Active power distribution network voltage sag inversion method based on two-dimensional feature joint solution

Abstract

The application discloses a voltage sag inversion method of an active power distribution network based on two-dimensional feature joint solving, and comprises the following steps: S1, establishing a voltage-controlled current source equivalent model of an inverter-type distributed power supply based on a low-voltage ride-through characteristic; S2, iteratively calculating a fault current and a voltage sag amplitude of the power distribution network according to a fault network superposition principle; S3, solving a voltage sag duration according to the fault current and the voltage sag amplitude obtained in S2, and combining an action characteristic of a protection device, so as to realize two-dimensional feature joint solving of a voltage sag event; and S4, realizing multi-stage voltage sag inversion based on the two-dimensional feature of the voltage sag event and combining a different inverter-type distributed power supply off-grid timing. The application proposes a voltage sag amplitude-duration two-dimensional feature joint solving and multi-stage voltage sag inversion method, improves voltage sag evaluation accuracy of the power distribution network, and lays a necessary model foundation for subsequent event tracing, diagnosis and treatment.

Description

Technical Field

[0001] This invention relates to the field of power quality technology, and in particular to a method for inverting voltage sags in active distribution networks based on joint solution of two-dimensional features. Background Technology

[0002] Voltage sag is essentially the voltage drop caused by a sudden large current being drawn from the power grid and flowing through the system impedance. Voltage sag inversion involves calculating the two-dimensional characteristics of the power grid at different node buses using short-circuit current calculations, specifically the voltage sag amplitude and duration. Currently, various methods are used for inverting voltage sag amplitude characteristics, such as the critical distance method, fault point method, analytical method, virtual node method, and Monte Carlo method. The critical distance method predicts the voltage sag amplitude based on a voltage distributor model; the fault point method typically selects a certain number of potential fault points in the system, performs simulations and short-circuit calculations on these fault points under various fault conditions, and then evaluates the system's voltage sag level; the analytical method obtains the evaluation result by calculating the residual voltage of node faults; the virtual node method obtains the sag characteristic of a fixed PCC node by treating all fault points as newly added virtual nodes in the system; the Monte Carlo method can utilize a small amount of system component fault probability information to calculate system voltage sag data for various fault types and multiple protection device parameters, thereby obtaining a probabilistic index of the system voltage sag. However, although the existing methods mentioned above can calculate the voltage sag amplitude characteristics, they all ignore the calculation of the duration characteristics, resulting in the voltage sag inversion only obtaining amplitude information, causing a lack of information in the inversion results.

[0003] With the large-scale, multi-point integration of distributed generation (DG) into distribution networks, the distribution characteristics of fault currents in distribution networks are altered by the influence of DG units. This further affects the voltage amplitude and duration characteristics perceived by different nodes in the distribution network. Existing studies on voltage amplitude characteristics often treat the fault equivalent model of the power source as a constant current source, neglecting the integration of DG and the low voltage ride-through (LVRT) requirements of grid-connected inverters. Furthermore, existing voltage sag assessment methods often only study the sag amplitude, assuming that the sag duration follows a probability distribution, ignoring the influence of short-circuit current on the duration, which is correlated with the sag amplitude.

[0004] In summary, existing methods often treat the two-dimensional characteristics of voltage sags as independent, lacking joint calculation of them, which leads to room for improvement in the accuracy of current voltage sag inversion results. Furthermore, existing feature characterization methods are only applicable to "single-stage" voltage sag events. With the large-scale integration of distributed generation into active distribution networks, the differences in their low-voltage ride-through characteristics can lead to the disconnection of some units during voltage sags, causing changes in fault current and voltage distribution, ultimately resulting in the successive disconnection of multiple units and multi-stage voltage sag events. However, there is currently a lack of systematic feature calculation methods for this type of voltage sag event. Summary of the Invention

[0005] To address the issues of existing voltage sag inversion methods not jointly solving for voltage amplitude and duration, and being applicable only to single-stage voltage sag events, resulting in low accuracy, this invention proposes an active distribution network voltage sag inversion method based on joint solution of two-dimensional features. This method solves the aforementioned problems through joint solution of two-dimensional features of voltage sag amplitude and duration, and a multi-stage voltage sag inversion method.

[0006] This application discloses a method for inverting voltage sag in active distribution networks based on joint solution of two-dimensional features, including the following steps:

[0007] S1. Establish an equivalent model of the voltage-controlled current source of the inverter-type distributed power source based on the low voltage ride-through characteristic;

[0008] S2. The fault current and voltage sag of the distribution network are obtained by iterative calculation based on the principle of fault network superposition.

