An energy system optimization method and device based on the optimization of the offline domain and the online domain

Abstract

The present invention discloses an energy system optimization method and device based on the optimization of the offline domain and the online domain, including: constructing a state perception module for real-time monitoring of the operating state and environmental changes of the multi-energy system, providing instant data support for the optimization decision-making of the RIES system and the CIES system; constructing a safety-economic optimization module, combining the Stackelberg game and the cooperative game strategy to construct an optimization model for the energy scheduling and resource allocation scheme; constructing an offline domain generation module, based on the multi-parameter programming theory, constructing the matrix form of the optimization model for the energy scheduling and resource allocation scheme, analyzing the correlation of scenarios, iterations and model parameters in the CIES system, and generating the scenario domain, iteration domain and model domain of the optimization problem based on the critical region; constructing an online domain optimization module, by matching the corresponding region, when it is recognized that the operating state of the current CIES system is in a certain critical region, solving through a linear equation, and determining the energy scheduling and resource allocation scheme based on the solution result of the linear equation of the online domain optimization module.

Description

Technical Field

[0001] The present invention relates to the technical field of integrated energy systems, and in particular, to an energy system optimization method and device based on offline domain and online domain optimization. Background Art

[0002] Regional Integrated Energy Systems (RIES) integrate various energies (including electricity, heat, and natural gas) within a region to promote the coordinated planning and optimal operation of energy subsystems. Community Integrated Energy Systems (CIES) focus on optimizing energy utilization within a community, improving energy efficiency, and ensuring reliable and secure energy supply at a more local scale. These two systems are crucial for improving overall energy efficiency, reducing environmental impacts, and ensuring the stable operation of the energy system.

[0003] RIES and CIES are operated by different entities, and each entity makes optimal decisions based on different goals. Game theory is widely used to solve multi-agent, multi-objective optimization problems. The interaction between the regional operator and CIES is usually modeled as a leader-follower strategy. A Stackelberg game strategy based on the KKT conditions was proposed in references [1] and [2] to solve the two-layer pricing model between the Distribution System Operator (DSO) and CIES, and a unique equilibrium solution was proven in reference [3]. In addition, the energy trading of RIES usually involves multiple CIES. From the perspective of cooperative game theory, the Shapley value is used for CIES optimization modeling [4][5] , and from the non-cooperative perspective, the Stackelberg game solves the unequal trading problem [6][7] . However, current hybrid game strategies often simplify energy operators to pure power grids, ignoring the roles of natural gas and heat networks. In addition, if a failure occurs in RIES, load shedding can be alleviated through the joint optimization of RIES and CIES [8] . These problems have not been incorporated into existing energy trading strategies.

[0004] The hybrid game framework has largely matured, but the computational complexity of the solution algorithm remains an important bottleneck. Distributed algorithms such as the Alternating Direction Method of Multipliers (ADMM) are usually used to solve cooperative and non-cooperative games in hybrid game problems involving multi-linear models [9]. Traditionally, the leader-follower game model is solved by transforming it into a single-level problem by applying the KKT conditions

[10] . However, this process can be time-consuming and memory-intensive, especially in the case of large models. Therefore, researchers are increasingly turning to heuristic algorithms

[11] . Although these methods are promising, they usually rely heavily on many hyperparameters and are prone to getting stuck in local optima. Some methods combine optimization techniques with heuristic algorithms. For example, Ref.

[12] proposed a hybrid algorithm that uses a genetic algorithm nested in quadratic programming. Although this method solves the local optimum problem, it still faces a heavy computational time burden.

[0005] References

[0006] [1] Lv C, Liang R, Zhang G, et al. Energy accommodation-oriented interaction of active distribution network and central energy station considering soft open points. Energy, 2023, 268: 126574.

[0007] [2] Jia S, Peng K, Zhang X, et al. Dynamic pricing strategy and regional energy consumption optimization based on different stakeholders. International Journal of Electrical Power & Energy Systems, 2022, 141: 108199.

[0008] [3] Zafar S, Blavette A, Camilleri G, et al. Decentralized optimal management of a large-scale EV fleet: optimality and computational complexity comparison between an Adaptive MAS and MILP. International Journal of Electrical Power & Energy Systems, 2023, 147: 108861.

[0009] [4] Cao M, Shao C, Hu B, et al. Reliability tracing of the integrated energy system using the improved shapley value. Energy, 2022, 260: 124997.

[0010] [5] Wang Y, Liu Z, Cai C, et al. Research on the optimization method of integrated energy system operation with multi - subject game. Energy, 2022, 245: 123305.

