A reactive power and voltage optimization method for renewable energy grid-connected power based on second-order cone

Abstract

The present invention proposes a reactive voltage optimization method for grid-connected renewable energy based on second-order cones. First, the mathematical models of static VAR compensator SVC and doubly-fed asynchronous wind turbine DFIG are introduced. Then, the traditional branch power flow model is subjected to phase angle relaxation and second-order cone relaxation to be converted into a second-order cone form. Then, the node voltage offset minimization and branch active network loss minimization are taken as the comprehensive goals, and the power balance, node voltage upper and lower limits, grouped capacitor capacity, SVC constraints and DFIG constraints are taken as constraints to obtain the optimal result that satisfies the constraints. Finally, the reactive output of the wind turbine, the number of parallel capacitor groups switched on and off, and the reactive output of the SVC are obtained. This strategy smoothes the equipment output and reduces the network loss while ensuring the voltage qualification rate, thereby improving the economy of voltage control.

Description

Technical Field

[0001] The present invention relates to the field of automatic voltage control, and in particular to a strategy study of participating in coordinated control of reactive power and voltage of a power grid after new energy sources are incorporated into the power grid. Background Art

[0002] In recent years, with the widespread use of coal, oil and natural gas, the energy crisis has become increasingly serious, and various fossil energy sources are an important support for the development of the power industry. In order to solve this problem, more and more new energy sources will be connected to the grid to replace traditional energy sources. According to statistics from the National Energy Administration, by the end of 2020, my country's installed capacity of renewable energy power generation reached 934 million kilowatts, an increase of about 17.5% over the previous year.

[0003] The distribution network is an important platform for carrying renewable energy. However, the access of various distributed energy sources (DG) will change the operation structure of the distribution network to a certain extent, thereby affecting the voltage operation level and network loss of the network. At the same time, due to the fluctuation and uncertainty of new energy, the voltage at the grid connection point will also be affected to a certain extent. The most common new energy clusters are photovoltaic power plants and wind power plants. The output of photovoltaic power plants is greatly affected by sunshine. The power output fluctuates with the irradiation of sunlight throughout the day, and even greater voltage fluctuations will be generated under the influence of extreme weather. The same is true for wind power generation. The variability of wind resources and the turnover characteristics of wind turbines themselves will make the output power of wind farms unstable, causing various voltage deviations and voltage flicker problems, which will directly affect the power quality of the power grid.

[0004] As the scale of renewable energy power generation entering the power grid continues to grow, the power balance and voltage stability of the power grid will be affected to varying degrees.

[0005] First, the voltage level of the power system is closely related to the distribution of reactive power. For example, if the reactive power is insufficient, the voltage will drop and the network loss will increase. When the reactive power is excessive, the voltage will rise. If the voltage fluctuation exceeds the upper and lower limits of the constraints, it will cause various stability problems. If you want to ensure the safety and economy of a high proportion of new energy access, you must optimize the reactive power. The double-fed units in the wind farm can work together with the reactive power regulation equipment in the power grid, participate in reactive power assistance, reduce the flow of reactive power, and maintain voltage stability and reduce network losses to a certain extent.

[0006] Secondly, a large number of renewable energy sources are connected to the distribution network, which increases the pressure on the operation and maintenance of the distribution network. The traditional distribution network is transformed into an intelligent active distribution network ADN (Active Distribution Network). The active distribution network adopts more effective control methods and measures to actively control and manage the distribution network, increase the proportion of distributed generation, reduce operating costs, and make the allocation of grid resources more reasonable. As one of the important tools for analyzing the control effect and economic benefits of the distribution network, its optimal power flow currently faces the following problems. Compared with the transmission network, the characteristics of the distribution network such as radiation and high impedance ratio make it difficult to solve the optimal power flow of the deterministic distribution network through more mature linearization methods such as DC power flow. After connecting distributed generation and many other active control devices to the distribution network, the complexity of the optimal power flow calculation of the distribution network will increase. Therefore, it is very necessary to find a more efficient and accurate calculation method to realize the optimal power flow calculation of the active distribution network.

