Power system reserve capacity measuring and calculating method based on multi-constraint security domain projection

Abstract

The invention provides a power system reserve capacity measuring and calculating method based on multi-constraint security domain projection, and relates to the technical field of power grid operation control. A power grid power flow fluctuation characteristic quantity is extracted by using an interval power flow method to construct a state space representing a power grid operation state, and the state space is linearly transformed to a two-dimensional projection plane; obtaining an operation domain and introducing section limit constraint to obtain a single-section limit constraint security domain, respectively depicting multi-section constraint security domain boundaries of two different operation scenes by taking maximization of the sum of global areas and maximization of a given power node operation margin as principles, and establishing a power node operation state random fluctuation model; an effective reserve formula is given according to the shortest distance and direction between the fluctuation boundary and the boundary of the multi-section limit constraint security domain corresponding to each section in any operation scene, so that the reserve capacity of the power system is measured and calculated, and the method can adapt to massive working condition combinations caused by high-proportion new energy output random fluctuation; and safe and orderly operation of the power grid under random fluctuation is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid operation control, and particularly relates to a method for calculating reserve capacity of a power system based on multi-constraint security domain projection. Background Art

[0002] In the context of high-proportion new energy access to the power grid, the output of new energy generation represented by wind power and photovoltaic power is directly coupled with meteorological parameters. Due to the strong uncertainty of weather change processes and the low accuracy and spatio-temporal refinement of existing meteorological information forecasting technologies, the output of a large amount of new energy shows great randomness, volatility, and intermittency, which are concentrated in the fact that the net injection power at each node of the power grid fluctuates randomly to varying degrees with the prediction error of new energy output.

[0003] When the power flow of new energy and load nodes in the power grid shows random fluctuations simultaneously, the number of their combination modes is astonishing, resulting in the inability to check each line congestion caused by the combination of node power flow fluctuations one by one. The incomplete pre-check of line congestion often leads to insufficient network-wide reserves or uncontrollable section limits, bringing serious accident losses to the power grid. Therefore, the application of traditional power grid dispatching theories and technologies in new power systems has become lagging and ineffective. Only checking typical working conditions with only statistical significance or reviewing and learning the control strategies for historical working conditions is far from being able to adapt to the massive working condition combinations caused by the random fluctuations of new energy output at present. Power dispatching operation control has gradually entered a "no man's land".

[0004] In summary, in a power system with high-proportion new energy access, the strong randomness, volatility, and intermittency of new energy output lead to a sharp increase in the number of power flow combination working conditions of power grid nodes. Existing methods are difficult to adapt to the massive working conditions, easily causing line congestion and insufficient reserve capacity, threatening the safe operation of the power grid.

[0005] Based on this, the present invention is proposed. Summary of the Invention

[0006] The object of the present invention is to provide a method for calculating reserve capacity of a power system based on multi-constraint security domain projection, which can adapt to the massive working condition combinations caused by the random fluctuations of high-proportion new energy output, avoid accidents such as insufficient network-wide reserve capacity or uncontrollable section limits caused by line congestion, and ensure the safe and orderly operation of the power grid under random fluctuations.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A method for calculating reserve capacity of a power system based on multi-constraint security domain projection, comprising:

[0009] S1: Extract the power flow fluctuation characteristic quantities of the power grid by the interval power flow method to construct a state space representing the operating state of the power grid, linearly transform the state space to a two-dimensional projection plane reflecting the relationship between section power and system power, and then obtain the operating domain satisfying the constraint conditions by determining the upper and lower boundary functions;

[0010] S2: Introduce the section limit constraint on the basis of the operating domain to obtain the single-section limit constraint safety domain;

[0011] S3: Combine the acquisition method of the single-section limit constraint safety domain and, with the principles of maximizing the sum of the global areas and maximizing the operating margin of the given power nodes, depict the boundaries of two multi-section limit constraint safety domains respectively for the operating scenarios of unknown current operating states of power nodes and known current operating states of power nodes;

[0012] S4: Establish a stochastic fluctuation model for the operating states of power nodes, give an effective reserve formula according to the shortest distance and direction between the fluctuation boundary of the model and the boundaries of the multi-section limit constraint safety domains corresponding to each section in any of the above-mentioned operating scenarios, and calculate the reserve capacity of the power system through the effective reserve formula.

[0013] The present invention provides a preferred solution. S1 includes: S11: Introduce interval variables to model the fluctuation ranges of new energy and load and the adjustable range of units to form a power flow equation containing interval variables, and define the nodes with new energy and / or load as power nodes; S12: Obtain the initial state points of the unit nodes and power nodes in space at any moment as the state space, and calculate the branch power flows between the nodes; S13: Construct a two-dimensional projection plane 1 with section power as the y-axis and system power as the x-axis according to the branch power flows, system total power and power flow equation, and set the section limit constraint and balance power constraint, and split and map the initial state points on the two-dimensional projection plane 1 to obtain the unit and power state points of the first projection; S14: Combine the transformation of the system power with the deformation of the power balance constraint, and split and map the initial state points on the two-dimensional projection plane 2 with section power as the y-axis and the transformed system power as the x-axis to obtain the unit and power state points of the second projection; S15: Define the upper and lower boundary functions of the unit and power state points of the second projection respectively, and obtain two piecewise linear and closed polygon envelopes, and the enclosed areas are the operating domains of the units and power nodes respectively.

[0014] More preferably, in S11, the modeling adopts the affine-linear optimization method, incorporates the adjustable range of units and the fluctuation ranges of new energy and load nodes into the node injection power respectively, and establishes an affine-linear equation representing the fluctuation.

[0015] The present invention provides a preferred solution. S2 includes: transforming the section power in combination with the deformation under the section limit constraint, performing a transformation on the unit node operating region obtained in S15 by first symmetrically flipping it about the x-axis and then translating it upward by a distance equal to the section limit size to obtain the transformed unit operating state, and completing multiple split mappings of the initial state point on the two-dimensional projection plane three with the transformed section power as the y-axis and the transformed system power as the x-axis, so as to obtain the unit and power state points of the three projections under the single-section constraint. Define the intersection of the region below the upper boundary function of the transformed unit operating state and the entire region of the power node operating region as the single-section limit constraint safety region.

[0016] The present invention provides a preferred solution. In S2, a unit scheduling scheme that maximizes the area of the single-section limit constraint safety region is given.

[0017] More preferably, in S2, the unit scheduling scheme that maximizes the area of the single-section limit constraint safety region adopts: the units increase their output in sequence on the upper boundary of the single-section limit constraint safety region.

[0018] The present invention provides a preferred solution. In S3, the method for maximizing the sum of the areas of the entire regions is the method for obtaining the boundary of the safety region with the largest area after bringing all sections into the same coordinate system.

