Two-stage robust optimization method for regional-level electricity-gas-heat system based on CCG algorithm

Abstract

The invention requests to protect a method based on Camp; the invention discloses a region-level electricity-gas-heat system two-stage robust optimization method based on a CG algorithm. The method comprises the following steps: determining a linearization model, an objective function and constraint conditions of a distribution network-level region electricity-gas-heat system; constructing a two-stage robust optimization model by considering a real-time extreme scene with the maximum day-ahead and real-time wind power and load prediction error; camp is adopted; the CG algorithm is used for decomposing the model for iterative solution; and obtaining an optimal strategy of robust optimization scheduling of the two stages of the regional-level electricity-gas-heat system according to a solving result. The method has effectiveness in the aspects of improving system robustness and wind power absorption capacity, resisting real-time compensation electricity price fluctuation and the like, the heat loss problem of the regional heat supply network is considered, the actual situation of a distribution network level regional electricity-gas-heat system is met, and the method is more economical.

Description

Technical Field

[0001] The present invention belongs to the field of optimized scheduling of integrated energy systems, and in particular relates to a two-stage robust optimized scheduling method capable of handling the coordinated optimized operation of multiple energy networks including electricity, gas and heat. Background Art

[0002] In modern energy systems, the coupling and coordinated optimization of different energy sources, such as electricity, gas, and heat, are key to achieving efficient energy utilization and meeting growing energy demand. With the large-scale integration of renewable energy and the diversification of energy consumption patterns, traditional energy system optimization methods face new challenges. Against this backdrop, a two-stage robust optimization method for regional electricity-gas-heat systems based on the C&CG (Chance-Constrained and Convex-Growth) algorithm has emerged. This two-stage robust optimization method is an effective means of addressing uncertainty. It allows the system to formulate a relatively conservative plan in the first stage and then optimize according to actual conditions in the second stage to account for the impact of uncertainty. The C&CG algorithm is one of the key technologies for implementing two-stage robust optimization. The Chance-Constrained algorithm can handle the influence of random variables on the optimization objective, while the Convex-Growth algorithm leverages convex optimization theory to simplify the problem, making it easier to solve. By combining these two algorithms, uncertainty and nonlinearity in electricity-gas-heat systems can be effectively addressed.

[0003] The power-gas-heat system is a complex, coupled multi-energy system that involves the conversion and supply of electricity, natural gas, and heat. This multi-energy coupling complicates system optimization, but it also offers the potential for improving energy efficiency and reducing costs. Regional-level power-gas-heat system optimization also needs to consider the energy distribution and demand characteristics within the region. Energy demand and supply conditions can vary significantly across regions, so optimization methods need to be adaptable to these specific characteristics. Regional-level optimization can better balance energy supply and demand across regions and improve overall system efficiency. Therefore, the two-stage robust optimization method for regional power-gas-heat systems based on the C&CG algorithm is an effective technology that adapts to current energy system development trends. It not only addresses the uncertainty and complexity of energy systems, but also improves energy efficiency and reduces operating costs, providing technical support for achieving sustainable energy systems.

[0004] After searching, application publication number CN114707783A was found. It describes a two-stage robust planning method for solar energy absorption equipment in a regional electric-thermal integrated energy system. First, the components and structure of the regional electric-thermal integrated energy system are clarified. Second, system parameters are input and initial values ​​are set for the algorithm. Subsequently, the main problem of the two-stage robust planning method for solar energy absorption equipment in the regional electric-thermal integrated energy system is solved. Finally, the difference between the upper and lower bounds is determined to be less than the allowable error to obtain the final solar energy absorption equipment planning scheme. By establishing an uncertainty set and adjusting the uncertainty of the uncertain factors, the present invention seeks a conservative economic optimal solution that still meets the system's safe operation constraints under harsh solar light intensity uncertainty conditions, providing investors with an economic optimization bottom line for solar energy absorption equipment investment projects.

[0005] The planning method in patent CN114707783A involves optimizing multiple stages and objective functions, resulting in high computational complexity and a long solution time. The two-stage robust planning method proposed in this patent has practical limitations in predicting and modeling uncertainty, particularly under complex and changing climate conditions. Furthermore, the algorithm's initial values ​​are sensitive, and improper selection of these values ​​can lead to inaccurate optimization results or even solution failure.