[0007] In summary, the present invention hopes to study the voltage stability and active network loss problems after a high proportion of renewable energy is connected to the power grid, explore the calculation method of the optimal power flow of the active distribution network, and discuss the strategy of the power grid's own reactive power regulation equipment and the regulation equipment connected to the new energy cluster to cooperate in the optimization of the reactive power and voltage regulation of the power grid, so as to solve practical problems. Summary of the invention

[0008] The present invention introduces second-order cone optimization into the new energy power grid for rapid voltage control, with the aim of achieving control over power grid voltage deviation and network loss, improving the overall voltage safety and stability of the power system with a high proportion of new energy integrated while improving economic benefits, solving the voltage problem faced by my country's power grid and solving the control problem.

[0009] The present invention is achieved through the following technical solutions:

[0010] A reactive voltage optimization method for renewable energy grid-connected power based on second-order cone, characterized by comprising:

[0011] Step 1: Collect the distribution network topology and line parameters and the voltage of each node, and check whether the voltage of each node exceeds the limit. If it exceeds the limit, execute step 2, otherwise enter the next collection cycle and repeat this step;

[0012] Step 2: Perform a second-order cone deformation on the branch power flow model, and then use the second-order cone optimization method to calculate the optimal power flow with the minimum voltage offset and the minimum active network loss as the comprehensive goal to obtain the optimization result.

[0013] Step 3: According to the optimization results, the reactive power output of the wind turbine and the reactive power output of the SVC are obtained.

[0014] In the above-mentioned reactive voltage optimization method for renewable energy grid connection based on second-order cone, in step 2, the optimization is based on the objective function and constraints:

[0015] Objective function: The comprehensive goal is to achieve the best voltage stability level at the wind turbine terminal, the smallest node voltage offset, and the smallest network active power loss.

[0016]

[0017]

[0018] The constraints are:

[0019]

[0020]

[0021] P j =P Gj -P Dj

[0022] Q j =Q Gj +Q SVC +Q DFIG -Q Dj

[0023] V J =V I -2(P ij r ij +Q ij x ij )+I IJ (r ij 2 +x ij 2 )

[0024] V imin ≤V i ≤V imax

[0025] I min ≤I≤I max

[0026] Q SVCmin ≤Q SVC ≤Q SVCmax

[0027] Q DFIGmin ≤Q DFIG ≤Q DFIGmax

[0028] V imin and V ima x represents the upper and lower limits of the node voltage, Vi is the actual voltage, v ref is the node reference voltage. For the convenience of calculation, the per-unit value is taken here, which is 1. N is the number of nodes that need to control the voltage.

[0029] U(i) is the set of nodes where node j is connected to node i. ij is the impedance of line ij, I ij is the current flowing through line ij, and n is the number of branches. ij is the phase angle difference between node i and node j; G ij and B ij is the conductance and susceptance of the branch at nodes i and j; P Gi and Q Gi is the active and reactive power generation of node i; P Di is the load active power of node i; Q Di represents the reactive power of the load at node i; Q SVC and Q DFIG are the reactive power injected into the grid by SVC and DFIG respectively. i Inject active power, Q i Inject reactive power into the node, I min and I max They represent the upper and lower limits of the branch current respectively, and I is the actual operating current. DFIGmin and Q DFIGmax They are the upper and lower limits of the wind turbine’s reactive power output. SVCmin and Q SVCmax They are the upper and lower limits of SVC reactive output respectively.

[0030] In the above-mentioned reactive voltage optimization method for renewable energy grid connection based on second-order cone, step 2 specifically includes:

[0031] Step 2.1, determine the mathematical model of the reactive power regulation equipment;

[0032] Step 2.2, taper the power flow model to the second order;

[0033] Step 2.3: Call the commercial solver CPLEX to solve the reactive power optimization problem.

[0034] In the above-mentioned reactive power voltage optimization method based on second-order cone for renewable energy grid connection, the power flow model is based on the following definition: ij is the current of branch ij, r ij +jx ij is the line impedance, P ij +jQ ij is the apparent power of line ij at node i, P jk +jQ jkis the apparent power of the line at node j, P i and Q i Inject active power and reactive power into node i. V i is the voltage at node i. The line voltage drop and current satisfy Ohm's law:

[0035] V i -V j =z ij I ij

[0036] Branch flow:

[0037]

[0038] The power balance equation is:

[0039]

[0040] Among them, k: j→k represents the set of terminal nodes of the branch with node j as the head node, and i: i→j represents the set of head nodes of the branch with power direction from i to j, that is, with j as the terminal node.