[0011] [6] Li Y, Wang B, Yang Z, et al. Hierarchical stochastic scheduling of multi - community integrated energy systems in uncertain environments via Stackelberg game. Applied Energy, 2022, 308: 118392.

[0012] [7] Li S, Zhang L, Nie L, et al. Trading strategy and benefit optimization of load aggregators in integrated energy systems considering integrated demand response: A hierarchical Stackelberg game. Energy, 2022, 249: 123678.

[0013] [8] Yan M, He Y, Shahidehpour M, et al. Coordinated regional-district operation of integrated energy systems for resilience enhancement in natural disasters. IEEE Transactions on Smart Grid, 2018, 10(5): 4881-4892.

[0014] [9] Zhao B, Duan P, Fen M, et al. Optimal operation of distribution networks and multiple community energy prosumers based on mixed game theory. Energy, 2023, 278: 128025.

[0015]

[10] Lei Z, Liu M, Shen Z. Analysis of strategic bidding of a DER aggregator in energy markets through the Stackelberg game model with the mixed-integer lower-level problem. International Journal of Electrical Power & Energy Systems, 2023, 152: 109237.

[0016]

[11] Liang Z, Mu L. Multi-agent low-carbon optimal dispatch of regional integrated energy system based on mixed game theory. Energy, 2024, 295: 130953.

[0017]

[12] Li K,Ye N,Li S,et al.Distributed collaborative operation strategies in multi-agent integrated energy system considering integrated demand response based on game theory.Energy,2023,273:127137. Summary of the Invention

[0018] The present invention provides an energy system optimization method and device based on offline domain and online domain optimization. The present invention improves the economic - security of the park integrated energy system and the regional integrated energy system, and improves the optimization scheduling efficiency, as described in detail below:

[0019] In a first aspect, an energy system optimization method based on offline domain and online domain optimization, the method includes:

[0020] Construct a state perception module for real - time monitoring of the operating state and environmental changes of the multi - energy system, providing instant data support for the optimization decision of the RIES system and the CIES system;

[0021] Construct a security - economy optimization module, combining Stackelberg game and cooperative game strategies to construct an optimization model for energy scheduling and resource allocation;

[0022] Construct an offline domain generation module. Based on multi - parameter programming theory, construct the matrix form of the optimization model for energy scheduling and resource allocation, analyze the correlation of scenarios, iterations, and model parameters in the CIES system, and generate the scenario domain, iteration domain, and model domain of the optimization problem based on the critical region;

[0023] Construct an online domain optimization module. During the operation of the CIES system, according to the instant data of the state perception module, identify the scenario domain, iteration domain, and model domain corresponding to the current operating state. When it is identified that the current operating state of the CIES system is in a certain critical region, solve it through a linear equation. Based on the solution result of the linear equation of the online domain optimization module, determine the energy scheduling and resource allocation plan.

[0024] Among them, the offline domain generation module is:

[0025] Iteration correlation:

[0026] min f=(c + △c) T x

[0027] s.t.Ax = b x≥0

[0028] where Δc is a random variable related to c;

[0029] Temporal correlation:

[0030] min f = c T x

[0031] s.t. Ax = b + Δb x ≥ 0

[0032] where Δb is a random variable related to b;

[0033] Model correlation:

[0034] min f = c T x

[0035] s.t. (A + ΔA)x = b x ≥ 0

[0036] where ΔA is a random variable related to A;

[0037] Define the iterative domain R related to CIES, Δc, Δb, and ΔA c , the scenario domain R b , and the model domain R A as follows:

[0038] R c = {(Δc)|c N + Δc N = A N T (A B -1 ) T (c B + Δc B )}

[0039] R b = {(Δb)|A B -1 (b + Δb) ≥ 0}

[0040]

[0041] where B is the critical region activation flag, and N is the non-critical region flag; c N represents the non-critical region vector corresponding to c; A N represents the non-critical region matrix corresponding to A; c B represents the critical region vector corresponding to c; A B represents the critical region matrix corresponding to A; Δc N represents the non-critical region vector corresponding to Δc; ΔA N represents the non-critical region matrix corresponding to ΔA; ΔcB Represents the critical region vector corresponding to Δc; △A B Represents the critical region matrix corresponding to ΔA.

[0042] Among them, the online domain optimization module is as follows:

[0043] Iterative correlation criterion:

[0044]

[0045] Time correlation criterion:

[0046]

[0047] Model similarity criterion:

[0048] c N T -c B (A B +△A B ) -1 (A N +△A N )≤0

[0049] (A B +△A B ) -1 b≥0

[0050] The offline domain space consists of several critical regions, and each critical region contains several optimization problems with the same parameters B and N. The mapping relationship between the optimization problems and the critical regions is constructed through the online domain optimization module to determine the corresponding critical region of the current optimization problem in the offline domain generation module.

