A regional integrated energy operation optimization method considering flexible adjustment of multi-energy networks

Abstract

This invention discloses a regional integrated energy operation optimization method considering the flexible adjustment of multi-energy networks, belonging to the field of energy operation optimization technology, and includes the following steps: S1, constructing an electric-thermal hybrid energy network cluster and a multi-DIES cluster; S2, constructing a regional distribution network model and a regional heating network model; S3, constructing an electric-thermal hybrid energy network cluster partitioning model based on the regional distribution network model and the regional heating network model; S4, constructing a two-layer collaborative operation optimization model for the multi-DIES cluster; S5, solving the electric-thermal hybrid energy network cluster partitioning model and the multi-DIES cluster two-layer collaborative operation optimization model to complete the optimization. The two-stage optimization method proposed in this invention can reduce the power demand of each DIES on the distribution network during peak load periods, reduce the risk of main transformer overload, and improve the economic efficiency of system operation.

Description

Technical Field

[0001] This invention relates to the field of energy operation optimization technology, and specifically to a regional integrated energy operation optimization method that takes into account the flexible adjustment of multi-energy networks. Background Technology

[0002] The existing integrated energy system is distributed in different nodes in the distribution network and regional heating network. Other electric and heat load nodes are also distributed in the distribution network and regional heating network. These nodes are supplied with heat and electricity through ESP. When the load in the distribution network is large, the power flowing into each distribution area through the main transformer of the distribution network will surge, causing the main transformer to overload. The overload of the main transformer will have a series of adverse consequences, mainly including the following aspects: 1. Increased heat loss and temperature rise of the transformer, specifically (1) Increased heat loss: Overload operation will significantly increase the heat loss of the transformer. This is because the transformer will generate heat due to resistance, core loss and other reasons when it is working. When the load exceeds the rated value, these losses will increase sharply, causing the transformer temperature to rise; (2) Temperature rise: Long-term overload operation will cause the internal temperature of the transformer to rise continuously, exceeding its maximum design allowable temperature. This will not only affect the insulation performance of the transformer, but may also accelerate the aging of the insulation material and shorten the service life of the transformer. 2. Voltage fluctuations and power quality degradation, specifically (1) Voltage drop: Overload operation will cause the transformer output voltage to drop, because when the transformer is overloaded, its internal resistance and leakage reactance will increase, thus affecting the stable output of voltage. Voltage fluctuations will affect the normal operation of power equipment and may even cause equipment failure; (2) Power quality degradation: Overload operation may also cause problems such as increased harmonics and three-phase imbalance, further reducing power quality. 3. Increased safety hazards, specifically (1) Equipment damage risk: Long-term overload operation will cause the internal components of the transformer to bear excessive stress, which may lead to winding deformation, insulation breakdown and other faults, and in severe cases, the transformer will burn out; (2) Fire and explosion risk: High temperature and insulation aging may cause serious accidents such as fire or explosion, posing a threat to personal and property safety. 4. Power grid stability is affected, specifically the risk of local power grid collapse: If the main transformer is severely overloaded and the overload lasts for a long time, it may affect the stability of the entire power grid. In extreme cases, it may lead to local power grid collapse, affecting the power supply of a large area.

[0003] In response to the overload problem of the main transformer in the distribution network, the existing countermeasures include the following aspects: (1) Strengthen equipment monitoring and maintenance: Regularly monitor and maintain key equipment such as transformers, and promptly identify and address potential problems. At the same time, establish a sound emergency plan and fault handling mechanism to deal with emergencies. (2) Improve equipment performance: Select high-performance and high-efficiency transformer equipment to reduce heat loss and improve power quality. At the same time, strengthen the heat dissipation design and maintenance management of equipment to reduce equipment temperature. (3) Strengthen load forecasting and management: Through scientific load forecasting and management, rationally arrange the load distribution of the power grid to avoid a single transformer bearing an excessive load. (4) Optimize the power grid structure: Rationally optimize the power grid structure to improve the flexibility and reliability of the power grid. For example, by adding transformers and adjusting the power grid layout, the load pressure of the main transformer can be shared.

[0004] However, conventional transformer overload switching devices have two major problems in actual operation: 1) They cannot automatically adapt to changes in the substation's operating mode, which may lead to the incorrect disconnection of some load lines; 2) Due to the adoption of an overly simplistic strategy for selecting load lines to be disconnected, unnecessary over-switching is often caused. Summary of the Invention

[0005] To address the above problems, this invention proposes a regional integrated energy operation optimization method that considers the flexible adjustment of multi-energy networks.

[0006] The technical solution of this invention is: a regional integrated energy operation optimization method considering the flexible adjustment of multi-energy networks, comprising the following steps:

[0007] S1. Construct a hybrid electric and thermal energy network cluster and a multi-DIES cluster;

[0008] S2. Construct regional power distribution network model and regional heating network model;

[0009] S3. Based on the regional power distribution network model and the regional heating network model, construct a cluster partitioning model for the hybrid electric and heating energy network;

[0010] S4. Construct a multi-DIES cluster two-layer collaborative operation optimization model;

[0011] S5. Solve the electrothermal hybrid energy network cluster partitioning model and the multi-DIES cluster two-layer collaborative operation optimization model to complete the optimization.

[0012] Furthermore, in S2, the expression for the regional distribution network model is:

[0013] ;

[0014] In the formula, F ij Indicates that by nodei With nodes j The power flow caused by changes in power transmission between them f 1i Indicates from node i The power flow changes on branch 1 caused by the power flowing in and out of the reference node. f 1j Represents a node j The power flow changes on branch 1 caused by the power flowing in and out of the reference node. f 2i Represents a node i The power flow changes on branch 2 caused by the power flowing in and out of the reference node. f 2j Represents a node j The power flow changes on branch 2 caused by the power flowing in and out of the reference node. f li Indicates from node i Branches caused by power flowing in and out of the reference node l Trend changes on the surface f lj Indicates from node j Branches caused by power flowing in and out of the reference node l Trend changes;

[0015] In S2, the regional heating network model includes the total heat loss of the heating pipeline and the head loss along the pipeline;

[0016] Total heat loss of heating pipelines E ij The expression is:

[0017] ;

[0018] In the formula, L ij Indicates from node i To the node j The length of the pipe, β Indicates the additional heat loss coefficient. q id This indicates the heat loss per unit length of pipe under standard operating conditions.

