A hybrid time-scale power distribution network system robust optimization method and system

Abstract

The application discloses a kind of hybrid time scale distribution network system robust optimization method and system, comprising the following steps: S1: introducing the flexible interconnection device combined by intelligent soft switch and energy storage element at the end of distribution network system;S2: the operation optimization model of distribution network system with flexible interconnection device is established;S3: the day-ahead optimization model and the day-ahead optimization model of distribution network system are established according to the operation optimization model;S4: using day-ahead optimization model to carry out day-ahead hour level optimization to distribution network system;S5: using day-ahead optimization model to carry out day-ahead minute level optimization to distribution network system;S6: establish real-time control strategy based on voltage amplitude sensitivity, according to real-time control strategy, real-time reactive power compensation is carried out to distribution network system, and the distribution network system with real-time optimal robustness is obtained.The method can cope with the risk brought by strong randomness of photovoltaic power generation, maintain system voltage stability, balance distribution network source and load distribution.

Description

Technical Field

[0001] This invention relates to the field of distribution network dispatch optimization technology, and more specifically, to a robust optimization method and system for active distribution networks with mixed time scales. Background Technology

[0002] With the escalating global energy crisis and environmental pollution, coupled with the vigorous development of new energy grid-connected technologies, the development of active distribution networks has received increasing attention. Against this backdrop, the penetration rate of photovoltaic (PV) power generation in distribution networks is continuously increasing. However, PV is inherently random and uncertain, significantly impacting distribution network operation. Furthermore, uneven regional development, large-scale and disorderly integration of PV into distribution networks, and unreasonable power supply distances for distribution lines lead to uneven distribution of power sources and loads within the distribution network. This can easily cause sharp fluctuations in feeder power and voltage exceeding limits, and in severe cases, even affect the operational stability of the distribution network.

[0003] Currently, the main methods for addressing the above problems are distribution network reconfiguration and voltage / reactive power control (VVC). Distribution network reconfiguration can balance loads and improve power supply voltage quality and reliability. Traditional distribution network reconfiguration relies on the cooperation of normally open points (NOPs) and sectionalizing switches to achieve power flow conversion. However, due to the limited number of operations of NOPs and sectionalizing switches, network reconfiguration is often performed only once a quarter or even longer in actual operation, failing to achieve rapid regulation of power flow distribution. VVC equipment mainly consists of on-load tap changers (OLTCs), switchable capacitor banks (SCs), and static var compensators (SVCs). However, these VVC devices cannot achieve continuous reactive power regulation, nor can they flexibly control the direction of power flow. Furthermore, most VVC devices have slow response speeds, making it difficult to meet the above challenges. To address the shortcomings of existing distribution network reconfiguration and VVC equipment, soft open point (SOP) technology has been proposed and is gradually being applied to distribution networks. A Standard Operating Procedure (SOP) is a highly controllable power electronic device that enables flexible connections between feeders and accurately controls active and reactive power flow as well as continuous voltage control. However, current SOPs are expensive and cannot completely replace traditional distribution network control equipment.

[0004] In view of the above, this application is hereby submitted. Summary of the Invention

[0005] The technical problem this invention aims to solve is that existing technologies cannot guarantee the stability of distribution network system operation. The objective is to provide a robust optimization method and system for active distribution networks using a hybrid time scale. This involves introducing a flexible interconnection device combining intelligent soft switches and energy storage components at the ends of distribution network lines. This establishes a day-ahead-intraday robust optimization model for the distribution network, incorporating various voltage regulation devices, and a voltage-sensitivity-based inverter droop control model. This addresses the risks posed by the high randomness of photovoltaic power generation, thereby maintaining system voltage stability and balancing the power and load distribution within the distribution network.

[0006] This invention is achieved through the following technical solution:

[0007] On the one hand, the present invention provides a robust optimization method for a distribution network system with mixed time scales, comprising the following steps:

[0008] S1: Introduce a flexible interconnection device at the end of the distribution network system, which is composed of intelligent soft switches and energy storage components;

[0009] S2: Establish an operation optimization model for the power distribution network system with the aforementioned flexible interconnection device;

[0010] S3: Establish the day-ahead optimization model and intraday optimization model of the distribution network system based on the aforementioned operation optimization model;

[0011] S4: Use the day-ahead optimization model to perform day-ahead hour-level optimization on the distribution network system to minimize the day-ahead losses of the distribution network system;

[0012] S5: Based on minimizing the daytime loss of the distribution network system, the intraday optimization model is used to perform intraday minute-level optimization of the distribution network system to minimize the intraday loss and intraday voltage deviation of the distribution network system.

[0013] S6: Based on minimizing the intraday loss and intraday voltage deviation of the distribution network system, establish a real-time control strategy for time-series voltage amplitude sensitivity under robust conditions, and perform real-time reactive power compensation on the distribution network system according to the real-time control strategy to obtain a distribution network system with optimal real-time robustness.

[0014] As a further description of the present invention,

[0015] The operation optimization model aims to minimize the losses of the distribution network system, the voltage deviation of the distribution network system, and the losses of the flexible interconnection device. It includes: a power flow model of the distribution network system equipped with distributed generation inverters, on-load tap changers, switchable capacitor banks, static var compensators, and the flexible interconnection device.

