Power distribution network operation stability evaluation method considering entropy space

Abstract

The invention discloses a power distribution network operation stability evaluation method considering an entropy space, and the method comprises the steps: constructing a distributed power supply dynamic model, and designing a distributed power supply energy function; based on a distributed power supply energy function, selecting an energy value of the dominant unstable equilibrium point as critical energy, and fitting a critical energy curved surface; constructing a system entropy index representing power flow balance and static voltage safety of the power distribution network, and dispersing the system entropy index to a distributed power supply level to obtain a distributed power supply entropy index; designing a distributed power supply entropy boundary condition, and constructing a distributed power supply entropy space representing safety and stability based on a boundary projection method; and rapidly solving a steady-state working point and an energy value under a given anticipated accident, rapidly positioning inside and outside an entropy space, and evaluating the operation stability of the power distribution network based on a positioning result and the entropy space of the distributed power supply. Compared with the prior art, the operation stability of the power distribution network is evaluated according to the entropy space of the distributed power supply, and a traditional safety and stability evaluation method is simplified.

Description

Technical Field

[0001] The invention relates to a distribution network operation stability evaluation method taking entropy space into account, and belongs to the field of distribution network optimization. Background Art

[0002] Distributed power sources have promoted the construction of distribution networks in new forms and promoted the clean and low-carbon transformation of distribution networks. At the same time, they have also had varying degrees of impact on the safe, economical and reliable operation of distribution networks. It is necessary to strengthen the construction of distribution networks in a targeted manner, evaluate the carrying capacity of distribution networks, and guide the scientific layout, orderly development, local access and local consumption of distributed new energy, so as to reduce the impact of the volatility and randomness of distributed power sources on the safety and stability of distribution networks. Summary of the invention

[0003] Purpose of the invention: In view of the shortcomings of the prior art, the present invention proposes a distribution network operation stability evaluation method taking into account the entropy space.

[0004] Technical solution: The present invention discloses a method for evaluating the operation stability of a distribution network taking into account entropy space, comprising the following steps:

[0005] Step 1: Construct a dynamic model of distributed generation, perform price reduction on it, and then use the first integration method to design a distributed generation energy function that characterizes the multi-dimensional stability characteristics of the distribution network;

[0006] Step 2: Based on the constructed distributed power energy function, the energy value of the dominant unstable equilibrium point is selected as the critical energy, and the critical energy surface at different working points is fitted based on the hyperplane piecewise affine method;

[0007] Step 3: Normalize the power flow balance entropy index and the voltage safety entropy index to obtain the system entropy index that characterizes the power flow balance and static voltage safety of the distribution network. Based on the branch and bound method, the system entropy index is dispersed to the distributed power level to obtain the distributed power entropy index that characterizes the power flow balance and static voltage safety.

[0008] Step 4: Design the entropy boundary conditions of the distributed power supply. According to the boundary conditions, construct the entropy boundary of the distributed power supply on the two-dimensional PQ plane. According to the critical energy surface fitted in step 2, construct the entropy space of the distributed power supply that characterizes the safety and stability based on the boundary projection method.

[0009] Step 5: Quickly solve the energy value of the steady-state operating point and the given expected accident, quickly locate it inside and outside the entropy space, and evaluate the stability of the distribution network operation based on the positioning results and the distributed power entropy space obtained in step 4.

[0010] Furthermore, the dynamic model of the distributed power supply in step 1 is specifically as follows:

[0011] The distribution network topology is analyzed and modeled as an undirected graph to describe the connection relationship between distributed power nodes. The distributed power node injection power considering the network structure characteristics is:

[0012]

[0013] Where Α represents the directed incidence matrix, Z v is the branch impedance parameter matrix, δ is the voltage phase angle vector, B ii represents the off-diagonal elements in the admittance matrix, V is the node voltage amplitude vector;

[0014] The relative gain matrix is ​​used to quantitatively evaluate the coupling strength between different loops in a multivariable control system, and the control variable Δx i and the controlled variable ΔS i The interaction strength z ij for:

[0015]

[0016] n is the number of distributed generation nodes.

[0017] Control variable Δx i and the controlled variable ΔS i The relative gain matrix R RGA for:

[0018]

[0019] Based on R RGA Quantitative evaluation of the interaction between the electrical quantities of distributed power sources in the distribution network:

[0020]

[0021] In the formula, γ ij The value is 1, indicating that there is a strong interaction between distributed power nodes, and their mutual influence cannot be ignored; γ ij A value of 0 indicates that the interaction between DG nodes is weak;

[0022] Ignore ij The interaction between distributed power nodes is 0, and the distributed power node i is subject to network coupling, that is, the transfer function relationship of the stabilizing branch state quantity is:

[0023]