[0019] The present invention provides a preferred solution. In S3, the method for maximizing the given power node operating margin includes: for the initial power state point in the initial state point, setting the section limit and constraint conditions. When there is a solution for the vector in the constraint conditions, setting the objective function and solving the objective function to obtain the optimal solution of the vector; then classifying all units into two segments according to the optimal solution of the vector, the first category being all the segments with increased output, and the second category being all the segments without increased output, and then forming the unit sorting methods for the first category and the second category respectively according to the single-section limit constraint safety region obtaining method, thus forming the boundary of the multi-section limit constraint safety region.

[0020] The present invention provides a preferred solution. S4 includes: S41: Assume that each power node changes with a power state point as the center at a volatility, and the change range is , all power node fluctuations are reflected in the operation domain. With the power state point as the center, the polygon envelope formed by the operation domain of power nodes with the side length reduced to 2λ of the original length is the fluctuation interval polygon envelope. When the volatility λ gradually increases, the fluctuation interval polygon envelope increases proportionally with the power state point as the center; S42: Assume that the center of the power state point function is located inside the safety domain corresponding to the observed cross-section. Connect the center of the power state point function with each vertex of the polygon envelope of the power node fluctuation interval and extend it. The obtained extension line intersects with the boundary of the safety domain to get the intersection point of the extension line and the safety domain boundary. The length of the connection line from the intersection point to the center of the power state point function is d 1 , d 1 and the length d of the connection line from the corresponding vertex to the center of the power state point function 0 When the ratio is the smallest, this d 1 is the shortest fluctuation distance from the center of the power node function to the function of the unit node; S43: Define the product of the shortest fluctuation distance and the reserve coefficient as the effective reserve of this cross-section . When considering all cross-sections of the system, take the minimum value of the effective reserve among them and define it as the system effective reserve R:

[0021] .

[0022] Compared with the prior art, the technical solution of the present invention has the following advantages:

[0023] The present invention deconstructs the operation scenario with a time-varying bounded fluctuation interval of the power flow of the power source nodes for load and new energy grid connection. Conducts linearized modeling for the volatility of load and new energy output, then uses the interval power flow method to examine the impact of the power flow changes of each node under the fluctuation model on the branch power flow, and further constructs a projection model of the safety domain in the two-dimensional state space of section power - balanced power. On this basis, a method for calculating the unit reserve considering load and new energy fluctuations is developed. It is a scientific and complete method for efficiently calculating the reserve capacity of a high-proportion new energy power system based on the projection of a multi-constraint safety domain, which can adapt to the massive working condition combinations caused by the random fluctuations of high-proportion new energy output, and can avoid accidents such as insufficient network-wide reserve capacity caused by line congestion or uncontrollable section limits, ensuring the safe and orderly operation of the power grid under random fluctuations. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0025] Figure 1Flow chart of the power system reserve capacity calculation method based on multi-constraint security domain projection provided by a specific embodiment of the present invention;

[0026] Figure 2 Schematic diagram of the operating domain obtained by the power system reserve capacity calculation method based on multi-constraint security domain projection provided by a specific embodiment of the present invention;

[0027] Figure 3 Schematic diagram of the security domain (single-section limit constraint) obtained by the power system reserve capacity calculation method based on multi-constraint security domain projection provided by a specific embodiment of the present invention;

[0028] Figure 4 Position relationship diagram of the operating domain in the inapplicable scenario of the method of maximizing the sum of the global areas in the power system reserve capacity calculation method based on multi-constraint security domain projection provided by a specific embodiment of the present invention;

[0029] Figure 5 Schematic diagram of the security domain delimitation process after obtaining the optimal solution of the unit by the method of maximizing the margin of a given power node in the power system reserve capacity calculation method based on multi-constraint security domain projection provided by a specific embodiment of the present invention;

[0030] Figure 6 Schematic diagram of the position relationship between the power state point fluctuation situation characterized by the polygon envelope of the fluctuation interval and the security domain in the power system reserve capacity calculation method based on multi-constraint security domain projection provided by a specific embodiment of the present invention.

[0031] Figure 7 Schematic diagram of the process of obtaining the effective reserve by the analytical geometry method in the power system reserve capacity calculation method based on multi-constraint security domain projection provided by a specific embodiment of the present invention. Specific embodiment

[0032] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0033] Please refer to Figure 1 , this embodiment provides a power system reserve capacity calculation method based on multi-constraint security domain projection, which is mainly implemented through the following steps:

[0034] S1: Extract the power flow fluctuation characteristic quantities of the power grid by the interval power flow method to construct a state space representing the operating state of the power grid, linearly transform the state space to a two-dimensional projection plane reflecting the relationship between section power and system power, and then obtain the operating domain satisfying the constraint conditions by determining the upper and lower boundary functions;

[0035] S2: Introduce the section limit constraint on the basis of the operating domain to obtain the single-section limit constraint safety domain;

[0036] S3: Combine the acquisition method of the single-section limit constraint safety domain and, taking the maximization of the sum of the global areas and the maximization of the operating margin of the given power nodes as the principles, respectively depict the boundaries of two multi-section limit constraint safety domains for application to the operating scenarios of unknown current operating states of power nodes and known current operating states of power nodes;

[0037] S4: Establish a stochastic fluctuation model for the operating states of power nodes, give an effective reserve formula according to the shortest distance and direction between the fluctuation boundary of the model and the boundaries of the multi-section limit constraint safety domains corresponding to each section in any of the above-mentioned operating scenarios, and measure the reserve capacity of the power system through the effective reserve formula.

[0038] This embodiment deconstructs the operating scenarios with time-varying bounded fluctuation intervals of the power flows of the load and new energy grid-connected power supply nodes, linearly models the fluctuations of the load and new energy output, then uses the interval power flow method to examine the influence of the power flow changes of each node under the fluctuation model on the branch power flow, and further constructs a projection model of the safety domain in the two-dimensional state space of section power - balanced power. On this basis, a method for measuring the unit reserve considering the fluctuations of the load and new energy is developed, which can efficiently measure the reserve capacity, adapt to the massive working condition combinations caused by the random fluctuations of high-proportion new energy output, avoid accidents such as insufficient reserve capacity of the whole network caused by line congestion or uncontrollable section limits, and ensure the safe and orderly operation of the power grid under random fluctuations.

[0039] In a preferred embodiment, on the basis of the above embodiment, step S1: Extract the power flow fluctuation characteristic quantities of the power grid by the interval power flow method to construct a state space representing the operating state of the power grid, linearly transform the state space to a two-dimensional projection plane reflecting the relationship between section power and system power, and then obtain the operating domain satisfying the constraint conditions by determining the upper and lower boundary functions, which is specifically implemented as follows:

[0040] S11: Introduce interval variables to model the fluctuation ranges of new energy and load and the adjustable range of the unit, form a power flow equation containing interval variables, and define the nodes with new energy and / or load as power nodes.