[0006] This patent proposes a two-stage robust optimization method for the electricity-gas-heat system, which takes into account the coupling and coordinated optimization of multiple energy forms such as electricity, gas, and heat, which helps to improve energy utilization efficiency and reduce costs. At the same time, the C&CG algorithm is adopted, which combines the opportunity constraint and the convex growth algorithm to improve the efficiency and accuracy of solving the optimization problem. The method of this patent also takes into account the two stages of day-ahead prediction and real-time scheduling. Through the two-stage scheduling strategy, it better handles the uncertainty problem and improves the robustness of the system. Finally, this patent also specifically considers the heat loss of the heating network, making the model closer to the actual operation situation and improving the economic efficiency. Summary of the Invention

[0007] The present invention aims to solve the above problems in the prior art. It proposes a two-stage robust optimization method for regional electric-gas-heat systems based on the C&CG algorithm. The technical solution of the present invention is as follows:

[0008] A two-stage robust optimization method for regional power-gas-heat systems based on the C&CG algorithm includes the following steps:

[0009] (1) Construct a linearized model of the regional electricity-gas-heat system at the distribution network level; the regional electricity-gas-heat system at the distribution network level includes the regional distribution network, regional gas distribution network, and regional heat distribution network;

[0010] (2) A two-stage robust optimization model is constructed considering the real-time extreme scenario with the largest errors in day-ahead and real-time wind power and load forecasting;

[0011] (3) Determine the objective function and constraints of the optimization model;

[0012] (4) Using the C&CG algorithm, the model is decomposed into the main problem MP of unit combination and economic dispatch joint optimization for the day-ahead forecast operation scenario and the real-time extreme operation scenario, and the sub-problem SP of safety detection for the main problem. The problems are solved iteratively to obtain the solution.

[0013] (5) Based on the analysis of the solution results, the optimal strategy for the two-stage robust optimization scheduling of the regional electricity-gas-heat system is obtained.

[0014] Furthermore, the regional distribution network constraints in step (1) include DistFlow branch flow constraints and operation safety constraints applicable to the distribution network;

[0015] The power flow constraint formula is:

[0016]

[0017] The operational safety constraint formula is:

[0018]

[0019] Where, subscript t represents time, Ω DPN is the DPN node set; e(j)(s(j)) is the starting (end) node set of the branch with node j as the end (start) node; P ij,t / Q ij,t and P ij,max / Q ij,max are the active / reactive power and upper limit of the branch ij respectively; I ij,t , I ij,min , I ij,max is the current flowing through branch ij and its lower / upper limit; r ij / x ij is the resistance / reactance of branch ij; is the net injected active / reactive power of node ij; Active power purchased from the main grid / reactive power support provided by the main grid; The active / reactive power output of node j connected to CHP; is the electric power consumed by PtG / WP at node j; is the active / reactive load of node j; U j,t and U j,max / U j,min are the voltage and upper / lower limit values ​​of node j respectively; The predicted value / actual value of wind power active output; Provide reactive power for wind power; and are auxiliary variables, which are the square terms of current and voltage respectively;

[0020] Furthermore, the gas distribution network constraints in step (1) mainly include steady-state pipeline flow constraints, node flow balance equations, node pressure constraint equations, gas vendor capacity constraint equations, and gas storage tank operation constraint equations. The steady-state pipeline flow constraints include the flow-pressure Weymouth equation for passive pipelines and the pressure limit constraint equation for active pipelines.

[0021] Weymouth equation for flow-pressure in a passive pipe:

[0022]

[0023]

[0024] The pressure limit constraint formula of the active pipeline is:

[0025]

[0026] Node traffic balancing formula:

[0027]

[0028] Node pressure constraint formula:

[0029]

[0030] Gas seller capacity constraint:

[0031]

[0032] Gas tank operation constraint type:

[0033]

[0034] Where, and is an auxiliary 0-1 variable: when π m,t ≥π n,t hour, On the contrary, when π m,t ≤π n,t hour, is the average flow rate of the passive pipe mn is the inflow / outflow of the passive pipe mn; is the Weymouth constant of the passive pipe mn, π n,t and πn,min / π n,max is the air pressure at node n and its minimum / maximum limits; π m,t is the air pressure at node m; is the compression ratio of the compressor in the active pipe mn; g(n) is the set of nodes connected to node n; is the net airflow injection amount at node n; and are the CHP gas consumption, load, gas sales volume, gas storage tank gas output / intake and PtG gas production of node n respectively; G mn,max is the flow limit of pipeline mn; e(n) is the set of starting nodes of pipelines with n nodes as the end nodes in the gas network; The minimum / maximum gas sales volume of the gas seller; It is the mutually exclusive state variable of the gas tank's inlet / outlet; is the gas storage capacity and its minimum / maximum limits; The lower / upper limit of the gas inlet / outlet volume of the gas tank; ε mn,t It is in the form of a product of bilinear terms;