[0051] In a second aspect, an energy system optimization device based on offline domain and online domain optimization, the device includes: a processor and a memory, and program instructions are stored in the memory. The processor calls the program instructions stored in the memory to enable the device to perform the following operations:

[0052] Real-time monitor the operating status and environmental changes of the multi-energy system, and provide instant data support for the optimization decisions of the RIES system and the CIES system;

[0053] Combined with the Stackelberg game and the cooperative game model, construct an optimization model for the energy scheduling and resource allocation scheme;

[0054] Based on the multi-parameter programming theory, construct the matrix form of the optimization model for the energy scheduling and resource allocation scheme, analyze the correlation of scenarios, iterations, and model parameters in the CIES system, and generate the scenario domain, iteration domain, and model domain of the optimization problem based on the critical region;

[0055] Based on the real-time data of the status perception module, identify the scenario domain, iteration domain, and model domain corresponding to the current operating status. By matching the corresponding regions, when it is identified that the operating status of the current CIES system is in a certain critical region, solve it through a linear equation;

[0056] Based on the solution result of the linear equation of the online domain optimization module, determine the energy scheduling and resource allocation scheme.

[0057] In a third aspect, a computer-readable storage medium stores a computer program, the computer program includes program instructions, and when the program instructions are executed by a processor, the processor is caused to execute the method described in any one of the claims in the first aspect.

[0058] The beneficial effects of the technical solution provided by the present invention are:

[0059] 1) Improve the operation reliability and economy of the multi-energy system: By establishing a reliable and economic multi-energy system optimization model, combining the strategies of offline domain generation and online domain optimization, improve the stable operation ability of the system in complex environments;

[0060] 2) Improve the solution efficiency of the optimization model: Through the strategies of offline domain generation and online domain optimization, reduce the computational amount in the online stage, and transform complex optimization problems into linear solutions, further improving the solution efficiency; especially in large-scale energy systems, various types of problems can be quickly solved to adapt to the dynamic changes of the energy system;

[0061] 3) Realize the collaborative optimization of multiple systems: Combining Stackelberg game and cooperative game, develop an energy support strategy based on price incentives, enabling CIES to form a micro energy network in fault scenarios and providing load support for RIES, enhancing the adaptability of the system under sudden faults or extreme conditions. Description of the Drawings

[0062] Figure 1 It is a schematic diagram of an energy system optimization method based on offline domain and online domain optimization;

[0063] Figure 2 It is a schematic diagram of energy trading between CIESs in a fault state;

[0064] Figure 3 It is a comparison chart of calculation times. Detailed Embodiments

[0065] To make the objectives, technical solutions, and advantages of the present invention clearer, the following further describes the embodiments of the present invention in detail.

[0066] Example 1

[0067] An embodiment of the present invention provides an energy system optimization method based on offline domain and online domain optimization. The method includes: a state perception module, an offline domain generation module, an online domain optimization module, and a critical region fast solution module. Among them:

[0068] Build a state perception module for real-time monitoring of the operating state of the multi-energy system and environmental changes, providing instant data support for the optimization decisions of the RIES system and the CIES system;

[0069] Build a safety-economic optimization module, and combine the Stackelberg game and cooperative game strategies to build an optimization model for energy scheduling and resource allocation;

[0070] Build an offline domain generation module. Based on the multi-parameter programming theory, build a matrix form of the optimization model for the energy scheduling and resource allocation scheme, analyze the correlation of scenarios, iterations, and model parameters in the CIES system, and generate a scenario domain, an iteration domain, and a model domain for the optimization problem based on the critical region;

[0071] Build an online domain optimization module. During the operation of the CIES system, according to the instant data of the state perception module, identify the scenario domain, iteration domain, and model domain corresponding to the current operating state. By matching the corresponding regions, when it is identified that the current operating state of the CIES system is in a certain critical region, solve it through a linear equation, and determine the energy scheduling and resource allocation scheme based on the solution result of the linear equation of the online domain optimization module.

[0072] In summary, the energy system optimization method based on offline domain and online domain optimization provided by the embodiment of the present invention ensures that the RIES and CIES systems can efficiently, economically, and reliably cope with complex environments and dynamic changes during operation by generating critical regions offline and combining online optimization strategies.

[0073] Embodiment 2

[0074] The following further introduces the solution in Embodiment 1 in combination with specific calculation formulas and examples. See the following description for details:

[0075] I. State Perception Module

[0076] Among them, the comprehensive energy system state perception module includes two parts: 1) RIES state perception sub-module; 2) CIES state perception sub-module.