[0041] In this embodiment, interval variables are introduced to model the adjustable range of units in the power system and the fluctuation ranges of new energy and load nodes, forming a power flow equation with interval variables ( and below). The solution set of the equation is presented in the form of upper and lower bound envelope lines, which includes all possible deterministic power flow solutions and directly reflects all possible operating states of the power grid. For the modeling form of interval variables, that is, the power flow equation with interval variables, in a preferred embodiment, the modeling method adopts the affine-linear optimization method. The adjustable range of units and the fluctuation ranges of new energy and load nodes are respectively incorporated into the nodal injection power to establish an affine-linear equation, and the expression is as follows:

[0042] ;

[0043] where represents the affine variable of the unit, new energy or load fluctuation at node i (the units mentioned throughout the text specifically refer to thermal power units), is the fluctuation amplitude of the power at node i, is the fluctuation noise element of the power at node i, and the value range of is [0, 1]. When represents the affine variable of unit fluctuation, is the base value of the unit output, and takes the lower bound when it is a positive value. Since from the perspective of dispatching, the fluctuations of load and new energy output both originate from prediction errors, in this embodiment, the load and new energy output at the same node in the power system are considered together and represented by the same affine-linear fluctuation quantity. A node with at least one of load and new energy is defined as a power node. When represents the affine variable of power node fluctuation, is the difference between the new energy output and the load base value at the node, and usually takes the absolute value upper bound when it is a negative value. Considering the characteristics of the fluctuation quantity of new energy units, since the installed capacity of new energy corresponding to each main transformer is fixed, the fluctuation amplitude caused by the change of new energy output is basically determinable. When a power node only contains new energy units, the maximum fluctuation amplitude can be characterized by a constant value available at each moment. Considering the characteristics of load fluctuation quantity, this embodiment mainly examines the influence of temperature change on the load fluctuation amplitude. Since the temperature change is relatively smooth and the temperature-sensitive load has a large hysteresis, it is applicable to determine the fluctuation interval on a monthly time scale.

[0044] In addition, it should be noted that the affine-linear equation is a representation method of the power of units, load and new energy nodes with two variables, and the power flow equation is an equation for calculating the section power flow based on the power variables of each node and the grid structure data.

[0045] S12: Obtain the initial state points of the unit nodes and power nodes in space at any moment as the state space, and calculate the branch power flow between each node.

[0046] Denote the number of grid units and power nodes as n and m respectively. According to the affine linear equation established in step S11, the operating state of the power grid at time t can be represented as the following state points in an (n + m)-dimensional space:

[0047] ;

[0048] When the reference curves of each node are determined, the state points of the power grid operating state at time t are points that fluctuate around the reference curve with fluctuations.

[0049] According to the DC power flow calculation method, let the inverse matrix of the admittance matrix Y of the power grid nodes (here the nodes refer to the nodes on the grid structure, and this node is a power node and / or a unit node) be B. The element in the k-th row and j-th column of the B matrix is represented by characterized by represents the element in the k-th row and j-th column of the B matrix, is the branch admittance coefficient between nodes kj and = - , to make the Y matrix non-singular, let be the reference node and = 0, then the branch power flow can be expressed as:

[0050] ;

[0051] Among them, is the branch admittance coefficient between nodes jk, represents the element in the k-th row and i-th column of the B matrix, represents the element in the j-th row and i-th column of the B matrix.

[0052] The total system power can be expressed as:

[0053] ;

[0054] S13: Construct a two-dimensional projection plane with the section power as the y-axis and the system power as the x-axis - , and set the section limit constraint and the balance power constraint, and split and map the initial state points on the two-dimensional projection plane - to obtain the unit and power state points of the first projection.

[0055] The affine linear equation representing the fluctuation Substitute into the above two equations ( and ), and denote the sensitivity coefficient of the power of node i to the power flow of branch kj as . is the branch admittance coefficient of the branch section taken for the two-dimensional projection (the y-axis of the two-dimensional projection plane refers to the sectional power flow of the specific section being observed, so each plane corresponds to only one branch). Construct a two-dimensional projection plane of the operating state space with the sectional power as the y-axis and the system power as the x-axis. The expression is as follows:

[0056] ;

[0057] ;

[0058] According to the sectional limit constraint and the power balance constraint, denote the sectional limit of the branch section taken for the two-dimensional projection as , and there is:

[0059] ;

[0060] ;

[0061] To obtain detailed information about the unit nodes and power nodes, in a preferred embodiment, the n unit nodes and m power nodes are separately examined. According to the affine linear equation established in step S11, define the initial unit state point as , the initial power state point as , and correspondingly split the sectional power and the system power into the unit sectional power , the unit system power , the power sectional power and the power system power . Among them, represents the base value of the output of unit node i, represents the fluctuation amplitude of unit node i, represents the fluctuation noise element of unit node i, represents the difference between the new energy output and the load base value at power node i, represents the fluctuation amplitude of power node i, represents the fluctuation noise element of power node i. Then the two constraint equations can be split as follows:

[0062] ;

[0063] ;

[0064] Denote as the influencing parts of unit regulation on the balanced power and section power respectively, as the influencing parts of power node fluctuations on the balanced power and section power respectively, where represents the sensitivity coefficient of the power of unit node i to the power flow of the observed branch, represents the sensitivity coefficient of the power of power node i to the power flow of the observed branch. Define the two-dimensional projection unit state point and power state point using the balanced power and section power, which represents the initial unit state point , the power state point is the projection point of the two-dimensional plane with the section power as the y-axis and the system power as the x-axis, that is, the unit and power state points of the first projection are obtained. Plane - The original constraint equation is simplified to:

[0065] ;

[0066] ;

[0067] where , , , The expressions are:

[0068] ;

[0069] ;

[0070] ;

[0071] ;

[0072] S14: Combine the transformation of the power balance constraint to transform the system power to obtain , and split and map the initial state point on the two-dimensional projection plane - to obtain the unit and power state points of the second projection.

[0073] Transform the power balance constraint expression to obtain the following expression:

[0074] ;

[0075] Denote , , this transformation aims to clear the base value of the unit output so as to intuitively reflect the adjustable amount of the unit at each moment, facilitating the dispatcher to observe and record the unit dispatching plan. After the above transformation, the two-dimensional projected unit state points and the power state points are remapped to the unit state points and the power state points , that is, the unit and power state points of the secondary projection are obtained.

[0076] S15: Define the upper and lower boundary functions of the unit and power state points of the secondary projection respectively, and obtain two piecewise linear and closed polygon envelopes. The enclosed areas are the operation domains of the unit and power nodes respectively.

[0077] For a general piecewise linear function , , is a constant.

[0078] ;

[0079] Denote as the piecewise linear function at , as the piecewise linear function at . It is easy to prove that for any x and any permutation method, there is , that is , constitutes the upper and lower envelopes of all paths of x from 0 to .