[0041] 5 In the above-mentioned reactive power voltage optimization method for renewable energy grid connection based on second-order cone, the power flow model is rewritten into the second-order cone form.

[0042] The power balance equation is known to be

[0043]

[0044] Among them, k: j→k represents the set of terminal nodes of the branch with node j as the head node, and i: i→j represents the set of head nodes of the branch with power direction from i to j, that is, with j as the terminal node. ij is the current in branch ij, z ij is the branch impedance. S ij Inject power into the node.

[0045] You can use S ij Split into P j +Q j , and substitute it into the above formula again to get the following formula.

[0046] After splitting, we can get:

[0047]

[0048]

[0049] because

[0050]

[0051] V J is the node voltage. ij +jx ij is the line impedance, P ij +jQ ij is the apparent power of line ij at node i, P jk +jQ jk is the apparent power of the line at node j, P i and Q i Inject active power and reactive power into node i.

[0052] By multiplying both sides by conjugate, we can get

[0053] V J =V I -2(P ij r ij +Q ij x ij )+I IJ (r ij 2 +x ij 2 )

[0054] Among them, V J =|V j | 2 , V I =|V i | 2 , I IJ =|I ij | 2

[0055] because

[0056]

[0057] Right now

[0058] P ij 2 +Q ij 2 =V I I IJ

[0059] So the current formula is

[0060]

[0061] Performing a second-order cone relaxation on the current formula, we can obtain

[0062]

[0063] Right now

[0064]

[0065] The present invention has the following advantages: The radial distribution network active-reactive coordination dynamic optimization model based on three-phase Distflow flow is essentially a mixed integer non-convex nonlinear programming problem, and this type of model lacks a strict and efficient solution method. To this end, the three-phase power flow equation is convexly relaxed by using the second-order cone relaxation technology, so that the optimization problem is converted into a mixed integer second-order cone programming model that can be effectively solved. This strategy smoothes the equipment output and reduces the network loss while ensuring the voltage qualification rate, thereby improving the economy of voltage control. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Attached Figure 1 It is a reactive voltage regulation method process for new energy grid-connected based on second-order cone.

[0067] Attached Figure 2 It is the basic structure of static VAR compensator SVC.

[0068] Attached Figure 3 It is the basic structure of the double-fed asynchronous wind turbine generator DFIG.

[0069] Attached Figure 4 It is the steady-state equivalent circuit of DFIG.

[0070] Attached Figure 5 It is a branch power flow model.

[0071] Attached Figure 6 It is a schematic diagram of the second-order cone relaxation process. DETAILED DESCRIPTION

[0072] The technical solution of the present invention is further specifically described below through embodiments and in conjunction with the accompanying drawings.

[0073] Example:

[0074] This embodiment specifically includes the following steps:

[0075] Step 1: Analyze the mathematical model of traditional reactive power regulation equipment.

[0076] 1. Static VAR Compensator SVC

[0077] Static VAR generator (SVC) introduces power electronic components into static VAR compensation, thereby achieving fast, continuous and smooth regulation of reactive power. There are many types of SVC structures, but the basic components are thyristor-controlled reactors (TCR) and thyristor-switched capacitors (TSC).

[0078] TCR is composed of a fixed inductor and a bidirectionally conductive thyristor in series. Since the thyristor is turned off at around 3-10KV and 3-6KA, multiple thyristors are often connected in series in actual applications to meet the capacity and voltage requirements.

[0079] TSC is composed of capacitors and bidirectional thyristors connected in series. It has two working modes, namely, thyristors are turned on, capacitors are put into operation, and capacitive reactive power is generated, or thyristors are turned off and capacitors are removed. Therefore, TSC has excellent reactive power compensation characteristics such as fast response speed and smooth regulation.

[0080] When all TCR branches are disconnected and TSC branches are enabled, the SVC generates the maximum capacitive reactive power Q cmax When all TCR branches are switched on and TSC branches are disconnected, the SVC generates the maximum inductive reactive power Q lmax .