[0077] 1) RIES State Perception Sub-module

[0078] The distribution network uses the Distflow model to achieve state perception.

[0079]

[0080] Among them, t is the time index; Ω b is the set of nodes; P t,ij and Q t,ij represent the active and reactive powers of branch ij; r ij and x ij represent the resistance and reactance of branch ij; V t,i and V t,j respectively represent the squared values of the voltage amplitudes of nodes i and j at time t; l t,ij is the squared value of the branch current amplitude; P t,i is the active power of the load carried by node i; Q t,i is the reactive power of the load carried by node i; Q t,jk is the reactive power of branch jk; P t,jk is the active power of branch jk; k is all the load nodes connected to node j.

[0081] The natural gas network realizes state perception based on the node pressure equation and the pipeline flow equation.

[0082]

[0083] pf i = ρ·pf j (6)

[0084] A G ·F G - G l ·D l - S = F L (7)

[0085] Among them, F ij represents the natural gas flow; c ij is the pipeline parameter; pf i and pf i respectively represent the gas pressures at nodes i and j. ρ represents the compression ratio; A G is the node - branch incidence matrix of the natural gas system; F G represents the branch flow matrix; G l is the compressor node incidence matrix; D l represents the compressor consumption flow; S and F L respectively represent the gas source output and the node load.

[0086] The thermal system realizes state perception based on the hydraulic model and the thermal model.

[0087] A H ·m = m q (8)

[0088] Bh f= BK|m| = 0 (9)

[0089] Η L = C p m q (T s - T o ) (10)

[0090]

[0091] (∑m out )T end = ∑(m in T start ) (12)

[0092] Among them, A H is the node-branch incidence matrix; m represents the mass flow rate vector; m q is the injection flow vector at the system node. h f is the vector of heat pressure loss; B is the branch incidence matrix; K represents the resistance coefficient of the heat pipeline. Η L represents the heat power, C p represents the specific heat capacity of water; T s and T o correspond to the injection and outflow temperatures at the node respectively. T end and T start are the inlet and outlet temperatures of the pipeline respectively. The ambient temperature is represented by T a , λ represents the heat transfer coefficient per unit length of the pipeline; m out and m in represent the mass flow velocities of the pipeline when leaving or entering the mixing node respectively, L represents the pipeline length, and m represents the mass flow scalar parameter.

[0093] 2) CIES Status Sensing Sub-module

[0094] The status sensing models of ground source heat pumps, air source heat pumps, gas boilers, gas turbines, compressed air refrigerators, and electric refrigerators can be expressed as:

[0095]

[0096]

[0097] Among them, P, H, and C represent electric, heat, and cold powers respectively; F represents the natural gas flow rate; η ASHP , η GHP , η GB , η GT , η AC and η EC all represent energy conversion coefficients; q gasis the calorific value of natural gas; t represents the time index; GHP represents the ground source heat pump; ASHP represents the air source heat pump; GB represents the gas boiler; GT represents the gas turbine; AC represents the compressed air refrigeration machine; EC represents the electric refrigeration machine.

[0098] The state perception model of the extraction condensing cogeneration unit (CHP) can be expressed as:

[0099]

[0100] Among them, Z CHP is the ratio of the change in electric output power to the heat output power. represents the power of the extraction condensing cogeneration unit in the full condensation mode; η g2e is the efficiency of converting natural gas into electricity by the cogeneration; η CHP represents the overall efficiency of the cogeneration; is the fuel consumption flow rate of the cogeneration at time t; F CHP is the rated fuel consumption flow rate of the cogeneration unit; is the output electric power of the cogeneration unit at time t; is the output heat power of the cogeneration unit at time t.

[0101] II. Safety - economic optimization module

[0102] 1) RIES optimization sub - module

[0103] As the leader of the Stackelberg game, RIES aims to maximize its revenue by conducting energy transactions with each lower - level CIES and ensure reliable and safe operation. The objective function of RIES is expressed as follows:

[0104]

[0105] Among them, and respectively represent the economic benefit and the reliability cost, which can be calculated using the following equations:

[0106]

[0107] Among them, n is the index of CIES; t is the time step; k represents the type of energy; e represents the type of load demand. b and s respectively represent the identifiers for CIES to buy and sell energy to RIES, while B and S represent whether RIES buys or sells energy from / to the superior energy network. and represent the energy price, and represent the energy trading volume, Represents the reliability cost of load type e, Represents the load shedding amount.