[0035] Furthermore, the heat distribution network operation characteristic equation is composed of a hydraulic branch characteristic equation, a temperature balance equation, and a heat balance equation, wherein the hydraulic branch characteristic equation includes a flow continuity equation and a loop pressure drop equation;

[0036] The flow continuity equation is:

[0037]

[0038] The loop pressure drop equation is:

[0039]

[0040] In the formula, the superscript q represents the circuit in the heating network; and are the node-pipeline association matrix and loop-pipeline association matrix of the heat network respectively; m u,t is the hot water flow matrix of node u; h uv,t is the head loss of pipeline uv; x uv and R uv are the length and inner diameter of the pipe uv respectively; v water and ρ water are the flow viscosity and density of water respectively; g is the acceleration due to gravity;

[0041] The heat loss calculation formula of pipeline uv is:

[0042]

[0043] Since the DHN transmission distance is short, the water temperature change in the pipe network is very small, ΔHuv,t Can be approximated as X uv =0, the first-order Taylor formula, that is:

[0044]

[0045] Where, is the ambient temperature of the pipeline uv; ∑R unit H is the thermal resistance per unit length from hot water to the ambient medium; u,t is the heat at node u; and are the supply water temperature and return water temperature at node u respectively; S uv,max and S uv are the maximum flow rate and cross-sectional area of ​​hot water in the pipe uv, respectively; and are the heat output and heat load of the heat source at node u respectively; τ uv,t is an auxiliary 0-1 variable, τ uv,t =1 indicates that the hot water flows from u to v, τ uv,t =0 indicates that the hot water flow direction is v→u; c water is the specific heat capacity of water;

[0046] Furthermore, the step (2) specifically refers to: constructing a two-stage robust optimization model considering the real-time extreme scenario with the largest day-ahead and real-time wind power and load forecast errors;

[0047] The min-max-min optimization model is:

[0048]

[0049] Where: C DA is the operating cost of the entire system; D RT is the operating deviation; H RT (x 0 ,u,x t ,Γ)=0 and G RT (x 0 ,u,x r ,Γ)≤0 are the equality / inequality constraints for the day-ahead and real-time phases respectively; (x 0 ,u) and x r are the strategy sets for the day-ahead and real-time phases, x and u are continuous variables and 0-1 variables, respectively; ζ and γ are the uncertainty factors and their sets in the real-time system, respectively; Φ is the strategy set for the real-time phase; Γ is the robust adjustment parameter used to balance the robustness and economy of the optimization results.

[0050] Furthermore, the step (3) specifically refers to: determining the objective function and constraints of the optimization model;

[0051] The objective function of the optimization model is:

[0052]

[0053] The constraints include linearization constraints for normal operation of REGHS in the day-ahead phase, and rescheduling constraints for DPN, DGN, DHN, and coupling units in the real-time phase:

[0054]

[0055] Where T is the scheduling duration (24h) and the scheduling interval is 1h; and They are electricity purchase price, wind curtailment cost, gas seller’s gas price and gas storage maintenance cost; The superscript r of the variables in the table indicates the value in the real-time stage; and are the uncertainties of load and wind power, respectively, and belong to the uncertainty set γ in the form of a convex polyhedron. and is the corresponding maximum prediction error; For a given u、 feasible decision space under are the electricity / heat output of CHP-a connected to node j of the power grid and node u of the heat network respectively; is the thermoelectric ratio of CHP; is a 0-1 operating state variable representing CHP / PtG; The capacity limitations of CHP and PtG; is the source and load side deviation of the system in the real-time stage; θ RO N is the maximum deviation that the system can accept; EL and N W is the electrical load and the number of fans; is an auxiliary 0-1 variable describing whether the actual load or wind is at the upper or lower boundary; Γ L and T W is the budget value of load and fan output prediction error; U N is the grid reference voltage; is the maximum up / down ramp rate of the CHP-a unit; is the redispatched output ratio of CHP-a / CHP-b units in the real-time phase;

[0056] Furthermore, the step (4) specifically means that the C&CG algorithm can decompose the original problem into MP and SP for iterative solution. The compact forms of MP and SP are:

[0057]

[0058]

[0059] Where, and are the objective functions of MP and SP respectively; is the kth extreme uncertainty factor obtained by SP in the real-time phase The strategy set under Ω k For extreme scene sets; is the solution transferred from MP to sP; c x,t and g x,t are the coefficient matrices corresponding to x respectively; A, B, C, D, F, J, K, L, M and N are the constraint coefficient matrices; b, e, d and f are parameter vectors. An auxiliary 0-1 variable to describe whether the actual load or wind is at the upper or lower boundary;