[0080] According to the above piecewise linear function theorem, denote = , and define the upper boundary function of all working conditions of the unit state point in the two-dimensional plane as:

[0081] ;

[0082] ;

[0083] Correspondingly, the lower boundary function is:

[0084] ;

[0085] ;

[0086] Similarly, according to the above piecewise linear function theorem, denote = , = , then , then the power state points are defined The upper boundary function for all operating conditions in the two-dimensional plane is:

[0087] ;

[0088] ;

[0089] Correspondingly, the lower boundary function is:

[0090] ;

[0091] ;

[0092] Due to the same starting and ending points, the piecewise linear envelopes of the unit and power node operating states determined by the above formulas can be - In the two-dimensional plane, they respectively enclose closed polygons. The regions enclosed by the two polygon envelopes will contain all operating states of the unit and power node within the fluctuation range. Therefore, the regions enclosed by the polygons are respectively called the unit node operating domain and the power node operating domain. Since the left and right sides of the power balance constraint are respectively the abscissas of the unit and power node, therefore, from the power balance constraint deformed above , for two points representing the operating states of the unit and power node at the same moment, they are always on the same vertical line. Then the system operating domain should satisfy: when = = , and hold simultaneously, that is, the equal parts in the two operating domains . At the same moment, the operating state points are not only within the operating domain envelope, but also need to satisfy the same abscissa. When the number of units and power nodes considered is large enough, this polygon can be approximately regarded as an ellipse, as shown in Figure 2 . Figure 2 In it, the yellow arc represents the upper boundary function of the unit node operation, the black arc represents the lower boundary function of the unit node operation, the red arc represents the upper boundary function of the power node operation, and the blue arc represents the lower boundary function of the power node operation. The enclosed part is the unit node operating domain, The enclosed part is the power node operating domain.

[0093] In a more preferred embodiment, S2: Based on the operating region, introduce section limit constraints to obtain the single-section limit constraint security region, and give the unit scheduling scheme that maximizes the area of the single-section limit constraint security region, which is mainly achieved through the following steps: Combine the deformation of the section limit constraints to transform the section power to obtain , perform a transformation on the unit node operating region obtained in S15 by first making a symmetric flip about the x-axis and then translating it upward by a distance equal to the section limit size to obtain the transformed unit operating state, and complete the transformation of the initial state point in the two-dimensional projection plane with the transformed section power as the y-axis and the transformed system power as the x-axis, and perform multiple split mappings on - to obtain the unit and power state points of the three projections of the single-section constraint. Define the upper boundary function of the transformed unit operating state, and the intersection of the following region and the entire region of the power node operating region is the single-section limit constraint security region. The more detailed steps are as follows:

[0094] Rearrange the section limit constraint inequality to get:

[0095] ;

[0096] Denote , , that is, there is:

[0097] ≥ ;

[0098] From the transformation relationship, it can be seen that the above transformation will only affect the unit operating region obtained in S15 (that is, the unit operating region formed by the envelope of the upper and lower boundary functions of the projection points of the initial unit state points on the two-dimensional plane - ). First, make a symmetric flip of the original unit operating region about the horizontal axis (x-axis), and then translate it upward by a distance , that is, change from Figures 2 to 3 . So far, after multiple rounds of mapping, the initial unit state point , the initial power state point are mapped to the unit state point and the power state point , that is, obtain the unit and power state points of the three projections, and their coordinates satisfy the following characteristics:

[0099] ≥ ;

[0100] ;

[0101] In the formula, is the cross-section power of the transformed unit, is the system power of the transformed unit, is the cross-section power of the transformed power grid, is the system power of the transformed power grid. The geometric meaning of the above expressions is that for any power state point , x = corresponding must exist and be above the power state point . Define the set of all operating state points where the operating state of the power node is within the polygon area enclosed by the upper and lower envelopes of the transformed operating region and satisfies the above position relationship as the "safe region". That is, as Figure 3 shown, the range of the "safe region" in the two-dimensional plane can be described as: the intersection of the area below the upper boundary function of the transformed unit operating state and the entire polygon of the power node operating state. The "safe region" obtained in this step is the single-section limit constraint safe region, and its essence is the characterization of the safe fluctuation range of the power node. Figure 3 In, the part enclosed by the yellow arc and the black arc is the operating region of the unit node before transformation, that is, the Figure 2 obtained in step S15 and enclosed part; the part enclosed by the red arc and the dark blue arc is the operating region of the power node, that is, the Figure 2 obtained in step S15 and enclosed part is the operating region of the power node; the part enclosed by the green arc and the light blue arc is the operating region of the unit node after transformation through step S2; the shaded part is the "safe region" obtained through step S2, that is, the function and enclosed part.

[0102] In a more preferred embodiment, S2: Based on the operating region, sectional limit constraints are introduced to obtain the single-sectional limit constraint security region, and then a unit scheduling scheme that maximizes the area of the single-sectional limit constraint security region is given. The single-sectional limit security region is formed by the intersection of the envelope of the unit operating region and the envelope of the power node operating region according to certain rules. This embodiment accordingly gives a sorting method for increasing the output of the units, and this sorting method is the way of increasing the output that constitutes the boundary of the unit operating region. The unit scheduling scheme that maximizes the area of the single-sectional limit constraint security region adopts: The units increase their output in sequence on the upper boundary of the single-sectional limit constraint security region. Specifically as follows: The upper boundary function of the transformed unit operating state can be considered as the maximum value of the sectional power flow that the observed section can bear caused by the fluctuations of the power nodes when the unit exerts its maximum ability to reduce the sectional power flow. The part less than this maximum value within the power node operating region is the safe range of fluctuations of the power node. For all unit scheduling schemes, it is easy to know that when the area of the security region is the largest, the safety margin of the power node fluctuations is the most sufficient. To make the security region the largest, the scheduling scheme for a single section should make the unit exert its maximum ability to reduce the sectional power flow, that is, take the upper boundary of the unit operating region after multiple projections. The unit scheduling scheme corresponding to this boundary can be described as: When the net load obtained by subtracting the new energy output from the power node load changes from low to high, the unit output starts from the lowest, and in sequence, the output is increased in ascending order according to the sensitivity coefficient from negative to positive and from small to large. In this way of scheduling, the upper boundary of the unit operating region after the second projection can be obtained, and at this time, the area of the unit operating region is the largest, and it has the largest safety margin for different power node fluctuations. According to the above description of the mathematical theorem of the envelope line of the piecewise linear function, for a cluster of piecewise linear functions, only by arranging the function segments in ascending order according to the coefficients from small to large can the upper envelope (upper boundary function) be formed, and the upper envelope corresponds to the above unit scheduling scheme.

[0103] The strategy of increasing the output of the units in sequence on the upper boundary of the single-sectional limit constraint security region ensures the continuity of the unit regulation. As long as the units are adjusted along the upper boundary, it can be ensured that the net load distribution of all power nodes within the security region will not exceed the limit. However, if the net load distribution of the power nodes is outside the security region, no matter how the units are adjusted, it will surely exceed the limit. Therefore, the distance between the power state point within the security region and the boundary can reflect the operating margin. For the security region under the single-section constraint, the boundary determined by the above method is globally optimal.