[0024] In the formula, G i (s) represents the closed-loop transfer function of distributed generation i acting alone, J ij(s) represents the network transfer function between distributed generation i and distributed generation j, J ji (s) represents the network transfer function between distributed generation j and distributed generation i; G j,j+1 (s) represents the mutual coupling between distributed generation j and distributed generation j+1;

[0025] The equivalent decoupling power of the distributed power node is:

[0026] ΔS′ i =M i (s)Δx i

[0027] The decoupled distributed power dynamic model is expressed as:

[0028]

[0029] In the formula, x t is the system state variable, including the frequency and phase angle state in the phase-locked loop, the state of the inner and outer loop PI controllers, the state of the filter inductor and capacitor, the state of the DC voltage, and the state variable after decoupling, y t is an algebraic variable, including the active power P and reactive power Q of the steady-state operating point of the distributed generation, and w is the disturbance.

[0030] Furthermore, the distributed power dynamic model reduction method in step 1 is as follows:

[0031] The dynamic links of distributed power supply include phase-locked loop, inner and outer loop PI controller, filter and DC side dynamics. Intrinsic orthogonal decomposition is used to intercept some modes corresponding to key state variables in the process of constructing reduced-order system, retain the key state variables of concern, and compensate for the influence of non-key state variables, so as to obtain the dynamic model of distributed power supply after reduction:

[0032]

[0033] In the formula, x r =[δ i ,ω i ,e i ] is the key state variable of the system retained after order reduction, δ i is the power angle of the distributed power source; ω i =2πf, f is the frequency of the distributed power supply, e i is the electromotive force of the distributed power source, y r is an algebraic variable;

[0034] Furthermore, the distributed generation energy function that characterizes the multi-dimensional stability characteristics of the distribution network is designed using the first integration method in step 1 as follows:

[0035] Distributed power structure maintenance model:

[0036]

[0037] Where, T i is the stiffness coefficient, M i is the equivalent inertia time constant, D i is the damping coefficient;

[0038] The first integration method is used to construct the distributed power energy function, which is as follows:

[0039]

[0040] In the formula, E Ki and E Pi are the kinetic energy and potential energy of the distributed power source, F i (δ i ,e i ) corresponds to the system after the disturbance, (ω is ,δ is ,e is ) corresponds to the state before the disturbance.

[0041] Furthermore, the step 2 of fitting the critical energy surfaces at different working points based on the hyperplane piecewise affine method comprises the following steps:

[0042] Based on the construction of energy function, the energy value of the dominant unstable equilibrium point is selected as the critical energy, and the homology method is used to set Π(x r )=f(x r ,y r )+w=0, introduce parameter t and auxiliary function Ψ(x r ), construct a homotopy mapping, and solve:

[0043]

[0044] Among them, x r0 is x r The initial value of x is used to define the initial condition x r (0);

[0045] The dominant unstable equilibrium point is obtained as (δ ub ,ω ub ,e ub ), the corresponding potential energy is the critical energy Ecr. If the energy of the distributed generation exceeds the critical energy E cr , that is, it is considered unstable; based on the critical energy solution method, different system dynamic models are calculated at different steady-state operating points to obtain different critical energies E cr,In order to obtain the critical energy surface at different working points, the scatter diagram of the critical energy is firstly solved at different working points, and then the critical energy surface is fitted by hyperplane piecewise affine;

[0046] Taking critical energy as the boundary value, a two-stage solution model is established to characterize the critical energy surface: the inner model determines whether the system is stable based on boundary condition constraints, and the outer model searches for boundary points along the power growth direction until the system loses stability, and the previous boundary point is taken as the critical stability point. The inner model needs to determine the operating conditions for solution, and the limit of the outer model is constrained by the stability margin of the inner model. The final critical energy surface is composed of multiple critical stability hyperplanes. The fitting function of the critical energy surface is expressed as:

[0047] E cr =g(P i ,Q i )

[0048] Among them, (P i ,Q i ) represents the steady-state operating point of distributed power source i, and g represents the polynomial function which generally takes the 2nd or 3rd order.

[0049] Furthermore, the entropy index of the distributed power source characterizing the power flow balance and static voltage safety in step 3 is specifically as follows:

[0050] H0=α1H′ r +α2H' V (8)

[0051] In the formula, α1 and α2 are the weights of power flow balance entropy index and voltage security entropy index respectively, and H′ r , H' V They are the normalized results of power flow balance entropy index and voltage safety entropy index respectively.

