Element collaboration strategy optimization method and system of electricity-gas-heat comprehensive energy system

Abstract

The invention provides an element collaboration strategy optimization method and system of an electricity-gas-heat comprehensive energy system, and belongs to the technical field of comprehensive energy system analysis. The method comprises the steps that power grid equipment elements, heat supply network equipment elements, gas network equipment elements, system dimension elements and operation dimension elements of the electricity-gas-heat comprehensive energy system are screened, a to-be-screened element set is constructed, when multiple implementation schemes exist in the elements, each implementation scheme is regarded as an independent element to be added into the to-be-screened element set, and the to-be-screened element set is constructed; calculating the reliability improvement sensitivity of all the elements and the required economic cost; constructing a multi-objective optimization model by taking maximization of reliability improvement sensitivity of all elements and minimization of economic cost as objectives; and solving the multi-objective optimization model to obtain a Pareto optimal solution set, and screening out a final reliability improvement scheme from the Pareto optimal solution set through multi-agent collaborative decision based on centralized training. And the comprehensiveness and systematicness of the reliability evaluation index of the integrated energy system are improved.

Description

Technical Field

[0001] This invention belongs to the field of integrated energy system analysis technology, and more specifically, relates to a method and system for optimizing the element coordination strategy of an integrated energy system of electricity, gas and heat. Background Technology

[0002] Existing technologies often employ mathematical linear programming models when establishing reliability assessment indicators to quantify the indicators that affect the probability of eventual failure. They also combine Monte Carlo simulations to determine the range of influence coefficients of the indicators and the accuracy of the assessment model. For example, when assessing the reliability of heating systems, existing technologies introduce probability models of failure and user comfort indicators, quantifying the user and system interaction experience in the final objective function of system reliability assessment.

[0003] Existing technologies utilize non-sequential Monte Carlo models to assess the reliability of multi-microgrid systems in active power distribution systems. However, model-driven methods suffer from limitations due to the excessive number of scenarios introduced; as the system hierarchy expands, the number of scenarios increases exponentially. This leads to significant shortcomings when dealing with complex integrated energy systems, making it difficult to guarantee computational speed and accuracy. Furthermore, with the increasing penetration of renewable energy and the diversification of load types in my country, traditional scenario generation models struggle to meet the requirements of integrated energy system reliability assessment models, further highlighting their limitations.

[0004] Data-driven models have certain disadvantages. Besides requiring a high quantity, accuracy, and type of historical operational data, their adaptability is limited. After modifying and maintaining an energy system, they cannot accurately assess the reliability of the current system. Furthermore, a trained model is often only applicable to its specific energy system, meaning the objective function lacks general applicability. Additionally, since most of the resulting models are nonlinear, there is a lack of sufficient technical guidance in explaining the mechanisms of parameter influence.

[0005] Existing technical document 1 (CN109784569A) discloses an optimized control method for a regional integrated energy system. Its shortcoming is that it does not take into account reliability requirements such as equipment aging, external environmental influences, fault repair time, and system backup. It only models the system's operating constraints under normal working conditions and ignores the reliability assessment requirements under extreme environments.

[0006] Prior art document 2 (CN110990785A) discloses a multi-objective-based optimization scheduling method for a smart park integrated energy system. Its shortcomings lie in that, although it quantifies the cost of equipment operation and maintenance, it still fails to address the issue of dedicated energy supply due to equipment failure or aging, and does not conduct an in-depth analysis of this. Furthermore, this invention focuses on economic optimization based on system safety and stability, and while it provides preliminary reliability index analysis for each energy source, it does not conduct a comprehensive evaluation of the overall system reliability. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a reliability improvement strategy model applicable to integrated electric-gas-heat energy systems. The purpose of establishing this model is to select the optimal factors for reliability improvement.

[0008] The present invention adopts the following technical solution.