[0060] Furthermore, the C&CG algorithm is used to decompose the model into a main problem of unit commitment and economic dispatch joint optimization for day-ahead forecast operation scenarios and real-time extreme operation scenarios, and a sub-problem of safety detection for the main problem, and the sub-problems are iteratively solved, specifically including:

[0061] 1) Initialization, let k = 0, Real-time phase offset Assign θ RO ;

[0062] 2) Solve the MP equation to obtain the optimal day-ahead dispatch plan and set the CHP and PTG units’ on / off states u t , and prediction scenarios Pass it to DSP for security testing;

[0063] 3) Step 3: Solve the DSP formula and update the maximum offset under the current scheduling scheme And the most extreme wind load scenario

[0064] 4) Determine the current scenario Is it true? If so, output the current day-ahead scheduling plan as the optimal plan: Otherwise, set k=k+1 and return to step 2).

[0065] The advantages and beneficial effects of the present invention are as follows:

[0066] The present invention proposes a two-stage robust optimization method for regional electricity-gas-heat systems based on the C&CG algorithm. Through two-stage optimization, the optimal scheduling scheme can be found under the day-ahead prediction scenario, while taking into account the operating deviations under real-time extreme scenarios, thereby improving the economy and safety of the system; a linearized model of the regional electricity-gas-heat system at the distribution network level is constructed, which includes detailed constraints and characteristic equations of the regional distribution network, regional gas distribution network and regional heating distribution network. This modeling method allows for more accurate simulation and optimization of the operation of the entire system, including the mutual conversion and supply of electricity, natural gas and thermal energy; the day-ahead prediction and real-time extreme scenarios are simultaneously considered in the optimization model, which can more comprehensively deal with the volatility and uncertainty of renewable energy such as wind power. This dual consideration enables the system to respond more flexibly to various situations in actual operation and reduce the impact of prediction errors. The objective function and constraints of the optimization model are clarified, including electricity purchase costs, wind curtailment penalty costs, gas purchase costs, and gas storage loss costs, which helps to maximize cost-effectiveness while ensuring the stable and safe operation of the system. The C&CG algorithm decomposes the original problem into a main problem (MP) and subproblems (SP), and obtains the optimal strategy through iterative solution. This decomposition and iteration method improves the convergence and accuracy of the algorithm, making the solution process more efficient.

[0067] This invention decomposes the complex multi-energy system optimization problem into manageable subproblems and effectively handles uncertainty through a two-stage robust optimization approach. Furthermore, the algorithm design allows for the identification of the lowest-cost operational strategy while ensuring safe system operation, which is highly valuable in practical energy system management. This approach can improve wind power utilization, reduce wind curtailment, and enhance the operational efficiency and economic efficiency of the entire energy network. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 It is a flow chart of a method according to a preferred embodiment of the present invention.

[0069] Figure 2 It is a schematic diagram of the structure of the regional-level electricity-gas-heat system described in the present invention.

[0070] Figure 3 It is a flow chart of the C&CG algorithm adopted by the method of the present invention. DETAILED DESCRIPTION

[0071] The following will describe the technical solutions in the embodiments of the present invention in detail with reference to the accompanying drawings. The described embodiments are only a part of the embodiments of the present invention.

[0072] The technical solution of the present invention to solve the above technical problems is:

[0073] The overall framework flow chart of the method of the present invention is as follows Figure 1 As shown, a two-stage robust optimization method for a regional electric-gas-heat system based on the C&CG algorithm is provided. The method comprises the following steps in sequence:

[0074] (1) Construct a linearized model of the regional electricity-gas-heat system at the distribution network level;

[0075] (2) A two-stage robust optimization model is constructed considering the real-time extreme scenario with the largest errors in day-ahead and real-time wind power and load forecasting;

[0076] (3) Determine the objective function and constraints of the optimization model;

[0077] (4) Using the C&CG algorithm, the model is decomposed into the main problem (MP) of unit combination and economic dispatch joint optimization for the day-ahead forecast operation scenario and the real-time extreme operation scenario, and the sub-problem (SP) of safety detection for the main problem. The results are solved iteratively to obtain the solution.

[0078] (5) Based on the analysis of the solution results, the optimal strategy for the two-stage robust optimization scheduling of the regional electricity-gas-heat system is obtained.

[0079] Step (1) specifically involves constructing a linearized model of a regional power-gas-heat system at the distribution network level, where the regional power-gas-heat system includes a regional distribution network, a regional gas distribution network, and a regional heat distribution network. The regional distribution network constraints include DistFlow branch flow constraints and operational safety constraints applicable to the distribution network.

