A method for enhancing the resilience of distribution networks through multi-resource collaboration between power sources, grids, and loads in ice storm scenarios.
By constructing an ice disaster impact mechanism model and formulating multi-resource collaboration strategy, establishing a two-stage mixed integer random planning model, combining robust optimization to deal with uncertainty, the coordination problem of the distribution network's resilience improvement in extreme ice disaster scenarios is solved, and the coordination problem of power distribution networks is achieved. Load supply and resilience increase during ice disasters.
Patent Information
- Application Number
- CN202411169149.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-08-23
AI Technical Summary
In the research on the resilience improvement of distribution networks in the extreme ice disaster scenarios, the strategies of each stage lack coordination, making it difficult to effectively deal with the negative impacts of long-term and large-scale ice disasters, especially in the uncertainty of new energy output and lines. In the case of failure, it is difficult to reflect the real state of the distribution network, and it is inefficient to solve the problem of large-scale random optimization.
A method for improving the resilience of distribution networks with multiple resources in ice disaster scenarios is proposed. By constructing a mechanism model for the influence of space-time characteristics of ice disasters, a distribution network line reinforcement, liquefied natural gas capacity configuration and maintenance team configuration plan is formulated, and multiple faults are established. The two-stage mixed integer random planning model of the scene uses the punishment-based Gaussian-Sedal method to decompose the problem, and combines robust optimization to deal with the uncertainty of the scenery output.
Maintain important load supply in the distribution network during the ice disaster, reduce load losses through coordinated multiple types of resources, improve the resilience of the distribution network, improve the model solution efficiency, and effectively reduce the load loss cost.
Smart Images

Figure CN118693820B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of improving the resilience of distribution networks, and develops a method for improving the resilience of distribution networks by coordinating multiple resources of sources, grids and loads under extreme ice disaster scenarios. Background Art
[0002] In recent years, extreme weather disasters have occurred frequently around the world, endangering the safe and stable operation of power systems. Compared with extreme weather such as typhoons and rainstorms, ice disasters cover a wider range and last longer, which has a negative impact on the source side, grid side, and load side of the power system, bringing greater risks to the safe operation of the power system. Therefore, in order to improve the power system's ability to resist extreme ice disaster events, it is necessary to effectively evaluate the resilience of the power system under ice disasters and propose corresponding resilience improvement strategies.
[0003] Power system resilience is defined as the ability of a power system to prevent, resist, and quickly restore loads against low-probability, high-loss extreme events. The distribution network is located near the terminal load and is a key link in the power supply of users. Due to its complex operating environment and topology, low network redundancy, and weak ability to resist disasters, research on improving the resilience of the distribution network has received widespread attention. For example, some studies have effectively improved the distribution network's ability to resist typhoon disasters by adopting line reinforcement and energy storage configuration measures. Some studies have pre-configured and dispatched flexible emergency resources such as mobile power generation vehicles and mobile energy storage, and achieved rapid recovery of the distribution network after the disaster. For ice disaster scenarios, existing studies have proposed a distribution network situation awareness framework and resilience assessment system for ice disaster scenarios, which are of great significance to the disaster prevention of distribution networks; some studies have proposed a multi-period ice melting strategy for distribution networks based on the flexible dispatch of mobile ice melting devices to limit line ice coverage within a safe range; some studies have proposed a post-ice disaster emergency repair and dispatch strategy for an electric-thermal coupling system that considers mobile energy storage and maintenance teams, which effectively reduces load shedding losses. However, most studies on improving the resilience of distribution networks under ice disaster scenarios only consider a single stage of disaster response or post-disaster recovery, resulting in a lack of coordination between resilience improvement strategies at each stage; at the same time, in order to effectively deal with the adverse effects of long-term and large-scale ice disasters, it is necessary to use a variety of coordinated resilience resources to achieve more effective resilience improvement. In addition, with the increase in the penetration rate of new energy in distribution networks, the impact of ice disasters on the output of new energy cannot be ignored. Wind power and photovoltaic power are directly exposed to ice disaster environments, and their unit output is affected, and their ability to provide emergency support is limited. Therefore, it is necessary to analyze and model the mechanism of the impact of ice disasters on the source-grid-load of the power system, while considering line fault conditions and the uncertainty of new energy output, otherwise it is difficult to reflect the true state of the distribution network under ice disaster scenarios.
[0004] In terms of considering uncertain factors, robust optimization and stochastic optimization are common methods for solving uncertain optimization problems. Robust optimization solves resilience improvement solutions under the most serious system failure scenarios. Since the most serious failure scenarios do not occur often, the resilience improvement solutions obtained by robust optimization may be conservative. Another idea is the stochastic optimization method, which describes the uncertainty of line failures based on a series of failure scenarios. However, solving large-scale stochastic optimization problems involving a large number of scenarios is very challenging. In order to improve the efficiency of problem solving, scenario reduction technology can be used to select some typical scenarios as representatives, or algorithms can be designed to decompose the original problem into multiple smaller sub-problems for iterative solution. In order to achieve resilience improvement under multi-resource collaboration, the optimization model contains more integer variables. How to design a decomposition algorithm suitable for mixed integer stochastic programming problems to ensure convergence effect and calculation accuracy is also a problem worthy of study. Summary of the invention
[0005] The present invention proposes a method for improving the resilience of a distribution network by coordinating multiple resources including source, grid and load in ice disaster scenarios. Based on the analysis of the spatiotemporal characteristics of ice disasters, the impact mechanism of ice disasters on the source, grid and load of the distribution network is revealed and a corresponding model is established; in view of the differentiated resource requirements for system prevention, response and recovery, distribution network line reinforcement, liquefied natural gas capacity configuration and maintenance team configuration plans are formulated in the pre-disaster stage, and during the disaster, resilience improvement strategies such as unit output scheduling, distribution network reconstruction, maintenance team scheduling and maintenance, and differentiated load shedding are adopted to minimize load losses. Based on the proposed strategy, a two-stage mixed integer stochastic programming model for multiple fault scenarios is established, and the penalty-based Gauss-Seidel method is used to decompose the large-scale stochastic mixed integer programming problem into multiple sub-problems that can be solved in parallel for iterative solution, thereby improving the solution efficiency of the model, and a robust optimization method is used to handle the uncertainty of wind and solar output. The method of the present invention can maintain the supply of important loads in the distribution network during ice disasters, and the coordination of multiple types of resources can effectively reduce load losses and improve the resilience of the distribution network.
[0006] In order to achieve the purpose of the present invention, the present invention provides a method for improving the resilience of a distribution network by synergizing multiple resources of source, grid and load in an ice disaster scenario, comprising the following steps:
[0007] Step 1: Combined with the spatiotemporal characteristics of ice disasters, a model of the impact mechanism of ice disaster weather on distribution network lines, distributed generators and residential loads is constructed; the spatiotemporal characteristics of ice disasters can be described by ice disaster trajectory parameters and ice disaster intensity parameters; the impact of ice disasters on distribution network lines is reflected in the probability of line failure; distributed generators include gas units, wind turbines and photovoltaic systems, and the impact of ice disasters on them is reflected in their output size; the impact of ice disasters on residential loads is reflected in their load increment;
[0008] Step 2: Develop resilience strategies based on differentiated resource requirements for system prevention, response, and recovery. Develop distribution network line reinforcement, LNG capacity configuration, and maintenance team configuration plans in the pre-disaster stage. Minimize load losses during disasters through unit output scheduling, distribution network reconstruction, maintenance team scheduling, and differentiated load shedding resilience strategies.
