Fuel gas hydrogen-doped multi-park two-stage robust optimization scheduling method considering electric energy interaction

Abstract

The invention discloses a fuel gas hydrogen-doped multi-park two-stage robust optimization scheduling method considering electric energy interaction, and the method comprises the following steps: obtaining initial power, heat and cold loads of a system, wind and light prediction power, and parameters of all equipment in the system; analyzing the utilization path of hydrogen in the electricity-to-gas conversion process, and establishing a hydrogen-doped gas unit, electricity-to-gas conversion and carbon capture model; a multi-park scheduling model composed of a photovoltaic park, a hydrogen-containing cogeneration park and a combined cooling heating and power park is established through a power interaction mode; establishing a fuel gas hydrogen-doped multi-park scheduling model considering electric energy interaction by taking the minimum total cost of the system as an optimization target; a two-stage robust optimization method is adopted to process the uncertainty of multi-park wind and light output, and Camp is used; and solving by a CG algorithm. Finally, the effect of the robust model is verified through example analysis, and the result shows that the two-stage robust optimization can effectively resist the risk caused by uncertain factor fluctuation, and the safety and stability of the system are improved; the introduction of the fuel gas doped with hydrogen can further improve the operation economy and reduce the carbon emission of the system.

Description

Technical Field

[0001] The present invention relates to the field of comprehensive energy optimization scheduling, and in particular to a two-stage robust optimization scheduling method for multi-parks of gas-hydrogen blending taking into account electric energy interaction. Background Art

[0002] With the development of energy internet technology, integrated energy systems that couple multiple energy networks have emerged. Compared with the operation of a single energy system, the integrated energy system optimizes the conversion and transmission of energy by leveraging the economic and technical characteristics of different energy sources, enabling coordinated planning and interactive response of multiple heterogeneous energy sources, and ultimately achieving a reasonable allocation of energy.

[0003] Multiple integrated energy systems are connected to the same local distribution area to form an integrated energy system group. The systems exchange electricity through energy interconnection lines. This measure can achieve the complementarity and mutual assistance of various heterogeneous energy flows between systems. However, the complex electricity interaction relationship between systems and the uncertainty of renewable energy output bring challenges to the centralized optimization and scheduling of the integrated energy system group. Therefore, it is particularly important to formulate a reasonable and effective scheduling plan for the multi-park integrated energy system.

[0004] The uncertainty of renewable energy sources such as wind power and photovoltaics poses a great challenge to the safe and stable operation of the system. The main methods for studying system uncertainty are stochastic optimization methods and robust optimization methods. However, stochastic optimization methods often cannot fully reflect the operating status of the system. Compared with stochastic optimization, robust optimization methods are more practical. Research on uncertainty mainly targets single independently operated integrated energy systems, and research on integrated energy system groups with complex interactions is relatively scarce.

[0005] Hydrogen energy is a clean and efficient secondary energy source. Mixing hydrogen produced by electricity-to-hydrogen conversion with natural gas and feeding it into gas-fired units can promote the recycling of hydrogen energy. Relevant studies have shown that when the hydrogen blending ratio of gas turbines is controlled within 10% to 20% and the hydrogen blending ratio of gas boilers is controlled within 2% to 20%, the safe and stable operation of gas-fired units can be guaranteed. A feasibility study was conducted on the operation of gas-fired units with hydrogen blending and the specific energy efficiency was calculated, but the use of gas-to-hydrogen blending technology in the integrated energy system group was not considered. Summary of the invention

[0006] In view of the uncertainty of wind and solar power output and the impact of hydrogen blending of gas on system scheduling, the present invention provides a two-stage robust optimization scheduling method for multi-parks with hydrogen blending of gas that takes into account electric energy interaction. The method takes into account the economic benefits of the system and the volatility of renewable energy output. First, a multi-park scheduling model containing a gas-hydrogen blending subsystem that takes into account electric energy interaction is established; secondly, a two-stage robust optimization is used to reduce the impact of fluctuations in renewable energy output; finally, the effectiveness of the model is verified through example simulation, and the impact of uncertain adjustment parameters and hydrogen blending ratio on system scheduling costs is measured. Two-stage robust optimization can effectively resist the risks brought by fluctuations in uncertain factors and improve system safety and stability; the introduction of hydrogen blending of gas can further improve operating economy and reduce system carbon emissions.

[0007] The technical solution adopted by the present invention is:

[0008] The two-stage robust optimization scheduling method for multi-park gas-hydrogen blending considering electric energy interaction includes the following steps:

[0009] Step 1: Obtain the system's initial electricity, heat, cooling loads and wind and solar power forecasts, as well as the parameters of each device in the system;

[0010] Step 2: Analyze the utilization pathways of hydrogen in the power-to-gas process and establish models for hydrogen-blended gas turbines, power-to-gas, and carbon capture;

[0011] Step 3: Establish a multi-park dispatching model consisting of a photovoltaic park, a hydrogen-containing cogeneration park, and a cooling, heating and power cogeneration park through a power interaction mode;

[0012] Step 4: Taking the minimum total system cost as the optimization goal, a multi-park scheduling model for gas-hydrogen blending considering power interaction is established;

[0013] Step 5: A two-stage robust optimization method is used to deal with the uncertainty of wind and solar power output in multiple parks, and the C&CG algorithm is used to solve it.

[0014] In step 2, the hydrogen-doped gas turbine, power-to-gas and carbon capture models are established as follows:

[0015] 1) Power-to-gas model:

[0016] Power-to-Gas (P2G) is composed of an electrolyzer (EL) and a methanogenization reactor (MR).

[0017]

[0018] Where: are the power consumption and hydrogen output of EL in park j during period t; ηEL is the hydrogen production efficiency of EL; are the methane power output, hydrogen power consumption, and CO2 consumption of MR in park j during period t, respectively; η MR is the conversion efficiency of MR; is the calorific value of CH4; is the standard density of CO2.