[0009] S3. Based on the fault current and voltage sag amplitude obtained in S2, and combined with the operating characteristics of the protection device, the duration of the voltage sag is solved to achieve the joint solution of the two-dimensional characteristics of the sag event.

[0010] S4. Based on the two-dimensional characteristics of voltage sag events and combined with the grid disconnection timing of different inverter-type distributed power sources, multi-stage voltage sag inversion is achieved.

[0011] Preferably, the low voltage ride-through characteristic satisfies:

[0012]

[0013] All quantities in the formula are per-unit values. This is the positive-sequence active current. It is the positive sequence reactive current. K represents the positive sequence voltage amplitude of the power grid after the voltage drop. q I is the reactive power compensation coefficient. max i represents the maximum fault current value of the inverter-type distributed power converter. d0 This represents the current value before a fault in an inverter-type distributed power source.

[0014] Preferably, step S2 includes the following steps:

[0015] Utilizing the principle of fault analysis superposition, the fault network of a multi-inverter distributed power source is decomposed into normal operation and fault component networks, where only short-circuit node current injection occurs. Each inverter-type distributed power source injects short-circuit current into the normal network as a voltage-controlled current source, influencing the short-circuit node current through the normal network voltage. By iteratively calculating the initial current, normal positive and negative sequence network node voltages, short-circuit composite sequence network voltages, fault point current, fault component node voltage, and power reference value, a short-circuit current value meeting accuracy requirements can be obtained, from which a high-precision sag value can be calculated.

[0016] S21. Let the number of iterations be k, the initial number of iterations be k = 0, set the convergence threshold ε, and initialize the injection current of the inverter-type distributed power source. This is the current value before the fault;

[0017] S22. Based on the symmetric component method, the unbalanced system is transformed into a positive-sequence network, a negative-sequence network, and a zero-sequence network. The nodal admittance matrix Y is constructed, and the nodal voltage equations of the positive-sequence network sequence can be expressed as:

[0018] YU′=I; (4)

[0019] Among them, U′=[U′ 1 ,U′ 2 ,U′ 0 ] T Represents the sequence voltage vector in a normal network, I = [I 1 ,I 2 ,I 0 The sequence current vector, where superscript 1 indicates a positive sequence, superscript 2 indicates a negative sequence, and superscript 0 indicates a zero sequence;

[0020] S23. Calculate the sequence voltage vector U′ in the normal network using the triangular decomposition method;

[0021] S24. Calculate the sequence current I at the fault point. f :

[0022]

[0023] Among them, Z f Here is the sequence impedance matrix at the fault point, E is the voltage source at the fault point, and U′ is the sequence impedance matrix at the fault point. f This represents the sequence voltage at the fault point in a normal network.

[0024] S25, Fault-point-based sequence current I f Determine the sequence voltages in the faulty component network:

[0025] ΔU=Z×(-I f (11)

[0026] Where Z is the sequence equivalent impedance matrix associated with the fault point f;

[0027] Based on the superposition principle, the voltage amplitudes at each node are calculated as follows:

[0028] U = U′ + ΔU; (12)

[0029] Based on the current iteration voltage U, the injection current of the inverter-type distributed power source is updated using a low-voltage ride-through control strategy.

[0030] S26. Calculate the maximum absolute difference between the current iteration voltage and the previous iteration voltage:

[0031] ΔU max =max|U (k+1) -U (k) |; (13)

[0032] If ΔU max If the value is less than ε, the iteration terminates; otherwise, return to S22 to continue iterative calculation until the short-circuit current value that meets the accuracy requirements is obtained. Then, the node voltage sag value is obtained through equation (4).

[0033] Preferably, step S23 includes the following steps:

[0034] The admittance matrix Y is decomposed into a product of upper and lower triangular matrices, i.e.:

[0035] Y = LW; (5)

[0036] Where L is an upper triangular matrix and W is a lower triangular matrix;

[0037] Introducing the intermediate variable x, we have:

[0038] Lx = I; (6)

[0039] in,

[0040] After obtaining the intermediate variable x, perform a reverse substitution to calculate the sequence voltage vector U′ in the normal network:

[0041] WU′=x; (7)

[0042] in,

[0043] Preferably, the operating characteristics of the protection device include staged current protection and inverse time overcurrent protection.

[0044] Preferably, the steps for determining the voltage sag duration of the staged current protection are as follows:

[0045] The time of the temporary landing event is denoted as T. initial .