[0019] Head loss along the pipeline H ij The expression is:

[0020] ;

[0021] In the formula, Y Indicates pipe roughness. r Indicates the density of the transmission medium.d ij Indicates the inner diameter of the pipe. G This indicates the flow rate of the transmission medium.

[0022] Furthermore, S3 includes the following sub-steps:

[0023] S31. Based on the regional distribution network model, calculate the distribution network modularity index of the electric-thermal hybrid energy network cluster;

[0024] S32. Based on the regional heat network model, calculate the heat network modularity index of the electric-thermal hybrid energy network cluster;

[0025] S33. Calculate the power balance index of the distribution network of the electric-thermal hybrid energy network cluster;

[0026] S34. Calculate the power balance index of the heat network cluster in the electric-thermal hybrid energy network cluster.

[0027] S35. Based on the modularity index of the power distribution network, the modularity index of the heat network, the power balance index of the power distribution network, and the power balance index of the heat network, construct a cluster partitioning model for the hybrid electric and heat energy network.

[0028] Furthermore, in S31, the modularity index of the power distribution network D E The expression is:

[0029] ;

[0030] In the formula, m This represents the sum of the weights of all edges in the network. w i Represents all nodes i The sum of the weights of connected edges. w j The table represents all nodes. j The sum of the weights of connected edges. F ij Indicates that by node i With nodes j The power flow caused by changes in power transmission between them d E ( i , j ) represents the factor for determining the affiliation of distribution network nodes;

[0031] In S32, the thermal network modularity index D H The expression is:

[0032] ;

[0033] In the formula, Hij This indicates the head loss along the pipeline. E ij This indicates the total heat loss of the heating pipeline. d H ( i , j () indicates the factor for determining the affiliation of a regional heating network node;

[0034] In S33, the power balance index of the distribution network Q P The expression is:

[0035] ;

[0036] In the formula, f P,k Indicates the first k Power balance index of a DIES cluster K Indicates the number of multi-DIES clusters;

[0037] In S34, the power balance index of the heating network Q H The expression is:

[0038] ;

[0039] In the formula, f H,k Indicates the first k Thermal power balance index of a DIES cluster;

[0040] In S35, a cluster partitioning model for an electrothermal hybrid energy network is constructed using a reinforcement learning model.

[0041] Action space partitioning model of electric-thermal hybrid energy network cluster Action The expression is:

[0042] ;

[0043] State space of the hybrid energy network cluster partitioning model State The expression is:

[0044] ;

[0045] In the formula, p k,m,t Represents a cluster k Middle node m The electricity load demand, h k,m,t,t+Δt Represents a cluster node m In time Thermal power requirements at that time Represents a cluster k exist t Electric power at any given moment Represents a cluster k exist t Thermal power at any moment M E Represents a cluster k The number of distribution network nodes in the system M H Represents a cluster k The number of nodes in the regional heating network;

[0046] Reward function of the hybrid energy network cluster partitioning model Reward The expression is:

[0047] ;

[0048] In the formula, w p This represents the weighting coefficient of two indicators in the distribution network. D E Indicates the modularity index of the distribution network. D H Indicates the modularity index of a thermal network. w H This represents the weighting coefficients of two indicators in the regional heating network, and max(·) represents the maximum value function;

[0049] Action value function of the hybrid energy network cluster partitioning model Q t ( S t , A t The expression for ) is:

[0050] ;

[0051] In the formula, R t Indicates an immediate reward. Q t ( S t+1 , A t+1 This represents the action value function for the next time period. S t express t The state at any given moment, A t express t Actions at any moment S t+1 Indicates the state in the next time period. A t+1Indicates the action to be taken in the next time period.

[0052] Furthermore, in S4, the multi-DIES cluster two-layer collaborative operation optimization model includes a lower-layer DIES cluster operation optimization model and an upper-layer DIES cluster operation optimization model.

[0053] Furthermore, the expression for the lower-level DIES cluster operation optimization model is:

[0054] ;

[0055] ;

[0056] ;

[0057] ;

[0058] ;

[0059] ;

[0060] In the formula, C IES,k This represents the operating cost of a multi-DIES cluster. C en,k This represents the energy cost of a multi-DIES cluster. C om,k This represents the maintenance cost of a multi-DIES cluster. C E,k This represents the interaction cost between a multi-DIES cluster and the power grid. C NG,k This represents the fuel cost of a multi-DIES cluster. C H,k This represents the thermal cost of a multi-DIES cluster. T This indicates the optimization time period. This indicates the price at which the DIES cluster sells electricity to ESP. This indicates the electricity price of ESP. Indicates the market price of heat. This represents the heat power purchased from the regional heating network by the multi-DIES cluster, and max(·) represents the maximum value function. P grid,k,t This indicates the power exchange capacity between the DIES cluster and the ESP. P CHP,k,t Indicates the first k Power of CHP units in a DIES cluster P GB,k,t Indicates the first k GB unit power in a DIES cluster cgas Indicates the price of natural gas. or CHP,k Indicates the first k Operating efficiency of CHP units in a DIES cluster. LHV ng Indicates the lower heating value of natural gas. or GB,k Indicates the first k Operating efficiency of GB units in a DIES cluster c w Indicates the first k Units in a DIES cluster w Unit operating power maintenance cost P k,w,t Indicates the first k Units in a DIES cluster w Operating power W Indicates the first k Number of units in a DIES cluster;

[0061] The expression for the constraints of the lower-level DIES cluster operation optimization model is as follows:

[0062] ;

[0063] ;

[0064] ;

[0065] ;

[0066] ;