[0108] 2) CIES optimization sub-module

[0109] The optimization of CIES aims to minimize the total operating cost. The objective function is expressed as follows:

[0110]

[0111] Among them, Represents the transaction cost related to RIES; Represents the energy transaction cost between multiple CIESs; Refers to the operating cost of the internal equipment of CIES n; Represents the cost related to IDR; Corresponds to the energy transmission cost between multiple CIESs. Represents the carbon emission cost of CIES n.

[0112] The specific expressions of each cost are provided in the following equations:

[0113]

[0114] Among them, m is the index of CIESs other than CIES n. Are the energy purchase and sale prices between CIESs respectively; Are the energy purchase and sale volumes between CIESs respectively; Is the transmission cost coefficient of the energy interconnection line; v dp And v dg Represent the operating cost coefficients of electrical equipment and gas equipment respectively. P dp,n,t And G dg,n,t Are the output powers of electrical equipment and gas equipment respectively; Is the transferable load cost of load e; And Are the increments and decrements of the transferred load at time t respectively; u Ca Represents the carbon price. And Represent the actual carbon emissions and carbon emission quotas of CIES n respectively.

[0115] III. Offline domain generation module

[0116] The proposed security-economic optimization module is divided into two stages: the security operation optimization of RIES and CIES based on Stackelberg game, and the economic optimization of CIES cooperation game. For clarity, first, the lower-layer optimization problem of the Stackelberg game model is expressed in matrix form.

[0117]

[0118] Among them, \(T\) is the transpose symbol; \(A\) and \(b\) are the technical coefficient matrix and the right - hand - side vector of model (32), which are determined by (13)-(21), \(c\) is the cost vector, which is determined by equations (25), (26)-(31), \(x\) is the optimization variable, and \(f\) is the objective function.

[0119] Define three types of parameter correlations:

[0120] 1. Iterative correlation: In each iteration, the upper - layer model transmits different energy - price vectors and to the lower - layer model, resulting in changes in \(c\). Therefore, the iterative similarity model can be expressed as:

[0121]

[0122] where \(\Delta c\) is a random variable related to \(c\).

[0123] 2. Temporal correlation: Over different time ranges, changes in the load level and renewable - energy output cause changes in the parameter \(b\). Therefore, the temporal similarity model can be further refined based on this.

[0124]

[0125] where \(\Delta b\) is a random variable related to \(b\).

[0126] 3. Model correlation: Within different CIESs, differences in equipment capacity and equipment technical parameters cause changes in the parameter \(A\). Therefore, model similarity can be expressed as:

[0127]

[0128] where \(\Delta A\) is a random variable related to \(A\).

[0129] The critical region (CR) is defined as a set of random variables that keep the optimal basis (i.e., the current set of active constraints) unchanged. In CIES, the \(A\) matrix remains fixed over time, ensuring that the dimension of the optimization model remains consistent. Therefore, in the offline stage, the critical regions \(R\) c 、\(R\) b 、\(R\) A related to \(\Delta c\), \(\Delta b\), and \(\Delta A\) for CIES can be formulated as follows:

[0130] \(R\) c =\(\{(\triangle c)|c\) N +\(\triangle c\) N =\(A\) N T (AB -1 ) T (c B +△c B )} (36)

[0131] R b ={(△b)|A B -1 (b + △b)≥0} (37)

[0132]

[0133] Among them, B is the critical region activation identifier, and N is the non-critical region identifier; c N represents the non-critical region vector corresponding to c; A N represents the non-critical region matrix corresponding to A; c B represents the critical region vector corresponding to c; A B represents the critical region matrix corresponding to A; △c N represents the non-critical region vector corresponding to Δc; △A N represents the non-critical region matrix corresponding to ΔA; △c B represents the critical region vector corresponding to Δc; △A B represents the critical region matrix corresponding to ΔA.

[0134] IV. Online Domain Optimization Module

[0135] Once the CR is generated, different optimization problems can be matched with the corresponding critical regions, and the optimization problems can be transformed into equation solving to determine the optimal strategy. The following gives the matching criteria for three optimization problems in the critical region:

[0136] 1. Iterative Correlation Criterion:

[0137] (c N +△c N ) T -(c B +△c B )A B -1 A N ≤0 (39)

[0138] 2. Time Correlation Criterion:

[0139] A B -1 (b + △b)≥0 (40)

[0140] 3. Model Similarity Criterion:

[0141] c N T -cB (A B +△A B ) -1 (A N +△A N )≤0 (41)

[0142] (A B +△A B ) -1 b≥0 (42)

[0143] The parameter space consists of several critical regions, and each region contains several optimization problems with the same parameters B and N. Therefore, during the online operation phase, based on the online domain optimization module, a mapping relationship between the optimization problems and the critical regions is constructed, that is, the critical region where the current optimization problem is located in the offline domain generation module is determined, and then the solution efficiency of the optimization problem is significantly accelerated during the online phase.