[0009] Step 3: Based on the failure probability of the distribution network line in each period, the Monte Carlo method is used to generate multiple distribution network line failure scenarios, and then the typical distribution network failure scenarios are obtained through scenario reduction;
[0010] Step 4: Based on the proposed resilience improvement strategy, a two-stage mixed integer stochastic programming model for multiple fault scenarios is established; the first-stage model aims to minimize the resource allocation cost and optimize the resilience resource allocation scheme, and the constraint condition is the upper limit of the number of resources; the second-stage model aims to minimize the load loss cost and achieve the optimal operation of the distribution network, and the constraints include unit operation constraints, distribution network reconstruction constraints, distribution network flow constraints, line maintenance constraints, and line status constraints;
[0011] Step 5: Use the robust optimization method to deal with the uncertainty of wind and solar power output, and use the penalty-based Gauss-Seidel method to solve the proposed mixed integer stochastic programming model.
[0012] The method for improving the resilience of a distribution network with source-grid-load multi-resource coordination in an ice disaster scenario, wherein in step 1, the spatiotemporal characteristic model of the ice disaster is:
[0013] (1)
[0014] (2)
[0015] (3)
[0016] (4)
[0017] in, The center of the ice disaster is t The geographical coordinates of the moment, represents the initial geographic coordinates of the center of the ice disaster, is the moving speed of the center of the ice disaster, is the angle between the moving direction of the ice disaster center and the true north direction (clockwise is positive), Represents a unit time interval; Represents coordinates ( x , y ) to the center of the ice disaster, and The coordinates are ( x ,y ), the freezing rain intensity and wind speed at , They represent the maximum freezing rain intensity and maximum wind speed of the ice disaster, is the radius of the ice disaster.
[0018] The method for improving the resilience of a distribution network with source-grid-load multi-resource collaboration under an ice disaster scenario, wherein in step 1, the impact mechanism model of the ice disaster on the distribution line is:
[0019] (5)
[0020] (6)
[0021] (7)
[0022] (8)
[0023] in, and are the ice growth rate and ice thickness of distribution conductors respectively; and are the freezing rain density and ice density, respectively, and their values are , ; W Represents the liquid water content in the air, which can be expressed by Calculated; , Indicates the threshold of ice thickness that the conductor can withstand; , Distribution conductors l With distribution lines ij exist t The probability of failure at time To form a distribution line ij A collection of power distribution wires.
[0024] The method for improving the resilience of a distribution network with source-grid-load multi-resource coordination under an ice disaster scenario, wherein, in step 1, the impact mechanism model of the ice disaster on distributed generators includes the impact mechanism models of gas units, wind turbines, and photovoltaic systems; for gas units, in ice disaster low temperature weather, the natural gas network pipeline is prone to freezing and blockage, thereby affecting the fuel supply of the gas turbine of the distribution network, causing the unit to be unable to output normally. It is assumed that during the ice disaster, the natural gas network cannot provide fuel supply for the gas turbine.
[0025] The impact mechanism model of ice disaster on wind turbines is as follows:
[0026] (9)
[0027] (10)
[0028] (11)
[0029] (12)
[0030] (13)
[0031] (14)
[0032] (15)
[0033] in, and are the ice growth rate and ice thickness of wind turbine blades, , are the natural wind speed and relative wind speed at the corresponding locations, , , The collision coefficient, capture coefficient and freezing coefficient are respectively, W is the liquid water content in the air, is the axial inductance, r is the fan blade cross-sectional radius, n is the speed of the fan blades. For safety reasons, when the ice thickness of the fan blades reaches the safety threshold, the fan needs to be shut down for de-icing. Indicates that the fan is t The running status at the moment, a value of 1 indicates running, and a value of 0 indicates shutdown; The upper limit of the safety threshold of ice thickness on wind turbine blades; , are the maximum output of wind turbines in normal operation and ice disaster scenarios, respectively. , , They represent the cut-in wind speed, cut-off wind speed and rated wind speed of the wind turbine respectively. Indicates the rated power of the wind turbine. is the wind turbine power loss coefficient, which is related to the ice thickness of the wind turbine blades Related, , is an empirical parameter.
[0034] The impact mechanism model of the ice disaster on the photovoltaic system is:
[0035] (16)
[0036] (17)
[0037] in, is the snow thickness on the photovoltaic panels, is the photovoltaic power loss coefficient, represents the radiation intensity reduction factor, , They are the maximum output of photovoltaic power generation under normal operation and ice disaster scenarios respectively.
[0038] The method for improving the resilience of a distribution network with source-grid-load multi-resource coordination under an ice disaster scenario, wherein in step 1, the impact mechanism model of the ice disaster on residential load is:
[0039] (18)
[0040] (19)
[0041] in, For Node j The original load at is the load increment under the influence of ice disaster, which is related to the body temperature Related, It is the residential load node.
[0042] The method for improving the resilience of a distribution network with the coordination of multiple resources of source, grid and load in an ice disaster scenario, wherein, in step 2, the resilience improvement strategy is that, in the first stage, the pre-disaster configuration of three resilience improvement resources is considered, including: line reinforcement (defense resources), liquefied natural gas (emergency response resources) and maintenance team (repair resources). Among them, line reinforcement can effectively reduce the probability of line failure and ensure the normal operation of important lines; liquefied natural gas (LNG) can be pre-configured at the gas unit as a temporary fuel to avoid the fuel shortage problem caused by the blockage of the natural gas pipeline; the maintenance team is used to repair the faulty line to achieve system recovery. The second stage is the optimized operation of the distribution network in the disaster under the coordination of multiple resources. After the ice disaster invades the system, as the line ice accumulates, some lines fail, and the output of distributed power sources drops sharply. In order to avoid a significant reduction in system performance, emergency power support is provided by dispatching gas units on the power supply side, and the grid topology is optimized by reconstructing the distribution network on the power grid side, and the maintenance team is dispatched to repair the line in time to avoid long-term power outages of the load; on the load side, the importance of different loads is considered, and differentiated load shedding is adopted to give priority to the supply of important loads. With the coordination of the above-mentioned multiple types of resources, the system performance can be reduced to a lesser extent and can return to normal earlier, and the resilience of the distribution network can be effectively improved.
[0043] The method for improving the resilience of a distribution network with source-grid-load multi-resource collaboration under an ice disaster scenario, wherein in step 3, the scenario generation and reduction method is:
[0044] First, calculate the failure probability of each distribution line at each time , in the scene s Next, for each line ij Generate a random number between [0,1] , and the random number is sequentially compared with each time period ( t =1,2,…, T ) contrast:
[0045] (1) If ,but ;
[0046] (2) If ,but .
[0047] in, Representation scene s Down t Timeline ij The fault state of the line is 1, which indicates a normal state, and 0 indicates that the line has failed. The Monte Carlo method is used for random simulation to generate multiple random scenarios of distribution network line failures. Different fault scenarios will cause different load loss situations. According to the load loss index, the K-means clustering method is used to cluster the generated scenarios. After clustering, one scenario can be selected from each set to represent all the scenarios in the set, and the number of samples obtained is k A collection of random failure scenarios .
[0048] The method for improving the resilience of a distribution network with source-grid-load multi-resource collaboration under an ice disaster scenario, wherein the mixed integer stochastic programming model includes a first stage and a second stage;
[0049] In the first stage, a pre-disaster configuration plan for various resilience resources is formulated. The objective function is to minimize the configuration cost. The optimization model for the first stage is:
[0050] (20)
[0051] in, For gas turbines i LNG capacity, The upper limit of LNG resource capacity; Indicates power distribution line ij Whether it is reinforced. A value of 1 indicates that the line is reinforced, and a value of 0 indicates that it is not reinforced. The upper limit of the number of lines that can be reinforced; Indicates the number of configured maintenance teams, The upper limit for the number of maintenance teams; , , They are LNG unit capacity configuration cost, line reinforcement cost and maintenance team configuration cost respectively.