[0067] In the formula, SOC ES,t+1 Indicates energy storage t The state of charge at time +1, SOC ES,t Indicates energy storage t State of charge at time t, α l Indicates the energy storage self-discharge loss rate. This indicates the charging power of the energy storage. Indicates energy storage charging efficiency. Indicates the rated charge / discharge power of energy storage, Δ t Indicates the optimization time interval. Indicates the rated capacity of energy storage. This indicates the discharge power of the stored energy. Indicates the energy storage discharge efficiency. M Represents the state variables of energy storage charging and discharging. Indicates the minimum state of charge of energy storage. Indicates the maximum state of charge of energy storage. SOC ES,1 The desired state of charge represents the initial state of charge of the energy storage. SOC ES,T This indicates the expected state of charge at the end of the energy storage dispatch cycle;

[0068] The expression for the upper-layer DIES cluster operation optimization model is:

[0069] ;

[0070] ;

[0071] In the formula, C ENO This represents the total operating cost, and max(·) represents the maximum value function. P Tr,t This indicates the output power of the main transformer. This indicates the price at which the distribution network purchases electricity from the upstream main grid. C over This represents the overload cost of the main transformer. This indicates the rated power of the main transformer. c Tr This represents the overload loss coefficient of the main transformer;

[0072] The expression for the constraints of the lower-level DIES cluster operation optimization model is as follows:

[0073] ;

[0074] In the formula, Indicates the first k Power trading between a DIES cluster and the distribution network P Cwt,t This indicates the power generation capacity of wind power clusters in the distribution network.

[0075] Furthermore, in S5, the xLSTM algorithm is used to determine the optimal action in a given state, thereby solving the cluster partitioning model of the electric-thermal hybrid energy network.

[0076] In S5, a genetic algorithm is used to solve the upper-level DIES cluster operation optimization model;

[0077] In S5, the Cplex commercial solver is used to solve the lower-level DIES cluster operation optimization model.

[0078] The beneficial effects of this invention are:

[0079] (1) This invention designs a power distribution network main transformer overload solution and a two-stage operation optimization framework that considers the flexible adjustment of power distribution and regional heating network structure. Compared with the existing topology adjustment strategy, it not only involves the topology and equipment related to the power distribution network, but also considers the regional heating network coupled with the power distribution network and related electric-thermal coupling equipment. It has more adjustable resources and more effective adjustment.

[0080] (2) This invention establishes a flexible model for power distribution and regional heating network structure. In view of the complex scenario of multiple devices such as distributed photovoltaic, electric thermal energy storage, electric boiler, heat pump and micro gas turbine and new loads being connected to the power distribution network at the present stage, a flexible model for hybrid electric thermal network structure is established, which can provide basic model support for the topology adjustment strategy of electric-thermal hybrid energy network.

[0081] (3) This invention establishes a reinforcement learning model for the cluster partitioning of an electric-thermal hybrid energy network that considers the coupling relationship between structural characteristics and power characteristics. Existing technologies have established distribution network cluster partitioning models or distribution network reconfiguration models based on traditional data models. This invention simultaneously considers distribution network topology reconfiguration, regional heating network topology optimization, and the coupling relationship between the distribution network and the regional heating network, and establishes a reinforcement learning model for the cluster partitioning of an electric-thermal hybrid energy network that considers the coupling relationship between structural characteristics and power characteristics. The cluster partitioning model is established through the reinforcement learning model.

[0082] (4) This invention establishes a two-layer collaborative operation optimization model that considers the collaborative operation of each distributed cluster and the electric and thermal multi-energy network. The existing technology does not consider incorporating the cost of main transformer overload into the optimization objective of the upper-level energy service provider. This invention establishes a two-layer collaborative operation optimization model that considers the collaborative operation of each distributed cluster and the electric and thermal multi-energy network, and incorporates the cost of main transformer overload into the optimization objective in the upper-level model, thereby solving the main transformer overload problem through the linkage optimization between the upper and lower levels.

[0083] (5) This invention designs a Q-learning algorithm based on an improved xLSTM architecture to solve the reinforcement learning model for the partitioning of electric and thermal hybrid energy networks. It is necessary to coordinate the aggregation of nodes in the distribution network and regional heating network topology, and there are many decision actions. The state of the partitioning of electric and thermal hybrid energy networks includes the load, power generation, heating power and number of nodes in each cluster, which is complex. The xLSTM algorithm is suitable for handling high-dimensional state and continuous action space, which fits the characteristics of the electric and thermal hybrid energy network partitioning problem. The algorithm can continuously learn and adapt to environmental changes, optimize the cluster partitioning action strategy and match the multi-state and multi-action electric and thermal hybrid energy network partitioning problem, and has strong application potential. When using the Q-learning algorithm to solve the reinforcement learning model for the partitioning of electric and thermal hybrid energy networks, this invention introduces the xLSTM algorithm. In the initial action selection, the xLSTM algorithm is used to learn and select the best action under the given state, which improves the accuracy and reliability of the Q-learning algorithm.

[0084] (6) This invention designs a distributed solution method for a two-layer collaborative optimization model of multiple DIES clusters. Considering the uninterrupted parameter transfer process between the upper and lower models, and that each DIES is an independent entity for operation and optimization, a distributed solution algorithm is designed to solve the two-layer optimization model. The upper-layer multi-DIES collaborative optimization model involves discontinuous variables, which are difficult to solve using traditional mathematical methods. Therefore, a genetic algorithm (GA) is used to solve the upper-layer model. The lower-layer DIES cluster operation optimization model is a typical convex optimization model, which is solved using the commercial solver Cplex. During the GA solution process of the upper-layer model, the Cplex solver of the lower-layer model is embedded to construct a distributed solution process and finally obtain the global optimal solution. Attached Figure Description

[0085] Figure 1 A flowchart for a regional integrated energy operation optimization method that considers the flexible adjustment of multi-energy networks;

[0086] Figure 2 A structural diagram of the research object and the adaptation scenario;

[0087] Figure 3 Diagram showing temperature transfer in the supply and return water pipes;

[0088] Figure 4 This is a diagram of the Markov decision process. Detailed Implementation

[0089] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0090] like Figure 1 As shown, this invention provides a regional integrated energy operation optimization method that considers the flexible adjustment of multi-energy networks, including the following steps:

[0091] S1. Construct a hybrid electric and thermal energy network cluster and a multi-DIES cluster;

[0092] S2. Construct regional power distribution network model and regional heating network model;

[0093] S3. Based on the regional power distribution network model and the regional heating network model, construct a cluster partitioning model for the hybrid electric and heating energy network;

[0094] S4. Construct a multi-DIES cluster two-layer collaborative operation optimization model;

[0095] S5. Solve the electrothermal hybrid energy network cluster partitioning model and the multi-DIES cluster two-layer collaborative operation optimization model to complete the optimization.