[0144] Embodiment 3

[0145] The best implementation manner of the embodiment of the present invention relates to an energy system optimization device based on offline domain and online domain optimization, aiming to achieve the efficient, intelligent and safe operation of the energy system through a highly integrated modular design. This implementation manner includes the following key components:

[0146] State perception module: This module provides real-time data support by monitoring the operating status and environmental changes of the RIES and CIES systems. It includes: two sub-modules, RIES and CIES. The RIES sub-module perceives the status of each energy sub-system based on the Distflow model of the distribution network, the pressure and flow equations of the natural gas network, and the hydraulic and thermal models of the thermal system. The CIES sub-module monitors the energy conversion status of various devices (such as heat pumps, gas boilers, gas turbines, etc.). The data is transmitted to the data storage module in real time through sensors for subsequent optimization decisions.

[0147] Safety-economic optimization module: This module combines the Stackelberg game and the cooperative game model to realize the energy trading optimization between the RIES and CIES. In the RIES optimization sub-module, the RIES acts as the leader and selects the best energy scheduling scheme through the game with the CIES to ensure the safe operation of the RIES system and improve economic benefits. The CIES optimization sub-module further optimizes the energy use by reducing the operating cost and carbon emissions. This module ensures that the RIES and CIES systems can respond to faults or extreme events in real time during the dynamic operation process.

[0148] Offline Domain Generation Module: In the offline stage, this module generates the critical regions for the operation of the CIES system by analyzing the changes in iterative and timing parameters and based on the state similarity method. By dividing scenarios, iterations, and model parameters into different domains, it generates scenario domains, iteration domains, and model domains, covering multiple optimization problems within the CIES system. Once the critical regions are generated, the optimization problems can be quickly solved through linear equations, reducing the complexity of online optimization.

[0149] Online Domain Optimization Module: During the operation of the CIES system, this module simplifies complex optimization problems into linear solutions by identifying the critical regions to which the current state belongs. When it is identified that the system is in a certain critical region, it directly matches the corresponding critical region and quickly obtains the optimal solution through a system of linear equations without the need for further optimization operations. This optimization method significantly improves the operating efficiency of the system and ensures that the energy system remains efficient and stable during real-time operation.

[0150] In this study, the RIES test system was constructed using a modified IEEE 33-bus distribution system, a 20-node gas system in Belgium, and a 6-node thermal system. The test system includes three CIESs. Each CIES includes a ground-source heat pump, an air-source heat pump, a gas boiler, a gas turbine, a compressed-air chiller, and an electric chiller to meet the regional demands for heat, electricity, and hydrogen. The IDR is set to 20% of the load demand.

[0151] To verify the effectiveness of the proposed REO-RMC framework, three scenarios were set for comparative analysis:

[0152] P1: The Stackelberg game between the RIES and multiple CIESs includes internal energy sharing but does not utilize cooperative games or consider demand response.

[0153] P2: The Stackelberg game between the RIES and the CIES coalition includes internal energy sharing, uses cooperative game theory for profit distribution, but does not consider demand response.

[0154] P3: On the basis of P3, the IDR strategy is further incorporated.

[0155] Through the cooperative game among multiple RIESs, the energy operation cost can be significantly reduced. Specifically, compared with P2, the operation cost of P1 is reduced by 12,003.7 yuan. More precisely, the costs of CIES1, CIES2, and CIES 3 are reduced by 7,702.82 yuan, 438.97 yuan, and 3,861.91 yuan respectively. In P3, the operation cost is further reduced by the IDR strategy of multiple CIESs. Compared with P2, the operation costs of CIES1, CIES2, and CIES 3 are decreased by 1,708.15 yuan, 1,074.48 yuan, and 1,523.47 yuan respectively. In addition, both the cooperative game strategy and the IDR strategy improve the economic benefits of RIESs and the operation costs of CIESs. Compared with P1, the RIES revenues of P3 and P3 increase by 5.58% and 7.73% respectively.

[0156] Table 1 Comparison of Economic and Technical Indicators

[0157]

[0158]

[0159] (1) The energy support strategy based on price incentive developed by combining the Stackelberg game and the cooperative game in the embodiments of the present invention enables the CIES to provide more than 50% energy support for the RIES in the fault scenario, significantly improving the operation reliability of the RIES. At the same time, this strategy effectively balances the interests of energy suppliers, increasing the economic benefits of the RIES by 7.73% and the economic benefits of the CIES by 16.4%, overall enhancing the economic benefits of the RIES and CIES systems.