[0052] Furthermore, the process of constructing the entropy space of the distributed power source characterizing the safety and stability in step 4 is as follows:

[0053] Design the entropy boundary condition of distributed power supply, which is expressed as:

[0054] f(P i ,Q i )=H0

[0055] According to the boundary conditions, the entropy boundary of the distributed power supply is constructed on the two-dimensional PQ plane. According to the critical energy surface obtained by fitting, the entropy boundary in the two-dimensional PQ plane is projected onto the critical energy surface in the three-dimensional PQE space. The entropy space of the distributed power supply after projection is expressed as:

[0056]

[0057] Furthermore, in step 5, the energy values ​​of the steady-state operating point and the given expected accident are quickly solved and quickly located inside and outside the entropy space, as follows:

[0058] (1) Based on the gradient boosting decision tree (GBDT), the steady-state operating point and the energy value under a given expected accident are quickly mapped, and rapid positioning inside and outside the entropy space is achieved. The input and output of the GBDT network are:

[0059]

[0060] In the formula, X represents the input of the decision tree network, Y is the output value of the decision tree mapping, (P i ,Q i ) represents the steady-state operating point of distributed generation i, w is the given expected disturbance, E i is the energy value at the corresponding steady-state operating point and expected disturbance, which measures the stability of the system. y is a 0-1 variable label used to determine whether the system is safe and stable.

[0061] Initialize the weak learner to be f0(x i ), the loss function is r ti , the strong learner is updated by the loss function:

[0062]

[0063] In the formula, x i is the input variable, y i is the corresponding true label, L(y i ,c) is the loss function, c is the output value of the learner, m is the number of rounds of training, f(x i ) is the decision tree function, f t (x) is the learner function for this round of training, f t-1 (x) is the function of the learner in the previous round of training, f t (x i ) is when the input is x i The learner is , J is the number of iterations, c tj is the learner output of t rounds and j iterations, and I is the compensation coefficient;

[0064] The system safety and stability judgment method is divided into two parts: offline and online. Offline training relies on historical data, which is obtained through distribution network measurement; online perception relies on real-time measurement or prediction data, which requires data of a single time section to obtain the position of the steady-state operating point in the entropy space in real time, and then judge whether the system is in a safe and stable state. For the entropy value, it is not necessary to give the expected disturbance, and the input is power, that is, X = [P i ,Q i ] T;

[0065] (2) Local linearization relationship of entropy value change trend based on least squares method

[0066] The entropy value is locally linearized. First, a number of small fluctuations are given to the mapping model of entropy value and power to obtain the corresponding entropy value fluctuations. The relationship between local entropy value and power is approximately expressed as:

[0067] ΔH=CΔP+DΔQ+θ H (14)

[0068] In the formula, ΔH is the entropy change, ΔP is the active power change, ΔQ is the reactive power change matrix, C and D are local linearization parameters, θ H is the high-order parameter of entropy;

[0069] The overdetermined equations are solved by the least squares method, that is, the local linear matrix under this operating state is obtained, and the established overdetermined equations are:

[0070]

[0071] Among them, Z ij is the matrix parameter of the matrix Z in row i and column j, M is the total number of rows, N is the total number of columns, β j is the corresponding coefficient; when solving, find β, introduce the residual square sum function F(β)=||Zβ-H||, find the maximum value of β differential, and get β=(Z T Z) - 1 Z T H; After obtaining the linear change trend, quickly locate in the entropy space and quickly obtain the safest, most stable and orderly position of the working point.

[0072] Beneficial effects:

[0073] 1. The present invention simplifies the traditional safety and stability assessment method by establishing a distributed power entropy space that characterizes "multi-dimensional stability problems such as system voltage stability and synchronous stability". Through energy function analysis, model decoupling and order reduction are adopted to simplify the analysis of the complex system characteristics of the distribution network. The energy function describes the ability of the power grid to absorb the accumulated energy during the disturbance after being disturbed. The conversion between the kinetic energy and potential energy of the system can well explain the complex motion trajectory of the system. The real critical transient energy of the system is close to the potential energy of the unstable equilibrium point. Under random disturbances, by analogy with the motion equation of the synchronous machine, a mapping relationship between the single kinetic energy of the distributed power source after decoupling from the system and the trajectory of the single machine is established, and its stability is described through the process of energy conversion of the distributed power source nodes.

[0074] 2. The distribution network stability assessment method proposed in the present invention reduces the error of the optimization scheduling model and improves the scheduling efficiency and the economy, safety and orderliness of the system by considering the multi-objective optimization of stability and adopting the mapping relationship positioning method. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 Schematic diagram of the entropy space of the present invention. DETAILED DESCRIPTION

[0076] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0077] The present invention discloses a method for evaluating the operation stability of a distribution network taking into account entropy space, comprising the following steps:

[0078] Step 1: Construct a dynamic model of distributed generation, perform price reduction on it, and then use the first integration method to design a distributed generation energy function that characterizes the multi-dimensional stability characteristics of the distribution network.

