Integrated energy system gridding planning method based on minimum spanning tree splitting

Abstract

The invention relates to an integrated energy system gridding planning method based on minimum spanning tree splitting. The method comprises the following steps: constructing a node load autocorrelation matrix and a node load cross-correlation matrix; based on the node load autocorrelation matrix, the node energy demand and the physical distance between the nodes, constructing a universal gravitation model-based integrated energy system minimum spanning tree and the longest path thereof; and constructing and solving an integrated energy system gridding planning model based on the minimum spanning tree and the longest path thereof to obtain an optimal planning scheme. According to the method, isolated load prediction in traditional integrated energy system planning is fundamentally avoided, and the precision of load prediction is improved; according to the method, the minimum spanning tree technology is introduced into the planning of the integrated energy system, the topology of the network is simplified by constructing the minimum spanning tree, the number of decision variables of the gridding planning model of the integrated energy system is greatly reduced, the planning difficulty is reduced, the solving efficiency is improved, and the intelligence and automation level of the planning method is remarkably improved.

Description

Technical Field

[0001] The invention relates to the technical field of integrated energy systems, and in particular to a grid planning method for integrated energy systems based on minimum spanning tree splitting. Background Art

[0002] The integrated energy system produces, transmits and consumes energy in various forms, realizes the interaction between various energy sources, and is a powerful tool for achieving low carbon. However, in the context of the energy revolution, the user-side energy market has been gradually opened up, and new forms and elements such as distributed photovoltaics, electric vehicles, multi-energy complementarity, and electric energy substitution on the user side have been booming, resulting in an increase in the types of loads in the integrated energy system and a large difference between them.

[0003] The traditional integrated energy system planning method has a relatively rough planning scale and is prone to a "one-size-fits-all" planning scheme. Therefore, a systematic planning method for integrated energy that takes leanness and differentiation into consideration has emerged. The primary issue of grid planning is to reasonably divide the area to be planned based on spatial load forecasts, and then construct and evaluate the target grid for each unit. Against the background of the increasingly large scale of distribution networks, the objects to be divided are becoming more and more complex. Manual division based entirely on experience is inefficient and it is difficult to obtain a relatively optimal division scheme. There is an urgent need for an optimization planning method for an automated integrated energy system based on load characteristics and requirements. Summary of the invention

[0004] In order to solve the problems that the existing technology performs manual division of the grid of the integrated energy system completely according to experience, is inefficient and difficult to obtain a relatively optimal division scheme, and fails to perform grid division of the integrated energy system based on load coupling matching characteristics, the purpose of the present invention is to provide a grid planning method for the integrated energy system based on minimum spanning tree splitting, which can effectively reduce the number of decision variables of the model, reduce the planning difficulty, and improve the solution efficiency, while significantly improving the intelligence and automation level of the planning method, and realizing the optimization planning of the integrated energy system.

[0005] To achieve the above object, the present invention adopts the following technical solution: a grid planning method for an integrated energy system based on minimum spanning tree splitting, the method comprising the following steps in order:

[0006] (1) Analyze the correlation between the cold, hot, and electric loads connected to each node in the integrated energy system network and construct the node load autocorrelation matrix c ii,n and the node load cross-correlation matrix c ij,n , where i and j are node numbers respectively; n = 1, 2, 3, 4 represent typical days in spring, summer, autumn and winter respectively;

[0007] (2) Based on the node load autocorrelation matrix cii,n , node energy demand and the physical distance between nodes, and construct the minimum spanning tree of the integrated energy system based on the universal gravitation model and its longest path E max ;

[0008] (3) Construct a minimum spanning tree and its longest path E max The grid planning model of the integrated energy system is proposed and solved to obtain the optimal planning scheme Ω.