[0080] The power flow constraint formula is:

[0081]

[0082]

[0083] The operational safety constraint formula is:

[0084]

[0085] Where, subscript t represents time, Ω DPN is the DPN node set; e(j)(s(j)) is the starting (end) node set of the branch with node j as the end (start) node; P ij,t / Q ij,t and P ij,max / Q ij,max are the active / reactive power and upper limit of the branch ij respectively; I ij,t , I ij,min , I ij,max is the current flowing through branch ij and its lower / upper limit; r ij / xij is the resistance / reactance of branch ij; is the net injected active / reactive power of node ij; Active power purchased from the main grid / reactive power support provided by the main grid; The active / reactive power output of node j connected to CHP; is the electric power consumed by PtG / WP at node j; is the active / reactive load of node j; U j,t and U j,max / U j,min are the voltage and upper / lower limit values ​​of node j respectively; The predicted value / actual value of wind power active output; Provide reactive power for wind power; and are auxiliary variables, which are the square terms of current and voltage respectively;

[0086] The gas distribution network constraints mainly include steady-state pipeline flow constraints (including the flow-pressure Weymouth equation for passive pipelines and the pressure limit constraint for active pipelines), node flow balance equations, node pressure constraint equations, gas seller capacity constraint equations, and gas storage tank operation constraint equations.

[0087] Weymouth equation for flow-pressure in a passive pipe:

[0088]

[0089]

[0090] The pressure limit constraint formula of the active pipeline is:

[0091]

[0092] Node traffic balancing formula:

[0093]

[0094] Node pressure constraint formula:

[0095]

[0096] Gas seller capacity constraint:

[0097]

[0098] Gas tank operation constraint type:

[0099]

[0100] Where, and is an auxiliary 0-1 variable: when π m,t ≥π n,t hour, On the contrary, when π m,t ≤π n,t hour, is the average flow rate of the passive pipe mn is the inflow / outflow of the passive pipe mn; is the Weymouth constant of the passive pipe mn, π n,t and π n,min / π n,max is the air pressure at node n and its minimum / maximum limits; π m,t is the air pressure at node m; is the compression ratio of the compressor in the active pipe mn; g(n) is the set of nodes connected to node n; is the net airflow injection amount at node n; and are the CHP gas consumption, load, gas sales volume, gas storage tank gas output / intake and PtG gas production of node n respectively; G mn,max is the flow limit of pipeline mn; e(n) is the set of starting nodes of pipelines with n nodes as the end nodes in the gas network; The minimum / maximum gas sales volume of the gas seller; It is the mutually exclusive state variable of the gas tank's inlet / outlet; is the gas storage capacity and its minimum / maximum limits; The lower / upper limit of the gas inlet / outlet volume of the gas tank; ε mn,t It is in the form of a product of bilinear terms;

[0101] The heat distribution network operating characteristic equation consists of the hydraulic branch characteristic equation, the temperature balance equation, and the heat balance equation. The hydraulic branch characteristic equation includes the flow continuity equation and the loop pressure drop equation.

[0102] The flow continuity equation is:

[0103]

[0104] The loop pressure drop equation is:

[0105]

[0106] In the formula, the superscript q represents the circuit in the heating network; and are the node-pipeline association matrix and loop-pipeline association matrix of the heat network respectively; m u,t is the hot water flow matrix of node u; h uv,t is the head loss of pipeline uv; x uvand R uv are the length and inner diameter of the pipe uv respectively; v water and ρ water are the flow viscosity and density of water (997kg / m 3 ); g is the acceleration due to gravity;

[0107] The heat loss calculation formula of pipeline uv is:

[0108]

[0109] Since the DHN transmission distance is short, the water temperature change in the pipe network is very small, ΔH uv,t Can be approximated as X uv =0, the first-order Taylor formula, that is:

[0110]

[0111] Where, is the ambient temperature of the pipeline uv; ∑R unit H is the thermal resistance per unit length from hot water to the ambient medium. u,t is the heat at node u; and are the supply water temperature (including the heat source outlet water temperature) and return water temperature at node u; S uv,max and S uv are the maximum flow rate and cross-sectional area of ​​hot water in the pipe uv, respectively; and are the heat output and heat load of the heat source at node u respectively; τ uv,t is an auxiliary 0-1 variable, τ uv,t =1 indicates that the hot water flows from u to v, τ uv,t =0 indicates that the hot water flow direction is v→u; c water is the specific heat capacity of water;

[0112] The step (2) specifically refers to: constructing a two-stage robust optimization model by considering the real-time extreme scenario with the largest day-ahead and real-time wind power and load forecast errors.

