme 3 is 132,600 yuan higher than the expected target cost Z of Scheme 2 in the deterministic scenario. This is because in an uncertain environment, the output of wind and solar power will develop in an unfavorable direction. In order to meet the power deficit of the microgrid, the output of diesel generators must be increased, resulting in a significant increase in the operating cost of the microgrid. On the other hand, the maximum pessimistic cost Z' of Scheme 3 is 228,700 yuan less than that of Scheme 2. This is because when wind and solar power fluctuate arbitrarily within a certain range, the obtained robust solution can meet the maximum pessimistic cost acceptable to the decision-maker, indicating that the decision solution has good adaptability.

[0134] Figure 6 This is a schematic diagram illustrating the operation results of the microgrid under a typical summer day deterministic scenario, as shown in Scheme 2. Figure 7 This is a schematic diagram illustrating the operation results of the microgrid under a typical summer day with uncertainty. Figure 6 and Figure 7 It can be observed that under typical summer days, in order to cope with the uncertainty of wind and solar power output, Scheme 3 requires one more period of diesel generator operation than Scheme 2, which leads to increased operating costs.

[0135] b. The impact of incentive-based DR on capacity allocation and cost

[0136] As shown in Table 3, compared to Scheme 2, Scheme 4, after introducing the incentive-based DR, reduces the proportion of diesel generator power generation by 2.34%, reduces the annual operating cost of the microgrid by RMB 198,300, and reduces the total cost by RMB 198,300. Compared to Scheme 4, Scheme 5 increases the installed capacity of wind turbines and photovoltaics by 100kW and 135kW respectively; simultaneously, the capacities of the electrolyzer, fuel cell, and hydrogen storage tank decrease by 167kW, 137kW, and 871kW·h respectively. In terms of operation, the proportion of diesel generator power generation decreases by 1.12%, diesel costs decrease by RMB 186,700, annual investment costs decrease by RMB 47,000, and annual operating costs decrease by RMB 176,200, a reduction of 4.67%, resulting in a total cost reduction of RMB 223,200. This is because the introduction of the incentive-based DR adjusts the working time of the time-shiftable load, making the load curve more consistent with the wind and solar power output curves, improving the utilization rate of renewable energy, and reducing the output of the hydrogen storage system and diesel generators needed to smooth out fluctuations in wind and solar power output.

[0137] Figure 8 This is a schematic diagram illustrating the operation results of a typical microgrid during winter after introducing an excitation-type DR in Scheme 4. Figure 9 The diagram shows the operation results of the microgrid on a typical winter day after the introduction of incentive-based DR in Scheme 5. In the operation results of Scheme 5 considering incentive-based DR on a typical winter day, the number of times the diesel generator set is activated is reduced by 1 compared to Scheme 4, thereby reducing the operating cost of the microgrid. This shows that the planning scheme obtained by microgrid decision-makers when incentive-based DR is taken into account during the planning stage is better than the planning scheme obtained by implementing incentive-based DR after the system is built.

[0138] Figure 10 The iterative convergence curves of Scheme 5 under the genetic algorithm with the elite retention strategy and the traditional genetic algorithm show that the genetic algorithm with the elite retention strategy converges faster, can suppress premature convergence, and avoids the algorithm getting trapped in local optima.

[0139] c. Sensitivity analysis of the risk aversion coefficient μ

[0140] Taking Option 5 as an example, the range of μ variation is 2% to 20%. Conservative decision-makers believe that since uncertainty may cause the goal to develop in an unfavorable direction, they hope to resist the possible uncertainty by increasing operating costs in order to minimize the adverse effects. Figure 11The diagram illustrates the impact of different μ values ​​on α and the annual operating cost of the microgrid. When the annual operating cost of the microgrid increases from RMB 3.3378 million to RMB 3.9269 million, the α that the microgrid can accommodate increases from 2.16% to 18.01%. Correspondingly, when wind and solar power output fluctuates within a certain range, it can be ensured that the system's operating cost is lower than the maximum pessimistic operating cost.

[0141] d. Validity analysis of the IGDT robust model

[0142] Figure 12 This diagram illustrates the annual operating cost of a microgrid under uncertain scenarios. Monte Carlo simulation was used to extract 1000 random wind and solar power output scenarios within the uncertainty set, with μ set to 20%, corresponding to a maximum pessimistic cost of 3.9269 million yuan. Figure 12 Within the fluctuation range, the operating cost of the microgrid can be between RMB 3.3378 million and RMB 3.9269 million for any wind and solar power output scenario. The maximum α that the microgrid can withstand at this time is 18.01%, which verifies the effectiveness of the model.

