Robust optimization scheduling method and device for integrated energy system based on regulatable capability

Abstract

The invention belongs to the technical field of energy optimization scheduling, and particularly relates to a comprehensive energy system robust optimization scheduling method and device based on regulatable ability, and the method comprises the steps: building a natural gas supply system management and storage model; establishing a power distribution network time shift regulation and control load model, a natural gas network time shift regulation and control load model and a power distribution network demand response load model; constructing an optimization objective function considering the operation cost of the power distribution network and the operation cost of the natural gas network; establishing a comprehensive energy system robust optimization model considering the load uncertainty based on the optimization objective function; and constructing a double-layer robust optimization algorithm, solving the worst operation scene by the sub-problem, optimizing the operation cost of the integrated energy system by the main problem, and solving to obtain an optimal scheduling scheme of the integrated energy system. According to the invention, the scheduling strategy can effectively deal with the complex operation problem of the power-gas coupling integrated energy system, and the flexibility and adaptability of the system are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of energy optimization and scheduling, and in particular relates to a robust optimization and scheduling method and device for an integrated energy system based on controllable capabilities. Background Art

[0002] As a complex network of multi-energy coupling, the Integrated Energy System (IES) has become a core technological path for achieving efficient energy utilization and low-carbon transformation through the coordinated optimization and complementarity of multiple energy sources such as electricity, heat, gas, and cooling. In recent years, with the increase in renewable energy penetration and the diversification of load demands, research on the optimized operation of IES has gradually expanded from a deterministic framework to uncertainty modeling, and attempts have been made to integrate demand-side resources to enhance system flexibility. However, existing research still has significant shortcomings: on the one hand, there is a lack of active management and coordinated optimization of controllable loads; on the other hand, load uncertainty modeling often focuses on a single energy form or simplified scenarios, making it difficult to characterize the complexity of the dynamic response of the demand side under multi-energy coupling.

[0003] For coupled electricity-hydrogen-heat systems, the paper "Research on Optimal Scheduling of Integrated Energy Microgrids Considering Multiple Uncertainties" uses a heat recovery model to improve energy efficiency. However, its optimization framework is still based on forecast data and does not quantify the impact of load fluctuations on system economics and robustness. Similarly, the paper "Source-Load Low-Carbon Economic Scheduling Method for Hydrogen-Containing Multi-Energy Systems Based on Carbon-Green Certificate Mutual Recognition and Flexible Electric and Thermal Loads" introduces a coupled device model, but load demand is often assumed to be rigid, failing to consider the dynamic adjustment potential of demand response strategies such as interruptible and shiftable loads. Although stochastic optimization and robust optimization methods have been partially applied to IES uncertainty analysis, their modeling focuses on wind and solar power output fluctuations, with insufficient attention paid to the spatiotemporal heterogeneity of user energy consumption behavior and the randomness of load responses. The paper "Master-Slave Game-Based Robust Optimal Scheduling of Integrated Energy Systems in Parks Considering Multi-Scenario Collaborative Carbon Reduction" addresses source-side uncertainties through a data-driven distributed robust optimization method, but fails to incorporate flexible resources on the demand side into the collaborative optimization framework, resulting in a scheduling solution that is difficult to adapt to real-time load fluctuations.

[0004] Regarding the integration of controllable loads, existing research has initially explored the role of price-based demand response in regulating load curves, but this is often limited to single energy sources or simple elasticity models. The papers "Master-Slave Game Strategy for Virtual Power Plants Based on Time-of-Use Electricity-Carbon Coupling Pricing" and "Two-Tier Optimization of Virtual Power Plants Considering Source-Load Coordinated Response and Dynamic Pricing" optimize electricity load curves through time-of-use electricity pricing strategies, but do not address the coordinated response mechanisms of thermal and gas loads. Furthermore, the load models ignore differences in user behavior and equipment operating constraints. Furthermore, modeling the dynamic characteristics of typical controllable loads such as electric vehicles and smart homes is still in its infancy. Existing research often employs fixed charging and discharging patterns or idealized response assumptions, making it difficult to reflect the interactive effects of user preferences and equipment physical constraints in real-world scenarios. This lack of modeling can lead to optimization results being invalidated in actual operation due to load response deviations, limiting the potential for synergy across the entire "source-grid-load-storage" chain within IES.

[0005] In summary, the existing schemes do not realize the research on the robust optimization scheduling method of the integrated energy system considering the controllable capacity. Therefore, it is necessary to study the robust optimization scheduling method of the integrated energy system considering the controllable capacity. Summary of the Invention

[0006] In view of the above deficiencies in the existing technology, the purpose of the present invention is to provide a robust optimization scheduling method and device for an integrated energy system based on controllable capabilities, so that the scheduling strategy can effectively cope with the complex operation problems of the electric-gas coupled integrated energy system and improve the flexibility and adaptability of the system.

