Electric-thermal-hydrogen multi-microgrid system optimization scheduling method suitable for considering uncertainty

Abstract

The invention discloses an electricity-heat-hydrogen multi-microgrid system optimization scheduling method suitable for considering uncertainty. According to the method, firstly, an electricity-heat-hydrogen multi-microgrid system architecture is constructed, and the system is divided into an industrial area and a residential area which have independent energy production and load requirements; an equipment model in the system is established, and uncertainty modeling is carried out on solar irradiance and wind speed by adopting beta distribution and Weibull distribution according to photovoltaic and fan output fluctuation. On the basis of the model, a cost objective function of the system is constructed, and the cost objective function covers productivity consumption, maintenance, faults and energy procurement cost. Finally, a robust optimization algorithm based on an alternating direction multiplier method is adopted, the global optimization problem is decomposed into local optimization problems of all the subnets, supply and demand balance is achieved through information exchange, and the adaptive output power of all the devices is obtained. According to the method, the influence of renewable energy source volatility on equipment output is fully considered, and the energy utilization efficiency is improved and the system operation cost is reduced by cooperatively scheduling power, heat and hydrogen resources.

Description

Technical Field

[0001] The present invention belongs to the technical field of hydrogen-containing integrated energy, and specifically relates to an optimization scheduling method for an electric-heat-hydrogen multi-microgrid system taking uncertainty into consideration. Background Art

[0002] With the increasingly diversified development of energy systems, traditional electric-thermal systems have many limitations in responding to flexible energy supply needs. Introducing hydrogen energy into traditional electric-thermal systems can not only meet the growing demand for hydrogen energy, but also achieve the coordinated use of multiple energy sources. Microgrid refers to an energy system composed of distributed power sources (photovoltaic power generation, wind turbine power generation, fuel cells, etc.), energy storage devices, loads and control devices, which can realize the supply of multiple energy forms to loads. Compared with a single microgrid system, multi-microgrids can conduct energy transactions between adjacent subgrids, which enhances the stability of the system and the flexibility of energy scheduling.

[0003] In the electric-heat-hydrogen multi-microgrid system, reasonable energy scheduling becomes the key to balancing supply and demand and reducing operating costs. However, there are still many challenges in the scheduling process, especially the volatility of renewable energy and the complex coupling relationship between multiple microgrids, including the dependence of power supply on unstable resources such as light radiation and wind, the dependence of hydrogen production on electricity, and the mutual conversion of electricity, heat and hydrogen in the system. These factors significantly affect the reliability and response speed of the scheduling process. In this context, optimizing the scheduling method of the electric-heat-hydrogen multi-microgrid system is of great significance to improving the economic benefits of the system and achieving stable operation. Summary of the invention

[0004] The purpose of the present invention is to address the above-mentioned problems. This study considers the uncertainty of renewable energy and proposes a method suitable for optimizing the scheduling of an electric-thermal-hydrogen multi-microgrid system. The specific steps are as follows:

[0005] Step 1: Build the electricity-heat-hydrogen multi-microgrid system architecture and establish the electricity, heat and hydrogen output models of each component of the system;

[0006] Step 2: Analyze the uncertainty of renewable energy and establish the uncertainty model of renewable energy;

[0007] Step 3: Analyze the cost composition of the capacity side in the energy system and establish a mathematical model of its capacity cost;

[0008] Step 4: Decompose the global optimization problem and design the alternating direction multiplier method (ADMM) robust optimization scheduling method.

[0009] Furthermore, for the construction of the electric-heat-hydrogen multi-microgrid system architecture in step 1, it is divided into industrial areas and residential areas, both of which have independent energy production and energy load capabilities, which can be purchased through the power grid (PG) and the hydrogen production plant (HPP). At the same time, electricity and hydrogen can be traded between subgrids to meet their respective needs. Depending on different geographical conditions and load requirements, the specific equipment and capacity of industrial areas and residential areas vary.

[0010] The equipment in the electric-heat-hydrogen multi-microgrid system is modeled. The specific model is as follows:

[0011] The photovoltaic power output model is expressed as:

[0012]

[0013] Where P pv,t is the output power of photovoltaic at time t (kw); P r,pv is the rated output power of photovoltaic (kw); G t ,G STC are solar irradiance (W / m2) and standard solar irradiance (1000W / m2); β is the temperature coefficient (0.00485 / K); T a,t is the air temperature (K); T NOCT is the standard operating temperature of photovoltaic cells (318.15K); T m is the test temperature (K).