[0019] 2) Carbon capture equipment model:

[0020]

[0021] Where: They are the total energy consumption and operation energy consumption of carbon capture during period t respectively; is the CO2 capture amount of carbon capture in period t; P g For fixed energy consumption; is the total CO2 production of the gas unit in period t; λ is the energy consumption required to capture unit CO2; η t is the carbon capture rate.

[0022] 3) Hydrogen-blended gas turbine model:

[0023]

[0024] Where: and denote the electrical efficiency and thermal efficiency of the gas turbine respectively; and are the electrical output and thermal output of the gas turbine in the park j at time t, respectively; and are the CH4 power and H2 power consumed by the gas turbine in park j at time t, respectively; is the calorific value of hydrogen; is the hydrogen blending ratio of the gas turbine.

[0025] 4) Hydrogen-blended gas boiler model:

[0026]

[0027] Where: η GB Indicates the thermal efficiency of gas boiler; is the heat output of the gas boiler in the park at time t; and They are the natural gas power and hydrogen power consumed by the gas boiler j in the park at time t respectively; is the relative molecular mass of hydrogen; is the relative molecular mass of natural gas; It is the hydrogen blending ratio of gas boiler.

[0028] In step 3, the multi-park dispatching model constructed includes three parks, namely, a photovoltaic park, a thermal power supply park, and a combined heat and power supply park; Park 1 is a photovoltaic park, representing the electricity consumption of urban and suburban residents, and is a surplus power park, supplying electricity to Parks 2 and 3. Park 2 is a hydrogen-containing thermal power supply park, which contains a gas-hydrogen blending subsystem and represents the electricity consumption of industrial users. Park 3 is a combined heat and power supply park, representing the electricity consumption of urban residents; the distribution network transmits electricity to each park, and there is a two-way power exchange between the energy storage power station and each park, and there is a two-way power exchange between parks.

[0029] In step 4, the multi-park scheduling model for gas-hydrogen blending is established, including:

[0030] a. Objective function:

[0031]

[0032] The total operating cost F consists of five parts: electricity purchase cost, energy storage cost, gas purchase cost, operation and maintenance cost, and wind and solar power abandonment cost;

[0033] The specific cost expression is as follows:

[0034]

[0035] Where: N represents the number of parks, which is 3 in the present invention; T represents the total scheduling time, which is 24h in the present invention; represents the power purchased by park j from the distribution network at time t; They represent the discharging and charging power of the energy storage power station to park j respectively; G j,t is the gas purchasing power of the park from the gas grid at time j; is the unit operation and maintenance power of device B in park j; They represent the wind curtailment and solar curtailment at time t respectively; is the unit price of electricity sold by the distribution network at time t; Respectively represent the electricity sales and purchase prices of energy storage power stations; is the natural gas price at time t; is the unit power operation and maintenance cost of device B, and They are expressed as the penalty cost per unit of wind and solar power curtailment respectively.

[0036] b. Constraints:

[0037] 1) Balance of electricity, heat, cold and hydrogen:

[0038]

[0039] Where: Indicates the power consumption of the air conditioner. Indicates photovoltaic output; the power output of the CHP unit in Park 2 and the CCHP unit in Park 3 of the present invention are both expressed as express; and They represent the power purchased and sold by park j to park i at time t respectively; Indicates the cooling output of CCHP unit; Indicates the cooling output of the air conditioner; Indicates wind power output;

[0040] represents the electrical load of park j; represents the power consumption of EL in park j; represents the total energy consumption of carbon capture in park j; represents the thermal output of gas turbine in park j; It indicates the heat output of gas boiler in the park j; represents the heat load of park j; represents the cooling load of park j.

[0041] 2) Inter-park power transmission constraints:

[0042]

[0043]

[0044]

[0045]

[0046]

[0047] Where: represents the purchased power of park j at time t; represents the electricity sales power of park j at time t; N represents the number of parks, which is 3 here; T represents the duration.

[0048] They are the power purchased by park j from the distribution network, the power purchased by park j from park i, and the power sold by park j to park i at time t; It represents the upper limit of the power purchased by park j from the distribution network at time t, is the state variable corresponding to the purchase of electricity from the distribution network by park j; represents the upper limit of the power purchased by park j from park i at time t; is the state variable corresponding to the purchase of electricity by park j from park i; represents the upper limit of the power sold by park j to park i at time t; It is the state variable corresponding to the electricity sales from park j to park i.

[0049] 3) Coupling equipment output constraints:

[0050]

[0051]

[0052]

[0053] Where: represents the output power of device B in park j at time t; Indicates the upper limit of the output power of device B in park j; The lower limit of the output power of device B in park j; The state variable representing the output of device B at time t; The state variable representing the output of device B in the previous period t; A state variable representing the startup behavior of device B; State variable representing the downtime behavior of device B.

[0054] 4) Energy storage constraints:

[0055]

[0056]

[0057]

[0058] Where: and They represent the maximum power of charging and discharging of the energy storage power station respectively; and They respectively represent the charging and discharging status of the energy storage power station to park j at time t.

[0059] 5) Wind and solar power output constraints:

[0060]

[0061]

[0062] In the formula, represents the predicted output of wind power in park j at time t; represents the predicted output of photovoltaic power station j at time t; represents the wind abandonment at time t; Indicates the amount of abandoned light at time t.

[0063] In step 5, in view of the uncertainty of source-measured wind and solar power output, the present invention adopts a two-stage robust optimization method to construct a model with a min-max-min structure:

[0064]

[0065] The optimization variables in the above formula are x, u, and y, which are optimized from the outside to the inside; x is the set of all state variables, x = [v, r, z] T ; T is the transpose of the matrix; u is an uncertain variable;

[0066] U is the uncertain set of wind power and photovoltaic output;

[0067] Ω(x,u) represents the feasible domain of optimizing variable y given a set of (x,u). The specific expression of Ω(x,u) is as follows:

[0068]

[0069] Where: γ, λ, ν, π represent the dual variables corresponding to each constraint in the second stage minimization problem; D, K, F, G and I u is the coefficient matrix of different constraints; d and h are constant column vectors.