[0007] To achieve the above objectives, the present invention provides a robust optimization scheduling method for an integrated energy system based on controllable capabilities, comprising the following steps: S1. Taking the natural gas supply system and natural gas storage device as the research objects, a natural gas supply system pipeline storage model is established, and the gas pressure constraints of the natural gas pipeline nodes, pipeline operation constraints, gas supply constraints of the natural gas gate station, and gas filling and release constraints of the natural gas storage device are established; S2. Taking the controllability of the integrated energy system including the distribution network and the natural gas network as the research object, a time-shiftable load model for the distribution network, a time-shiftable load model for the natural gas network, and a demand response load model for the distribution network are established; S3. Taking the operating cost of the integrated energy system as the research target, an optimization objective function is constructed that takes into account the operating costs of the distribution network and the natural gas network; S4. Based on the distribution network operation constraints, energy storage constraints, natural gas gate station adjustment constraints and the constraints in S1, a robust optimization model of the integrated energy system considering load uncertainty is established based on the optimization objective function in S3; S5. Construct a two-layer robust optimization algorithm. The sub-problem solves the worst operating scenario, and the main problem optimizes the operating cost of the integrated energy system, and the optimal scheduling plan for the integrated energy system is obtained.

[0008] As a preferred embodiment of the present invention, in S1, the process of establishing the pipeline storage model of the natural gas supply system is as follows: In the natural gas supply system, the pipeline storage state of the natural gas network is described based on the Weymouth equation: ; Where, is the average natural gas flow rate of natural gas pipeline mn during period t; and are the gas pressures at nodes m and n of the natural gas pipeline during period t; is the Weymouth constant; The flow balance of the natural gas network must meet the following requirements: ; Where, is the set of natural gas pipelines connected to node n; is the set of natural gas storage devices connected to node n; is the set of natural gas gate stations connected to node n; is the set of natural gas loads connected to node n; is the set of gas turbines connected to node n; and are the gas flow rates at the end and head end of the natural gas pipeline mn during period t; and are the gas release and gas intake of the natural gas storage device s during period t, respectively; is the natural gas flow rate at the natural gas gate station c during period t; is the gas flow of natural gas load l connected to node n during period t; is the gas flow of the gas turbine g connected to the node n during period t; A linear pipeline storage model is used to describe the dynamic effects of the natural gas pipeline and to establish the pipeline storage model of the natural gas supply system. The expression is as follows: ; ; ; ; Where, 、 are the pipeline stocks of natural gas pipeline mn in periods t and t-1 respectively; is the average pressure of natural gas pipeline mn; is the length of the natural gas pipeline mn.

[0009] As a preferred solution of the present invention, in S1, the gas pressure constraint of the natural gas pipeline node is: ; Where, 、 、 are the gas pressure and its upper and lower limits at the natural gas pipeline node during period t; The pipeline operation constraints are: ; Where, 、 are the upper and lower limits of the pipeline inventory of natural gas pipeline mn in period t respectively; The gas supply constraints of natural gas gate stations are: ; Where, 、 are the upper and lower limits of the natural gas flow rate at the natural gas gate station c during period t; The gas charging and releasing constraints of natural gas storage devices are: ; ; ; ; ; ; Where, 、 are the gas storage capacity of the natural gas storage device s at time periods t and t-1 respectively; 、 are binary variables, representing the inflation and degassing states of the gas storage device during period t; 、 are the charging and releasing efficiencies of the natural gas storage device s, respectively; 、 are the maximum and minimum gas release amounts of the gas storage device s during period t, respectively; 、 are the maximum and minimum air filling volumes of the air storage device s during period t respectively.

[0010] As a preferred solution of the present invention, in S2, there is a time-shiftable load inside the distribution network, and the time-shiftable load control model of the distribution network is: ; ; Where, is the total load that can be time-shifted in a day; is the access amount of time-shiftable load in period t; 、 are the upper and lower limits of the load during time period t, respectively; T is the number of time periods, with a value of 24, representing one day; There are time-shiftable loads in the natural gas network. The time-shiftable load control model of the natural gas network is: ; ; Where, is the total load of the time-shiftable gas load in a day; is the access capacity of the time-shiftable gas load in period t; 、 are the upper and lower limits of the gas load during period t respectively; The distribution network demand response load model is: ; ; Where, represents the load reduction level of the zth curtailable load in period t; represents the maximum load reduction power during period t; 、 、 They represent the total amount of electric energy that can be cut for the zth load and its upper and lower limits respectively.

[0011] As a preferred solution of the present invention, in S3, the optimization objective function considering the distribution network operating costs and the natural gas network operating costs is: ; Where F is the operating cost of the integrated energy system; The operating cost of the distribution network; For the running costs of the gas network; The calculation method is: ; Where, The cost of purchasing electricity from the distribution network; Subsidy for time-shifting loads; Subsidy charges for demand response loads; The calculation method is: ; Where, Gas purchase costs for the natural gas network; Subsidy for time-shifting load on the natural gas network.