[0009] The first aspect of the present invention provides a method for optimizing the element coordination strategy of an integrated electric-gas-heat energy system, comprising: Screen the power grid equipment elements, heating network equipment elements, gas network equipment elements, system dimension elements, and operation and maintenance dimension elements of the integrated energy system of electricity, gas, and heat, and construct a set of elements to be screened; When there are multiple implementation schemes for an element in the set of elements to be screened, each implementation scheme is treated as an independent element and added to the set of elements to be screened. The reliability improvement sensitivity and the required economic cost of all elements are calculated. A multi-objective optimization model is constructed with the goal of maximizing the reliability of all factors to improve sensitivity and minimizing economic costs. The NSGA-II algorithm is used to solve the multi-objective optimization model to obtain the Pareto optimal solution set. Based on centralized training, the final reliability improvement scheme is selected from the Pareto optimal solution set through multi-agent collaborative decision-making.

[0010] Preferably, calculating the reliability improvement sensitivity of all elements includes: The sensitivity of input factor reliability improvement is obtained by subtracting the reliability index after input factor from the reliability index before input factor, and then dividing the difference by the reliability index before input factor.

[0011] Preferably, the first objective function of the multi-objective optimization model is to maximize the reliability improvement sensitivity of all factors, expressed by the following formula:

[0012] In the formula, This represents an indicator of the overall benefits of improved reliability. This indicates the maximum number of pre-selected elements that can be set. Represents the saturation coefficient, 0≤ <1, This indicates that the reliability of element i is improved by enhancing its sensitivity. Indicates the total number of elements. Represents decision variables, =1 indicates that input element i is selected in the final optimized solution. =0 indicates that element i is not selected in the final optimized solution.

[0013] Preferably, the second objective function of the multi-objective optimization model is to minimize the economic cost, expressed by the following formula:

[0014] In the formula, Indicators representing economic cost This represents the total economic cost required for factor i. This represents the cost weight of element i.

[0015] Preferably, the multi-objective optimization model uses mutual exclusion constraints as constraints, expressed by the following formula:

[0016] In the formula, This represents the k-th element group, which describes a set of different implementation schemes for the same reliability enhancement element. This means that for the same element, at most one implementation scheme can be selected.

[0017] Preferably, based on centralized training, the final reliability improvement scheme is selected from the Pareto optimal solution set through multi-agent collaborative decision-making, including: Define electric, gaseous, and thermal intelligent agents and their state space, action space, and reward function; The MADDPG algorithm is used to obtain the trained agent policy network by centrally training the agent's policy network and critic network to maximize the expected cumulative reward. For each solution, a comprehensive score is independently generated for each agent based on the trained agent policy network; The Pareto solution with the highest overall score was selected as the final reliability improvement scheme.

[0018] Preferably, the reward function is expressed by the following formula:

[0019] In the formula, , and These represent the weighting coefficients for reliability, economy, and synergy, respectively. Indicates the synergy index; This represents the state vector corresponding to the j-th solution. , Representing the solution The overall benefit index of reliability improvement Representing the solution The economic cost indicators Describe the solutions respectively Select sub-vectors for electrical, gas, and thermal elements; Representing the solutions of electric, gaseous, and thermal intelligent agents respectively The rating is 0 for disapproval and 1 for approval.

[0020] Preferably, The synergy index is expressed by the following formula:

[0021] In the formula, , This represents the average score of the three agents.

[0022] Preferably, the comprehensive score is expressed by the following formula:

[0023] In the formula, Solution Overall score , , Represents the weights of electric, thermal, and gaseous intelligent agents. This represents the weight of the synergistic effect.

[0024] The second aspect of the present invention provides a component coordination strategy optimization system for an integrated electric-gas-heat energy system, which runs the component coordination strategy optimization method for an integrated electric-gas-heat energy system described in the first aspect, including: The module for generating a set of elements to be screened is used to screen the grid equipment elements, heating network equipment elements, gas network equipment elements, system dimension elements, and operation and maintenance dimension elements of the integrated energy system of electricity, gas and heat, and to construct a set of elements to be screened. The sensitivity and economic cost solution module is used when there are multiple implementation schemes for the elements in the set of elements to be screened. Each implementation scheme is treated as an independent element and added to the set of elements to be screened. The module calculates the reliability improvement sensitivity and the required economic cost for all elements. The model building module is used to build a multi-objective optimization model with the goal of maximizing the reliability of all factors to improve sensitivity and minimizing economic costs. The solution-solving module is used to solve the multi-objective optimization model using the NSGA-II algorithm to obtain the Pareto optimal solution set. Based on centralized training, the final reliability improvement scheme is selected from the Pareto optimal solution set through multi-agent collaborative decision-making.

