Wind power cluster access power grid-oriented double-layer reactive power optimization method considering reactive power potential of wind power plant

Through the double-layer reactive power optimization method and the improved whale optimization algorithm, the reactive power problem when the wind power cluster is connected to the power grid is solved, the grid operation efficiency is improved and the voltage is maintained stably, and the reactive power potential of the wind turbine is fully utilized.

CN118983889APending Publication Date: 2024-11-19STATE GRID GANSU ELECTRIC POWER RESEARCH INSTITUTE
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411042326.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-31
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

The existing technology is difficult to effectively optimize the reactive power problem when wind power clusters are connected to the power grid, resulting in reduced grid operation efficiency and voltage fluctuations, and failing to fully utilize the reactive power support potential of wind power units.

Method used

Using a double-layer reactive power optimization method for wind power clusters, a top-level optimization model with the smallest active power loss and voltage deviation is established by finely analyzing the reactive power loss and margin of the wind power cluster. The improved whale optimization algorithm is used to solve it, combined with the wind power cluster. The lower-level model with the highest utilization rate of reactive margins optimizes the reactive power output plan within the wind power cluster.

Benefits of technology

It effectively reduces the grid loss and voltage deviation of the power system, maximizes the utilization of wind power reactive potential, and improves the operating efficiency and voltage stability of the power grid.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118983889A_ABST
    Figure CN118983889A_ABST
Patent Text Reader

Abstract

The invention discloses a wind power cluster access power grid-oriented double-layer reactive power optimization method considering the reactive potential of a wind power plant. The method comprises the steps of performing fine analysis on reactive power in the wind power cluster plant to obtain reactive power loss of a power transmission line and a transformer; carrying out refined analysis by taking a wind turbine generator set and a collection line in the wind power cluster field as centers to obtain a reactive power margin of the wind power cluster; establishing a reactive power optimization upper-layer model with minimum active power loss and voltage deviation as targets; solving the upper-layer model by adopting an improved whale optimization algorithm to obtain reactive power output values of each wind power cluster and multiple reactive power sources in an optimization time period; establishing a reactive power optimization lower-layer model with the highest reactive power margin utilization rate as a target, and inputting a reactive power output value; and solving the lower-layer model by adopting an improved whale optimization algorithm to obtain a reactive power output scheme in the wind power cluster. The problem that the operation efficiency of an existing power grid is reduced is solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of power grid planning, and in particular relates to a double-layer reactive power optimization method for accessing a wind power cluster to a power grid and taking into account the reactive power potential of a wind farm. Background Art

[0002] With the gradual escalation of the energy crisis and increasingly severe environmental problems, sustainable development has become the primary goal of future power grid construction. Energy low-carbon transformation is an inevitable trend in global energy development, and new energy is an important prerequisite for leading the green and low-carbon development and transformation of the power system, and will occupy a dominant position in the future power system. As the proportion of new energy in the power grid increases, the volatility and intermittency of new energy have a greater impact on the power grid, causing frequent fluctuations in the power grid voltage, bringing new challenges to reactive power support and voltage safety.

[0003] To this end, domestic and foreign scholars have conducted a lot of research on reactive power optimization of power systems and reactive power potential analysis of wind power clusters. Reactive power optimization of power systems is a necessary means to improve the voltage level of power grids and improve the economic efficiency of system operation. The optimal value of each variable operation is obtained through algorithm optimization, and the optimal reactive power flow distribution is realized, so as to achieve economic, safe and efficient operation of the power grid. The core idea of ​​reactive power potential analysis of wind power clusters is to fully tap the reactive support potential of wind turbines. In essence, it is to simply superimpose the reactive power limits of each unit to obtain the reactive output capacity of the wind farm. However, in engineering applications, uncertain factors such as active output of wind farms, voltage fluctuations at the grid connection point, and network flow constraints should also be considered to obtain a reliable reactive power support range of wind farms. At present, there are few studies on combining reactive power optimization and reactive power potential analysis of wind power clusters to improve the voltage level of power grids and improve the economic efficiency of system operation. Summary of the invention

[0004] The purpose of the present invention is to provide a double-layer reactive power optimization method for wind power cluster access to a power grid taking into account the reactive power potential of wind farms, thereby solving the problem of reduced operating efficiency of existing power grids.

[0005] The technical solution adopted by the present invention is: a double-layer reactive power optimization method considering the reactive power potential of wind farms for wind power cluster access to the power grid, comprising the following steps:

[0006] Step 1: Perform a detailed analysis of the reactive power in the wind farm cluster to obtain the reactive loss of the transmission line and transformer;

[0007] Step 2: Based on the reactive power loss obtained in step 1, a refined analysis is performed with the wind turbines and the collection lines in the wind power cluster as the center to obtain the reactive power margin of the wind power cluster;

[0008] Step 3: Establish a reactive power optimization upper layer model with the goal of minimizing active power loss and voltage deviation;

[0009] Step 4: Use the improved whale optimization algorithm to solve the upper model and obtain the reactive output values ​​of each wind power cluster and multiple reactive sources within the optimization time period;

[0010] Step 5, establishing a reactive power optimization lower layer model with the goal of maximizing the reactive power margin utilization rate obtained in step 2, and inputting the reactive power output value obtained in step 4;

[0011] Step 6: Use the improved whale optimization algorithm to solve the lower model and obtain the reactive power output plan within the wind power cluster.