[0070] In step 5,

[0071] 1) Uncertainty model:

[0072] In order to describe the uncertainty of renewable energy output, the present invention constructs an uncertainty set. When the wind power and photovoltaic output reach the lower limit of the fluctuation range, the system operation cost is the highest, and the scenario is the worst at this time. Therefore, the scope of the uncertainty set ignores the upward fluctuation of renewable energy and only considers its downward fluctuation. The uncertainty set is expressed as follows:

[0073]

[0074] Where: B = [B wt (t),B pv (t)] is a binary variable; U represents the uncertainty set; represents the predicted value of wind power output at time t; represents the predicted value of photovoltaic output at time t; Indicates the maximum value of wind power output fluctuation at time t; represents the maximum value of the photovoltaic output fluctuation at time t; u = [u wt (t),u pv (t)] is a set of variables, where u wt (t) and u pv(t) respectively represent the wind power and photovoltaic power output process quantities at time t; B = [B wt (t),B pv (t)] is a binary variable. When the value is 1, the uncertain variable of the corresponding time period reaches the boundary of the interval.

[0075] In order to describe the conservativeness of the dispatching scheme, the uncertain adjustment parameters Γ of wind power and photovoltaic power are introduced: wt and Γ pv , the physical meaning is the number of hours that the output of renewable energy reaches the lower limit; the larger the uncertain adjustment parameter, the more conservative the plan.

[0076] 2) Solving algorithm:

[0077] The compact form of the two-stage robust optimization model can be expressed as:

[0078]

[0079] Where: c is c T y coefficient column vector; D, K, F, G, and I u is the coefficient matrix; d and h are constant column vectors; Represents a set of uncertain variables;

[0080] The first row represents inequality constraints; the second row represents equality constraints; the third row corresponds to inequalities containing state variables. The values ​​of wind power and photovoltaic output in the fourth row should be the predicted values ​​corresponding to each time period.

[0081] The problem of solving the two-stage robust model is a multi-layer problem, and the solution is relatively complex. The present invention adopts the C&CG algorithm to solve it. The algorithm decomposes the above formula into a main problem and a sub-problem, and iterates the two types of problems to obtain the optimal solution. The main problem and the sub-problem are as follows:

[0082]

[0083]

[0084] Where: k is the number of iterations; y l is the solution of the subproblem after the lth iteration; is the value of u in the worst scenario after the lth iteration; α represents the main problem, Represents an arbitrary number of iterations;

[0085] According to the strong duality theory, the innermost minimization of the subproblem is transformed into a maximization form, and then merged with the outermost max, resulting in the following problem:

[0086]

[0087] In the formula, D, K, F, G and Iu is the coefficient matrix; D T , K T , G T are the transposes of the D, K, and G matrices respectively, and c is c T y is the coefficient column vector; d and h are constant column vectors; d T is the transpose of the d matrix; γ, λ, ν, π represent the dual variables corresponding to each constraint in the minimization problem.

[0088] u T π is a bilinear term. When the formula of the dual problem takes the maximum value, the uncertain variable u takes the boundary of the fluctuation interval described by the uncertainty set U. By introducing auxiliary variables and linearizing the above formula, the final expression of the subproblem can be obtained:

[0089]

[0090] Where: Indicates the upper limit of wind power output fluctuation, Indicates the upper limit of PV output fluctuation.

[0091] B' is an auxiliary variable introduced; B' = [B' WT (t),B' PV (t)] T , B' WT (t) represents the auxiliary variable of wind power introduced, B' PV (t) represents the introduced photovoltaic auxiliary variable; is an upper bound on the dual variable.

[0092] After the above derivation, the two-stage robust model is finally transformed into a main problem and a sub-problem, which is then solved iteratively using the C&CG algorithm, including the following steps:

[0093] Step 1: Set the initial worst scenario for the uncertain variables and set the upper and lower bounds of the operating cost;

[0094] Step 2: Solve the main problem and update the optimal operating cost lower bound;

[0095] Step 3: Substitute the equipment decision variable values ​​solved by the main problem into the sub-problem for solution. If the sub-problem has a solution, update the upper bound of the optimal operating cost. The sub-problem returns the worst scenario and adds variables and constraints to the main problem.

[0096] Step 4: Given a convergence threshold ε, if the absolute value of the difference between the upper and lower limits of the cost is less than the threshold, stop the iteration; otherwise, continue to add constraints to the variables and iterate until the algorithm converges.

[0097] The present invention provides a two-stage robust optimization scheduling method for multi-parks of gas-hydrogen blending considering electric energy interaction, and the beneficial effects are as follows:

[0098] 1) The two-stage robust method proposed in the present invention can effectively resist the risks brought by the fluctuation of uncertain factors, and can improve the safety stability and overall economic benefits of the system. At the same time, by selecting appropriate uncertain adjustment parameters, the conservativeness of the optimization scheme can be adjusted.

[0099] 2) In the present invention, compared with the independent operation of each park, the joint operation of multiple parks can save costs through power sharing, optimize the output of each unit, and achieve mutual benefit and win-win results for each park.

[0100] 3) The present invention verifies the effect of hydrogen blending of fuel gas and analyzes that the system operating cost decreases with the increase of hydrogen blending ratio of the gas unit. When the hydrogen blending ratio increases to 16%, the system operating cost is reduced to the minimum, and the hydrogen blending ratio of the gas turbine is optimal at this time. BRIEF DESCRIPTION OF THE DRAWINGS

[0101] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0102] Figure 1 This is a structural diagram of the multi-park integrated energy system of the present invention.

[0103] Figure 2 This is a structural diagram of the fuel gas hydrogen blending subsystem of the present invention.

[0104] Figure 3 This is a flow chart of the C&CG algorithm of the present invention.

[0105] Figure 4 This is the predicted wind power of Park 1 of the present invention.

[0106] Figure 5 This is the predicted wind power of Park 2 of the present invention.

[0107] Figure 6 This is the predicted wind power of Park 3 of the present invention.

[0108] Figure 7 The photovoltaic power prediction of Park 1 of the present invention.

[0109] Figure 8 The photovoltaic power prediction for Park 2 of the present invention.