[0025] Compared with the prior art, the beneficial effects of the present invention include at least the following: This invention addresses the problem in existing technologies that fail to consider reliability requirements such as equipment aging, external environmental impact, fault repair time, and system backup by constructing a multi-dimensional quantitative screening system encompassing equipment, system, and operation and maintenance levels. It establishes a comprehensive reliability assessment standard covering both normal and extreme fault environments, enhancing the comprehensiveness and systematic nature of integrated energy system reliability assessment indicators. Through quantitative screening across equipment, system, and operation and maintenance dimensions, it establishes a complete assessment standard covering both normal and extreme fault scenarios, reducing the risk of blind spots and biases in reliability assessments caused by neglecting key factors (such as equipment aging, external environment, and fault repair time), and reducing over-reliance on single-dimensional indicators. This invention improves sensitivity and economic cost by quantifying the reliability of each element, solving the problems of excessive reliance on historical data, low adaptability, and poor generalization of data-driven models. It establishes a general evaluation framework that can adapt to system modification, improving the adaptability, transparency, and generalization ability of the reliability evaluation model. By accurately quantifying the sensitivity and cost of each element, a general evaluation framework that is not bound by specific historical datasets is established, reducing the model's dependence on massive amounts of high-precision historical data, as well as the cost and difficulty of retraining the model after system modification or upgrade. This invention establishes a multi-objective optimization model with the goals of maximizing reliability improvement and minimizing economic cost, and solves the problem using the NSGA-II algorithm. This solves the "curse of dimensionality" problem caused by too many scenarios in the model-driven method, improving the solution efficiency while ensuring computational accuracy. It enhances the solution efficiency and decision effectiveness of multi-objective optimization problems in complex scenarios. By using the NSGA-II algorithm, the trade-off between reliability improvement and economic cost is effectively handled while ensuring the accuracy of the Pareto solution set. This reduces the computational complexity and computational resource consumption caused by the explosion of scenario combinations in the model-driven method, and avoids the difficult trade-off between solution speed and accuracy in traditional methods. This invention utilizes a multi-agent collaborative decision-making framework based on CTDE to collaboratively evaluate the electrical, gas, and thermal subsystems as agents. This addresses the problem in existing technologies that lack comprehensive evaluation of overall system reliability, achieving a comprehensive assessment considering multi-energy coupling and improving the overall synergy, practical operability, and global optimality of the final decision scheme. Through multi-agent collaborative evaluation, a balance of interests among the electrical, gas, and thermal subsystems is achieved, ensuring the overall optimality of the solution at the system level. This reduces subjective arbitrariness in the decision-making process and local conflicts between the optimization objectives of various subsystems, avoiding the problem of low overall efficiency due to a lack of collaborative mechanisms. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the element coordination strategy optimization process of the integrated energy system of electricity, gas and heat provided in accordance with the embodiments of the present invention; Figure 2 This is a schematic diagram of the reliability improvement elements of a multi-dimensional integrated electric-gas-heat energy system provided in accordance with embodiments of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0028] like Figure 1 As shown, Embodiment 1 of the present invention provides a reliability improvement strategy and method for an integrated electric-gas-heat energy system, comprising the following steps: Step 1: Screen the grid equipment elements, heating network equipment elements, gas network equipment elements, system-dimensional elements, and operation and maintenance-dimensional elements of the integrated electricity-gas-heat energy system to construct a set of elements to be screened, such as... Figure 2 As shown.

[0029] Step 1.1: Obtain the power grid equipment elements, heating network equipment elements, and gas network equipment elements.

[0030] From an equipment perspective, the goal is to extend equipment lifespan while considering economic efficiency, and optimize system reliability through load sharing. The objective function is to minimize the overall economic loss of the system caused by load reduction, expressed by the following formula:

[0031] In the formula, These are the sets of electrical load nodes, thermal load nodes, and gas load nodes in the IES, respectively. The unit load reduction costs for electrical load node i, gas load node k, and heat load node l are respectively. These are the importance coefficients of the corresponding load nodes. These represent the load reduction amounts for the corresponding nodes.

[0032] To extend equipment lifespan, additional operational constraints related to electricity, heat, and gas networks are imposed on the equipment in addition to economic targets.