[0104] S3: Combine the acquisition method of the single-section limit constraint security region, and respectively depict the boundaries of two multi-section limit constraint security regions based on the principles of maximizing the sum of the global areas and maximizing the operating margin of the given power nodes, so as to be respectively applied to the depiction of the security region boundaries in the operating scenarios of unknown and known current operating states of power nodes. Considering the method for obtaining the security region boundary under multi-section constraints, the difficulty lies in that at this time, the unit regulation must simultaneously consider the collaborative impact on all section limits, that is, the maximum delimitation method of the security region is equivalent to the optimal regulation strategy of the unit under the condition of the maximum safety margin of all considered sections. The present invention provides two methods for delimiting the unit security region under multi-section constraints.

[0105] First is the maximum area delimitation method, that is, maximizing the sum of the global areas. That is, after all sections are incorporated into the same coordinate system, the method for obtaining the security region boundary with the largest area is adopted. Therefore, the global region corresponds to all sections. This method can be applied to the depiction of the security region boundary in the operating scenario of unknown current operating state of power nodes. In a preferred embodiment, for S3, the method of maximizing the sum of the global areas is specifically implemented through the following steps:

[0106] S31: Normalize the unit sensitivity coefficients of all sections, and then add the powers of the q sections after normalization to obtain the added section power , whereby the unit state points of the three projections of all single-section limit constraints are mapped to the new two-dimensional plane - . This method can directly observe which power node net load distributions have insufficient margins for a certain unit dispatching method, providing a basis for continuing to deduce the supplementary measures, and is applicable to the safety margin verification during monthly and weekly mode pre-arrangements.

[0107] For the unit three-mapping (projection) section powers of all sections, that is, the transformed unit section powers , denoted as = , first divide by to complete the normalization of the unit sensitivity coefficients, and then add the powers of the q sections after normalization. Denote the added unit section power as , then the expression is as follows: ;

[0108] Among them, represents the sensitivity coefficient of unit node i to section j after normalization, denoted as the joint sensitivity coefficient, , , respectively represent the output base value, fluctuation amplitude and fluctuation noise element of unit node i affecting section j.

[0109] According to the above definition, all the state points of single-section units are mapped to the two-dimensional plane of multi-section collaborative planning - in which each unit state point is mapped to .

[0110] S32: In the two-dimensional plane of multi-section collaborative planning - the unit sensitivity coefficients of all sections after normalization are jointly sorted as the combined unit sensitivity coefficient, that is, the sorting of multi-section normalized sensitivity coefficients is denoted as .

[0111] S33: According to the piecewise linear function theorem, the upper boundary function and the lower boundary function are defined to obtain the unit operation domain in the two-dimensional plane - That is, the unit operation domain in the two-dimensional plane can be obtained according to the method of determining the upper and lower boundary functions in step S15 - .

[0112] S34: Select the unit with the largest combined sensitivity coefficient to increase its output first. When the output of this unit reaches the maximum value, all the sensitivity coefficients corresponding to this unit are removed from the combined sorting, and the combined sorting is updated. Specifically: select the unit with the largest combined sensitivity coefficient to increase its output first, denote the serial number of this unit as u. When the output of this unit reaches the maximum value, all the sensitivity coefficients corresponding to this unit are removed from the combined sorting, and the combined sorting is updated from to .

[0113] S35: Repeat S34 until all the sensitivity coefficients corresponding to all units are removed, and a unit scheduling scheme with the largest sum of global area (also known as: the largest comprehensive safety margin, the safety domain boundary under multi-section constraints) is obtained, which is the best strategy for unit regulation. It can be considered that the above unit scheduling scheme corresponds to the largest area of the unit combined operation domain enclosed by the head-to-tail connection of the line segments in the two-dimensional plane, that is, it contains the most combinations of balance power and section power, and the working conditions that can be represented have the most extensive impact on the power grid. Therefore, it is considered that the The upper bound is the optimal strategy for unit regulation, that is, the boundary of the security region under multi-section constraints. This boundary is the global optimal scheduling scheme for the unit if and only if the upper boundaries of the security regions of all sections are completely inside the operating region of the power node, that is, when the unit regulation margin is low, the obtained scheduling scheme is the optimal scheduling scheme. Otherwise, when the unit regulation margin is high, this scheme will cause margin waste. Regarding the unit regulation margin, the following is a brief explanation: As Figure 4 shown, the green shaded area is the margin waste area, and the blue shaded area is the ideal security region. x 1 , x 2 , x 3 , x 4 , x 5 are the abscissas of the divided intervals. By dividing into four intervals, the position situations of the two operating regions where margin waste will occur and the position situations of the two operating regions where margin waste will not occur are compared and explained. In Region III of the interval [x 3 , x 4 , the security region is completely inside the operating region of the power node, and its upper boundary is completely composed of the unit node function. For intervals II and IV, the upper boundaries of the two sections are different, and the unit node function is greater than the power node function. Using the maximum area method to determine the unit scheduling scheme in these two interval ranges will take the unit node function as the upper boundary of the security region. In fact, limited by the upper boundary of the operating state of the power node, the fluctuation range of the power node is far from the determined upper boundary of the security region. The resulting security capacity margin can be regarded as a waste of unit scheduling resources. The security region obtained by the above maximum area delimitation method is not ideal and is redundant, that is, the solved unit scheduling scheme is not the global optimal scheduling scheme. Therefore, when using the maximum area delimitation method, when the unit regulation margin is low, the obtained scheduling scheme is the optimal scheduling scheme. Since the optimization is complex when the abundance is surplus (the shape of the power node function needs to be considered) and the improvement space is limited, the above maximum area delimitation method can still be used as a relatively fast engineering method under this condition, or as an initial feasible solution for further optimization.

[0114] Another joint delimitation method is the maximum delimitation method for the margin of a given power node, that is, to maximize the operating margin of the power node. This method can be applied to the characterization of the security domain boundary in the operating scenario where the current operating state of the power node is known, that is, this method can quickly identify the operating risks existing in the known (predicted) power nodes and is applicable to the daily security margin check and the intraday rolling check. In a preferred implementation manner, for S3, the method for maximizing the operating margin of a given power node is achieved through the following steps: for the initial power state point in the initial state points, set the section limit and constraint conditions. When there is a solution for the vector in the constraint conditions, set the objective function and solve the objective function to obtain the optimal solution of the vector; then classify all the units into two segments according to the optimal solution of the vector, the first category is all the segments with increased output, and the second category is all the segments without increased output. Then, according to the method for obtaining (solving) the security domain by the single-section limit constraint, form the unit ranking methods for the first category and the second category respectively. Thus, the security domain boundary with multi-section limit constraints is formed. More specifically:

[0115] For any initial power state point , consider the power state point , denote the power state point of the kth section at time t as , the section limit is , and let K be the total number of sections. It should be noted that , different sections are distinguished by the upper right superscript. The lowercase k represents the kth section, which is the section number label, and K is the total number of sections and also the maximum value of k; The section limit is a general term because it does not involve multiple sections. When there are truly multiple sections, many variables need to be distinguished according to the section. The section limit is when there are multiple sections, and many variables need to be distinguished according to the section. All points should satisfy the following constraint conditions. The following constraint conditions are the expansion of the section limit constraint and are the constraints that should be naturally satisfied for each section. The following constraint conditions are the preparations for solving the following optimization problem:

[0116] Among them, represents the influence function of the adjustable part of the unit on the power flow of the kth section except for the baseline output, represents the sensitivity coefficient of the power of unit node i to the power flow of the kth section, represents the section limit of the kth section, represents the section limits of K sections, represents the influence part of all power nodes on the power flow of the kth section, is the solution set vector of the fluctuation noise elements of the power of the unit node, that is, the solution set of the dispatching scheme of the adjustable part of the unit.

