Networking type energy storage hierarchical optimization site selection method considering multi-scale stability influence

Abstract

The invention discloses a network construction type energy storage hierarchical optimization site selection method considering multi-scale stability influence. The method comprises the following steps: designing a GFM-ESS hierarchical site selection strategy based on different voltage levels; establishing a three-level site selection model based on a GFM-ESS hierarchical site selection strategy, wherein the three-level site selection model comprises a GFM-ESS system-level site selection planning model, a GFM-ESS station-level site selection planning model and a GFM-ESS equipment-level site selection planning model; solving the three-level site selection model to obtain an optimal site selection scheme; according to the method, the GFM-ESS installation requirements are set from the system level, the station level and the equipment level through the constructed three-level site selection model, the installation position is optimized while the economic benefits are considered, the voltage stability of the power system is effectively enhanced, the small signal stability of the GFM station is guaranteed, and the supporting capacity of the equipment is improved to the maximum extent.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a hierarchical optimization site selection method for grid-type energy storage considering the influence of multi-scale stability. Background Art

[0002] Wind and solar energy, as inverter-based resources (IBRs), face two major challenges when applied to the power grid on a large scale. On the one hand, they cannot provide inertia and damping. On the other hand, the increase in the electrical distance between renewable energy equipment and the system weakens the strength of the grid. Unlike traditional IBRs, the Grid-Forming Energy Storage System (GFM-ESS) can simulate the inertia and damping characteristics of synchronous machines, provide voltage and frequency support to the grid, and has excellent adaptability to weak grids. Therefore, GFM-ESS is considered to be an effective means to cope with a series of challenges brought about by the high penetration of IBRs.

[0003] At present, the research on GFM-ESS mainly focuses on the equipment itself, such as its advanced control strategy and operation stability analysis, while the application of GFM-ESS in power systems is insufficient. In particular, in terms of power system planning, the optimal location of GFM-ESS is still an unresolved issue. Existing research on energy storage planning mainly focuses on traditional energy storage. For example, in terms of cost-benefit analysis, energy arbitrage is achieved by peak shaving and valley filling, revenue is generated by participating in the power auxiliary service market, and battery life is extended to reduce costs. In terms of participating in system operation, the peak-shaving and frequency-regulating characteristics of ESS are considered to optimize the configuration, and the spatial and temporal distribution characteristics of ESS virtual inertia contribution are considered to optimize the location. However, the above research mainly focuses on traditional grid-following (GFL)-ESS, and does not consider the specific functions of GFM-ESS. There is a relative lack of research on the optimal location of GFM-ESS. Unlike GFL-ESS, GFL-ESS has higher synchronization stability in weak grids or low SCR systems, which can improve SCR and strengthen weak grid strength. Recently, research on GFM-ESS planning has gradually begun to focus on improving the small signal stability of the system. However, in addition to considering the contribution of GFM-ESS to small-signal stability, the inherent voltage and frequency support characteristics of GFM-ESS should also be considered, and its proper placement can also significantly improve large-disturbance stability. Therefore, determining the optimal layout of GFM-ESS should comprehensively consider the stability of GFM-ESS itself and its contribution to the stability of power system operation. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a hierarchical optimization site selection method for grid-type energy storage that takes into account the influence of multi-scale stability, which aims to solve the problems mentioned in the background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a hierarchical optimization site selection method for grid-type energy storage considering the influence of multi-scale stability, comprising the following steps:

[0006] Step S1: Designing a hierarchical site selection strategy for GFM-ESS based on different voltage levels;

[0007] Step S2: Establish a three-level site selection model based on the GFM-ESS hierarchical site selection strategy, including a GFM-ESS system-level site selection planning model, a GFM-ESS station-level site selection planning model, and a GFM-ESS equipment-level site selection planning model;

[0008] Step S3: Solve the three-level site selection model to obtain the best site selection plan;

[0009] The GFM-ESS hierarchical site selection strategy includes:

[0010] First, considering the stable distribution characteristics of static voltage, the effect of voltage reactive power compensation, and the network loss of the power system, identify the nodes in the 220kV transmission network that are weaker than the specified value;

[0011] Article 2. Based on the result of Article 1, select the node whose GFM-ESS synchronization stability is higher than that of GFL-ESS;

[0012] Article 3. From the perspective of GFM-ESS enhancing the transient stability of the power system and considering the comprehensive impact on the dynamic frequency and voltage support of the power grid, GFM-ESS is installed at nodes with a voltage level of 35kV.

[0013] Furthermore, GFM is a virtual synchronous generation VSG, and the control loop of VSG consists of an active and reactive synchronization loop and a voltage and current control double closed loop structure;

[0014] The power angle swing equation of the active synchronous control loop is expressed as:

[0015]

[0016] Where, T j represents the network inertial time constant; d represents differential; t represents time; P ref Indicates the grid-type active power reference value; P e Indicates the electromagnetic power of the power grid; D p represents the network damping coefficient; ω represents the angular frequency; ω N Indicates the rated angular frequency of the power grid; δ GFM Indicates the grid-type output power angle;

[0017] Simplify formula (1):

[0018]

[0019]

[0020] In the formula, Represents the second-order derivative of the grid-type output power angle; represents the first-order derivative of the output power angle of the network; E represents the internal potential of the network; E g Indicates the grid voltage; X f represents the internal reactance of the network; X g Indicates the grid reactance;

[0021] According to the voltage vector relationship, Replaced by the second-order derivative of the potential phase within the network The first-order derivative of the potential phase in the network It is expressed as:

[0022]

[0023] T j ′=T j f(δ e )(5);

[0024]

[0025] Where, T j ′ represents the equivalent inertia constant; D p ′ represents the equivalent damping coefficient; P e ′ represents the equivalent electromagnetic power; f(·) represents the function; δ e Indicates the potential phase within the network;

[0026] The reactive power-voltage control equation of the reactive synchronization control loop is expressed as:

[0027] E=E N +D q (Q ref -Q e ) (8);

[0028]

[0029] D′ q =D q f(δ e ) (10);

[0030] In the formula, E N Indicates the rated voltage of the power grid; D qRepresents reactive power droop coefficient; Q ref Indicates reactive power reference value; Q e Indicates the instantaneous reactive power of the power grid; D q ′ represents the equivalent reactive power droop coefficient.

[0031] Furthermore, based on the impact of GFM-ESS access on the static voltage stability of the power system, the pre-site selection principle and the three-level site selection optimization model are used to jointly plan the ESS site selection. With the goals of improving the static voltage stability margin, maximizing the static voltage regulation sensitivity and minimizing the branch power loss, a GFM-ESS system-level site selection planning model is established.

[0032] Furthermore, the indicators of the pre-site selection principle include:

[0033] Voltage sensitivity index, expressed as:

[0034]

[0035] In the formula, It represents the derivative of active power; Represents the derivative of reactive power; J Pθ represents the Jacobian matrix about Pθ; J Qθ represents the Jacobian matrix about Qθ; J PV represents the Jacobian matrix about PV; J QV represents the Jacobian matrix about QV; represents the derivative of the phase angle; represents the derivative of voltage;

[0036] Assuming that the active power disturbance is zero, the sensitivity of voltage to reactive power is calculated and expressed as:

[0037]

[0038] In the formula, ΔP represents the active power increment; Indicates about The Jacobian matrix of ;

[0039] Assuming that the reactive power disturbance is zero, the active power sensitivity is obtained, which is expressed as:

[0040]

[0041] In the formula, ΔQ represents the reactive power increment; Indicates about The Jacobian matrix of ;

[0042] Considering both active sensitivity and reactive sensitivity, the voltage sensitivity index VSI calculation formula is expressed as:

[0043]

[0044] Furthermore, with the goal of improving the static voltage stability margin, maximizing the static voltage regulation sensitivity and minimizing the branch power loss, the specific process of establishing the GFM-ESS system-level site selection planning model is as follows:

[0045] GFM-ESS and IBRs are connected to the power system through PCC;

[0046] set up Among them, U p Indicates the voltage at PCC; represents the derivative of the grid voltage; represents the derivative of the voltage at PCC; θ p Indicates the phase of the voltage at the PCC. When GFM-ESS is not connected, the power output of IBRs to the grid is expressed as:

[0047]

[0048] Where P re Indicates the active power output by IBRs; Q re Represents the reactive power output by IBRs; R g Indicates the grid resistance;

[0049] Ignore R g , eliminating θ in equation (21) p , then the power P at PCC re With voltage U p The relationship is:

[0050]

[0051] After connecting to GFM-ESS, the power output of PCC to the grid becomes P pcc =P re +P es , Q pcc =Q re +Q es , where P pcc represents the active power of PCC when GFM-ESS is connected, P es Indicates the active power output of the grid-connected energy storage, Q pcc represents the reactive power at PCC, Q es Indicates the reactive power output of the grid-type energy storage; P pcc and Q pcc Substituting P in formula (22) re and Q re , establish the power P of PCC under GFM-ESS access pcc and voltage U pthe relationship between;