[0046] The branch short-circuit current I is calculated using the following formula:

[0047] YU = I; (14)

[0048] Where U is the node voltage when the short-circuit current value that meets the accuracy requirements is obtained;

[0049] Let the main protection current setting value be I. pp The backup protection current setting value is I. backup When I ≥ I pp Under normal circumstances, when the main protection trips to clear the fault, the voltage dip duration D numerically satisfies the relationship D = T. pp -T initial When the main protection at one or both ends of the line fails to operate, the backup protection operates, and the voltage dip duration D numerically satisfies the relationship D = T. backup -T initial T backup This refers to the triggering time of the backup protection. A variable `r` is introduced: `r = 0` when the main protection operates normally; `r = 1` when the main protection fails to operate at one or both ends of the line, and the backup protection operates. When I... pp >I≥I backup When the main protection does not operate, the backup protection operates, and the voltage dip duration D numerically satisfies the relationship D = T. backup -T initial I backup When the value is >1, it indicates that there is no fault point on this line, and the fault occurs on another line. In summary, the formula for calculating the voltage sag duration of staged current protection is as follows:

[0050]

[0051] Among them, T pp When the main protection is triggered, T initial T is the time when the voltage sag event occurs. backup For the backup protection trigger time, r is an introduced variable, I pp Main protection current setting value, I backup This is the backup protection current setting value.

[0052] Preferably, the steps for determining the voltage sag duration of inverse-time overcurrent protection are as follows:

[0053] The time of the temporary landing event is denoted as T. initial .

[0054] The branch short-circuit current I is calculated using the formula YU = I.

[0055] Solving for the voltage sag duration D: when I p At that time, the protection will not be activated;

[0056] When I>I p At that time, the voltage sag duration of the inverse-time overcurrent protection is:

[0057] D = min(t, T) backup -T initial (17)

[0058] Where t is the protection action time, T backup This is the trigger time for backup protection.

[0059] Preferably, step S4 includes the following steps:

[0060] S41. Calculate and record the node-end voltage U of the non-grid-connected inverter-type distributed power generation. IIDG With the limiting voltage U of the inverter-type distributed power source lim Compare, if U IIDG lim If so, it is marked as a unit to be disconnected from the grid;

[0061] S42. Denote the node terminal voltage of the s-th unit to be disconnected from the grid as U. IIDG_s And calculate the estimated disconnection time T for the s-th unit to be disconnected from the grid. trip_s ;

[0062] S43. Record the estimated disconnection time of all units to be disconnected from the grid in T. trip The minimum expected disconnection time T is obtained by comparison. trip_min ;

[0063] S44, Minimum expected offline time T trip_min The current of the corresponding generating unit and the affected forced tripping generating unit is set to zero, and S42-S43 is repeated;

[0064] S45. Record the information of all affected units that are forced to trip during the iteration process, as well as the voltage drop and corresponding duration after the units trip at different stages. Output a multi-stage sag two-dimensional feature array C containing the voltage sag amplitude and duration. i :

[0065] C i ={(U1,D1),(U2,D2),…,(U k D k )}; (twenty two)

[0066] Among them, U k ​​Let D be the voltage sag amplitude in the k-th stage. k The duration of the voltage dip in the k-th stage.

[0067] Preferably, the formula for calculating the limiting voltage of the inverter-type distributed power source is as follows:

[0068]

[0069] Among them, U max U is the maximum voltage on the curve. min T is the minimum voltage on the curve. max T represents the longest time period corresponding to the maximum voltage on the curve. min This represents the shortest time corresponding to the minimum voltage on the curve.

[0070] Preferably, the formula for calculating the estimated disconnection time of the s-th unit to be disconnected from the grid is as follows:

[0071]

[0072] Among them, U IIDG_s Let be the node-end voltage of the s-th generator unit to be disconnected from the grid.

[0073] The beneficial effects of this invention are:

[0074] 1. This invention proposes for the first time a joint solution method for two-dimensional features of voltage sag amplitude and duration, which solves the problem of decreased accuracy caused by the assumption of mutual independence of two-dimensional features in traditional methods.

[0075] 2. This invention proposes a multi-stage voltage sag inversion method, which solves the problem that traditional methods can only be applied to single-stage voltage sag events.

[0076] 3. This invention accurately reflects the impact of grid disconnection events on the power grid, improves the accuracy of voltage sag assessment in distribution networks, and lays a necessary model foundation for subsequent event tracing, diagnosis, and management. Attached Figure Description

[0077] Figure 1 This is a flowchart of the active distribution network voltage sag inversion method based on two-dimensional feature joint solution according to an embodiment of the present invention.

[0078] Figure 2 This is a schematic diagram of a three-phase PV inverter connected to the power grid according to an embodiment of the present invention.