[0016] As a further description of the present invention, S4 includes:

[0017] Solving the day-ahead optimization model yields the day-ahead control method for the on-load tap changer and the switchable capacitor bank;

[0018] The day-ahead control method described above is used to perform initial optimization of the distribution network system on an hourly basis, so as to minimize the day-ahead network loss of the distribution network system.

[0019] As a further description of the present invention, the method for solving the day-ahead optimization model includes: a two-stage robust optimization method and a column constraint generation method.

[0020] As a further description of the present invention, S5 includes:

[0021] Solving the intraday optimization model yields the intraday control method for the distributed power inverter, the static var compensator, and the flexible interconnection device;

[0022] Based on minimizing the daytime losses of the distribution network system, the distribution network system is further optimized in minutes according to the intraday control method to minimize the intraday losses and intraday voltage deviation of the distribution network system.

[0023] As a further description of the present invention, the secondary optimization includes: adjusting the reactive power of the distributed power inverter, the reactive power of the static var compensator, and the active and reactive power of the flexible interconnection device.

[0024] As a further description of the present invention, S6 includes:

[0025] An improved droop control method for the distributed power inverter and the flexible interconnection device is obtained, resulting in an improved droop control model.

[0026] Based on the modified equations for power flow calculation using the Newton-Raphson method, a model is established to model the relationship between reactive power and voltage amplitude at nodes in a distribution network system, as well as a voltage sensitivity model.

[0027] For each node in the distribution network system, perform the following steps:

[0028] If a voltage overshoot occurs at a node, the voltage amplitude is converted between nodes using the time-series voltage sensitivity model under robust conditions. Then, based on the improved droop control model, the distributed power inverter or flexible interconnect device closest to the node where the voltage overshoot occurred is activated to perform droop control.

[0029] As a further description of the present invention,

[0030] Before step S3, the constraints of the on-load voltage regulator and the switchable capacitor bank in the operating model are subjected to model convex optimization processing, and the constraints of the flexible interconnection device and the distributed power inverter in the operating model are transformed into corresponding second-order cone constraints.

[0031] On the other hand, the present invention provides a robust optimization system for a distribution network system with mixed time scales, comprising:

[0032] The distribution network system creation module is used to create a virtual distribution network system and introduce a flexible interconnection device composed of intelligent soft switches and energy storage elements at the end of the created virtual distribution network system.

[0033] An operation optimization model creation module is used to establish an operation optimization model for a distribution network system with the aforementioned flexible interconnection device;

[0034] The day-ahead optimization model creation module is used to establish a day-ahead optimization model for the distribution network system based on the operational optimization model.

[0035] The intraday optimization model creation module is used to establish an intraday optimization model for the distribution network system based on the operational optimization model.

[0036] The day-ahead optimization module is used to perform day-ahead hour-level optimization of the distribution network system using the day-ahead optimization model to minimize the day-ahead losses of the distribution network system.

[0037] The intraday optimization module is used to perform intraday minute-level optimization of the distribution network system using the intraday optimization model, so as to minimize the intraday loss and intraday voltage deviation of the distribution network system.

[0038] The real-time control strategy creation module is used to establish a real-time control strategy for time-series voltage amplitude sensitivity based on robust conditions.

[0039] The real-time optimization module is used to perform real-time reactive power compensation on the distribution network system according to the real-time control strategy, so as to obtain a distribution network system with real-time optimal robustness.

[0040] As a further description of the present invention, the system also includes:

[0041] The model convex optimization processing module is used to perform model convex optimization processing on the constraints of the on-load voltage regulator and the switchable capacitor bank in the running model.

[0042] The second-order cone constraint conversion module is used to convert the constraints of the flexible interconnection device and the distributed power inverter in the operation model into corresponding second-order cone constraints.

[0043] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0044] 1. The present invention provides a robust optimization method and system for a hybrid time-scale distribution network system. By introducing a flexible interconnection device combining intelligent soft switches and energy storage elements at the end of the distribution network line, the stability of the distribution network is increased, and the system network loss can be effectively reduced and the system voltage can be improved.

[0045] 2. The present invention provides a robust optimization method and system for a distribution network system with mixed time scales. It establishes a day-ahead-intraday robust optimization model for the distribution network with various voltage regulating devices and a time-series voltage sensitivity inverter droop control model based on robust conditions. It can effectively cope with the risks brought about by the strong randomness of photovoltaic power generation and achieve the purpose of maintaining system voltage stability and balancing the power and load distribution of the distribution network.

[0046] 3. The robust optimization method and system for a hybrid time-scale distribution network system provided in this embodiment of the invention can effectively cope with sudden situations in the distribution network based on voltage sensitivity droop control, so that the system does not experience voltage over-limit. Attached Figure Description

[0047] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a schematic diagram of the E-SOP structure provided in Embodiment 1 of the present invention;

[0049] Figure 2 This is a schematic diagram of the multi-timescale optimization framework provided in Embodiment 1 of the present invention;

[0050] Figure 3 A schematic diagram of the improved droop control characteristics provided in Embodiment 1 of the present invention;

[0051] Figure 4 This is a schematic diagram of an improved 33-node system provided in Embodiment 2 of the present invention;

[0052] Figure 5 This is a load and photovoltaic forecast curve provided in Embodiment 2 of the present invention;

[0053] Figure 6 This is a schematic diagram of the system voltage at maximum and minimum photovoltaic output provided in Embodiment 2 of the present invention;

