Steel park micro-grid low-carbon optimization operation method based on distributed robust opportunity constraint

Abstract

A distributed robust opportunity constraint-based steel park micro-grid low-carbon optimization operation method belongs to the technical field of electric power systems, constructs a DRCC-based steel park micro-grid material-energy consumption-carbon emission model, performs low-carbon optimization operation, facilitates comprehensive mining of industrial emission reduction potential, optimizes and upgrades a production mode, and improves production efficiency. Decision reference considering reliability and economical efficiency is provided for production and operation of the micro-grid in the steel park, risk preferences and actual demands in production, operation and transaction are accurately balanced, an optimal implementation scheme is formulated, then the economic benefit and operation reliability of the micro-grid in the steel park are improved, and a low-carbon and efficient economic operation mode is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of power systems, and in particular relates to a low-carbon optimized operation method of a steel park microgrid based on distributed robust opportunity constraints. Background Art

[0002] With the rapid development of society and the economy, climate change caused by the ever-increasing demand for energy is one of the greatest global challenges facing human society today. Against the backdrop of an increasingly tense global energy situation and increasingly prominent environmental issues, energy conservation and carbon reduction are particularly important for high-energy-consuming industries, as major energy consumers and emitters. The steel industry, as a traditionally energy-intensive pillar industry, produces large amounts of carbon dioxide and other pollutants. Therefore, a low-carbon transformation of the steel industry is urgently needed. Due to the inherent complexity of steel production processes, the low-carbon transformation strategies of steel companies further exacerbate the multi-energy coupling characteristics of their internal integrated energy systems, making the dynamic balance of energy, information, and material flows even more difficult to control. This poses unprecedented challenges to the refinement and intelligence of production scheduling.

[0003] Furthermore, with the growing global emphasis on sustainable development, steel park microgrids, as typical high-energy-consuming and high-emission enterprises, are facing increasing environmental pressure. The stable and efficient operation of photovoltaic systems will undoubtedly provide strong support for reducing carbon emissions and cultivating a positive corporate image for steel park microgrids. However, the inherent uncertainties of photovoltaic systems also pose challenges to the production continuity of steel park microgrids. This increases the cost and difficulty of operation and maintenance management, and can even reduce the return on investment, thereby hindering the successful achievement of environmental protection goals.

[0004] Currently, stochastic optimization and robust optimization are widely used to deal with the uncertainty of renewable energy output. Stochastic optimization methods need to assume that uncertainties follow a specific distribution, but in practice it is difficult to obtain the exact probability distribution of random variables, and the estimation error of its probability distribution may have a significant impact on the optimization results. Robust optimization does not require an exact probability distribution for uncertain variables, but only considering the worst case in the optimization process often leads to overly conservative results, thereby increasing the production and operation costs of the steel park microgrid. However, distributed robust chance constraints (DRCC) deal with photovoltaic uncertainties. It does not require an exact probability distribution of uncertainties, but makes decisions based on statistical characteristics, thereby reducing the conservatism of decisions. It can provide a decision-making reference for the production and operation of steel park microgrids that takes into account both reliability and economy, and balance the risk preferences and needs of production and operation and trading of steel park microgrids.

[0005] Therefore, the prior art urgently needs a new technical solution to solve the above problems. Summary of the Invention

[0006] The technical problem to be solved by the present invention is: to provide a low-carbon optimization operation method for a steel park microgrid based on distributed robust opportunity constraints, to construct a material-energy-carbon emission model of a steel park microgrid based on DRCC, and to perform low-carbon optimization operation, which will help to fully tap the industry's emission reduction potential, optimize and upgrade the production model, and provide a decision-making reference for the production and operation of the steel park microgrid that takes into account both reliability and economy, accurately balance the risk preferences and actual needs in production operations and transactions, and formulate the most optimized implementation plan, thereby improving the economic benefits and operational reliability of the steel park microgrid and realizing a low-carbon and efficient economic operation model.

[0007] A low-carbon optimization operation method for a steel park microgrid based on distributed robust opportunity constraints includes the following steps:

[0008] Step 1: Build a material flow model for the steel park microgrid process. Model the materials in the electric arc furnace steelmaking process, build an electric arc steelmaking start-stop model and model constraints. Based on the steel park microgrid steelmaking process material model, build an energy consumption characteristic model for the steel park microgrid and model constraints.

[0009] Step 2: Build a product storage model and an air compression system model; build a gas system model to be used as fuel input to the CHP unit; and build a CHP unit and CDQ waste heat recovery technology model to be used as electrical and thermal energy output.