[0014] The power output model of the fan is expressed as:

[0015]

[0016] Where P wt,t is the output power of the fan at time t (kw); P r,wt is the rated output power of the fan (kw); v c ,v r ,v f They are the cut-in wind speed, rated wind speed and cut-off wind speed of the fan (m / s) respectively.

[0017] The power output model of the electrolyzer is expressed as:

[0018]

[0019] Where W el,t is the mass of hydrogen output by the electrolyzer at time t (kg); P el,t is the electric power input to the electrolytic cell at time t (kw); η el is the energy conversion efficiency of the electrolyzer; κh is the conversion factor between hydrogen mass and energy, which is 39.54 kWh / kg.

[0020] The power output model of the fuel cell is expressed as:

[0021]

[0022] Where P hfc,t is the output power of the fuel cell at time t (kw); H hfc,t is the output thermal power of the fuel cell at time t (kW); W hfc,t is the hydrogen input mass of the fuel cell at time t (kg); η hfce is the power conversion efficiency of the fuel cell; η hfch is the thermal energy conversion efficiency of the fuel cell.

[0023] The power output model of the electric boiler is expressed as:

[0024] Q eb,t =P eb,t η eb

[0025] In the formula, Q eb,t is the electric power input to the electric boiler at time t (kw); P eb,t is the thermal power output of the electric boiler at time t; η eb is the thermoelectric conversion efficiency.

[0026] The model of the energy storage device is expressed as:

[0027]

[0028] Where, k = 1, 2, 3, respectively, are the energy storage devices for electricity, heat, and hydrogen; E k,t ,E k,t-1 are the energies stored in the energy storage device at time t and t-1 respectively; σ self,k is the self-discharge rate of energy storage device k; P ch,k,t and P dis,k,t are the charging and discharging powers of energy storage device k at time t; η cha,k and η dis,k are the charging and discharging efficiencies of the energy storage device k, respectively.

[0029] Furthermore, in step 2, considering the volatility of renewable energy, solar irradiance and wind speed have a direct impact on the output power of photovoltaic and wind turbines respectively, and uncertainty modeling is performed on solar irradiance and wind speed. Beta distribution is used to model the uncertainty of solar irradiance, and the model is as follows:

[0030]

[0031]

[0032] Where f(G) is the distribution of solar irradiance; Γ(x) is the gamma function; α, β are parameters in the distribution; μ, σ are the mean and standard deviation of solar irradiance, respectively.

[0033] The uncertainty of wind speed is modeled using Weibull distribution, as follows:

[0034]

[0035]

[0036] In the formula, v c is the wind speed (m / s); λ, k are the parameters of Weibull distribution; are the mean and standard deviation of wind speed, respectively.

[0037] The output fluctuations of photovoltaic and wind turbines in industrial and residential areas are analyzed based on the distribution of solar irradiance and wind speed, so as to more accurately describe the impact of uncertainty on system operating costs.

[0038] Furthermore, in step 3, the power constraints of the devices in the system and the supply and demand balance constraints of the system are considered. The power constraints of non-energy storage devices in the system are:

[0039]

[0040] Where P pv,min ,P pv,max are the minimum and maximum power of photovoltaic power generation; P wt,min ,P wt,max are the minimum and maximum power generated by the wind turbine respectively; P hfc,min ,P hfc,max are the minimum and maximum power generated by the fuel cell; P el,min ,P el,max are the minimum and maximum power of hydrogen production by the electrolyzer; Q hfc,min ,Q hfc,max are the minimum and maximum power of fuel cell heat generation; Q eb,min ,Q eb,max They are the minimum and maximum power of electric boiler heat generation respectively.

[0041] The power constraint of the energy storage device in the system is:

[0042]

[0043] In the formula, The upper and lower limits of the capacity of energy storage device k respectively; X k ,Yk are the state variables of charging and discharging of energy storage device k respectively.