[0143] The system energy storage solution adopts lithium-ion batteries, which are currently the most widely used type, without considering wind and solar uncertainties or incentive-driven DR. The capacity configuration and cost calculation results are shown in Table 4.

[0144] Table 4. Capacity configuration and cost calculation results for lithium-ion battery energy storage.

[0145]

[0146] Comparing Scheme 1 in Table 3, it can be seen that although the required energy storage capacity using hydrogen energy storage is 2220 kWh higher than that using lithium-ion battery energy storage, the total cost of the hydrogen energy storage scheme is still 725,100 yuan lower than that of the lithium-ion battery energy storage scheme. This is because hydrogen energy storage has the characteristics of high energy density and large capacity, its unit capacity cost is much lower than that of lithium-ion batteries, and hydrogen can be stored for a long time under appropriate conditions without energy loss. The lifespan of hydrogen energy storage is also longer than that of lithium-ion battery energy storage, and its contribution to long-term regulation capability is more significant.

[0147] The specific embodiments used in this invention have provided a detailed description of the invention, but are not limited to these embodiments. Any obvious modifications made by those skilled in the art based on the teachings of this invention are within the scope of protection of this invention.

Claims

1. A robust planning method for an independent microgrid IGDT containing hydrogen energy storage, characterized by: The steps are as follows, and the steps are performed in sequence. Step 1: Establish an independent microgrid model of the hydrogen energy storage system, which includes an electrolyzer model, a fuel cell model, and a hydrogen storage tank model. Step 2: Based on the IGDT robust optimization method, establish an independent microgrid operation scenario model, and use the entropy weight method to evaluate the weights of the uncertain parameters wind power and photovoltaic output; Step 3: By considering the outer and inner optimization models, a two-layer robust optimization planning model for independent microgrids that takes into account IGDT is established. A genetic algorithm with an elite retention strategy is used to solve the two-layer robust optimization planning model to obtain a planning scheme for independent microgrids with hydrogen energy storage. The entropy weight method is used to calculate the weight values ​​of wind power output and photovoltaic power output, and the importance of their uncertainty factors is evaluated, as shown in equation (13). （13） In the formula, For the first i The first indicator in the t The standardized value among the alternative options; V i,t For the first t The first alternative option i The proportion of each evaluation indicator; H i For the first i The entropy value of each indicator; w i For the first i Entropy weight of each indicator; Each uncertain variable is regarded as an indicator of system uncertainty. The corresponding entropy weight is calculated using the predicted values ​​of wind and solar power. Based on the entropy weight, the proportion of the corresponding uncertain variable in the total uncertainty is calculated, as shown in equation (14). （14） In the formula, The proportion of uncertainty in wind turbine output; The proportion of uncertainty in the output of the photovoltaic array; The method for constructing the independent microgrid two-layer robust optimization planning model based on IGDT in step three is as follows: ① Outer layer optimization model The outer layer aims to minimize the annual comprehensive cost Z of the microgrid. The set of decision variables x for outer layer optimization represents the rated capacity of fuel cells, electrolyzers, hydrogen storage tanks, photovoltaic arrays, and wind turbines. In a deterministic scenario, the expected target cost function Z of the outer layer is: （15） In the formula, Z inv Annual value of equipment investment cost; Z ope The annual operating cost of the system; When the wind and solar power output develops in an unfavorable direction, after the introduction of IGDT, the maximum pessimistic operating cost and total cost accepted by the microgrid decision-maker can be rewritten as equations (16) and (17). （16） （17） In the formula, The maximum pessimistic operating cost acceptable to decision-makers; This represents the annual value of the comprehensive cost of a microgrid. Z inv The calculation formulas include the average annual investment cost and annual maintenance cost of each piece of equipment, as shown in equations (18) and (19). （18） （19） In the formula, k For equipment type, k =1, 2, ..., 5 represent wind turbine, photovoltaic array, electrolyzer, fuel cell and hydrogen storage tank respectively; R For annual value operators; L The system's operational lifespan; Z f,k No. k Initial investment cost per unit capacity of this type of equipment; Z m,k Annual maintenance cost of the equipment; ε k The number or capacity of the equipment installed; λ The discount rate; Due to the influence of the site construction area, the inequality constraints on the investment capacity of various equipment are