[0096] This invention proposes an operation optimization method for regional integrated energy systems that considers the flexible adjustment of multi-energy networks. The invention is applied to regional integrated energy systems, and the research object and applicable scenarios are as follows: Figure 2 As shown.

[0097] like Figure 2 As shown, the Regional Integrated Energy System (RIES) consists of multiple distribution area integrated energy systems (DIES) within the Energy Management System (EMS) and the 10kV medium-voltage Regional Distribution Network (RDN) located within the power supply range of the 110kV substation, as well as distributed photovoltaic clusters and decentralized wind power clusters, and is connected to the Urban Energy System (UES). The distribution area integrated energy systems are located at the end of the 10kV / 380V distribution transformer area and independently equip themselves with power and heating subsystems to supply electricity and heat to users. Matching this is the Regional Heat Network (RHN), which undertakes the task of heat transmission and distribution.

[0098] The equipment in the power subsystem of the integrated energy system at the district level includes distributed photovoltaic (PV) and electric energy storage (EES) devices. The heating subsystem includes various equipment such as electric boilers (EB), heat pumps (HP), gas boilers (GB), combined heat and power (CHP) units, and heat energy storage (HES). The energy service provider (ESP) is the leader / collaborator in optimizing the operation of RIES, responsible for the overall low-carbon economic operation of the system and the operation and maintenance of the electricity / heat energy network. During system operation, the ESP can optimize the collaborative operation strategy with various DIES, PV clusters, and WT (Wind Turbine) clusters by setting reasonable electricity prices, and optimize the power interaction strategy between RIES, the power grid, and the heating company.

[0099] like Figure 2 As shown, the integrated energy systems of each distribution area are distributed at different nodes in the distribution network and the regional heating network. Other electric and heat load nodes are also distributed in the distribution network and the regional heating network. These nodes are supplied with heat and electricity through ESP.

[0100] The first phase targets the hybrid energy network containing multiple transformer substations and integrated energy systems. It establishes a modularity index to characterize the density of the energy network structure and a supply-demand balance index to characterize the matching degree of power supply and demand for electricity and heat. By weighting the indexes, a cluster partitioning optimization model is constructed to divide the hybrid energy network into multiple clusters centered on DIES. DIES is responsible for supplying power and heat to the user load nodes within the cluster.

[0101] The second phase, targeting the divided DIES clusters, constructs a two-layer collaborative operation optimization model aimed at reducing the overload risk of distribution network main transformers. The lower-layer model establishes an operation optimization model for a single DIES cluster, optimizing its operation strategy. The upper-layer model establishes a collaborative operation optimization model for multiple DIES clusters, with ESP dynamically adjusting distribution network electricity prices to optimize the total power demand of the distribution network from multiple DIES clusters. Through iterative optimization between the upper and lower-layer models, the optimal operation strategy for each DIES cluster and the optimal operation strategy for the distribution network main transformer are obtained.

[0102] In this embodiment of the invention, in S2, the expression for the regional distribution network model is:

[0103] ;

[0104] In the formula,F ij Indicates that by node i With nodes j The power flow caused by changes in power transmission between them f 1i Indicates from node i The power flow changes on branch 1 caused by the power flowing in and out of the reference node. f 1j Represents a node j The power flow changes on branch 1 caused by the power flowing in and out of the reference node. f 2i Represents a node i The power flow changes on branch 2 caused by the power flowing in and out of the reference node. f 2j Represents a node j The power flow changes on branch 2 caused by the power flowing in and out of the reference node. f li Indicates from node i Branches caused by power flowing in and out of the reference node l Trend changes on the surface f lj Indicates from node j Branches caused by power flowing in and out of the reference node l Trend changes;

[0105] In S2, the regional heating network model includes the total heat loss of the heating pipeline and the head loss along the pipeline;

[0106] Regional heating networks experience significant energy losses during long-distance heat transmission, primarily including heat loss and head loss. Figure 3 This is for temperature transfer in supply and return pipelines. Due to the temperature difference between the inside and outside of the pipeline, the transport medium in the pipeline undergoes a heat exchange process. Therefore, during the heat transfer process, there will be some heat loss along the pipeline, resulting in the pipeline outlet temperature being lower than the inlet temperature. T s- This indicates the inlet temperature of the heating pipe. T s+ This indicates the outlet temperature of the heating pipe. T R- This indicates the inlet temperature of the hot reflux pipe. T R+ This indicates the outlet temperature of the hot reflux pipe. L Indicates the length of the pipe.

[0107] Total heat loss of heating pipelines E ij The expression is:

[0108] ;

[0109] In the formula, L ij Indicates from node i To the node j The length of the pipe, β Indicates the additional heat loss coefficient. q id This indicates the heat loss per unit length of pipe under standard operating conditions.

[0110] Besides heat loss, head loss is another important indicator describing pressure loss within a pipeline. Head loss is the mechanical energy loss per unit mass of liquid during water flow, caused by external resistance to the flow. Head loss includes friction head loss and local head loss. For long-distance heat transport in pipelines, local head loss can be neglected. The head loss along the pipeline can be obtained using the Darcy-Weisbach equation.