[0117] i) If the vector has a solution, then set the objective function as:

[0118] ;

[0119] wherein, represents the sensitivity coefficient of the power of unit node i to the power flow of the jth section, is the fluctuation amplitude of the power of node i, is the fluctuation noise element of the power of unit node i.

[0120] After solving the optimization objective, the obtained can be considered as the optimal solution that simultaneously satisfies the power flow boundary after unit output compensation, the sum of the power flow distances (i.e., the sum of the power flow margins) caused by the load of power nodes and new energy on all sections, and satisfies the constraint equations. Then, according to the solution of the vector all units are divided into two segments and classified. The first category is all the output added sections of the adjustable part of the units at the current moment , and the second category is all the sections that have not added output of the adjustable part of the units at the current moment 1- . Then, according to the single-section limit constraint safety domain solution method, the unit sorting methods for the first category and the second category are respectively formed, that is, the multi-section limit constraint safety domain boundary (the part surrounded by the yellow arc and the black arc and the part surrounded by the dark blue arc and the red arc) obtained by the maximum power node margin determination method is formed. Given the current unit dispatch , the formation process of the safety domain is as shown in Figure 5 . In the figure, the part surrounded by the yellow arc and the black arc and the part surrounded by the dark blue arc and the red arc, that is, the function and surrounded part and and surrounded part, is the multi-section limit constraint safety domain obtained by the maximum power node margin determination method at the known current unit state point. The part surrounded by the green arc and the light blue arc is the multi-section limit constraint safety domain obtained by the maximum area determination method, that is, the function and surrounded part. Among them, is the sum of all section limits . , is: the multi-section safety domain envelope obtained by sequentially reducing / increasing the output of each unit according to the sorting of the multi-section normalized sensitivity coefficients of the units that can reduce / add output according to the current output.

[0121] ii) If the vector has no solution, denote:

[0122] = ;

[0123] ;

[0124] ;

[0125] ;

[0126] ;

[0127] Among them, is the solution set vector of the fluctuation noise elements of the unit node power, is the fluctuation noise element of the power of unit node i, A represents the sensitivity coefficient matrix, represents the sensitivity coefficient of the power of unit node j to the power flow of the i-th section, represents the matrix of the fluctuation amplitude of the power, is the fluctuation amplitude of the power of node i, represents the vector of the section limit adjustment capabilities required for each section of the unit under the influence of the current power state point, represents the section limit of the k-th section, represents the base value of the output of unit node i, represents the influence part of all power nodes on the power flow of the k-th section, represents the product of the sensitivity coefficient matrix and the diagonal matrix of the fluctuation amplitude, which belongs to an intermediate variable.

[0128] In case (ii), the original constraint condition has no solution, which means that the current power operation state point is outside the safety domain corresponding to the current unit state. There is a situation where no matter how much the output is increased, the power operation state point cannot be within the safe range. Therefore, by finding the minimum norm method, the shortest distance from the current power operation state point outside the domain to the safety domain can be found, and then the point on the safety domain closest to the power operation state point outside the domain can be found by this distance. This point is the best scheduling plan that the current unit can adopt to enclose the power operation state point.

[0129] Therefore, taking the minimum norm with respect to the distance from y as the objective function:

[0130] ;

[0131] Among them, , is the new matrix formed by inserting a row of all 1s at the end of A (the optimization problem expressed in matrix form originally lacks the constraint condition of power balance. Therefore, the power balance constraint is added in the form of interpolation), is added As a new matrix formed by the last element, for the matrix generated by the above method , if the sensitivity of the section constraint is linearly related to the added row, then the row where the section is located is removed from the matrix . If the sensitivities of the two section constraints are linearly related, they are combined and considered as the same section. After the above processing, at this time should be row full rank, and the optimization problem can obtain a unique solution, and its solution expression is:

[0132] ;

[0133] Through the above method, the initial value of the unit solution closest to the net load distribution of the power nodes can be given. Then, using the method in the aforementioned i), after forming the boundary of the security region, make up for all the overstepping parts, and a new security region that envelopes the fluctuation range of the power nodes can be formed. The new security region here is a means to forcibly expand the scope of the security region by taking further make-up measures when the original power state point is not within the security region, that is, when the original constraint conditions have no solution, in order to enclose this power state point. It belongs to the subsequent security region make-up measures and is not covered by the technical problems solved by the present invention. Therefore, the specific means are not discussed in the present invention.

[0134] S4: Establish a stochastic fluctuation model for the operating state of power nodes, and give an effective reserve formula according to the shortest distance and direction between the fluctuation boundary of the model (when the model makes an affine transformation outward from the geometric center, a polygonal fluctuation domain is formed, and the boundary of this polygonal fluctuation domain is the fluctuation boundary of the model) and the boundary of the multi-section limit constraint security region corresponding to each section, and measure the reserve capacity of the power system through the effective reserve formula. Calculate an effective reserve for each single-section security region, and then take the minimum value of all reserves. It is specifically implemented through the following steps:

[0135] In the scenario considering the stochastic fluctuation of the net load of power nodes, for each power operating state point in the power operating domain at each moment, there will be a stochastic fluctuation range.

[0136] S41: Assume that each power node changes with a volatility of centered on the power state point , where . Denote as the fluctuation amplitude of the power of node i, and its variation range is . Then, all the fluctuations of the power nodes are reflected in the operating domain, and the fluctuation interval is: centered on the power state point , a polygonal envelope formed by the power node operating domain with the side length reduced to 2 of the original length, which is the polygonal envelope of the fluctuation interval, that is Figure 6The polygonal fluctuation domain geometrically similar to the power node operation domain shown in []. Denote the probability that the power operation state point has a volatility of λ as Φ. For a given probability Φ, the volatility λ corresponding to different time scales is different, that is, the distributions of the power node fluctuation variables at different time scales are different. The smaller the prediction time scale, the smaller the volatility λ. Therefore, the fluctuation interval range of monthly prediction will be much larger than that of day-ahead prediction (considering temperature-sensitive loads), and the day-ahead prediction will be larger than the two-hour-ahead intra-day prediction (only considering the fluctuation of new energy output). In the real-time state, the fluctuation interval shrinks to a state point. It should be noted that d is the long-term fluctuation amplitude of the node originally based on long-term statistics. However, when observing only short-term fluctuations, its fluctuation amplitude is often shorter than the long-term statistical quantity. Therefore, λ represents the ratio relationship of the short-term fluctuation amplitude relative to the long-term statistical amplitude. Since the difference between the long-term and short-term fluctuation scenarios lies only in that the fluctuation amplitude becomes 2λ times the original, and the formation process of the power node operation domain is the same, the side length of the short-term node operation domain relative to the original envelope only becomes 2λ times the original. According to the above characterization of the power node fluctuation in the two-dimensional plane of the operation domain, it can be considered that the power node fluctuation interval can be characterized by a polygon envelope of a piecewise linear boundary function. When the volatility λ gradually increases, the polygon envelope of the fluctuation interval expands proportionally with the power state point as the center, that is, an affine transformation with as the center, as shown in Figure 6 .