[0044] The supply and demand balance constraint of the system is:

[0045]

[0046] Where M ind ,M res It is a collection of equipment for industrial and residential areas; j is the energy production of device j in the corresponding subnet; p buy,m,ind ,P buy,m,res P is the energy purchased from the power grid and hydrogen supply plant by industrial and residential areas respectively; load,ind ,P load,res They are the load demands of industrial areas and residential areas respectively.

[0047] The supply and demand balance constraints within industrial areas and residential areas are:

[0048]

[0049] Where P pv,t ,P wt,t ,P hfc,t ,P el,t ,P eb,t are the energy production of photovoltaic, wind turbine, fuel cell, electrolyzer and electric boiler at time t; P buy,pg,t ,P buy,hpp,t ,P buy,res,t ,P buy,ind,t are the energy purchased from the power grid, hydrogen supply plant, residential area and industrial area respectively; P es,ind ,P es,res are the energy storage in industrial and residential areas respectively; P load,ind ,P load,res They are the load demands of industrial areas and residential areas respectively.

[0050] Establish a relationship model between the power and cost of equipment in the system, as follows:

[0051]

[0052] In the formula, C total is the total system cost; n is the number of subnets; j∈M i Indicates device j belonging to subnet i; are the capacity consumption cost, regular maintenance cost, failure cost, and energy purchase cost of device j in subnet i; p j is the output power of equipment j at time t; a j ,b j ,c jis the capacity cost coefficient of equipment j; ω j is the maintenance cost factor of equipment j; is the average assembly cost of one equipment replacement; T d is the downtime; c j is the output energy price of equipment j; l is other regional subgrids, power grids and hydrogen supply plants; c i,l,t is the price of energy purchased by region i from region l at time t; E i,k,t is the amount of energy purchased by region i from region l at time t.

[0053] Furthermore, in step 4, a robust optimization algorithm is designed to solve the cost optimization model and obtain the optimal solution. In view of the complex problems of multi-microgrid systems, the global optimization problem is decomposed into local optimization problems in each microgrid area. The ADMM algorithm is used to independently optimize the cost model of each microgrid. At the same time, the global supply and demand balance is coordinated through information exchange. The overall objective function is expressed as:

[0054] L global =C total -λ∈ total

[0055] Where, L global is the global optimization objective; λ is the Lagrange multiplier; ∈ total is the supply and demand constraint deviation of the overall system.

[0056] The steps to solve the overall objective function are as follows:

[0057] Step 1: Decompose the global optimization problem into multiple local sub-problems, and independently optimize the local variables in each sub-problem. The local objective function of each sub-network is as follows:

[0058]

[0059] In the formula, C total,i is the cost target of regional subnet i; is the Lagrange multiplier; ρ is the step size parameter; ∈ cons,i is the power and storage constraints in subnet i; ∈ total ,∈ i They represent the deviation of supply and demand constraints of the whole network and the corresponding subnetwork respectively; It is the auxiliary variable of ADMM.

[0060] Step 2: Optimize the variable P in each subproblem i , find the local,i Minimize P i :

[0061] (1) Find partial derivatives: Calculate the local objective function with respect to P i The partial derivative of is calculated as:

[0062]

[0063] (2) Update local variables: Use the gradient descent method to adjust the variables so that the local objective function converges to the minimum value. The calculation formula is:

[0064]

[0065] In the formula, α is the learning rate and k is the number of iterations.

[0066] Step 3: Update the Lagrange multipliers λ and The calculation formulas are:

[0067] λ (k+1) =λ (k) +ρ∈ total

[0068]

[0069] Step 4: Introducing auxiliary variables through ADMM Update auxiliary variables, the calculation formula is:

[0070]

[0071] Step 5: Determine whether the convergence condition is met. If it is met, stop the iteration and output the optimal solution. The convergence condition is as follows:

[0072] ||∈ total ||≤∈ threshold

[0073] In the formula, ∈ total is the deviation of load supply and demand; ∈ threshold is the preset threshold.

[0074] If the convergence condition is not met, return to step 2 and step 3 to continue optimization and updating until the convergence condition is met.