[0052] The method for improving the resilience of a distribution network with source-grid-load multi-resource coordination under an ice disaster scenario, the second-stage model includes an objective function and constraint conditions;
[0053] The objective function of the second stage model is:
[0054] (twenty one)
[0055] in, Representation scene s Next t Time Node j The load reduction at is the load loss cost, For Node j The weight factor of the load.
[0056] The operating constraints of the gas unit are:
[0057] (twenty two)
[0058] (twenty three)
[0059] (twenty four)
[0060] in, Representation scene s Gas Generator Set i exist t The effort of the moment, is the calorific value of natural gas, is the power generation efficiency of the gas unit, is the conversion factor between LNG and gaseous natural gas, is the ramp rate of the gas unit and is the upper limit of the gas unit output.
[0061] The distribution network reconstruction constraints are:
[0062] (25)
[0063] (26)
[0064] (27)
[0065] (28)
[0066] in, Indicates the number of system nodes. For the scene s Medium t time line ij Virtual goods flow on For the scene s middle t Timeline ij The virtual connection status, For the scene s middle t Timeline ij The actual connection status, is the maximum virtual commodity flow on the line, taken as Formula (25) limits the number of closed lines in the system, formula (26) represents the commodity flow balance of the line, formula (27) limits the range of the commodity flow of the line, and formula (28) represents the relationship between the virtual connection state and the actual connection state of the line.
[0067] The distribution network operation constraints are:
[0068] (29)
[0069] (30)
[0070] (31)
[0071] (32)
[0072] (33)
[0073] (34)
[0074] (35)
[0075] Among them, the collection , Node j The child node set and parent node set of , are the node set and the line set respectively; , Respectively for scenes s middle t Time flow through the line ij Active power and reactive power; , Respectively for scenes s middle t The active power and reactive power injected by the substation into the distribution network at all times; , They are t Time Node j Active and reactive load demands at the location; , Respectively for scenes s middle t Time Node j Active and reactive power reduction; , Respectively for scenes s middle t Time Node i , j The voltage amplitude; is the reference voltage; M is a large positive number; , Distribution lines ij resistance and reactance; , are the upper and lower limits of node voltage amplitude respectively, , are the upper limits of active and reactive power of the line, respectively. Equations (29)-(31) are the system power balance constraints, Equation (32) is the voltage drop constraint based on the LinDistFlow model, and Equations (33)-(35) limit the node voltage, line power, and load shedding, respectively.
[0076] The line maintenance constraints are:
[0077] (36)
[0078] (37)
[0079] (38)
[0080] (39)
[0081] (40)
[0082] in, Representation scene s Middle Line ij exist t Whether to start maintenance at any time, is the line maintenance time. Formula (36) limits each line to be repaired at most once, Formulas (37)-(38) indicate that the number of lines under maintenance at each moment does not exceed the number of maintenance teams, and Formulas (39)-(40) indicate that maintenance is only carried out after a line failure. Due to the continuous impact of the ice disaster, ice may still appear after the line maintenance is completed. In order to ensure that the repaired line does not suffer a secondary failure, the ice thickness of the repaired line is required to be limited to the first threshold to ensure that the failure probability of the repaired line is always 0. Therefore, the earliest time point when the faulty line can start maintenance is The constraint can be expressed as:
[0083] (41)
[0084] (42)
[0085] Comprehensively consider the impact of line reinforcement, line failure and line maintenance, and the distribution network line status The constraint can be expressed as:
[0086] (43)
[0087] (44)
[0088] in, and They represent the normal line set and the segment switch set respectively.
[0089] The method for improving the resilience of a distribution network with source-grid-load multi-resource coordination under an ice disaster scenario, wherein in step 5, the wind-solar output robust optimization model is:
[0090] (45)
[0091] (46)
[0092] (47)
[0093] (48)
[0094] in, , They are the maximum wind power and photovoltaic output under ice disaster scenarios, respectively. , are the actual grid-connected power of wind power and photovoltaic power, respectively. Affected by the ice disaster, the actual maximum output of the new energy unit may deviate from the theoretical value, and there is a strong uncertainty. The uncertainty of wind and photovoltaic output is limited by a polyhedron uncertainty set, as shown in formula (47). represents an uncertain variable, which refers to the maximum wind and solar output at each time period under the ice disaster scenario ( , ); is the random fluctuation parameter, which represents the fluctuation degree of the uncertain variable; Indicates the fluctuation deviation value of the uncertain variable; is a robust risk indicator with a value range of [0,1]. This indicator determines the fluctuation range of uncertain variables. =1 when the fluctuation is the largest. = 0, the robust model degenerates into a deterministic model. For equations (45) and (46), robust equivalent transformation is used to obtain equation (48), where: , , , Auxiliary variables introduced for robust equivalence transformation, It is the upper limit of the fluctuation of new energy output.
[0095] The method for improving the resilience of a distribution network with source-grid-load multi-resource collaboration in an ice disaster scenario, wherein in step 5, the penalty-based Gauss-Seidel method is as follows:
[0096] The two-stage mixed integer stochastic programming model to be established can be expressed in the following compact form:
[0097] (49)
[0098] (50)
[0099] (51)
[0100] in, and denote the variables in the first and second stages respectively, For the scene s The probability of, Equations (51) and (52) represent the constraints of the first stage and the second stage respectively.
[0101] In order to make the original problem easier to decompose, the original model is rewritten as:
[0102] (52)
[0103] (53)
[0104] (54)
[0105] Among them, formula (53) is an unexpected constraint, which provides a copy of the first-stage variables for each scenario , while introducing auxiliary variables z To ensure that the variables solved in the first stage are applicable to all scenarios in the second stage. Then, the unexpected constraints are relaxed using augmented Lagrangian relaxation to obtain the following model:
[0106] (55)
[0107] In the formula, is the Lagrange multiplier. To ensure that the augmented Lagrange relaxation (55) has a bounded optimal solution, the dual feasible condition must be satisfied, that is, ,therefore Can be set to 0
[34] . is the penalty term. When the penalty term is formula (56), the strong duality of the model can be guaranteed to hold.
[0108] (56)
[0109] in, is a constant greater than 1, For the n The multiplier factor of the iteration, which increases with the number of iterations to improve the convergence of the algorithm; is the penalty factor; , which can be linearized as:
[0110] (57)
[0111] After the above processing, the model is finally transformed into:
[0112] (58)
[0113] Each scene s The corresponding sub-problem can be expressed as:
[0114] (59)
[0115] The specific implementation steps of the penalty-based Gauss-Seidel (PBGS) algorithm are as follows:
[0116] Compared with the prior art, the present invention can achieve the following beneficial effects:
[0117] The present invention has developed a distribution network resilience improvement method that coordinates multiple resources of source, grid and load in ice disaster scenarios. In terms of the impact mechanism of ice disasters, the present invention comprehensively considers the multi-dimensional impact of ice disasters on the source, grid and load of the distribution network, and establishes a corresponding impact mechanism model to better reflect the real state of the distribution network in ice disaster scenarios; in terms of resilience improvement strategies, the present invention combines multiple types of resources of source, grid and load, formulates distribution network line reinforcement, liquefied natural gas capacity configuration and maintenance team configuration plans in the pre-disaster stage, and reduces load losses through unit output scheduling, distribution network reconstruction, maintenance team scheduling and maintenance, and differentiated load shedding resilience improvement strategies during disasters; in terms of model establishment and solution, a two-stage mixed integer stochastic programming model for multiple fault scenarios is established based on the proposed strategy, and the penalty-based Gauss-Seidel method is used to decompose the large-scale mixed integer stochastic programming problem into multiple sub-problems that can be solved in parallel for iterative solution, thereby improving the solution efficiency of the model, and a robust optimization method is used to handle the uncertainty of wind and solar output. The present invention helps to maintain the supply of important loads in the distribution network during ice disasters, alleviate the adverse effects of long-term and large-scale ice disasters, and through the coordination of multiple resilience resources, it can effectively reduce load loss costs and improve the resilience of the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS
[0118] Figure 1 To improve the IEEE-33 node distribution network structure diagram.