[0137] S42: Assume that the center of the power state point function (power state point) is located inside the safety domain corresponding to the observed cross-section. Connect the center of the power state point function with each vertex of the polygon envelope of the power node fluctuation interval and extend it. The extended line intersects the safety domain boundary to obtain the intersection point of the extended line and the safety domain boundary. The length of the line connecting the intersection point to the center of the power state point function is d 1 , and d 1 and the length of the line connecting the corresponding vertex to the center of the power state point function d 0 . When the ratio is the smallest, this d 1 is the shortest fluctuation distance from the center of the power node function to the center of the unit node function. The product of the shortest fluctuation distance and the reserve coefficient is defined as the effective reserve of the observed cross-section. The solution process of the effective reserve by the analytical geometry method under a single cross-section is shown in Figure 7 , as shown in Figure 7 , Indicates the effective reserve of the observed section (section effective reserve), which is represented as a line segment in the figure; i represents the power node number, that is, power node i, which can be understood as any power node. Power node i is the intersection point of the extension line and the fluctuation boundary (i.e., the envelope polygon of the power node fluctuation range in the figure), which is represented as a point in the figure; j' represents the clockwise numbering of the unit nodes determined by the maximum bounding method according to the given power node margin corresponding to the net load growth direction of the power nodes in the kth section, which is the unit node number corresponding to the intersection point of the extension line and the safety domain boundary, and is represented as a point on the figure. When considering all sections of the system, take the minimum value of the section effective reserve, which is defined as the system effective reserve R:

[0138] ;

[0139] According to the above solution process of the section effective reserve By using the analytical geometry method to obtain , The expression can be represented as follows:

[0140] ;

[0141] ;

[0142] ;

[0143] ;

[0144] ;

[0145] ;

[0146] ;

[0147] ;

[0148] Among them, is the horizontal angle between the center point of the operating area boundary function of the power nodes in the kth section and the corresponding point of the ith power node, is the unit reserve coefficient, is the fluctuation amplitude of the ith power node, is the maximum peak shaving capacity of the unit at the ith node, is the value of the fluctuation noise element of the ith node at time t, G is the admittance coefficient of the kth section, is the sensitivity coefficient of the ith power node to the kth section, is the sensitivity coefficient of the unit at the ith node to the kth section, represents the sensitivity coefficient of the unit at the node to the kth section is the upper bound of the net load of the power node, is the lower bound of the unit output, is the difference between the active power corresponding to the th node of the unit node operating region boundary function of the kth section and the active power corresponding to the center point of the power node operating region boundary function, is the difference between the section active power corresponding to the th node of the unit node operating region boundary function of the kth section and the section active power corresponding to the center point of the power node operating region boundary function. m is the total number of power nodes, n is the total number of unit nodes, K is the total number of sections, and j' is the clockwise numbering of the unit nodes under the unit scheduling scheme determined by the maximum bounding method according to the given power node margin corresponding to the net load growth direction of the power node in the kth section.

[0149] It should be noted that the power node operating state discussed in the present invention is carried out on the assumption that the net load of each power node has not reached the extreme value in advance. This can be used as a simple model during monthly checking, but the impact when the net load of the power node fluctuates and touches the limit needs to be considered during day-ahead or intra-day checking. Because the net load of the power node that reaches the extreme value will not continue to fluctuate according to the ideal fluctuation range, thus changing the shape of the random fluctuation range. In this case, the desired effective reserve direction .

[0150] In summary, based on the above embodiments, compared with the prior art, the present invention can achieve the following beneficial effects:

[0151] 1. The present invention models the new energy output, load fluctuation and unit adjustable range through interval variables, forming an affine linear equation solution set covering all possible operating states. This method can comprehensively capture the randomness, volatility of new energy and load uncertainty, avoid the checking omission problems caused by traditional methods ignoring extreme working conditions or combined scenarios, and significantly improve the system robustness. At the same time, by obtaining the operating region (unit and power node operating regions), the distribution states of new energy fluctuations and load fluctuations and the changes in unit output states can be incorporated into a unified operating state domain, and the influence of unit output combinations and random fluctuations such as new energy and load on the section power flow is quantified. By mapping the operating boundaries accommodating all unit operating states to the two-dimensional plane of section power - balanced power, it can provide theoretical support for the safety monitoring of unit and power node operating states.

[0152] 2. The present invention linearly maps a high-dimensional state space to a two-dimensional plane reflecting section power and balanced power, and constructs a safety region satisfying multiple constraints through geometric operations such as flipping and translation. It transforms a complex high-dimensional optimization problem into an intuitive two-dimensional geometric analysis, greatly reducing the computational complexity. At the same time, through the visual boundary of the safety region, it is convenient for dispatchers to monitor the system safety margin in real time. Meanwhile, the proposal of the safety region enables the effective quantification and visualization of the safety margins under single-section and multi-section conditions considering unit regulation, new energy, and load fluctuations. By giving two regulation methods with the maximum safety margins of the units, the regulation efficiency of the units for section power flow is improved. The interval power flow method provides full-condition coverage, and the two-dimensional projection simplifies the computational complexity. The combination of the two achieves efficient analysis while ensuring accuracy.

[0153] 3. The present invention proposes two principles for depicting the safety region boundary. One is the maximum area delimitation method: through the joint sorting of normalized sensitivity coefficients, the units with the greatest impact on multiple sections are preferentially regulated to ensure the maximum total safety margin of the whole region, which is applicable to medium- and long-term planning. The other is the maximum power node margin delimitation method: using the minimum norm optimization to approximate the net load distribution of power nodes, quickly identifying and compensating the risk points of key sections, which is applicable to short-term rolling checking. The two methods are respectively applied to the operation scenarios of unknown and known current operating states of power nodes, taking into account global optimization and local risk prevention and control, and improving the regulation efficiency and safety under multi-section constraints. Moreover, the maximum area method is applicable to long-term planning, and the given margin method is suitable for short-term checking. The two methods cooperate to meet the reserve requirements of different time scales.

[0154] 4. The introduction and solution of the effective reserve concept can not only be used to obtain the reserve direction with the greatest overstep possibility under the net load fluctuation of power nodes, so that the added reserve can improve the overstep situation of the observed section to the greatest extent, but also make the risk perception of grid section over-limit more refined. By introducing the reserve with section limit constraints, the chain reaction caused by section over-limit accidents due to the overly optimistic traditional reserve calculation method can be avoided, and the dispatching loss caused by the hard reserve gap can be avoided.