[0052] The static voltage stability margin improvement index is expressed as:

[0053]

[0054] Where U MAX is the critical bus voltage stability limit of the current power system, U L0 is the voltage value of the current operating point, VSM is the voltage stability margin, VSMI is i is the deviation value of VSM at node i before and after installing GFM-ESS; VSM ess，i Indicates the voltage stability margin after grid-connected energy storage; VSM 0，i Indicates the voltage stability margin before grid-connected energy storage is connected;

[0055] The static voltage regulation sensitivity index is expressed as:

[0056]

[0057] Where n is the total number of power system nodes, U iref is the node voltage reference value, U i is the node voltage amplitude, ∑ΔU base VRI is the total offset of the node voltage before GFM-ESS is connected; i represents the static voltage regulation sensitivity at node i in the power system;

[0058] Branch power loss index, which is expressed as:

[0059]

[0060] Where b is the total number of system branches; R l is the branch resistance, I l is the branch current; l represents the lth branch in the system; U i，i+1 represents the voltage of the busbars at both ends of the lth branch; Z i，i+1 PLI represents the impedance between the busbars at both ends of the lth branch; i represents the branch power loss at node i in the power system;

[0061] The GFM-ESS system-level site selection planning model is established and expressed as:

[0062] Minf1(x)=-VSMI×VRI (26);

[0063] Minf2(x)=PLI (27);

[0064] J ess，k ∈γ see(28);

[0065]

[0066] In the formula, J ess，k is the node connected to unit k in GFM-ESS, γ see The set of pre-selected sites obtained by sorting the VSIs. and are the upper and lower limits of energy storage installed capacity, and are the upper and lower limits of the energy storage installed power respectively; Minf1(x) represents the minimum value of f1(x); Minf2(x) represents the minimum value of f2(x); f1 represents the function expression of the multiplication of VSMI and VRI; f2 represents the function expression of PLI; x represents the function variable; Indicates the nominal capacity of grid-type energy storage; Indicates the nominal power of grid-type energy storage; Represents the minimum value of the voltage at node i; represents the maximum value of the voltage at node i; Indicates the minimum value of branch b power; P b represents the power of branch b; Indicates the maximum power of branch b.

[0067] Furthermore, considering the impact of GFM-ESS, the definition of gSCR is extended, and based on satisfying the stability constraints of GFM-ESS itself, a GFM-ESS site selection planning model is constructed with the goal of maximizing the improvement of small disturbance stability after GFM-ESS access. Specifically:

[0068] Calculate the small signal stability of the multi-infeed power electronic system MIPES after connecting to the GFM-ESS;

[0069] MIPES includes the device side and the network side. The dynamic equation of the network side is expressed as:

[0070]

[0071] Where Y Net (s) is the AC grid admittance matrix, B∈R n×n is the Thevenin equivalent admittance matrix, B=1 / (ω0L), α(s)=ω0 2 / (s 2 +ω0 2 ),β(s)=sω0 / (s 2 +ω0 2 ), ω0 represents the synchronization frequency; α(s), β(s) represent complex functions about s; L represents the inductance of the line; s represents the Laplace operator; ΔIx , ΔI y are matrices composed of the increments of the x-axis and y-axis components of the current injected into node i in the global coordinate system; ΔU x , ΔU y are matrices composed of the increments of the x-axis and y-axis components of the voltage at node i in the global coordinate system; γ(s) represents the matrix composed of α(s) and β(s);

[0072] The dynamic equation on the equipment side is expressed as:

[0073]

[0074] In the formula, ΔI xi , ΔI yi They represent the increments of the x-axis and y-axis components of the current injected into node i in the global coordinate system; De_xy (s) represents the admittance transfer function matrix of the device in the x, y global coordinate system; ΔU xi , ΔU yi They represent the increments of the x-axis and y-axis components of the voltage at node i in the global coordinate system; T i Represented by cosθ i 、sinθ i The matrix composed of B Indicates the rated capacity of the equipment; G dd (s), G dq (s), G qd (s), G qq (s) represents the input-output transfer function of the device at the dq axis port; Y De_dq (s) represents the admittance transfer function matrix of the device in the dq global coordinate system; θ i represents the phase angle of node i;

[0075] According to the multivariable feedback control theory, the closed-loop characteristic equation of MIPES is expressed as:

[0076]

[0077] In the formula, I 2n represents the 2n-order identity matrix;

[0078] By performing an equivalent transformation on the closed-loop characteristic equation, MIPES is decoupled into n independent equivalent SIPES, which can be expressed as:

[0079]

[0080] In the formula, I n represents the n-order identity matrix;

[0081] The closed-loop characteristic equation is equivalently transformed into:

[0082]

[0083] In the formula, λ i Represents the matrix Y eq The i-th eigenvalue of ;

[0084] make 0<λ1≤λ2≤λ i , …, ≤λ n , i=1,…,n;

[0085] Small signal stability depends on the matrix Y eq The SIPES corresponding to the minimum eigenvalue λ1; define λ1 as gSCR;

[0086] Install capacity S on node n VSG The GFM-ESS is equivalent to adding a parallel susceptance S to the ground branch. VSG B p ; Because the Thevenin equivalent admittance matrix B is modified to:

[0087]

[0088] In the formula, B′ represents the modified B;

[0089] Expand gSCR to the extended admittance matrix Y′ eq The minimum eigenvalue of is expressed as:

[0090]

[0091] In the formula, λ + (·) is the positive eigenvalue of the matrix; It is the rated capacity matrix after adding grid-type energy storage;

[0092] The GFM-ESS site selection planning model is expressed as:

[0093]

[0094] Furthermore, based on the time-domain target characteristics of the multi-scale control response characteristics of GFM-ESS, multiple evaluation indicators of GFM-ESS are obtained, and a GFM-ESS equipment-level site selection planning model that takes into account voltage and frequency support capabilities is constructed; specifically:

[0095] The GFM-ESS is considered as a “black box” or “grey box” model, and the dynamic support capability is evaluated by observing the power, voltage and frequency characteristics at multiple time scales at the PCC point;

[0096] GFM-ESS adjusts voltage deviation through reactive power-voltage control.D During the period, the average reactive power compensation provided by GFM-ESS at PCC is defined as the equivalent droop coefficient of GFM-ESS station, expressed as:

[0097]

[0098] In the formula, Indicates that node i is in F k Equivalent voltage droop capability under disturbance; Indicates t0 to t f The average value of U t0 and Q t0 are the initial voltage and reactive power value respectively; U t，i and Q t，i They represent the voltage and reactive power value of node i at time t respectively; F k For large disturbance events; is the expected failure set;

[0099] In addition, the transient voltage peak value and the transient voltage recovery degree are defined as indicators for evaluating the transient voltage stability under large disturbances, as shown in equations (43) and (44):

[0100]

[0101] In the formula, Indicates that node i is in F k Transient voltage peak under disturbance; U max，i Indicates the maximum value of transient voltage; Indicates that node i is in F k The degree of transient voltage recovery under disturbance; t c Indicates the time corresponding to when the voltage returns to a stable value;

[0102] By calculation and The sensitivity of is used to evaluate the improvement of transient voltage stability after the implementation of GFM-ESS, which is expressed as:

[0103]

[0104] In the formula, Indicates that after adding grid-type energy storage, node i is in F k The degree of improvement in transient voltage stability under disturbance; Indicates that after adding grid-type energy storage, node i is in F k Transient voltage peak under disturbance; Indicates that node i is in F after not adding grid-connected energy storage k Transient voltage peak under disturbance; Indicates that after adding grid-type energy storage, node i is in F kThe degree of transient voltage recovery under disturbance; Indicates that node i is in F after not adding grid-connected energy storage k The degree of transient voltage recovery under disturbance; a and b represent and The weight coefficient of

[0105] The equivalent inertia index is expressed as:

[0106]

[0107] In the formula, represents the impact of GFM-ESS power injection at node i on the RoCoF of the power system; RoCoP represents the power change rate; RoCoF represents the frequency change rate; ΔP represents the active power increment; Δf represents the frequency value; Indicates that node i is at t m Active power increment at ; Indicates that node i is at t m The frequency value at ; Indicates the active power increment at the initial moment; Indicates the frequency value at the initial moment;

[0108] The GFM-ESS damping effectiveness index is defined as:

[0109]

[0110] In the formula, Indicates that node i is in F k Equivalent damping under disturbance; t m represents the time corresponding to the lowest frequency point; t0 represents the initial moment of the fault; f t，i represents the frequency value of node i at time t; f t0 Frequency value at the initial moment;