[0054] Figure 7 This is a comparison chart of network loss and voltage under different cases provided in Embodiment 2 of the present invention;

[0055] Figure 8 This is a schematic diagram of the scheduling results of the IDG provided in Embodiment 2 of the present invention;

[0056] Figure 9 This is a schematic diagram of the scheduling results of the E-SOP provided in Embodiment 2 of the present invention;

[0057] Figure 10 This is a schematic diagram of the hourly scheduling strategy for LOT and SC provided in Embodiment 2 of the present invention;

[0058] Figure 11 This is a schematic diagram of the E-SOP active power scheduling quantity provided in Embodiment 2 of the present invention;

[0059] Figure 12 This is a schematic diagram of the reactive power output of SVC, PV and E-SOP provided in Embodiment 2 of the present invention;

[0060] Figure 13 This is a schematic diagram of inverter droop control parameters provided in Embodiment 2 of the present invention;

[0061] Figure 14 This is a schematic diagram of PV output under cloud cover scenario provided in Embodiment 2 of the present invention;

[0062] Figure 15 This is a voltage curve diagram under cloud cover provided in Embodiment 2 of the present invention;

[0063] Figure 16 This is a voltage curve diagram after improved droop control provided in Embodiment 2 of the present invention;

[0064] Figure 17 This is a schematic diagram of the voltage sensitivity of node 16 provided in Embodiment 2 of the present invention;

[0065] Figure 18 This is a voltage curve graph based on voltage sensitivity calculation provided in Embodiment 2 of the present invention;

[0066] Figure 19 This is a comparison diagram of voltage before and after adjustment provided in Embodiment 2 of the present invention;

[0067] Figure 20 This is a schematic diagram of the reactive power output of IDG and E-SOP provided in Embodiment 2 of the present invention;

[0068] Figure 21 This is a voltage comparison curve of three VVC methods provided in Embodiment 2 of the present invention;

[0069] Figure 22 This is a comparison chart of the optimization effects of the three VVC methods provided in Embodiment 2 of the present invention. Detailed Implementation

[0070] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0071] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be apparent to those skilled in the art that these specific details are not necessary to practice the invention. In other embodiments, well-known structures, circuits, materials, or methods have not been specifically described in order to avoid obscuring the invention.

[0072] Throughout this specification, references to "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in connection with that embodiment or example is included in at least one embodiment of the invention. Therefore, the phrases "an embodiment," "an example," "an example," or "an example" appearing in various places throughout the specification do not necessarily refer to the same embodiment or example. Furthermore, specific features, structures, or characteristics can be combined in one or more embodiments or examples in any suitable combination and / or sub-combination. Moreover, those skilled in the art will understand that the illustrations provided herein are for illustrative purposes and are not necessarily drawn to scale. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0073] In the description of this invention, the terms "front", "rear", "left", "right", "up", "down", "vertical", "horizontal", "high", "low", "inner", and "outer" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting the scope of protection of this invention.

[0074] Example 1

[0075] To address the issue of distribution network operational stability being affected by existing distribution network reconfiguration and VVC equipment deficiencies, this embodiment provides a hybrid time-scale robust optimization method for distribution network systems. This method introduces a flexible interconnection device combining intelligent soft switches and energy storage components at the ends of distribution network lines. It also establishes a day-ahead-intraday robust optimization model for the distribution network, incorporating various voltage regulation devices, and an inverter droop control model based on voltage sensitivity. This addresses the risks posed by the high randomness of photovoltaic power generation, aiming to maintain system voltage stability and balance the power and load distribution within the distribution network. Specifically, the hybrid time-scale robust optimization method for distribution network systems includes the following steps:

[0076] S1: Introduce a flexible interconnection device at the end of the distribution network system, which is composed of intelligent soft switches and energy storage components.

[0077] A smart soft switch (hereinafter referred to as SOP) consists of two voltage-source converters (VSCs) and is used to replace tie switches in a distribution network. Since the SOP has a DC branch, this embodiment incorporates an energy storage element (hereinafter referred to as ESS) into the smart soft switch system to form a flexible interconnection device (hereinafter referred to as E-SOP). The structure of the E-SOP is as follows... Figure 1 As shown.

[0078] For E-SOP, the following constraints are established in this embodiment:

[0079] The active power output constraint of E-SOP has the following model expression:

[0080] P SOP,i,t +P SOP,j,t +P SOP,loss,i,t +P SOP,loss,i,t +P dis,t +P ch,t =0 (1),

[0081]

[0082] In the formula, P SOP,i,t and P SOP,j,t P represents the active power injected by SOP into nodes i and j at time t, respectively. SOP,loss,i,t and P SOP,loss,j,t Let A represent the losses at nodes i and j at time SOPt, respectively. SOP,i and A SOP,j The distribution represents the loss coefficient of the SOP, Q SOP,i,t and Q SOP,j,t Let SOP represent the reactive power compensated by SOP to nodes i and j at time t, respectively.

[0083] The reactive power output constraint of E-SOP has the following model expression:

[0084]

[0085] In the formula, Q SOP,i,min and Q SOP,i,max Q represents the minimum and maximum reactive power injected by SOP into node i, respectively. SOP,j,min and Q SOP,j,max These represent the minimum and maximum reactive power injected by SOP into node j, respectively.