[0010] Step 3: Construct a comprehensive energy consumption and carbon emission model for the steel enterprise park microgrid, including an indirect carbon emission model for purchased electricity consumption, a direct carbon emission model for coke consumption, a carbon emission model for the production process after reduction, and a carbon emission model for by-product gas.

[0011] Step 4: Construct a distributed robust chance constraint model. Based on the distributed robust chance constraint method of moment information, fuzzy sets are used to deal with the uncertainty of photovoltaic power generation. Through the moment information fuzzy set with known mean and variance:

[0012]

[0013] Where: V t is the chance constraint function; Z t is the probability that the HP-DGMG active power inequality holds true;

[0014] P t is the photovoltaic information fuzzy set; E t is the mean and variance of the photovoltaic information fuzzy set; μ m,t and Π m,t 2 are random variables The mean and variance of is the set of all probability distributions of the support set Ω;

[0015] Through the fuzzy set, the distributed robust chance constraint of the active power balance of the steel park microgrid is obtained:

[0016]

[0017] Where: is the photovoltaic output power, is the output power of the CHP unit, Purchase power for the upper power grid, is the waste heat power generation power, P total,t is the total production load of the process, P load,t Operate the remaining electrical load for the steel park microgrid;

[0018] The general formula for conservative approximation using the worst-case conditional value at risk (CVaR) constraint is:

[0019]

[0020] y(x t )=-1

[0021] y 0 (x t )=P ca,t +P total,t +P load,t -(P t chp +P t grid +P t cdq )

[0022] Z t The CVaR of the risk level ε under the distribution is defined as:

[0023]

[0024] Where: (·) + is max{·,0}; E t Z t expected value under the distribution;

[0025] The conservative approximation for the chance constraint is:

[0026]

[0027] According to the saddle point theorem and duality theory, the conservative approximation of the chance constraint is transformed:

[0028]

[0029] Introducing random variables κ t The mean and variance of y(x t )μ t and (y(x t )Π t ) 2 ; The optimization of the inner sup problem in the formula is equivalent to:

[0030]

[0031] Where: w 1,t 、w 2,t and w 3,t The dual variables of the constraints are respectively; Γ is the decision variable of The cone of upper nonnegative Borel measure;

[0032] The final transformation obtained through duality theory is as the distributed robust opportunity constraint of the steel park microgrid,

[0033]

[0034] Where: p t ,q t , r t is an auxiliary variable;

[0035] Construct a distributed robust opportunity-constrained microgrid energy-carbon emission characteristic model for a steel park to optimize low-carbon operation.

[0036]

[0037] Where: C min Minimize the sum of the power purchase cost of the upper power grid, the carbon emission cost of the park, the production raw material cost, and the risk cost brought by the uncertainty of photovoltaics; C grid,t is the time-of-use electricity price; E pei,t is the free carbon emission quota (which is inconsistent with the subscript in the formula and needs to be represented by a unified letter); C carbon To fix the carbon price; is the total carbon emissions of steel enterprises; C j is the price per ton of production raw material j; j is the production raw material, i.e., coking coal, iron ore, and scrap steel; βt is the photovoltaic power loss caused by photovoltaic uncertainty; C risk For risk cost.

[0038] The material flow model of the steel park microgrid process in step 1 includes building material models for coking, sintering, pelletizing, blast furnace ironmaking, converter steelmaking, electric arc furnace steelmaking, and steel rolling processes;

[0039]

[0040] Where: Y N,k,t is the output of process k at time t; λk is a numerical matrix whose dimension represents the material conversion rate of each process; M R,k,t Y represents the quantity of production material R input to process k at time t, where k is the process flow, namely coking, sintering, pelletizing, blast furnace ironmaking, converter steelmaking, electric arc furnace steelmaking, and rolling; max,k and Y min,k are the upper and lower limits of the output of process k; jh is the coking production process, sj is the sintering production process; qt is the pelletizing production process; gl is the blast furnace ironmaking production process; zl is the converter steelmaking production process; dhl is the electric arc furnace steelmaking production process; and zg is the steel rolling production process.

[0041] Through the above design scheme, the present invention can bring the following beneficial effects:

[0042] Currently, stochastic optimization and robust optimization methods have become the two mainstream strategies for addressing the uncertainty of renewable energy generation. Stochastic optimization strategies presuppose that uncertainties follow specific probability distributions. However, in practice, accurately obtaining the probability distribution of these random variables is extremely difficult, and errors in the estimated distribution often significantly negatively impact optimization effectiveness. In contrast, robust optimization does not rely on a precise probabilistic description of uncertainties. While focusing on worst-case scenarios for planning enhances robustness, it inevitably leads to overly conservative solutions, which in turn increases production costs for steel companies. However, the Distributed Robust Chance Constraint (DRCC) approach has opened up a new path for managing the uncertainty of photovoltaic power generation. Similarly, it does not require precise probability distribution information for uncertainties, but instead cleverly uses statistical properties to guide decision-making, effectively mitigating the conservative bias in decision outcomes. This approach can provide steel companies with reliable and cost-effective decision-making for production operations, skillfully balancing operational stability with trading risk appetite and practical needs.