[0012] As a preferred solution of the present invention, in S4, the distribution network operation constraints are: ; ; ; ; ; ; Where, is the square of the voltage amplitude at node i in the distribution network; 、 are the upper and lower limits of the voltage amplitude of the distribution network nodes respectively; is the square of the current amplitude flowing through branch ij; is the upper limit of the square of the current amplitude flowing through branch ij; is the active power flowing through branch ij during period t; is the number of the gas turbine; is the number of gas turbines; is a binary variable indicating whether the gas turbine is installed at the distribution network node j; For gas turbines The power value; is the resistance of branch ij; is the active load of node j connected to the distribution network during period t; is the active power flowing through branch ju during period t; Indicates that distribution network node u is connected to distribution network node j; is the number of the adjustable load; is the number of adjustable loads; is a binary variable, indicating whether the controllable load is connected to the distribution network node j; The load that can be controlled during period t The value of the charging load. The adjustable load includes time-shiftable adjustable load, energy storage load, and demand response load; is the reactive power flowing through branch ij during period t; is the reactance of branch ij; is the reactive load of node j connected to the distribution network during period t; Indicates that distribution network node h is connected to distribution network node j; is the reactive power flowing through branch jh during period t; is the square of the voltage amplitude at node j in the distribution network; The energy storage constraint is: ; ; ; ; ; Where, 、 are the remaining energy storage capacity in periods t and t-1 respectively; The charging and discharging efficiency of energy storage; 、 are binary variables representing the charge and discharge states of energy storage; 、 They are the upper and lower limits of the remaining energy storage capacity respectively; 、 、 are the energy storage charging power and its upper and lower limits in period t respectively; 、 、 are the energy storage discharge power and its upper and lower limits in period t respectively; The natural gas gate station adjustment constraints include natural gas supply flow constraints, gas supply quantity adjustment constraints, gas supply quantity adjustment frequency constraints, and gas supply flow limit constraints, which are: ; ; ; ; ; ; ; Where, is the natural gas flow rate of natural gas gate station c during period t-1; 、 are the minimum and maximum gas supply adjustment amounts of natural gas gate station c respectively; 、 are binary variables representing the increase in gas supply at the natural gas gate station during periods t and t-1, respectively; 、 are binary variables for the reduction of gas supply by the natural gas gate station during periods t and t-1, respectively; The maximum number of times the gas supply volume can be adjusted for a natural gas gate station.

[0013] As a preferred solution of the present invention, in S4, the process of establishing the robust optimization model of the integrated energy system considering load uncertainty is as follows: During the operation of the integrated energy system, both the electric load and the gas load are uncertain. The uncertainty of the load is defined using a boxed uncertainty set. For the distribution network load: ; ; Where, is the load of the distribution network during period t; is the predicted value of distribution network load in period t; is the fluctuation deviation of distribution network load; The regulation parameters introduced to deal with the uncertainty of distribution network load; For natural gas network load: ; ; Where, is the load of the natural gas network during period t; is the predicted value of the natural gas network load during period t; is the fluctuation deviation of natural gas network load; Adjustment parameters introduced to account for uncertainty in natural gas network load; The optimization objective function is transformed into a robust optimization scheduling model, and a robust optimization model of the integrated energy system considering load uncertainty is obtained: .

[0014] As a preferred solution of the present invention, in S5, the process of constructing a two-layer robust optimization algorithm is: After considering the uncertain set of distribution network loads and natural gas network loads, the solution of the optimization objective function changes from a one-stage process to a two-stage process. In order to more effectively explain the solution process and solution method, a compact model is designed as follows: ; ; ; ; ; ; Where x is the optimization variable in the first stage, i.e., the time-of-use charging price of each controllable load; the optimization variables in the second stage are v and y, i.e., the optimized solutions of the load uncertainty variables and the distribution network flow; V is the set of load uncertainty variables; and Y is the set of optimized solutions of the distribution network flow. and is the coefficient matrix of the optimization objective function, the superscript T indicates transposition; d, o, a are constant column vectors; A, H, B, D, E, , G, is the sparse matrix of constraints corresponding to different optimization stages; Decompose the original problem into a main problem and sub-problems. The specific form of the main problem is: Objective function: ; Where, is the solution to the subproblem; Constraints: ; ; ; ; ; ; ; Where k is the index of the number of iterations; is the maximum number of iterations; is y at the kth iteration; is v at the kth iteration; The specific form of the sub-problem is: Objective function: ; Where, is a set of solutions for the variable x; Constraints: ; ; Given a set of uncertainty variables v, the inner min problem becomes a second-order cone programming problem. According to the strong duality theory, it is transformed into a max form of a dual problem and combined with the outer max problem to obtain the following form: Objective function: ; Where, is the dual variable in the subproblem; Constraints: ; ; ; Where, Number the norm constraint coefficient matrix; is the number of norm constraints; For the The coefficient matrix of the norm constraint; constrain the dual variable for the norm of the dual variable; The dual variable The coefficient matrix of the norm constraint; 、 is the dual variable; represents the L2 norm; b is a constant vector; Since the objective function in Since it is a bilinear problem, the binary expansion method and the big M method are used for linearization, and the subproblem is reformulated as a mixed integer second-order cone programming problem: ; Where M is the constant in the big M method; is the dual variable corresponding to the part with uncertain variables; is the upper bound of the dual variable; is the lower bound of the dual variable; is an auxiliary variable, which is a binary variable; is an auxiliary variable, which is a binary variable; is the uncertainty adjustment parameter.

[0015] As a preferred solution of the present invention, in S5, the solution process is as follows: S5.1. Given a set of v values ​​as the initial worst-case scenario, set the lower bound of the operating cost to , the upper bound of the operating cost is , the number of iterations is k=1; S5.2. Based on the worst scenario Solve the main problem and obtain the optimal solution of the main problem at the kth iteration and , obtained from the main problem value as the new lower bound, ; S5.3. Substitute the solution of the main problem into the subproblem and obtain the optimal solution of the subproblem at the kth iteration and the worst-case scenario , update the upper bound ; S5.4. Settings is the preset convergence judgment margin, if , it indicates that the optimization calculation has reached the optimal solution and the iteration stops; otherwise, add variables and the following constraints: ; ; ; ; Where, is y at the k+1th iteration; S5.5. Set k = k + 1, return to S5.2 and continue iterating until the algorithm converges.