[0110] Fig. 9 The photovoltaic power prediction for Park 3 of the present invention.

[0111] Fig.10 This is a comparison chart of the operating costs of the present invention at different hydrogen blending ratios.

[0112] Fig.11This is the result of the power dispatching in Park 1 invented.

[0113] Fig.12 This is the result of the invented Park 2 power dispatching.

[0114] Fig.13 This is the thermal scheduling result of Park 2 invented.

[0115] Fig.14 This is the result of the invented Park 3 power dispatching.

[0116] Fig.15 This is the thermal scheduling result of Park 3 invented.

[0117] Fig.16 This is the cold dispatch result of Park 3 invented.

[0118] Fig.17 This is a curve diagram showing the influence of the uncertain adjustment parameters on the robust optimization of the present invention. DETAILED DESCRIPTION

[0119] In order to reduce the impact of renewable energy output fluctuations on multi-park dispatch results and promote the economic operation of the system, the present invention proposes an economic dispatch model for a multi-park integrated energy system with gas-hydrogen blending based on robust optimization. First, according to the utilization of hydrogen in the power-to-gas process, a mathematical model of hydrogen-doped gas units, two-stage power-to-gas, and carbon capture is established. Secondly, a multi-park dispatch model consisting of a photovoltaic park, a hydrogen-containing cogeneration park, and a combined heat, cooling, and power park is established through a power interaction mode. Then, with the optimization goal of minimizing the sum of the total costs of multiple parks, a two-stage robust optimization dispatch model is established, and the column-and-constraint generation (C&CG) algorithm is used to solve the model. Finally, the effect of the robust model is verified by example analysis, and the impact of different hydrogen blending ratios on the system cost is compared. The results show that the two-stage robust optimization can effectively resist the risks brought by the fluctuation of uncertain factors and improve the safety and stability of the system; the introduction of gas-hydrogen blending can further improve the economic efficiency of operation and reduce the carbon emissions of the system.

[0120] The two-stage robust optimization scheduling method for multi-park gas-hydrogen blending considering electric energy interaction includes the following steps:

[0121] Step 1: Obtain the system's initial electricity, heat, and cooling loads, wind and solar power forecasts, and the parameters of each device in the system; 24-hour initial electricity, heat, and cooling loads, wind and solar power forecasts. Other model parameters include natural gas unit calorific value price, time-of-use electricity price data, gas turbine conversion efficiency, aging loss coefficient, etc.

[0122] Step 2: Divide the power-to-gas process into two stages: power-to-hydrogen and methanation. Part of the hydrogen produced by electrolysis is directly fed into the gas turbine and gas boiler. The carbon capture equipment captures the CO2 produced by the gas boiler and gas turbine, and then feeds the other part of the hydrogen and CO2 produced by electrolysis into the methane reactor for methanation. The fuel consumed by the gas boiler and gas turbine is provided by hydrogen and natural gas produced by electrolysis and gas purchased from the gas grid. On this basis, a mathematical model of hydrogen-blended gas units, power-to-gas and carbon capture is established, as follows:

[0123] 1) Power-to-gas model

[0124] Power-to-Gas (P2G) is composed of an electrolyzer (EL) and a methanogenization reactor (MR).

[0125]

[0126] Where: are the power consumption and hydrogen output of EL in park j during period t; η EL is the hydrogen production efficiency of EL; are the methane power output, hydrogen power consumption, and CO2 consumption of MR in park j during period t, respectively; η MR is the conversion efficiency of MR; is the calorific value of CH4; is the standard density of CO2.

[0127] 2) Carbon capture equipment model

[0128]

[0129] Where: are the total energy consumption and operation energy consumption of carbon capture during period t, is the CO2 capture amount of carbon capture in period t; P g For fixed energy consumption; is the total CO2 production of the gas unit in period t; λ is the energy consumption required to capture unit CO2; η t is the carbon capture rate.

[0130] 3) Hydrogen-blended gas turbine model

[0131]

[0132] Where: and denote the electrical efficiency and thermal efficiency of the gas turbine respectively; and are the electrical output and thermal output of the gas turbine in the park j at time t, respectively; and are the CH4 power and H2 power consumed by the gas turbine in the park j at time t, is the calorific value of hydrogen, is the hydrogen blending ratio of the gas turbine.

[0133] 4) Hydrogen-blended gas boiler model.

[0134]

[0135] Where: η GB Indicates the thermal efficiency of gas boiler; is the heat output of the gas boiler in the park at time t; and are the natural gas power and hydrogen power consumed by the gas boiler j in the park at time t, is the relative molecular mass of hydrogen, is the relative molecular mass of natural gas, It is the hydrogen blending ratio of gas boiler.

[0136] Step 3: Establish a multi-park scheduling model consisting of a photovoltaic park, a combined heat and power park, and a combined cooling, heating and power park through the power interaction mode, as follows:

[0137] The constructed multi-park dispatching model includes three parks, namely the photovoltaic park, the thermal power supply park, and the combined heat and power supply park. Park 1 is a photovoltaic park, which represents the electricity consumption of urban and suburban residential users. It is a surplus power park and supplies power to parks 2 and 3. Park 2 is a hydrogen-containing thermal power supply park, which contains a gas-hydrogen blending subsystem and represents the electricity consumption of industrial users. Park 3 is a combined heat and power supply park, which represents the electricity consumption of urban residential users. The distribution network transmits electricity to each park, and there is a two-way power exchange between the energy storage power station and each park, and there is a two-way power exchange between parks.

[0138] Step 4: Taking the minimum total system cost as the optimization goal, establish a multi-park scheduling model for gas-hydrogen blending considering power interaction:

[0139] a. Objective function:

[0140]

[0141] The total operating cost F consists of five parts: electricity purchase cost, energy storage cost, gas purchase cost, operation and maintenance cost, and wind and solar power abandonment cost. The specific cost expression is as follows:

[0142]

[0143] Where: N represents the number of parks, which is 3 in the present invention; T represents the total scheduling time, which is 24h in the present invention; represents the power purchased by park j from the distribution network at time t, They represent the discharging and charging power of the energy storage power station to park j respectively; G j,t is the gas purchasing power of the park from the gas grid at time j; is the unit operation and maintenance power of device B in park j, They represent the wind curtailment and solar curtailment at time t, respectively. is the electricity selling price of the distribution network at time t, They represent the electricity sales and purchase prices of energy storage power stations respectively, is the natural gas price at time t, is the unit power operation and maintenance cost of device B, and They are expressed as the penalty cost per unit of wind and solar power abandonment, respectively.