[0043] 2. Currently, existing research focuses on the domestic electricity market, carbon market, and ancillary services market, but has yet to fully examine the actual development of multiple markets and the current state of industry policy. Analysis of potential barriers to interaction between multiple markets is insufficient, and the mechanisms for connecting these markets have yet to be fully elaborated. This paper aims to fill this gap, facilitate further research on the mechanism construction and mutual interaction of high-energy-consuming industries in multiple markets, and promote further development of theory and practice in related fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The present invention will be further described below with reference to the accompanying drawings and specific embodiments:

[0045] Figure 1 This is a flow chart of a low-carbon optimization operation method for a steel park microgrid based on distributed robust opportunity constraints according to the present invention;

[0046] Figure 2 Schematic diagram of the energy consumption framework model of a long-process steel enterprise, showing a specific implementation method of a steel park microgrid low-carbon optimization operation method based on distributed robust opportunity constraints according to the present invention;

[0047] Figure 3 A schematic diagram of the carbon emission accounting boundary of a steel park microgrid based on distributed robust opportunity constraints is provided as a specific embodiment of the present invention;

[0048] Figure 4 Schematic diagram of the material-energy-carbon emission characteristic model of the steel park microgrid based on distributed robust opportunity constraints;

[0049] Figure 5 A schematic diagram of time-of-use electricity prices and grid dynamic carbon emission factor parameters for a specific implementation of a steel park microgrid low-carbon optimization operation method based on distributed robust opportunity constraints of the present invention;

[0050] Figure 6 A schematic diagram of the power purchase from the upper power grid after low-carbon optimization of the steel park microgrid;

[0051] Figure 7 A schematic diagram of carbon emissions of a steel park microgrid after low-carbon optimization is shown in the following figure:

[0052] Figure 8 Detailed description of the invention of a method for low-carbon optimized operation of a steel park microgrid based on distributed robust opportunity constraints. Schematic diagram of power loads of various processes after low-carbon optimization of the steel park microgrid.

[0053] Figure 9 This is a schematic diagram of the electricity purchase cost of the steel park microgrid participating in power market transactions under different confidence levels, showing a specific implementation method of the low-carbon optimized operation method of the steel park microgrid based on distributed robust opportunity constraints of the present invention. DETAILED DESCRIPTION

[0054] In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention are described in further detail below.

[0055] Example 1

[0056] A low-carbon optimization operation method for steel park microgrid based on distributed robust opportunity constraints, see Figure 1 , the method comprises the following steps:

[0057] First, based on actual production data from a long-process steel enterprise in Jiangsu Province and taking a daily steel delivery volume of approximately 27,500 tons as an example, this paper outlines the steel enterprise's processes, energy usage framework, and carbon emission accounting boundaries. The paper introduces the various industrial production processes within a steel park microgrid, categorizing typical production processes into continuous and discrete processes, and constructs a material model for the steel park microgrid's production processes. Furthermore, based on the energy consumption constraints of the long-process steel enterprise, a product storage model, an air compression system model, a gas system model, and a CHP unit and dry coke quenching waste heat recovery technology model are constructed to construct a material-energy characteristic model for the steel park microgrid. Secondly, considering the carbon emission accounting boundaries of the steel park microgrid, a material-energy-carbon emission characteristic model for the steel park microgrid is constructed based on the steel park microgrid's material-energy-carbon emission characteristic model. A distributed robust opportunity constraint approach is used to address the output uncertainty of the photovoltaic system within the steel park microgrid, and a distributed robust opportunity constraint-based material-energy-carbon emission characteristic model for the steel park microgrid is constructed. Finally, the objective function is to minimize the sum of the risk costs brought by the power purchase cost of the upper power grid, the carbon emission cost of the park, production raw materials, and photovoltaic uncertainty, in order to construct a low-carbon optimization operation model of the steel park microgrid based on distributed robust opportunity constraints. The model is solved, and the regulation and management of the steel park microgrid is realized under the premise of considering the risk cost, providing a decision-making reference that takes into account both reliability and economy for the production and operation of the steel park microgrid.