[0081] The power output by TSC is adjustable in steps, while TCR is continuously adjustable, so SVC can theoretically combine the two to achieve smooth regulation. The reactive power output is:

[0082]

[0083] Where U is the effective value of bus voltage, and n is the number of groups of TSC input. When 0≤β≤π, SVC can output continuous reactive power.

[0084] 2. Doubly-fed wind turbine DFIG

[0085] In a doubly-fed wind turbine, the stator is directly connected to the grid, and the rotor is connected to the grid through an AC-DC-AC converter. The AC-DC-AC converter consists of two back-to-back connected voltage-type PWM converters: the one close to the rotor is called the rotor-side converter, and the one close to the grid is called the grid-side converter.

[0086] P mec is the mechanical power input to the wind turbine, P s , Q s is the active and reactive power generated by the stator, P c , Q c is the active and reactive power input from the grid to the grid-side converter, P g , Q g It refers to the active and reactive power flowing into the grid from the wind turbine.

[0087] Ignoring the losses in the stator and rotor windings,

[0088] P mec =P s -P r

[0089] Ps =P mec / (1-s)

[0090] P r =sP mec / (1-s)

[0091] Ignoring the switching loss and line loss of the power converter, according to the principle of energy conservation, we can get:

[0092] P c =P r

[0093] Ignoring system losses, according to the power flow direction specified in the figure:

[0094] P g =P s -P c =P mec

[0095] The reactive power of the doubly-fed induction generator is:

[0096] Q g =Q s -Q c

[0097] The operating range of reactive power on the stator side is usually constrained by two aspects: the rotor-side converter current and the stator current.

[0098] Analyzing the equivalent circuit, we can get:

[0099] U s =-(R s +jX s )I s -jX m I r

[0100] U r =-(R r +jsX r )I r -jsX m I s

[0101] U s is the stator side voltage, U r is the rotor side voltage, I s is the stator side current, I r is the rotor side current, R s , X s is the equivalent resistance and reactance on the stator side, R r , X r is the equivalent resistance and reactance on the rotor side, X m is the magnetizing reactance.

[0102]

[0103]

[0104] Set the maximum allowable current of the sub-side to I smax , then the reactive power range is:

[0105]

[0106]

[0107]

[0108]

[0109] Where σ is and U s Angle.

[0110] According to the above formula, we can get

[0111]

[0112] Assume that the maximum allowable rotor current is I rmax , which is generally 1.2 times the rated current of the rotor. Then we can get P on the stator side s When determined, the reactive power range on the stator side is:

[0113]

[0114]

[0115] Due to the existence of a DC link in the bidirectional converter, there is only active power exchange between the grid-side converter and the rotor-side converter, and there is no exchange of reactive power. The active and reactive powers are decoupled from each other. Since the converter that provides external power to the rotor generally operates at a unit power factor, the capacity is set according to the maximum slip power. However, the wind speed changes continuously during actual operation, so the converter power is not always at the maximum slip power. When the wind speed is relatively low, the grid-side converter can operate in a non-power factor mode, providing or absorbing some reactive power to the system. However, this part of the power is relatively small compared to the stator-side input and can be appropriately ignored and not included in the final result. That is, it is considered that:

[0116] Q c =0

[0117] Q g =Q s

[0118] Pg =P s -P c

[0119] P c =P r

[0120] Therefore, the active power and reactive power injected into the grid by the doubly-fed wind turbine can be obtained.

[0121] Q smin =max=(Q s1min , Q s2min )

[0122] Q smax =min(Q s1max , Q s2max )

[0123] Thus, the extreme value of reactive power output of DFIG can be obtained.

[0124] Step 2: Establish an optimal power flow model based on second-order cone programming

[0125] First, let’s introduce the definition of a second-order cone. In Euclidean space, for any pair of points in a set, every point on the line segment connecting the pair of points is also in the set. Such a set is called a convex set.

[0126] Let S∈R n , x0∈S, if for all x∈S and λ>0, there is x0+λx∈S, then S is called a cone with x0 as the vertex. If S is also a convex set, then S is called a convex cone.

[0127] The conditions for an n-dimensional canonical cone are:

[0128] S={(x,t)|||x||2≤t,x∈R n , t∈R}

[0129] This x is the 2-norm.

[0130] After a finite-dimensional linear space is given an inner product, the linear space has a norm and an angle and becomes a Euclidean space.