[0016] A robust optimization scheduling device for an integrated energy system based on controllable capabilities includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor. The above method is implemented by executing the computer program through the processor.

[0017] The beneficial effects of the present invention are: Optimize scheduling flexibility: This invention establishes a natural gas supply system pipeline storage model, a power grid time-shiftable load model, a gas network time-shiftable load model, and a power grid demand response load model. It comprehensively considers the operating constraints of the natural gas supply system and gas storage device, as well as the controllable capabilities of the power grid and gas network, so that the scheduling strategy can effectively cope with the complex operating problems of the electricity-gas coupled integrated energy system and enhance the flexibility and adaptability of the system.

[0018] Reducing operating costs: This paper targets the operating costs of an integrated energy system and constructs an optimization objective function that considers both the distribution network and the natural gas grid. By considering load uncertainty, establishing a robust optimization model, and employing a two-layer robust optimization algorithm, this approach effectively reduces the operating costs of the integrated energy system and improves its economic benefits while ensuring stable system operation.

[0019] Improving system stability: This paper constructs a robust optimization model for an integrated energy system by considering distribution network and natural gas grid operating constraints, energy storage capacity, gas storage capacity, charging and discharging constraints, and gas charging and releasing constraints. This model effectively addresses load uncertainty, improves system stability and reliability, and reduces operational risks associated with load fluctuations.

[0020] Practicality and Innovation: This invention integrates controllable load resources and employs robust optimization methods to address the dynamic response of the demand side under multi-energy coupling. Simulations based on real-world cases demonstrate its effectiveness in reducing active power losses in distribution networks and operating costs of integrated energy systems, demonstrating its high practical value and innovation. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the process of the present invention; Figure 2It is a topological structure diagram of the integrated energy system during the verification process of the present invention; Figure 3 It is the voltage distribution diagram of the distribution network in scenario 1 during the verification process of the present invention; Figure 4 It is the voltage distribution diagram of the distribution network in scenario 2 during the verification process of the present invention; Figure 5 It is the voltage distribution diagram of the distribution network in scenario 3 during the verification process of the present invention. DETAILED DESCRIPTION

[0022] The embodiments of the present invention are further described below with reference to the accompanying drawings: like Figure 1 As shown in FIG, the robust optimization scheduling method of the integrated energy system based on controllable capability includes the following steps: S1. Taking the natural gas supply system and natural gas storage device as the research objects, a natural gas supply system pipeline storage model is established, and the gas pressure constraints of the natural gas pipeline nodes, pipeline operation constraints, gas supply constraints of the natural gas gate station, and gas filling and release constraints of the natural gas storage device are established; S2. Taking the controllability of the integrated energy system including the distribution network and the natural gas network as the research object, a time-shiftable load model for the distribution network, a time-shiftable load model for the natural gas network, and a demand response load model for the distribution network are established; S3. Taking the operating cost of the integrated energy system as the research target, an optimization objective function is constructed that takes into account the operating costs of the distribution network and the natural gas network; S4. Based on the distribution network operation constraints, energy storage constraints, natural gas gate station adjustment constraints and the constraints in S1, a robust optimization model of the integrated energy system considering load uncertainty is established based on the optimization objective function in S3; S5. Construct a two-layer robust optimization algorithm. The sub-problem solves the worst operating scenario, and the main problem optimizes the operating cost of the integrated energy system, and the optimal scheduling plan for the integrated energy system is obtained.

[0023] In S1, the process of establishing the pipeline storage model of the natural gas supply system is as follows: In the natural gas supply system, the pipeline storage state of the natural gas network is described based on the Weymouth equation: ; Where, is the average natural gas flow rate of natural gas pipeline mn during period t; and are the gas pressures at nodes m and n of the natural gas pipeline during period t; is the Weymouth constant, which is related to the physical properties of the pipe, such as temperature, length, diameter, and friction; The flow balance of the natural gas network must meet the following requirements: ; Where, is the set of natural gas pipelines connected to node n; is the set of natural gas storage devices connected to node n; is the set of natural gas gate stations connected to node n; is the set of natural gas loads connected to node n; is the set of gas turbines connected to node n; and are the gas flow rates at the end and head end of the natural gas pipeline mn during period t; and are the gas release and gas intake of the natural gas storage device s during period t, respectively; is the natural gas flow rate at the natural gas gate station c during period t; is the gas flow of natural gas load l connected to node n during period t; is the gas flow of the gas turbine g connected to the node n during period t; A linear pipeline storage model is used to describe the dynamic effects of the natural gas pipeline and to establish the pipeline storage model of the natural gas supply system. The expression is as follows: ; ; ; ; Where, 、 are the pipeline stocks of natural gas pipeline mn in periods t and t-1 respectively; is the average pressure of natural gas pipeline mn; is the length of the natural gas pipeline mn.

[0024] The gas pressure constraint at the natural gas pipeline node is: ; Where, 、 、 are the gas pressure and its upper and lower limits at the natural gas pipeline node during period t; The pipeline operation constraints are: ; Where, 、 are the upper and lower limits of the pipeline inventory of natural gas pipeline mn in period t respectively; The gas supply constraints of natural gas gate stations are: ; Where, 、 are the upper and lower limits of the natural gas flow rate at the natural gas gate station c during period t; The gas charging and releasing constraints of natural gas storage devices are: ; ; ; ; ; ; Where, 、 are the gas storage capacity of the natural gas storage device s at time periods t and t-1 respectively; 、 are binary variables, representing the inflation and degassing states of the gas storage device during period t, with a value of 1 indicating inflation or degassing, and the same applies to the other binary variables; 、 are the charging and releasing efficiencies of the natural gas storage device s, respectively; 、 are the maximum and minimum gas release amounts of the gas storage device s during period t, respectively; 、 are the maximum and minimum air filling volumes of the air storage device s during period t respectively.