[0111] The indicators that characterize frequency stability include frequency change rate, transient frequency deviation and dynamic recovery frequency deviation, as shown in equations (48) to (50):

[0112]

[0113] In the formula, Indicates that node i is in F k The rate of change of frequency under disturbance; Indicates that node i is in F k Frequency transient deviation under disturbance; Indicates that node i is in F k Frequency dynamic deviation under disturbance; f max，i Indicates the maximum frequency deviation; f d，i Indicates the frequency dynamic deviation value;

[0114] The transient frequency stability improvement effect index after connecting to GFM-ESS is shown in formula (51);

[0115]

[0116] In the formula, Indicates that after the grid-connected energy storage is connected, node i is in F k The transient frequency stability improvement effect under disturbance; Indicates that after the grid-connected energy storage is connected, node i is in F k The rate of change of frequency under disturbance; Indicates that node i is in F before grid-connected energy storage is connected k The rate of change of frequency under disturbance; Indicates that after the grid-connected energy storage is connected, node i is in F k Frequency transient deviation under disturbance; Indicates that node i is in F before grid-connected energy storage is connected k Frequency transient deviation under disturbance; Indicates that after the grid-connected energy storage is connected, node i is in F k Frequency dynamic deviation under disturbance; Indicates that node i is in F before grid-connected energy storage is connected k The frequency dynamic deviation under disturbance; c represents the ... The weight coefficient determined by the importance of

[0117] The GFM-ESS equipment-level site selection planning model is constructed. The GFM-ESS equipment-level site selection planning model uses the proposed evaluation indicators and numerical calculations to obtain the final site selection results.

[0118] Furthermore, the specific process of step S3 is: using the NSGA-II multi-objective optimization algorithm to solve the GFM-ESS system-level site selection planning model to obtain the optimal site selection result of the GFM-ESS system-level site selection planning model, calling the power system simulation software to iteratively solve the GFM-ESS station-level site selection planning model to obtain the optimal site selection result of the GFM-ESS station-level site selection planning model; using the improved GRA-VIKOR evaluation algorithm to solve the GFM-ESS equipment-level site selection planning model to obtain the optimal site selection result of the GFM-ESS equipment-level site selection planning model.

[0119] Furthermore, the improved GRA-VIKOR evaluation algorithm is used to solve the GFM-ESS equipment-level site selection planning model. The specific process of obtaining the optimal site selection result of the GFM-ESS equipment-level site selection planning model is as follows:

[0120] The subjective weights were calculated using Bayesian BWM;

[0121] First, k experts are invited to select the evaluation index (A={A1,A2,…,A n}) to determine the optimal index A B and the worst indicator A W , A n is the nth evaluation index; using the 1-9 scale method, A B and A W As a reference comparison object, construct the best and worst comparison matrix C B ={a B1 ,a B2 ,…,a Bn}, C W ={a 1W ,a 2W ,…,a nW}, a Bn are the n optimal evaluation indicators determined by experts, a nW The n worst evaluation indicators determined by experts;

[0122] Second, use the multinomial distribution to distribute the input A B、 A W Modeling, calculation input A B、 A W The multinomial probability distribution of , and the relationship between weight and expert score are obtained;

[0123] Finally, based on A B、 A W The input and output of the model are obtained by formulating the Bayesian hierarchical model to obtain the subjective weight vector w1;

[0124] The objective weights were calculated using the CRITIC method;

[0125] First, the standard deviation is used to reflect the variability of the evaluation index V i ;

[0126] Secondly, we need to find the conflict S of the evaluation index i , construct an expression to characterize the conflict of indicators based on the correlation coefficient;

[0127] Finally, calculate the objective weight of the evaluation index and convert V i With S i Multiply to get the information amount G of the i-th evaluation index i , based on G i Get the indicator objective weight vector w2;

[0128] Use game weighting method to determine the comprehensive weight;

[0129] First, construct the combined weight vector w3:

[0130] w3=A1w1+A2w2(52);

[0131] In the formula, A1 and A2 are the coefficients corresponding to w1 and w2 respectively;

[0132] Secondly, the idea of ​​game theory is introduced, which can be expressed as:

[0133]

[0134] In the formula, A v is the coefficient variable, v = 1, 2; w v A v The weight of Z Represents an index;

[0135] Finally, the optimal combination weight is obtained;

[0136] Determine the optimal ranking;

[0137] Through the GRA idea, the grey correlation between each index value and the positive and negative ideal solutions is obtained, and the final group utility value Q is calculated. i and individual regret value R i , and according to formula (54), the comprehensive evaluation index T of GFM-ESS installed at different sites is obtained i :

[0138]

[0139] According to T i , determine the best GFM-ESS installation node: let the best site selection scheme be P1, the second best site selection scheme be P2, and according to the maximum group utility value Q i Sort different site selection plans; when the site selection plan meets the preset conditions, it is the best solution.

[0140] Furthermore, the preconditions are:

[0141] Condition (1), T(P1)-T(P2)≥1 / n-1;

[0142] Condition (2), P1 in the group utility value Q i and individual regret R i The first in the ranking;

[0143] The standard expression of the best proportion solution is:

[0144] When conditions (1) and (2) are met, P1 is the best site selection solution;

[0145] When condition (1) is met, P1 and P2 are the best site selection options;

[0146] When condition (2) is met, any site selection plan that does not meet condition (1) is the best site selection plan.

[0147] Compared with the existing technology, the present invention has the following beneficial effects:

[0148] (1) The present invention sets the installation requirements of GFM-ESS from the three levels of system, site and equipment by constructing a three-level site selection model. While considering economic benefits, it optimizes the installation location, effectively enhances the voltage stability of the power system, ensures the small signal stability of the GFM station, and maximizes the support capacity of the equipment.

[0149] (2) Based on the time domain target characteristics of the multi-scale control response characteristics of GFM-ESS, the present invention proposes evaluation indicators of the inertia, damping and voltage support capabilities of GFM-ESS, and quantifies the active support capability of GFM-ESS.

[0150] (3) The present invention reduces the selection range from the large-scale bus of the entire system at the beginning to the specific 35kv site at the end through the designed solution method, which greatly simplifies the site selection process and reduces the calculation complexity. BRIEF DESCRIPTION OF THE DRAWINGS

[0151] Figure 1 The figure is a flow chart of the method of the present invention.

[0152] Figure 2 This is the GFM control topology diagram.

[0153] Figure 3 PV curves before and after GFM-ESS access.

[0154] Figure 4 Schematic diagram of the gSCR derivation of MIPES.

[0155] Figure 5 This is the reactive power response diagram under voltage change. DETAILED DESCRIPTION

[0156] like Figure 1 As shown, the present invention provides a technical solution: a hierarchical optimization site selection method for grid-type energy storage considering the influence of multi-scale stability, comprising the following steps:

[0157] Step S1: Design a hierarchical site selection strategy for GFM-ESS based on different voltage levels.

[0158] The GFM-ESS hierarchical site selection strategy includes:

[0159] First, considering the static voltage stability distribution characteristics, voltage reactive power compensation effect, power system network loss and other factors, the aim is to economically identify the weaker nodes in the 220kV transmission network and enhance the static voltage stability by the best site selection method.

[0160] Article 2. Based on the result of Article 1, select the node where the synchronization stability of GFM-ESS is higher than that of GFL-ESS to ensure the safe operation of ESS at the voltage level of 110kV.

[0161] Article 3. From the perspective of GFM-ESS enhancing the transient stability of the power system and taking into account the comprehensive impact on the dynamic frequency and voltage support of the power grid, it is specified to install GFM-ESS at the 35kV voltage level node to ensure that the power system obtains the maximum support from the GFM-ESS investment.

[0162] Step S2: Establish a three-level site selection model based on the GFM-ESS hierarchical site selection strategy.

[0163] The grid-type GFM is externally manifested as a controllable voltage source with low series impedance, which can actively form an output voltage at the point of common coupling (PCC). The GFM of the present invention is virtual synchronous generation (VSG), and its control structure is as follows: Figure 2 The control loop of VSG is mainly composed of active and reactive synchronization loops and voltage and current control double closed loop structures.

[0164] The power angle swing equation of the active synchronous control loop is expressed as:

[0165]

[0166] Where, T j represents the network inertial time constant; d represents differential; t represents time; P ref Indicates the grid-type active power reference value; P e Indicates the electromagnetic power of the power grid; D p represents the network damping coefficient; ω represents the angular frequency; ω N Indicates the rated angular frequency of the power grid; δ GFM Indicates the output power angle of the network.

[0167] Simplify formula (1):

[0168]

[0169] In the formula, Represents the second-order derivative of the grid-type output power angle; represents the first-order derivative of the output power angle of the network; E represents the internal potential of the network; E g Indicates the grid voltage; X f represents the internal reactance of the network; X g Represents the grid reactance.