[0119] Figure 2 This is the system load prediction curve.
[0120] Figure 3 This is a comparison chart of out-of-sample performance under different strategies.
[0121] Figure 4 Iteration times and computation time of the PBGS algorithm solution process.
[0122] Figure 5 It is a step flowchart of the present invention. DETAILED DESCRIPTION
[0123] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0124] like Figure 5 As shown, the present invention provides a method for improving the resilience of a distribution network by synergizing multiple resources of source, grid and load in an ice disaster scenario, including:
[0125] Step 1: Combined with the spatiotemporal characteristics of ice disasters, a model of the impact mechanism of ice disaster weather on distribution network lines, distributed generators and residential loads is constructed; the spatiotemporal characteristics of ice disasters can be described by ice disaster trajectory parameters and ice disaster intensity parameters; the impact of ice disasters on distribution network lines is reflected in the probability of line failure; distributed generators include gas units, wind turbines and photovoltaic systems, and the impact of ice disasters on them is reflected in their output size; the impact of ice disasters on residential loads is reflected in their load increment;
[0126] In the present invention, step 1 specifically includes the following sub-steps:
[0127] Step 1.1: Establish the spatiotemporal characteristic model of ice disaster. Equation (1) describes the movement trajectory of the ice disaster center, and equations (2)-(4) describe the distribution of wind speed and freezing rain intensity.
[0128] (1)
[0129] (2)
[0130] (3)
[0131] (4)
[0132] in, The center of the ice disaster is t The geographical coordinates of the moment, represents the initial geographic coordinates of the center of the ice disaster, is the moving speed of the center of the ice disaster, is the angle between the moving direction of the ice disaster center and the true north direction (clockwise is positive), Represents a unit time interval; Represents coordinates ( x , y ) to the center of the ice disaster, and The coordinates are ( x , y ), the freezing rain intensity and wind speed at , They represent the maximum freezing rain intensity and maximum wind speed of the ice disaster, is the radius of the ice disaster.
[0133] Step 1.2: Establish a mechanism model for the impact of ice disasters on distribution lines. Under the influence of ice disasters, the ice growth rate of distribution conductors mainly depends on the wind speed and freezing rain intensity they are subjected to in each period, as shown in formula (5). The ice thickness is obtained by accumulating the ice growth rate, as shown in formula (6). When the combined force borne by the distribution conductor is greater than the upper limit of the combined force borne by the conductor itself, it may cause the distribution conductor to fail. The failure rate of the distribution conductor can be expressed by formula (7). Since the distribution line is composed of multiple sections of conductors connected in series, the failure of any section of the conductor will cause the failure of the entire distribution line. Assuming that the failures of each distribution conductor are independent of each other, the failure probability of the distribution line can be expressed as formula (8);
[0134] (5)
[0135] (6)
[0136] (7)
[0137] (8)
[0138] in, and are the ice growth rate and ice thickness of distribution conductors respectively; and are the freezing rain density and ice density, respectively, and their values are , ; W Represents the liquid water content in the air, which can be expressed by Calculated; , Indicates the threshold of ice thickness that the conductor can withstand; , Distribution conductors l With distribution lines ij exist t The probability of failure at time To form a distribution line ij A collection of power distribution wires.
[0139] Step 1.3: Establish a model of the impact mechanism of ice disasters on distributed generators. Distributed generators include gas generators, wind turbines, and photovoltaic systems.
[0140] For gas units, in low-temperature weather such as ice disasters, natural gas network pipelines are prone to freezing and blockage, which in turn affects the fuel supply of gas turbines in the distribution network, causing the units to be unable to output normally. Assume that during the ice disaster, the natural gas network cannot provide fuel supply for the gas turbine.
[0141] For wind turbines, ice accumulation is prone to occur on the blades under low temperature and high humidity conditions. This phenomenon is related to factors such as meteorological parameters and physical properties of the equipment, and the physical process is relatively complex. Based on the Makkonen theory, a prediction model for the ice thickness of wind turbine blades is established, as shown in equations (9)-(11). For safety reasons, when the ice thickness of the wind turbine blades reaches the safety threshold, the wind turbine needs to be shut down for deicing. The relevant constraints are shown in equation (12). Equation (13) describes the maximum output of the wind turbine in normal operation, and equations (14)-(15) describe the maximum output of the wind turbine in an ice disaster scenario.
[0142] (9)
[0143] (10)
[0144] (11)
[0145] (12)
[0146] (13)
[0147] (14)
[0148] (15)
[0149] in, and are the ice growth rate and ice thickness of wind turbine blades, , are the natural wind speed and relative wind speed at the corresponding locations, , , The collision coefficient, capture coefficient and freezing coefficient are respectively, W is the liquid water content in the air, is the axial inductance, r is the fan blade cross-sectional radius, n is the speed of the fan blades. For safety reasons, when the ice thickness of the fan blades reaches the safety threshold, the fan needs to be shut down for de-icing. Indicates that the fan is t The running status at the moment, a value of 1 indicates running, and a value of 0 indicates shutdown; The upper limit of the safety threshold of ice thickness on wind turbine blades; , are the maximum output of wind turbines in normal operation and ice disaster scenarios, respectively. , , They represent the cut-in wind speed, cut-off wind speed and rated wind speed of the wind turbine respectively. Indicates the rated power of the wind turbine. is the wind turbine power loss coefficient, which is related to the ice thickness of the wind turbine blades Related, , is an empirical parameter.
[0150] For photovoltaic systems, affected by ice disasters, the thickness of snow on the surface of photovoltaic panels will significantly reduce the radiation intensity absorbed by the photovoltaic panels. The photovoltaic output model under the condition of snow accumulation on photovoltaic panels can be expressed as equations (16)-(17).
[0151] (16)
[0152] (17)
[0153] in, is the snow thickness on the photovoltaic panels, is the photovoltaic power loss coefficient, represents the radiation intensity reduction factor, , They are the maximum output of photovoltaic power generation under normal operation and ice disaster scenarios respectively.
[0154] Step 1.4: Establish a mechanism model for the impact of ice disasters on residential loads. During ice disasters, the temperature decreases, leading to an increase in users’ heating demand and a corresponding increase in residential electricity load, which can be described by equations (18)-(19).
[0155] (18)
[0156] (19)
[0157] in, For Node j The original load at is the load increment under the influence of ice disaster, which is related to the body temperature Related, It is the residential load node.
[0158] Step 2: Develop resilience-enhancing strategies based on the differentiated resource requirements for system prevention, response, and recovery.