[0009] The step (1) specifically comprises the following steps:

[0010] (1a) Assume that the electricity, cooling and heating loads of node i in a typical day n are e, h, and c represent the electrical load, heating load, and cooling load, respectively. The maximum load in a typical day is defined as After normalization, the 24-hour electricity, cooling, and heating loads of node i in a typical day n are:

[0011]

[0012] In the formula, in the formula, is the normalized electric, heating and cooling load vector of node i in a typical day n;

[0013] (1b) Construct node load autocorrelation matrix c ii,n , as shown below:

[0014]

[0015] in, and denote the correlations between the electric-electric load, heat-heat load, and cooling-cooling load of node i in a typical day n, and They represent the correlation between the electricity-heating load, electricity-cooling load, and heat-cooling load of node i in a typical day n, and the matrix c ii,n The specific parameters are expressed as:

[0016]

[0017] In the formula, is the normalized electricity, heating and cooling loads of node i at time t in a typical day n; (1c) construct the node load correlation matrix c ij,n , as shown below:

[0018]

[0019] in, and denote the correlations between the electric-electric load, heat-heat load, and cooling-cooling load of nodes i and j in a typical day n, respectively. and represents the correlation between the electric-heating load, electric-cooling load, and heat-cooling load of node i and node j in a typical day n, respectively. The matrix c ij,n Specific parameters within It is expressed as:

[0020]

[0021] In the formula, are the normalized electric, heating and cooling loads of node j at time t in a typical day n.

[0022] The step (2) specifically comprises the following steps:

[0023] (2a) Consider the two nodes i and j in the integrated energy system network as two mass points in space, each of which has its own inherent mass and the similarity of load complementary matching; let x i,n,max represents the maximum value of the electricity, cooling and heating loads of node i in a typical day n, represents the maximum value of the electrical load of node i in a typical day n, represents the maximum cooling load of node i in a typical day n, represents the maximum value of the heat load of node i in a typical day n. According to the universal gravitational model, the energy gravitational force f between two nodes i and j is defined as ij ' ,n for:

[0024]

[0025] Where, L ij is the network distance between two nodes i and j; c ii,n and c jj,n are all node load autocorrelation matrices; x j,n,max is the maximum value of the electricity, cooling and heating loads of node j in a typical day n; and are the magnitudes of the attractive forces of electricity, heat and cooling loads between two nodes i and j respectively;

[0026] Introducing network weight coefficients on different typical days e, h, and c represent the electric load, heat load, and cooling load, respectively. The energy attraction between two nodes i and j under different typical days is calculated to obtain the comprehensive energy attraction f between nodes i and j. ij :

[0027]

[0028]

[0029] In the formula, and They represent the network weight coefficients of the electric network, the heat network, and the cold network in a typical day n, respectively;

[0030] make ω ij It is regarded as the weight of the branch ij with two end nodes i and j. Obviously, the smaller the weight of the branch, the higher the similarity of the two nodes it connects, and the more likely it is to be divided into the same grid;

[0031] (2b) Arrange all possible branches in the integrated energy system network according to their weights from small to large, and then select each edge in order. If the nodes at both ends of the branch do not belong to the same set, merge them until all nodes belong to the same set, thus forming a minimum tree. Specifically, it includes the following four steps:

[0032] (2b1) The set of all nodes in the entire planning area is regarded as a forest G, and each node is regarded as an independent tree;

[0033] (2b2) ​​Add all branches with weights to the set

[0034] (2b3) Take the branch (i, j) with the smallest weight from the set S, that is, ω ij is the minimum value in the set S. If (i, j) is not in the same tree, then connect i and j to merge the two trees, and add (i, j) to the branch set E of the spanning tree.

[0035] (2b4) Repeat step (2b3) until all nodes belong to the same tree and the branch set E is a minimum spanning tree;

[0036] (2c) Select any node k in the forest G, perform a depth-first traversal based on the connectivity relationship of the branch set E, and accumulate the weights of its branches to obtain the longest path. The branch set composed of the longest path is recorded as E. k Similarly, traverse all nodes in the forest G and select the longest branch set E from them max ,Right now E max This is the longest path of the minimum spanning tree.