[0144] b. Constraints:

[0145] 1) Balance of electricity, heat, cold and hydrogen

[0146]

[0147] Where: Indicates the power consumption of the air conditioner. The photovoltaic output is represented by the CHP unit in Park 2 and the CCHP unit in Park 3 of the present invention. express, and They represent the power purchased and sold by park j to park i at time t, Indicates the cooling output of CCHP unit. Indicates the cooling output of the air conditioner.

[0148] 2) Inter-park power transmission constraints

[0149]

[0150]

[0151]

[0152]

[0153]

[0154] Where: They are the power purchased by park j from the distribution network at time t and the power purchased and sold by park j to park i; and They respectively represent the upper limit of the power purchased by park j from the distribution network at time t, the upper limit of the power purchased and sold to park i, and the state variables of this behavior.

[0155] 3) Coupling equipment output constraints

[0156]

[0157]

[0158]

[0159] Where: They respectively represent the output power, upper and lower limits of the output power, the state variables of the behavior, and the state variables of the start and stop behavior of the device B in the park j at time t.

[0160] 4) Energy storage constraints

[0161]

[0162]

[0163]

[0164] Where: and They represent the maximum power of charging and discharging of the energy storage station, and They respectively represent the charging and discharging status of the energy storage power station to park j at time t.

[0165] 5) Wind and solar power output constraints

[0166]

[0167]

[0168] In the formula, represents the predicted output of wind power in park j at time t, It represents the predicted output of photovoltaic power plant j at time t. Step 5: A two-stage robust optimization method is used to deal with the uncertainty of wind and solar power output, and the C&CG algorithm is used to solve it.

[0169] Aiming at the uncertainty of source-measured wind and solar power output, the present invention adopts a two-stage robust optimization method to construct a model with a min-max-min structure.

[0170]

[0171] The optimization variables in the above formula are x, u, and y, which are optimized from the outside to the inside. x is the set of all state variables, U is the uncertain set of wind power and photovoltaic output, Ω(x,u) represents the feasible domain of optimizing variable y when a set (x,u) is given, and the specific expression of Ω(x,u) is as follows:

[0172]

[0173] Where γ, λ, ν, and π represent the dual variables corresponding to each constraint in the minimization problem of the second stage.

[0174] 1) Uncertainty Model

[0175] In order to describe the uncertainty of renewable energy output, the present invention constructs an uncertainty set. When the wind power and photovoltaic output reach the lower limit of the fluctuation range, the system operation cost is the highest, and the scenario is the worst at this time. Therefore, the scope of the uncertainty set ignores the upward fluctuation of renewable energy and only considers its downward fluctuation. The uncertainty set is expressed as follows:

[0176]

[0177] Where: are the predicted power values ​​and the maximum fluctuation values ​​of wind power and photovoltaic power at time t respectively. wt (t),B pv (t)] is a binary variable. To describe the conservativeness of the dispatching scheme, the uncertain adjustment parameters Γ of wind power and photovoltaic power are introduced wt and Γ pv , which physically means the number of hours when the output of renewable energy reaches the lower limit. The larger the uncertain adjustment parameter, the more conservative the solution.

[0178] 2) Solution algorithm

[0179] The robust optimization model can be expressed as:

[0180]

[0181] Where: c is c T y coefficient column vector; D, K, F, G, and I uis a coefficient matrix; d and h are constant column vectors. The first row represents inequality constraints; the second row represents equality constraints; the third row corresponds to inequalities containing state variables. The values ​​of wind power and photovoltaic output in the fourth row should be the predicted values ​​corresponding to each time period. This problem is a multi-layer problem and is relatively complex to solve. The present invention uses the C&CG algorithm to solve it. The algorithm decomposes the above formula into a main problem and sub-problems, and iterates the two types of problems to obtain the optimal solution. The main problem and sub-problems are as follows:

[0182]

[0183]

[0184] k is the number of iterations; y l is the solution of the subproblem after the lth iteration; is the value of u in the worst scenario obtained after the lth iteration.

[0185] According to the strong duality theory, the innermost minimization of the subproblem is transformed into a maximization form, and then merged with the outermost max, resulting in the following problem:

[0186]

[0187] In the formula, u T π is a bilinear term. When the formula of the dual problem takes the maximum value, the uncertain variable u takes the boundary of the fluctuation interval described by the uncertainty set U. By introducing auxiliary variables and linearizing the above formula, the final expression of the subproblem can be obtained:

[0188]

[0189] Where: B' is an auxiliary variable introduced, B'=[B' WT (t),B' PV (t)] T , is an upper bound on the dual variable.

[0190] The structure diagram of the multi-park integrated energy system of the present invention is as follows: Figure 1 As shown, the structure diagram of the fuel gas hydrogen blending subsystem of the present invention is as follows Figure 2 As shown, the C&CG algorithm flow chart of the present invention is as follows Figure 3 shown.

[0191] Example:

[0192] In order to verify the effectiveness of the two-stage robust model and algorithm built in this invention in optimizing scheduling and resisting the risk of uncertain variable fluctuations, Figure 1Taking the multi-park integrated energy system shown in the figure as an example, Matlab is used to call the Cplex solver for solution. The present invention sets the fluctuation deviation of wind power and photovoltaic output to be 15% of the predicted value, and the uncertain adjustment parameter Γ is set to 10. The wind power and photovoltaic prediction data of the three parks are shown in Figures 4 to 9 The specific parameters of each device in the multi-park system and the energy time-sharing price are shown in Table 1 and Table 2.