[0033] In the formula, They provide power from electricity, heat, and gas sources, respectively. The power outputs are respectively from the electrical, thermal, and pneumatic coupling elements; These are the normal load values, The total load to be reduced is as follows: These are network losses.

[0034] Inequality constraints on the power grid side:

[0035] In the formula, Indicates the node voltage amplitude. Indicates the amplitude of the branch current; , This represents the active and reactive power output of generator at node i; and These represent the upper and lower limits of the node voltage amplitude, respectively. and These represent the upper and lower limits of the branch current amplitude, respectively. and These represent the upper and lower limits of the active power output of generator i, respectively. and These represent the upper and lower limits of the reactive power output of generator i, respectively.

[0036] Inequality constraints on the heating network side:

[0037] In the formula, , Indicates the supply and regeneration temperatures; Indicates the mass flow rate of the pipeline; Indicates the amount of heat load reduction; This represents the steady-state average heat load.

[0038] Gas network side inequality constraints:

[0039] In the formula, This represents the squared value of the nodal pressure. Indicates the gas flow rate in the pipeline. This refers to the gas flow rate of the compressor. This refers to the compressor's boost ratio. For gas load Reduction amount, This represents the average gas load.

[0040] Step 1.2: Filter system dimension elements, which include: the grid planning structure, system reserve level, line connection mode, and line power supply radius; The grid structure of the heating network, system reserve level, load classification and zoning, and supply and return temperature difference flow rate are planned. The grid is divided and classified in a reasonable manner according to the load distribution. The supply and return heat temperature and temperature difference of each node of the heating network, and the mass flow rate level of the overall network are set. The planning structure of the gas network, the system backup level, the compressor configuration scheme and the gas pressure and flow settings, the reasonable selection and setting of compressor locations, and the appropriate node gas pressure and pipeline flow.

[0041] Step 1.3: Screening Operation and Maintenance Dimensions. Operation and maintenance dimensions include external environmental interference to the power grid, such as extreme weather, human damage, flora and fauna, pollution, etc.; the automation level of the power system, the level of emergency repair operations, and the comprehensive power outage management of the dispatching department; the regional ambient temperature of the heating network, the regional heating standards, the heat reserve coefficient of buildings, and comprehensive heating outage management; and the external environmental interference of the gas network, the regional gas supply standards, i.e., the setting of the minimum allowable gas pressure at nodes, the gas storage performance of pipelines after a fault, and the comprehensive gas outage management level after a fault, such as valve on / off control, ring network airflow control, and compressor operating status adjustment.

[0042] Step 1.4: Generate a set of elements to be screened based on the power grid equipment elements, heating network equipment elements, gas network equipment elements, system dimension elements, and operation and maintenance dimension elements. The power grid equipment elements include main transformers, the heating network equipment elements include regional heating pipelines, the gas network equipment elements include natural gas pressure regulating stations, the system dimension elements include network topology optimization strategies, and the operation and maintenance dimension elements include preventive maintenance strategies.

[0043] Step 2: If an element has multiple implementation schemes, such as, but not limited to, compressors having two models, A and B, then each implementation scheme is treated as an independent element, added to the set of elements to be screened, and the reliability improvement sensitivity and required economic cost of all elements are calculated.

[0044] The reliability improvement sensitivity of the input factor is obtained by subtracting the reliability index after the input factor from the reliability index before the input factor, and then dividing the difference by the reliability index before the input factor; the input cost of the factor includes the entire life cycle cost of equipment purchase, installation, operation and maintenance.

[0045] Step 3: Construct a multi-objective optimization model based on the reliability, sensitivity, and required economic costs of all factors.

[0046] Step 3.1: Construct the first objective function with the goal of maximizing reliability to improve sensitivity, expressed by the following formula:

[0047] In the formula, This represents an indicator of the overall benefits of improved reliability. This indicates the maximum number of pre-selected elements that can be set. Represents the saturation coefficient, 0≤ <1 is used to quantify the diminishing marginal returns that may occur when multiple factors are input simultaneously. This indicates that the reliability of element i is improved by enhancing its sensitivity. Indicates the total number of elements. Represents decision variables, =1 indicates that in the final optimized solution, input element i is selected. =0 indicates that element i is not selected in the final optimized solution.