[0037] The step (3) specifically comprises the following steps:

[0038] (3a) According to the area and energy consumption of the planned area, the upper and lower limits of the number of grids in the integrated energy system in the planned area are given; assuming that the area of ​​the planned area is S plan , r min and r max are the minimum and maximum power supply radius of the integrated energy system grid, respectively. The expression of the grid number z of the integrated energy system is:

[0039]

[0040] Where: represents rounding up; the value of the grid number z of the integrated energy system is between z min and z max Between; Set the initial value of the optimization target Let the number of grids in the integrated energy system of the planned area be z = z min ;

[0041] (3b) The longest path through E max According to the requirement of the number of grids in the integrated energy system, it is decomposed into z parts, z-1 split points are selected, and the minimum spanning tree is divided into z independent parts. Let the node set and branch set of each part be G k,0 and E k,0 , k=1,2,…,z, and E k The sum of the weights of the internal branches is w E,k ;

[0042] Dividing the minimum spanning tree into z independent parts must meet the requirements of the following formula:

[0043] min(|ω E,2 -ω E,1 |+|ω E,3 -ω E,2 |+...+|ω E,z -ω E,z-1 |) (10)

[0044] (3c) The node set G k,0 With the remaining node set Perform k-means clustering analysis to obtain z clustering results, that is, the grid division results of the integrated energy system. At this time, the integrated energy system is divided into z grids, and the expression of the node set of each grid is as follows:

[0045]

[0046] in: It is represented as the kth node in the zth grid, n1, n2 and nz represent the number of nodes in the 1st, 2nd and zth grids respectively;

[0047] (3d) A grid-based planning model for an integrated energy system is constructed, and its objective function is as follows:

[0048] f ∑ =min(f I +f 0 ) (12)

[0049] Where: f I Total annual cost for grid and equipment planning; 0 The annual operation and maintenance cost of the grid and equipment;

[0050] f I The expression is as follows:

[0051]

[0052] Where: r 0 is the annual interest rate; T l and T q are the service life cycles of branch l and integrated energy equipment q in the grid respectively; C l is the investment cost per unit length of branch l, C q is the unit capacity investment cost of the integrated energy equipment q in the grid; L l is the actual length of branch l after taking into account geographical obstacles; q is the capacity of the integrated energy equipment q;

[0053] f 0 The expression is as follows:

[0054]

[0055] Where: C 0 is the grid electricity price; ΔP l is the active power loss of branch l; τ max,l is the annual maximum load loss hours of branch l;

[0056] (3e) The constraints of the grid-based planning model for the integrated energy system include:

[0057] (3e1) Voltage U at node k k or air pressure p k The voltage quality constraints must be met, namely:

[0058]

[0059] Where: U k,max and U k,min are the upper and lower limits of the voltage at node k during normal operation; p k,max and p k,minare the upper and lower limits of the air pressure at node k during normal operation;

[0060] (3e2) The current I in branch lm lm Or gas flow fl lm The upper limit constraint must be satisfied, namely:

[0061]

[0062] Where: I lm,max is the upper limit of the current of branch lm during normal operation; fl lm,max It is the upper limit of the airflow of the branch 1m during normal operation;

[0063] (3e3) Network radiation constraint: Ensure that there is no loop when the system is operating normally, and satisfy the following formula:

[0064]

[0065] Where: B is the total number of branches in the network; b is the branch number; x b is a 0-1 variable representing the state of the b-th branch. When the branch is closed, the value is 1, and when the branch is open, the value is 0; N is the total number of nodes in the network; Ns is the number of power sources in the network; M d is the number of branches in the dth loop, d = 1, 2, ..., L; L is the total number of loops in the network; x dm is the state of the mth branch in the dth loop;

[0066] (3f) Solve the grid planning model of the integrated energy system and obtain the optimal value of the objective function f ∑ , and For comparison: If And z <z max , then let Update the optimal planning scheme to Ω, set z = z + 1, and then return to step (3b) to continue solving until z = z max , at this time, f ∑ The corresponding optimized planning scheme is the optimal planning scheme Ω.