[0012] The present invention is also characterized in that:

[0013] Step 1 is as follows: Assuming that the system is in normal operation, the voltage of each node is maintained at a per-unit value of 1, and assuming that the power factor of the series branch port is 1, the mathematical relationship is as follows:

[0014]

[0015] ΔQ Z1 ≈P 2 X 1

[0016] In the formula, X 1 Represents the reactance in the line, ΔQ Z1 is the reactive power consumed by the line impedance, P is the active power in the line, and U p is the voltage at the beginning of the transmission line, I is the current flowing through the line, is the power factor;

[0017] The reactive power generated by the distributed capacitance in the transmission line is:

[0018]

[0019] In the formula, B 1 Indicates the susceptance in the line, Q c is the reactive power generated by the line distributed capacitance, U p is the voltage at the beginning of the transmission line, U q is the voltage at the end of the transmission line;

[0020] Combining the above formula, the reactive power loss of the line is:

[0021]

[0022] In the formula, X 1 Represents the reactance in the line, B 1 It represents the susceptance in the line, P is the active power in the line, Q c is the reactive power generated by the distributed capacitance of the line, ΔQ Z1 The reactive power consumed by the line impedance;

[0023] The reactive power loss of the line impedance element is:

[0024]

[0025] In the formula, is the reactive power consumed by the transformer equivalent line impedance, k is the non-standard transformation ratio of the transformer under the standard value, X t It represents the reactance in the equivalent line impedance, and P is the active power in the line;

[0026] The sum of reactive power absorbed by the ground branch is:

[0027]

[0028] In the formula, X t represents the reactance in the equivalent line impedance, k is the non-standard transformation ratio of the transformer under the standard value, U p is the voltage at the beginning of the transmission line, U q is the voltage at the end of the transmission line;

[0029] The reactive loss of ground excitation admittance is:

[0030]

[0031] In the formula, B t Indicates the susceptance in the equivalent line impedance, U q is the voltage at the end of the transmission line;

[0032] The reactive power loss of the transformer is:

[0033]

[0034] Where P is the active power in the line, k is the non-standard transformation ratio of the transformer under the standard value, and B t represents the susceptance in the equivalent line impedance, X t Represents the reactance in the equivalent line impedance.

[0035] Step 2 is specifically as follows: define the line between the jth wind turbine and the j-1th wind turbine in the i-th feeder in the wind power cluster as ij, and define the reactive partition ij centered on the wind turbine, which includes the wind turbine ij and its corresponding box transformer, and half the length of the transmission line i,j and i,j-1;

[0036] The reactive power loss in the upstream transmission line is:

[0037]

[0038] In the formula, X l,i,j , B l,i,jrepresents the reactance and susceptance in the upstream line i,j, P i,b is the active power of the bth wind turbine in the i-th feeder, b is the wind turbine number, and n is the total number of wind turbines in a single feeder;

[0039] The reactive power loss in the downstream transmission line is:

[0040]

[0041] Where X l,i,j+1 , B l,i,j+1 represents the reactance and susceptance in the downstream line i,j+1;

[0042] The reactive power loss of the box-type transformer connected to a single fan is:

[0043]

[0044] Where X t.i,j , B t.i,j represents the reactance and susceptance of the box-type transformer connected to wind turbines i and j, k t.i,j is the box-type transformer ratio, P i,j is the active power of the jth wind turbine in the i-th feeder;

[0045] Therefore, the reactive power loss of each wind turbine area in the wind power cluster is:

[0046]

[0047] The topology of the wind power cluster collection line includes the step-up transformer at the collection station, half of the length of the first feeder section of each line, and the total reactive power of the wind cluster collection and transmission line;

[0048] The reactive power loss of line i,1 is:

[0049]

[0050] Where, X l,i,1 , B l,i,1 is the reactance and susceptance of the first section of the line;

[0051] The reactive power loss of the step-up transformer is:

[0052]

[0053] Where, X T , B T is the reactance and susceptance of the step-up transformer in the collection line, k T is the ratio of the step-up transformer, a is the feeder number, m is the fan number, P a,b is the active power of the bth wind turbine in the ath feeder;

[0054] The reactive power loss of the collection line is:

[0055]

[0056] Where, X L , B L is the reactance and susceptance of the collection line;

[0057] The total reactive power loss of the collection line is:

[0058]

[0059] In the formula, Q l.a,1 is the reactive power loss of the first feeder section of each feeder;

[0060] The total reactive power loss of the wind farm is the sum of the reactive power of the corresponding area of ​​each wind turbine and the reactive power of the collection line. The specific calculation formula is:

[0061]

[0062] In the formula, Q l.a,b is the reactive power loss of each section of each feeder, Q t.a,b is the reactive power loss of the wind box transformer in each section of the feeder;

[0063] When the wind farm is operating normally, the total reactive power output of the wind power cluster and the calculated total reactive power loss Q in the wind farm are WT The difference is the actual reactive capacity of the wind farm:

[0064]

[0065] Step 3 is as follows: Establish an hourly optimization model with the minimum system active network loss and voltage deviation as the objective function:

[0066]

[0067] ΔU imax =U imax -U imin

[0068] Where U i is the voltage value of node i in the system, N is the set of system nodes, U* i is the voltage reference value of node i in the system, ΔU imax The maximum voltage deviation allowed at node i in the system, ΔU is the voltage deviation, U imax is the maximum voltage value of node i, U imin is the minimum voltage value of node i;

[0069]

[0070] Where P loss is the total network loss of the system, U i is the voltage value of node i in the system, U j is the voltage value of node j in the system, G ij is the conductance between nodes i and j, θ ij is the voltage phase difference between nodes i and j, N c is the total number of reactive power compensation devices in the system;

[0071] The power balance constraints of the upper optimization model are as follows:

[0072]

[0073] Where P Gi is the active power of node i in the system, P li is the active load of node i in the system, Q Gi is the reactive power of node i in the system, Q li is the reactive load of node i in the system, Q Ci is the reactive power provided by the reactive compensation device at node i in the system, G ij and B ij are the real and imaginary parts of the node admittance matrix respectively;

[0074] The grid safety operation constraints are as follows:

[0075] U gimin ≤U gi ≤U gimax i∈N g

[0076] P gi,min ≤P gi ≤P gi,max i∈N g

[0077] Q gi,min ≤Q gi ≤Q gi,max i∈N g

[0078] Q ci,min ≤Q ci ≤Q ci,max i∈N c

[0079] T kmin ≤T k ≤T kmax i∈N t

[0080] Where U gimax , U giminis the terminal voltage U of the i-th generator gi The upper and lower limits of P gimax , P gimin is the active output P of the i-th generator gi The upper and lower limits of Q gimax , Q gimin is the reactive power output Q of the i-th generator gi The upper and lower limits of Q cimax , Q cimin is the switching capacity Q of the i-th reactive power compensation device ci The upper and lower limits of T kimax 、T kimin is the transformation ratio T of the i-th voltage regulating transformer k The upper and lower limits of N g is the total number of generators in the system, N c is the total number of reactive power compensation devices in the system, N t is the total number of voltage regulating transformers in the system;