[0086] The PQ constraint of E-SOP has the following model expression:

[0087]

[0088] In the formula, S SOP,ij,max This indicates the maximum apparent output power of the SOP;

[0089] The constraints of energy storage elements in E-SOP are expressed in the following model:

[0090] E SOP,t+Δt =E SOP,t +(ζ ch P ch,t -P dis,t / ζ dis )Δt (5),

[0091] 0≤P ch,t ≤(1-δ t )P ch,max (6),

[0092] 0≤P dis,t ≤δ t P dis,max (7)

[0093] E SOP,T =E SOP,1 (8),

[0094] E bat,min ≤E SOP,t ≤E bat,max (9),

[0095] In the formula, E SOP,t ζ represents the charge of ESS at time t. ch and ζ dis P represents the charge / discharge efficiency of the ESS. ch,t and P dis,t Let P represent the charging and discharging power of the ESS at time t, respectively. ch,max and P dis,max δ represents the maximum charging and discharging power of the ESS, respectively. t This is the state of charge / discharge value, when δ t When E = 1, it is in the discharge state; otherwise, it is in the charging state. bat,min and E bat,max These represent the minimum and maximum battery levels of the ESS, respectively.

[0096] S2: Establish an operation optimization model for the distribution network system with the E-SOP.

[0097] Since photovoltaic power generation requires power conversion via an inverter before it can be connected to the grid, this embodiment first establishes a model of a distributed power inverter (hereinafter referred to as IDG) as follows:

[0098]

[0099] In the formula, P pv,j,t and Q pv,j,t Let P be the active and reactive power injected by IDG into node j at time t. pv,max and Q pv,max S represents the maximum active and reactive power output capacity of IDG. G,t This represents the apparent power of the IDG.

[0100] Then, based on the objective of the robust optimization method for distribution network systems with mixed time scales—namely, to adjust various VVC devices, IDGs, and E-SOPs in the distribution network system to minimize network losses, voltage deviations, and equipment losses—the objective function is established as follows:

[0101]

[0102]

[0103]

[0104] Equation (11) represents the minimum system network loss, Equation (12) represents the minimum E-SOP loss, and Equation (13) represents the minimum voltage deviation; ij,t and r ij This represents the sum of the squares of the currents in branch ij at time t, and its resistance; Ω SOP The set of nodes connected to the SOP; T is the number of time periods in a scheduling cycle; N is the number of nodes in the distribution network; V i,t U is the square of the voltage magnitude at node i at time t; ref The root node reference voltage value is set to 1.0 pu.

[0105] Accordingly, a disflow power model for this distribution network system is established, and its expression is as follows:

[0106]

[0107]

[0108]

[0109]

[0110]

[0111]

[0112]

[0113] Q SVC,j,min ≤QSVC,j,t ≤Q SVC,j,max (twenty one),

[0114] In the formula, π(j) represents the set of branches whose head is node j; P ij,t Q ij,t Let x represent the active and reactive power flowing through branch ij at time t, respectively; ij Q represents the reactance of branch ij at time t; j,t and P j,t Inject reactive and active power into node j at time t; Q SC,j,t Q represents the output of SC connected to node j at time t; SVC,j,t Let be the output of the SVC connected to node j at time t; This indicates a branch that does not have an on-load tap changer installed. H represents a branch line with an on-load tap changer installed; ij The transformer ratio for the on-load tap changer on branch ij; and These are the minimum and maximum voltage values ​​at node i, respectively; Q is the maximum value of the branch current ij; SVC,j,max and Q SVC,j,min These represent the upper and lower limits of SVC output.

[0115] S3: Perform convex optimization on the constraints of the on-load tap changer (hereinafter referred to as OLTC) and the switchable capacitor bank (hereinafter referred to as SC) in the operating model, and transform the constraints of E-SOP and IDG in the operating model into corresponding second-order cone constraints.

[0116] (1) Convex optimization treatment of OLTC constraints

[0117] Since equation (17) is a non-convex constraint, it needs to be transformed as follows to convert equation (17) into a 0-1 integer programming problem:

[0118]

[0119]

[0120]

[0121]

[0122] In the formula, m ij,k,t and h j,k,t The auxiliary variable in the transformation; M is an arbitrarily large real number; n is the number of taps of the on-load tap changer; N T,max This represents the maximum number of times a transformer can be switched on and off within a single scheduling cycle.

[0123] (2) Convex optimization treatment of SC constraints

[0124] SC is a reactive power compensation device that switches on and off according to the number of groups. The specific model is shown in equations (26)-(29):

[0125] Q SC,j,t =K SC,j,t ΔQ SC (26),

[0126] Q SC,j,min ≤Q SC,j,t ≤Q SC,j,max (27),

[0127]

[0128]

[0129] In the formula, K SC,j,t ΔQ is the number of capacitor groups switched on / off at node j at time t; SC Q is the reactive power compensation of a single capacitor. SC,j,min and Q SC,j,max For node j, the minimum and maximum values ​​of the switching capacitor bank are; k is determined based on the specific number of capacitor banks installed in SC; K SC,j,t,max δ0, δ1, ..., δ is the upper limit of the number of SC switching groups at node j; k For binary auxiliary variables; ε 1j and ε 2j N is a binary auxiliary variable; SC,max This represents the maximum number of times SC can be switched.