[0079] Figure 3 This is the IIDG equivalent model during the fault period in an embodiment of the present invention.

[0080] Figure 4 This is a schematic diagram of the LVRT curve of the IIDG and the IIDG disconnection timing during a fault, according to an embodiment of the present invention. Detailed Implementation

[0081] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided with reference to the accompanying drawings and embodiments.

[0082] This invention discloses a method for inverting voltage sags in active distribution networks based on joint solution of two-dimensional features. The specific process is as follows: Figure 1 As shown in the figure. This invention proposes for the first time a joint solution scheme for the two-dimensional features of voltage sag amplitude and duration, solving the accuracy problem caused by the assumption of mutual independence of the two-dimensional features in traditional methods. Existing voltage sag assessment methods often only study the feature quantity of sag amplitude, while assuming that the sag duration follows a probability distribution, ignoring the coupling relationship between duration, voltage sag amplitude and fault current. In fact, the duration of a voltage sag event is usually determined by the duration of the power system fault, which in turn depends on the operating time of the protection device and the operating time of the circuit breaker (i.e., protection operating time). Since overcurrent protection is commonly used in distribution networks, the duration of the sag event is affected by the short-circuit current value. In addition, the connection of distributed generation causes changes in the fault current distribution characteristics, increasing the risk that the protection device may not operate correctly, thus changing the duration of the fault. Existing methods often treat different features of voltage sag as independent, studying only one feature, lacking a joint solution for the two-dimensional features of voltage sag events, resulting in the need to improve the accuracy of existing voltage sag assessment results.

[0083] Furthermore, this invention also proposes a multi-stage voltage sag inversion method, solving the problem that traditional methods can only be applied to single-stage voltage sag events. With the large-scale decentralized integration of distributed power sources in active distribution networks, different generating units have varying low-voltage ride-through capabilities, and their grid connection voltages differ during faults. This can lead to some units potentially disconnecting from the grid during a sag, resulting in a multi-stage evolution of the sag event. However, traditional sag analysis methods do not consider changes in the grid connection status of new energy generating units, and the inversion results cannot reflect the actual system.

[0084] S1. An equivalent model of the voltage-controlled current source for an inverter-interfaced distributed generator (IIDG) is established based on its low-voltage ride-through (LVRT) characteristics. During a fault, the IIDG outputs a fault current coupled to the grid connection point voltage according to the LVRT control strategy. The location of the IIDG is determined based on its position in the actual system, and the total number of IIDG nodes depends on the actual system. A fault is randomly generated, which leads to a voltage sag event.

[0085] Inverter-type distributed power sources are connected to the grid via grid-connected inverters. During normal operation, the inverters are configured to explicitly provide a specific amount of active power to the grid. Taking a photovoltaic (PV) system as an example, the basic diagram of grid connection via a three-phase three-wire inverter is shown below. Figure 2 As shown.

[0086] Figure 2 In the image, C represents the photovoltaic DC voltage regulator capacitor, and U... dc Let R and L be the voltages across the inverter and the equivalent impedances of the lines between the inverter and the grid; i(i a i b i c ) represents the three-phase current, e(e a e b e c ) represents the three-phase voltage on the grid side; dq / abc is a coordinate transformation, changing i and e from the three-phase coordinate system to the dq coordinate system, i.e., i d i q e d e q P and Q represent active and reactive power, respectively.

[0087] During voltage drops, IIDGs must help support the terminal voltage to prevent off-grid operation. When a drop in the positive-sequence voltage at the point of grid connection is detected to be below 90%, according to LVRT requirements, during grid faults, the positive-sequence active current is adjusted based on the magnitude of the voltage drop. and reactive current To perform adaptive adjustments, that is:

[0088]

[0089] All quantities in the formula are per-unit values. K represents the positive sequence voltage amplitude of the power grid after the voltage drop. q I is the reactive power compensation coefficient. max i represents the maximum fault current value of the IIDGs converter. d0 This refers to the pre-fault current value for IIDGs. The active current command is used to issue the maximum permissible active power without overcurrent in the inverter. Take i d0 and The smaller of the two values.

[0090] Under asymmetrical fault conditions, the corresponding fault current reference value can be calculated as follows:

[0091]

[0092] Where p0 is the average active power, q0 is the average reactive power, and p c2p is the cosine component of the active power oscillation. s2 This represents the sinusoidal component of active power oscillation. Reference values ​​for the positive and negative sequence components of the current can be calculated based on the objective of eliminating active power oscillations under asymmetrical fault conditions. The reference value p for active power oscillation... c2 and p s2 It should be set to 0. For and e represents the grid-side voltage, i represents the grid-side current, the superscript + indicates positive sequence, the superscript - indicates negative sequence, the subscript d represents the d-axis component of the dq decomposition method, and the subscript q represents the q-axis component of the dq decomposition method. and The reference value can be generated based on the severity of the voltage drop and the inverter current limiting.