[0058] In summary, the embodiments of the present invention, through the above steps, can more accurately determine the optimal production and operation plan, improving the system's economic efficiency and energy efficiency. This helps to further tap into the steel industry's enormous potential for emission reduction, promotes the optimization and upgrading of production models, and provides reliable and economical decision-making support for the production and operation of steel park microgrids. Through this innovation, we can tailor more efficient and environmentally friendly production plans for the steel industry, contributing to the development of a green, low-carbon, and sustainable development path.

[0059] The above scheme is described in detail below with reference to specific calculation formulas and examples.

[0060] First, combine Figure 2 This paper briefly introduces the production process model of a long-process steel enterprise in Jiangsu Province, including coking, sintering, pelletizing, blast furnace ironmaking, converter steelmaking, electric arc furnace steelmaking, rolling, and the oxygen production process of the air compressor that provides oxygen in the corresponding processes.

[0061] (1) Coking process

[0062] The calculation scope of the coking process starts from the entry of raw materials into the coking process and ends when the product coke, coke oven gas and by-product tar and steam are output.

[0063] (2) Sintering process

[0064] The calculation scope of the sintering process starts from the entry of raw materials such as iron ore, coke, and blast furnace gas into the sintering plant to the output of sintered ore products.

[0065] (3) Pelletizing process

[0066] The calculation scope of the pelletizing process starts from the time when the raw materials of iron ore, coke, blast furnace gas and coke oven gas enter the pelletizing plant and ends when the finished pelletized ore is produced.

[0067] (4) Blast furnace ironmaking process

[0068] The calculation scope of the blast furnace ironmaking process is from the moment when the raw materials sintered ore, pelletized ore, as well as electricity, oxygen, steam, etc. enter the blast furnace ironmaking plant, to the output of finished molten iron and the cessation of blast furnace gas supply.

[0069] (5) Converter steelmaking process

[0070] The calculation scope of the converter steelmaking process starts from the time when 100% molten iron and by-products such as gas, oxygen, and electricity enter the steelmaking workshop, and ends with the output of steel billets and the external supply of by-products such as gas and steam.

[0071] (6) Electric arc furnace steelmaking process

[0072] The calculation scope of the electric arc furnace steelmaking process starts when the raw materials such as 100% scrap steel, by-product gas, oxygen and electricity enter the electric arc furnace workshop and ends when the steel billet is produced.

[0073] (7) Steel rolling process

[0074] The calculation scope of the steel rolling process starts when the raw steel billets, by-product gas and electricity and other resources enter the steel rolling workshop and ends when the finished steel products are output.

[0075] (8) Air compressor oxygenation process

[0076] Provide oxygen required for each process.

[0077] In summary, the above processes can be classified into the following categories: coking and scrap steel electric arc furnace steelmaking processes are intermittent output processes; sintering, pelletizing, blast furnace ironmaking, converter steelmaking and steel rolling processes are continuous output processes; and air compressor oxygen production is an auxiliary process.

[0078] Based on the continuity constraints, minimum operating time constraints, maximum operating time constraints, minimum downtime constraints of the electric arc furnace steelmaking production process, and the material flow constraints between each process, a material model of the steel park microgrid production process is constructed.

[0079] Among them, including:

[0080] 1) Construct material models for coking, sintering, pelletizing, blast furnace ironmaking, converter steelmaking, and steel rolling processes:

[0081]

[0082] Where: Y N,k,t is the output of process k at time t; k is a numerical matrix whose dimensions represent the material conversion rate of each process; M R,k,t Y represents the quantity of production material R input to process k at time t, where k is the process flow, namely coking, sintering, pelletizing, blast furnace ironmaking, converter steelmaking, electric arc furnace steelmaking, and rolling; max,k and Y min,k are the upper and lower limits of the output of process k.

[0083] 2) Establish continuity constraints for the start and stop of the electric arc furnace steelmaking production process:

[0084] ds dhl,t +dc dhl,t ≤1 (3)

[0085] ds dhl,t -dc dhl,t =op dhl,t -op dhl,t-1 (4)

[0086] Among them: ds dhl,t is the starting variable of the production process; dc dhl,t is the stop variable of the production process; op dhl,t It is the operating variable of the production process.

[0087] 3) Construct the minimum operating time constraint for the electric arc furnace steelmaking production process:

[0088]

[0089] Among them: Minop dhl The minimum running time of the production process.

[0090] 4) Construct the maximum operating time constraint for the electric arc furnace steelmaking production process:

[0091]

[0092] Among them: Maxop dhl The maximum running time of the production process.

[0093] 5) Establish the minimum downtime constraint for the electric arc furnace steelmaking production process:

[0094]

[0095] Where: Dp dhl The minimum downtime of the production process.