[0048] Step 3.2: Construct a second objective function with the goal of minimizing economic costs, expressed by the following formula:

[0049] In the formula, Indicators representing economic cost This represents the total economic cost required for factor i. This represents the cost weight of element i.

[0050] Step 3.3, using mutual exclusion constraints as the constraint conditions, is expressed by the following formula:

[0051] In the formula, This represents the k-th element group, which is a set of different implementation schemes for the same reliability improvement element. For example, for the element of compressor upgrade, different compressor models A, B, and C constitute an element group. , This means that for the same element, at most one implementation scheme can be selected.

[0052] Step 4: The NSGA-II algorithm is used to solve the multi-objective optimization model to obtain the Pareto optimal solution set. Based on CTDE (Centralized Training with Decentralized Execution), the final reliability improvement scheme is selected from the Pareto optimal solution set through multi-agent collaborative decision-making.

[0053] Step 4.1: Solve the multi-objective optimization model using the NSGA-II algorithm to output the Pareto optimal solution set P={v1, v2, ..., v...} m}, each solution v in the set j =(v j1 v j2 , ..., v jn ) is a feature selection scheme, v j It is the j-th Pareto optimal solution, v ji Is the solution v j The selection state of the i-th element is 0, where 0 indicates no selection and 1 indicates selection. m represents the number of Pareto solutions and n represents the total number of elements.

[0054] Step 4.2: Define the electric, gaseous, and thermal intelligent agents and their state space, action space, and reward function.

[0055] Define a set of intelligent agents ,in, These represent intelligent agents for the power grid, gas grid, and heating grid, respectively, with e, g, and h representing subsystems of the power grid, gas grid, and heating grid, respectively. Constructing the state space ,in, This represents the state vector corresponding to the j-th solution. Solution The overall benefit index of reliability improvement Solution The economic cost indicators Solution Select sub-vectors for electrical, gas, and thermal elements; Action space ,in, Represents intelligent agents Solution The rating is 0 for disapproval and 1 for approval. The reward function is expressed by the following formula:

[0056] In the formula, , and These represent the weighting coefficients for reliability, economy, and synergy, respectively. Representing the solutions of electric, gaseous, and thermal intelligent agents respectively The rating is 0 for disapproval and 1 for approval. The coordination index, used to measure the consistency of agent ratings, is expressed by the following formula:

[0057] In the formula, , This represents the average score of the three agents; Step 4.3: Using the MADDPG algorithm, the agent's policy network and critic network are trained in a centralized manner to maximize the expected cumulative reward, thus obtaining the trained agent policy network.

[0058] The agent policy network is trained using the MADDPG algorithm. and the network of critics ,in, Represents intelligent agents A policy network is used to generate scoring actions. This represents a centralized network of critics used to evaluate the value of collaborative actions. The optimized expected cumulative reward is expressed by the following formula:

[0059] In the formula, Policy networks representing electric, gaseous, and thermal intelligent agents. This indicates the expected cumulative reward.

[0060] Step 4.4, for each solution Based on the trained agent policy network, a comprehensive score is independently generated for each agent, expressed by the following formula:

[0061] In the formula, Representing state China's intelligent agents Observations.

[0062] The overall score is calculated using the following formula:

[0063] In the formula, Solution Overall score , , Represents the weights of electric, thermal, and gaseous intelligent agents. This indicates the weight of the synergy effect, controlling the degree of influence of synergy on the overall score.

[0064] Step 4.5: Select the Pareto solution with the highest overall score as the final reliability improvement scheme. It can be expressed by the following formula:

[0065] Embodiment 2 of the present invention provides a component coordination strategy optimization system for an integrated electric-gas-heat energy system, which runs the component coordination strategy optimization method for an integrated electric-gas-heat energy system described in Embodiment 1, including: The module for generating a set of elements to be screened is used to screen the grid equipment elements, heating network equipment elements, gas network equipment elements, system dimension elements, and operation and maintenance dimension elements of the integrated energy system of electricity, gas and heat, and to construct a set of elements to be screened. The sensitivity and economic cost solution module is used when there are multiple implementation schemes for the elements in the set of elements to be screened. Each implementation scheme is treated as an independent element and added to the set of elements to be screened. The module calculates the reliability improvement sensitivity and the required economic cost for all elements. The model building module is used to build a multi-objective optimization model with the goal of maximizing the reliability of all factors to improve sensitivity and minimizing economic costs. The solution-solving module is used to solve the multi-objective optimization model using the NSGA-II algorithm to obtain the Pareto optimal solution set. Based on centralized training, the final reliability improvement scheme is selected from the Pareto optimal solution set through multi-agent collaborative decision-making.