[0075] The present invention has the following beneficial effects:

[0076] (1) According to the geographical conditions and load requirements of different regions, the architecture of the electricity-heat-hydrogen multi-microgrid system was constructed;

[0077] (2) In view of the volatility of photovoltaic and wind turbine output, the present invention uses Beta distribution and Weibull distribution to model the uncertainty of solar irradiance and wind speed, so as to more accurately describe the impact of uncertainty on system operating costs;

[0078] (3) Based on the robust optimization algorithm of ADMM, the present invention decomposes the global optimization problem into local optimization problems of each subnet, realizes the supply and demand balance through information exchange, and obtains the adaptive output power of each device, thereby effectively improving the energy utilization efficiency and reducing the system operation cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 is a schematic diagram of the structure of an electric-heat-hydrogen multi-microgrid system in an example of the present invention;

[0080] Figure 2 is a flow chart of equipment modeling in an example of the present invention;

[0081] Figure 3 It is a flowchart of the ADMM robust algorithm solution in the example of the present invention. DETAILED DESCRIPTION

[0082] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is further elaborated in detail below in combination with the implementation modes and the accompanying drawings.

[0083] The optimization scheduling method provided by the present invention has the following specific steps:

[0084] Step 1: Build the electricity-heat-hydrogen multi-microgrid system architecture and establish the electricity, heat and hydrogen output models of each component of the system;

[0085] Step 2: Analyze the uncertainty of renewable energy and establish the uncertainty model of renewable energy;

[0086] Step 3: Analyze the cost composition of the capacity side in the energy system and establish a mathematical model of its capacity cost;

[0087] Step 4: Decompose the global optimization problem and design the ADMM robust optimization scheduling method.

[0088] Furthermore, for the electricity-heat-hydrogen multi-microgrid system architecture built in step 1, Figure 1 As shown in the figure, according to different geographical conditions and load requirements, the two subgrids are represented as industrial areas and residential areas, both of which have independent energy production and energy load capabilities, which can be purchased through the power grid (PG) and the hydrogen production plant (HPP). At the same time, the subgrids can trade electricity and hydrogen to meet their respective needs. The equipment in the industrial area includes photovoltaics, wind turbines, fuel cells, electrolyzers and energy storage equipment, and the equipment in the residential area includes photovoltaics, electrolyzers, electric boilers and energy storage equipment.

[0089] Furthermore, the output model of the equipment in different areas and the uncertainty model of renewable energy are established, and on this basis, the cost composition of industrial and residential areas is analyzed, such as Figure 2 As shown, the details are as follows:

[0090] The output model of photovoltaic is expressed as:

[0091]

[0092] Where P pv,t is the output power of photovoltaic at time t (kw); P r,pv is the rated output power of photovoltaic (kw); G t ,G STC are solar irradiance (W / m2) and standard solar irradiance (1000W / m2); β is the temperature coefficient (0.00485 / K); T a,t is the air temperature (K); T NOCT is the standard operating temperature of photovoltaic cells (318.15K), T m is the test temperature (K).

[0093] The output model of the fan is expressed as:

[0094]

[0095] Where P wt,t is the output power of the fan at time t (kw); P r,wt is the rated output power of the fan (kw); v c ,v r ,v f They are the cut-in wind speed, rated wind speed and cut-off wind speed of the fan (m / s) respectively.

[0096] The power output model of the electrolyzer is expressed as:

[0097]

[0098] Where W el,t is the mass of hydrogen output by the electrolyzer at time t (kg); P el,t is the electric power input to the electrolytic cell at time t (kw); η el is the energy conversion efficiency of the electrolyzer; κ h is the conversion factor between hydrogen mass and energy, which is 39.54 kWh / kg.

[0099] The power output model of the fuel cell is expressed as:

[0100]

[0101] Where P hfc,tis the output power of the fuel cell at time t (kw); H hfc,t is the output thermal power of the fuel cell at time t (kW); W hfc,t is the hydrogen input mass of the fuel cell at time t (kg); η hfce is the power conversion efficiency of the fuel cell; η hfch is the thermal energy conversion efficiency of the fuel cell.

[0102] The power output model of the electric boiler is expressed as:

[0103] Q eb,t =P eb,t η eb

[0104] In the formula, Q eb,t is the electric power input to the electric boiler at time t (kw), P eb,t is the thermal power output of the electric boiler at time t, η eb is the thermoelectric conversion efficiency.