[0096] 6) Constructing material flow constraints for the electric arc furnace steelmaking process:

[0097]

[0098] Where: Y 6,dhl,t Y is the output of the product from the electric arc furnace steelmaking process; dhl It is the unit output of electric arc furnace steelmaking process.

[0099] Based on the above constraints, the relevant production process material model is constructed.

[0100] Based on the above models, the product storage model, air compression system model, gas system model, and CHP unit and dry quenching waste heat recovery technology model are constructed.

[0101] Among them, including:

[0102] 7) Build warehouse storage model:

[0103]

[0104] Where: Equations (9) and (10) are the balance equations of the storage link; Equation (11) is the upper and lower limit equations of the storage link. C is the warehouse number; S 0,c is the initial storage capacity of warehouse c; M m,c,t is the output of the mth process before warehouse c at time t; Y n,c,t S is the material required by the nth process after warehouse c at time t; c,t is the capacity of warehouse c at time t; S min,c is the lower limit of the storage capacity of the c-th warehouse; S max,c is the upper limit of the storage capacity of the cth warehouse. A total of three warehouses are constructed, namely the coke warehouse, the sintered ore warehouse, and the pellet warehouse.

[0105] 8) Constructing an air compression system model:

[0106] SA ca,t =SA ca,t-1 -SU ca,t Δt+α ca P ca,tΔt (12)

[0107] 0.8SA ca,ini ≤SA ca,end ≤1.2SA ca,ini (13)

[0108] V ca p ca,min ≤SA ca,t ≤V ca p ca,max (14)

[0109] SA ca,t +op ca,t V ca p ca,max ≤1.1V ca p ca,max (15)

[0110] 1.1V ca p ca,min ≤SA ca,t +op ca,t V ca p ca,min (16)

[0111] Where: Formula (12) is the air storage balance equation; Formulas (13) and (14) are the upper and lower limit equations of air storage; Formulas (15) and (16) are the air compressor start-stop equations. ca,t is the gas storage capacity of the system gas tank at time t; SU ca,t is the gas consumption of the system at time t; a ca is the air compressor efficiency; P ca,t is the air compressor output power at time t; SA ca,ini SA is the gas storage capacity of the gas tank at the initial moment; ca,end is the gas storage capacity of the gas tank at the end time; Vca is the volume of the gas tank; p ca,min and p ca,max The maximum and minimum pressures allowed in the gas tank; op ca,t is the air compressor operating variable at time t.

[0112] 9) Build a gas system model:

[0113]

[0114] V min,o ≤V o,t ≤V max,o (twenty one)

[0115] Where: μ COG , μ BFG , μ LDGare the by-product gas yields, which are proportional to the corresponding production process load; o is the by-product gas type (COG, BFG, LDG); V o,t For gas reserves; is the gas output; f j,o,t is the demand for gas o by process j; V is the gas volume input to the CHP unit; min,o and V max,o It corresponds to the upper and lower limits of the gas storage tank capacity.

[0116] 10) Construct CHP unit model:

[0117]

[0118] Where: Equations (22), (23), and (24) represent the CHP unit using by-product gas for energy supply. is the steam heat generated by the CHP unit; P chp t is the electric power of the CHP unit; δ is the length of the optimization cycle; and The upper and lower limits of the CHP unit output; θ chp is the electrical efficiency of the CHP unit; θ chp,hea t is the thermal efficiency of the CHP unit; β o is the calorific value of gas o.

[0119] 11) Constructing a low-carbon process CDQ waste heat recovery technology model:

[0120] P t cdq =λ cdp P jh,t (25)

[0121]

[0122] in: is the waste heat power generation power; cdq is the power generation coefficient; P jh,t is the coking power; and These are the upper and lower limits of the CDQ waste heat unit output.

[0123] Based on the above model, the energy consumption characteristic model of the steel park microgrid is constructed, including electrical energy, thermal energy, and oxygen energy, as follows:

[0124] 12) Energy Consumption Characteristics Constraints of Microgrids in Steel Parks:

[0125]

[0126] P tpv +P t chp +P t grid +P t cdq =P ca,t +P total,t +P load,t (28)

[0127]

[0128] Y 7,zg,t ≥F sum,task (32)

[0129] Among them: Equations (27) and (28) are active power balance constraints; Equation (29) is the gas energy balance constraint; Equation (30) is the thermal energy balance constraint; Equation (31) is the compressed air energy balance constraint; and Equation (32) is the total production task constraint. total,t is the total production load of the process; P k is the fixed load for unit production of process k; is the photovoltaic output power; is the output power of the CHP unit; Power purchased from the upper power grid; P load,t The remaining electrical loads are operated by the microgrid in the steel park; H k is the heat energy required for process k; H load,t The remaining heat load for the steel park microgrid; ca A is the efficiency of converting electricity to gas; k is the unit gas consumption of process k; F sum,task The total task volume of the process, the heat, power data, and oxygen consumption used in the above-mentioned steel park microgrid production process are all based on the actual data of a steel park microgrid in Jiangsu Province.