[0093] Therefore, the positive-sequence fault current and the negative-sequence fault current can be calculated as follows:

[0094]

[0095] in, This is the positive sequence fault current. It is a negative sequence fault current.

[0096] Because the fault current of an IIDG is coupled to the corresponding terminal voltage, the fault equivalent model of an IIDG must be configured as a voltage-controlled current source in a positive-sequence or negative-sequence network, such as... Figure 3 As shown in the diagram. U represents the voltage across IIDG, and I is the output current of the current source.

[0097] S2. The fault current and voltage sag of the distribution network are obtained by iterative calculation based on the principle of fault network superposition.

[0098] Utilizing the principle of fault analysis superposition, the fault network of multi-inverter distributed generation (IIDGs) is decomposed into normal operation and fault component networks, where only short-circuit node current is injected. Each IIDG injects short-circuit current into the normal network as a voltage-controlled current source, influencing the short-circuit node current through the normal network voltage. By iteratively calculating the initial current, normal positive and negative sequence network node voltages, short-circuit composite sequence network voltages, fault point current, fault component node voltage, and power reference value, a short-circuit current value meeting accuracy requirements can be obtained, from which a high-precision sag value can be calculated.

[0099] The specific steps are as follows:

[0100] S21. Let the number of iterations be k, and the initial number of iterations be k = 0. Set the convergence threshold ε (10 in this embodiment). -6 Initialize the injection current of the inverter-type distributed power source. This is the current value before the fault.

[0101] S22. Based on the symmetric component method, the unbalanced system is transformed into a positive-sequence network, a negative-sequence network, and a zero-sequence network. The nodal admittance matrix Y (containing each sequence component) is constructed, and the nodal voltage equations of the positive-sequence network sequence can be expressed as:

[0102] YU′=I; (4)

[0103] Among them, U′=[U′ 1 ,U′ 2 ,U′ 0 ] T Represents the sequence voltage vector in a normal network, I = [I 1 ,I 2 ,I 0 The sequence current vector is ], where superscript 1 indicates a positive sequence, superscript 2 indicates a negative sequence, and superscript 0 indicates a zero sequence.

[0104] S23. Calculate the sequence voltage vector U′ in the normal network using the triangular decomposition method.

[0105] The admittance matrix Y is decomposed into a product of upper and lower triangular matrices, i.e.:

[0106] Y = LW; (5)

[0107] Where L is an upper triangular matrix and W is a lower triangular matrix.

[0108] Introducing the intermediate variable x, we have:

[0109] Lx = I; (6)

[0110] in, The subscripts i and j are used to distinguish different elements in the matrix.

[0111] After obtaining the intermediate variable x, perform a reverse substitution to calculate the sequence voltage vector U′ in the normal network:

[0112] WU′=x; (7)

[0113] in,

[0114] S24. Based on the fault boundary conditions and the Thevenin equivalent circuit, calculate the sequence current I at the fault point. f :

[0115]

[0116] Among them, Z f Here is the sequence impedance matrix at the fault point, E is the voltage source at the fault point, and U′ is the sequence impedance matrix at the fault point. f This represents the sequence voltage at the fault point in a normal network.

[0117] The sequence equivalent impedance matrix z associated with the fault point f can be obtained as follows:

[0118]

[0119] In the formula Z i The impedance matrix associated with node i is expressed as:

[0120]

[0121] Solving for matrix z is similar to solving for the expression for U′. Introducing an intermediate variable m satisfies the following equation:

[0122]

[0123] In the formula,

[0124] After obtaining the variable m, a reverse substitution is performed, and the equivalent impedance matrix Z is solved using the following formula:

[0125] WZ = m;

[0126] In the formula,

[0127] S25. Calculate the sequence current I at the fault point. f Subsequently, the sequence voltages in the faulty component network can be determined as follows:

[0128] ΔU=Z×(-I f (11)

[0129] Based on the superposition principle, the voltage amplitudes at each node are calculated as follows:

[0130] U = U′ + ΔU; (12)

[0131] Based on the current iteration voltage U, the injection current of the inverter-type distributed power source is updated using a low-voltage ride-through control strategy.