[0067] It can be seen from the above technical scheme that the beneficial effects of the present invention are as follows: First, the present invention comprehensively considers the correlation between the cold, hot and electric loads of each node in the integrated energy system, constructs the node load autocorrelation matrix and the node load cross-correlation matrix to reflect the possibility of mutual conversion between different energy sources, fundamentally eliminates the isolated load forecasting in the traditional integrated energy system planning, and improves the accuracy of load forecasting; Second, the present invention introduces the universal gravitation model into the calculation of load coupling characteristics, realizes the accurate characterization of load characteristics such as mutual conversion and replacement between different energy loads, and effectively solves the coupling and matching degree of loads of different energy forms between different nodes taking into account the influence of spatial distance; Third, the present invention introduces the minimum spanning tree technology into the planning of the integrated energy system, and assigns similarity as the weight of the adjacency graph to each edge, retains the most important adjacency relationship, and simplifies the topology of the network by constructing the minimum spanning tree, preliminarily realizes the grid planning of the integrated energy, greatly reduces the number of decision variables in the grid planning model of the integrated energy system, reduces the planning difficulty, improves the solution efficiency, and significantly improves the intelligence and automation level of the planning method. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 is a flow chart of the method of the present invention;

[0069] Figure 2 is the minimum spanning tree and its longest path;

[0070] Figure 3 Schematic diagram of minimum spanning tree splitting. DETAILED DESCRIPTION

[0071] like Figure 1 As shown, a grid-based planning method for an integrated energy system based on minimum spanning tree splitting includes the following steps in order:

[0072] (1) Analyze the correlation between the cold, hot, and electric loads connected to each node in the integrated energy system network and construct the node load autocorrelation matrix c ii,n and the node load cross-correlation matrix c ij,n , where i and j are node numbers respectively; n = 1, 2, 3, 4 represent typical days in spring, summer, autumn and winter respectively;

[0073] (2) Based on the node load autocorrelation matrix c ii,n , node energy demand and the physical distance between nodes, and construct the minimum spanning tree of the integrated energy system based on the universal gravitation model and its longest path E max ;

[0074] (3) Construct a minimum spanning tree and its longest path E maxThe grid planning model of the integrated energy system is proposed and solved to obtain the optimal planning scheme Ω.

[0075] The step (1) specifically comprises the following steps:

[0076] (1a) Assume that the electricity, cooling and heating loads of node i in a typical day n are e, h, and c represent the electrical load, heating load, and cooling load, respectively. The maximum load in a typical day is defined as After normalization, the 24-hour electricity, cooling, and heating loads of node i in a typical day n are:

[0077]

[0078] In the formula, in the formula, is the normalized electric, heating and cooling load vector of node i in a typical day n;

[0079] (1b) Construct node load autocorrelation matrix c ii,n , as shown below:

[0080]

[0081] in, and denote the correlations between the electric-electric load, heat-heat load, and cooling-cooling load of node i in a typical day n, and They represent the correlation between the electricity-heating load, electricity-cooling load, and heat-cooling load of node i in a typical day n, and the matrix c ii,n The specific parameters are expressed as:

[0082]

[0083] In the formula, is the normalized electricity, heating and cooling loads of node i at time t in a typical day n; (1c) construct the node load correlation matrix c ij,n , as shown below:

[0084]

[0085] in, and denote the correlations between the electric-electric load, heat-heat load, and cooling-cooling load of nodes i and j in a typical day n, respectively. and represents the correlation between the electric-heating load, electric-cooling load, and heat-cooling load of node i and node j in a typical day n, respectively. The matrix c ij,nSpecific parameters within It is expressed as:

[0086]

[0087] In the formula, are the normalized electric, heating and cooling loads of node j at time t in a typical day n.

[0088] The step (2) specifically comprises the following steps:

[0089] (2a) Consider the two nodes i and j in the integrated energy system network as two mass points in space, each of which has its own inherent mass and the similarity of load complementary matching; let x i,n,max represents the maximum value of the electricity, cooling and heating loads of node i in a typical day n, represents the maximum value of the electrical load of node i in a typical day n, represents the maximum cooling load of node i in a typical day n, represents the maximum value of the heat load of node i in a typical day n. According to the universal gravitational model, the energy gravitational force f′ between two nodes i and j is defined as ij,n for:

[0090]

[0091] Where, L ij is the network distance between two nodes i and j; c ii,n and c jj,n are all node load autocorrelation matrices; x j,n,max is the maximum value of the electricity, cooling and heating loads of node j in a typical day n; and are the magnitudes of the attractive forces of electricity, heat and cooling loads between two nodes i and j respectively;