[0025] In S2, there are time-shiftable loads in the distribution network, and the time-shiftable load control model of the distribution network is: ; ; Where, is the total load that can be time-shifted in a day; is the access amount of time-shiftable load in period t; 、 are the upper and lower limits of the load during time period t, respectively; T is the number of time periods, with a value of 24, representing one day; There are time-shiftable loads in the natural gas network. The time-shiftable load control model of the natural gas network is: ; ; Where, is the total load of the time-shiftable gas load in a day; is the access capacity of the time-shiftable gas load in period t; 、 are the upper and lower limits of the gas load during period t respectively; The distribution network demand response load model is: ; ; Where, represents the load reduction level of the zth curtailable load in period t; represents the maximum load reduction power during period t; 、 、 They represent the total amount of electric energy that can be reduced for the zth load and its upper and lower limits respectively.

[0026] In S3, the optimization objective function considering the distribution network operating costs and the natural gas network operating costs is: ; Where F is the operating cost of the integrated energy system; The operating cost of the distribution network; For the running costs of the gas network; The calculation method is: ; Where, The cost of purchasing electricity from the distribution network; Subsidy for time-shifting loads; Subsidy charges for demand response loads; The calculation method is: ; Where, Gas purchase costs for the natural gas network; Subsidy for time-shifting load on the natural gas network.

[0027] In S4, the distribution network operation constraints are: ; ; ; ; ; ; Where, is the square of the voltage amplitude at node i in the distribution network; 、 are the upper and lower limits of the voltage amplitude of the distribution network nodes respectively; is the square of the current amplitude flowing through branch ij; is the upper limit of the square of the current amplitude flowing through branch ij; is the active power flowing through branch ij during period t; is the number of the gas turbine; is the number of gas turbines; is a binary variable indicating whether the gas turbine is installed at the distribution network node j; For gas turbines The power value; is the resistance of branch ij; is the active load of node j connected to the distribution network during period t; is the active power flowing through branch ju during period t; Indicates that distribution network node u is connected to distribution network node j; is the number of the adjustable load; is the number of adjustable loads; is a binary variable, indicating whether the controllable load is connected to the distribution network node j; The load that can be controlled during period t The value of the charging load. The adjustable load includes time-shiftable adjustable load, energy storage load, and demand response load; is the reactive power flowing through branch ij during period t; is the reactance of branch ij; is the reactive load of node j connected to the distribution network during period t; Indicates that distribution network node h is connected to distribution network node j; is the reactive power flowing through branch jh during period t; is the square of the voltage amplitude at node j in the distribution network; The energy storage constraint is: ; ; ; ; ; Where, 、 are the remaining energy storage capacity in periods t and t-1 respectively; The charging and discharging efficiency of energy storage; 、 are binary variables representing the charge and discharge states of energy storage; 、 They are the upper and lower limits of the remaining energy storage capacity respectively; 、 、 are the energy storage charging power and its upper and lower limits in period t respectively; 、 、 are the energy storage discharge power and its upper and lower limits in period t respectively; The natural gas gate station adjustment constraints include natural gas supply flow constraints, gas supply quantity adjustment constraints, gas supply quantity adjustment frequency constraints, and gas supply flow limit constraints, which are: ; ; ; ; ; ; ; Where, is the natural gas flow rate of natural gas gate station c during period t-1; 、 are the minimum and maximum gas supply adjustment amounts of natural gas gate station c respectively; 、 are binary variables representing the increase in gas supply at the natural gas gate station during periods t and t-1, respectively; 、 are binary variables for the reduction of gas supply by the natural gas gate station during periods t and t-1, respectively; The maximum number of times the gas supply volume can be adjusted for a natural gas gate station.

[0028] The process of establishing the robust optimization model of the integrated energy system considering load uncertainty is as follows: During the operation of the integrated energy system, both the electric load and the gas load are uncertain. The uncertainty of the load is defined using a boxed uncertainty set. For the distribution network load: ; ; Where, is the load of the distribution network during period t; is the predicted value of distribution network load in period t; is the fluctuation deviation of distribution network load; The regulation parameters introduced to deal with the uncertainty of distribution network load; is the load number of the distribution network; For natural gas network load: ; ; Where, is the load of the natural gas network during period t; is the predicted value of the natural gas network load during period t; is the fluctuation deviation of natural gas network load; Adjustment parameters introduced to account for uncertainty in natural gas network load; The uncertainty parameter represents the number of time periods with the maximum or minimum values ​​of the fluctuation range described by the distribution network load and the natural gas network load during the scheduling period. It can adjust the conservatism of the optimal solution. The larger the value, the more conservative the optimization solution.

[0029] The optimization objective function is transformed into a robust optimization scheduling model, and a robust optimization model of the integrated energy system considering load uncertainty is obtained: .