[0131] If S∈R n is a non-empty convex set, f: for S→R, if If f(λx1+(1-λ)x2)≤λf(x1)+(1-λ)f(x2), then the function f is called a convex function on S.

[0132] If S∈R n is a convex set, and the function f is a convex function on S, then the planning problem minx∈ D f(x) is a convex program.

[0133] The local optimal value of convex programming is the global optimal value. When the objective function of convex programming is a strictly convex function, if there is an optimal solution, this solution will also be the only global optimal solution.

[0134] Second-order cone programming is derived from linear programming. It has been widely used in various fields and the corresponding solution technology has become very mature. Second-order cone programming is a special type of convex programming. Therefore, the result of solving the second-order cone optimization will theoretically be a unique global optimal solution.

[0135] The standard form of second-order cone programming is:

[0136] Min{C T x|Ax=b,x i ∈S, i=1, 2, ... N}

[0137] In the formula, the variable x∈R N , b∈R M , c∈R N , A M×N ∈R M×N , S is a second-order cone or a rotated second-order cone.

[0138]

[0139] The specific principles of establishing the optimal power flow model based on second-order cone programming are introduced below.

[0140] I ij is the current of branch ij, r ij +jx ij is the line impedance, P ij +jQ ij is the apparent power of line ij at node i, P jk +jQ jk is the apparent power of the line at node j, P i and Q i Inject active power and reactive power into node i. V i is the voltage at node i. The line voltage drop and current satisfy Ohm's law:

[0141] V i -V j =z ij I ij

[0142] Branch flow:

[0143]

[0144] The power balance equation is:

[0145]

[0146] Among them, k: j→k represents the set of terminal nodes of the branch with node j as the head node, i: i→j represents the set of head nodes of the branch with power direction from i to j, that is, with j as the terminal node.

[0147] To perform second-order cone planning, we must first rewrite the traditional branch-based power flow model into a second-order cone form.

[0148] The power balance equation is known to be

[0149]

[0150] Among them, k:j→k represents the set of terminal nodes of the branch with node j as the head node, i:i→j represents the set of head nodes of the branch with power direction from i to j, that is, with j as the terminal node. ij is the current in branch ij, z ij is the branch impedance. S ij Inject power into the node.

[0151] You can use S ij Split into P j +Q j , and substitute it into the above formula again to get the following formula.

[0152] After splitting, we can get:

[0153]

[0154]

[0155] because

[0156]

[0157] V J is the node voltage. ij +jX ij is the line impedance, P ij +jQ ij is the apparent power of line ij at node i, P jk +jQ jk is the apparent power of the line at node j, P i and Q i Inject active power and reactive power into node i.

[0158] By multiplying both sides by conjugate, we can get

[0159] V J =VI -2(P ij r ij +Q ij x ij )+I IJ (r ij 2 +x ij 2 )

[0160] Among them, V J =|V j | 2 , V I =|V i | 2 , I IJ =|I ij | 2

[0161] because

[0162]

[0163] Right now

[0164] P ij 2 +Q ij 2 =V I I IJ

[0165] So the current formula is

[0166]

[0167] Performing a second-order cone relaxation on the current formula, we can obtain

[0168]

[0169] Right now

[0170]

[0171] Through the above steps, we transformed the original planning model into a second-order cone programming model.

[0172] Step 3: Establish the reactive power optimization model with objective function and constraints.

[0173] Objective function:

[0174]

[0175]

[0176] Constraints:

[0177]

[0178]

[0179] P j =P Gj -P Dj

[0180] Q j =Q Gj +Q SVC +Q DFIG -Q Dj

[0181] V J =V I -2(P ij r ij +Q ij x ij )+I IJ (r ij 2 +x ij 2 )

[0182] V imin ≤V i ≤V imax

[0183] I min ≤I≤I max

[0184] Q SVCmin ≤Q SVC ≤Q SVCmax

[0185] Q DFIGmin ≤Q DFIG ≤Q DFIGmax

[0186] The upper and lower limits of the node voltage are set to 1.05 and 0.95.