[0092] Introducing network weight coefficients on different typical days e, h, and c represent the electric load, heat load, and cooling load, respectively. The energy attraction between two nodes i and j under different typical days is calculated to obtain the comprehensive energy attraction f between nodes i and j. ij :

[0093]

[0094]

[0095] In the formula, and They represent the network weight coefficients of the power network, heat network and cold network in a typical day n respectively; considering that the connection between different energy grids in the rural integrated energy system mainly depends on the power network, in order to strengthen the connection between different energy grids in rural areas, it is recommended The value is around 0.8. and The value is around 0.1; in urban areas, it is recommended The value is around 0.6. and The value is around 0.2.

[0096] make ω ij It is regarded as the weight of the branch ij with two end nodes i and j. Obviously, the smaller the weight of the branch, the higher the similarity of the two nodes it connects, and the more likely it is to be divided into the same grid;

[0097] (2b) Arrange all possible branches in the integrated energy system network according to their weights from small to large, and then select each edge in order. If the nodes at both ends of the branch do not belong to the same set, merge them until all nodes belong to the same set, thus forming a minimum tree. Specifically, it includes the following four steps:

[0098] (2b1) The set of all nodes in the entire planning area is regarded as a forest G, and each node is regarded as an independent tree;

[0099] (2b2) ​​Add all branches with weights to the set

[0100] (2b3) Take the branch (i, j) with the smallest weight from the set S, that is, ω ij is the minimum value in the set S. If (i, j) is not in the same tree, then connect i and j to merge the two trees, and add (i, j) to the branch set E of the spanning tree.

[0101] (2b4) Repeat step (2b3) until all nodes belong to the same tree and the branch set E is a minimum spanning tree;

[0102] (2c) Select any node k in the forest G, perform a depth-first traversal based on the connectivity relationship of the branch set E, and accumulate the weights of its branches to obtain the longest path. The branch set composed of the longest path is recorded as E. k Similarly, traverse all nodes in the forest G and select the longest branch set E from them max ,Right now E max This is the longest path of the minimum spanning tree. Figure 2 As shown, one of the minimum spanning trees includes all nodes, that is, Figure 2 Each line segment of the solid and dashed lines is an element in the branch set E, and the longest path branch set E max only Figure 2 The solid line is composed of E max ={(1,2)(2,3),…,(7,8)}, the line (i,j) indicates that the nodes at its beginning and end are numbered i and j respectively.

[0103] The step (3) specifically comprises the following steps:

[0104] (3a) According to the area and energy consumption of the planned area, the upper and lower limits of the number of grids in the integrated energy system in the planned area are given; assuming that the area of ​​the planned area is S plan , r min and r max are the minimum and maximum power supply radius of the integrated energy system grid, respectively. The expression of the grid number z of the integrated energy system is:

[0105]

[0106] Where: represents rounding up; the value of the grid number z of the integrated energy system is between z min and z max Between; Set the initial value of the optimization target Let the number of grids in the integrated energy system of the planned area be z = z min ;

[0107] (3b) The longest path through E max According to the requirement of the number of grids in the integrated energy system, it is decomposed into z parts, z-1 split points are selected, and the minimum spanning tree is divided into z independent parts. Let the node set and branch set of each part be G k,0 and E k,0 , k=1,2,…,z, and E k The sum of the weights of the internal branches is w E,k ;

[0108] Dividing the minimum spanning tree into z independent parts must meet the requirements of the following formula, that is, the weights of each independent part should be as close as possible, and can be split into different numbers of parts as required:

[0109] min(|ω E,2 -ω E,1 |+|ω E,3 -ω E,2 |+...+|ω E,z -ω E,z-1 |) (10)

[0110] Attached Figure 3 As shown, the minimum spanning tree and its longest path are split into two parts, and the line segment (3,4) is decoupled. The load matching degree of the left and right parts meets the requirements of formula (10), and the line segment (3,4) can be used as a connecting line, which is disconnected in normal conditions and can be closed in the event of unexpected situations such as failures to ensure the reliability of the entire network function.