[0170] According to the voltage vector relationship, Replaced by the second-order derivative of the potential phase within the network The first-order derivative of the potential phase in the network It is expressed as:

[0171]

[0172] T j ′=T j f(δ e )(5);

[0173]

[0174] Where, T j ′ represents the equivalent inertia constant; D p ′ represents the equivalent damping coefficient; P e ′ represents the equivalent electromagnetic power; f(·) represents the function of (·); δ e Represents the potential phase within the network.

[0175] The reactive power-voltage control equation of the reactive synchronization control loop is as follows:

[0176] E=E N +D q (Q ref -Q e ) (8);

[0177]

[0178] D′ q =D q f(δ e ) (10);

[0179] In the formula, E N Indicates the rated voltage of the power grid; D q Represents reactive power droop coefficient; Q ref Indicates reactive power reference value; Q e Indicates the instantaneous reactive power of the power grid; D q ′ represents the equivalent reactive power droop coefficient.

[0180] Step S2.1: Based on the impact of GFM-ESS access on the static voltage stability of the power system, the pre-site selection principle and the three-level site selection optimization model are used to jointly plan the ESS site selection. With the goals of improving the static voltage stability margin, maximizing the static voltage regulation sensitivity and minimizing the branch power loss, a GFM-ESS system-level site selection planning model is established.

[0181] Step S2.1.1, the indicators of the pre-site selection principle include:

[0182] Voltage sensitivity index, which is an important parameter for ranking the busbar's robustness and power variation;

[0183] The sensitivity between power and bus voltage changes can be expressed by the elements of the Jacobian matrix:

[0184]

[0185] In the formula, It represents the derivative of active power; Represents the derivative of reactive power; J Pθ represents the Jacobian matrix about Pθ; J Qθ represents the Jacobian matrix about Qθ; J PV represents the Jacobian matrix about PV; J QV represents the Jacobian matrix about QV; represents the derivative of the phase angle; Represents the derivative of voltage.

[0186] Assuming that the active power disturbance is zero, the sensitivity of voltage to reactive power is calculated and expressed as:

[0187]

[0188] In the formula, ΔP represents the active power increment; Indicates about The Jacobian matrix of .

[0189] Assuming that the reactive power disturbance is zero, the active power sensitivity is obtained, which is expressed as:

[0190]

[0191] In the formula, ΔQ represents the reactive power increment; Indicates about The Jacobian matrix of .

[0192] Considering both active sensitivity and reactive sensitivity, the voltage sensitivity index VSI calculation formula is as follows:

[0193]

[0194] The voltage sensitivity index (VSI) reflects the static voltage stability of a node and can help determine areas in the grid that need to be strengthened. The higher the VSI value, the weaker the node, and the node becomes a pre-selected target.

[0195] Step S2.1.2: With the goal of improving the static voltage stability margin, maximizing the static voltage regulation sensitivity and minimizing the branch power loss, the specific process of establishing the GFM-ESS system-level site selection planning model is as follows:

[0196] GFM-ESS and IBRs are connected to the power system through PCC. The following analysis illustrates the impact of connecting GFM-ESS on the static voltage stability limit of IBRs system.

[0197] Without GFM-ESS, the power output of IBRs to the grid is as follows (assuming Among them, U p Indicates the voltage at PCC; represents the derivative of the grid voltage; represents the derivative of the voltage at PCC; θ p Indicates the phase of the voltage at PCC):

[0198]

[0199] Where P re Indicates the active power output by IBRs; Q re Represents the reactive power output by IBRs; R g Indicates the grid resistance.

[0200] Ignore R g , eliminating θ in equation (21) p , then the power P at PCC re With voltage U p The relationship is:

[0201]

[0202] After connecting to GFM-ESS, the power output of PCC to the grid becomes P pcc =P re +P es , Q pcc =Q re +Q es , where P pcc represents the active power of PCC when GFM-ESS is connected, P es Indicates the active power output of the grid-connected energy storage, Q pcc represents the reactive power at PCC, Q es Represents the reactive power output of the grid-connected energy storage; therefore, P pcc and Q pcc Substituting P in formula (22) re and Q re , establish the power P of PCC under GFM-ESS access pcc and voltage Up The relationship between.

[0203] It can be inferred from formula (22) that the access of GFM-ESS will change the power-voltage relationship in the IBRs system. According to formula (22), the PV curves before and after the access of GFM-ESS and the different X g The PV curve under the value is as follows Figure 3 shown.

[0204] from Figure 3 It can be seen that as the load increases, the system voltage approaches the saddle bifurcation point or the limit-induced bifurcation (LIB) point. The comparison of the PV curves before and after the GFM-ESS connection shows that the GFM-ESS increases the output active power under the static voltage stability limit and improves the active power margin. In addition, the grid impedance X g The increase of reduces the static stability limit, resulting in a decrease in the active output capacity of the PCC. Therefore, by calculating the static voltage stability margin index after the GFM-ESS is connected to different nodes, the improvement of the static voltage stability of the system can be evaluated.

[0205] Considering that the access of GFM-ESS is equivalent to introducing an additional voltage source into the system, which affects the overall power flow and static characteristics of the system, the site selection indicators at this stage also include static voltage regulation sensitivity and branch power loss.

[0206] The static voltage stability margin improvement index reflects the improvement effect of GFM-ESS access on static voltage stability, which is expressed as:

[0207]

[0208] Where U MAX is the critical bus voltage stability limit of the current power system, U L0 is the voltage value of the current operating point, VSM is the voltage stability margin, VSMI is i is the deviation value of VSM at node i before and after installing GFM-ESS; VSM ess，i Indicates the voltage stability margin after grid-connected energy storage; VSM 0，i Indicates the voltage stability margin before grid-connected energy storage is connected.

[0209] VSMI i The larger it is, the greater the enhancement to the voltage stability margin.

[0210] Static voltage regulation sensitivity index, which indicates the impact of connecting GFM-ESS at a specific node on the voltage stability of the entire system, is expressed as:

[0211]

[0212] Where n is the total number of power system nodes, U iref is the node voltage reference value, U i is the node voltage amplitude, ∑ΔU base VRI is the total offset of the node voltage before GFM-ESS is connected; i Represents the static voltage regulation sensitivity at node i in the power system.

[0213] VRI i The smaller the value, the higher the sensitivity of the node in regulating voltage.

[0214] The branch power loss index indicates the impact of accessing the GFM-ESS at a specific node on the change in the loss of the entire system, expressed as:

[0215]

[0216] Where b is the total number of system branches; R l is the branch resistance, I l is the branch current; l represents the lth branch in the system; U i，i= represents the voltage of the busbars at both ends of the lth branch; Z i，i+1 PLI represents the impedance between the busbars at both ends of the lth branch; i Represents the branch power loss at node i in the power system.

[0217] Indicator PLI i The smaller the value, the smaller the power loss of the entire network.

[0218] The GFM-ESS system-level site selection planning model is established and expressed as:

[0219] Minf1(x)=-VSMI×VRI(26);

[0220] Minf2(x)=PLI(27);

[0221] J ess，k ∈γ see (28);

[0222]

[0223] In the formula, J ess，k is the node connected to unit k in GFM-ESS, Υ see The set of pre-selected sites obtained by sorting the VSIs. and are the upper and lower limits of energy storage installed capacity, and are the upper and lower limits of the energy storage installed power respectively; Minf1(x) represents the minimum value of f1(x); Minf2(x) represents the minimum value of f2(x); f1 represents the function expression of the multiplication of VSMI and VRI; f2 represents the function expression of PLI; x represents the function variable; Indicates the nominal capacity of grid-type energy storage; Indicates the nominal power of grid-type energy storage; Represents the minimum value of the voltage at node i; represents the maximum value of the voltage at node i; Indicates the minimum value of branch b power; P b represents the power of branch b; Indicates the maximum power of branch b.

[0224] Step S2.2: Considering the impact of GFM-ESS, the definition of gSCR is extended, and based on satisfying the stability constraints of GFM-ESS itself, a GFM-ESS site-level site selection planning model is constructed with the goal of maximizing the improvement of small disturbance stability after GFM-ESS access.

[0225] The focus of this step is to calculate the small signal stability of the multi-infeed power electronic system (MIPES) after connecting to the GFM-ESS. First, the closed-loop characteristic equation of MIPES is derived through impedance analysis, and it is decoupled into multiple independent and equivalent single-infeed power electronic systems (SIPES). Then, the stability of the MIPES system is analyzed based on the minimum eigenvalue of SIPES, and the concept of gSCR is introduced to evaluate the small signal stability of MIPES. Secondly, the impact of installing GFM-ESS on the small signal stability of the system is discussed, and the definition of gSCR is expanded to consider the impact of GFM-ESS.