[0193] Table 1 System equipment parameters

[0194]

[0195] Table 2 Energy time-of-use prices

[0196]

[0197] In order to fully illustrate the effectiveness of the proposed multi-park scheduling model that considers the uncertainty of wind and solar power and has power interaction, three schemes are set up for comparative analysis. The costs corresponding to different schemes are shown in Table 3.

[0198] Option 1: Each park does not consider the uncertainty of renewable energy output, and there is no power exchange between parks.

[0199] Option 2: Each park does not consider the uncertainty of renewable energy output, and there is power exchange between parks.

[0200] Option 3: Each park takes the uncertainty of renewable energy output into consideration, and there is power exchange between parks.

[0201] Table 3 Scheduling results of schemes 1-3

[0202]

[0203] Comparing Scheme 1 and Scheme 2, it can be seen that after the power exchange between the parks, the cost of each park was reduced by 34.65 yuan, 1227.98 yuan and 4523.59 yuan respectively, and the total cost was reduced by 5786.22 yuan. The results show that compared with the independent operation of each park, the multi-park scheduling model with power exchange can save costs through power sharing, help to fully utilize energy, and thus improve the overall economic benefits.

[0204] Comparing Scheme 2 and Scheme 3, it can be seen that considering the uncertainty of renewable energy output will increase the operating cost. This is because the actual value of renewable energy output is lower than the predicted value. In order to resist the risk brought by renewable energy fluctuations, the system needs to purchase more electricity from the distribution network, so the cost increases. Compared with Scheme 2, the total cost of Scheme 3 increased by 2,285.78 yuan, and the costs of each park increased by 388.53 yuan, 834.53 yuan, and 1,062.72 yuan respectively.

[0205] In order to verify the effectiveness of hydrogen-blended gas in improving economic benefits and reducing carbon emissions, Scheme 4 is added for comparative analysis based on Scheme 3, which takes into account the uncertainty of renewable energy and the interaction of electric energy between parks. Scheme 4: Considering that Park 2 contains a hydrogen-blended gas system with a hydrogen blending ratio of 15%, the corresponding costs of the two schemes are shown in Table 4:

[0206] Table 4 Scheduling results under different hydrogen blending scenarios

[0207]

[0208] As can be seen from Table 4, after the introduction of hydrogen blending in Scheme 4 compared to Scheme 3, the cost of Park 2 was reduced by 2,624.59 yuan and the carbon emissions were reduced by 1.09 tons. This is because after the introduction of the gas-hydrogen blending system, the hydrogen produced by the electrolyzer is mixed with natural gas and sent to the hydrogen-blended gas unit, which increases the power output of the hydrogen-blended gas turbine to meet the power consumption of electrolysis and reduces the system's power purchase. After hydrogen blending, the purchase of gas increases, and the carbon emissions of the gas turbine increase accordingly. However, due to the fixed heat-to-electricity ratio, the increase in the thermal output of the gas turbine leads to a decrease in the thermal output of the gas boiler, so the carbon emissions of the gas boiler are reduced. Since the unit carbon emissions of the gas turbine are relatively small, the overall carbon emissions of the system are reduced. In summary, the operation of the gas unit with hydrogen blending can effectively reduce the operating costs and carbon emissions of the system.

[0209] The present invention mainly discusses the influence of hydrogen blending ratio on the operation cost, and draws the relationship curve between the operation cost of Park 2 and the hydrogen blending ratio, such as Fig.10 As shown. Fig.10 It can be seen that if the hydrogen blending ratio of the gas boiler remains unchanged, when the hydrogen blending ratio of the gas turbine is relatively small, the operating cost of Park 2 decreases with the increase of the hydrogen blending ratio of the gas turbine, and the decrease is obvious. When the hydrogen blending ratio of the gas turbine is greater than 16%, the cost reduction is not obvious, and the curve tends to be flat. At the same time, with the increase of the hydrogen blending ratio of the gas boiler, the cost is further reduced. The above changes are because after the gas unit is operated with hydrogen blending, the hydrogen produced by electricity to hydrogen is mixed with natural gas and sent to the gas unit to meet the demand for electric and thermal loads, thereby reducing the purchase of electricity from the distribution network and saving costs. However, with the increase of the hydrogen blending ratio, the electric energy consumed by the electrolyzer increases, the output increase of the gas unit increases, and the increase in the system gas purchase cost increases, so the operating cost changes little and the curve tends to be flat.

[0210] The dispatch results of each park are as follows: Figures 11 to 16 shown.

[0211] In order to explore the influence of uncertain adjustment parameters on the robust optimization results, the uncertainty adjustment parameter Γ is set to a value of 0-24. The trend of the day-ahead operation cost corresponding to different uncertain adjustment parameters is shown in the following figure. Fig.17 Analysis Fig.17It can be seen that as the uncertain adjustment parameter increases, the system's total cost a day ahead shows a trend of first increasing and then flattening out. This is because the larger the value of the uncertain adjustment parameter, the worse the corresponding day-ahead forecast scenario, the more conservative the solution obtained, and the higher the day-ahead scheduling cost. However, when the uncertain adjustment parameter is very large, the range of variation in the severity of the scenario is relatively small, so the curve tends to be flat.

[0212] Table 5 Cost comparison of different optimization methods

[0213]

[0214] In order to verify the effect of two-stage robust optimization, the present invention compares the day-ahead dispatching cost and actual cost obtained by the robust optimization method and the deterministic optimization method. As shown in Table 5, the day-ahead dispatching scheme obtained by the two-stage robust optimization method has a higher day-ahead prediction cost. This is because the method considers the influence of uncertain parameters and predicts the worst scenario, so the cost is higher. The day-ahead prediction cost of the conventional deterministic optimization method is lower, because the method does not consider the influence of uncertain adjustment parameters. The volatility of wind and solar power leads to a certain error between the day-ahead predicted power generation and actual power generation and consumption, which needs to be compensated by balancing costs. In the actual operation results, the cost of the robust optimization method is lower than that of the deterministic optimization method. This is because the robust optimization method considers the risks brought by the fluctuation of uncertain factors, which can be compensated with only a small balancing cost. For the conventional deterministic optimization method, since the purchase price of electricity in the real-time market is generally higher than that in the day-ahead market, and when the power is insufficient, a certain compensation is required, which makes the actual operation cost of the conventional deterministic optimization method higher. As shown in Table 5, the total cost of the robust optimization method used in the present invention is reduced by 1.6% compared with the deterministic optimization method.