[0113] The min-max-min optimization model is:

[0114]

[0115] Where: C DA is the operating cost of the entire system; D RT is the operating deviation; H RT (x 0 ,u,x t ,Γ)=0 and G RT (x 0 ,u,x r,Γ)≤0 are the equality / inequality constraints for the day-ahead and real-time phases respectively; (x 0 ,u) and x r are the strategy sets for the day-ahead and real-time phases, x and u are continuous variables and 0-1 variables, respectively; ζ and γ are the uncertainty factors and their sets in the real-time system, respectively; Φ is the strategy set for the real-time phase; Γ is the robust adjustment parameter used to balance the robustness and economy of the optimization results;

[0116] The step (3) specifically refers to: determining the objective function and constraint conditions of the optimization model.

[0117] The objective function of the optimization model is:

[0118]

[0119] The constraints include linearization constraints for normal operation of REGHS in the day-ahead phase, and rescheduling constraints for DPN, DGN, DHN, and coupling units in the real-time phase:

[0120]

[0121] Where T is the scheduling duration (24h) and the scheduling interval is 1h; and They are electricity purchase price, wind curtailment cost, gas seller’s gas price and gas storage maintenance cost; The superscript r of the variables in the table indicates the value in the real-time stage; and are the uncertainties of load and wind power, respectively, and belong to the uncertainty set γ in the form of a convex polyhedron. and is the corresponding maximum prediction error; For a given u、 feasible decision space under are the electricity / heat output of CHP-a connected to node j of the power grid and node u of the heat network respectively; is the thermoelectric ratio of CHP; is a 0-1 operating state variable representing CHP / PtG; The capacity limitations of CHP and PtG; is the source and load side deviation of the system in the real-time stage; θ RO N is the maximum deviation that the system can accept; EL and N W is the electrical load and the number of fans; is an auxiliary 0-1 variable describing whether the actual load or wind is at the upper or lower boundary; Γ L and T W is the budget value of load and fan output prediction error; UN is the grid reference voltage; is the maximum up / down ramp rate of the CHP-a unit; is the redispatched output ratio of CHP-a / CHP-b units in the real-time phase;

[0122] The step (4) specifically means that the C&CG algorithm can decompose the original problem into MP and SP for iterative solution. The compact forms of MP and SP are:

[0123]

[0124]

[0125] Where, and are the objective functions of MP and SP respectively; is the kth extreme uncertainty factor obtained by SP in the real-time phase The strategy set under Ω k For extreme scene sets; is the solution transferred from MP to sP; c x,t and g x,t are the coefficient matrices corresponding to x respectively; A, B, C, D, F, J, K, L, M and N are the constraint coefficient matrices; b, e, d and f are parameter vectors; An auxiliary 0-1 variable to describe whether the actual load or wind is at the upper or lower boundary;

[0126] The C&CG algorithm flow chart used in the method of the present invention is as follows: Figure 3 As shown in the figure, the C&CG algorithm is used to decompose the model into the main problem of unit commitment and economic dispatch joint optimization for the day-ahead forecast operation scenario and the real-time extreme operation scenario, and the sub-problem of safety detection of the main problem. The sub-problems are solved iteratively, including:

[0127] 1) Initialization, let k = 0, Real-time phase offset Assign θ RO ;

[0128] 2) Solve the MP equation to obtain the optimal day-ahead dispatch plan and set the CHP and PTG units’ on / off states u t , and prediction scenarios Pass it to DSP for security testing;

[0129] 3) Step 3: Solve the DSP formula and update the maximum offset under the current scheduling scheme And the most extreme wind load scenario

[0130] 4) Determine the current scenario Is it true? If so, output the current day-ahead scheduling plan as the optimal plan: Otherwise, set k=k+1 and return to step 2.

[0131] This paper utilizes the C&CG algorithm, establishes a linearized multi-energy flow model, and applies two-stage robust optimization theory to construct an optimization framework that can adapt to both day-ahead and real-time scheduling requirements. This two-stage robust optimization effectively addresses forecast errors, optimizes system operation strategies, and improves system economics and safety, thereby optimizing the overall performance of the integrated energy system and ensuring user satisfaction.

[0132] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions.

[0133] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0134] The above embodiments should be understood as merely illustrating the present invention and not as limiting the scope of protection of the present invention. After reading the contents of the present invention, technicians may make various changes or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims

1. A two-stage robust optimization method for regional electricity-gas-heat system based on C&CG algorithm, characterized in that: The following steps are involved: (1) Construct a linearized model of the regional electricity-gas-heat system at the distribution network level; the regional electricity-gas-heat system at the distribution network level includes the regional distribution network, the regional gas distribution network, and the regional heat distribution network; (2) A two-stage robust optimization model is constructed by considering the real-time extreme scenario with the largest errors in day-ahead and real-time wind power and load forecasting; (3) Determine the objective function and constraints of the optimization model; (4) The C&CG algorithm is used to decompose the model into the main problem MP of unit combination-economic dispatch joint optimization for the day-ahead forecast operation scenario and the real-time extreme operation scenario and the sub-problem SP of safety detection of the main problem, and the solution is obtained by iterative solution. (5) Based on the analysis of the solution results, the optimal strategy for the two-stage robust optimization scheduling of the regional electricity-gas-heat system is obtained.