[0132] S26. Calculate the maximum absolute difference between the current iteration voltage and the previous iteration voltage:

[0133] ΔU max =max|U (k+1) -U (k) |; (13)

[0134] If ΔU max If the result is less than ε, the iteration terminates; otherwise, return to S22 and continue iterating until the required accuracy (ΔU) is obtained. max The short-circuit current value of <ε) (sequence current I at the fault point) fThen, the node voltage sag amplitude (sequence voltage vector U′ in the normal network) is obtained through equation (4).

[0135] S3. A method for estimating the duration of voltage sag considering the impact of IIDG access. Based on the fault current and voltage sag amplitude obtained in S2, the duration of voltage sag is solved by combining the operating characteristics of the protection device, thus realizing the joint solution of two-dimensional features of the sag event.

[0136] In distribution networks, the duration of voltage sags is closely related to the operating characteristics of protection devices. Traditional voltage sag assessment methods typically assume a fixed protection operating time, neglecting the impact of the protection device's operating time characteristics on the sag duration, leading to insufficient assessment accuracy. Commonly used protection devices in distribution networks include staged current protection and inverse-time overcurrent protection. Depending on the operating time characteristics of different types of protection, the correlation between the duration of voltage sags and fault current varies. The operating characteristics of protection devices include staged current protection and inverse-time overcurrent protection.

[0137] For solving the voltage dip duration problem in staged current protection, instantaneous overcurrent protection, time-limited instantaneous overcurrent protection, and overcurrent protection are all protection devices that operate when the current exceeds the limit. To ensure selective fault clearing, these protections are often combined to form staged current protection. In practical applications, two-stage protection is usually used, such as instantaneous overcurrent protection plus overcurrent protection, or time-limited instantaneous overcurrent protection plus overcurrent protection. The instantaneous overcurrent protection or time-limited instantaneous overcurrent protection that triggers first is used as the main line protection, and the triggering time is denoted as T. pp Time-limit overcurrent protection, because its setting current is the maximum load current, has the maximum protection range. As backup protection, it operates when the main protection fails to operate; the triggering time is denoted as T. backup .

[0138] The steps for determining the voltage sag duration of staged current protection are as follows:

[0139] The time of the temporary landing event is denoted as T. initial .

[0140] The branch short-circuit current I is calculated using the following formula:

[0141] YU = I; (14)

[0142] Where U is the node voltage when the short-circuit current value that meets the accuracy requirements is obtained.

[0143] Let the main protection current setting value be I. pp The backup protection current setting value is I. backup When I ≥ I pp Under normal circumstances, when the main protection trips to clear the fault, the voltage dip duration D numerically satisfies the relationship D = T.pp -T initial When the main protection at one or both ends of the line fails to operate, the backup protection operates, and the voltage dip duration D numerically satisfies the relationship D = T. backup -T initial T backup This refers to the triggering time of the backup protection. A variable `r` is introduced: `r = 0` when the main protection operates normally; `r = 1` when the main protection fails to operate at one or both ends of the line, and the backup protection operates. When I... pp >I≥I backup When the main protection does not operate, the backup protection operates, and the voltage dip duration D numerically satisfies the relationship D = T. backup -T initial I backup When the value is >1, it indicates that there is no fault point on this line, and the fault occurs on another line. In summary, the formula for calculating the voltage sag duration of staged current protection is as follows:

[0144]

[0145] To solve the voltage sag duration problem of inverse-time overcurrent protection, considering that the inverse-time overcurrent protection of the distribution network has the characteristics of "short operating time when the current is large and long operating time when the current is small", and its operating time is negatively correlated with the fault current, the operating characteristics of the inverse-time overcurrent protection are combined with the voltage sag duration to improve the accuracy of the assessment.

[0146] According to the IEEE C37.112-2018 standard, the operating characteristic equation for inverse-time overcurrent protection is:

[0147]

[0148] Among them, I p In this embodiment, the starting current setting is 1.2 × I. rated I rated Where t is the rated current of the line, I is the protection action time, TMS is the time multiplier, and α is a constant, which is taken as α = 0.02 in this embodiment.

[0149] The steps for determining the voltage sag duration of inverse-time overcurrent protection are as follows:

[0150] The time of the temporary landing event is denoted as T. initial .

[0151] The branch short-circuit current i is calculated using the formula YU=i.

[0152] Solving for the voltage sag duration D: when i p At that time, the protection will not be activated;

[0153] When I>I​p At that time, the voltage sag duration of the inverse-time overcurrent protection is:

[0154] D = min(t, T) backup -T initial (17)

[0155] Where t is the protection action time, T backup This is the trigger time for backup protection.