[0160] (2) The embodiments of the present invention can effectively reduce the reliability safety risk cost of the RIES by 50.23% by combining the Stackelberg game; after introducing the price incentive strategy, the reliability safety risk can be further reduced by about 20%; the energy sharing strategy among CIES systems can reduce the reliability risk among CIESs by about 8%.

[0161] (3) The method proposed in the embodiments of the present invention reduces the calculation time of the bilevel optimization problem to less than 60% of the original method. Through the combination of offline critical region generation and online critical region identification framework, the solution efficiency of the proposed method is further improved by about 50%. As the number of iterative scenarios and time series scenarios increases, this efficiency advantage becomes more obvious.

[0162] Example 4

[0163] An energy system optimization device based on the optimization of the offline domain and the online domain, the device includes: a processor and a memory, and program instructions are stored in the memory, and the processor calls the program instructions stored in the memory to enable the device to perform the following operations:

[0164] Monitor the operating status and environmental changes of the multi-energy system in real time, and provide instant data support for the optimization decisions of the RIES system and the CIES system;

[0165] Combined with the Stackelberg game and the cooperative game model, construct an optimization model for the energy scheduling and resource allocation scheme;

[0166] Based on the multi-parameter programming theory, construct the matrix form of the optimization model for the energy scheduling and resource allocation scheme, analyze the correlation of scenarios, iterations, and model parameters in the CIES system, and generate the scenario domain, iteration domain, and model domain of the optimization problem based on the critical region;

[0167] According to the instant data of the state perception module, identify the scenario domain, iteration domain, and model domain corresponding to the current operating state. By matching the corresponding regions, when it is identified that the operating state of the current CIES system is in a certain critical region, solve it through a linear equation;

[0168] Based on the solution results of the linear equation of the online domain optimization module, determine the energy scheduling and resource allocation scheme.

[0169] Among them, the offline domain generation module is:

[0170] Iterative correlation:

[0171] min f=(c + △c) T x

[0172] s.t. Ax = b x≥0

[0173] Among them, Δc is a random variable related to c;

[0174] Time correlation:

[0175] min f = c T x

[0176] s.t. Ax = b + △b x≥0

[0177] Among them, Δb is a random variable related to b;

[0178] Model correlation:

[0179] min f = c T x

[0180] s.t. (A + △A)x = b x≥0

[0181] where ΔA is a random variable related to A;

[0182] Define the iteration domain R related to CIES, Δc, Δb, and ΔA c , the scenario domain R b , the model domain R A as follows:

[0183] R c = {(△c)|c N +△c N = A N T (A B -1 ) T (c B +△c B )}

[0184] R b = {(△b)|A B -1 (b + △b) ≥ 0}

[0185]

[0186] where B is the critical region activation flag, N is the non-critical region flag; c N represents the non-critical region vector corresponding to c; A N represents the non-critical region matrix corresponding to A; c B represents the critical region vector corresponding to c; A B represents the critical region matrix corresponding to A; △c N represents the non-critical region vector corresponding to Δc; △A N represents the non-critical region matrix corresponding to ΔA; △c B represents the critical region vector corresponding to Δc; △A B represents the critical region matrix corresponding to ΔA.

[0187] Among them, the online domain optimization module is:

[0188] Iterative correlation criterion:

[0189]

[0190] Temporal correlation criterion:

[0191]

[0192] Model similarity criterion:

[0193] c N T -c B (AB +△A B ) -1 (A N +△A N )≤0

[0194] (A B +△A B ) -1 b≥0

[0195] The offline domain space consists of several critical regions, and each critical region contains several optimization problems with the same parameters B and N. The mapping relationship between the optimization problems and the critical regions is constructed through the online domain optimization module to determine the corresponding critical region of the current optimization problem in the offline domain generation module.

[0196] It should be noted here that the device description in the above embodiments corresponds to the method description in the embodiments, and the embodiments of the present invention will not be elaborated herein.

[0197] The execution subjects of the above-mentioned processor and memory can be devices with computing functions such as a computer, a single-chip microcomputer, and a microcontroller. In specific implementation, the embodiments of the present invention do not limit the execution subject, and it is selected according to the needs in actual applications.

[0198] Data signals are transmitted between the memory and the processor through a bus, and the embodiments of the present invention will not be elaborated herein.

[0199] Based on the same inventive concept, the embodiments of the present invention also provide a computer-readable storage medium. The storage medium includes a stored program, and when the program runs, it controls the device where the storage medium is located to execute the method steps in the above embodiments.

[0200] The computer-readable storage medium includes but is not limited to flash memory, hard disk, solid-state drive, etc.