Claims

1. A two-stage robust optimization scheduling method for multi-park gas-hydrogen blending considering electric energy interaction, characterized by The following steps are involved: Step 1: Obtain the system's initial electricity, heat, cooling loads and wind and solar power forecasts, as well as the parameters of each device in the system; Step 2: Analyze the utilization pathways of hydrogen in the power-to-gas process and establish models for hydrogen-blended gas turbines, power-to-gas, and carbon capture; Step 3: Establish a multi-park dispatching model consisting of a photovoltaic park, a hydrogen-containing cogeneration park, and a cooling, heating and power cogeneration park through a power interaction mode; Step 4: Taking the minimum total system cost as the optimization goal, a multi-park scheduling model for gas-hydrogen blending considering power interaction is established; Step 5: A two-stage robust optimization method is used to deal with the uncertainty of wind and solar power output in multiple parks, and the C&CG algorithm is used to solve it.

2. According to claim 1, the method for multi-park two-stage robust optimization scheduling of gas-hydrogen blending considering electric energy interaction is characterized by: In step 2, the hydrogen-doped gas turbine, power-to-gas and carbon capture models are established as follows: 1) Power-to-gas model: Power-to-Gas (P2G) is composed of an electrolyzer (EL) and a methanogenization reactor (MR); Where: are the power consumption and hydrogen output of EL in park j during period t; η EL is the hydrogen production efficiency of EL; are the methane power output, hydrogen power consumption, and CO2 consumption of MR in park j during period t, respectively; η MR is the conversion efficiency of MR; is the calorific value of CH4; is the standard density of CO2; 2) Carbon capture equipment model: Where: They are the total energy consumption and operation energy consumption of carbon capture during period t respectively; is the CO2 capture amount of carbon capture in period t; P g For fixed energy consumption; is the total CO2 production of the gas unit in period t; λ is the energy consumption required to capture unit CO2; η t is the carbon capture rate; 3) Hydrogen-blended gas turbine model: Where: and denote the electrical efficiency and thermal efficiency of the gas turbine respectively; and are the electrical output and thermal output of the gas turbine in the park j at time t, respectively; and are the CH4 power and H2 power consumed by the gas turbine in park j at time t, respectively; is the calorific value of hydrogen; is the gas turbine hydrogen blending ratio; 4) Hydrogen-blended gas boiler model: Where: η GB Indicates the thermal efficiency of gas boiler; is the heat output of the gas boiler in the park at time t; and are the natural gas power and hydrogen power consumed by the gas boiler j in the park at time t respectively; is the relative molecular mass of hydrogen; is the relative molecular mass of natural gas; It is the hydrogen blending ratio of gas boiler.

3. According to claim 1, the method for multi-park two-stage robust optimization scheduling of gas-hydrogen blending considering electric energy interaction is characterized by: In step 3, the constructed multi-park scheduling model includes three parks, namely a photovoltaic park, a thermal power supply park, and a combined heat, power and cooling park; Park 1 is a photovoltaic park, representing the electricity consumption of urban and suburban residents, and is a surplus power park, supplying electricity to Parks 2 and 3; Park 2 is a hydrogen-containing thermal power supply park, which includes a gas-hydrogen blending subsystem and represents the electricity consumption of industrial users; Park 3 is a combined heat, power and cooling park, representing the electricity consumption of urban residents; the distribution network transmits electricity to each park, and there is a two-way electricity exchange between the energy storage power station and each park, and there is a two-way electricity exchange between the parks.

4. According to claim 1, the multi-park two-stage robust optimization scheduling method for gas-hydrogen blending considering electric energy interaction is characterized by: In step 4, the multi-park scheduling model for gas-hydrogen blending is established, including: a. Objective function: The total operating cost F consists of five parts: electricity purchase cost, energy storage cost, gas purchase cost, operation and maintenance cost, and wind and solar power abandonment cost; The specific cost expression is as follows: Where: N represents the number of parks; T represents the total scheduling time; represents the power purchased by park j from the distribution network at time t; They represent the discharging and charging power of the energy storage power station to park j respectively; G j,t is the gas purchasing power of the park from the gas grid at time j; is the unit operation and maintenance power of device B in park j; They represent the wind curtailment and solar curtailment at time t respectively; is the unit price of electricity sold by the distribution network at time t; Respectively represent the electricity sales and purchase prices of energy storage power stations; is the natural gas price at time t; is the unit power operation and maintenance cost of device B, and They are expressed as the penalty cost per unit of wind and solar curtailment respectively; b. Constraints: 1) Balance of electricity, heat, cold and hydrogen: Where: Indicates the power consumption of the air conditioner. Indicates photovoltaic output; the CHP units in Park 2 and the CCHP units in Park 3 are both used for power output express; and They represent the power purchased and sold by park j to park i at time t respectively; Indicates the cooling output of CCHP unit; Indicates the cooling output of the air conditioner; Indicates wind power output; represents the electrical load of park j; represents the power consumption of EL in park j; represents the total energy consumption of carbon capture in park j; represents the thermal output of gas turbine in park j; It indicates the heat output of gas boiler in the park j; represents the heat load of park j; represents the cooling load of park j; 2) Inter-park power transmission constraints: Where: represents the purchased power of park j at time t; represents the power sold by park j at time t; N represents the number of parks, which is 3 here; T represents the duration; They are the power purchased by park j from the distribution network, the power purchased by park j from park i, and the power sold by park j to park i at time t; It represents the upper limit of the power purchased by park j from the distribution network at time t, is the state variable corresponding to the purchase of electricity from the distribution network by park j; represents the upper limit of the power purchased by park j from park i at time t; is the state variable corresponding to the purchase of electricity by park j from park i; represents the upper limit of the power sold by park j to park i at time t; is the state variable corresponding to the electricity sales from park j to park i; 3) Coupling equipment output constraints: Where: represents the output power of device B in park j at time t; Indicates the upper limit of the output power of device B in park j; The lower limit of the output power of device B in park j; The state variable representing the output of device B at time t; The state variable representing the output of device B in the previous period t; A state variable representing the startup behavior of device B; The state variable representing the shutdown behavior of device B; 4) Energy storage constraints: Where: and They represent the maximum power of charging and discharging of the energy storage power station respectively; and They respectively represent the charging and discharging status of the energy storage power station to park j at time t; 5) Wind and solar power output constraints: In the formula, represents the predicted output of wind power in park j at time t; represents the predicted output of photovoltaic power station j at time t; represents the amount of wind abandoned at time t; Indicates the amount of abandoned light at time t.