[0130] (3) Regarding the transformation of E-SOP and IDG constraints into corresponding second-order cone constraints

[0131] Since equations (2), (4) and (10) are quadratic nonlinear constraints, they can be transformed into the following second-order cone constraints:

[0132]

[0133]

[0134]

[0135] S4: Establish the day-ahead optimization model and intraday optimization model of the distribution network system based on the operation optimization model.

[0136] Due to the randomness of load and distributed generation, this embodiment decomposes the distribution network optimization operation problem into a mixed time scale optimization problem of "daytime hour-intraday minute-real-time".

[0137] In distribution network systems, OLTC and SC are characterized by long response times and limited number of actions, so they are set as "day-ahead hour-level" optimization variables in mixed time-scale optimization. SVC, IDG, and E-SOP, as power electronic devices, have rapid impact speeds, so the second-level dispatching selects to perform 15-minute-level "intraday minute-level" control on the reactive power regulation capabilities of SVC, IDG, and E-SOP based on the first-level optimization strategy. Both "hour-level" and "minute-level" control are based on predictive curves, but distribution network operation always experiences sudden disturbances, such as sudden changes in PV output caused by weather changes, which can easily lead to voltage exceeding limits. Since the droop control of E-SOP and IDG has the advantage of local real-time adjustment, the third level will adjust the "real-time" droop control of E-SOP and IDG. The final designed three-level control system structure framework is as follows: Figure 2 As shown.

[0138] Current stage:

[0139] The day-ahead optimization phase is an optimization strategy based on the day-ahead PV and load forecast curves. Since the actual operation of the distribution network differs significantly from the day-ahead forecast curves, and considering the impact of uncertainties in PV and load, this embodiment utilizes a two-stage robust optimization algorithm to solve the day-ahead optimization model of the distribution network. Two-stage robust optimization is a three-level optimization method of "min-max-min". The min problems in the first and second stages are used to find the cost that minimizes the objective function. The max problem in the second stage aims to find the worst-case scenario, i.e., a scenario where the optimal solution is applicable to all possible values ​​within the uncertainty set. Two-stage robust optimization can be solved using a column constraint generation algorithm.

[0140] The focus in the day-ahead phase is on minimizing system losses; therefore, the objective function for the day-ahead phase is:

[0141] f day-ahead =ω1f loss +ω2f SOP,loss (33),

[0142] In the formula, ω1 and ω2 are weighting coefficients, which can be adjusted according to actual needs.

[0143] The day-ahead two-stage robust optimization model containing E-SOP is shown in equation (34):

[0144]

[0145] In equation (34),

[0146] In equation (35), and These represent the upper and lower limits of load fluctuations. and The upper and lower limits of photovoltaic fluctuations are given. The constraints for the first-level min problem are (1), (3), (5)-(9), (14)-(20) and (22)-(32); the constraints for the second-level max problem are (35); and the constraints for the third-level min problem are (21).

[0147] Intraday phase:

[0148] The day-ahead decision-making is based on the worst-case scenario of load and PV output fluctuations, therefore its economic efficiency is relatively poor. It requires readjustment of flexibly controllable equipment during the intraday phase. During the intraday phase, based on the forecast value one hour in advance, deterministic optimization of PV and SVC reactive power, as well as E-SOP reactive and active power, is performed at 15-minute intervals. The intraday scheduling phase focuses on system losses and system voltage deviation; therefore, the objective function for intraday scheduling is:

[0149] f day-in =λ1f loss +λ2f SOP,loss +λ3f v (36),

[0150] The constraints are equations (1), (3), (5)-(10), (14)-(21) and (30)-(32).

[0151] In the formula, λ1, λ2 and λ3 are weighting coefficients, which can be adjusted according to actual needs.

[0152] Real-time phase:

[0153] Although the first two stages can maintain the distribution network's normal operation along the predicted curve, the network may still encounter unforeseen circumstances during the 15-minute dispatch period, such as sudden weather changes like cloud cover. This could put the distribution network at risk of voltage exceedances. Therefore, in the real-time stage, it is necessary to adjust the system's reactive power as quickly as possible to cope with sudden voltage exceedances.

[0154] In the real-time phase, with voltage deviation as the target, when the system voltage exceeds the limit, IDG and E-SOP will activate QV droop control to promptly supplement reactive power to the distribution network and ensure system voltage stability.

[0155] This embodiment proposes a novel real-time voltage control strategy. First, when node i's voltage is the most severely over-limited in the entire network, the IDG or E-SOP closest to node i is selected for reactive power regulation. Second, the voltage regulation value ΔU of node i is adjusted using a voltage sensitivity matrix. i,tTransformed into the voltage regulation value ΔU of the target node j equipped with IDG or E-SOP i,j Finally, based on the droop control curve, the IDG or E-SOP is scheduled for reactive power adjustment.

[0156] Considering the volatility of load time-series characteristic curves and risks such as PV disconnection, this embodiment employs a robust time-series voltage sensitivity analysis to enhance the control capability of the IDG / E-SOP over the distribution network, aiming to correct the reactive power regulation of the IDG / E-SOP and ensure voltage stability at all nodes. The calculation of the voltage sensitivity matrix relies on Newton-Lafarge power flow calculations. If the voltage sensitivity matrix were adjusted in real-time based on distribution network load and PV output, it would increase the computational complexity of the system and affect real-time distribution network regulation. Therefore, this paper combines voltage sensitivity calculation with the day-ahead optimization phase, using a 24-hour period as one cycle and a 1-hour time scale. The worst-case load scenario obtained from the robust optimization in the two day-ahead phases is used as the load time-series characteristic curve, thereby obtaining the voltage sensitivity matrix under robust conditions.