[0066] Compared with the prior art, the beneficial effects of the present invention include at least the following: This invention addresses the problem in existing technologies that fail to consider reliability requirements such as equipment aging, external environmental impact, fault repair time, and system backup by constructing a multi-dimensional quantitative screening system encompassing equipment, system, and operation and maintenance levels. It establishes a comprehensive reliability assessment standard covering both normal and extreme fault environments, enhancing the comprehensiveness and systematic nature of integrated energy system reliability assessment indicators. Through quantitative screening across equipment, system, and operation and maintenance dimensions, it establishes a complete assessment standard covering both normal and extreme fault scenarios, reducing the risk of blind spots and biases in reliability assessments caused by neglecting key factors (such as equipment aging, external environment, and fault repair time), and reducing over-reliance on single-dimensional indicators. This invention improves sensitivity and economic cost by quantifying the reliability of each element, solving the problems of excessive reliance on historical data, low adaptability, and poor generalization of data-driven models. It establishes a general evaluation framework that can adapt to system modification, improving the adaptability, transparency, and generalization ability of the reliability evaluation model. By accurately quantifying the sensitivity and cost of each element, a general evaluation framework that is not bound by specific historical datasets is established, reducing the model's dependence on massive amounts of high-precision historical data, as well as the cost and difficulty of retraining the model after system modification or upgrade. This invention establishes a multi-objective optimization model with the goals of maximizing reliability improvement and minimizing economic cost, and solves the problem using the NSGA-II algorithm. This solves the "curse of dimensionality" problem caused by too many scenarios in the model-driven method, improving the solution efficiency while ensuring computational accuracy. It enhances the solution efficiency and decision effectiveness of multi-objective optimization problems in complex scenarios. By using the NSGA-II algorithm, the trade-off between reliability improvement and economic cost is effectively handled while ensuring the accuracy of the Pareto solution set. This reduces the computational complexity and computational resource consumption caused by the explosion of scenario combinations in the model-driven method, and avoids the difficult trade-off between solution speed and accuracy in traditional methods. This invention utilizes a multi-agent collaborative decision-making framework based on CTDE to collaboratively evaluate the electrical, gas, and thermal subsystems as agents. This addresses the problem in existing technologies that lack comprehensive evaluation of overall system reliability, achieving a comprehensive assessment considering multi-energy coupling and improving the overall synergy, practical operability, and global optimality of the final decision scheme. Through multi-agent collaborative evaluation, a balance of interests among the electrical, gas, and thermal subsystems is achieved, ensuring the overall optimality of the solution at the system level. This reduces subjective arbitrariness in the decision-making process and local conflicts between the optimization objectives of various subsystems, avoiding the problem of low overall efficiency due to a lack of collaborative mechanisms.

[0067] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0068] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for optimizing the synergistic strategy of elements in an integrated electric-gas-heat energy system, characterized in that: Screen the power grid equipment elements, heating network equipment elements, gas network equipment elements, system dimension elements, and operation and maintenance dimension elements of the integrated energy system of electricity, gas, and heat, and construct a set of elements to be screened; When there are multiple implementation schemes for an element in the set of elements to be screened, each implementation scheme is treated as an independent element and added to the set of elements to be screened. The reliability improvement sensitivity and the required economic cost of all elements are calculated. A multi-objective optimization model is constructed with the goal of maximizing the reliability of all factors to improve sensitivity and minimizing economic costs. The NSGA-II algorithm is used to solve the multi-objective optimization model to obtain the Pareto optimal solution set. Based on centralized training, the final reliability improvement scheme is selected from the Pareto optimal solution set through multi-agent collaborative decision-making.

2. The method for optimizing the element coordination strategy of an integrated electric-gas-heat energy system according to claim 1, characterized in that: The reliability enhancement sensitivity calculation for all elements includes: The sensitivity of input factor reliability improvement is obtained by subtracting the reliability index after input factor from the reliability index before input factor, and then dividing the difference by the reliability index before input factor.