[0130] Combine Figure 3 This paper introduces the carbon emission accounting boundaries of steel enterprises, including indirect carbon emissions caused by purchased electricity consumption, direct carbon emissions caused by fossil energy (coke, coal, etc.) consumption, and direct carbon emissions caused by other raw materials in the production process.

[0131] Based on this, a carbon emission model of the steel park microgrid is constructed:

[0132] Table 1 Carbon emission coefficients of steel enterprise processes / materials

[0133]

[0134] 1) Construct an indirect carbon emission model for purchased electricity consumption:

[0135]

[0136] in: is the carbon emission of purchased electricity; ρ elec,t is the carbon content coefficient of purchased electricity.

[0137] 2) Constructing a direct carbon emission model for coke consumption:

[0138]

[0139] Since coke is produced by the coking process, its carbon emissions need to be calculated separately. The carbon emissions generated by other fuels can be attributed to the fixed carbon emission coefficient of the production process through the calculation method provided by WSA.

[0140] in: is the carbon emission of coke; ρ coke is the carbon content coefficient of coke; M coke,t The amount of coke used.

[0141] 3) Construct a carbon emission model for the production process after calculation:

[0142]

[0143] in: is the carbon emission of the production process; ρ k is the fixed carbon emission coefficient of process j.

[0144] 4) Constructing a carbon emission model for by-product gas:

[0145]

[0146] in: Carbon emissions from by-product coal gas; is the carbon content coefficient of by-product gas.

[0147] 5) Constructing a total carbon emissions model for steel enterprises:

[0148]

[0149] in: is the total carbon emissions of steel enterprises.

[0150] The carbon emission coefficient data used in the production process of the above-mentioned steel park microgrid is taken as an example to construct a carbon emission model of the steel park microgrid, taking the actual data of a steel park microgrid in Jiangsu Province as an example.

[0151] An 80MW photovoltaic system is installed in a microgrid plant in a steel park in Jiangsu Province. The uncertainty of the photovoltaic system is reflected in the uncertainty of the photovoltaic output. Therefore, it will bring risks to the production, operation and transaction of the steel park microgrid. Therefore, a distributed robust opportunity constraint is adopted to deal with the active power balance constraint of the steel park microgrid.

[0152] Based on this, a distributed robust chance constraint model is constructed:

[0153] The uncertainty of photovoltaic power generation is handled by using a distributed robust method based on moment information and fuzzy sets. The moment information fuzzy set with known mean and variance is expressed as follows:

[0154]

[0155] Where: V t is the chance constraint function; Z t is the probability that the HP-DGMG active power inequality holds true; P t is the photovoltaic information fuzzy set; E t is the mean and variance of the photovoltaic information fuzzy set; μ m,t and Π m,t 2 are random variables The mean and variance of is the set of all probability distributions with support set Ω.

[0156] Based on the above fuzzy sets, the distributed robust opportunity constraints for active power balance in the steel park microgrid are obtained:

[0157]

[0158] At the same time, for the convenience of calculation, the worst-case conditional value at risk (CVaR) constraint is used to conservatively approximate the opportunity constraint, and its general form is expressed as follows:

[0159]

[0160] y(x t )=-1 (41)

[0161] y 0 (x t )=P ca,t +P total,t +P load,t -(P t chp +P t grid +P t cdq ) (42)

[0162] Z t The CVaR of the risk level ε under the distribution is defined as:

[0163]

[0164] in:(·)+ is max{·,0}; E t Z t Expected value under the distribution.

[0165] Therefore, a conservative approximation to the chance constraint can be expressed as:

[0166]

[0167] According to the saddle point theorem and duality theory, Equation (44) is equivalent to a series of constraints containing second-order cone programming, which can be expressed as follows:

[0168]

[0169] Introducing random variable κ t =y(x t )P t pv , κ t The mean and variance of y(x t )μ t and (y(x t )Π t ) 2 .

[0170] The inner layer sup problem in the above formula can be optimized to be equivalent to:

[0171]

[0172] Where: w 1,t 、w 2,t and w 3,t are the dual variables of the constraints in Eq. (46); Γ is the decision variable of The cone of nonnegative Borel measure on .