[0159] Specifically, three types of resilience-enhancing resource configuration plans are formulated in the pre-disaster stage, including: line reinforcement (defense resources), liquefied natural gas (emergency response resources) and maintenance teams (repair resources). Among them, line reinforcement can effectively reduce the probability of line failure and ensure the normal operation of important lines; liquefied natural gas (LNG) can be pre-configured at gas units as temporary fuel to avoid fuel shortages caused by gas pipeline blockage; maintenance teams are used to repair faulty lines to achieve system recovery. During the disaster, when the ice disaster invades the system, as the ice accumulates on the lines, some lines fail, and the output of distributed power sources drops sharply. In order to avoid a significant reduction in system performance, emergency power support is provided by gas unit scheduling on the power supply side, and the grid topology is optimized by reconstructing the distribution network on the power grid side. The maintenance team is dispatched to repair the line in time to avoid long-term power outages of the load; on the load side, the importance of different loads is considered, and differentiated load shedding is adopted to prioritize the supply of important loads. With the cooperation of the above-mentioned multi-type resources, the system load loss is reduced, the system performance is reduced to a lesser extent, and it can be restored to normal earlier, and the resilience of the distribution network can be effectively improved.
[0160] Step 3: Based on the failure probability of the distribution network line in each period, the Monte Carlo method is used to generate multiple distribution network line failure scenarios, and then the typical distribution network failure scenarios are obtained through scenario reduction.
[0161] First, calculate the failure probability of each distribution line at each time , in the scene s Next, for each line ij Generate a random number between [0,1] , and the random number is sequentially compared with each time period ( t =1,2,…, T ) contrast:
[0162] (1) If ,but ;
[0163] (2) If ,but .
[0164] in, Representation scene s Down t Timeline ijThe fault state of the line is 1, which indicates a normal state, and 0 indicates that the line has failed. The Monte Carlo method is used for random simulation to generate multiple random scenarios of distribution network line failures. Different fault scenarios will cause different load loss situations. According to the load loss index, the K-means clustering method is used to cluster the generated scenarios. After clustering, one scenario can be selected from each set to represent all the scenarios in the set, and the number of samples obtained is k A collection of random failure scenarios .
[0165] Step 4: Based on the proposed resilience improvement strategy, a two-stage mixed integer stochastic programming model for multiple fault scenarios is established; the first-stage model aims to minimize the resource allocation cost and optimize the resilience resource allocation scheme, and the constraint condition is the upper limit of the number of resources; the second-stage model aims to minimize the load loss cost and achieve the optimal operation of the distribution network. The constraints include unit operation constraints, distribution network reconstruction constraints, distribution network flow constraints, line maintenance constraints, and line status constraints.
[0166] In the present invention, step 4 specifically includes the following sub-steps:
[0167] Step 4.1: Establish the first stage optimization model.
[0168] In the first stage, a pre-disaster configuration plan for various resilience resources is formulated. The objective function is to minimize the configuration cost. The optimization model for the first stage is:
[0169] (20)
[0170] in, For gas turbines i LNG capacity, The upper limit of LNG resource capacity; Indicates power distribution line ij Whether it is reinforced. A value of 1 indicates that the line is reinforced, and a value of 0 indicates that it is not reinforced. The upper limit of the number of lines that can be reinforced; Indicates the number of configured maintenance teams, The upper limit for the number of maintenance teams; , , They are LNG unit capacity configuration cost, line reinforcement cost and maintenance team configuration cost respectively.
[0171] Step 4.2: Establish the second stage optimization model, including the loss cost objective function and the unit operation constraints, distribution network reconstruction constraints, distribution network flow constraints, line maintenance constraints, and line status constraints.
[0172] The objective function of the second stage model is:
[0173] (twenty one)
[0174] in, Representation scene s Next t Time Node j The load reduction at is the load loss cost, For Node j The weight factor of the load.
[0175] The operating constraints of the gas unit are:
[0176] (twenty two)
[0177] (twenty three)
[0178] (twenty four)
[0179] in, Representation scene s Gas Generator Set i exist t The effort of the moment, is the calorific value of natural gas, is the power generation efficiency of the gas unit, is the conversion factor between LNG and gaseous natural gas, is the ramp rate of the gas unit, It is the upper limit of gas unit output.
[0180] When some lines of the system fail, the distribution network can be reconstructed by changing the state of the distribution network segment switches to provide power supply paths for isolated load nodes and reduce load losses. Since the distribution network usually operates in a radial manner, a single commodity flow model is used to ensure that the distribution network still operates in a radial manner after the switch state is changed, as shown in Equations (25)-(28).
[0181] (25)
[0182] (26)
[0183] (27)
[0184] (28)
[0185] in, Indicates the number of system nodes. For the scene s Medium t time line ijVirtual goods flow on For the scene s middle t Timeline ij The virtual connection status, For the scene s middle t Timeline ij The actual connection status, is the maximum virtual commodity flow on the line, taken as Formula (25) limits the number of closed lines in the system, formula (26) represents the commodity flow balance of the line, formula (27) limits the range of the commodity flow of the line, and formula (28) represents the relationship between the virtual connection state and the actual connection state of the line.
[0186] The distribution network operation constraints are:
[0187] (29)
[0188] (30)
[0189] (31)
[0190] (32)
[0191] (33)
[0192] (34)
[0193] (35)
[0194] Among them, the collection , Node j The child node set and parent node set of , are the node set and the line set respectively; , Respectively for scenes s middle t Time flow through the line ij Active power and reactive power; , Respectively for scenes s middle t The active power and reactive power injected by the substation into the distribution network at all times; , They are t Time Node j Active and reactive load demands at the location; , Respectively for scenes s middle t Time Node j Active and reactive power reduction; , Respectively for scenes s middle t Time Node i , j The voltage amplitude; is the reference voltage; M is a large positive number; , Distribution lines ij resistance and reactance; , are the upper and lower limits of node voltage amplitude respectively, , are the upper limits of active and reactive power of the line, respectively. Equations (29)-(31) are the system power balance constraints, Equation (32) is the voltage drop constraint based on the LinDistFlow model, and Equations (33)-(35) limit the node voltage, line power, and load shedding, respectively.
[0195] For extreme disasters with a short duration, such as typhoons, the system can generally be restored after the disaster ends. However, ice disasters last for a long time. If line maintenance is carried out after the ice disaster is completely over, a large number of loads will be cut off for a long time. Therefore, the present invention applies line maintenance to the mid-disaster stage, that is, the faulty line can be put back into operation after the maintenance is completed, so as to reduce the load loss during the disaster. The relevant constraints are equations (36)-(40).
[0196] (36)
[0197] (37)
[0198] (38)
[0199] (39)
[0200] (40)
[0201] in, Representation scene s Middle Line ij exist t Whether to start maintenance at any time, is the line maintenance time. Formula (36) limits each line to be maintained at most once, Formulas (37)-(38) indicate that the number of lines under maintenance at each moment does not exceed the number of maintenance teams, and Formulas (39)-(40) indicate that maintenance is only carried out after a line failure occurs.
[0202] Due to the continuous impact of the ice disaster, ice may still appear after the line maintenance is completed. In order to ensure that the repaired line does not have a secondary failure, the ice thickness of the repaired line is required to be limited to the first threshold to ensure that the failure probability of the repaired line is always 0. Therefore, the earliest time point when the faulty line can be repaired is The constraint can be expressed as:
[0203] (41)
[0204] (42)
[0205] Comprehensively consider the impact of line reinforcement, line failure and line maintenance, and the distribution network line status The constraint can be expressed as:
[0206] (43)
[0207] (44)
[0208] in and They represent the normal line set and the segment switch set respectively.