[0111] (3c) The node set G k,0 With the remaining node set Perform k-means clustering analysis to obtain z clustering results, that is, the grid division results of the integrated energy system. At this time, the integrated energy system is divided into z grids, and the expression of the node set of each grid is as follows:

[0112]

[0113] in: It is represented as the kth node in the zth grid, n1, n2 and nz represent the number of nodes in the 1st, 2nd and zth grids respectively;

[0114] (3d) A grid-based planning model for an integrated energy system is constructed, and its objective function is as follows:

[0115] f ∑ =min(f I +f 0 ) (12)

[0116] Where: f I Total annual cost for grid and equipment planning; 0 The annual operation and maintenance cost of the grid and equipment;

[0117] f I The expression is as follows:

[0118]

[0119] Where: r 0 is the annual interest rate; T l and T q are the service life cycles of branch l and integrated energy equipment q in the grid respectively; C l is the investment cost per unit length of branch l, C q is the unit capacity investment cost of the integrated energy equipment q in the grid; L l is the actual length of branch l after taking into account geographical obstacles; q is the capacity of the integrated energy equipment q;

[0120] f 0 The expression is as follows:

[0121]

[0122] Where: C 0 is the grid electricity price; ΔP l is the active power loss of branch l; τ max,l is the annual maximum load loss hours of branch l;

[0123] (3e) The constraints of the grid-based planning model for the integrated energy system include:

[0124] (3e1) Voltage U at node k k or air pressure p k The voltage quality constraints must be met, namely:

[0125]

[0126] Where: U k,max and U k,min are the upper and lower limits of the voltage at node k during normal operation; p k,max and p k,min are the upper and lower limits of the air pressure at node k during normal operation;

[0127] (3e2) The current I in branch lm lm Or gas flow fl lm The upper limit constraint must be satisfied, namely:

[0128]

[0129] Where: I lm,max is the upper limit of the current of branch lm during normal operation; fl lm,max It is the upper limit of the airflow of the branch 1m during normal operation;

[0130] (3e3) Network radiation constraint: Ensure that there is no loop when the system is operating normally, and satisfy the following formula:

[0131]

[0132] Where: B is the total number of branches in the network; b is the branch number; x b is a 0-1 variable representing the state of the b-th branch. When the branch is closed, the value is 1, and when the branch is open, the value is 0; N is the total number of nodes in the network; Ns is the number of power sources in the network; M d is the number of branches in the dth loop, d = 1, 2, ..., L; L is the total number of loops in the network; x dm is the state of the mth branch in the dth loop;

[0133] (3f) Solve the grid planning model of the integrated energy system and obtain the optimal value of the objective function f ∑ , and For comparison: If And z <z max , then let Update the optimal planning scheme to Ω, set z = z + 1, and then return to step (3b) to continue solving until z = z max , at this time, f ∑ The corresponding optimized planning scheme is the optimal planning scheme Ω.

[0134] In summary, the present invention comprehensively considers the correlation between the cold, hot and electric loads of each node in the integrated energy system, constructs a node load autocorrelation matrix and a node load cross-correlation matrix to reflect the possibility of mutual conversion between different energy sources, fundamentally eliminates the isolated load forecasting in the traditional integrated energy system planning, and improves the accuracy of load forecasting; the present invention introduces the minimum spanning tree technology into the planning of the integrated energy system, and assigns similarity as the weight of the adjacency graph to each edge, retains the most important adjacency relationship, simplifies the topology of the network by constructing a minimum spanning tree, and preliminarily realizes the grid planning of integrated energy, greatly reduces the number of decision variables in the grid planning model of the integrated energy system, reduces the planning difficulty, improves the solution efficiency, and significantly improves the intelligence and automation level of the planning method.

Claims

1. A grid-based planning method for an integrated energy system based on minimum spanning tree splitting, characterized in that: The method comprises the following steps in order: (1) Analyze the correlation between the cold, hot, and electric loads connected to each node in the integrated energy system network and construct the node load autocorrelation matrix c ii,n and the node load cross-correlation matrix c ij,n , where i and j are node numbers respectively; n = 1, 2, 3, 4 represent typical days in spring, summer, autumn and winter respectively; (2) Based on the node load autocorrelation matrix c ii,n , node energy demand and the physical distance between nodes, and construct the minimum spanning tree of the integrated energy system based on the universal gravitation model and its longest path E max ; (3) Construct a minimum spanning tree and its longest path E max The grid planning model of the integrated energy system is proposed and solved to obtain the optimal planning scheme Ω.