[0105] The model of the energy storage device is expressed as:

[0106]

[0107] Where, k = 1, 2, 3, respectively, are the energy storage devices for electricity, heat, and hydrogen; E k,t ,E k,t-1 are the energies stored in the energy storage device at time t and t1 respectively; σ self,k is the self-discharge rate of energy storage device k; P ch,k,t and P dis,k,t are the charging and discharging powers of energy storage device k at time t; η cha,k and η dis,k are the charging and discharging efficiencies of the energy storage device k, respectively.

[0108] Furthermore, in step 2, the volatility of renewable energy is considered. Solar irradiance and wind speed have a direct impact on the output power of photovoltaic and wind turbines. In order to improve the accuracy of scheduling optimization and meet the needs of users in a timely manner, the distribution of solar irradiance and wind speed is modeled. The model is as follows:

[0109] Considering the volatility of renewable energy, solar irradiance and wind speed have a direct impact on the output power of photovoltaic and wind turbines respectively, and uncertainty modeling of solar irradiance and wind speed is performed.

[0110] The uncertainty of solar irradiance is modeled using the Beta distribution as follows:

[0111]

[0112] Where f(G) is the distribution of solar irradiance; Γ(x) is the gamma function; α, β are parameters in the distribution; μ, σ are the mean and standard deviation of solar irradiance, respectively.

[0113] The uncertainty of wind speed is modeled using Weibull distribution, as follows:

[0114]

[0115] In the formula, v c is the wind speed (m / s); λ, k are the parameters of Weibull distribution; are the mean and standard deviation of wind speed respectively.

[0116] Furthermore, in step 3, the power constraints of the devices in the system and the supply and demand balance constraints of the system are considered. The power constraints of non-energy storage devices in the system are:

[0117]

[0118] Where P pv,min ,P pv,max are the minimum and maximum power of photovoltaic power generation; P wt,min ,P wt,max are the minimum and maximum power generated by the wind turbine respectively; P hfc,min ,P hfc,max are the minimum and maximum power generated by the fuel cell; P el,min ,P el,max are the minimum and maximum power of hydrogen production by the electrolyzer; Q hfc,min ,Q hfc,max are the minimum and maximum power of fuel cell heat generation; Q eb,min ,Q eb,max They are the minimum and maximum power of electric boiler heat generation respectively.

[0119] The power constraint of the energy storage device in the system is:

[0120]

[0121] In the formula, The upper and lower limits of the capacity of energy storage device k respectively; X k ,Y k are the state variables of charging and discharging of energy storage device k respectively.

[0122] The supply and demand balance constraint of the system is:

[0123]

[0124] Where M ind ,M res It is a collection of equipment for industrial and residential areas;j is the energy production of device j in the corresponding subnet; p buy,m,ind ,P buy,m,res P is the energy purchased from the power grid and hydrogen supply plant by industrial and residential areas respectively; load,ind ,P load,res They are the load demands of industrial areas and residential areas respectively.

[0125] The supply and demand balance constraints within industrial areas and residential areas are:

[0126]

[0127] Where P pv,t ,P wt,t ,P hfc,t ,P el,t ,P eb,t are the energy production of photovoltaic, wind turbine, fuel cell, electrolyzer and electric boiler at time t; P buy,pg,t ,P buy,hpp,t ,P buy,res,t ,P buy,ind,t are the energy purchased from the power grid, hydrogen supply plant, residential area and industrial area respectively; P es,ind ,P es,res are the energy storage in industrial and residential areas respectively; P load,ind ,P load,res They are the load demands of industrial areas and residential areas respectively.

[0128] Establish a relationship model between the power and cost of equipment in the system, as follows:

[0129]

[0130]

[0131]

[0132]

[0133]

[0134] In the formula, C total is the total system cost; n is the number of subnets; j∈M i Indicates device j belonging to subnet i; are the capacity consumption cost, regular maintenance cost, failure cost, and energy purchase cost of device j in subnet i; p j is the output power of equipment j at time t; a j ,b j ,c j is the capacity cost coefficient of equipment j; ω jis the maintenance cost factor of equipment j; is the average assembly cost of one equipment replacement; T d is the downtime; c j is the output energy price of equipment j; l is other regional subgrids, power grids and hydrogen supply plants; c i,k,t is the price of energy purchased by region i from region l at time t; E i,l,t is the amount of energy purchased by region i from region l at time t.