[0209] Step 5: Use the robust optimization method to deal with the uncertainty of wind and solar power output, and use the penalty-based Gauss-Seidel method to solve the proposed mixed integer stochastic programming model.
[0210] In the present invention, step 5 includes the following sub-steps:
[0211] Step 5.1: Establish a robust optimization model for wind and solar power output.
[0212] Affected by the ice disaster, the actual maximum output of the new energy unit may deviate from the theoretical value and there is a strong uncertainty. The uncertainty of wind and solar output is limited by the polyhedron uncertainty set, as shown in formula (47): represents an uncertain variable, which refers to the maximum wind and solar power output at each time period under the ice disaster scenario; is the random fluctuation parameter, which represents the fluctuation degree of the uncertain variable; Indicates the fluctuation deviation value of the uncertain variable; is a robust risk indicator with a value range of [0,1]. This indicator determines the fluctuation range of uncertain variables. =1 when the fluctuation is the largest. = 0, the robust model degenerates into a deterministic model. For equations (45) and (46), robust equivalent transformation is used to obtain equation (48),
[0213] (45)
[0214] (46)
[0215] (47)
[0216] (48)
[0217] in, , They are the maximum wind power and photovoltaic output under ice disaster scenarios, respectively. , are the actual grid-connected power of wind power and photovoltaic power respectively. , , , Auxiliary variables introduced for robust equivalence transformation, It is the upper limit of the fluctuation of new energy output.
[0218] Step 5.2: Use penalty-based Gauss-Seidel to solve the proposed mixed integer random programming model.
[0219] The two-stage mixed integer stochastic programming model to be established can be expressed in the following compact form:
[0220] (49)
[0221] (50)
[0222] (51)
[0223] in, and denote the variables in the first and second stages respectively, For the scene s The probability of, Equations (51) and (52) represent the constraints of the first stage and the second stage respectively.
[0224] In order to make the original problem easier to decompose, the original model is rewritten as:
[0225] (52)
[0226] (53)
[0227] (54)
[0228] Among them, formula (53) is an unexpected constraint, which provides a copy of the first-stage variables for each scenario , while introducing auxiliary variables z To ensure that the variables solved in the first stage are applicable to all scenarios in the second stage. Then, the unexpected constraints are relaxed using augmented Lagrangian relaxation to obtain the following model:
[0229] (55)
[0230] In the formula, is the Lagrange multiplier. To ensure that the augmented Lagrange relaxation (55) has a bounded optimal solution, the dual feasible condition must be satisfied, that is, ,therefore Can be set to 0
[34] . is the penalty term. When the penalty term is formula (56), the strong duality of the model can be guaranteed to hold.
[0231] (56)
[0232] in, is a constant greater than 1, For the n The multiplier factor of the iteration, which increases with the number of iterations to improve the convergence of the algorithm; is the penalty factor; , which can be linearized as:
[0233] (57)
[0234] After the above processing, the model is finally transformed into:
[0235] (58)
[0236] Each scene s The corresponding sub-problem can be expressed as:
[0237] (59)
[0238] The specific implementation steps of the penalty-based Gauss-Seidel (PBGS) algorithm are as follows:
[0239] Algorithm 1: PBGS algorithm
[0240] Input Parameters , ,
[0241] for ,
[0242] ,
[0243] for ,
[0244] according to Solve the sub-problems in each scenario in parallel ,get ,
[0245] ,according to Solution ,get ,
[0246] if , then
[0247] , break
[0248] ,
[0249] if or ,
[0250] return ,
[0251] else
[0252] Update penalty factor :
[0253] ,
[0254] .
[0255] Step 6: Apply the proposed distribution network resilience improvement method with source-grid-load multi-resource coordination under ice disaster scenarios to an improved IEEE33-node distribution network for simulation analysis.
[0256] This step specifically includes:
[0257] Step 6.1: Parameter settings
[0258] The simulation example is based on the improved IEEE 33-node system. Figure 1As shown in the figure, the system includes 2 wind turbines, 2 photovoltaics and 3 gas turbines. The rated power of the photovoltaic and wind turbines is 600kW, and the gas turbine capacity is 800kW. The first-level load node is shown as a red dot in the figure. The second-level load is mainly residential electricity load, as shown in the blue shadow in the figure. The rest is the third-level load. The weights of the three loads are set to 1, 0.4, and 0.2 respectively. Assume that the initial center coordinates of the ice disaster are (-60km, -60km), and the center of the ice disaster is along the x-axis. Direction Speed movement, the maximum wind speed of the ice disaster is The maximum freezing rain intensity is The total duration of the ice disaster is T =32h, =1h. Considering the increase in residential electricity load, the system load forecast curve is as follows Figure 2 As shown in Table 1, the resource configuration related parameters are shown in Table 1. Assuming that there are 18 overhead lines in the distribution system, which may fail due to ice disasters, the number of typical scenarios is 20 through line failure scenario generation and reduction. In the PBGS algorithm, and Take it as 0.01, Take it as 1.5, , Take it as 1500.
[0259] Table 1 Resilience resource configuration parameters
[0260]
[0261] Step 6.2: Policy Settings
[0262] In order to verify the effectiveness of the method for improving the resilience of distribution network with multi-resource coordination of sources, grids and loads proposed in this invention, six strategies as shown in Table 2 are designed based on different combinations of various resources.
[0263] Table 2 Policy settings
[0264]
[0265] Step 6.3: Simulate the six strategies separately
[0266] The simulation results are shown in Table 3. Strategy S1 and Strategy S2 only consider single resource configuration, and the load loss cost is high and the total cost is high; Strategy S3-S5 selects two of the three resilience resources for configuration, and the expected load loss cost is less than S1 and S2. Since S3 does not consider line reinforcement measures, the line failure is more serious, and the load loss is greater than S6, and more maintenance teams are needed to restore the line as soon as possible; S4 considers line reinforcement and maintenance team configuration, effectively improving the system's physical defense capability and recovery speed, but the system's distributed power supply cannot provide sufficient emergency power support for the system, causing a certain load loss; S5 has no maintenance team to participate in line recovery, and the line has been in an open state after the fault, causing a long-term power outage for some loads. A larger capacity of LNG needs to be configured to provide more power support and alleviate load losses. Strategy S6 considers the effective coordination of the three resources. Although the resource configuration cost is the highest, the expected load loss cost is much lower than other strategies, and the total cost is the lowest, indicating that this strategy can significantly enhance the resilience of the distribution network.
[0267] Table 3 Comparison of resource allocation results and costs under different strategies
[0268]
[0269] In order to further verify the superiority of the method (S6) of the present invention, 200 additional distribution network fault scenarios are generated as test samples. The configuration results obtained by S3-S6 are applied to each fault scenario to test the out-of-sample performance of the proposed method. Figure 3 The load loss cost box plot is shown in Figure 2. As can be seen from the figure, the load loss cost distribution of each scenario under strategy S6 is significantly better than that of S3-S5, indicating that the proposed multi-resource coordination strategy can better cope with various fault scenarios and has a better resilience improvement effect.
[0270] The proposed two-stage mixed integer stochastic programming model is decomposed into a series of smaller and independent sub-problems for parallel solution by using the scenario decomposition PBGS method. To illustrate the advantages of the algorithm, the PBGS method is compared with the direct call solver solution. When the number of scenarios is 20, the direct call solver to solve the proposed large-scale mixed integer stochastic programming problem cannot obtain the optimization result within 60,000 seconds; when the PBGS algorithm is used for solution, the number of iterations and the solution time of each iteration of this method are as follows Figure 4 As shown in the figure, the number of outer layer iterations is 7, each outer layer iteration contains 2 to 3 inner layer iterations, and the total solution time is 1971 seconds. It can be seen that the solution performance of the PBGS algorithm is significantly better than that of directly calling the solver.