[0157] The Newton-Layer power balance equations under robust conditions are modified to equations (37) and (38):

[0158]

[0159]

[0160] In the formula, ΔP i,t Inject the change in active power at node i at time t; ΔQ i,t P represents the change in reactive power injected into node i at time t. i,t,worst Q represents the active power load of node i in the worst-case scenario at time t during the current phase. i,t,worst U represents the reactive power load of node i at time t under the worst-case scenario. i,t Let G be the voltage value of node i at time t; j∈i represents the node adjacent to node i; ij and B ij These represent the conductance and susceptance of branch ij, respectively; θ ij Let be the voltage phase angle difference between nodes i and j.

[0161] Taking the partial derivatives of equations (37) and (38), the relationship between the voltage change ΔU and ΔQ is shown in equation (39):

[0162]

[0163] In the formula: Δθ i,t Let ΔU be the change in voltage phase angle at node i at time t. i,t Let be the change in voltage amplitude at node i at time t.

[0164] ΔU i,t =α ij,t ΔU j,t (40),

[0165]

[0166] d ij,t =-lg(α) ij,t *α ij,t (42),

[0167] In the formula, α ij,t q represents the voltage sensitivity of node i and node j between time t; j,t d represents the reactive power injected at node j at time t; ij,t This represents the electrical distance between node i and node j at time t.

[0168] Therefore, the improved droop control characteristics of this invention are as follows: Figure 3 As shown, the model is as shown in equations (43)-(45):

[0169]

[0170]

[0171] In the formula, Let IDG / E-SOP be the reactive power output value of node j at time t. For the IDG / E-SOP reactive power adjustment of node j, and These are the upper and lower limits of the reactive power output of node j (IDG / E-SOP); U i (t-1) represents the voltage value of node i before adjustment. and These are the upper and lower limits of the voltage value at node i; To control the slope downwards; The voltage value of node i when no reactive power is provided to IDG / E-SOP.

[0172] Among them, the droop control slope As shown in equation (45):

[0173] Example 2

[0174] This embodiment utilizes, as Figure 4 The IEEE 33-bus system shown is used for simulation calculations. The predicted curves for load and PV output are as follows: Figure 5As shown, the load and PV fluctuation ranges are both set to [0.85-1.15]. The root node voltage is set to 1.0 pu; the maximum current of each branch is 500A; seven sets of capacitors are installed at nodes 11 and 21, each with a reactive power capacity of 40 kVar; a soft switch with energy storage is installed between nodes 18 and 33, with a SOP capacity of 1000 kVA, a maximum reactive power compensation of 400 kVar, and a loss factor of 0.02; the ESS has a maximum capacity of 1000 kW, a maximum charging and discharging power of 400 kW, and a charging and discharging efficiency of 90%; PV inverters are installed at nodes 9, 11, 27, and 29, with a maximum reactive power compensation capacity of 500 kVar; static var compensators are installed at nodes 3, 6, and 23, with a reactive power compensation range of [-300 kVar, 300 kVar]; the OLTC is set to 11 levels, with a transformation ratio of 0.01 for each level.

[0175] On a computer configured with a main frequency of 2.30GHz, 16.0GB of memory, and a Windows 10 operating system, the example was solved by calling the Gurobi solver through Yalmip, based on the Matlab platform.

[0176] (1) Comparative analysis of optimized operation of E-SOP connected to active distribution network

[0177] To compare the advantages of E-SOP access to the distribution network, five cases are set up for comparative analysis: Case 1 is without any optimization measures; Case 2 adds PV to Case 1; Case 3 adds OLTC, SC and SVC coordinated control to Case 2; Case 4 adds SOP to Case 3; Case 5 is the solution of this invention.

[0178] Optimizing the control of the above five schemes, the system voltage at the maximum photovoltaic output (14:00) and minimum photovoltaic output (22:00) is as follows: Figure 6 As shown, the key indicators of network loss and voltage in each case are compared to, for example... Figure 9 As shown.

[0179] Depend on Figure 6 and Figure 7 It can be seen that after E-SOP is connected to the distribution network, the distribution network loss and voltage amplitude indicators are significantly improved, and the power quality is significantly improved.

[0180] The scheduling results of IDG and E-SOP in the present invention are as follows: Figure 8-9 As shown.

[0181] (2) Analysis of scheduling results under mixed time scales

[0182] (2.1) Hourly optimizations

[0183] Given the PV and load curves for a given day, a two-stage robust optimization is performed to obtain the "hourly" scheduling strategies for OLTC and SC, as follows: Figure 10 As shown.

[0184] The average network loss for day-ahead hourly scheduling is 172.44 kW, and the average voltage is 0.9660 pu. Since day-ahead hourly optimization is performed under the worst-case scenario, which is relatively rare, the actual network loss and voltage figures should be better than the day-ahead hourly scheduling results.

[0185] (2.2) Intraday minute-level optimization

[0186] This stage adjusts the active and reactive power outputs of E-SOP, as well as the reactive power outputs of SVC and IDG, with the results as follows: Figure 11-12 As shown.