[0079] 1) Construct a dynamic model of distributed generation taking into account the coupling mechanism of distribution networks

[0080] First, the topological structure of the distribution network is analyzed. The concept of graph theory is used to model the topological structure of the distribution network as an undirected graph to describe the connection relationship between distributed power nodes. The injected power of distributed power nodes considering the network structure characteristics is:

[0081] P=ΑZ v sin(Α T δ)

[0082] Q=-|Α|Z v cos(Α T δ)-[V]diag[B ii ]V

[0083] Where Α represents the directed incidence matrix, Z v is the branch impedance parameter matrix, B ii represents the off-diagonal elements in the admittance matrix, δ is the voltage phase angle vector, and V is the node voltage amplitude vector.

[0084] Secondly, in view of the complex interactions between distributed power sources, the relative gain matrix is ​​used to quantitatively evaluate the coupling strength between different loops of the multivariable control system. i and the controlled variable ΔS iThe interaction strength between

[0085]

[0086] n is the number of distributed generation nodes.

[0087] Control variable Δx i and the controlled variable ΔS i The relative gain matrix R RGA for:

[0088]

[0089] Based on R RGA Quantitative evaluation of the interaction between the electrical quantities of distributed power sources in the distribution network:

[0090]

[0091] In the formula, γ ij The value is 1, indicating that there is a strong interaction between distributed power nodes, and their mutual influence cannot be ignored; γ ij A value of 0 indicates that the interaction between distributed generation nodes is weak and can be ignored.

[0092] Ignoring the weaker interaction between distributed power nodes, distributed power node i is subject to network coupling, that is, the transfer function relationship of the stabilizing branch state quantity is:

[0093]

[0094] In the formula, G i (s) represents the closed-loop transfer function of distributed generation i acting alone, J ij (s) represents the network transfer function between distributed generation i and distributed generation j, J ji (s) represents the network transfer function between distributed generation j and distributed generation i; G j,j+1 (s) represents the mutual coupling part between distributed generation j and distributed generation j+1.

[0095] The equivalent decoupling power of the distributed power node is: ΔS i '=M i (s)Δx i The decoupled distributed power dynamic model can be expressed as: In the formula, x t is the system state variable, including the frequency and phase angle state in the phase-locked loop, the state of the inner and outer loop PI controllers, the state of the filter inductor and capacitor, the state of the DC voltage, and the state variable after decoupling, y tis an algebraic variable, including the steady-state operating point of the distributed power source, such as active power P, reactive power Q, etc., and w is the disturbance.

[0096] The method of reducing the order of the distributed power dynamic model is as follows:

[0097] The dynamic links of distributed power supply include phase-locked loop, inner and outer loop PI controller, filter and DC side dynamics. Intrinsic orthogonal decomposition is used to intercept some modes corresponding to key state variables in the process of constructing reduced-order system, retain the key state variables of concern, and compensate for the influence of non-key state variables, so as to obtain the dynamic model of distributed power supply after reduction:

[0098]

[0099] In the formula, x r =[δ i ,ω i ,e i ] is the key state variable of the system retained after order reduction, δ i is the power angle of the distributed power source; ω i =2πf, f is the frequency of the distributed power supply, e i is the electromotive force of the distributed power source, y r is an algebraic variable.

[0100] 2) Design of distributed power energy function based on the first integration method: The energy function describes the ability of the power grid to absorb the accumulated energy during the disturbance after being disturbed. The conversion between the kinetic energy and potential energy of the system can well explain the complex motion trajectory of the system. The real critical transient energy of the system is close to the potential energy of the unstable equilibrium point. Under random disturbances, by analogy with the motion equation of the synchronous machine, a mapping relationship between the single kinetic energy of the distributed power source after decoupling from the system and the trajectory of the single machine is established, and its stability is described through the process of energy conversion of the distributed power source node. Distributed power structure retention model:

[0101]

[0102] Where, T i is the stiffness coefficient, M i is the equivalent inertia time constant, D i is the damping coefficient.

[0103] Considering the physical interpretation of the state quantity mapped to the energy space, during the process of system state change, part of the energy in the system is converted into work, while the other part is dissipated as heat in a disordered form. The definition of the energy function reflects the part of energy transfer that is converted into work. According to the corresponding relationship between the key state variables of the system and the system energy, the first integration method is used to construct the distributed power energy function, which is in the following form:

[0104]

[0105] In the formula, E Ki and E Pi are the kinetic energy and potential energy of the distributed power source, F i (δ i ,e i ) corresponds to the system after the disturbance, (ω is ,δ is ,e is ) corresponds to the state before the disturbance.

[0106] Step 2: Based on the constructed distributed power energy function, the energy value of the dominant unstable equilibrium point is selected as the critical energy, and the critical energy surface at different working points is fitted based on the hyperplane piecewise affine method.

[0107] On the basis of the construction of the energy function, the energy value of the dominant unstable equilibrium point is selected as the critical energy. Considering that the equilibrium point calculation method based on Newton's method has high requirements on the selection of initial values ​​for iteration, the homology method is used to overcome the sensitivity of Newton's iteration method to the initial values ​​(δ0, ω0, e0).