3. The method for optimizing the element coordination strategy of an integrated electric-gas-heat energy system according to claim 1, characterized in that: The first objective function of the multi-objective optimization model is to maximize the reliability improvement sensitivity of all factors, expressed by the following formula: In the formula, This represents an indicator of the overall benefits of improved reliability. This indicates the maximum number of pre-selected elements that can be set. Represents the saturation coefficient, 0≤ < 1, This indicates that the reliability of element i is improved by enhancing its sensitivity. Indicates the total number of elements. Represents decision variables, =1 indicates that input element i is selected in the final optimized solution. =0 indicates that element i is not selected in the final optimized solution.

4. The method for optimizing the element coordination strategy of an integrated electric-gas-heat energy system according to claim 3, characterized in that: The second objective function of the multi-objective optimization model is to minimize the economic cost, expressed by the following formula: In the formula, Indicators representing economic cost This represents the total economic cost required for factor i. This represents the cost weight of element i.

5. The method for optimizing the element coordination strategy of an integrated electric-gas-heat energy system according to claim 4, characterized in that: The multi-objective optimization model uses mutual exclusion constraints as conditions, and is expressed by the following formula: In the formula, This represents the k-th element group, which describes a set of different implementation schemes for the same reliability enhancement element. This means that for the same element, at most one implementation scheme can be selected.

6. The method for optimizing the element coordination strategy of an integrated electric-gas-heat energy system according to claim 1, characterized in that: Based on centralized training, the final reliability improvement scheme is selected from the Pareto optimal solution set through multi-agent collaborative decision-making, including: Define electric, gaseous, and thermal intelligent agents and their state space, action space, and reward function; The MADDPG algorithm is used to obtain the trained agent policy network by centrally training the agent's policy network and critic network to maximize the expected cumulative reward. For each solution, a comprehensive score is independently generated for each agent based on the trained agent policy network; The Pareto solution with the highest overall score was selected as the final reliability improvement scheme.

7. The method for optimizing the element coordination strategy of an integrated electric-gas-heat energy system according to claim 6, characterized in that: The reward function is expressed by the following formula: In the formula, , and These represent the weighting coefficients for reliability, economy, and synergy, respectively. Indicates the synergy index; This represents the state vector corresponding to the j-th solution. , Representing the solution The comprehensive benefit index of reliability improvement Solution The economic cost indicators Describe the solutions respectively Select sub-vectors for electrical, gas, and heat elements; Representing the solutions of electric, gaseous, and thermal intelligent agents respectively The rating is 0 for disapproval and 1 for approval.

8. The method for optimizing the element coordination strategy of an integrated electric-gas-heat energy system according to claim 7, characterized in that: The synergy index is expressed by the following formula: In the formula, , This represents the average score of the three agents.

9. The method for optimizing the element coordination strategy of an integrated electric-gas-heat energy system according to claim 8, characterized in that: The overall score is expressed by the following formula: In the formula, Solution Overall score , , Represents the weights of electric, thermal, and gaseous intelligent agents. This represents the weight of the synergistic effect.

10. A component coordination strategy optimization system for an integrated electric-gas-heat energy system, comprising the component coordination strategy optimization method for an integrated electric-gas-heat energy system as described in any one of claims 1-9, characterized in that: The module for generating a set of elements to be screened is used to screen the grid equipment elements, heating network equipment elements, gas network equipment elements, system dimension elements, and operation and maintenance dimension elements of the integrated energy system of electricity, gas and heat, and to construct a set of elements to be screened. The sensitivity and economic cost solution module is used when there are multiple implementation schemes for the elements in the set of elements to be screened. Each implementation scheme is treated as an independent element and added to the set of elements to be screened. The module calculates the reliability improvement sensitivity and the required economic cost for all elements. The model building module is used to build a multi-objective optimization model with the goal of maximizing the reliability of all factors to improve sensitivity and minimizing economic costs. The solution-solving module is used to solve the multi-objective optimization model using the NSGA-II algorithm to obtain the Pareto optimal solution set. Based on centralized training, the final reliability improvement scheme is selected from the Pareto optimal solution set through multi-agent collaborative decision-making.

Images