5. According to claim 1, the multi-park two-stage robust optimization scheduling method for gas-hydrogen blending considering electric energy interaction is characterized by: In step 5, a two-stage robust optimization method is used to construct a min-max-min model in view of the uncertainty of source-measured wind and solar power output: The optimization variables in the above formula are x, u, and y, which are optimized from the outside to the inside; x is the set of all state variables, x = [v, r, z] T ; T is the transpose of the matrix; u is an uncertain variable; U is the uncertain set of wind power and photovoltaic output; Ω(x,u) represents the feasible domain of optimizing variable y given a set of (x,u). The specific expression of Ω(x,u) is as follows: Where: γ, λ, ν, π represent the dual variables corresponding to each constraint in the second stage minimization problem; D, K, F, G and I u is the coefficient matrix of different constraints; d and h are constant column vectors.

6. The two-stage robust optimization scheduling method for multi-parks of gas-hydrogen blending considering electric energy interaction according to claim 5 is characterized by: Step 5 includes: 1) Uncertainty model: In order to describe the uncertainty of renewable energy output, an uncertainty set is constructed. When the wind power and photovoltaic output reach the lower limit of the fluctuation range, the system operation cost is the highest, and the scenario is the worst at this time. Therefore, the scope of the uncertainty set ignores the upward fluctuation of renewable energy and only considers its downward fluctuation. The uncertainty set is expressed as follows: Where: B = [B wt (t),B pv (t)] is a binary variable; U represents the uncertainty set; represents the predicted value of wind power output at time t; represents the predicted value of photovoltaic output at time t; Indicates the maximum value of wind power output fluctuation at time t; represents the maximum value of the photovoltaic output fluctuation at time t; u = [u wt (t),u pv (t)] is a set of variables, where u wt (t) and u pv (t) respectively represent the wind power and photovoltaic power output process quantities at time t; B = [B wt (t),B pv (t)] is a binary variable. When the value is 1, the uncertain variable of the corresponding period is at the boundary of the interval; In order to describe the conservativeness of the dispatching scheme, the uncertain adjustment parameters Γ of wind power and photovoltaic power are introduced: wt and Γ pv , which indicates the number of hours when the output of renewable energy reaches the lower limit; the larger the uncertain adjustment parameter, the more conservative the scheme; 2) Solving algorithm: The compact form of the two-stage robust optimization model is: Where: c is c T y coefficient column vector; D, K, F, G, and I u is the coefficient matrix; d and h are constant column vectors; Represents a set of uncertain variables; The first row represents inequality constraints; the second row represents equality constraints; the third row corresponds to inequalities containing state variables; the values ​​of wind power and photovoltaic output in the fourth row should be the predicted values ​​corresponding to each time period; The solution of the two-stage robust model is a multi-layer problem, which is solved by the C&CG algorithm. The algorithm decomposes the above equation into a main problem and sub-problems, and iterates the two types of problems to obtain the optimal solution. The main problem and sub-problems are as follows: Where: k is the number of iterations; y l is the solution of the subproblem after the lth iteration; is the value of u in the worst scenario after the lth iteration; α represents the main problem, Represents an arbitrary number of iterations; According to the strong duality theory, the innermost minimization of the subproblem is transformed into a maximization form, and then merged with the outermost max, resulting in the following problem: In the formula, D, K, F, G and I u is the coefficient matrix; D T , K T , G T are the transposes of the D, K, and G matrices respectively, and c is c T y is the coefficient column vector; d and h are constant column vectors; d T is the transpose of the d matrix; γ, λ, ν, π represent the dual variables corresponding to each constraint in the minimization problem; u T π is a bilinear term. When the formula of the dual problem takes the maximum value, the uncertain variable u takes the boundary of the fluctuation interval described by the uncertainty set U. By introducing auxiliary variables and linearizing the above formula, the final expression of the subproblem can be obtained: Where: Indicates the upper limit of wind power output fluctuation, Indicates the upper limit of photovoltaic output fluctuation; B' is an auxiliary variable introduced; B'=[B' WT (t),B' PV (t)] T , B' WT (t) represents the auxiliary variable of wind power introduced, B' PV (t) represents the introduced photovoltaic auxiliary variable; is the upper bound of the dual variable; After the above derivation, the two-stage robust model is finally transformed into a main problem and sub-problems, which are then solved iteratively using the C&CG algorithm.

7. The two-stage robust optimization scheduling method for multi-parks of gas-hydrogen blending considering electric energy interaction according to claim 6 is characterized by: The iterative solution of the C&CG algorithm includes the following steps: Step 1: Set the initial worst scenario for the uncertain variables and set the upper and lower bounds of the operating cost; Step 2: Solve the main problem and update the optimal operating cost lower bound; Step 3: Substitute the equipment decision variable values ​​solved by the main problem into the sub-problem for solution. If the sub-problem has a solution, update the upper bound of the optimal operating cost. The sub-problem returns the worst scenario and adds variables and constraints to the main problem. Step 4: Given a convergence threshold ε, if the absolute value of the difference between the upper and lower limits of the cost is less than the threshold, stop the iteration; otherwise, continue to add constraints to the variables and iterate until the algorithm converges.

Images