[0081] OLTC-related constraints:

[0082] u i,t =k ij,t u j,t

[0083] k ij,t =k ij,0 +K ij,t Δk ij

[0084]

[0085] In the formula, k ij,t is the transformation ratio of OLTC at time t, k ij,0 is the transformation ratio of OLTC at the initial moment, Δk ij, is the change in ratio, K ij,t is the gear position at time t, The highest gear, is the maximum number of tap change steps in OLTC, u i,t is the voltage of node i at time t, u j,t is the voltage of node j at time t, Z is the number of adjustable gears of OLTC, and T is the time;

[0086] CB related constraints:

[0087]

[0088] In the formula, is the reactive power output of each CB, is the number of CBs of node i at time t, is the reactive power output of the CB of node i at time t, is the maximum number of CB switching changes, It is the maximum number of switching groups of CB.

[0089] Step 4 is as follows: input the original data of the upper model, including the network component parameters, the active and reactive output values ​​of conventional power sources, conventional loads and their power factors, use the improved whale optimization algorithm to solve the upper model, find a set of Pareto optimal solutions, and select the optimal compromise solution, that is, the reactive output values ​​of each wind power cluster and multiple reactive sources within the optimization time period; the improved whale optimization algorithm includes:

[0090] 1) Searching for prey

[0091] The position of each whale in the search space is a solution. Let the position of the whale group currently in the best hunting position be the target prey or close to the best target prey:

[0092]

[0093] Where t is the current iteration number, X is the current whale position vector, and X rand is the position vector randomly selected by the whale population, A and C are matrix coefficients, and D is the distance between the whale population and the prey;

[0094]

[0095] In the formula, r is a random number between [0,1], t m is the upper limit of the number of algorithm iterations, and a is a value that gradually decreases between [2,0] during the algorithm iteration process;

[0096] 2) Surround the prey

[0097] The other whales in the group all surround the optimal position to update their own positions:

[0098]

[0099] Where X* is the optimal position vector of the global group;

[0100] 3) Spiral bubble attack

[0101] Humpback whales will shrink and encircle the prey in a spiral line, with a probability of 0.5 to perform the surrounding prey behavior and spiral renewal behavior. The mathematical model is as follows:

[0102]

[0103] b is a constant used to define the shape of the spiral; l is a random number between [-1,1]; P is a random number between [0,1], and D' is the latest distance between the whale group and the prey;

[0104] Each iteration passes through the individual extreme value p best and the global extremum g best Update particles and optimize the positions of particles with poor fitness during the iteration process. The formulas for speed and position update are as follows:

[0105]

[0106] ω is the inertia weight factor, is the velocity of particle i in the jth dimension at the t+1th iteration, is the position of particle i in the jth dimension at the t+1th iteration, is the position of particle i in the jth dimension at the tth iteration, is the position of the jth dimension of whale individual i at the tth iteration, c 1 、c 2 is the learning factor, r 1 、r 2 is a random number distributed in [0,1].

[0107] Step 5 is specifically as follows: establishing a lower-level model, whose objective function is to maximize the reactive margin utilization rate of the wind power cluster, and the input of the model includes the original parameters inside the wind power cluster, the active output of wind power, and the reactive output value obtained by solving the upper-level model;

[0108] The lower layer establishes a 15-minute rolling optimization model with the highest reactive margin utilization rate of the wind power cluster as the objective function. The active output value of a group of wind turbines is input every 15 minutes, and the reactive output Q of the upper layer at 24 moments is calculated. set As the initial parameters for lower layer optimization:

[0109]

[0110] In the formula, Q WFi represents the reactive capacity of the i-th wind turbine, Q WF Represents the total reactive capacity of the wind farm, Q i represents the actual reactive power output value of the i-th wind turbine, Q set Optimize the reactive power output of wind power transmitted to the upper layer;

[0111] Establish relevant constraints, including:

[0112] Power flow constraints:

[0113]

[0114] Where P WTi , Q WTi are the active output and reactive output of the i-th wind turbine generator set, Q WTi_max , Q WTi_minis the upper and lower limits of wind turbine output, P WTi_max , P WTi_min is the upper and lower limits of the active output of the wind turbine, U i_max , U i_min are the access point voltage U of the i-th wind turbine group respectively. i Upper and lower limits.

[0115] Step 6 is as follows: Use the improved whale optimization algorithm to solve the lower model, obtain the reactive power output plan within the wind power cluster, and use the optimized results to update the upper and lower limits of the reactive power output of the upper wind power cluster as the boundaries of a new round of optimization for the next iterative calculation.

[0116] The beneficial effects of the present invention are as follows: the present invention adopts a double-layer reactive power optimization method for wind farms connected to the power grid, which considers the reactive power potential of wind farms, and analyzes the reactive power potential of wind farms in a refined manner, and proposes an upper-layer multi-objective reactive power optimization model that minimizes voltage deviation and network loss. The reactive power output value of the wind farm is obtained by optimizing the number of capacitor groups switched on and off, the output of static reactive power compensation devices, and the ratio of voltage-regulating transformers; on this basis, the lower layer establishes an optimization model with the highest reactive power margin utilization rate of the wind farm as the goal, and takes the reactive power output of each wind turbine unit as the quantity to be optimized, thereby effectively reducing the active power loss. The present invention effectively reduces the network loss and voltage deviation of the power system when the reactive power potential of wind power is maximized, thereby solving the problem of reduced grid operation efficiency caused by increased network loss and voltage over-limit caused by the access of wind farms to the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0117] Figure 1 It is a flow chart of a double-layer reactive power optimization method for wind power cluster access to a power grid taking into account the reactive power potential of wind farms according to the present invention;

[0118] Figure 2 It is a modified IEEE39 node system network topology diagram in the double-layer reactive power optimization method for wind power cluster access to the power grid taking into account the reactive power potential of the wind farm of the present invention;

[0119] Figure 3 It is a structural topology diagram of a wind power cluster in a lower optimization system in a double-layer reactive power optimization method for wind power cluster access to a power grid taking into account reactive power potential of wind farms according to the present invention;