[0156] S4. Based on the two-dimensional characteristics of voltage sag events and combined with the grid disconnection timing of different inverter-type distributed power sources, multi-stage voltage sag inversion is achieved.

[0157] like Figure 4 As shown, based on the LVRT characteristics of IIDG, different types of low voltage ride-through curve parameters are set, and T in the figure... clear This represents the clearing time corresponding to the limiting voltage.

[0158] Substituting the voltage sag duration D into the curve equation, the limiting voltage U of each IIDG is calculated. lim Terminal voltage U IIDG lim Then mark it as a unit to be disconnected from the grid.

[0159]

[0160] Among them, U max U is the maximum voltage on the curve. mim T is the minimum voltage on the curve. max T represents the longest time corresponding to the maximum voltage on the curve. min This represents the shortest time corresponding to the minimum voltage on the curve.

[0161] The specific steps are as follows:

[0162] S41. Calculate and record the node-end voltage U of the non-grid-connected inverter-type distributed power generation. IIDG With the limiting voltage U of the inverter-type distributed power source lim Compare, if U IIDG lim If so, it is marked as a unit to be disconnected from the grid.

[0163] S42. Denote the node terminal voltage of the s-th unit to be disconnected from the grid as U. IIDG_s And calculate the estimated disconnection time T for the s-th unit to be disconnected from the grid. trip_s :

[0164]

[0165] S43. Record the estimated disconnection time of all units to be disconnected from the grid in T. trip :​​

[0166] T trip =[T trip_1 ,T trip_2 ,……,T trip_s ,……]; (20)

[0167] The minimum expected disconnection time T is obtained by comparison. trip_min ;

[0168] T trip_min =min(T) trip ); (twenty one)

[0169] S44, Minimum expected offline time T trip_min The current of the corresponding generating unit and the affected forced tripping generating unit is set to zero, and S42-S43 is repeated.

[0170] S45. Record the information of all affected units that are forced to trip during the iteration process, as well as the voltage drop and corresponding duration after the units trip at different stages. Output a multi-stage sag two-dimensional feature array C containing the voltage sag amplitude and duration. i :

[0171] C i ={(U1,D1),(U2,D2),…,(U k D k )}; (twenty two)

[0172] Among them, U k Let D be the voltage sag amplitude in the k-th stage. k The duration of the voltage dip in the k-th stage.

[0173] In summary, considering the low-voltage ride-through characteristics of distributed generation (DG), a voltage-controlled current source model of inverter-type DG is constructed to iteratively calculate the high-precision fault current of the distribution network. Simultaneously, the sag amplitude is calculated, and the duration of the sag event is predicted by combining the operating characteristic equations of the protection device, achieving a joint solution of the two-dimensional characteristics of the sag event. Based on this, by estimating the disconnection time series of different inverter-type DGs, the disconnection sequence of the inverter-type DGs and the voltage distribution of the grid after disconnection are obtained, thus yielding multi-stage voltage sag characteristics and achieving multi-stage voltage sag inversion. The scheme proposed in this application accurately reflects the impact of disconnection events on the grid, improves the accuracy of distribution network voltage sag assessment, and lays a necessary model foundation for subsequent event tracing, diagnosis, and mitigation.

[0174] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for inverting voltage sag in active distribution networks based on joint solution of two-dimensional features, characterized in that, Includes the following steps: S1. Establish an equivalent model of the voltage-controlled current source of the inverter-type distributed power source based on the low voltage ride-through characteristic; S2. The fault current and voltage sag of the distribution network are obtained by iterative calculation based on the principle of fault network superposition. S3. Based on the fault current and voltage sag amplitude obtained in S2, and combined with the operating characteristics of the protection device, the duration of the voltage sag is solved to achieve the joint solution of the two-dimensional characteristics of the sag event. S4. Based on the two-dimensional characteristics of voltage sag events and combined with the grid disconnection timing of different inverter-type distributed power sources, multi-stage voltage sag inversion is achieved. S41. Calculate and record the node-end voltage of the non-grid-connected inverter-type distributed generation. With the limiting voltage of inverter-type distributed power sources If a comparison is made, If so, it is marked as a unit to be disconnected from the grid; S42, the first The node terminal voltage of the unit to be disconnected from the grid is denoted as And calculate the first Expected disconnection time for the Taiwan-based units awaiting disconnection ; S43. Record the estimated disconnection time of all units to be disconnected from the grid. The minimum expected disconnection time is obtained through comparison. ; S44, Minimum estimated disconnection time The current of the corresponding generating unit and the affected forced tripping generating unit is set to zero, and S42-S43 is repeated; S45. Record the information of all affected forced tripping units during the iteration process, as well as the voltage sag and corresponding duration after tripping at different stages. Output a multi-stage sag two-dimensional feature array containing the voltage sag amplitude and duration. : （22） in, For the first The voltage sag amplitude at each stage, For the first The duration of voltage dips in each stage.