[0187] With the secondary scheduling of E-SOP, IDG and SVC, after the minute-level optimization within the day, the average network loss of the distribution network is 107.6501kW, the average voltage value is 0.9906pu, and the maximum voltage deviation value is 0.0399pu, which verifies the effectiveness of the minute-level scheduling within the day.

[0188] (2.3) Real-time adjustment phase

[0189] To verify the effectiveness of the real-time control model of this invention, the test period was selected as 12:00-12:15 am. First, two sudden scenarios were set up during this period: Scenario A is that during normal PV power generation, cloud cover suddenly occurs, causing a sudden drop in PV power generation; Scenario B is that cloud cover and the root node of the distribution network suddenly change, causing the distribution network voltage to exceed the limit.

[0190] The inverter droop control parameters for this time period are as follows: Figure 13 As shown.

[0191] Scenario A: In this scenario, the PV output suddenly drops. Figure 14 The PV output curve is shown for the period from 11:35 to 12:45. The corresponding system voltage curve for this stage is shown below. Figure 15 As shown. Since the voltage at node 18 exceeds the limit, if improved droop control without a voltage sensitivity matrix is ​​used, the resulting system voltage is as follows. Figure 16 As shown. By Figure 16It can be seen that the voltage value of node 18 just exceeds the minimum voltage limit, but nodes 14-17 still do not meet the voltage constraint. The lowest voltage is at node 16, with a value of 0.9487 pu. Therefore, this invention utilizes voltage sensitivity to improve the IDG and SOP droop control methods. First, sensitivity analysis is used to obtain the voltage sensitivity of node 16 and each node as follows: Figure 17 As shown, droop control is then performed based on this voltage sensitivity, and the final system voltage curve is as follows. Figure 18 As shown.

[0192] Depend on Figure 14-18 As can be seen, the method of this invention increases the voltage of node 16 from 0.9487 pu to 0.9501 pu, thereby ensuring the stability of the system voltage and verifying the effectiveness of the algorithm.

[0193] Scenario B: Assume that in this scenario, the root node voltage decreases from 1.0 pu to 0.985 pu. A voltage comparison analysis is performed on node 18, connected to the E-SOP, and node 16, which has the lowest voltage. Based on the method proposed in this invention, a transformer comparison before and after reactive power adjustment is obtained. Figure 19 As shown, the reactive power outputs of IDG and E-SOP are as follows: Figure 20 As shown.

[0194] Depend on Figures 19-20 It can be seen that when the distribution network encounters both root node voltage drift and cloud cover, E-SOP and IDG will perform droop control as needed to compensate the system for reactive power in a timely manner, thereby ensuring system reliability.

[0195] (3) Robustness analysis

[0196] To verify the effectiveness of the proposed hybrid time-scale optimization method, this invention compares and analyzes traditional centralized VVC, stochastic programming-based VVC, and the proposed method. Centralized optimization is a deterministic optimization method that adjusts various VVC devices only once using predictive information. Stochastic optimization describes the uncertainty of PV and load by generating 500 scenarios randomly in Monte Carlo during the day-ahead phase. Then, it selects the scenario with the largest network loss without optimization for day-ahead scheduling, retains OLTC and SC results, and coordinates and controls various VVCs during the intraday phase.

[0197] The final average voltage optimization results of the three methods are as follows: Figure 21 and 22 As shown.

[0198] Depend on Figure 21 and 22As can be seen, the hybrid time-scale robust optimization method proposed in this invention has the fewest voltage limit violations and the smallest voltage deviation compared to stochastic and centralized optimization methods. Therefore, the method of this invention has strong robustness and security.

[0199] Example 3

[0200] This embodiment provides a robust optimization system for a distribution network system with mixed time scales, including:

[0201] The distribution network system creation module is used to create a virtual distribution network system and introduce a flexible interconnection device composed of intelligent soft switches and energy storage elements at the end of the created virtual distribution network system.

[0202] An operation optimization model creation module is used to establish an operation optimization model for a distribution network system with the aforementioned flexible interconnection device;

[0203] The day-ahead optimization model creation module is used to establish a day-ahead optimization model for the distribution network system based on the operational optimization model.

[0204] The intraday optimization model creation module is used to establish an intraday optimization model for the distribution network system based on the operational optimization model.

[0205] The day-ahead optimization module is used to perform day-ahead hour-level optimization of the distribution network system using the day-ahead optimization model to minimize the day-ahead losses of the distribution network system.

[0206] The intraday optimization module is used to perform intraday minute-level optimization of the distribution network system using the intraday optimization model, so as to minimize the intraday loss and intraday voltage deviation of the distribution network system.

[0207] The real-time control strategy creation module is used to establish a real-time control strategy for time-series voltage amplitude sensitivity based on robust conditions.

[0208] The real-time optimization module is used to perform real-time reactive power compensation on the distribution network system according to the real-time control strategy, so as to obtain a distribution network system with real-time optimal robustness.

[0209] The model convex optimization processing module is used to perform model convex optimization processing on the constraints of the on-load voltage regulator and the switchable capacitor bank in the running model.

[0210] The second-order cone constraint conversion module is used to convert the constraints of the flexible interconnection device and the distributed power inverter in the operation model into corresponding second-order cone constraints.