[0120] Figure 4 It is a typical topological structure diagram inside a wind power cluster in the double-layer reactive power optimization method for wind power cluster access to the power grid taking into account the reactive power potential of the wind farm according to the present invention;

[0121] Figure 5 It is a comparison diagram of system network loss before and after optimization in the double-layer reactive power optimization method considering reactive power potential of wind farms for access of wind power clusters to the power grid of the present invention;

[0122] Figure 6 It is a comparison diagram of voltage amplitudes at various moments before and after optimization in a double-layer reactive power optimization method considering reactive power potential of wind farms for access of wind power clusters to a power grid according to the present invention;

[0123] Figure 7 It is a comparison diagram of active power loss at each moment after optimization of three scenarios in the double-layer reactive power optimization method considering reactive power potential of wind farms for access of wind power clusters to the power grid of the present invention. DETAILED DESCRIPTION

[0124] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0125] Example 1

[0126] The present invention provides a double-layer reactive power optimization method for wind farms connected to the power grid, which takes into account the reactive power potential of wind farms. First, an upper-layer optimization model with the system operating active network loss and voltage deviation as the minimum is established, and the model is solved to effectively reduce the system network loss and realize the safe and energy-saving operation of the power grid. Secondly, the lower layer conducts a refined analysis of the wind power cluster, estimates the reactive power potential of the wind power cluster, takes the reactive margin utilization rate of the wind farm as the objective function, takes the reactive output of each wind turbine unit as the quantity to be optimized, coordinates the reactive power output within the wind power cluster, and adopts an improved whale optimization algorithm to solve the double-layer optimization model. Finally, it is verified on the improved IEEE39 node system that the proposed optimization strategy and algorithm can effectively reduce the network loss and voltage deviation of the power system when the reactive power potential of wind power is maximized, which proves the effectiveness and rationality of the present invention.

[0127] Example 2

[0128] The present invention provides a double-layer reactive power optimization method considering the reactive power potential of wind farms for wind power cluster access to a power grid, which specifically includes the following steps:

[0129] Step 1: Conduct a detailed analysis of the reactive power within the cluster wind farm.

[0130] During the operation of a wind farm, the power output by wind turbines is transmitted through transmission lines and various transformers. Therefore, when conducting a detailed analysis of the interior of a wind farm, the reactive power losses of transmission lines and transformers must be considered.

[0131] In order to focus on the relationship between active power and reactive power loss, it is assumed that the system is in normal operation, the voltage of each node is roughly maintained at a per-unit value of about 1, and the power factor of the series branch port is assumed to be about 1. The mathematical relationship is as follows:

[0132]

[0133] ΔQ Z1 ≈P 2X 1

[0134] Where X 1 Represents the reactance in the line, ΔQ Z1 is the reactive power consumed by the line impedance, P is the active power in the line, and U p is the voltage at the beginning of the transmission line, I is the current flowing through the line, is the power factor.

[0135] The reactive power generated by the distributed capacitance in the transmission line is:

[0136]

[0137] Where B 1 Indicates the susceptance in the line, Q c is the reactive power generated by the line distributed capacitance, U p is the voltage at the beginning of the transmission line, U q is the voltage at the end of the transmission line.

[0138] Combining the above formula, the reactive power loss of the line can be obtained as:

[0139]

[0140] Where X 1 Represents the reactance in the line, B 1 It represents the susceptance in the line, P is the active power in the line, Q c is the reactive power generated by the distributed capacitance of the line, ΔQ Z1 The reactive power consumed by the line impedance.

[0141] The reactive power loss of the line impedance element is:

[0142]

[0143] Where ΔQ Zt is the reactive power consumed by the transformer equivalent line impedance, k is the non-standard transformation ratio of the transformer under the standard value, X t It represents the reactance in the equivalent line impedance, and P is the active power in the line.

[0144] The sum of reactive power absorbed by the ground branch is:

[0145]

[0146] Where X t represents the reactance in the equivalent line impedance, k is the non-standard transformation ratio of the transformer under the standard value, U p is the voltage at the beginning of the transmission line, U q is the voltage at the end of the transmission line.

[0147] The reactive loss of ground excitation admittance is:

[0148]

[0149] Where B t Indicates the susceptance in the equivalent line impedance, U q is the voltage at the end of the transmission line.

[0150] The transformer reactive loss can be obtained as:

[0151]

[0152] Where P is the active power in the line, k is the non-standard transformation ratio of the transformer under the standard value, and B t represents the susceptance in the equivalent line impedance, X t Represents the reactance in the equivalent line impedance.

[0153] Step 2: Perform a detailed reactive power analysis centered on the wind turbines and collection lines within the cluster wind farm.

[0154] Define the line between the jth wind turbine and the j-1th wind turbine in the i-th feeder in the wind farm cluster as ij, and define the reactive zone ij centered on the wind turbine, which includes the wind turbine ij, the corresponding box transformer of the wind turbine, the transmission line ij and half the length of ij-1.

[0155] The reactive power loss in the upstream transmission line is:

[0156]

[0157] Where X l,i,j , B l,i,j represents the reactance and susceptance in the upstream line i,j, P i,b is the active power of the bth wind turbine in the i-th feeder, b is the wind turbine number, and n is the total number of wind turbines in a single feeder.

[0158] The reactive power loss in the downstream transmission line is:

[0159]

[0160] Where X l,i,j+1 , B l,i,j+1 represents the reactance and susceptance in the downstream line i,j+1.

[0161] The reactive power loss of the box-type transformer connected to a single fan is:

[0162]

[0163] Where X t.i,j , B t.i,jrepresents the reactance and susceptance of the box-type transformer connected to wind turbines i and j, k t.i,j is the transformation ratio of the box-type transformer, P i,j is the active power of the jth wind turbine in the i-th feeder.

[0164] Therefore, the reactive power loss of each wind turbine area in the wind power cluster is:

[0165]

[0166] The topology of the wind power cluster collection line includes the step-up transformer at the collection station, half of the length of the first feeder section of each line, and the total reactive power of the wind cluster collection and transmission line.