2. The two-stage robust optimization method for regional electric-gas-heat systems based on the C&CG algorithm according to claim 1 is characterized in that: The regional distribution network constraints in step (1) include DistFlow branch flow constraints and operation safety constraints applicable to the distribution network; The power flow constraint formula is: The operational safety constraint formula is: In the formula, subscript t represents time, Ω DPN is the DPN node set; e(j)(s(j)) is the starting (end) node set of the branch with node j as the end (start) node; P ij,t / Q ij,t and P ij,max / Q ij,max are respectively the active / reactive power and upper limit value at the beginning of branch ij; I ij,t ,I ij,min ,I ij,max is the current flowing through branch ij and the lower / upper limit; r ij / x ij is the resistance / reactance of branch ij; is the net injected active / reactive power of node ij; Active power purchased by the main grid / reactive power support provided by the main grid; The active / reactive power output of node j connected to CHP; is the electrical power consumed by PtG / WP at node j; is the active / reactive load of node j; U j,t and U j,max / U j,min are the voltage and upper / lower limit values ​​of node j respectively; It is the predicted value / actual value of wind power active output; Provide reactive power for wind power; and are auxiliary variables, which are the square terms of current and voltage respectively.

3. The two-stage robust optimization method for regional electric-gas-heat systems based on the C&CG algorithm according to claim 1 is characterized in that: The gas distribution network constraints in step (1) mainly include steady-state pipeline flow constraints, node flow balance formula, node pressure constraint formula, gas seller capacity constraint formula and gas tank operation constraint formula, and the steady-state pipeline flow constraints include the flow-pressure Weymouth equation of the passive pipeline and the pressure limit constraint formula of the active pipeline; The flow-pressure Weymouth equation for a passive pipe is: Pressure limit constraint formula for active pipeline: Node traffic balancing formula: Node pressure constraint formula: Gas seller capacity constraint: Gas tank operation restraint type: In the formula, and is an auxiliary 0-1 variable: when π m,t ≥π n,t hour, On the contrary, when π m,t ≤π n,t hour, is the average flow rate of the passive pipe mn is the inflow / outflow of the passive pipe mn; is the Weymouth constant for the passive pipe mn, π n,t and π n,min / π n,max is the air pressure at node n and its minimum / maximum limits; π m,t is the air pressure at node m; is the compression ratio of the compressor in the active pipeline mn; g(n) is the set of nodes connected to node n; is the net airflow injection amount at node n; and G are the CHP gas consumption, load, gas sales volume, gas tank gas output / intake volume and PtG gas production of node n respectively; mn,max is the flow limit of pipeline mn; e(n) is the set of starting nodes of pipelines with n nodes as the terminal nodes in the gas network; The minimum / maximum gas sales volume of the gas seller; It is the mutually exclusive state variable of the gas tank's air intake / exhaust; is the gas storage capacity and its minimum / maximum limits; The lower / upper limit of the gas inlet / outlet volume of the gas storage tank; ε mn,t It is in the form of a product of bilinear terms.

4. The two-stage robust optimization method for regional electric-gas-heat systems based on the C&CG algorithm according to claim 1 is characterized in that: The heat distribution network operation characteristic equation is composed of a hydraulic branch characteristic equation, a temperature balance equation and a heat balance equation, and the hydraulic branch characteristic equation includes a flow continuity equation and a loop pressure drop equation; The flow continuity equation is: The loop pressure drop equation is: In the formula, the superscript q represents the circuit in the heating network; and are the node-pipeline association matrix and loop-pipeline association matrix of the heat network respectively; m u,t is the hot water flow matrix of node u; h uv,t is the head loss of pipeline uv; x uv and R uv are the length and inner diameter of the pipe uv; v water and ρ water are the flow viscosity and density of water respectively; g is the acceleration due to gravity; The heat loss calculation formula of pipeline uv is: Since the DHN transmission distance is short, the water temperature in the pipe network changes very little, ΔH uv,t Can be approximated as X uv =0, the first-order Taylor formula, that is: In the formula, is the ambient temperature of the pipeline uv; ∑R unit H is the thermal resistance per unit length from hot water to the ambient medium; u,t is the heat at node u; and are the supply water temperature and return water temperature at node u respectively; S uv,max and S uv are the maximum flow rate and cross-sectional area of ​​hot water in the pipe uv respectively; and are the heat output and heat load of the heat source at node u respectively; τ uv,t is an auxiliary 0-1 variable, τ uv,t =1 indicates that the hot water flows from u to v, τ uv,t =0 indicates that the hot water flows in the direction of v→u; c water is the specific heat capacity of water.