[0030] In S5, the process of constructing a two-layer robust optimization algorithm is as follows: After considering the uncertain set of distribution network loads and natural gas network loads, the solution of the optimization objective function changes from a one-stage process to a two-stage process. In order to more effectively explain the solution process and solution method, a compact model is designed as follows: ; ; ; ; ; ; Where x is the optimization variable in the first stage, i.e., the time-of-use charging price of each controllable load; the optimization variables in the second stage are v and y, i.e., the optimized solutions of the load uncertainty variables and the distribution network flow; V is the set of load uncertainty variables; and Y is the set of optimized solutions of the distribution network flow. and is the coefficient matrix of the optimization objective function, the superscript T indicates transposition; d, o, a are constant column vectors; A, H, B, D, E, , G, is the sparse matrix of constraints corresponding to different optimization stages; Decompose the original problem into a main problem and sub-problems. The specific form of the main problem is: Objective function: ; Where, is the solution to the subproblem; Constraints: ; ; ; ; ; ; ; Where k is the index of the number of iterations; is the maximum number of iterations; is y at the kth iteration; is v at the kth iteration; The specific form of the sub-problem is: Objective function: ; Where, is a set of solutions for the variable x; Constraints: ; ; Given a set of uncertainty variables v, the inner min problem becomes a second-order cone programming problem. According to the strong duality theory, it is transformed into a max form of a dual problem and combined with the outer max problem to obtain the following form: Objective function: ; Where, is the dual variable in the subproblem; Constraints: ; ; ; Where, Number the norm constraint coefficient matrix; is the number of norm constraints; For the The coefficient matrix of the norm constraint; constrain the dual variable for the norm of the dual variable; The dual variable The coefficient matrix of the norm constraint; 、 is the dual variable; represents the L2 norm; b is a constant vector; Since the objective function in Since it is a bilinear problem, the binary expansion method and the big M method are used for linearization, and the subproblem is reformulated as a mixed integer second-order cone programming problem: ; Where M is the constant in the big M method; is the dual variable corresponding to the part with uncertain variables; is the upper bound of the dual variable; is the lower bound of the dual variable; is an auxiliary variable, which is a binary variable; is an auxiliary variable, which is a binary variable; is the uncertainty adjustment parameter.

[0031] The solution process is as follows: S5.1. Given a set of v values ​​as the initial worst-case scenario, set the lower bound of the operating cost to , the upper bound of the operating cost is , the number of iterations is k=1; S5.2. Based on the worst scenario Solve the main problem and obtain the optimal solution of the main problem at the kth iteration and ( is the solution of the variable, is the value of the objective function), and the main problem is obtained value as the new lower bound, ; S5.3. Substitute the solution of the main problem into the subproblem and obtain the optimal solution of the subproblem at the kth iteration and the worst-case scenario (The optimal solution is the kth iteration, and then the optimal solution will get a worst-case scenario, which will be brought into the k+1th iteration), update the upper bound ; S5.4. Settings is the preset convergence judgment margin, if , it indicates that the optimization calculation has reached the optimal solution and the iteration stops; otherwise, add variables and the following constraints: ; ; ; ; Where, is y at the k+1th iteration; S5.5. Set k = k + 1 and return to S5.2 to continue iterating until the algorithm converges.

[0032] The verification process is based on Figure 2 The integrated energy system topology shown was simulated and verified using Matlab, and solved using Gurobi. The verification example settings are shown in Table 1, and the simulation comparison scenario settings are shown in Table 2.

[0033] Table 1 Verification example settings

[0034] Table 2 Simulation comparison scenario settings

[0035] Figure 3-Figure 5 Table 3 compares the active power losses of the distribution network for each scenario. Table 4 compares the operating costs of the integrated energy system for each scenario.

[0036] Table 3 Comparison of active power losses in distribution networks

[0037] Table 4 Comparative analysis of operating costs of integrated energy systems

[0038] According to Table 3, Table 4 and Figure 3-5 It can be seen that by adopting the method proposed in this embodiment, the minimum operating costs of the distribution network and the natural gas network are achieved, and the voltage distribution is more stable.

[0039] Example 2: A robust optimization scheduling device for an integrated energy system based on controllable capabilities includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor. The method in Example 1 is implemented by executing the computer program through the processor.

Claims

1. A robust optimization scheduling method for integrated energy systems based on controllable capabilities, characterized by The following steps are involved: S1. Taking the natural gas supply system and natural gas storage device as the research objects, a natural gas supply system pipeline storage model is established, and the gas pressure constraints of the natural gas pipeline nodes, pipeline operation constraints, gas supply constraints of the natural gas gate station, and gas filling and release constraints of the natural gas storage device are established; S2. Taking the controllability of the integrated energy system including the distribution network and the natural gas network as the research object, a time-shiftable load model for the distribution network, a time-shiftable load model for the natural gas network, and a demand response load model for the distribution network are established; S3. Taking the operating cost of the integrated energy system as the research target, an optimization objective function is constructed that takes into account the operating costs of the distribution network and the natural gas network; S4. Based on the distribution network operation constraints, energy storage constraints, natural gas gate station adjustment constraints and the constraints in S1, a robust optimization model of the integrated energy system considering load uncertainty is established based on the optimization objective function in S3; S5. Construct a two-layer robust optimization algorithm. The sub-problem solves the worst operating scenario, and the main problem optimizes the operating cost of the integrated energy system, and the optimal scheduling plan for the integrated energy system is obtained.