[0187] The specific embodiments described herein are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in similar ways, but they will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

Claims

1. A reactive voltage optimization method for renewable energy grid-connected power based on second-order cone, characterized in that: include: Step 1: Collect the distribution network topology and line parameters and the voltage of each node, and check whether the voltage of each node exceeds the limit. If it exceeds the limit, execute step 2, otherwise enter the next collection cycle and repeat this step; Step 2: The branch power flow model is transformed into a second-order cone, and then the optimal power flow is calculated using the second-order cone optimization method with the minimum voltage offset and the minimum active network loss as the comprehensive goal to obtain the optimization result; Step 3: According to the optimization results, the reactive power output of the wind turbine and the reactive power output of the SVC are obtained; In step 2, the optimization is based on the objective function and constraints: Objective function: The comprehensive goal is to achieve the best voltage stability at the wind turbine terminal, the smallest node voltage offset, and the smallest network active power loss; The constraints are: P j =P Gj -P Dj Q j =Q Gj +Q SVC +Q DFIG -Q Dj V J =V I -2(P ij r ij +Q ij x ij )+I IJ (r ij 2 +x ij 2 ) In imin ≤V i ≤V imax I min ≤I≤I max Q SVCmin ≤Q SVC ≤Q SVCmax Q DFIGmin ≤Q DFIG ≤Q DFIGmax V imin and V imax Represent the upper and lower limits of the node voltage, V i is the actual voltage, v ref is the node reference voltage. For the convenience of calculation, the per-unit value is taken here, which is 1. N is the number of nodes that need to control the voltage. U(i) is the set of nodes where node j is connected to node i; ij is the impedance of line ij, I ij is the current flowing through line ij, n is the number of branches; θ ij is the phase angle difference between node i and node j; G ij and B ij is the conductance and susceptance of the branch at nodes i and j; P Gi and Q Gi is the active and reactive power generation of node i; P Di is the load active power of node i; Q Di represents the reactive power of the load at node i; Q SVC and Q DFIG are the reactive power injected into the grid by SVC and DFIG respectively; P i Inject active power, Q i Inject reactive power into the node, I min and I max They represent the upper and lower limits of the branch current respectively, I is the actual operating current; Q DFIGmin and Q DFIGmax They are the upper and lower limits of the wind turbine’s reactive power output; Q SVCmin and Q SVCmax They are the upper and lower limits of SVC reactive output respectively; Step 2 specifically includes: Step 2.1, determine the mathematical model of the reactive power regulation equipment; Step 2.2, taper the power flow model to the second order; Step 2.3, call the commercial solver CPLEX to solve the reactive power optimization problem; The power flow model is rewritten in second-order cone form; The power balance equation is known to be Among them, K: j→k represents the set of terminal nodes of the branch with node j as the head node, i: i→j represents the set of head nodes of the branch with power direction from i to j, that is, j as the terminal node; I ij is the current in branch ij, z ij is the branch impedance; S ij Inject power into the node; You can use S ij Split into P j +Q j , and then substitute it into the above formula to get the following formula; After splitting, we can get: because V J is the node voltage; r ij +jx ij is the impedance of the line, P ij +jQ ij is the apparent power of line ij at node i, P jk +jQ jk is the apparent power of the line at node j, P i and Q i Inject active power and reactive power into node i; By multiplying both sides by conjugate, we can get V J =V I -2(P ij r ij +Q ij x ij )+I IJ (r ij 2 +x ij 2 ) Among them, V J =|V j | 2 , V I =|V i | 2 , I IJ =|I ij | 2 because Right now P ij 2 +Q ij 2 =V I I IJ So the current formula is Performing a second-order cone relaxation on the current formula, we can obtain Right now 2. The method for optimizing reactive voltage of new energy grid-connected based on second-order cone according to claim 1 is characterized in that: The power flow model is based on the following definitions: ij is the current of branch ij, r ij +jx ij is the impedance of the line, P ij +jQ ij is the apparent power of line ij at node i, P jk +jQ jk is the apparent power of the line at node j, P i and Q i Inject active power and reactive power into node i; V i is the voltage at node i; the line voltage drop and current satisfy Ohm's law: V i -V j =z ij I ij Branch flow: The power balance equation is: Among them, k: j→k represents the set of terminal nodes of the branch with node j as the head node, and i: i→j represents the set of head nodes of the branch with power direction from i to j, that is, the set of head nodes of the branch with j as the terminal node.

Images