[0211] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A hybrid time-scale power distribution network system robust optimization method, characterized in that, Includes the following steps: S1: Introduce a flexible interconnection device at the end of the distribution network system, which is composed of intelligent soft switches and energy storage components; S2: Establish an operation optimization model for the distribution network system with the flexible interconnection device; wherein, the operation optimization model aims to minimize the losses of the distribution network system, the voltage deviation of the distribution network system, and the losses of the flexible interconnection device, and includes: a power flow model of the distribution network system equipped with distributed power inverters, on-load tap changers, switchable capacitor banks, static var compensators, and the flexible interconnection device. S3: Establish the day-ahead optimization model and intraday optimization model of the distribution network system based on the aforementioned operation optimization model; S4: Solve the day-ahead optimization model to obtain the day-ahead control method for the on-load tap changer and the switchable capacitor bank; the method for solving the day-ahead optimization model includes: a two-stage robust optimization method and a column constraint generation method; perform initial optimization of the distribution network system on an hourly basis according to the day-ahead control method to minimize the day-ahead loss of the distribution network system; S5: Solve the intraday optimization model to obtain the intraday control method for the distributed power inverter, the static var compensator, and the flexible interconnection device; based on minimizing the day-ahead loss of the distribution network system, perform secondary optimization of the distribution network system in minutes according to the intraday control method to minimize the intraday loss and intraday voltage deviation of the distribution network system. S6: Based on minimizing the intraday loss and intraday voltage deviation of the distribution network system, a real-time control strategy based on the time-series voltage sensitivity model under robust conditions is established. The reactive power compensation of the distribution network system is performed in real time according to the real-time control strategy to obtain a distribution network system with optimal real-time robustness. The voltage sensitivity calculation is combined with the day-ahead optimization stage, with the day-ahead 24 hours as a cycle and 1 hour as a time scale. The worst load scenario obtained by the robust optimization in the two day-ahead stages is used as the load time-series characteristic curve, thereby obtaining the voltage sensitivity matrix under robust conditions. S6 includes: improving the droop control method for the distributed power inverter and the flexible interconnection device to obtain an improved droop control model; the improved droop control model is specifically as follows: , In the formula, For nodes exist The reactive power output value of the distributed power inverter / flexible interconnect device at any given time. For nodes Reactive power adjustment of distributed power inverters / flexible interconnection devices and They are nodes Upper and lower limits of reactive power output of distributed power inverters / flexible interconnection devices; For nodes The voltage value before adjustment and For nodes Upper and lower limits of voltage value; To control the slope downwards; When the distributed power inverter / flexible interconnect device does not provide reactive power, the node The voltage value; Represents a node With nodes Voltage sensitivity between; droop control slope As shown in the following formula: ; Based on the modified equations for power flow calculation using the Newton-Raphson method, a model is established to model the relationship between reactive power and voltage amplitude at nodes in a distribution network system, as well as a voltage sensitivity model. For each node in the distribution network system, the following steps are performed: determine whether a voltage over-limit has occurred at the node; if a voltage over-limit has occurred, then the voltage amplitude is converted between nodes using the time-series voltage sensitivity model under robust conditions, and the distributed power inverter or flexible interconnection device closest to the node where the voltage over-limit occurred is mobilized for droop control according to the improved droop control model.

2. The robust optimization method for a distribution network system with mixed time scales according to claim 1, characterized in that, The secondary optimization includes adjusting the reactive power of the distributed power inverter, the reactive power of the static var compensator, and the active and reactive power of the flexible interconnection device.

3. The robust optimization method for a distribution network system with mixed time scales according to claim 1, characterized in that, Before step S3, the constraints of the on-load voltage regulator and the switchable capacitor bank in the operation optimization model are subjected to model convex optimization processing, and the constraints of the flexible interconnection device and the distributed power inverter in the operation optimization model are transformed into corresponding second-order cone constraints.

4. A robust optimization system for a distribution network system with mixed time scales, characterized in that, A robust optimization method for a distribution network system with mixed time scales as described in any one of claims 1-3 includes: The distribution network system creation module is used to create a virtual distribution network system and introduce a flexible interconnection device composed of intelligent soft switches and energy storage elements at the end of the created virtual distribution network system. An operation optimization model creation module is used to establish an operation optimization model for a distribution network system with the aforementioned flexible interconnection device; The day-ahead optimization model creation module is used to establish a day-ahead optimization model for the distribution network system based on the operational optimization model. The intraday optimization model creation module is used to establish an intraday optimization model for the distribution network system based on the operational optimization model. The day-ahead optimization module is used to perform day-ahead hour-level optimization of the distribution network system using the day-ahead optimization model to minimize the day-ahead losses of the distribution network system. The intraday optimization module is used to perform intraday minute-level optimization of the distribution network system using the intraday optimization model, so as to minimize the intraday loss and intraday voltage deviation of the distribution network system. The real-time control strategy creation module is used to establish a real-time control strategy based on a time-series voltage sensitivity model under robust conditions. The real-time optimization module is used to perform real-time reactive power compensation on the distribution network system according to the real-time control strategy, so as to obtain a distribution network system with real-time optimal robustness.

5. A robust optimization system for a distribution network system with mixed time scales according to claim 4, characterized in that, Also includes: The model convex optimization processing module is used to perform model convex optimization processing on the constraints of the on-load voltage regulator and the switchable capacitor bank in the running optimization model. The second-order cone constraint transformation module is used to transform the constraints of the flexible interconnection device and the distributed power inverter in the operation optimization model into corresponding second-order cone constraints.

Images