2. The robust optimization scheduling method for an integrated energy system based on controllable capability according to claim 1 is characterized in that: In S1, the process of establishing the pipeline storage model of the natural gas supply system is as follows: In the natural gas supply system, the pipeline storage state of the natural gas network is described based on the Weymouth equation: ； Where, is the average natural gas flow rate of natural gas pipeline mn during period t; and are the gas pressures at nodes m and n of the natural gas pipeline during period t; is the Weymouth constant; The flow balance of the natural gas network must meet the following requirements: ； Where, is the set of natural gas pipelines connected to node n; is the set of natural gas storage devices connected to node n; is the set of natural gas gate stations connected to node n; is the set of natural gas loads connected to node n; is the set of gas turbines connected to node n; and are the gas flow rates at the end and head end of the natural gas pipeline mn during period t; and are the gas release and gas intake of the natural gas storage device s during period t, respectively; is the natural gas flow rate at the natural gas gate station c during period t; is the gas flow of natural gas load l connected to node n during period t; is the gas flow of the gas turbine g connected to the node n during period t; A linear pipeline storage model is used to describe the dynamic effects of the natural gas pipeline and to establish the pipeline storage model of the natural gas supply system. The expression is as follows: ； ； ； ； Where, 、 are the pipeline stocks of natural gas pipeline mn in periods t and t-1 respectively; is the average pressure of natural gas pipeline mn; is the length of the natural gas pipeline mn.

3. The robust optimization scheduling method for an integrated energy system based on controllable capacity according to claim 2 is characterized in that: In S1, the gas pressure constraint of the natural gas pipeline node is: ； Where, 、 、 are the gas pressure and its upper and lower limits at the natural gas pipeline node during period t; The pipeline operation constraints are: ； Where, 、 are the upper and lower limits of the pipeline inventory of natural gas pipeline mn in period t respectively; The gas supply constraints of natural gas gate stations are: ； Where, 、 are the upper and lower limits of the natural gas flow rate at the natural gas gate station c during period t; The gas charging and releasing constraints of natural gas storage devices are: ； ； ； ； ； ； Where, 、 are the gas storage capacity of the natural gas storage device s at time periods t and t-1 respectively; 、 are binary variables, representing the inflation and degassing states of the gas storage device during period t; 、 are the charging and releasing efficiencies of the natural gas storage device s, respectively; 、 are the maximum and minimum gas release amounts of the gas storage device s during period t, respectively; 、 are the maximum and minimum air filling volumes of the air storage device s during period t respectively.

4. The robust optimization scheduling method for an integrated energy system based on controllable capability according to claim 1 is characterized in that: In S2, there is a time-shiftable load inside the distribution network, and the time-shiftable load control model of the distribution network is: ； ； Where, is the total load that can be time-shifted in a day; is the access amount of time-shiftable load in period t; 、 are the upper and lower limits of the load during time period t, respectively; T is the number of time periods, with a value of 24, representing one day; There are time-shiftable loads in the natural gas network. The time-shiftable load control model of the natural gas network is: ； ； Where, is the total load of the time-shiftable gas load in a day; is the access capacity of the time-shiftable gas load in period t; 、 are the upper and lower limits of the gas load during period t respectively; The distribution network demand response load model is: ； ； Where, represents the load reduction level of the zth curtailable load in period t; represents the maximum load reduction power during period t; 、 、 They represent the total amount of electric energy that can be reduced for the zth load and its upper and lower limits respectively.

5. The robust optimization scheduling method for an integrated energy system based on controllable capability according to claim 3 is characterized in that: In S3, the optimization objective function considering the distribution network operating costs and the natural gas network operating costs is: ； Where F is the operating cost of the integrated energy system; The operating cost of the distribution network; For the running costs of the gas network; The calculation method is: ； Where, The cost of purchasing electricity from the distribution network; Subsidy for time-shifting loads; Subsidy charges for demand response loads; The calculation method is: ； Where, Gas purchase costs for the natural gas network; Subsidy for time-shifting load on the natural gas network.

6. The robust optimization scheduling method for an integrated energy system based on controllable capability according to claim 5 is characterized in that: In S4, the distribution network operation constraints are: ； ； ； ； ； ； Where, is the square of the voltage amplitude at node i in the distribution network; 、 are the upper and lower limits of the voltage amplitude of the distribution network nodes respectively; is the square of the current amplitude flowing through branch ij; is the upper limit of the square of the current amplitude flowing through branch ij; is the active power flowing through branch ij during period t; is the number of the gas turbine; is the number of gas turbines; is a binary variable indicating whether the gas turbine is installed at the distribution network node j; For gas turbines The power value; is the resistance of branch ij; is the active load of node j connected to the distribution network during period t; is the active power flowing through branch ju during period t; Indicates that distribution network node u is connected to distribution network node j; is the number of the adjustable load; is the number of adjustable loads; is a binary variable, indicating whether the controllable load is connected to the distribution network node j; The load that can be controlled during period t The value of the charging load. The adjustable load includes time-shiftable adjustable load, energy storage load, and demand response load; is the reactive power flowing through branch ij during period t; is the reactance of branch ij; is the reactive load of node j connected to the distribution network during period t; Indicates that distribution network node h is connected to distribution network node j; is the reactive power flowing through branch jh during period t; is the square of the voltage amplitude at node j in the distribution network; The energy storage constraint is: ； ； ； ； ； Where, 、 are the remaining energy storage capacity in periods t and t-1 respectively; The charging and discharging efficiency of energy storage; 、 are binary variables representing the charge and discharge states of energy storage; 、 They are the upper and lower limits of the remaining energy storage capacity respectively; 、 、 are the energy storage charging power and its upper and lower limits in period t respectively; 、 、 are the energy storage discharge power and its upper and lower limits in period t respectively; The natural gas gate station adjustment constraints include natural gas supply flow constraints, gas supply quantity adjustment constraints, gas supply quantity adjustment frequency constraints, and gas supply flow limit constraints, which are: ； ； ； ； ； ； ； Where, is the natural gas flow rate of natural gas gate station c during period t-1; 、 are the minimum and maximum gas supply adjustment amounts of natural gas gate station c respectively; 、 are binary variables representing the increase in gas supply at the natural gas gate station during periods t and t-1, respectively; 、 are binary variables for the reduction of gas supply by the natural gas gate station during periods t and t-1, respectively; The maximum number of times the gas supply volume can be adjusted for a natural gas gate station.