[0271] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for improving the resilience of a distribution network by synergizing multiple resources including source, grid and load in an ice disaster scenario, characterized in that: The following steps are involved: Step 1: Combined with the spatiotemporal characteristics of ice disasters, a model of the impact mechanism of ice disaster weather on distribution network lines, distributed generators and residential loads is constructed; the spatiotemporal characteristics of ice disasters are described by ice disaster trajectory parameters and ice disaster intensity parameters; the impact of ice disasters on distribution network lines is reflected in the probability of line failure; distributed generators include gas units, wind turbines and photovoltaic systems, and the impact of ice disasters on them is reflected in their output size; the impact of ice disasters on residential loads is reflected in their load increment; Step 2: Develop resilience strategies based on differentiated resource requirements for system prevention, response, and recovery. Develop distribution network line reinforcement, LNG capacity configuration, and maintenance team configuration plans in the pre-disaster stage. Minimize load losses during disasters through unit output scheduling, distribution network reconstruction, maintenance team scheduling, and differentiated load shedding resilience strategies. Step 3: Based on the failure probability of the distribution network line in each period, the Monte Carlo method is used to generate multiple distribution network line failure scenarios, and then the typical distribution network failure scenarios are obtained through scenario reduction; Step 4: Based on the proposed resilience improvement strategy, a two-stage mixed integer stochastic programming model for multiple fault scenarios is established; the first-stage model aims to minimize the resource allocation cost and optimize the resilience resource allocation scheme, and the constraint condition is the upper limit of the number of resources; the second-stage model aims to minimize the load loss cost and achieve the optimal operation of the distribution network, and the constraints include unit operation constraints, distribution network reconstruction constraints, distribution network flow constraints, line maintenance constraints, and line status constraints; Step 5: Use a robust optimization method to deal with the uncertainty of wind and solar power output, and use the penalty-based Gauss-Seidel method to solve the proposed mixed integer stochastic programming model; In step 1, the impact mechanism model of the ice disaster on the distributed generator includes the impact mechanism models of the gas unit, the wind turbine unit and the photovoltaic system; for the gas unit, in the low temperature weather of the ice disaster, the natural gas network pipeline is prone to freezing and blockage, thereby affecting the fuel supply of the gas turbine of the distribution network, resulting in the unit being unable to output normally; it is assumed that during the ice disaster period, the natural gas network cannot provide fuel supply for the gas turbine; The impact mechanism model of ice disaster on wind turbines is as follows: (9) (10) (11) (12) (13) (14) (15) in, and are the ice growth rate and ice thickness of wind turbine blades, , are the natural wind speed and relative wind speed at the corresponding locations, , , are the collision coefficient, capture coefficient and freezing coefficient respectively, W is the liquid water content in the air, is the axial inductance, r is the fan blade cross-sectional radius, n is the speed of the fan blades. For safety reasons, when the ice thickness of the fan blades reaches the safety threshold, the fan needs to be shut down for de-icing. Indicates that the fan is t The running status at the moment, a value of 1 indicates running, and a value of 0 indicates shutdown; The upper limit of the safety threshold of ice thickness on wind turbine blades; , are the maximum output of wind turbines in normal operation and ice disaster scenarios, respectively. , , They represent the cut-in wind speed, cut-off wind speed and rated wind speed of the wind turbine respectively. Indicates the rated power of the wind turbine. is the wind turbine power loss coefficient, which is related to the ice thickness of the wind turbine blades Related, , is an empirical parameter; The impact mechanism model of the ice disaster on the photovoltaic system is: (16) (17) in, is the snow thickness on the photovoltaic panels, is the photovoltaic power loss coefficient, represents the radiation intensity reduction factor, , They are the maximum output of photovoltaic power generation under normal operation and ice disaster scenarios respectively.
2. According to claim 1, a method for improving the resilience of a distribution network by synergizing multiple resources of source, grid and load in an ice disaster scenario, characterized in that: In step 1, the temporal and spatial characteristic model of ice disaster is: (1) (2) (3) (4) in, The center of the ice disaster is t The geographical coordinates of the moment, represents the initial geographic coordinates of the center of the ice disaster, is the moving speed of the center of the ice disaster; It is the angle between the moving direction of the ice disaster center and the true north direction, with the clockwise direction being positive; Represents a unit time interval; Represents coordinates ( x , y ) to the center of the ice disaster, and The coordinates are ( x , y ), the freezing rain intensity and wind speed at , They represent the maximum freezing rain intensity and maximum wind speed of the ice disaster, is the radius of the ice disaster.
3. According to claim 1, a method for improving the resilience of a distribution network by synergizing multiple resources of source, grid and load in an ice disaster scenario, characterized in that: In step 1, the impact mechanism model of the ice disaster on the distribution line is: (5) (6) (7) (8) in, and are the ice growth rate and ice thickness of distribution conductors respectively; and are the freezing rain density and ice density, respectively, and their values are , ; W Represents the liquid water content in the air, which can be expressed by Calculated; , Indicates the threshold of ice thickness that the conductor can withstand; , Distribution conductors l With distribution lines ij exist t The probability of failure at time To form a distribution line ij A collection of power distribution wires.
4. According to the method for improving the resilience of a distribution network with source-grid-load multi-resource coordination in an ice disaster scenario in claim 1, in step 1, the impact mechanism model of the ice disaster on residential load is: (18) (19) in, For Node j The original load at is the load increment under the influence of ice disaster, which is related to the body temperature Related, It is the residential load node.
5. According to claim 1, a method for improving the resilience of a distribution network by synergizing multiple resources of source, grid and load in an ice disaster scenario, characterized in that: In step 2, the resilience enhancement strategy is that, in the first stage, the pre-disaster configuration of three resilience enhancement resources is considered, including: line reinforcement, liquefied natural gas and maintenance team; among them, line reinforcement can effectively reduce the probability of line failure and ensure the normal operation of important lines; liquefied natural gas can be pre-configured at the gas unit as temporary fuel to avoid fuel shortage caused by natural gas pipeline blockage; the maintenance team is used to repair the faulty line to achieve system recovery; the second stage is the optimized operation of the distribution network during the disaster under the coordination of multiple resources. After the ice disaster invades the system, as the ice accumulates on the line, some lines fail, and the output of distributed power sources drops sharply. In order to avoid a significant reduction in system performance, emergency power support is provided by dispatching gas units on the power supply side, and the grid topology is optimized by reconstructing the distribution network on the power grid side, and the maintenance team is dispatched to repair the line in time to avoid long-term power outages of the load; on the load side, the importance of different loads is considered, and differentiated load shedding is adopted to give priority to the supply of important loads; with the cooperation of multiple resources, the system performance is less degraded and can be restored to normal earlier, and the resilience of the distribution network can be effectively improved.
6. The method for improving the resilience of a distribution network by synergizing multiple resources of source, grid and load in an ice disaster scenario according to claim 1 is characterized in that: In step 3, the scene generation and reduction method is: First, calculate the failure probability of each distribution line at each time , in the scene s Next, for each line ij Generate a random number between [0,1] , and the random number is sequentially compared with each time period ( t =1,2,…, T ) contrast: (1) If ,but ; (2) If ,but ; in, Representation scene s Down t Timeline ij The fault state is 1, which indicates a normal state, and 0 indicates that the line has failed. The Monte Carlo method is used for random simulation to generate multiple random scenarios of distribution network line failures. Different fault scenarios will cause different load loss situations. According to the load loss index, the K-means clustering method is used to cluster the generated scenarios. After clustering, one scenario can be selected from each set to represent all the scenarios in the set, and the number of samples obtained is k A collection of random failure scenarios .