[0155] In summary, the present invention can cover the full-dimensional uncertainties of new energy fluctuations, load changes, and unit regulation, and solve the problem of massive working condition checking. Through two-dimensional projection and geometric operations, it transforms complex optimization problems into intuitive calculations, significantly improving the efficiency of reserve capacity measurement. In addition, multi-section collaborative optimization and effective reserve directional configuration ensure the safe and stable operation of the system in high-proportion new energy scenarios. Moreover, through the effective reserve calculation of the present invention, the reserve demand can be accurately quantified, redundant configuration can be reduced, and the grid operation cost can be lowered.

[0156] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above method embodiments. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM), etc.

[0157] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification. Moreover, the above embodiments only represent several implementation manners of the present invention, and the description is relatively specific and detailed, but it cannot be understood as a limitation on the scope of the present invention. For those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention.

Claims

1. A method for calculating the reserve capacity of a power system based on multi-constrained security region projection, characterized in that: include: S1: Use the interval power flow method to extract the power grid power flow fluctuation characteristics to construct a state space that characterizes the power grid operation state, linearly transform the state space into a two-dimensional projection plane that reflects the relationship between the section power and the system power, and then obtain the operation domain that meets the constraint conditions by determining the upper and lower boundary functions; S2: Introduce section limit constraints on the basis of the operation domain to obtain a single section limit constraint safety domain; S3: Combining the acquisition method of the single-section limit constraint safety domain and based on the principle of maximizing the sum of the whole domain area and maximizing the operating margin of a given power node, two multi-section limit constraint safety domain boundaries are depicted respectively, so as to be applied to the operation scenarios of unknown current power node operation status and known current power node operation status respectively; S4: Establish a random fluctuation model of the operating status of power nodes, and give an effective backup formula according to the fluctuation boundary of the model and the shortest distance and direction of the multi-section limit constraint safety domain boundary corresponding to each section under any of the operating scenarios, and calculate the power system backup capacity through the effective backup formula.

2. The method for calculating the reserve capacity of a power system based on multi-constrained security region projection according to claim 1 is characterized in that: S1 includes: S11: Introduce interval variables to model the fluctuation range of new energy and load and the adjustable range of the unit, form a power flow equation containing interval variables, and define nodes with new energy and / or load as power nodes; S12: obtaining the initial state point of the unit node and the power node in the space at any time as the state space, and calculating the branch flow between each node; S13: construct a two-dimensional projection plane 1 with the section power as the y-axis and the system power as the x-axis according to the branch power flow, the total system power and the power flow equation, set the section limit constraint and the balanced power constraint, split and map the initial state point on the two-dimensional projection plane 1, and obtain the unit and power state point of the first projection; S14: transforming the system power in combination with the deformation of the power balance constraint, splitting and mapping the initial state point on a two-dimensional projection plane 2 with the section power as the y-axis and the transformed system power as the x-axis, to obtain the secondary projected unit and power state point; S15: respectively define the upper and lower boundary functions of the secondary projected unit and power state point, and obtain two piecewise linear and closed polygonal envelopes, and the enclosed areas are the unit and power node operation domains respectively.

3. The method for calculating the reserve capacity of a power system based on multi-constrained security region projection according to claim 2 is characterized in that: In S11, the modeling adopts the affine-linear optimization method, and the adjustable range of the unit and the fluctuation range of the new energy and load nodes are respectively included in the node injection power to establish an affine linear equation that characterizes the fluctuation.

4. The method for calculating the reserve capacity of a power system based on multi-constrained security region projection according to claim 2 is characterized in that: S2 includes: The section power is transformed in combination with the deformation of the section limit constraint. The unit node operation domain obtained by S15 is first symmetrically flipped about the x-axis and then translated upward. The translation distance is the section limit size to obtain the transformed unit operation state. The initial state point is split and mapped multiple times on the two-dimensional projection plane with the transformed section power as the y-axis and the transformed system power as the x-axis to obtain the three-projected unit and power state points of the single section constraint. The intersection of the area below the upper boundary function of the transformed unit operation state and the entire domain of the power node operation domain is defined as the single section limit constraint safety domain.

5. The method for calculating the reserve capacity of a power system based on multi-constrained security region projection according to claim 2 is characterized in that: In S2, a unit dispatching plan is given to maximize the single-section limit constraint safety domain area.

6. The method for calculating the reserve capacity of a power system based on multi-constrained security region projection according to claim 5 is characterized in that: In S2, the unit dispatching plan that maximizes the area of ​​the single-section limit constraint safety domain is: the units increase their output in sequence at the upper boundary of the single-section limit constraint safety domain.

7. The method for calculating the reserve capacity of a power system based on multi-constrained security region projection according to claim 4 is characterized in that: In S3, the method for maximizing the sum of the whole area is to include all cross sections in the same coordinate system and then obtain the boundary of the safety zone with the largest area.

8. The method for calculating the reserve capacity of a power system based on multi-constrained security region projection according to claim 2 is characterized in that: In S3, the method for maximizing the operating margin of a given power node includes: for the initial power state point in the initial state point, setting section limits and constraints, and when the vector in the constraint condition has a solution, setting an objective function, and solving the objective function to obtain the vector optimal solution; then, according to the vector optimal solution, all units are divided into two sections and classified, the first category is all the sections with increased output, and the second category is all the sections without increased output, and then, according to the single section limit constraint safety domain acquisition method, the first and second category unit sorting methods are formed respectively, thereby forming the multi-section limit constraint safety domain boundary.

9. The method for calculating the reserve capacity of a power system based on multi-constrained security domain projection according to claim 2, characterized in that S4 include: S41: Assume that each power node is centered on the power state point. Volatility changes, ranging from , then all power node fluctuations are reflected in the operation domain. The polygonal envelope formed by the power node operation domain with the power state point as the center and the side length reduced to the original length 2λ is the fluctuation range polygonal envelope; when the volatility λ gradually increases, the fluctuation range polygonal envelope increases proportionally with the power state point as the center; S42: Assuming that the center of the power state point function is located inside the safety domain corresponding to the observed section, connect the center of the power state point function and each vertex of the envelope polygon of the power node fluctuation interval and extend it, and the obtained extension line intersects with the boundary of the safety domain, and obtains the intersection of the extension line and the safety and boundary. The length of the line from the intersection to the center of the power state point function is d1. When the ratio of d1 to the length d0 of the line from the corresponding vertex to the center of the power state point function is the smallest, d1 is the shortest fluctuation distance from the center of the power node function to the unit node function; S43: The shortest fluctuation distance and the reserve coefficient The product of is defined as the effective reserve of the section , when considering all sections of the system, take the minimum effective reserve value, which is defined as the system effective reserve R: 。 10. The method for calculating the reserve capacity of a power system based on multi-constrained security region projection according to claim 9, characterized in that: According to S42 and S43, the effective reserve is obtained by analytic geometry method. .

Images