[0167] The reactive power loss of line i,1 is:

[0168]

[0169] Where X l,i,1 , B l,i,1 is the reactance and susceptance of the first section of the line.

[0170] The reactive power loss of the step-up transformer is:

[0171]

[0172] Where X T , B T is the reactance and susceptance of the step-up transformer in the collection line, k T is the ratio of the step-up transformer, a is the feeder number, m is the fan number, P a,b is the active power of the bth wind turbine in the ath feeder.

[0173] The reactive power loss of the collection line is:

[0174]

[0175] Where X L , B L is the reactance and susceptance of the collection line.

[0176] The total reactive power loss of the collection line is:

[0177]

[0178] In the formula, Q l.a,1 is the reactive power loss of the first feeder section of each feeder.

[0179] The total reactive power loss of the wind farm is the sum of the reactive power of the corresponding area of ​​each wind turbine and the reactive power of the collection line. The specific calculation formula is:

[0180]

[0181] When the wind farm is operating normally, the total reactive power output of the wind power cluster and the total reactive power loss Q in the wind farm calculated by this formula are WT The difference is the actual reactive capacity of the wind farm:

[0182]

[0183] Step 3: Establish an upper-level reactive power optimization mathematical model with the goal of minimizing active power loss and voltage deviation.

[0184] Specifically, in the reactive power optimization model of large-scale wind power clusters connected to the power grid, usually only the total active network loss of the system operation is considered, but the impact of the optimized voltage deviation on the network and the important impact of reactive power regulation on the system network loss and voltage are ignored. Therefore, in the upper model, an hourly optimization model is established, with the minimum system active network loss and voltage deviation as the objective function.

[0185]

[0186] ΔU imax =U imax -U imin

[0187] Where: U i is the voltage value of node i in the system; N is the set of system nodes; U*i is the voltage reference value of node i in the system, which is generally set to 1.0pu; ΔU imax The maximum voltage deviation allowed at node i in the system, ΔU is the voltage deviation, U imax is the maximum voltage value of node i, U imin is the minimum voltage value of node i.

[0188]

[0189] Where: P loss is the total network loss of the system; U ij is the voltage value of nodes i and j in the system; G ij is the conductance between nodes i and j; θ ij is the voltage phase difference between nodes i and j.

[0190] The power balance constraints of the upper optimization model are as follows:

[0191]

[0192] Where P Gi is the active power supply of node i in the system; P li is the active load of node i in the system; Q Gi is the reactive power of node i in the system; Qli is the reactive load of node i in the system; Q Ci is the reactive power provided by the reactive compensation device at node i in the system, G ij and B ij are the real and imaginary parts of the node admittance matrix, respectively.

[0193] The grid safety operation constraints are as follows:

[0194] U gimin ≤U gi ≤U gimax i∈N g

[0195] P gi,min ≤P gi ≤P gi,max i∈N g

[0196] Q gi,min ≤Q gi ≤Q gi,max i∈N g

[0197] Q ci,min ≤Q ci ≤Q ci,max i∈N c

[0198] T kmin ≤T k ≤T kmax i∈N t

[0199] Where: U gimax , U gimin is the upper and lower limits of the terminal voltage of the i-th generator; P gimax , P gimin is the upper and lower limits of the active output of the i-th generator; Q gimax , Q gimin is the upper and lower limits of the reactive power output of the i-th generator; Q cimax , Q cimin is the upper and lower limits of the switching capacity of the i-th reactive power compensation device; T kimax 、T kimin N is the upper and lower limits of the transformation ratio of the i-th voltage regulating transformer; g is the total number of generators in the system; N c N is the total number of reactive power compensation devices in the system; t is the total number of voltage regulating transformers in the system; h m min ,h m max They are the upper and lower limits of the power flow of branch m respectively.

[0200] OLTC-related constraints:

[0201] u i,t =k ij,t u j,t

[0202] k ij,t =k ij,0 +K ij,t Δk ij

[0203]

[0204] In the above formula, k ij,t is the transformation ratio of OLTC at time t; k ij,0 is the transformation ratio of OLTC at the initial moment; Δk ij,t K is the change in the transformation ratio; ij,t is the gear position at time t; Kmax ij is the highest gear position; Δmax OLTC is the maximum tap change step number of OLTC, u i,t is the voltage of node i at time t, u j,t is the voltage of node j at time t, Z is the number of adjustable gears of OLTC, and T is the time.

[0205] CB related constraints:

[0206]

[0207] in, It is the reactive power output of each CB; is the number of CBs of node i at time t; is the reactive power output of CB of node i at time t, is the maximum number of CB switching changes, It is the maximum number of switching groups of CB.

[0208] Step 4: Input the original data of the upper model, including the network component parameters, the active and reactive output values ​​of conventional power sources, conventional loads and their power factors. The improved whale algorithm is used to solve the upper model, and a set of Pareto optimal solutions is obtained. The optimal compromise solution is selected, that is, the reactive output values ​​of each wind power cluster and multiple reactive sources within the optimization time period, and it is used as the input of the lower optimization model.

[0209] The whale optimization algorithm solves the optimization problem by simulating the predation behavior of humpback whales. It has the advantages of simple operation and strong optimization ability, but it also has the defects of low convergence accuracy and easy to fall into local optimality. To address these shortcomings, the present invention adopts an improved whale algorithm.

[0210] 1) Search for prey. The position of each whale in the search space is a solution, and the position of the whale group currently in the best hunting position is the target prey or close to the best target prey.

[0211]

[0212] t is the current iteration number; X is the current whale position vector; X rand is the position vector randomly selected by the whale population; A and C are the matrix coefficients, and D is the distance between the whale group and the prey.

[0213]

[0214] r is a random number between [0,1]; t m is the upper limit of the number of algorithm iterations, and a is a value that gradually decreases between [2,0] during the algorithm iteration process.

[0215] 2) Surround the prey

[0216] The other whales in the group all surround the optimal position to update their own positions:

[0217]

[0218] X* is the optimal position vector of the global population.