[0111] Head loss along the pipeline H ij The expression is:

[0112] ;

[0113] In the formula, Y Indicates pipe roughness. r Indicates the density of the transmission medium. d ij Indicates the inner diameter of the pipe. G This indicates the flow rate of the transmission medium.

[0114] In this embodiment of the invention, S3 includes the following sub-steps:

[0115] S31. Based on the regional distribution network model, calculate the distribution network modularity index of the electric-thermal hybrid energy network cluster;

[0116] S32. Based on the regional heat network model, calculate the heat network modularity index of the electric-thermal hybrid energy network cluster;

[0117] S33. Calculate the power balance index of the distribution network of the electric-thermal hybrid energy network cluster;

[0118] S34. Calculate the power balance index of the heat network cluster in the electric-thermal hybrid energy network cluster.

[0119] S35. Based on the modularity index of the power distribution network, the modularity index of the heat network, the power balance index of the power distribution network, and the power balance index of the heat network, construct a cluster partitioning model for the hybrid electric and heat energy network.

[0120] Cluster partitioning primarily aims to improve the matching degree between the output power of coupled devices in a DIES cluster and the actual load demand by fully leveraging the coordination and complementarity among nodes. Simultaneously, the partitioned clusters should also ensure sufficient internal structural strength. Therefore, this invention proposes cluster partitioning indices based on modularity and power balance metrics.

[0121] In this embodiment of the invention, in S31, for the distribution network, the weight matrix for distribution network partitioning can be defined based on the power flow changes caused by power transmission between different nodes. According to the distribution network model, all branches with small power flow changes are partitioned into the same cluster, while branches with large power flow changes are partitioned into different clusters. When partitioning the distribution network, the smaller the power flow change, the larger the edge weight. Distribution network modularity index. D E The expression is:

[0122] ;

[0123] In the formula, m This represents the sum of the weights of all edges in the network. w i Represents all nodes i The sum of the weights of connected edges. w j The table represents all nodes. j The sum of the weights of connected edges. F ij Indicates that by node i With nodes j The power flow caused by changes in power transmission between them d E ( i , j ) represents the factor for determining the affiliation of distribution network nodes;

[0124] In S32, for a regional heat network, the heat transfer medium in the pipes needs to overcome resistance during flow, resulting in temperature and head losses. Therefore, the weight matrix for regional heat network partitioning can be described by temperature and head losses. Based on the established regional heat network model, pipes with lower temperature and head losses are grouped into the same cluster, while pipes with higher temperature and head losses are grouped into different clusters. When partitioning the regional heat network, the smaller the temperature and head losses of a pipe, the larger its edge weight. Heat network modularity index D H The expression is:

[0125] ;

[0126] In the formula, Hij This indicates the head loss along the pipeline. E ij This indicates the total heat loss of the heating pipeline. d H ( i , j () indicates the factor for determining the affiliation of a regional heating network node;

[0127] In S33, a power balance index is used to characterize the degree of matching between the electrical / thermal power output and the actual electrical / thermal load demand within a defined cluster. Distribution network power balance index Q P The expression is:

[0128] ;

[0129] In the formula, f P,k Indicates the first k Power balance index of a DIES cluster K Indicates the number of multi-DIES clusters;

[0130] In S34, the power balance index of the heating network Q H The expression is:

[0131] ;

[0132] In the formula, f H,k Indicates the first k Thermal power balance index of a DIES cluster;

[0133] In S35, firstly, under the premise of ensuring close connections between nodes within the formed cluster and that the cluster's supply capacity meets internal needs, optimization objectives for the initial heating network and power grid-level cluster partitioning are established using modularity and supply-demand matching indices, respectively. The cluster partitioning results must, on the one hand, ensure the connection strength between nodes within the cluster; that is, the larger the modularity index D value, the closer the connections within the cluster and the better the structural performance. On the other hand, it must ensure sufficient energy supply capacity within the cluster; that is, the larger the power balance index Q, the stronger the energy supply capacity within the cluster and the better the functional performance.

[0134] like Figure 4 As shown, this invention, based on Markov decision processes, models the first-stage electric-thermal hybrid energy network cluster partitioning problem as a reinforcement learning model. The reinforcement learning model for electric-thermal energy network cluster partitioning typically consists of a quadruple (...). Action , State , Reward , π )definition, πThe policy set of the agent is represented by the state space. State To the action space Action The mapping, Reward The reward function is represented. A cluster partitioning model for an electrothermal hybrid energy network is constructed using a reinforcement learning model.

[0135] For the process of partitioning the electric and thermal energy network cluster, the decision variable is the node partitioning decision in the distribution network / regional heating network, that is, the node affiliation determination factor in the modularity index. d ( i , j The system aggregates nodes and determines whether they should be grouped into the same cluster. That is, when a node... i and j When located in the same cluster d ( i , j )=1, otherwise d ( i , j =0. Action space of the hybrid electric and thermal energy network cluster partitioning model. Action The expression is:

[0136] ;

[0137] For the process of dividing the electric and thermal energy network clusters, in terms of the operational space of the distribution network / regional heating network Action After execution, the distribution network / regional heating network is divided into multiple clusters centered on DIES. At this time, the load, power generation, heating power, and number of nodes in each cluster will change, and these changes will directly affect the agent's reward function. Reward The calculation results. Therefore, this invention defines the state space. State The state space of the hybrid energy network cluster partitioning model is defined by the set of loads, power generation, heating capacity, and number of nodes within the partitioned cluster. State The expression is:

[0138] ;

[0139] In the formula, p k,m,t Represents a cluster k Middle node m The electricity load demand, h k,m,t,t+Δt Represents a cluster node m In time Thermal power requirements at that time Represents a cluster k exist t Electric power at any given moment Represents a cluster k exist t Thermal power at any moment M E Represents a cluster k The number of distribution network nodes in the system M H Represents a cluster k The number of nodes in the regional heating network;