[0135] Furthermore, in step 4, a robust optimization algorithm is designed to solve the cost optimization model and obtain the optimal solution. In view of the complex problems of multi-microgrid systems, the global optimization problem is decomposed into local optimization problems in each microgrid area. The ADMM algorithm is used to independently optimize the cost model of each microgrid. At the same time, the global supply and demand balance is coordinated through information exchange. The overall objective function is expressed as:

[0136] L global =C total -λ∈ total

[0137] Where, L global is the global optimization objective; λ is the Lagrange multiplier; ∈ total is the supply and demand constraint deviation of the overall system.

[0138] exist Figure 3 The solution process of the objective function is shown in Figure 1. The specific solution steps are as follows:

[0139] Step 1: Decompose the global optimization problem into multiple local sub-problems, and independently optimize the local variables in each sub-problem. The local objective function of each sub-network is as follows:

[0140]

[0141] In the formula, C total,i is the cost target of regional subnet i; is the Lagrange multiplier; ρ is the step size parameter; ∈ cons,i is the power and storage constraints in subnet i; ∈ total ,∈ i They represent the deviation of supply and demand constraints of the whole network and the corresponding subnetwork respectively; It is the auxiliary variable of ADMM.

[0142] Step 2: Optimize the variable P in each subproblem i , find the local,i Minimize P i :

[0143] (1) Find partial derivatives: Calculate the local objective function with respect to P i The partial derivative of is calculated as:

[0144]

[0145] (2) Update local variables: Use the gradient descent method to adjust the variables so that the local objective function converges to the minimum value. The calculation formula is:

[0146]

[0147] In the formula, α is the learning rate and k is the number of iterations.

[0148] Step 3: Update the Lagrange multipliers λ and The calculation formulas are:

[0149] λ (k+1) =λ (k) +ρ∈ total

[0150]

[0151] Step 4: Introducing auxiliary variables through ADMM Update auxiliary variables, the calculation formula is:

[0152]

[0153] Step 5: Determine whether the convergence condition is met. If it is met, stop the iteration and output the optimal solution. The convergence condition is as follows:

[0154] ||∈ total ||≤∈ threshold

[0155] In the formula, ∈ total is the deviation of load supply and demand; ∈ threshold is the preset threshold.

[0156] If the convergence condition is not met, return to step 2 and step 3 to continue optimization and updating until the convergence condition is met.

[0157] Based on the above analysis, the present invention establishes a targeted energy scheduling model in the electric-heat-hydrogen multi-microgrid system, combining the different needs of industrial areas and residential areas. The industrial area focuses on large-scale energy consumption and is equipped with a variety of equipment such as photovoltaics, fans, electrolyzers, fuel cells, etc. to cope with the comprehensive needs of electricity, heat, and hydrogen; while the residential area is mainly based on electric boilers and small energy storage equipment to meet the daily electricity and hot water supply. The system flexibly configures electricity, heat, and hydrogen resources according to the load characteristics and demand differences of these two types of areas to ensure that the production capacity can efficiently respond to load demands on a low-cost basis. By dynamically scheduling the energy demand and production capacity of industrial and residential areas, the system cost is minimized and the energy utilization is maximized.

[0158] The method proposed in the present invention is suitable for the optimal scheduling of electric-heat-hydrogen multi-microgrid systems taking into account uncertainty. The power generation party can perform effective scheduling based on the demand data for electricity, heat and hydrogen energy of users in different regions, so as to meet the load needs of users with low-cost consumption.

[0159] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0160] The above are only some embodiments of the present invention. For those skilled in the art, several modifications and improvements can be made without departing from the creative concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. An optimization scheduling method for an electric-thermal-hydrogen multi-microgrid system considering uncertainty, characterized in that: The following steps are involved: Step 1: Build the electricity-heat-hydrogen multi-microgrid system architecture and establish the electricity, heat and hydrogen output models of each component of the system; Step 2: Analyze the uncertainty of renewable energy and establish a renewable energy uncertainty model; Step 3: Analyze the cost composition of the capacity side in the energy system and establish a mathematical model of its capacity cost; Step 4: Decompose the global optimization problem and design the alternating direction method of multipliers (ADMM) robust optimization scheduling method.