[0219] 3) Spiral bubble attack

[0220] Humpback whales will shrink and encircle the prey in a spiral line, with a probability of 0.5 to perform the surrounding prey behavior and spiral renewal behavior. The mathematical model is as follows:

[0221]

[0222] b is a constant used to limit the shape of the spiral; l is a random number between [-1,1]; P is a random number between [0,1], and D' is the latest distance between the whale group and the prey.

[0223] The advantage of particle swarm algorithm is introduced in the whale algorithm. In each iteration, the individual extreme value p is obtained. best and the global extremum g best Update particles and optimize the positions of particles with poor fitness during the iteration process. The improved algorithm retains the search strategy of the whale algorithm to update its own position by surrounding the optimal position and shrinking around the spiral line, and adds the individual experience of the particle swarm algorithm and the guidance of the optimal particles. The formula for speed and position update in the algorithm is as follows:

[0224]

[0225] ω is the inertia weight factor, is the velocity of particle i in the jth dimension at the t+1th iteration, is the position of particle i in the jth dimension at the t+1th iteration, is the position of particle i in the jth dimension at the tth iteration, is the position of the jth dimension of whale individual i at the tth iteration, c 1 、c 2 is the learning factor, r 1 、r 2 is a random number distributed in [0,1].

[0226] Step 5: Establish a lower-level reactive power optimization model, whose objective function is to maximize the reactive power margin utilization rate of the wind power cluster. The input of this model includes the original parameters inside the wind power cluster, the active power output of wind power, and the reactive power output value obtained by solving the upper-level model.

[0227] The lower layer establishes a 15-minute rolling optimization model with the highest reactive margin utilization rate of the wind power cluster as the objective function. The active output value of a group of wind turbines is input every 15 minutes, and the reactive output Q of the upper layer at 24 moments is input. set As the initial parameters for lower layer optimization.

[0228]

[0229] In the formula, Q WFi represents the reactive capacity of the i-th wind turbine, Q WF Represents the total reactive capacity of the wind farm, Q i represents the actual reactive power output value of the i-th wind turbine, Q set Optimize the reactive power output of wind power transmitted to the upper layer.

[0230] Establish relevant constraints, including:

[0231] Power flow constraints:

[0232]

[0233] Where P WTi , Q WTi are the active output and reactive output of the i-th wind turbine generator set, Q WTi_max , Q WTi_min is the upper and lower limits of wind turbine output, P WTi_max , P WTi_min U is the upper and lower limits of the active output of the wind turbine; i_max , U i_min are the upper and lower limits of the voltage at the access point of the i-th wind turbine group respectively.

[0234] Step 6: Use the improved whale algorithm to solve the lower optimization model and obtain the reactive power output plan within the wind power cluster. Use the optimized results to update the upper and lower limits of the reactive power output of the upper wind power cluster as the boundaries of the new round of optimization for the next iterative calculation.

[0235] Example 3

[0236] The present invention performs simulation calculation on the modified IEEE 39-node system. The system structure is as follows: Figure 2 As shown in the figure, the base power of the system is 100MVA. In the system, nodes 7, 26, 20, and 29 are connected to wind power clusters WT1, WT2, WT3, and WT4 with installed capacity of 45MW respectively; on-load tap-changing transformers are configured in the four lines 2-30, 11-12, 22-35, and 29-38, with a total of 8 gears and an adjustment step of 0.025pu; switchable capacitor groups (CB) are connected to nodes 8 and 27, with a maximum number of switchable groups of 5 and a single group capacity of 20Mvar; static VAR compensation devices (SVC) with a capacity of 100Mvar are connected to nodes 4, 12, and 21 respectively.

[0237] Figure 3 This is the topological diagram of the wind power cluster structure in the lower-level optimization system. The wind power cluster includes 10 wind turbines with a rated capacity of 1.875MW, which are connected to the grid through high-voltage transmission lines after boosting.

[0238] Figure 4 This is a typical topological structure diagram inside a wind power cluster. There are m lines connected to the low-voltage busbar of the wind farm. There are n wind turbines on each line. Each line is connected to the power system through a step-up transformer and a transmission line.

[0239] Figure 5 The comparison chart of system network loss before and after optimization shows that the active network loss of the optimized system has decreased significantly at every moment. Before optimization, the average network loss power of the system in 24 hours a day was 104.6538MW, and the average voltage deviation was 11.31pu. After reactive power optimization, the average network loss power of the system was 95.8948MW, the average voltage deviation was 5.49pu, and the loss reduction rate reached 8.65%.

[0240] Figure 6 The voltage amplitude comparison diagram of the system at each moment before and after optimization shows that before optimization, the voltage of some nodes has exceeded the safety range of the system. By comparison, it can be concluded that the voltage amplitude has decreased overall after optimization and still remains within a reasonable range. This optimization measure is conducive to reducing system network losses and maintaining voltage stability.

[0241] The optimization scenarios are set. Scenario 1 does not consider the impact of reactive power compensation and the transformer ratio adjustment system, as well as the impact of the internal structure of the wind power cluster and the reactive power potential of the wind farm on the active power loss. Scenario 2 does not consider the impact of reactive power compensation and the transformer ratio adjustment system, but considers the impact of the internal structure of the wind power cluster and the reactive power potential of the wind farm on the active power loss. Scenario 3 considers the impact of reactive power compensation and the transformer ratio adjustment system, as well as the impact of the internal structure of the wind power cluster and the reactive power potential of the wind farm on the active power loss. By solving the optimization model, the optimization results of the three scenarios are as follows: Figure 7 As shown in the table, it is easy to find that the active network loss in optimization scenario 1 is 107.4MW. The active network loss in optimization scenario 2 is 104.9MW. The active loss in scenario 3 is 95.4MW. The loss reduction results are better when considering the impact of reactive power compensation and adjusting the transformer ratio on the system and the impact of the internal structure of the wind power cluster and the reactive power potential of the wind farm on the active power loss.

[0242] The double-layer reactive power optimization method for wind farms connected to the power grid adopted by the present invention taking into account the reactive power potential of wind farms, in this embodiment, formulates the switching power of the reactive compensation device according to the optimization results, so as to enhance the system voltage stability, effectively reduce the power loss in the system, coordinate the reactive power output of the wind turbines within the cluster, and improve the loss reduction level.