[0173] Using the duality theory, the dual problem of formula (46) is as follows:

[0174]

[0175] in:<a,b> is the inner product of a and b.

[0176] Further, transform the above formula into

[0177]

[0178] stw 1,t +w 2,t κ t +w 3,t κ t 2 ≥0 (49)

[0179] w 1,t +w 2,t κ t +w 3,t κ t 2 ≥y 0 (x t )+κ t -β t (50)

[0180] w 3,t >0 (51)

[0181] Based on the strong duality theory, equations (49) and (50) can be transformed into:

[0182]

[0183] Introduce auxiliary variable p t ,q t and r t >0, so that

[0184]

[0185] Finally, formula (46) can be transformed into the following form

[0186]

[0187] Therefore, combined with the above transformation process, the active power balance constraint of the steel park microgrid is finally converted into Equation (57).

[0188] In summary, based on the DRCC-based steel park microgrid energy consumption and carbon emission model constructed above, a low-carbon optimized operation solution is implemented. The objective function is to minimize the sum of the power purchase cost of the upper power grid, the carbon emission cost of the park, the cost of production raw materials, and the risk cost caused by photovoltaic uncertainty:

[0189]

[0190] Where: C min Minimize the sum of the power purchase cost of the upper power grid, the carbon emission cost of the park, the production raw material cost, and the risk cost brought by the uncertainty of photovoltaics; C grid,t is the time-of-use electricity price; E pei,t Free carbon emission quota; C carbon is a fixed carbon price; C j is the price per ton of production raw material j; j is the production raw material, i.e. coking coal, iron ore, scrap steel; β t is the photovoltaic loss power caused by photovoltaic uncertainty; C risk For risk cost.

[0191] Considering the uncertainty of photovoltaic output, based on the energy consumption and carbon emission characteristic model of the steel park microgrid, a low-carbon optimization operation model of the steel park microgrid based on distributed robust opportunity constraints is constructed. The details are described below:

[0192] Each production process in the steel park microgrid consumes heat and electricity. Based on actual data, taking the production of 1 ton of steel as an example, an energy consumption-carbon emission characteristic model is constructed, as shown in Tables 2 and 3.

[0193] Table 2 Energy consumption parameters of steel enterprises in producing 1 ton of steel

[0194]

[0195] Table 3 Energy consumption and carbon emission data of steel enterprises in producing 1 ton of steel

[0196]

[0197]

[0198] At the same time, the active power balance constraint of the steel park microgrid is converted into a distributed robust opportunity constraint. Under the premise of responding to the time-of-use electricity price signal and the carbon price signal, the steel park microgrid is regulated and managed, providing a decision-making reference that takes into account both reliability and economy for the production and operation of the steel park microgrid, balancing the risk preferences and demands of the production and operation and transactions of the steel park microgrid, and incentivizing the utilization rate of the internal photovoltaic system. The example takes a long-process steel enterprise in Jiangsu Province that delivers about 27,500 tons of steel per day as an example. The objective function is to minimize the power purchase cost of the upper power grid, the carbon emission cost of the park, the material cost, and the risk cost. In response to the time-of-use electricity price signal and the carbon price signal, low-carbon optimization scheduling is carried out. By regulating the material flow and the start and stop of the arc furnace, the internal regulation and management of the steel park microgrid is realized. The time-of-use electricity price parameters and the upper power grid dynamic carbon emission factor parameters are as follows: Figure 6 As shown; the final result of the power purchase of the upper power grid is as follows Figure 7 As shown, analysis Figure 7 The steel park is powered by self-contained CHP units, photovoltaic equipment, electricity purchased from the upper power grid, and dry quenching waste heat power generation technology. Since the electric arc furnace steelmaking process replaces part of the blast furnace-converter production capacity to produce 7,532 tons of molten steel, the power purchase from the upper power grid has increased significantly. When the production load is concentrated in the low electricity price period, the by-product gas reserves are insufficient, and the CHP units for this part of the demand can no longer provide enough electricity. The remaining demand is met by purchasing electricity from the upper power grid; while in the high electricity price period, the total production load is reduced, and the CHP units and other power generation equipment are sufficient to support the load. The total carbon emissions of the park are as follows: Figure 8 As shown in the figure; the electricity purchase cost of the steel park microgrid participating in the power market transaction under different confidence levels is as follows: Figure 9As shown, as the confidence level decreases, the steel park microgrid's electricity purchase cost decreases when participating in power market transactions. This means that more reliable, safe, and stable operation is achieved by increasing the steel park microgrid's electricity purchase cost. Notably, when the confidence level decreases from 0.85 to 0.80, the reduction in the steel park microgrid's electricity purchase cost is smaller. This low-carbon optimization method will explore the carbon emission reduction potential of steel park microgrids, optimize and upgrade production models, and provide a decision-making reference for the production and operation of steel park microgrids that balances reliability and economy. It will also facilitate further research on the construction and interaction of multiple market mechanisms.