shown in equation (20). （20） In the formula, C de,N , C wt,N and C pv,N This refers to the rated installed capacity of diesel generator sets, wind turbine generator sets, and photovoltaic arrays. and These represent the minimum and maximum installed capacity of the photovoltaic array, respectively. and These are the minimum and maximum values ​​of the wind turbine assembly capacity, respectively. and These are the minimum and maximum values ​​of the diesel engine assembly capacity, respectively. and These are the minimum and maximum values ​​of the electrolytic cell capacity, respectively. and These represent the minimum and maximum values ​​of the fuel cell capacity, respectively. and These are the minimum and maximum values ​​of the hydrogen storage tank capacity, respectively. ② Inner layer optimization model The annual operating cost of the inner layer includes fuel cost and DR compensation cost. The decision variables of the inner layer are the charging and discharging power of wind turbines, photovoltaic arrays, diesel generator sets and hydrogen energy storage systems in each period of the year. The inner layer optimization model is shown in Equation (21). （21） In the formula, Z dr This is the annual load transfer compensation cost; Z fuel The annual fuel cost for the diesel generator set; The wind speed, solar irradiance, and load curves of typical days in summer, winter, spring, and autumn are obtained using the K-means clustering algorithm to replace the wind speed, solar irradiance, and load data for each day of the season, as shown in Equation (22). （22） In the formula, S Let s be the number of typical days in a year, where s = 1, 2, ..., S; n s This represents the number of days in a typical year, specifically the s-th day. Z dr,s and Z fuel,s This is expressed as the load transfer compensation cost and diesel generator set fuel cost on the s-th typical day; The fuel cost of diesel generator sets is shown in equations (23) and (24). （23） （24） In the formula, β This refers to the unit price of diesel fuel; Ω s It is a collection of three typical days in spring, autumn, summer, and winter; ω s,t This is a binary variable; a value of 0 indicates the s-th typical day. t The diesel generator set is shut down during a specific period, with 1 hour indicating operation. C s,t This refers to the fuel consumption of the diesel generator set. P de,s,t For diesel generator sets in s typical days t The effort required at that time; and , representing the maximum and minimum output of the diesel generator set, respectively; b represents the slope of the fuel consumption-power curve; c represents the intercept coefficient. Incentive-based DR is introduced, with the minimum sum of the absolute values ​​of the differences between renewable energy generation and load demand at each time point as the optimization objective of incentive-based DR. Incentive-based DR is shown in Equation (25). （25） In the formula, P pv,s,t , P wt,s,t Each is the s-th typical day t The output power of photovoltaic arrays and wind turbines at all times; P L0,s,t , P L1,s,t These are the pre- and post-activation DR systems, respectively. t Load power at any given time; The s-th typical day t The load transfer amount at any given time is either greater than 0 (load transferred in) or less than 0 (load transferred out). After considering DR, the main cost is the compensation cost that needs to be given to users to adjust the time-shiftable load, as shown in equation (26). （26） In the formula, C dr The cost of transferring and compensating for unit electricity load; ③ Establish a two-layer robust optimization planning model for independent microgrids that takes into account IGDT, as shown in equation (33). （33） The inner optimization model satisfies the following constraints: a. Power balance constraints （27） In the formula, and The first and second are respectively the electrolyzer and the fuel cell. s A typical day t Power at any given moment; b. Output constraints of distributed power sources （28） In the formula, , and These are the rated outputs of the photovoltaic array, wind turbine, and diesel generator, respectively. c. Constraints of hydrogen energy storage systems （29） In the formula, For typical days within s t Energy storage capacity of the time-limited hydrogen storage tank; and These are the upper and lower limits of the energy storage capacity of hydrogen storage tanks; and For typical days within s t Maximum power of fuel cells and electrolyzers during specific time periods; and These are the upper and lower limits of the state of charge of the hydrogen storage tank; S SOC,tank,s,t For typical days within s t State of charge of the hydrogen storage tank during a given period; and These represent the state of charge of the hydrogen storage tank at the beginning and end of a single operating cycle. t 0 and t N These represent the beginning and end periods of the operating cycle; d. Constraints on the proportion of electricity generated by diesel generator sets （30） e. Load transfer constraints （31） （32） In the formula, For the s-th typical day t Maximum load transfer amount at any given time; t 0 This is the period when the load is transferred in; t 1 This is the period for load transfer out; T n This is the nth running cycle.