[0108] Let Π(x r )=f(x r ,y r )+w=0, introduce parameter t and auxiliary function Ψ(x r ), construct a homotopy mapping, and solve:

[0109]

[0110] Among them, x r0 is x r The initial value of x is used to define the initial condition x r (0).

[0111] The dominant unstable equilibrium point is obtained as (δ ub ,ω ub ,e ub ), the corresponding potential energy is the critical energy Ecr. If the energy of the distributed generation exceeds the critical energy E cr Based on the critical energy solution method, different critical energies E can be calculated at different steady-state operating points corresponding to different system dynamic models. cr In order to obtain the critical energy surface at different working points, the scatter plot of the critical energy is first obtained by solving at different working points, and then the critical energy surface is fitted by hyperplane piecewise affine.

[0112] With critical energy as the boundary value, a two-stage solution model is established to characterize the critical energy surface. The inner model determines whether the system is stable based on boundary condition constraints, and the outer model searches for boundary points along the power growth direction until the system loses stability, and the previous boundary point is used as the critical stability point. The inner model needs to determine the operating conditions for solution, and the limit of the outer model is constrained by the stability margin of the inner model. The final critical energy surface is composed of multiple critical stability hyperplanes.

[0113] The fitting function of the critical energy surface can be expressed as: E cr =g(P i ,Q i ), (P i ,Q i ) represents the steady-state operating point of distributed power source i, and g represents the polynomial function, which is generally of 2-3 order.

[0114] Step 3: Normalize the power flow balance entropy index and the voltage safety entropy index to obtain the system entropy index that characterizes the power flow balance and static voltage safety of the distribution network. Based on the branch and bound method, the system entropy index is dispersed to the distributed power level to obtain the distributed power entropy index that characterizes the power flow balance and static voltage safety.

[0115] The entropy of distributed power sources that characterizes the safety and order of the distribution network system is constructed as follows:

[0116] In the field of distribution network system research, the information entropy is combined with the static security concepts related to the distribution network, and a unique entropy concept for the distribution network system is proposed to characterize its static orderliness and security.

[0117] The mathematical expression of the power flow equilibrium entropy of the distribution network is: In the formula, is the number of branches in the system whose load rate is within the safe range (A, B) among all branches. i The ratio of the total number of branches is calculated as follows: When the load rates of all branches are in the same interval, the power flow equilibrium entropy is the smallest; when the load rates of all branches are in different intervals, the power flow equilibrium entropy is the largest. From the perspective of load rate, the power flow equilibrium entropy quantitatively describes the balance of system power flow distribution among various lines under a certain power flow section.

[0118] Static voltage safety usually considers the distribution and level of voltage during system operation. Distributed power supplies including power electronic equipment have low voltage withstand capability and there is a risk of voltage exceeding the limit under faults or disturbances. The voltage safety entropy is defined as: The number of nodes in the distribution network system that deviate from the rated operating voltage value and are in the safe range (C, D) The ratio of the total number of nodes is calculated as: The larger the voltage safety entropy value, the more uneven the node voltage distribution is, and a few nodes may have larger over-limit values; the smaller the voltage safety entropy value, the smaller the difference between node voltages and the more uniform the voltage distribution is. However, when the node voltage deviates from the rated operating voltage by a large margin, the voltage collapse of the entire distribution network is likely to occur.

[0119] The power flow balance entropy index and voltage safety entropy index are normalized to obtain the system entropy index that characterizes the power flow balance and static voltage safety of the distribution network. Based on the branch and bound method, the system-level index is dispersed to the distributed power level to obtain the distributed power entropy index that characterizes the power flow balance and static voltage safety:

[0120] H0=α1H r '+α2H' V

[0121] In the formula, α1, α2 and H r '、H' V They are the weights and normalized results of power flow balance entropy index and voltage safety entropy index respectively.

[0122] Step 4: Design the entropy boundary conditions of the distributed power supply. According to the boundary conditions, construct the entropy boundary of the distributed power supply on the two-dimensional PQ plane. According to the critical energy surface fitted in step 2, construct the entropy space of the distributed power supply that characterizes the safety and stability based on the boundary projection method.