[0226] For MIPES, it can be divided into the device side and the network side. Through impedance analysis, the dynamic equation of the network side can be expressed as:

[0227]

[0228] Where Y Net (s) is the AC grid admittance matrix, B∈R n×n is the Thevenin equivalent admittance matrix, B=1 / (ω0L), α(s)=ω0 2 / (s 2 +ω0 2 ),β(s)=sω0 / (s 2 +ω0 2), ω0 represents the synchronization frequency; α(s), β(s) represent complex functions about s; L represents the inductance of the line; s represents the Laplace operator; ΔI x , ΔI y are matrices composed of the increments of the x-axis and y-axis components of the current injected into node i in the global coordinate system; ΔU x , ΔU y are the matrices composed of the increments of the x-axis and y-axis components of the voltage at node i in the global coordinate system; γ(s) represents the matrix composed of α(s) and β(s).

[0229] The dynamic equation on the equipment side can be expressed as:

[0230]

[0231] In the formula, ΔI xi , ΔI yi They represent the increments of the x-axis and y-axis components of the current injected into node i in the global coordinate system; De_xy (s) represents the admittance transfer function matrix of the device in the x, y global coordinate system; ΔU xi , ΔU yi They represent the increments of the x-axis and y-axis components of the voltage at node i in the global coordinate system; T i Represented by cosθ i 、sinθ i The matrix composed of B Indicates the rated capacity of the equipment; G dd (s), G dq (s), G qd (s), G qq (s) represents the input-output transfer function of the device at the dq axis port, describing the response of the current to the voltage disturbance; Y De_dq (s) represents the admittance transfer function matrix of the device in the dq global coordinate system; θ i represents the phase angle of node i.

[0232] According to the multivariable feedback control theory, the closed-loop characteristic equation of MIPES can be expressed as:

[0233]

[0234] In the formula, I 2n represents the 2n-order identity matrix.

[0235] By performing an equivalent transformation on the closed-loop characteristic equation, MIPES is decoupled into n independent equivalent SIPES, which can be expressed as:

[0236]

[0237] In the formula, In Represents the n-th order identity matrix.

[0238] In formula (37), each term is a diagonal block matrix, which can transform the closed-loop characteristic equation of the system into:

[0239]

[0240] In the formula, λ i Represents the matrix Y eq The i-th eigenvalue of .

[0241] make 0<λ1≤λ2≤λ i , …, ≤λ n , i = 1, ..., n. Equation (38) shows that the closed-loop characteristic equation of MIPES can be regarded as the product of n equivalent closed-loop characteristic equations of SIPES, such as Figure 4 shown.

[0242] like Figure 4 As shown, the small signal stability of the system depends on the matrix Y eq The minimum eigenvalue λ1 of MIPES corresponds to this minimum eigenvalue λ1. This minimum eigenvalue λ1 is defined as gSCR, which can evaluate the small signal stability of MIPES.

[0243] Install capacity S on node n VSG The GFM-ESS is equivalent to adding a parallel susceptance S to the ground branch. VSG B p Therefore, the Thevenin equivalent admittance matrix B is modified as follows:

[0244]

[0245] Wherein, B′ represents the modified B.

[0246] It can be seen that the access of GFM-ESS has no effect on the dynamic characteristics of the system equipment side; it only modifies the admittance matrix on the network side. Considering the influence of GFM-ESS, gSCR is expanded into the extended admittance matrix Y′ eq The minimum eigenvalue of is expressed as:

[0247]

[0248] In the formula, λ + (·) is the positive eigenvalue of the matrix; is the rated capacity matrix after adding grid-connected energy storage. At this time, all new energy sources in MIPES operate in discharge mode.

[0249] Since the installation location of GFM-ESS in the power grid is different, the corresponding extended admittance matrix is ​​also different, which means that the installation of GFM-ESS at different nodes has different effects on the small-signal stability of the power system. The primary goal at the station level is to find the node that provides the best small-signal stability in the 110kV network under the stability boundary of GFM-ESS. Considering that the small-signal stability decreases with the increase of SCR in SIPES using GFM-ESS, a maximum SCR called critical short circuit ratio (CSCR) is required to maintain small-signal stability. When the SCR is lower than the CSCR, the ESS must be controlled in the form of GFM to ensure the stability of operation. CSCR is the same as the critical value of the short-circuit ratio of a single machine. It can be obtained by analytical calculation or single-machine simulation experiment. Further, the small-signal stability of the power system using GFM-ESS is evaluated using gSCR. The higher the gSCR, the better the small-signal stability of the power system. Therefore, CSCR and gSCR can be used as site selection indicators to select nodes where GFM-ESS is better than GFL-ESS.

[0250] CSCR provides a stable boundary for the small signal stability of GFM-ESS, while gSCR evaluates the small signal stability of GFM-ESS. Therefore, under the constraint of CSCR, a GFM-ESS installation site with the maximum gSCR is selected. The GFM-ESS site selection planning model is expressed as:

[0251]

[0252] Step S2.3: Based on the time domain target characteristics of the multi-scale control response characteristics of the GFM-ESS, multiple evaluation indicators of the GFM-ESS are obtained, and a GFM-ESS equipment-level site selection planning model that takes into account the voltage and frequency support capabilities is constructed.

[0253] When a large disturbance occurs at a certain location in the power grid, the distribution of disturbance energy in the power grid is uneven due to factors such as generator inertia, transmission line parameters, and spin reserve. This results in different transient frequency and voltage responses at different nodes. Therefore, each potential installation node has different support requirements for frequency and voltage. Nodes with more severe interference and weaker regulation resources have higher support requirements for frequency and voltage.

[0254] For grid operators, GFM-ESS can be regarded as a “black box” or “grey box” model to evaluate its dynamic support capability by observing the power, voltage and frequency characteristics at multiple time scales at the PCC point. Figure 5The overall outline of the reactive power trajectory under the voltage event is shown, with key features marked, which can be used to evaluate transient voltage stability and dynamic support capability. t0 , Q t0 is the initial voltage and reactive power value, ΔU max is the maximum transient voltage deviation, t0 is the time when the fault occurs, t f is the fault clearing time, t c is the voltage recovery time.

[0255] GFM-ESS adjusts the voltage deviation through reactive power-voltage control. D During the period, the average reactive power compensation provided by GFM-ESS at PCC is defined as the equivalent droop coefficient of GFM-ESS station, expressed as:

[0256]

[0257] In the formula, Indicates that node i is in F k Equivalent voltage droop capability under disturbance; Indicates t0 to t f The average value of U t0 and Q t0 are the initial voltage and reactive power value respectively; U t，i and Q t，i They represent the voltage and reactive power value of node i at time t respectively; F k For major disturbance events, such as generator outages, load shedding, and short circuit faults; is the expected failure set.

[0258] In addition, the transient voltage peak value and the transient voltage recovery degree are defined as indicators for evaluating the transient voltage stability under large disturbances, as shown in (43) and (44):

[0259]

[0260] In the formula, Indicates that node i is in F k Transient voltage peak under disturbance; U max，i Indicates the maximum value of transient voltage; Indicates that node i is in F k The degree of transient voltage recovery under disturbance; t c Indicates the time it takes for the voltage to return to a stable value.

[0261] and The higher the value, the higher the voltage support requirement of node i.

[0262] By calculation and The sensitivity of is used to evaluate the improvement of transient voltage stability after the implementation of GFM-ESS, which is expressed as:

[0263]

[0264] In the formula, Indicates that after adding grid-type energy storage, node i is in F k The degree of improvement in transient voltage stability under disturbance; Indicates that after adding grid-type energy storage, node i is in F k Transient voltage peak under disturbance; Indicates that node i is in F after not adding grid-connected energy storage k Transient voltage peak under disturbance; Indicates that after adding grid-type energy storage, node i is in F k The degree of transient voltage recovery under disturbance; Indicates that node i is in F after not adding grid-connected energy storage k The degree of transient voltage recovery under disturbance; a and b represent and The weight coefficient of .

[0265] The larger it is, the higher the effectiveness of transient voltage support.

[0266] The equivalent inertia index is used to describe the impact of GFM-ESS power injection at node i on the RoCoF of the power system and is expressed as:

[0267]

[0268] In the formula, represents the impact of GFM-ESS power injection at node i on the RoCoF of the power system; RoCoP represents the power change rate; RoCoF represents the frequency change rate; ΔP represents the active power increment; Δf represents the frequency value; Indicates that node i is at t m Active power increment at ; Indicates that node i is at t m The frequency value at ; Indicates the active power increment at the initial moment; Indicates the frequency value at the initial moment.

[0269] The higher the value, the stronger the inertial response capability of GFM-ESS.

[0270] The GFM-ESS damping effectiveness index is defined as:

[0271]

[0272] In the formula, Indicates that node i is in F k Equivalent damping under disturbance; t m represents the time corresponding to the lowest frequency point; t0 represents the initial moment of the fault; f t，i represents the frequency value of node i at time t; f t0 The frequency value at the initial time.