Claims

1. A low-carbon optimization operation method for a steel park microgrid based on distributed robust opportunity constraints, characterized by: The following steps are involved: Step 1: Build a material flow model for the steel park microgrid process. Model the materials in the electric arc furnace steelmaking process, build an electric arc steelmaking start-stop model and model constraints. Based on the steel park microgrid steelmaking process material model, build an energy consumption characteristic model for the steel park microgrid and model constraints. Step 2: Build a product storage model and an air compression system model; build a gas system model to be used as fuel input to the CHP unit; and build a CHP unit and CDQ waste heat recovery technology model to be used as electrical and thermal energy output. Step 3: Construct a comprehensive energy consumption and carbon emission model for the steel enterprise park microgrid, including an indirect carbon emission model for purchased electricity consumption, a direct carbon emission model for coke consumption, a carbon emission model for the production process after reduction, and a carbon emission model for by-product gas. Step 4: Construct a distributed robust chance constraint model. Based on the distributed robust chance constraint method of moment information, fuzzy sets are used to deal with the uncertainty of photovoltaic power generation. Through the moment information fuzzy set with known mean and variance: Where: V t is the chance constraint function; Z t is the probability that the HP-DGMG active power inequality holds true; P t is the photovoltaic information fuzzy set; E t is the mean and variance of the photovoltaic information fuzzy set; μ m,t and Π m,t 2 are random variables The mean and variance of is the set of all probability distributions of the support set Ω; Through the fuzzy set, the distributed robust chance constraint of the active power balance of the steel park microgrid is obtained: Where: is the photovoltaic output power, is the output power of the CHP unit, Purchase power for the upper power grid, is the waste heat power generation power, P total,t is the total production load of the process, P load,t Operate the remaining electrical load for the steel park microgrid; The general formula for conservative approximation using the worst-case conditional value at risk (CVaR) constraint is: y(x t )=-1 y 0 (x t )=P ca,t +P total,t +P load,t -(P t chp +P t grid +P t cdq ) Z t The CVaR of the risk level ε under the distribution is defined as: Where: (·) + is max{·,0}; E t Z t expected value under the distribution; The conservative approximation for the chance constraint is: According to the saddle point theorem and duality theory, the conservative approximation of the chance constraint is transformed: Introducing random variables κ t The mean and variance of y(x t )μ t and (y(x t )Π t ) 2 ; The optimization of the inner sup problem in the formula is equivalent to: Where: w 1,t 、w 2,t and w 3,t The dual variables of the constraints are respectively; Γ is the decision variable of The cone of upper nonnegative Borel measure; The final transformation obtained through duality theory is as the distributed robust opportunity constraint of the steel park microgrid, Where: p t ,q t , r t is an auxiliary variable; Construct a distributed robust opportunity-constrained microgrid energy-carbon emission characteristic model for a steel park to optimize low-carbon operation. Where: C min Minimize the sum of the power purchase cost of the upper power grid, the carbon emission cost of the park, the production raw material cost, and the risk cost brought by the uncertainty of photovoltaics; C grid,t is the time-of-use electricity price; E pei,t Free carbon emission quota; C carbon To fix the carbon price; is the total carbon emissions of steel enterprises; C j is the price per ton of production raw material j; j is the production raw material, i.e. coking coal, iron ore, scrap steel; β t is the photovoltaic loss power caused by photovoltaic uncertainty; C risk For risk cost.

2. The method for low-carbon optimization operation of a steel park microgrid based on distributed robust opportunity constraints according to claim 1 is characterized by: The material flow model of the steel park microgrid process includes building material models for coking, sintering, pelletizing, blast furnace ironmaking, converter steelmaking, electric arc furnace steelmaking, and steel rolling processes; Where: Y N,k,t is the output of process k at time t; k is a numerical matrix whose dimensions represent the material conversion rate of each process; M R,k,t Y represents the quantity of production material R input to process k at time t, where k is the process flow, namely coking, sintering, pelletizing, blast furnace ironmaking, converter steelmaking, electric arc furnace steelmaking, and rolling; max,k and Y min,k are the upper and lower limits of the output of process k; jh is the coking production process, sj is the sintering production process; qt is the pelletizing production process; gl is the blast furnace ironmaking production process; zl is the converter steelmaking production process; dhl is the electric arc furnace steelmaking production process; and zg is the steel rolling production process.

Images