5. The two-stage robust optimization method for regional electric-gas-heat systems based on the C&CG algorithm according to claim 1 is characterized in that: The step (2) specifically refers to: constructing a two-stage robust optimization model by considering the real-time extreme scenario with the largest day-ahead and real-time wind power and load forecast errors; The min-max-min optimization model is: Where: C DA is the operating cost of the entire system; D RT is the running deviation; H RT (x 0 ,u,x t ,Γ)=0 and G RT (x 0 ,u,x r ,Γ)≤0 are the equality / inequality constraints for the day-ahead and real-time stages respectively; (x 0 ,u) and x r are the strategy sets for the day-ahead and real-time stages respectively; x and u are continuous variables and 0-1 variables respectively; ζ and γ are the uncertainty factors and their sets in the real-time system respectively; Φ is the real-time stage strategy set; Γ is the robust adjustment parameter, which is used to weigh the robustness and economy of the optimization results.

6. The two-stage robust optimization method for regional electric-gas-heat systems based on the C&CG algorithm according to claim 1 is characterized in that: The step (3) specifically refers to: determining the objective function and constraint conditions of the optimization model; The objective function of the optimization model is: The constraints include linearization constraints for normal operation of REGHS in the day-ahead phase, and rescheduling constraints for DPN, DGN, DHN and coupling units in the real-time phase: In the formula, T is the scheduling duration (24h), and the scheduling interval is 1h; and They are electricity purchase price, wind curtailment cost, gas seller’s gas price and gas storage maintenance cost; The superscript r of the variables in the table indicates the value in the real-time stage; and are the uncertainties of load and wind power, respectively, and belong to the uncertainty set γ in the form of a convex polyhedron. and is the corresponding maximum prediction error; For a given u、 feasible decision space under are the electricity / heat output of CHP-a connected to the grid node j and the heat network node u respectively; is the thermal-to-electric ratio of CHP; is a 0-1 operating state variable representing CHP / PtG; The capacity limitations of CHP and PtG; is the source and load side deviation of the system in the real-time stage; θ RO N is the maximum deviation that the system can accept; EL and N W is the number of electrical loads and fans; An auxiliary 0-1 variable describing whether the actual load or wind is at the upper or lower boundary; Γ L and T W is the budget value of load and fan output prediction error; U N is the grid reference voltage; is the maximum up / down ramp rate of the CHP-a unit; It is the redispatching output ratio of CHP-a / CHP-b units in the real-time stage.

7. The two-stage robust optimization method for regional electric-gas-heat systems based on the C&CG algorithm according to claim 1 is characterized in that: The step (4) specifically means that the C&CG algorithm can decompose the original problem into MP and SP for iterative solution. The compact forms of MP and SP are: In the formula, and are the objective functions of MP and SP respectively; is the kth extreme uncertainty factor obtained by SP in the real-time stage The strategy set under Ω k For extreme scene sets; is the solution transmitted from MP to sP; c x,t and g x,t are the coefficient matrices corresponding to x respectively; A, B, C, D, F, J, K, L, M and N are the constraint coefficient matrices; b, e, d and f are parameter vectors; An auxiliary 0-1 variable to describe whether the actual load or wind is at the upper or lower boundary.

8. The two-stage robust optimization method for regional electric-gas-heat systems based on the C&CG algorithm according to claim 1 is characterized in that: The C&CG algorithm is used to decompose the model into a main problem of unit combination-economic dispatch joint optimization for day-ahead forecast operation scenarios and real-time extreme operation scenarios and a sub-problem of safety detection of the main problem for iterative solution, specifically including: 1) Initialization, set k = 0, Real-time phase offset Assign θ RO ; 2) Solve the MP equation to obtain the optimal day-ahead dispatching plan and set the CHP and PTG units’ on / off states u t , and the prediction scenario Pass it to DSP for security testing; 3) Step 3: Solve the DSP formula and update the maximum offset under the current scheduling scheme And the most extreme wind load scenario 4) Determine the current scenario Is it true? If so, output the current day-ahead scheduling plan as the optimal plan: Otherwise, set k=k+1 and return to step 2).

Images