7. The robust optimization scheduling method for an integrated energy system based on controllable capability according to claim 6 is characterized in that: In S4, the process of establishing the robust optimization model of the integrated energy system considering load uncertainty is as follows: During the operation of the integrated energy system, both the electric load and the gas load are uncertain. The uncertainty of the load is defined using a boxed uncertainty set. For the distribution network load: ； ； Where, is the load of the distribution network during period t; is the predicted value of distribution network load in period t; is the fluctuation deviation of distribution network load; The regulation parameters introduced to deal with the uncertainty of distribution network load; For natural gas network load: ； ； Where, is the load of the natural gas network during period t; is the predicted value of the natural gas network load during period t; is the fluctuation deviation of natural gas network load; Adjustment parameters introduced to account for uncertainty in natural gas network load; The optimization objective function is transformed into a robust optimization scheduling model, and a robust optimization model of the integrated energy system considering load uncertainty is obtained: 。 8. The robust optimization scheduling method for an integrated energy system based on controllable capability according to claim 7 is characterized in that: In S5, the process of constructing a two-layer robust optimization algorithm is as follows: After considering the uncertain set of distribution network loads and natural gas network loads, the solution of the optimization objective function changes from a one-stage process to a two-stage process. In order to more effectively explain the solution process and solution method, a compact model is designed as follows: ； ； ； ； ； ； Where x is the optimization variable in the first stage, i.e., the time-of-use charging price of each controllable load; the optimization variables in the second stage are v and y, i.e., the optimized solutions of the load uncertainty variables and the distribution network flow; V is the set of load uncertainty variables; and Y is the set of optimized solutions of the distribution network flow. and is the coefficient matrix of the optimization objective function, the superscript T indicates transposition; d, o, a are constant column vectors; A, H, B, D, E, , G, is the sparse matrix of constraints corresponding to different optimization stages; Decompose the original problem into a main problem and sub-problems. The specific form of the main problem is: Objective function: ； Where, is the solution to the subproblem; Constraints: ； ； ； ； ； ； ； Where k is the index of the number of iterations; is the maximum number of iterations; is y at the kth iteration; is v at the kth iteration; The specific form of the sub-problem is: Objective function: ； Where, is a set of solutions for the variable x; Constraints: ； ； Given a set of uncertainty variables v, the inner min problem becomes a second-order cone programming problem. According to the strong duality theory, it is transformed into a max form of a dual problem and combined with the outer max problem to obtain the following form: Objective function: ； Where, is the dual variable in the subproblem; Constraints: ； ； ； Where, Number the norm constraint coefficient matrix; is the number of norm constraints; For the The coefficient matrix of the norm constraint; constrain the dual variable for the norm of the dual variable; The dual variable The coefficient matrix of the norm constraint; 、 is the dual variable; represents the L2 norm; b is a constant vector; Since the objective function in Since it is a bilinear problem, the binary expansion method and the big M method are used for linearization, and the subproblem is reformulated as a mixed integer second-order cone programming problem: ； Where M is the constant in the big M method; is the dual variable corresponding to the part with uncertain variables; is the upper bound of the dual variable; is the lower bound of the dual variable; is an auxiliary variable, which is a binary variable; is an auxiliary variable, which is a binary variable; is the uncertainty adjustment parameter.

9. The robust optimization scheduling method for an integrated energy system based on controllable capability according to claim 8 is characterized in that: In the above S5, the solution process is as follows: S5.

1. Given a set of v values ​​as the initial worst-case scenario, set the lower bound of the operating cost to , the upper bound of the operating cost is , the number of iterations is k=1; S5.

2. Based on the worst scenario Solve the main problem and obtain the optimal solution of the main problem at the kth iteration and , obtained from the main problem value as the new lower bound, ; S5.

3. Substitute the solution of the main problem into the subproblem and obtain the optimal solution of the subproblem at the kth iteration and the worst-case scenario , update the upper bound ; S5.

4. Settings is the preset convergence judgment margin, if , it indicates that the optimization calculation has reached the optimal solution and the iteration stops; otherwise, add variables and the following constraints: ； ； ； ； Where, is y at the k+1th iteration; S5.

5. Set k = k + 1 and return to S5.2 to continue iterating until the algorithm converges.

10. A robust optimization scheduling device for an integrated energy system based on controllable capabilities, characterized by: The method comprises a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the method according to any one of claims 1 to 9 is implemented by executing the computer program by the processor.

Images