Claims

1. A two-layer reactive power optimization method considering the reactive power potential of wind farms for wind power cluster access to the power grid, characterized in that: The following steps are involved: Step 1: Perform a detailed analysis of the reactive power in the wind farm cluster to obtain the reactive loss of the transmission line and transformer; Step 2: Based on the reactive power loss obtained in step 1, a refined analysis is performed with the wind turbines and the collection lines in the wind power cluster as the center to obtain the reactive power margin of the wind power cluster; Step 3: Establish a reactive power optimization upper layer model with the goal of minimizing active power loss and voltage deviation; Step 4: Use the improved whale optimization algorithm to solve the upper model and obtain the reactive output values ​​of each wind power cluster and multiple reactive sources within the optimization time period; Step 5, establishing a reactive power optimization lower layer model with the goal of maximizing the reactive power margin utilization rate obtained in step 2, and inputting the reactive power output value obtained in step 4; Step 6: Use the improved whale optimization algorithm to solve the lower model and obtain the reactive power output plan within the wind power cluster.

2. The double-layer reactive power optimization method considering the reactive power potential of wind farms for wind power cluster access to the power grid as claimed in claim 1, characterized in that: The step 1 is specifically as follows: assuming that the system is in a normal operating state, the voltage of each node is maintained at a per-unit value of 1, and assuming that the power factor of the series branch port is 1, the mathematical relationship is as follows: ΔQ Z1 ≈P 2 X1 In the formula, X1 represents the reactance in the line, ΔQ Z1 is the reactive power consumed by the line impedance, P is the active power in the line, and U p is the voltage at the beginning of the transmission line, I is the current flowing through the line, is the power factor; The reactive power generated by the distributed capacitance in the transmission line is: In the formula, B1 represents the susceptance in the line, Q c is the reactive power generated by the line distributed capacitance, U p is the voltage at the beginning of the transmission line, U q is the voltage at the end of the transmission line; Combining the above formula, the reactive power loss of the line is: In the formula, X1 represents the reactance in the line, B1 represents the susceptance in the line, P is the active power in the line, Q c is the reactive power generated by the line distributed capacitance, ΔQ Z1 The reactive power consumed by the line impedance; The reactive power loss of the line impedance element is: In the formula, ΔQz t is the reactive power consumed by the transformer equivalent line impedance, k is the non-standard transformation ratio of the transformer under the standard value, X t It represents the reactance in the equivalent line impedance, and P is the active power in the line; The sum of reactive power absorbed by the ground branch is: In the formula, X t represents the reactance in the equivalent line impedance, k is the non-standard transformation ratio of the transformer under the standard value, U p is the voltage at the beginning of the transmission line, U q is the voltage at the end of the transmission line; The reactive loss of ground excitation admittance is: In the formula, B t Indicates the susceptance in the equivalent line impedance, U q is the voltage at the end of the transmission line; The reactive power loss of the transformer is obtained as: Where P is the active power in the line, k is the non-standard transformation ratio of the transformer under the standard value, and B t represents the susceptance in the equivalent line impedance, X t Represents the reactance in the equivalent line impedance.

3. The double-layer reactive power optimization method considering the reactive power potential of wind farms for wind power cluster access to the power grid as claimed in claim 1, characterized in that: The step 2 is specifically as follows: define the line between the j-th wind turbine and the j-1-th wind turbine in the i-th feeder in the wind power cluster as ij, and define a reactive zone ij centered on the wind turbine, which includes the wind turbine ij and its corresponding box-type transformer, and half the length of the transmission line i,j and i,j-1; The reactive power loss in the upstream transmission line is: In the formula, X l,i,j , B l,i,j represents the reactance and susceptance in the upstream line i,j, P i,b is the active power of the bth wind turbine in the i-th feeder, b is the wind turbine number, and n is the total number of wind turbines in a single feeder; The reactive power loss in the downstream transmission line is: Where X l,i,j+1 , B l,i,j+1 represents the reactance and susceptance in the downstream line i,j+1; The reactive power loss of the box-type transformer connected to a single fan is: Where X t.i,j , B t.i,j represents the reactance and susceptance of the box-type transformer connected to wind turbines i and j, k t.i,j is the box-type transformer ratio, P i,j is the active power of the jth wind turbine in the i-th feeder; Therefore, the reactive power loss of each wind turbine area in the wind power cluster is: The topology of the wind power cluster collection line includes the step-up transformer at the collection station, half of the length of the first feeder section of each line, and the total reactive power of the wind cluster collection and transmission line; The reactive power loss of line i,1 is: In the formula, X l,i,1 , B l,i,1 is the reactance and susceptance of the first section of the line; The reactive power loss of the step-up transformer is: In the formula, X T , B T is the reactance and susceptance of the step-up transformer in the collection line, k T is the ratio of the step-up transformer, a is the feeder number, m is the fan number, P a,b is the active power of the bth wind turbine in the ath feeder; The reactive power loss of the collection line is: In the formula, X L , B L is the reactance and susceptance of the collection line; The total reactive power loss of the collection line is: In the formula, Q l.a,1 is the reactive power loss of the first feeder section of each feeder; The total reactive power loss of the wind farm is the sum of the reactive power of the corresponding area of ​​each wind turbine and the reactive power of the collection line. The specific calculation formula is: In the formula, Q l.a,b is the reactive power loss of each section of each feeder, Q t.a,b is the reactive power loss of the wind box transformer in each section of the feeder; When the wind farm is operating normally, the total reactive power output of the wind power cluster and the calculated total reactive power loss Q in the wind farm are WT The difference is the actual reactive capacity of the wind farm:

4. The double-layer reactive power optimization method considering the reactive power potential of wind farms for wind power cluster access to the power grid as claimed in claim 1, characterized in that: The step 3 is specifically as follows: establishing an hourly optimization model, with the objective function being to minimize the system active network loss and voltage deviation: ΔU imax =U imax -U imin Where U i is the voltage value of node i in the system, N is the set of system nodes, U* i is the voltage reference value of node i in the system, ΔU imax The maximum voltage deviation allowed at node i in the system, ΔU is the voltage deviation, U imax is the maximum voltage value of node i, U imin is the minimum voltage value of node i; Where P loss is the total network loss of the system, U i is the voltage value of node i in the system, U j is the voltage value of node j in the system, G ij is the conductance between nodes i and j, θ ij is the voltage phase difference between nodes i and j, N c is the total number of reactive power compensation devices in the system; The power balance constraints of the upper optimization model are as follows: Where P Gi is the active power of node i in the system, P li is the active load of node i in the system, Q Gi is the reactive power of node i in the system, Q li is the reactive load of node i in the system, Q Ci is the reactive power provided by the reactive compensation device at node i in the system, G ij and B ij are the real and imaginary parts of the node admittance matrix respectively; The grid safety operation constraints are as follows: IN gimin ≤U gi ≤U gimax i∈N g P gi,min ≤P gi ≤P gi,max i∈N g Q gi,min ≤Q gi ≤Q gi,max i∈N g Q ci,min ≤Q ci ≤Q ci,max i∈N c T kmin ≤T k ≤T kmax i∈N t Where U gimax , U gimin is the terminal voltage U of the i-th generator gi The upper and lower limits of P gimax , P gimin is the active output P of the i-th generator gi The upper and lower limits of Q gimax , Q gimin is the reactive power output Q of the i-th generator gi The upper and lower limits of Q cimax , Q cimin is the switching capacity Q of the i-th reactive power compensation device ci The upper and lower limits of T kimax 、T kimin is the transformation ratio T of the i-th voltage regulating transformer k The upper and lower limits of N g is the total number of generators in the system, N c is the total number of reactive power compensation devices in the system, N t is the total number of voltage regulating transformers in the system; OLTC-related constraints: in i,t =k ij,t in j,t k ij,t =k ij,0 +K ij,t Δk ij In the formula, k ij,t is the transformation ratio of OLTC at time t, k ij,0 is the transformation ratio of OLTC at the initial moment, Δk ij , is the change in ratio, K ij,t is the gear position at time t, The highest gear, is the maximum number of tap change steps in OLTC, u i,t is the voltage of node i at time t, u j,t is the voltage of node j at time t, Z is the number of adjustable gears of OLTC, and T is the time; CB related constraints: In the formula, is the reactive power output of each CB, is the number of CBs of node i at time t, is the reactive power output of the CB of node i at time t, is the maximum number of CB switching changes, It is the maximum number of switching groups of CB.

5. The double-layer reactive power optimization method considering the reactive power potential of wind farms for wind power cluster access to the power grid as claimed in claim 1, characterized in that: The step 4 specifically includes: inputting the original data of the upper model, including the component parameters of the network, the active and reactive output values ​​of the conventional power source, the conventional load and its power factor, solving the upper model by using the improved whale optimization algorithm, optimizing a set of Pareto optimal solutions, and selecting the optimal compromise solution, that is, optimizing the reactive output values ​​of each wind power cluster and multiple reactive sources within the time period; Improvements to the whale optimization algorithm include: 1) Searching for prey The position of each whale in the search space is a solution. Let the position of the whale group currently in the best hunting position be the target prey or close to the best target prey: Where t is the current iteration number, X is the current whale position vector, and X rand is the position vector randomly selected by the whale population, A and C are matrix coefficients, and D is the distance between the whale population and the prey; In the formula, r is a random number between [0,1], t m is the upper limit of the number of algorithm iterations, and a is a value that gradually decreases between [2,0] during the algorithm iteration process; 2) Surround the prey The other whales in the group all surround the optimal position to update their own positions: Where X* is the optimal position vector of the global group; 3) Spiral bubble attack Humpback whales will shrink and encircle the prey in a spiral line, with a probability of 0.5 to perform the surrounding prey behavior and spiral renewal behavior. The mathematical model is as follows: b is a constant used to define the shape of the spiral; l is a random number between [-1,1]; P is a random number between [0,1], and D' is the latest distance between the whale group and the prey; Each iteration passes through the individual extreme value p best and the global extremum g best Update particles and optimize the positions of particles with poor fitness during the iteration process. The formulas for speed and position update are as follows: ω is the inertia weight factor, is the velocity of particle i in the jth dimension at the t+1th iteration, is the position of particle i in the jth dimension at the t+1th iteration, is the position of particle i in the jth dimension at the tth iteration, is the position of the j-th dimension of whale individual i at the t-th iteration, c1 and c2 are learning factors, and r1 and r2 are random numbers distributed in [0,1].

6. The double-layer reactive power optimization method considering the reactive power potential of wind farms for wind power cluster access to the power grid as claimed in claim 1, characterized in that: The step 5 specifically includes: establishing a lower-level model, whose objective function is to maximize the reactive margin utilization rate of the wind power cluster, and the input of the model includes the original parameters inside the wind power cluster, the active output of wind power, and the reactive output value obtained by solving the upper-level model; The lower layer establishes a 15-minute rolling optimization model with the highest reactive margin utilization rate of the wind power cluster as the objective function. The active output value of a group of wind turbines is input every 15 minutes, and the reactive output Q of the upper layer at 24 moments is calculated. set As the initial parameters for lower layer optimization: In the formula, Q WFi represents the reactive capacity of the i-th wind turbine, Q WF Represents the total reactive capacity of the wind farm, Q i represents the actual reactive power output value of the i-th wind turbine, Q set Optimize the reactive power output of wind power transmitted to the upper layer; Establish relevant constraints, including: Power flow constraints: Where P WTi , Q WTi are the active output and reactive output of the i-th wind turbine generator set, Q WTi_max , Q WTi_min is the upper and lower limits of wind turbine output, P WTi_max , P WTi_min is the upper and lower limits of the active output of the wind turbine, U i_max , U i_min are the access point voltage U of the i-th wind turbine group respectively. i Upper and lower limits.

7. The double-layer reactive power optimization method considering the reactive power potential of wind farms for wind power cluster access to the power grid as claimed in claim 1, characterized in that: The step 6 is specifically as follows: using the improved whale optimization algorithm to solve the lower model, obtain the reactive power output plan within the wind power cluster, and use the optimized results to update the upper and lower limits of the reactive power output of the upper wind power cluster as the boundaries of a new round of optimization, and perform the next iterative calculation.