2. The method according to claim 1, characterized in that In step 1, a multi-microgrid system consisting of an industrial area and a residential area is constructed. Among them, both the industrial area and the residential area subgrid have independent energy production and energy load capabilities, and can also be purchased through the power grid (Power Grid, PG) and the hydrogen production plant (HPP). At the same time, the subgrids can trade electricity and hydrogen energy to meet their respective needs. In the electric-heat-hydrogen multi-microgrid system, the equipment composition includes: photovoltaic system (PV), wind turbine (WT), fuel cell (HFC), electrolyzer (EL), electric boiler (EB), lithium battery storage system (Electric Energy Storage System, ES), hydrogen storage system (HST) and thermal energy storage system (TES). According to the geographical conditions and load requirements of different regions, the specific equipment and capacity of industrial areas and residential areas vary. According to the operating characteristics of different devices in the system, the corresponding output model is established as follows: The photovoltaic power output model is expressed as: Where P pv,t is the output power of photovoltaic at time t (kw); P r,pv is the rated output power of photovoltaic (kw); G t ,G STC are solar irradiance (W / m2) and standard solar irradiance (1000W / m2); β is the temperature coefficient (0.00485 / K); T a,t is the temperature (K); T NOCT is the standard operating temperature of photovoltaic cells (318.15K); T m is the test temperature (K). The power output model of the fan is expressed as: Where P wt,t is the output power of the fan at time t (kw); P r,wt is the rated output power of the fan (kw); v c ,v r ,v f They are the cut-in wind speed, rated wind speed and cut-off wind speed of the fan (m / s) respectively. The power output model of the electrolyzer is expressed as: Where W el,t is the mass of hydrogen output by the electrolyzer at time t (kg); P el,t is the electric power input to the electrolytic cell at time t (kw); η el is the energy conversion efficiency of the electrolyzer; κ h is the conversion factor between hydrogen mass and energy, which is 39.54 kWh / kg. The power output model of the fuel cell is expressed as: Where P hfc,t is the output power of the fuel cell at time t (kw); H hfc,t is the output thermal power of the fuel cell at time t (kW); W hfc,t is the hydrogen input mass of the fuel cell at time t (kg); η hfce is the power conversion efficiency of the fuel cell; η hfch is the thermal energy conversion efficiency of the fuel cell. The power output model of the electric boiler is expressed as: Q eb,t =P eb,t the eb In the formula, Q eb,t is the electric power input to the electric boiler at time t (kw); P eb,t is the thermal power output of the electric boiler at time t; η eb is the thermoelectric conversion efficiency. The model of the energy storage device is expressed as: Where, k = 1, 2, 3, respectively, are the energy storage devices for electricity, heat, and hydrogen; E k,t ,E k,t-1 are the energies stored in the energy storage device at time t and t1 respectively; σ self,k is the self-discharge rate of energy storage device k; P ch,k,t and P dis,k,t are the charging and discharging powers of energy storage device k at time t; η cha,k and η dis,k are the charging and discharging efficiencies of the energy storage device k, respectively.

3. The method according to claim 2, characterized in that The output power of photovoltaic and wind turbine equipment is directly related to solar irradiance and wind speed. In step 2, the volatility of renewable energy is considered and an uncertain model is established from the perspective of solar irradiance and wind speed. Beta distribution is used to describe the distribution of solar irradiance, and the uncertainty of solar irradiance is modeled as follows: Where f(G) is the distribution of solar irradiance; Γ(x) is the gamma function; α, β are parameters in the distribution; μ, σ are the mean and standard deviation of solar irradiance, respectively. The Weibull distribution is used to describe the distribution of wind speed, and the uncertainty of wind speed is modeled as follows: In the formula, v c is the wind speed (m / s); λ, k are the parameters of Weibull distribution; are the mean and standard deviation of wind speed, respectively.