[0201] It should be noted here that the description of the readable storage medium in the above embodiments corresponds to the method description in the embodiments, and the embodiments of the present invention will not be elaborated herein.

[0202] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on the computer, the processes or functions according to the embodiments of the present invention are generated in whole or in part.

[0203] The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted through a computer-readable storage medium. The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or a data center that includes one or more integrated available media. The available medium can be a magnetic medium or a semiconductor medium, etc. In the embodiments of the present invention, except for special descriptions of the models of each device, the models of other devices are not limited, as long as the devices can perform the above functions.

[0204] Those skilled in the art can understand that the drawings are only schematic diagrams of a preferred embodiment, and the serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0205] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An energy system optimization method based on offline domain and online domain optimization, characterized in that: The method comprises: Build a state perception module to monitor the operating status and environmental changes of multi-energy systems in real time, and provide instant data support for RIES and CIES system optimization decisions; Construct a safety-economy optimization module, combine Stackelberg game and cooperative game strategies, and build an energy scheduling and resource allocation optimization model; Construct an offline domain generation module, construct a matrix form of the energy scheduling and resource allocation scheme optimization model based on multi-parameter planning theory, analyze the correlation between scenarios, iterations and model parameters in the CIES system, and generate scenario domains, iteration domains and model domains of the optimization problem based on critical regions; An online domain optimization module is constructed to identify the scenario domain, iteration domain, and model domain corresponding to the current operating state according to the real-time data of the state perception module during the operation of the CIES system. By matching the corresponding areas, when it is identified that the operating state of the current CIES system is in a critical area, the linear equation is solved, and the energy scheduling and resource allocation schemes are determined based on the linear equation solution results of the online domain optimization module.

2. The energy system optimization method based on offline domain and online domain optimization according to claim 1 is characterized in that: The offline domain generation module is: Iteration Dependencies: min f=(c+△c) T x stAx=bx≥0 Where Δc is a random variable related to c; Time Dependency: min f=c T x stAx=b+△bx≥0 Where Δb is a random variable related to b; Model Dependencies: min f=c T x st(A+△A)x=bx≥0 Where ΔA is a random variable related to A; Formulate the iteration domain R that relates CIES to Δc, Δb, and ΔA c , scene domain R b , model domain R A It is expressed as follows: R c ={(△c)|c N +△c N =A N T (A B -1 ) T (c B +△c B )} R b ={(△b)|A B -1 (b+△b)≥0} Among them, B is the critical area activation mark, N is the non-critical area mark; c N represents the non-critical region vector corresponding to c; A N represents the non-critical region matrix corresponding to A; c B represents the critical region vector corresponding to c; A B represents the critical region matrix corresponding to A; △c N represents the non-critical region vector corresponding to Δc; ΔA N represents the non-critical area matrix corresponding to ΔA; △c B represents the critical region vector corresponding to Δc; ΔA B Represents the critical region matrix corresponding to ΔA.

3. The energy system optimization method based on offline domain and online domain optimization according to claim 2 is characterized in that: The online domain optimization module is: Iterative correlation criterion: Time correlation criterion: Model similarity criteria: c N T -c B (A B +△A B ) -1 (A N +△A N )≤0 (A B +△A B ) -1 b≥0 The offline domain space consists of several critical regions, each of which contains several optimization problems with the same parameters B and N. The mapping relationship between the optimization problem and the critical region is constructed in the online domain optimization module to determine the corresponding critical region of the current optimization problem in the offline domain generation module.

4. An energy system optimization device based on offline domain and online domain optimization, characterized in that: The device comprises: a processor and a memory, wherein the memory stores program instructions, and the processor calls the program instructions stored in the memory to enable the device to perform the following operations: Real-time monitoring of the operating status and environmental changes of multi-energy systems, providing instant data support for RIES and CIES system optimization decisions; Combining Stackelberg game and cooperative game models, an energy scheduling and resource allocation optimization model is constructed; Based on the multi-parameter planning theory, the matrix form of the energy scheduling and resource allocation scheme optimization model is constructed, the scenarios, iterations and correlations of model parameters in the CIES system are analyzed, and based on the critical regions, the scenario domain, iteration domain and model domain of the optimization problem are generated; According to the real-time data of the state perception module, the scenario domain, iteration domain and model domain corresponding to the current operating state are identified. By matching the corresponding areas, when it is identified that the operating state of the current CIES system is in a critical area, a linear equation is used to solve it. Based on the linear equation solution results of the online domain optimization module, the energy scheduling and resource allocation schemes are determined.

5. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor executes the method according to any one of claims 1 to 3.

Images