[0273] The smaller the value, the stronger the damping response capability of GFM-ESS.

[0274] The indicators that characterize frequency stability include frequency change rate, transient frequency deviation and dynamic recovery frequency deviation, as shown in equations (48) to (50):

[0275]

[0276] In the formula, Indicates that node i is in F k The rate of change of frequency under disturbance; Indicates that node i is in F k Frequency transient deviation under disturbance; Indicates that node i is in F k Frequency dynamic deviation under disturbance; f max，i Indicates the maximum frequency deviation; f d，i Indicates the frequency dynamic deviation value.

[0277] and The smaller the index value is, the better the frequency stabilization effect is after GFM-ESS is installed on node i.

[0278] The transient frequency stability improvement effect index after connecting to GFM-ESS is shown in formula (51);

[0279]

[0280] In the formula, Indicates that after the grid-connected energy storage is connected, node i is in F k The transient frequency stability improvement effect under disturbance;

[0281] Indicates that after the grid-connected energy storage is connected, node i is in F k The rate of change of frequency under disturbance; Indicates that node i is in F before grid-connected energy storage is connected k The rate of change of frequency under disturbance; Indicates that after the grid-connected energy storage is connected, node i is in F k Frequency transient deviation under disturbance; Indicates that node i is in F before grid-connected energy storage is connectedk Frequency transient deviation under disturbance; Indicates that after the grid-connected energy storage is connected, node i is in F k Frequency dynamic deviation under disturbance; Indicates that node i is in F before grid-connected energy storage is connected k The frequency dynamic deviation under disturbance; c represents the The weight coefficient determined by the importance of .

[0282] The larger it is, the more effective the transient frequency support is.

[0283] The GFM-ESS equipment-level site selection planning model is constructed. The GFM-ESS equipment-level site selection planning model uses the proposed evaluation indicators and numerical calculations to obtain the final site selection results.

[0284] Step S3: Solve the three-level site selection model to obtain the best site selection plan.

[0285] Step S3.1: Use the NSGA-II multi-objective optimization algorithm to solve the GFM-ESS system-level site selection planning model.

[0286] Since the GFM-ESS system-level site selection and planning model involves the node and line characteristics of the entire 220kV power grid, it presents high-order nonlinear characteristics. The NSGA-II multi-objective optimization algorithm can flexibly adapt to large-scale optimization problems and handle the complexity of high-order models through parallel population-based search. It is an ideal solution to the GFM-ESS system-level site selection and planning model. The NSGA-II multi-objective optimization algorithm explores the solution space through crossover and mutation operations, and uses fast non-dominated sorting, elitism, and crowding distance mechanisms to ensure uniform distribution of the Pareto frontier. In the multi-objective optimization process, after obtaining a set of non-dominated solutions, the best decision is made based on the non-dominated solutions. The pseudo-weight method improves the traditional weighted sum method and finds the optimal solution by iteratively adjusting the weights. It can be used as a multi-criteria decision-making method for selection.

[0287] Step S3.2: Use Python to call the power system simulation software to iteratively solve the GFM-ESS site selection planning model.

[0288] First, use Python to write the iterative calculation process algorithm of the GFM-ESS site selection and planning model, and set the algorithm parameters. Then call the DIgSILENT simulation software, and pass the decision variables of the GFM-ESS site selection and planning model to the DIgSILENT simulation software. Establish a case simulation system on the DIgSILENT simulation software, set fault parameters, perform time domain simulation, and return the simulation results to Python. Repeat this iterative process until the maximum number of iterations is reached.

[0289] Step S3.3: Use the improved GRA-VIKOR evaluation algorithm to solve the GFM-ESS equipment-level site selection planning model.

[0290] According to the evaluation index, the numerical results of the evaluation index are input into the improved GRA-VIKOR evaluation algorithm, and the output is the optimal site selection result of the GFM-ESS device-level site selection planning model.

[0291] Firstly, the Bayesian BWM and CRITIC methods are used to determine the basic weights of the evaluation indicators ω i , combined with the principles of game theory, i Integrate it into the improved GRA-VIKOR evaluation algorithm to determine the optimal location for the GFM-ESS equipment-level site selection planning model; the specific process is:

[0292] 1. Use Bayesian BWM to calculate subjective weights.

[0293] First, k experts are invited to select the evaluation index (A={A1,A2,…,A n}) to determine the optimal index A B and the worst indicator A W , A n is the nth evaluation index; using the 1-9 scale method, A B and A W As a reference comparison object, construct the best and worst comparison matrix C B ={a B1 ,a B2 ,…,a Bn}, C W ={a 1W ,a 2W ,…,a nW}, a Bn are the n optimal evaluation indicators determined by experts, a nW are the n worst evaluation indicators determined by experts.

[0294] Second, use the multinomial distribution to distribute the input A B、 A W Modeling, taking the target layer as a random event and the indicator layer weight as the probability of the random event, calculate the input A B、 A W The multinomial probability distribution of is used to obtain the relationship between the weight and the expert score.

[0295] Finally, based on A B、 A W The input and output of the model,by formulating the Bayesian hierarchical model, the subjective weight vector,w1,can be obtained.

[0296] 2. Use the CRITIC method to calculate the objective weight.

[0297] First, the standard deviation is used to reflect the variability of the evaluation index V i , that is, the difference in the values ​​of the same evaluation indicator under different schemes is used to illustrate the evaluation strength of the evaluation indicator;

[0298] Secondly, we need to find the conflict S of the evaluation index i , construct an expression to characterize the conflict of indicators based on the correlation coefficient;

[0299] Finally, calculate the objective weight of the evaluation index and convert V i With S i Multiplying them together, we can get the information amount G of the i-th evaluation index i , G i The larger it is, the higher the amount of information contained in the evaluation index, and the larger the weight value assigned should be, thus obtaining the objective weight vector w2 of the index.

[0300] 3. Use game weighting method to determine the comprehensive weight.

[0301] First, construct the combined weight vector w3:

[0302] w3=A1w1+A2w2(52);

[0303] Where A1 and A2 are the coefficients corresponding to w1 and w2 respectively.

[0304] Secondly, the idea of ​​game theory is introduced, which can be expressed as:

[0305]

[0306] In the formula, A v is the coefficient variable, v = 1, 2; w v A v The weight of Z Represents an index.

[0307] Finally, the optimal combination weight is obtained.

[0308] 4. Determine the optimal ranking.

[0309] Through the GRA idea, the grey correlation between each index value and the positive and negative ideal solutions is obtained, and the final group utility value Q is calculated. i and individual regret value R i , and according to formula (54), the comprehensive evaluation index T of GFM-ESS installed at different sites can be obtained i :

[0310]

[0311] According to T i, the best GFM-ESS installation node can be determined by the following rules; assuming that the best site selection scheme is P1, the second best site selection scheme is P2, and the maximum group utility value Q i Sort different site selection schemes; the site selection scheme is the best solution when it meets the following two conditions:

[0312] Condition (1), T(P1)-T(P2)≥1 / n-1.

[0313] Condition (2), P1 in the group utility value Q i and individual regret R i The one that is ranked first in the list.

[0314] The criteria for the best ratio solution based on the above two conditions are as follows:

[0315] When conditions (1) and (2) are met, P1 is the best site selection option.

[0316] When condition (1) is met, P1 and P2 are the best site selection options.

[0317] When condition (2) is met, any site selection plan that does not meet condition (1) is the best site selection plan.

[0318] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A hierarchical optimization site selection method for grid-type energy storage considering the influence of multi-scale stability, characterized in that: The steps include: Step S1: Designing a hierarchical site selection strategy for GFM-ESS based on different voltage levels; Step S2: Establish a three-level site selection model based on the GFM-ESS hierarchical site selection strategy, including a GFM-ESS system-level site selection planning model, a GFM-ESS station-level site selection planning model, and a GFM-ESS equipment-level site selection planning model; Step S3: Solve the three-level site selection model to obtain the best site selection plan; The GFM-ESS hierarchical site selection strategy includes: First, considering the stable distribution characteristics of static voltage, the effect of voltage reactive power compensation, and the network loss of the power system, identify the nodes in the 220kV transmission network that are weaker than the specified value; Article 2. Based on the result of Article 1, select the node whose GFM-ESS synchronization stability is higher than that of GFL-ESS; Article 3. From the perspective of GFM-ESS enhancing the transient stability of the power system and considering the comprehensive impact on the dynamic frequency and voltage support of the power grid, GFM-ESS is installed at nodes with a voltage level of 35kV.