[0140] For the process of dividing the electric and thermal energy network clusters, in terms of the operational space of the distribution network / regional heating network Action After execution, the load, power generation, heating power, and number of nodes within each cluster of the distribution network / regional heating network will change, leading to changes in both the modularity index and the power balance index. Therefore, this invention constructs a reward function that simultaneously includes both the modularity index and the power balance index by setting weighting coefficients. Reward function for the cluster partitioning model of the hybrid electric-thermal energy network. Reward The expression is:

[0141] ;

[0142] In the formula, w p This represents the weighting coefficient of two indicators in the distribution network. D E Indicates the modularity index of the distribution network. D H Indicates the modularity index of a thermal network. w H This represents the weighting coefficients of two indicators in the regional heating network, and max(·) represents the maximum value function;

[0143] Agent policy set reward function π ( S t , A t ) represents the state that the agent faces. S t Select action at time A t The present invention employs the ε-greedy method, defining an exploration rate ε between 0 and 1. When selecting an action, the agent has an ε probability of selecting a random action and a (1-ε) probability of selecting the optimal action with the highest action value. The action value is determined by the immediate reward. R t Composed of expected rewards. Action-value function of the electrothermal hybrid energy network cluster partitioning model. Q t ( S t , A tThe expression for ) is:

[0144] ;

[0145] In the formula, R t Indicates an immediate reward. Q t ( S t+1 , A t+1 This represents the action value function for the next time period. S t express t The state at any given moment, A t express t Actions at any moment S t+1 Indicates the state in the next time period. A t+1 Indicates the action to be taken in the next time period.

[0146] In this embodiment of the invention, in S4, the multi-DIES cluster two-layer collaborative operation optimization model includes a lower-layer DIES cluster operation optimization model and an upper-layer DIES cluster operation optimization model.

[0147] In this embodiment of the invention, the lower-level model establishes an economic operation optimization model with a single DIES cluster as the optimization object, optimizing the DIES cluster operation strategy. The lower-level optimization model takes minimizing operating costs as its objective function. The operating costs of the DIES cluster include operation and maintenance costs, fuel costs, electricity purchase costs, and heat purchase costs. The expression of the lower-level DIES cluster operation optimization model is:

[0148] ;

[0149] ;

[0150] ;

[0151] ;

[0152] ;

[0153] ;

[0154] In the formula, C IES,k This represents the operating cost of a multi-DIES cluster. C en,k This represents the energy cost of a multi-DIES cluster. C om,kThis represents the maintenance cost of a multi-DIES cluster. C E,k This represents the interaction cost between a multi-DIES cluster and the power grid. C NG,k This represents the fuel cost of a multi-DIES cluster. C H,k This represents the thermal cost of a multi-DIES cluster. T This indicates the optimization time period. This indicates the price at which the DIES cluster sells electricity to ESP. This indicates the electricity price of ESP. Indicates the market price of heat. This represents the heat power purchased from the regional heating network by the multi-DIES cluster, and max(·) represents the maximum value function. P grid,k,t This indicates the power exchange capacity between the DIES cluster and the ESP. P CHP,k,t Indicates the first k Power of CHP units in a DIES cluster P GB,k,t Indicates the first k GB unit power in a DIES cluster c gas Indicates the price of natural gas. or CHP,k Indicates the first k Operating efficiency of CHP units in a DIES cluster. LHV ng Indicates the lower heating value of natural gas. or GB,k Indicates the first k Operating efficiency of GB units in a DIES cluster c w Indicates the first k Units in a DIES cluster w Unit operating power maintenance cost P k,w,t Indicates the first k Units in a DIES cluster w Operating power W Indicates the first k Number of units in a DIES cluster;

[0155] The constraints that the lower-level DIES operation optimization model needs to satisfy include electrical / thermal power balance constraints and unit operation constraints. The expressions for the constraints of the lower-level DIES cluster operation optimization model are as follows:

[0156] ;

[0157] ;

[0158] ;

[0159] ;

[0160] ;

[0161] In the formula, SOC ES,t+1 Indicates energy storage t The state of charge at time +1, SOC ES,t Indicates energy storage t State of charge at time t, α l Indicates the energy storage self-discharge loss rate. This indicates the charging power of the energy storage. Indicates energy storage charging efficiency. Indicates the rated charge / discharge power of energy storage, Δ t Indicates the optimization time interval. Indicates the rated capacity of energy storage. This indicates the discharge power of the stored energy. Indicates the energy storage discharge efficiency. M Represents the state variables of energy storage charging and discharging. Indicates the minimum state of charge of energy storage. Indicates the maximum state of charge of energy storage. SOC ES,1 The desired state of charge represents the initial state of charge of the energy storage. SOC ES,T This indicates the expected state of charge at the end of the energy storage dispatch cycle;

[0162] The upper-level model takes multiple DIES clusters as objects and establishes a collaborative operation optimization model. The ESP optimizes the total power demand of the distribution network from multiple DIES clusters by dynamically adjusting the distribution network electricity price. The objective function of the upper-level multi-DIES cluster collaborative operation optimization model is to minimize the system operating cost, which includes the transaction costs between the ESP and the grid, and the overload cost of the main transformer. The expression of the upper-level DIES cluster operation optimization model is:

[0163] ;

[0164] ;

[0165] In the formula, C ENO This represents the total operating cost, and max(·) represents the maximum value function. P Tr,t This indicates the output power of the main transformer. This indicates the price at which the distribution network purchases electricity from the upstream main grid. C over This represents the overload cost of the main transformer. This indicates the rated power of the main transformer. c Tr This represents the overload loss coefficient of the main transformer;

[0166] The expression for the constraints of the lower-level DIES cluster operation optimization model is as follows:

[0167] ;

[0168] In the formula, Indicates the first k Power trading between a DIES cluster and the distribution network P Cwt,t This indicates the power generation capacity of wind power clusters in the distribution network.

[0169] In this embodiment of the invention, in S5, the xLSTM algorithm is used to determine the optimal action in a given state to complete the solution of the electric-thermal hybrid energy network cluster partitioning model.

[0170] In S5, a genetic algorithm is used to solve the upper-level DIES cluster operation optimization model;

[0171] In S5, the Cplex commercial solver is used to solve the lower-level DIES cluster operation optimization model.