2. The robust planning method for an independent microgrid IGDT containing hydrogen energy storage according to claim 1, characterized in that: The independent microgrid structure in step one includes a photovoltaic array, a diesel generator set, a wind turbine generator set, loads, and a hydrogen energy storage system consisting of a hydrogen storage tank, a fuel cell, and an electrolyzer. The hydrogen energy storage system, the wind turbine generator set, and the photovoltaic array are connected to the AC bus through different converters. The user load consists of time-shiftable loads and rigid loads. In hydrogen energy storage systems, when wind and solar power generation exceeds load demand, the surplus electricity is used in an electrolyzer to electrolyze water, and the generated hydrogen is stored in a hydrogen storage tank. When wind and solar power generation is insufficient to meet load demand, a fuel cell uses the hydrogen in the storage tank as fuel to generate electricity to meet load demand; ① Electrolyzer model An alkaline electrolyzer is used to electrolyze water to produce hydrogen and oxygen. The output power is shown in equation (1). （1） In the formula, P el,in This refers to the input power of the electrolytic cell; P el,out This refers to the output power of the electrolytic cell; η el The efficiency of the electrolytic cell; The maximum input power of the electrolyzer is limited by its capacity and the remaining energy storage capacity of the hydrogen storage tank. The maximum input power of the electrolyzer is shown in equations (2) and (3). （2） （3） In the formula, This is the maximum input power of the electrolytic cell; C el,N This refers to the rated capacity of the electrolytic cell; This represents the maximum energy storage capacity of the hydrogen storage tank. C tank,N This refers to the rated capacity of the hydrogen storage tank. E tank,t for t The energy stored in the hydrogen storage tank at all times; The time interval is used; analogous to the state of charge of a battery, the state of charge of a hydrogen storage tank is defined as follows: S SOC,tank , This represents the maximum state of charge of the hydrogen storage tank. ② Fuel Cell Model Fuel cells release energy by burning hydrogen and oxygen, and their output power is shown in equation (4). （4） In the formula, P fc,out This refers to the output power of the fuel cell; P fc,in This refers to the input power of the fuel cell; η fc To improve the working efficiency of fuel cells; The maximum output power of a fuel cell is also limited by its capacity and the remaining capacity of the hydrogen storage tank, as shown in equations (5) and (6). （5） （6） In the formula, This represents the maximum output power of the fuel cell; This represents the minimum state of charge of the hydrogen storage tank. This represents the minimum energy storage capacity of the hydrogen storage tank. C fc,N This refers to the rated capacity of the fuel cell; ③ Hydrogen storage tank model The hydrogen storage tank is used to store hydrogen produced by the electrolyzer and to supply hydrogen to the fuel cell, as shown in equation (7). （7） In the formula, η tank To improve the working efficiency of the hydrogen storage tank.

3. The robust planning method for an independent microgrid IGDT containing hydrogen energy storage according to claim 1, characterized in that: The independent microgrid operation scenario model in step two is as follows: （8） In the formula, α The fluctuation range of the uncertain parameter; y For decision variables; Y These are uncertain parameters; These are the predicted values ​​of uncertain parameters; Uncertain parameters Y The fluctuation range does not exceed ; μ The risk aversion coefficient represents the percentage increase in costs caused by uncertainty that microgrid decision-makers can accept. The larger the aversion coefficient, the larger the budget that the decision-maker can afford for the planning scheme, and the stronger the planning model's ability to cope with uncertainty. H ( Y,y )and G ( Y,y These are equality constraints and inequality constraints, respectively. F For deterministic models Y Pick The target cost value at that time; For pessimistic costs; (1+ μ ) F This represents the maximum pessimistic cost target value that decision-makers can accept. The uncertain parameters are the output of both wind power and photovoltaic power, and their fluctuation range is shown in equations (9) and (10). （9） （10） In the formula, and These are the predicted output values ​​for wind turbines and photovoltaic arrays, respectively. and These are the actual output values ​​of the wind turbine and the photovoltaic array, respectively. α wt and α pv These represent the fluctuation ranges in the output of wind turbines and photovoltaic arrays, respectively, and their values ​​are directly related to the prediction error. The set of fluctuation ranges in wind turbine output; This represents the set of fluctuation ranges for photovoltaic arrays. The system operating cost is maximized when the actual output of wind and solar power is much lower than the predicted value, and its value is shown in equations (11) and (12). （11） （12）。

Images