2. The method for voltage sag inversion in active distribution networks based on joint solution of two-dimensional features according to claim 1, characterized in that, The low voltage ride-through characteristic satisfies: （1） All quantities in the formula are per-unit values. This is the positive-sequence active current. It is the positive sequence reactive current. This represents the positive-sequence voltage amplitude of the power grid after the voltage drop. This is the reactive power compensation coefficient. This represents the maximum fault current value for an inverter-type distributed power converter. This represents the current value before a fault in an inverter-type distributed power source.

3. The method for inverting voltage sag in active distribution networks based on joint solution of two-dimensional features according to claim 2, characterized in that, S2 includes the following steps: S21. Let the number of iterations be... The initial number of iterations is Set a convergence threshold Initialize the injection current of the inverter-type distributed power source This is the current value before the fault; S22. Based on the symmetric component method, the unbalanced system is transformed into a positive-order network, a negative-order network, and a zero-order network, and the node admittance matrix is ​​constructed. The voltage equations of the nodes in the positive component network sequence can be expressed as: （4） in, This represents the sequence voltage vector in a normal network. The sequence current vector, where superscript 1 indicates a positive sequence, superscript 2 indicates a negative sequence, and superscript 0 indicates a zero sequence; S23. Calculate the sequence voltage vector in the normal network using the triangular decomposition method. ; S24. Calculate the sequence current at the fault point. : （8） in, The sequence impedance matrix at the fault point. The voltage source at the fault point This represents the sequence voltage at the fault point in a normal network. S25, Sequence current based on fault point Determine the sequence voltages in the faulty component network: （11） in, To the fault point The associated sequence equivalent impedance matrix; Based on the superposition principle, the voltage amplitudes at each node are calculated as follows: （12） Based on the current iteration voltage The injection current of the inverter-type distributed power source is updated through a low-voltage ride-through control strategy. ; S26. Calculate the maximum absolute difference between the current iteration voltage and the previous iteration voltage: （13） like If the iteration terminates, the iteration returns to S22 and continues until the short-circuit current value that meets the accuracy requirements is obtained. Then, the node voltage sag value is obtained through equation (4).

4. The active distribution network voltage sag inversion method based on joint solution of two-dimensional features according to claim 3, characterized in that, S23 includes the following steps: Admittance matrix It can be decomposed into the product of upper and lower triangular matrices, that is: （5） in, It is an upper triangular matrix. It is a lower triangular matrix; Introducing intermediate variables Then we have: （6） in, ; Obtaining intermediate variables Then, reverse substitution is performed to calculate the sequence voltage vector in the normal network. : （7） in, .

5. The active distribution network voltage sag inversion method based on joint solution of two-dimensional features according to claim 4, characterized in that, The operating characteristics of the protection device include staged current protection and inverse time overcurrent protection.

6. The active distribution network voltage sag inversion method based on joint solution of two-dimensional features according to claim 5, characterized in that, The voltage sag duration of the staged current protection is as follows: （15） in, When the main protection is triggered, The time when the voltage sag event occurs. For the backup protection trigger time, For the introduced variables, The main protection current setting value. This is the backup protection current setting value.

7. The active distribution network voltage sag inversion method based on joint solution of two-dimensional features according to claim 6, characterized in that, The voltage sag duration of inverse-time overcurrent protection is as follows: set up The starting current setting value, when At that time, the protection will not be activated; when At that time, the voltage sag duration of the inverse-time overcurrent protection is: （17） in, To protect the action time, This is the trigger time for backup protection.

8. The method for voltage sag inversion in active distribution networks based on joint solution of two-dimensional features according to claim 7, characterized in that, The formula for calculating the limiting voltage of the inverter-type distributed power source is as follows: （18） in, This represents the maximum voltage on the curve. The minimum voltage on the curve, This represents the longest time corresponding to the maximum voltage on the curve. This represents the shortest time corresponding to the minimum voltage on the curve.

9. The method for inverting voltage sag in active distribution networks based on joint solution of two-dimensional features according to claim 8, characterized in that, The first The formula for calculating the estimated disconnection time of the units awaiting disconnection is as follows: （19） in, For the first The node-end voltage of the unit waiting to be disconnected from the grid.

Images