[0172] In this embodiment of the invention, Q-learning is a value-based reinforcement learning algorithm that is not based on an environment model. The main idea of ​​Q-learning is to define a state-action value function, i.e., the Q-function, and iteratively learn the Q-function by substituting observed data into the following update formula. The Q-learning algorithm learns the optimal policy through continuous trial and error and updating the Q-table. It does not require knowledge of a complete model of the environment, making it a model-free reinforcement learning method. The core of the Q-learning algorithm lies in the application of Bellman equations and the use of ε-greedy policies, which together enhance the agent's exploration and utilization capabilities, enabling the agent to find optimal or near-optimal behavioral policies in complex environments.

[0173] xLSTM (Extended Long Short-Term Memory) is an extension and improvement of the traditional LSTM model, effectively enhancing the performance of LSTM in processing time series data and sequence prediction tasks. xLSTM can scale to billions of parameters, unlocking new possibilities for language modeling and sequence processing tasks. In multiple benchmark tests, xLSTM has demonstrated capabilities surpassing conventional models, especially in scenarios requiring the understanding and generation of texts with high contextual relevance and complex logic.

[0174] The reinforcement learning model for cluster partitioning of hybrid electric and thermal energy networks established in this invention requires coordinating the aggregation of nodes in the distribution network and regional heating network topology, resulting in a large number of decision actions. The states of hybrid electric and thermal energy network cluster partitioning include the load, power generation, heating power, number of nodes within the cluster, and complex states of each cluster. The xLSTM algorithm is suitable for handling high-dimensional states and continuous action spaces, which aligns with the characteristics of the hybrid electric and thermal energy network cluster partitioning problem. This algorithm can continuously learn and adapt to environmental changes, optimizing the cluster partitioning action strategy to match the multi-state, multi-action hybrid electric and thermal energy network cluster partitioning problem, demonstrating strong application potential.

[0175] Therefore, when using the Q-learning algorithm to solve the reinforcement learning model of the electrothermal hybrid energy network cluster partitioning, this invention introduces the xLSTM algorithm. When selecting the initial action A1, the xLSTM algorithm is used to learn and select the best action A1 under a given state S1, thereby improving the accuracy and reliability of the Q-learning algorithm.

[0176] There is a continuous parameter transfer process between the upper and lower level models, and each DIES operates and optimizes independently. Therefore, a distributed solution algorithm is constructed to solve the two-level optimization model. The upper-level multi-DIES collaborative optimization model involves discontinuous variables, which are difficult to solve using traditional mathematical methods. Therefore, a Genetic Algorithm (GA) is used to solve the upper-level model. The lower-level DIES cluster operation optimization model is a typical convex optimization model, solved using the commercial solver Cplex. During the GA solution process of the upper-level model, the Cplex solver for the lower-level model is embedded to construct a distributed solution process. First, an ESP price is randomly generated to guide each lower-level DIES in operation optimization through Cplex, obtaining the interaction power between each DIES and the distribution network. Then, GA is executed to solve the upper-level multi-DIES collaborative optimization model, optimizing the operating strategies of WT, PV, and main transformers in the distribution network. Through continuous iteration using GA and Cplex, the global optimal solution is finally obtained.

[0177] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A regional integrated energy operation optimization method considering the flexible adjustment of multi-energy networks, characterized in that, Includes the following steps: S1. Construct a hybrid electric and thermal energy network cluster and a multi-DIES cluster; S2. Construct regional power distribution network model and regional heating network model; S3. Based on the regional power distribution network model and the regional heating network model, construct a cluster partitioning model for the hybrid electric and heating energy network; S4. Construct a multi-DIES cluster two-layer collaborative operation optimization model; S5. Solve the electrothermal hybrid energy network cluster partitioning model and the multi-DIES cluster two-layer collaborative operation optimization model, and complete the optimization. S3 includes the following sub-steps: S31. Based on the regional distribution network model, calculate the distribution network modularity index of the electric-thermal hybrid energy network cluster; S32. Based on the regional heat network model, calculate the heat network modularity index of the electric-thermal hybrid energy network cluster; S33. Calculate the power balance index of the distribution network of the electric-thermal hybrid energy network cluster; S34. Calculate the power balance index of the heat network cluster in the electric-thermal hybrid energy network cluster. S35. Based on the modularity index of the power distribution network, the modularity index of the heat network, the power balance index of the power distribution network, and the power balance index of the heat network, construct a cluster partitioning model for the hybrid electric and heat energy network. In S4, the multi-DIES cluster two-layer collaborative operation optimization model includes a lower-layer DIES cluster operation optimization model and an upper-layer DIES cluster operation optimization model. The expression for the lower-level DIES cluster operation optimization model is: ； ； ； ； ； ； In the formula, C IES,k This represents the operating cost of a multi-DIES cluster. C en,k This represents the energy cost of a multi-DIES cluster. C om,k This represents the maintenance cost of a multi-DIES cluster. C E,k This represents the interaction cost between a multi-DIES cluster and the power grid. C NG,k This represents the fuel cost of a multi-DIES cluster. C H,k This represents the thermal cost of a multi-DIES cluster. T This indicates the optimization time period. This indicates the price at which the DIES cluster sells electricity to ESP. This indicates the electricity price of ESP. Indicates the market price of heat. This represents the heat power purchased from the regional heating network by the multi-DIES cluster, and max(·) represents the maximum value function. P grid,k,t This indicates the power exchange capacity between the DIES cluster and the ESP. P CHP,k,t Indicates the first k Power of CHP units in a DIES cluster P GB,k,t Indicates the first k GB unit power in a DIES cluster c gas Indicates the price of natural gas. η CHP,k Indicates the first k Operating efficiency of CHP units in a DIES cluster. LHV ng Indicates the lower heating value of natural gas. η GB,k Indicates the first k Operating efficiency of GB units in a DIES cluster c w Indicates the first k Units in a DIES cluster w Unit operating power maintenance cost P k,w,t Indicates the first k Units in a DIES cluster w Operating power W Indicates the first k Number of units in a DIES cluster; The expression for the constraint conditions of the lower-level DIES cluster operation optimization model is as follows: ； ； ； ； ； In the formula, SOC ES,t+1 Indicates energy storage t The state of charge at time +1, SOC ES,t Indicates energy storage t State of charge at time t, α l Indicates the energy storage self-discharge loss rate. This indicates the charging power of the energy storage. Indicates energy storage charging efficiency. Indicates the rated charge / discharge power of energy storage, Δ t Indicates the optimization time interval. Indicates the rated capacity of energy storage. This indicates the discharge power of the stored energy. Indicates the energy storage discharge efficiency. M Represents the state variables of energy storage charging and discharging. Indicates the minimum state of charge of energy storage. Indicates the maximum state of charge of energy storage. SOC ES,1 The desired state of charge represents the initial state of charge of the energy storage. SOC ES,T This indicates the expected state of charge at the end of the energy storage dispatch cycle; The expression for the upper-layer DIES cluster operation optimization model is: ； ； In the formula, C ENO This represents the total operating cost, and max(·) represents the maximum value function. P Tr,t This indicates the output power of the main transformer. This indicates the price at which the distribution network purchases electricity from the upstream main grid. C over This represents the overload cost of the main transformer. This indicates the rated power of the main transformer. c Tr This represents the overload loss coefficient of the main transformer; The expression for the constraint conditions of the upper-layer DIES cluster operation optimization model is as follows: ； In the formula, Indicates the first k Power trading between a DIES cluster and the distribution network P Cwt,t This indicates the power generation capacity of wind power clusters in the distribution network.