2. The method for grid planning of an integrated energy system based on minimum spanning tree splitting according to claim 1 is characterized in that: The step (1) specifically comprises the following steps: (1a) Assume that the electricity, cooling and heating loads of node i in a typical day n are e, h, and c represent the electrical load, heating load, and cooling load, respectively. The maximum load in a typical day is defined as After normalization, the 24-hour electricity, cooling, and heating loads of node i in a typical day n are: In the formula, in the formula, is the normalized electric, heating and cooling load vector of node i in a typical day n; (1b) Construct node load autocorrelation matrix c ii,n , as shown below: in, and denote the correlations between the electric-electric load, heat-heat load, and cooling-cooling load of node i in a typical day n, and They represent the correlation between the electricity-heating load, electricity-cooling load, and heat-cooling load of node i in a typical day n, and the matrix c ii,n The specific parameters are expressed as: In the formula, is the normalized electricity, heating and cooling loads of node i at time t in a typical day n; (1c) construct the node load correlation matrix c ij,n , as shown below: in, and denote the correlations between the electric-electric load, heat-heat load, and cooling-cooling load of nodes i and j in a typical day n, respectively. and represents the correlation between the electric-heating load, electric-cooling load, and heat-cooling load of node i and node j in a typical day n, respectively. The matrix c ij,n Specific parameters within It is expressed as: In the formula, are the normalized electric, heating and cooling loads of node j at time t in a typical day n.

3. The grid planning method for an integrated energy system based on minimum spanning tree splitting according to claim 1 is characterized in that: The step (2) specifically comprises the following steps: (2a) Consider the two nodes i and j in the integrated energy system network as two mass points in space, each of which has its own inherent mass and the similarity of load complementary matching; let x i,n,max represents the maximum value of the electricity, cooling and heating loads of node i in a typical day n, represents the maximum value of the electrical load of node i in a typical day n, represents the maximum cooling load of node i in a typical day n, represents the maximum value of the heat load of node i in a typical day n. According to the universal gravitational model, the energy gravitational force f between two nodes i and j is defined as ij ' ,n for: Where, L ij is the network distance between two nodes i and j; c ii,n and c jj,n are all node load autocorrelation matrices; x j,n,max is the maximum value of the electricity, cooling and heating loads of node j in a typical day n; and are the magnitudes of the attractive forces of electricity, heat and cooling loads between two nodes i and j respectively; Introducing network weight coefficients on different typical days e, h, and c represent the electric load, heat load, and cooling load, respectively. The energy attraction between two nodes i and j under different typical days is calculated to obtain the comprehensive energy attraction f between nodes i and j. ij : In the formula, and They represent the network weight coefficients of the electric network, the heat network, and the cold network in a typical day n, respectively; make ω ij It is regarded as the weight of the branch ij with two end nodes i and j. Obviously, the smaller the weight of the branch, the higher the similarity of the two nodes it connects, and the more likely it is to be divided into the same grid; (2b) Arrange all possible branches in the integrated energy system network according to their weights from small to large, and then select each edge in order. If the nodes at both ends of the branch do not belong to the same set, merge them until all nodes belong to the same set, thus forming a minimum tree. Specifically, it includes the following four steps: (2b1) The set of all nodes in the entire area to be planned is regarded as a forest G, and each node is regarded as an independent tree; (2b2) ​​Add all branches with weights to the set S = {ω ij } (2b3) Take the branch (i, j) with the smallest weight from the set S, that is, ω ij is the minimum value in the set S. If (i, j) is not in the same tree, then connect i and j to merge the two trees, and add (i, j) to the branch set E of the spanning tree. (2b4) Repeat step (2b3) until all nodes belong to the same tree and the branch set E is a minimum spanning tree; (2c) Select any node k in the forest G, perform a depth-first traversal based on the connectivity relationship of the branch set E, and accumulate the weights of its branches to obtain the longest path. The branch set composed of the longest path is recorded as E. k Similarly, traverse all nodes in the forest G and select the longest branch set E from them max ,Right now E max This is the longest path of the minimum spanning tree.