[0123] The method for constructing the entropy space of distributed power supply based on boundary projection is as follows:

[0124] Based on the definition of distributed power entropy, the boundary condition of distributed power entropy is designed, which can be expressed as: f(P i ,Q i )=H0. According to the boundary condition, the entropy boundary of the distributed power supply is constructed on the two-dimensional PQ plane. According to the critical energy surface obtained by fitting, the entropy space of the distributed power supply that characterizes the safety and stability is constructed, and the entropy boundary in the two-dimensional PQ plane is projected onto the critical energy surface in the three-dimensional PQE space. The projected entropy space of the distributed power supply can be expressed as:

[0125] E cr =g(P i ,Q i )

[0126] f(P i ,Q i )-H0≤0

[0127] The physical meaning of the distributed power entropy space is: the system's steady-state operating point simultaneously satisfies the distributed power entropy constraints of safety and orderliness and the energy constraints of stability. As long as the system's steady-state operating point and its corresponding energy under the expected disturbance are within the space, the system must satisfy the defined safety, orderliness and stability constraints.

[0128] Step 5: Quickly solve the energy value of the steady-state operating point and the given expected accident, quickly locate it inside and outside the entropy space, and evaluate the stability of the distribution network operation based on the positioning results and the distributed power entropy space obtained in step 4.

[0129] 1) Rapid positioning in entropy space

[0130] Considering that it is difficult to quickly solve the energy function value under a given expected accident, in order to improve the speed of safety and stability judgment, the energy value under the steady-state operating point and the given expected accident is quickly mapped based on the gradient boosting decision tree (GBDT), and rapid positioning inside and outside the entropy space is achieved. The input and output of the GBDT network are:

[0131]

[0132] In the formula, X represents the input of the decision tree network, Y is the output value of the decision tree mapping, (P i ,Q i ) represents the steady-state operating point of distributed generation i, w is the given expected disturbance, E i is the energy value under the corresponding steady-state operating point and expected disturbance, which measures the stability of i. y is a 0-1 variable label, which is used to judge whether the system is safe and stable.

[0133] Initialize the weak learner to be f0(x i ), whose loss function is r ti , the strong learner is updated by the loss function:

[0134]

[0135] In the formula, x i is the input variable, y i is the corresponding true label, L(y i ,c) is the loss function, c is the output value of the learner, m is the number of rounds of training, f(x i ) is the decision tree function, f t (x) is the learner function for this round of training, f t-1 (x) is the function of the learner in the previous round of training, f t (x i ) is when the input is x iThe learner is , J is the number of iterations, c tj is the learner output of t rounds and j iterations, and I is the compensation coefficient.

[0136] The method for judging the safety and stability of the system is divided into two parts: offline and online. Offline training relies on historical data, which can be obtained through distribution network measurement. Online perception relies on real-time measurement or prediction data. It only needs data from a single time section to obtain the position of the steady-state operating point in the entropy space in real time, and then judge whether the system is in a safe and stable state. For the entropy value, there is no need to give the expected disturbance, and the input is power, that is, X = [P i ,Q i ] T .

[0137] 2) Local linearization relationship of entropy value change trend based on least squares method

[0138] Considering that the entropy value needs to be as small as possible, the changes near the same working point need to find the direction of entropy minimization, and the entropy value is locally linearized. First, the mapping model of entropy value and power is given several groups of small fluctuations to obtain the corresponding entropy value fluctuations. The relationship between local entropy value and power can be approximately expressed as:

[0139] ΔH=CΔP+DΔQ+θ H

[0140] In the formula, ΔH is the entropy change, ΔP is the active power change, ΔQ is the reactive power change matrix, C and D are local linearization parameters, θ H is a high-order parameter of entropy.

[0141] By solving the overdetermined equations by the least squares method, the local linear matrix under this operating state can be obtained. The overdetermined equations are established as:

[0142]

[0143] Among them, Z ij is the matrix parameter of the matrix Z in row i and column j, M is the total number of rows, N is the total number of columns, β j is the corresponding coefficient. When solving this equation, in order to find a suitable β to make the equation as valid as possible, the residual sum of squares function F(β)=||Zβ-H|| is introduced, and the maximum value of β is obtained by differential. After sorting, we can get β=(Z T Z) -1 Z T H. After obtaining the linear change trend, it is possible to quickly locate the working point in the entropy space and quickly obtain the safest, most stable and orderly position.

[0144] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0145] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. A distribution network operation stability evaluation method taking into account entropy space, characterized in that: The following steps are involved: Step 1: Construct a dynamic model of distributed generation, perform price reduction on it, and then use the first integration method to design a distributed generation energy function that characterizes the multi-dimensional stability characteristics of the distribution network; Step 2: Based on the constructed distributed power energy function, the energy value of the dominant unstable equilibrium point is selected as the critical energy, and the critical energy surface at different working points is fitted based on the hyperplane piecewise affine method; Step 3: Normalize the power flow balance entropy index and the voltage safety entropy index to obtain the system entropy index that characterizes the power flow balance and static voltage safety of the distribution network. Based on the branch and bound method, the system entropy index is dispersed to the distributed power level to obtain the distributed power entropy index that characterizes the power flow balance and static voltage safety. Step 4: Design the entropy boundary conditions of the distributed power supply. According to the boundary conditions, construct the entropy boundary of the distributed power supply on the two-dimensional PQ plane. According to the critical energy surface fitted in step 2, construct the entropy space of the distributed power supply that characterizes the safety and stability based on the boundary projection method. Step 5: Quickly solve the energy value of the steady-state operating point and the given expected accident, quickly locate it inside and outside the entropy space, and evaluate the stability of the distribution network operation based on the positioning results and the distributed power entropy space obtained in step 4.