2. The regional integrated energy operation optimization method considering flexible adjustment of multi-energy networks according to claim 1, characterized in that, In S2, the expression for the regional distribution network model is: ； In the formula, F ij Indicates that by node i With nodes j The power flow caused by changes in power transmission between them f 1i Indicates from node i The power flow changes on branch 1 caused by the power flowing in and out of the reference node. f 1j Represents a node j The power flow changes on branch 1 caused by the power flowing in and out of the reference node. f 2i Represents a node i The power flow changes on branch 2 caused by the power flowing in and out of the reference node. f 2j Represents a node j The power flow changes on branch 2 caused by the power flowing in and out of the reference node. f li Indicates from node i Branches caused by power flowing in and out of the reference node l Trend changes on the surface f lj Indicates from node j Branches caused by power flowing in and out of the reference node l Trend changes; In S2, the regional heating network model includes the total heat loss of the heating pipeline and the head loss along the pipeline. The total heat loss of the heating pipeline E ij The expression is: ； In the formula, L ij Indicates from node i To the node j The length of the pipe, β Indicates the additional heat loss coefficient. q id This indicates the heat loss per unit length of pipe under standard operating conditions. Head loss along the pipeline H ij The expression is: ； In the formula, Y Indicates pipe roughness. ρ Indicates the density of the transmission medium. d ij Indicates the inner diameter of the pipe. G This indicates the flow rate of the transmission medium.

3. The regional integrated energy operation optimization method considering flexible adjustment of multi-energy networks according to claim 1, characterized in that, In S31, the power distribution network modularity index D E The expression is: ； In the formula, m This represents the sum of the weights of all edges in the network. w i Represents all nodes i The sum of the weights of connected edges. w j The table represents all nodes. j The sum of the weights of connected edges. F ij Indicates that by node i With nodes j The power flow caused by changes in power transmission between them δ E ( i , j ) represents the factor for determining the affiliation of distribution network nodes; In S32, the thermal network modularity index D H The expression is: ； In the formula, H ij This indicates the head loss along the pipeline. E ij This indicates the total heat loss of the heating pipeline. δ H ( i , j () indicates the factor for determining the affiliation of a regional heating network node; In S33, the power balance index of the distribution network Q P The expression is: ； In the formula, φ P,k Indicates the first k Power balance index of a DIES cluster K Indicates the number of multi-DIES clusters; In S34, the power balance index of the heating network Q H The expression is: ； In the formula, φ H,k Indicates the first k Thermal power balance index of a DIES cluster; In S35, a cluster partitioning model for the electrothermal hybrid energy network is constructed using a reinforcement learning model. The action space of the electrothermal hybrid energy network cluster partitioning model Action The expression is: ； The state space of the electric-thermal hybrid energy network cluster partitioning model State The expression is: ； In the formula, p k,m,t Represents a cluster k Middle node m The electricity load demand, h k,m,t,t+Δt Represents a cluster node m In time Thermal power requirements at that time Represents a cluster k exist t Electric power at any given moment Represents a cluster k exist t Thermal power at any moment M E Represents a cluster k The number of distribution network nodes in the system M H Represents a cluster k The number of nodes in the regional heating network; The reward function of the electrothermal hybrid energy network cluster partitioning model Reward The expression is: ； In the formula, w p This represents the weighting coefficient of two indicators in the distribution network. D E Indicates the modularity index of the distribution network. D H Indicates the modularity index of a thermal network. w H This represents the weighting coefficients of two indicators in the regional heating network, and max(·) represents the maximum value function; Action value function of the electrothermal hybrid energy network cluster partitioning model Q t ( S t , A t The expression for ) is: ； In the formula, R t Indicates an immediate reward. Q t ( S t+1 , A t+1 This represents the action value function for the next time period. S t express t The state at any given moment, A t express t Actions at any moment S t+1 Indicates the state in the next time period. A t+1 Indicates the action to be taken in the next time period.

4. The regional integrated energy operation optimization method considering flexible adjustment of multi-energy networks according to claim 1, characterized in that, In S5, the xLSTM algorithm is used to determine the optimal action in a given state to solve the cluster partitioning model of the electric-thermal hybrid energy network. In S5, a genetic algorithm is used to solve the upper-level DIES cluster operation optimization model; In S5, the Cplex commercial solver is used to solve the lower-level DIES cluster operation optimization model.

Images