4. The method according to claim 3, characterized in that The equipment output and supply and demand balance constraints in the system are as follows: (1) The power constraint of non-energy storage equipment is: Where P pv,min ,P pv,max are the minimum and maximum power of photovoltaic power generation; P wt,min ,P wt,max are the minimum and maximum power generated by the wind turbine respectively; P hfc,min ,P hfc,max are the minimum and maximum power generated by the fuel cell; P el,min ,P el,max are the minimum and maximum power of hydrogen production by the electrolyzer; Q hfc,min ,Q hfc,max are the minimum and maximum power of fuel cell heat generation; Q eb,min ,Q eb,max They are the minimum and maximum power of electric boiler heat generation respectively. (2) The constraints of energy storage equipment are: In the formula, The upper and lower limits of the capacity of energy storage device k respectively; X k ,Y k are the state variables of charging and discharging of energy storage device k respectively. (3) The constraints of supply and demand balance are: Different subgrids can trade electricity and hydrogen energy, and can also purchase energy from the power grid and hydrogen supply plants to meet their own needs. The supply and demand balance constraints satisfied by the system are: Where M ind ,M res It is a collection of equipment for industrial and residential areas; j is the energy production of device j in the corresponding subnet; p buy,m,ind ,P buy,m,res P is the energy purchased from the power grid and hydrogen supply plant by industrial and residential areas respectively; load,ind ,P load,res They are the load demands of industrial areas and residential areas respectively. The supply and demand balance constraints satisfied within industrial and residential areas are: Where P pv,t ,P wt,t ,P hfc,t ,P el,t ,P eb,t are the energy production of photovoltaic, wind turbine, fuel cell, electrolyzer and electric boiler at time t; P buy,pg,t ,P buy,hpp,t ,P buy,res,t ,P buy,ind,t are the energy purchased from the power grid, hydrogen supply plant, residential area and industrial area respectively; P es,ind ,P es,res are the energy storage in industrial and residential areas respectively; P load,ind ,P load,res They are the load demands of industrial areas and residential areas respectively.

5. The method according to claim 1, characterized in that In step 3, a relationship model between the power and cost of equipment in the system is established, as follows: In the formula, C total is the total system cost; n is the number of subnets; j∈M i Indicates device j belonging to subnet i; are the capacity consumption cost, regular maintenance cost, failure cost, and energy purchase cost of device j in subnet i; p j is the output power of equipment j at time t; a j ,b j ,c j is the capacity cost coefficient of equipment j; ω j is the maintenance cost factor of equipment j; is the average assembly cost of one equipment replacement; T d is the downtime; c j is the output energy price of equipment j; l is other regional subgrids, power grids and hydrogen supply plants; c i,l,t is the price of energy purchased by region i from region l at time t; E i,k,t is the amount of energy purchased by region i from region l at time t.

6. The method according to claim 1, characterized in that In step 4, a robust optimization algorithm is designed to solve the cost optimization model and obtain the optimal solution. In view of the complex problems of multi-microgrid systems, the global optimization problem is decomposed into local optimization problems in each microgrid area. The ADMM algorithm is used to independently optimize the cost model of each microgrid. At the same time, the global supply and demand balance is coordinated through information exchange. The overall objective function is expressed as: L global =C total -λ∈ total Where, L global is the global optimization objective; λ is the Lagrange multiplier; ∈ total is the supply and demand constraint deviation of the overall system. The steps to solve the overall objective function are as follows: Step 1: Decompose the global optimization problem into multiple local sub-problems, and independently optimize the local variables in each sub-problem. The local objective function of each sub-network is as follows: In the formula, C total,i is the cost target of regional subnet i; is the Lagrange multiplier; ρ is the step size parameter; ∈ cons,i is the power and storage constraints in subnet i; ∈ total ,∈ i They represent the deviation of supply and demand constraints of the whole network and the corresponding subnetwork respectively; It is the auxiliary variable of ADMM. Step 2: Optimize the variable P in each subproblem i , find the local,i Minimize P i : (1) Find partial derivatives: Calculate the local objective function with respect to P i The partial derivative of is calculated as: (2) Update local variables: Use the gradient descent method to adjust the variables so that the local objective function converges to the minimum value. The calculation formula is: In the formula, α is the learning rate and k is the number of iterations. Step 3: Update the Lagrange multipliers λ and The calculation formulas are: l (k+1) =λ (k) +ρ∈ total Step 4: Introducing auxiliary variables through ADMM Update auxiliary variables, the calculation formula is: Step 5: Determine whether the convergence condition is met. If it is met, stop the iteration and output the optimal solution. The convergence condition is as follows: ||∈ total ||≤∈ threshold In the formula, ∈ total is the deviation of load supply and demand; ∈ threshold is the preset threshold. If the convergence condition is not met, return to step 2 and step 3 to continue optimization and updating until the convergence condition is met.

Images