2. A distribution network operation stability evaluation method taking into account entropy space according to claim 1, characterized in that: The dynamic model of distributed power supply in step 1 is as follows: The distribution network topology is analyzed and modeled as an undirected graph to describe the connection relationship between distributed power nodes. The distributed power node injection power considering the network structure characteristics is: Where Α represents the directed incidence matrix, Z v is the branch impedance parameter matrix, δ is the voltage phase angle vector, B ii represents the off-diagonal elements in the admittance matrix, V is the node voltage amplitude vector; The relative gain matrix is ​​used to quantitatively evaluate the coupling strength between different loops in a multivariable control system, and the control variable Δx i and the controlled variable ΔS i The interaction strength z ij for: n is the number of distributed generation nodes. Control variable Δx i and the controlled variable ΔS i The relative gain matrix R RGA for: Based on R RGA Quantitative evaluation of the interaction between the electrical quantities of distributed power sources in the distribution network: In the formula, γ ij A value of 1 indicates that there is a strong interaction between distributed generation nodes, and their mutual influence cannot be ignored; γ ij A value of 0 indicates that the interaction between DG nodes is weak; Ignore ij The interaction between distributed power nodes is 0, and the distributed power node i is subject to network coupling, that is, the transfer function relationship of the stabilizing branch state quantity is: In the formula, G i (s) represents the closed-loop transfer function of distributed generation i acting alone, J ij (s) represents the network transfer function between distributed generation i and distributed generation j, J ji (s) represents the network transfer function between distributed generation j and distributed generation i; G j,j+1 (s) represents the mutual coupling between distributed generation j and distributed generation j+1; The equivalent decoupling power of the distributed power node is: ΔS i '=M i (s)Δx i The decoupled distributed power dynamic model is expressed as: In the formula, x t is the system state variable, including the frequency and phase angle state in the phase-locked loop, the state of the inner and outer loop PI controllers, the state of the filter inductor and capacitor, the state of the DC voltage, and the state variable after decoupling, y t is an algebraic variable, including the active power P and reactive power Q of the steady-state operating point of the distributed generation, and w is the disturbance.

3. A distribution network operation stability evaluation method taking into account entropy space according to claim 2, characterized in that: The method for reducing the order of the distributed power dynamic model in step 1 is as follows: The dynamic links of distributed power supply include phase-locked loop, inner and outer loop PI controller, filter and DC side dynamics. Intrinsic orthogonal decomposition is used to intercept some modes corresponding to key state variables in the process of constructing reduced-order system, retain the key state variables of concern, and compensate for the influence of non-key state variables, so as to obtain the dynamic model of distributed power supply after reduction: In the formula, x r =[δ i ,ω i ,e i ] is the key state variable of the system retained after order reduction, δ i is the power angle of the distributed power source; ω i =2πf, f is the frequency of the distributed power supply, e i is the electromotive force of the distributed power source, y r is an algebraic variable.

4. A method for evaluating the operation stability of a distribution network taking into account entropy space according to claim 3, characterized in that: The distributed generation energy function that characterizes the multi-dimensional stability characteristics of the distribution network is designed using the first integration method in step 1 as follows: Distributed power structure preservation model: Where, T i is the stiffness coefficient, M i is the equivalent inertia time constant, D i is the damping coefficient; The first integration method is used to construct the distributed power energy function, which is as follows: In the formula, E Ki and E Pi are the kinetic energy and potential energy of the distributed power source, F i (δ i ,e i ) corresponds to the system after the disturbance, (ω is ,δ is ,e is ) corresponds to the state before the disturbance.