2. The method for hierarchical optimization of site selection for grid-type energy storage considering the influence of multi-scale stability according to claim 1 is characterized in that: GFM is a virtual synchronous generation VSG, and the control loop of VSG consists of active and reactive synchronization loops and voltage and current control double closed loop structures; The power angle swing equation of the active synchronous control loop is expressed as: Where, T j represents the network inertial time constant; d represents differential; t represents time; P ref Indicates the grid-type active power reference value; P e Indicates the electromagnetic power of the power grid; D p represents the network damping coefficient; ω represents the angular frequency; ω N Indicates the rated angular frequency of the power grid; δ GFM Indicates the grid-type output power angle; Simplify formula (1): In the formula, Represents the second-order derivative of the grid-type output power angle; represents the first-order derivative of the output power angle of the network; E represents the internal potential of the network; E g Indicates the grid voltage; X f represents the internal reactance of the network; X g Indicates the grid reactance; According to the voltage vector relationship, Replaced by the second-order derivative of the potential phase within the network The first-order derivative of the potential phase in the network It is expressed as: T′ j =T j f(δ e ) (5); In the formula, T′ j Denotes the equivalent inertia constant; D′ p represents the equivalent damping coefficient; P′ e represents the equivalent electromagnetic power; f(·) represents the function; δ e Indicates the potential phase within the network; The reactive power-voltage control equation of the reactive synchronization control loop is expressed as: E=E N +D q (Q ref -Q e ) (8); D′ q =D q f(δ e ) (10); In the formula, E N Indicates the rated voltage of the power grid; D q Represents reactive power droop coefficient; Q ref Indicates reactive power reference value; Q e Indicates the instantaneous reactive power of the power grid; D′ q Represents the equivalent reactive power droop coefficient.

3. The method for hierarchical optimization site selection of grid-type energy storage considering the influence of multi-scale stability according to claim 2 is characterized in that: Based on the impact of GFM-ESS access on the static voltage stability of the power system, the pre-site selection principle and the three-level site selection optimization model are used to jointly plan the ESS site selection. With the goals of improving the static voltage stability margin, maximizing the static voltage regulation sensitivity and minimizing the branch power loss, a system-level site selection planning model for GFM-ESS is established.

4. The method for hierarchical optimization site selection of grid-type energy storage considering the influence of multi-scale stability according to claim 3 is characterized in that: The indicators of pre-site selection principles include: Voltage sensitivity index, expressed as: In the formula, It represents the derivative of active power; Represents the derivative of reactive power; J Pθ represents the Jacobian matrix about Pθ; J Qθ represents the Jacobian matrix about Qθ; J PV represents the Jacobian matrix about PV; J QV represents the Jacobian matrix about QV; represents the derivative of the phase angle; represents the derivative of voltage; Assuming that the active power disturbance is zero, the sensitivity of voltage to reactive power is calculated and expressed as: In the formula, ΔP represents the active power increment; Indicates about The Jacobian matrix of ; Assuming that the reactive power disturbance is zero, the active power sensitivity is obtained, which is expressed as: In the formula, ΔQ represents the reactive power increment; Indicates about The Jacobian matrix of ; Considering both active sensitivity and reactive sensitivity, the voltage sensitivity index VSI calculation formula is expressed as:

5. The method for hierarchical optimization site selection of grid-type energy storage considering the influence of multi-scale stability according to claim 4 is characterized in that: With the goal of improving the static voltage stability margin, maximizing the static voltage regulation sensitivity and minimizing the branch power loss, the specific process of establishing the GFM-ESS system-level site selection planning model is as follows: GFM-ESS and IBRs are connected to the power system through PCC; set up Among them, U p Indicates the voltage at PCC; represents the derivative of the grid voltage; represents the derivative of the voltage at PCC; θ p Indicates the phase of the voltage at the PCC. When GFM-ESS is not connected, the power output of IBRs to the grid is expressed as: Where P re Indicates the active power output by IBRs; Q re Represents the reactive power output by IBRs; R g Indicates the grid resistance; Ignore R g , eliminating θ in equation (21) p , then the power P at PCC re With voltage U p The relationship is: After connecting to GFM-ESS, the power output of PCC to the grid becomes P pcc =P re +P es , Q pcc =Q re +Q es , where P pcc represents the active power of PCC when GFM-ESS is connected, P es Indicates the active power output of the grid-connected energy storage, Q pcc represents the reactive power at PCC, Q es Indicates the reactive power output of the grid-type energy storage; P pcc and Q pcc Substituting P in formula (22) re and Q re , establish the power P of PCC under GFM-ESS access pcc and voltage U p the relationship between; The static voltage stability margin improvement index is expressed as: Where U MAX is the critical bus voltage stability limit of the current power system, U L0 is the voltage value of the current operating point, VSM is the voltage stability margin, VSMI is i is the deviation value of VSM at node i before and after installing GFM-ESS; VSM ess，i Indicates the voltage stability margin after grid-connected energy storage; VSM 0，i Indicates the voltage stability margin before grid-connected energy storage is connected; The static voltage regulation sensitivity index is expressed as: Where n is the total number of power system nodes, U iref is the node voltage reference value, U i is the node voltage amplitude, ∑ΔU base VRI is the total offset of the node voltage before GFM-ESS is connected; i represents the static voltage regulation sensitivity at node i in the power system; Branch power loss index, which is expressed as: Where b is the total number of system branches; R l is the branch resistance, I l is the branch current; l represents the lth branch in the system; U i，i+ represents the voltage of the busbars at both ends of the lth branch; Z i，i++ PLI represents the impedance between the busbars at both ends of the lth branch; i represents the branch power loss at node i in the power system; The GFM-ESS system-level site selection planning model is established and expressed as: Min f1(x)=-VSMI×VRI (26); Min f2(x) = PLI (27); J ess，k ∈γ see (28); In the formula, J ess，k is the node connected to unit k in GFM-ESS, γ see The set of pre-selected sites obtained by sorting the VSIs, and are the upper and lower limits of energy storage installed capacity, and are the upper and lower limits of the energy storage installed power respectively; Minf1(x) represents the minimum value of f1(x); Minf2(x) represents the minimum value of f2(x); f1 represents the function expression of the multiplication of VSMI and VRI; f2 represents the function expression of PLI; x represents the function variable; Indicates the nominal capacity of grid-type energy storage; Indicates the nominal power of grid-type energy storage; Represents the minimum value of the voltage at node i; represents the maximum value of the voltage at node i; Indicates the minimum value of branch b power; P b represents the power of branch b; Indicates the maximum power of branch b.

6. The method for hierarchical optimization site selection of grid-type energy storage considering the influence of multi-scale stability according to claim 5 is characterized in that: Considering the impact of GFM-ESS, the definition of gSCR is extended, and based on satisfying the stability constraints of GFM-ESS itself, a GFM-ESS site selection planning model is constructed with the goal of maximizing the improvement of small disturbance stability after GFM-ESS access. Specifically: Calculate the small signal stability of the multi-infeed power electronic system MIPES after connecting to the GFM-ESS; MIPES includes the device side and the network side. The dynamic equation of the network side is expressed as: Where Y Net (s) is the AC grid admittance matrix, B∈R n×n is the Thevenin equivalent admittance matrix, B=1 / (ω0L), α(s)=ω0 2 / (s 2 +ω0 2 ),β(s)=sω0 / (s 2 +ω0 2 ), ω0 represents the synchronization frequency; α(s), β(s) represent complex functions about s; L represents the inductance of the line; s represents the Laplace operator; ΔI x , ΔI y are matrices composed of the increments of the x-axis and y-axis components of the current injected into node i in the global coordinate system; ΔU x , ΔU y are matrices composed of the increments of the x-axis and y-axis components of the voltage at node i in the global coordinate system; γ(s) represents the matrix composed of α(s) and β(s); The dynamic equation on the equipment side is expressed as: In the formula, ΔI xi , ΔI yi They represent the increments of the x-axis and y-axis components of the current injected into node i in the global coordinate system; De_xy (s) represents the admittance transfer function matrix of the device in the x, y global coordinate system; ΔU xi , ΔU yi They represent the increments of the x-axis and y-axis components of the voltage at node i in the global coordinate system respectively; T i Represented by cosθ i 、sinθ i The matrix composed of B Indicates the rated capacity of the equipment; G dd (s), G dq (s), G qd (s), G qq (s) represents the input-output transfer function of the device at the dq axis port; Y De_dq (s) represents the admittance transfer function matrix of the device in the dq global coordinate system; θ i represents the phase angle of node i; According to the multivariable feedback control theory, the closed-loop characteristic equation of MIPES is expressed as: In the formula, I 2n represents the 2n-order identity matrix; By performing an equivalent transformation on the closed-loop characteristic equation, MIPES is decoupled into n independent equivalent SIPES, which can be expressed as: In the formula, I n represents the n-order identity matrix; The closed-loop characteristic equation is equivalently transformed into: In the formula, λ i Represents the matrix Y eq The i-th eigenvalue of ; order 0<λ1≤λ2≤λ i , …, ≤λ n , i = 1, ..., n; Small signal stability depends on the matrix Y eq The SIPES corresponding to the minimum eigenvalue λ1; Define λ1 as gSCR; Install capacity S on node n VSG The GFM-ESS is equivalent to adding a parallel susceptance S to the ground branch. VSG B p ; Because the Thevenin equivalent admittance matrix B is modified to: In the formula, B′ represents the modified B; Expand gSCR to the extended admittance matrix Y′ eq The minimum eigenvalue of is expressed as: In the formula, λ + (·) is the positive eigenvalue of the matrix; It is the rated capacity matrix after adding grid-type energy storage; The GFM-ESS site selection planning model is expressed as:

7. The method for hierarchical optimization site selection of grid-type energy storage considering the influence of multi-scale stability according to claim 6 is characterized in that: Based on the time domain target characteristics of the multi-scale control response characteristics of GFM-ESS, multiple evaluation indicators of GFM-ESS are obtained, and a GFM-ESS equipment-level site selection planning model that takes into account voltage and frequency support capabilities is constructed; specifically: The GFM-ESS is considered as a "black box" or "grey box" model, and the dynamic support capability is evaluated by observing the power, voltage and frequency characteristics at multiple time scales at the PCC point; GFM-ESS adjusts voltage deviation through reactive power-voltage control. D During the period, the average reactive power compensation provided by GFM-ESS at PCC is defined as the equivalent droop coefficient of GFM-ESS station, expressed as: In the formula, Indicates that node i is in F k Equivalent voltage droop capability under disturbance; Indicates t0 to t f The average value of U t0 and Q t0 are the initial voltage and reactive power value respectively; U t，i and Q t，i They represent the voltage and reactive power value of node i at time t respectively; F k For large disturbance events; is the expected failure set; In addition, the transient voltage peak value and the transient voltage recovery degree are defined as indicators for evaluating the transient voltage stability under large disturbances, as shown in equations (43) and (44): In the formula, Indicates that node i is in F k Transient voltage peak under disturbance; U max，i Indicates the maximum value of transient voltage; Indicates that node i is in F k The degree of transient voltage recovery under disturbance; t c Indicates the time corresponding to when the voltage returns to a stable value; By calculation and The sensitivity of is used to evaluate the improvement of transient voltage stability after the implementation of GFM-ESS, which is expressed as: In the formula, Indicates that after adding grid-type energy storage, node i is in F k The degree of improvement in transient voltage stability under disturbance; Indicates that after adding grid-type energy storage, node i is in F k Transient voltage peak under disturbance; Indicates that node i is in F after not adding grid-connected energy storage k Transient voltage peak under disturbance; Indicates that after adding grid-type energy storage, node i is in F k The degree of transient voltage recovery under disturbance; Indicates that node i is in F after not adding grid-connected energy storage k The degree of transient voltage recovery under disturbance; a and b represent and The weight coefficient of The equivalent inertia index is expressed as: In the formula, represents the impact of GFM-ESS power injection at node i on the RoCoF of the power system; RoCoP represents the power change rate; RoCoF represents the frequency change rate; ΔP represents the active power increment; Δf represents the frequency value; Indicates that node i is at t m Active power increment at ; Indicates that node i is at t m The frequency value at ; Indicates the active power increment at the initial moment; Indicates the frequency value at the initial moment; The GFM-ESS damping effectiveness index is defined as: In the formula, Indicates that node i is in F k Equivalent damping under disturbance; t m Indicates the time corresponding to the lowest frequency point; t0 represents the initial time of the fault; f t，i represents the frequency value of node i at time t; f t0 Frequency value at the initial moment; The indicators that characterize frequency stability include frequency change rate, transient frequency deviation and dynamic recovery frequency deviation, as shown in equations (48) to (50): In the formula, Indicates that node i is in F k The rate of change of frequency under disturbance; Indicates that node i is in F k Frequency transient deviation under disturbance; Indicates that node i is in F k Frequency dynamic deviation under disturbance; f max，i Indicates the maximum frequency deviation; f d，i Indicates the frequency dynamic deviation value; The transient frequency stability improvement effect index after connecting to GFM-ESS is shown in formula (51); In the formula, Indicates that after the grid-connected energy storage is connected, node i is in F k The transient frequency stability improvement effect under disturbance; Indicates that after the grid-connected energy storage is connected, node i is in F k The rate of change of frequency under disturbance; Indicates that node i is in F before grid-connected energy storage is connected k The rate of change of frequency under disturbance; Indicates that after the grid-connected energy storage is connected, node i is in F k Frequency transient deviation under disturbance; Indicates that node i is in F before grid-connected energy storage is connected k Frequency transient deviation under disturbance; Indicates that after the grid-connected energy storage is connected, node i is in F k Frequency dynamic deviation under disturbance; Indicates that node i is in F before grid-connected energy storage is connected k The frequency dynamic deviation under disturbance; c represents the ... The weight coefficient determined by the importance of The GFM-ESS equipment-level site selection planning model is constructed. The GFM-ESS equipment-level site selection planning model uses the proposed evaluation indicators and numerical calculations to obtain the final site selection results.

8. The method for hierarchical optimization site selection of grid-type energy storage considering the influence of multi-scale stability according to claim 7 is characterized in that: The specific process of step S3 is: using the NSGA-II multi-objective optimization algorithm to solve the GFM-ESS system-level site selection planning model to obtain the optimal site selection result of the GFM-ESS system-level site selection planning model, calling the power system simulation software to iteratively solve the GFM-ESS station-level site selection planning model to obtain the optimal site selection result of the GFM-ESS station-level site selection planning model; using the improved GRA-VIKOR evaluation algorithm to solve the GFM-ESS equipment-level site selection planning model to obtain the optimal site selection result of the GFM-ESS equipment-level site selection planning model.

9. The method for hierarchical optimization site selection of grid-type energy storage considering the influence of multi-scale stability according to claim 8 is characterized in that: The improved GRA-VIKOR evaluation algorithm is used to solve the GFM-ESS equipment-level site selection planning model. The specific process of obtaining the optimal site selection result of the GFM-ESS equipment-level site selection planning model is as follows: The subjective weights were calculated using Bayesian BWM; First, k experts are invited to select the evaluation index (A={A1,A2,…,A n }) to determine the optimal index A B and the worst indicator A W , A n is the nth evaluation index; using the 1-9 scale method, A B and A W As a reference comparison object, construct the best and worst comparison matrix C B ={a B1 ,a B2 ,…,a Bn }, C W ={a 1W ,a 2W ,…,a nW }, a Bn are the n optimal evaluation indicators determined by experts, a nW The n worst evaluation indicators determined by experts; Second, use the multinomial distribution to distribute the input A B、 A W Modeling, calculation input A B、 A W The multinomial probability distribution of , and the relationship between weight and expert score are obtained; Finally, based on A B、 A W The input and output of the model are obtained by formulating the Bayesian hierarchical model to obtain the subjective weight vector w1; The objective weights were calculated using the CRITIC method; First, the standard deviation is used to reflect the variability of the evaluation index V i ; Secondly, we need to find the conflict S of the evaluation index i , construct an expression to characterize the conflict of indicators based on the correlation coefficient; Finally, calculate the objective weight of the evaluation index and convert V i With S i Multiply to get the information amount G of the i-th evaluation index i , based on G i Get the indicator objective weight vector w2; Use game weighting method to determine the comprehensive weight; First, construct the combined weight vector w3: w3=A1w1+A2w2(52); In the formula, A1 and A2 are the coefficients corresponding to w1 and w2 respectively; Secondly, the idea of ​​game theory is introduced, which can be expressed as: In the formula, A v is the coefficient variable, v = 1, 2; w v A v The weight of Z Represents an index; Finally, the optimal combination weight is obtained; Determine the optimal ranking; Through the GRA idea, the grey correlation between each index value and the positive and negative ideal solutions is obtained, and the final group utility value Q is calculated. i and individual regret value R i , and according to formula (54), the comprehensive evaluation index T of GFM-ESS installed at different sites is obtained i : According to T i , determine the best GFM-ESS installation node: let the best site selection scheme be P1, the second best site selection scheme be P2, and according to the maximum group utility value Q i Sort different site selection plans; when the site selection plan meets the preset conditions, it is the best solution.

10. The method for hierarchical optimization site selection of grid-type energy storage considering the influence of multi-scale stability according to claim 9, characterized in that: The default conditions are: Condition (1), T(P1)-T(P2)≥1 / n-1; Condition (2), P1 in the group utility value Q i and individual regret R i The first in the ranking; The standard expression of the best proportion solution is: When conditions (1) and (2) are met, P1 is the best site selection solution; When condition (1) is met, P1 and P2 are the best site selection options; When condition (2) is met, any site selection plan that does not meet condition (1) is the best site selection plan.

Images