7. The method for improving the resilience of a distribution network by synergizing multiple resources of source, grid and load in an ice disaster scenario according to claim 1 is characterized in that: In step 4, the mixed integer stochastic programming model includes a first stage and a second stage; In the first stage, a pre-disaster configuration plan for various resilience resources is formulated. The objective function is to minimize the configuration cost. The optimization model for the first stage is: (20) in, For gas turbines i LNG capacity, The upper limit of LNG resource capacity; Indicates power distribution line ij Whether it is reinforced. A value of 1 indicates that the line is reinforced, and a value of 0 indicates that it is not reinforced. The upper limit of the number of lines that can be reinforced; Indicates the number of configured maintenance teams, The upper limit for the number of maintenance teams; , , They are LNG unit capacity configuration cost, line reinforcement cost and maintenance team configuration cost respectively.
8. The method for improving the resilience of a distribution network by synergizing multiple resources of source, grid and load in an ice disaster scenario according to claim 7 is characterized in that: The second-stage model includes an objective function and constraints; The objective function of the second stage model is: (21) in, Representation scene s Next t Time Node j The load reduction at is the load loss cost, For Node j The weight factor of the load; The operating constraints of the gas unit are: (22) (23) (24) in, Representation scene s Gas Generator Set i exist t The effort of the moment, is the calorific value of natural gas, is the power generation efficiency of the gas unit, is the conversion factor between LNG and gaseous natural gas, is the ramp rate of the gas unit, The upper limit of gas unit output; The distribution network reconstruction constraints are: (25) (26) (27) (28) in, Indicates the number of system nodes. For the scene s Medium t time line ij Virtual commodity flow on For the scene s middle t Timeline ij The virtual connection status, For the scene s middle t Timeline ij The actual connection status, is the maximum virtual commodity flow on the line, taken as ; Formula (25) limits the number of closed lines in the system, Formula (26) represents the commodity flow balance of the line, Formula (27) limits the range of the commodity flow of the line, and Formula (28) represents the relationship between the virtual connection state and the actual connection state of the line; The distribution network operation constraints are: (29) (30) (31) (32) (33) (34) (35) Among them, the collection , Node j The child node set and parent node set of , are the node set and the line set respectively; , Respectively for scenes s middle t Time flow through the line ij Active power and reactive power; , Respectively for scenes s middle t The active power and reactive power injected by the substation into the distribution network at all times; , They are t Time Node j Active and reactive load demands at the location; , Respectively for scenes s middle t Time Node j Active and reactive power reduction; , Respectively for scenes s middle t Time Node i , j The voltage amplitude of is the reference voltage; M is a large positive number; , Distribution lines ij resistance and reactance; , are the upper and lower limits of node voltage amplitude respectively, , are the upper limits of active and reactive power of the line respectively; Equations (29)-(31) are the system power balance constraints, Equation (32) is the voltage drop constraint based on the LinDistFlow model, and Equations (33)-(35) respectively limit the node voltage, line power and load shedding size; The line maintenance constraints are: (36) (37) (38) (39) (40) in, Representation scene s Middle Line ij exist t Whether to start maintenance at any time. is the line maintenance time; Formula (36) limits each line to be repaired at most once, Formulas (37)-(38) indicate that the number of lines under maintenance at each time does not exceed the number of maintenance teams, and Formulas (39)-(40) indicate that maintenance is only carried out after a line failure occurs; Due to the continuous impact of ice disasters, ice may still appear after the line maintenance is completed; In order to ensure that the repaired line does not suffer a secondary failure, the ice thickness of the repaired line is required to be limited to the first threshold to ensure that the failure probability of the repaired line is always 0; Therefore, the earliest time point for the faulty line to start maintenance is The constraint expression is: (41) (42) Comprehensively consider the impact of line reinforcement, line failure and line maintenance, and the distribution network line status The constraint can be expressed as: (43) (44) in and They represent the normal line set and the segment switch set respectively.
9. The method for improving the resilience of a distribution network by synergizing multiple resources of source, grid and load in an ice disaster scenario according to claim 1 is characterized in that: In step 5, the wind and solar output robust optimization model is: (45) (46) (47) (48) in, , They are the maximum wind power and photovoltaic output under ice disaster scenarios, respectively. , are the actual grid-connected power of wind power and photovoltaic power, respectively. Affected by the ice disaster, the actual maximum output of the new energy unit may deviate from the theoretical value, and there is a strong uncertainty. The polyhedron uncertainty set is used to limit the uncertainty of wind and photovoltaic output, as shown in formula (47). represents an uncertain variable, which refers to the maximum wind and solar power output at each time period under the ice disaster scenario; is the random fluctuation parameter, which represents the fluctuation degree of the uncertain variable; Indicates the fluctuation deviation value of the uncertain variable; is a robust risk indicator with a value range of [0,1]. This indicator determines the fluctuation range of uncertain variables. =1 when the fluctuation is the largest. = 0, the robust model degenerates into a deterministic model; for equations (45) and (46), robust equivalent transformation is used to obtain equation (48), where: , , , Auxiliary variables introduced for robust equivalence conversion, It is the upper limit of the fluctuation of new energy output.
10. The method for improving the resilience of a distribution network by synergizing multiple resources of source, grid and load in an ice disaster scenario according to claim 1, characterized in that: In step 5, the penalty-based Gauss-Seidel method is as follows: The established two-stage mixed integer stochastic programming model is expressed in the following compact form: (49) (50) (51) in, and denote the variables in the first and second stages respectively, For the scene s The probability of, Equations (51) and (52) represent the constraints of the first stage and the second stage respectively; In order to make the original problem easier to decompose, the original model is rewritten as: (52) (53) (54) Among them, formula (53) is an unexpected constraint, which provides a copy of the first-stage variables for each scenario , while introducing auxiliary variables z To ensure that the variables solved in the first stage are applicable to all scenarios in the second stage; then, the unexpected constraints are relaxed using augmented Lagrangian relaxation to obtain the following model: (55) In the formula, is the Lagrange multiplier. To ensure that the augmented Lagrange relaxation (55) has a bounded optimal solution, the dual feasible condition must be satisfied, that is, ,therefore Can be set to 0 [34] ; is the penalty term. When the penalty term is formula (56), the strong duality of the model can be guaranteed to hold. (56) in, is a constant greater than 1, For the n The multiplier factor of the iteration, which increases with the number of iterations to improve the convergence of the algorithm; is the penalty factor; , which can be linearized as: (57) After the above processing, the model is finally transformed into: (58) Each scene s The corresponding sub-problem can be expressed as: (59) The specific implementation steps of the penalty-based Gauss-Seidel (PBGS) algorithm are as follows: Algorithm 1: PBGS algorithm Input Parameters , , , , , ; for , , for , according to Solve the sub-problems in each scenario in parallel ,get , ,according to Solution ,get , if , then , break , if or , return , else Update penalty factor : , 。
Citation Information
Patent Citations
All-time toughness enhancing method for power transmission system under ice disaster
CN112001625A
Toughness control method of power transmission and distribution system in ice disaster scene
CN115833090A
Power system toughness control and improvement method based on extreme ice disaster
CN115907269A
Resource optimization configuration method for multi-stage improvement of toughness of power distribution-traffic system
CN116720358A
Wind energy forecasting method with extreme wind speed prediction function
US20120046917A1