5. A distribution network operation stability evaluation method taking into account entropy space according to claim 4, characterized in that: The step 2 of fitting the critical energy surfaces at different working points based on the hyperplane piecewise affine method comprises the following steps: Based on the construction of energy function, the energy value of the dominant unstable equilibrium point is selected as the critical energy, and the homology method is used to set Π(x r )=f(x r ,y r )+w=0, introduce parameter t and auxiliary function Ψ(x r ), construct a homotopy mapping, and solve: Among them, x r0 is x r The initial value of x is used to define the initial condition x r (0); The dominant unstable equilibrium point is obtained as (δ ub ,ω ub ,e ub ), the corresponding potential energy is the critical energy E cr , if the energy of the distributed generation exceeds the critical energy E cr , that is, it is considered unstable; based on the critical energy solution method, different system dynamic models are calculated at different steady-state operating points to obtain different critical energies E cr ,In order to obtain the critical energy surface at different working points, the scatter diagram of the critical energy is firstly solved at different working points, and then the critical energy surface is fitted by hyperplane piecewise affine; Taking critical energy as the boundary value, a two-stage solution model is established to characterize the critical energy surface: the inner model determines whether the system is stable based on boundary condition constraints, and the outer model searches for boundary points along the power growth direction until the system loses stability, and the previous boundary point is taken as the critical stability point. The inner model needs to determine the operating conditions for solution, and the limit of the outer model is constrained by the stability margin of the inner model. The final critical energy surface is composed of multiple critical stability hyperplanes. The fitting function of the critical energy surface is expressed as: E cr =g(P i ,Q i ) Among them, (P i ,Q i ) represents the steady-state operating point of distributed power source i, and g represents the polynomial function which generally takes the 2nd to 3rd order.

6. A method for evaluating the operation stability of a distribution network taking into account entropy space according to claim 5, characterized in that: The specific entropy index of the distributed power supply characterizing the power flow balance and static voltage safety in step 3 is as follows: H0=α1H r '+α2H' V (8) In the formula, α1 and α2 are the weights of power flow balance entropy index and voltage security entropy index respectively, H r '、H' V They are the normalized results of power flow balance entropy index and voltage safety entropy index respectively.

7. A method for evaluating the operation stability of a distribution network taking into account entropy space according to claim 6, characterized in that: The process of constructing the entropy space of distributed power sources that characterizes safety and stability in step 4 is as follows: Design the entropy boundary condition of distributed power supply, which is expressed as: f(P i ,Q i )=H0 According to the boundary conditions, the entropy boundary of the distributed power supply is constructed on the two-dimensional PQ plane. According to the critical energy surface obtained by fitting, the entropy boundary in the two-dimensional PQ plane is projected onto the critical energy surface in the three-dimensional PQE space. The entropy space of the distributed power supply after projection is expressed as:

8. A method for evaluating the operation stability of a distribution network taking into account entropy space according to claim 1, characterized in that: In step 5, the energy values ​​of the steady-state operating point and the given anticipated accident are quickly solved and quickly located inside and outside the entropy space, as follows: (1) Based on the gradient boosting decision tree (GBDT), the steady-state operating point and the energy value under a given expected accident are quickly mapped, and rapid positioning inside and outside the entropy space is achieved. The input and output of the GBDT network are: In the formula, X represents the input of the decision tree network, Y is the output value of the decision tree mapping, (P i ,Q i ) represents the steady-state operating point of distributed generation i, w is the given expected disturbance, E i is the energy value at the corresponding steady-state operating point and expected disturbance, which measures the stability of the system. y is a 0-1 variable label used to determine whether the system is safe and stable. Initialize the weak learner to be f0(x i ), the loss function is r ti , the strong learner is updated by the loss function: In the formula, x i is the input variable, y i is the corresponding true label, L(y i ,c) is the loss function, c is the output value of the learner, m is the number of rounds of training, f(x i ) is the decision tree function, f t (x) is the learner function for this round of training, f t-1 (x) is the function of the learner in the previous round of training, f t (x i ) is when the input is x i The learner is , J is the number of iterations, c tj is the learner output of t rounds and j iterations, and I is the compensation coefficient; The system safety and stability judgment method is divided into two parts: offline and online. Offline training relies on historical data, which is obtained through distribution network measurement; online perception relies on real-time measurement or prediction data, which requires data of a single time section to obtain the position of the steady-state operating point in the entropy space in real time, and then judge whether the system is in a safe and stable state. For the entropy value, it is not necessary to give the expected disturbance, and the input is power, that is, X = [P i ,Q i ] T ; (2) Local linearization relationship of entropy value change trend based on least squares method The entropy value is locally linearized. First, a number of small fluctuations are given to the mapping model of entropy value and power to obtain the corresponding entropy value fluctuations. The relationship between local entropy value and power is approximately expressed as: ΔH=CΔP+DΔQ+θ H (14) In the formula, ΔH is the entropy change, ΔP is the active power change, ΔQ is the reactive power change matrix, C and D are local linearization parameters, θ H is the high-order parameter of entropy; The overdetermined equations are solved by the least squares method, that is, the local linear matrix under this operating state is obtained, and the established overdetermined equations are: Among them, Z ij is the matrix parameter of the matrix Z in row i and column j, M is the total number of rows, N is the total number of columns, β j is the corresponding coefficient; when solving, find β, introduce the residual square sum function F(β)=||Zβ-H||, find the maximum value of β differential, and get β=(Z T Z) -1 Z T H; After obtaining the linear change trend, quickly locate in the entropy space and quickly obtain the safest, most stable and orderly position of the working point.

Images