4. The method for grid planning of an integrated energy system based on minimum spanning tree splitting according to claim 1 is characterized in that: The step (3) specifically comprises the following steps: (3a) According to the area and energy consumption of the planned area, the upper and lower limits of the number of grids in the integrated energy system in the planned area are given; assuming that the area of ​​the planned area is S plan , r min and r max are the minimum and maximum power supply radius of the integrated energy system grid, respectively. The expression of the grid number z of the integrated energy system is: Where: represents rounding up; the value of the grid number z of the integrated energy system is between z min and z max Between; Set the initial value of the optimization target Let the number of grids in the integrated energy system of the planned area be z = z min ; (3b) The longest path through E max According to the requirement of the number of grids in the integrated energy system, it is decomposed into z parts, z-1 split points are selected, and the minimum spanning tree is divided into z independent parts. Let the node set and branch set of each part be G k,0 and E k,0 , k=1,2,…,z, and E k The sum of the weights of the internal branches is w E,k ; Dividing the minimum spanning tree into z independent parts must meet the requirements of the following formula: min(|ω E,2 -oh E,1 |+|ω E,3 -oh E,2 |+...+|ω E,z -oh E,z-1 |) (10) (3c) The node set G k,0 With the remaining node set Perform k-means clustering analysis to obtain z clustering results, that is, the grid division results of the integrated energy system. At this time, the integrated energy system is divided into z grids, and the expression of the node set of each grid is as follows: in: It is represented as the kth node in the zth grid, n1, n2 and nz represent the number of nodes in the 1st, 2nd and zth grids respectively; (3d) A grid-based planning model for an integrated energy system is constructed, and its objective function is as follows: f ∑ =min(f I +f0) (12) Where: f I is the annual total cost of grid and equipment planning; f0 is the annual operation and maintenance cost of grid and equipment; f I The expression is as follows: Where: r0 is the annual interest rate; T l and T q are the service life cycles of branch l and integrated energy equipment q in the grid respectively; C l is the investment cost per unit length of branch l, C q is the unit capacity investment cost of the integrated energy equipment q in the grid; L l is the actual length of branch l after taking into account geographical obstacles; q is the capacity of the integrated energy equipment q; The expression of f0 is as follows: Where: C0 is the power grid price; ΔP l is the active power loss of branch l; τ max,l is the annual maximum load loss hours of branch l; (3e) The constraints of the grid-based planning model for the integrated energy system include: (3e1) Voltage U at node k k or air pressure p k The voltage quality constraints must be met, namely: Where: U k,max and U k,min are the upper and lower limits of the voltage at node k during normal operation; p k,max and p k,min are the upper and lower limits of the air pressure at node k during normal operation; (3e2) The current I in branch lm lm Or gas flow fl lm The upper limit constraint must be satisfied, namely: Where: I lm,max is the upper limit of the current of branch lm during normal operation; fl lm,max It is the upper limit of the airflow of the branch circuit 1m during normal operation; (3e3) Network radiation constraint: Ensure that there is no loop when the system is operating normally, and satisfy the following formula: Where: B is the total number of branches in the network; b is the branch number; x b is a 0-1 variable representing the state of the b-th branch. When the branch is closed, the value is 1, and when the branch is open, the value is 0; N is the total number of nodes in the network; Ns is the number of power sources in the network; M d is the number of branches in the dth loop, d = 1, 2, ..., L; L is the total number of loops in the network; x dm is the state of the mth branch in the dth loop; (3f) Solve the grid planning model of the integrated energy system and obtain the optimal value of the objective function f ∑ , and For comparison: If And z <z max , then let Update the optimal planning scheme to Ω, set z = z + 1, and then return to step (3b) to continue solving until z = z max , at this time, f ∑ The corresponding optimized planning scheme is the optimal planning scheme Ω.

Images