A joint optimization scheduling method, system, device and medium

Abstract

The present invention discloses a joint optimization scheduling method, system, equipment and medium, which relate to the field of power system optimization scheduling; the method comprises: obtaining power grid data of the power system's distribution network after a disaster and road fault information in a traffic network corresponding to the distribution network; constructing an optimization scheduling model; solving a first objective function according to a first constraint condition to obtain a first solution value; inputting the first solution value into a correction model, solving a second objective function according to a second constraint condition to obtain a first solution correction value; adjusting the power grid data of the power system using the first solution correction value to obtain a scheduling plan; the scheduling plan is used to allocate photovoltaic output and dispatchable resource configuration information to restore power supply after a disaster; the present invention quickly realizes the restoration of power supply after a disaster and reduces losses caused by power outages by jointly scheduling dispatchable resource configuration information and photovoltaic prediction information.

Description

Technical Field

[0001] The present invention relates to the field of power system optimization and dispatching, and in particular to a joint optimization and dispatching method, system, equipment and medium. Background Art

[0002] Large-scale power outages often result in the loss of main grid power supply to distribution networks and the failure of multiple lines within the network, leading to the loss of power to a significant number of critical and common loads. To mitigate the harm caused by large-scale power outages in distribution networks, it is necessary to develop reasonable and effective post-disaster emergency dispatch methods for distribution networks. Current research on post-disaster emergency dispatch for distribution networks typically focuses on distributed generation (DG) such as wind and solar power, as well as mobile emergency resources such as repair crews (RC) and mobile energy storage systems (MESS). These methods often lack consideration of DG output uncertainty and incorporate intelligent soft open points (SOPs) for optimal dispatch. This results in a weak ability of these dispatch methods to address DG output uncertainty and overlooks the SOP's ability to rapidly adjust power and provide voltage and reactive power support. Furthermore, in terms of time, given the varying response speeds of different dispatchable resources and the varying accuracy of DG output predictions on long and short time scales, emergency dispatch based on a single time scale struggles to coordinate multiple emergency resources to participate in distribution network power restoration, resulting in a slow restoration process and increased losses from power outages. Summary of the Invention

[0003] The purpose of the present invention is to provide a joint optimization scheduling method, system, device and medium, which can quickly restore power supply after a disaster and reduce losses caused by power outages by jointly scheduling schedulable resource configuration information and photovoltaic forecast information.

[0004] To achieve the above object, the present invention provides the following solutions:

[0005] A joint optimization scheduling method, the method comprising:

[0006] Obtaining post-disaster grid data of the power system's distribution network and road fault information in the transportation network corresponding to the distribution network; the grid data includes: dispatchable resource configuration information, photovoltaic forecast information at a first time scale, and photovoltaic forecast information at a second time scale; the period of the first time scale is greater than the period of the second time scale; the photovoltaic forecast information includes: photovoltaic output and load demand; the dispatchable resource configuration information includes: mobile energy storage, number of emergency repair personnel, and converter configuration;

[0007] Build an optimized scheduling model;

[0008] The optimization scheduling model includes: an optimization model and a correction model; the optimization model includes: a first objective function and a first constraint condition; the correction model includes: a second objective function and a second constraint condition;

[0009] The first objective function is constructed based on the photovoltaic forecast information at the first time scale and the dispatchable resource configuration information, with the goal of minimizing the total cost of the power system; the first constraint conditions include: dispatchable resource configuration information constraint, distribution network topology constraint, and distribution network flow constraint;

[0010] The second objective function is constructed based on the photovoltaic forecast information at the second time scale and the dispatchable resource configuration information, with the goal of minimizing power outage losses, minimizing the correction amount of mobile energy storage, and minimizing the correction amount of converter configuration; the second constraint conditions include: dispatchable resource configuration information constraint, distribution network topology constraint, and distribution network power flow constraint;

[0011] Solving the first objective function according to the first constraint condition to obtain a first solution value; the first solution value is the corresponding mobile energy storage output and the transmission power of the converter configuration when the total cost of the power system is minimized;

[0012] Inputting the first solution value into the correction model, solving the second objective function with the second constraint condition, and obtaining a first solution correction value;

[0013] The first solved correction value is used to adjust the grid data of the power system to obtain a scheduling plan; the scheduling plan is used to allocate photovoltaic output and dispatchable resource configuration information to restore power supply after the disaster.

[0014] Optionally, the dispatchable resource configuration information constraints include: mobile energy storage scheduling constraints, mobile energy storage operation constraints, mobile energy storage energy constraints, mobile energy storage state of charge upper and lower limit constraints, mobile energy storage charging and discharging constraints, emergency repair personnel scheduling constraints, fault repair situation constraints and converter configuration operation constraints.

[0015] Optionally, the expression of the first objective function is:

[0016]

[0017] Where Z is the integer decision variable output by the optimization model; is the photovoltaic output of load node i at time t; Ω T is the time period set; I is the load node set; ω i,c is the importance coefficient of load node i; α i,t represents the load shedding state at load node i at time t; represents the active power reduction of load node i at time t; B is the branch set of the distribution network; ε is the tie switch operation cost coefficient; α i-j,t is the open or closed state of the line (ij) between the load node i and the adjacent load node j at time t; W is the decision variable output by the optimization model.

[0018] Optionally, the expression of the second objective function is:

[0019]

[0020] Among them, Y is the decision variable output by the correction model; T S is the number of time periods in a cycle; k′ is the time sequence number in the scheduling process; t′ is a time in the cycle; i is the load node; I is the load node set; ω i,c is the importance coefficient of node i; Θ is the grid data set; l is the number of the grid data; κ is the penalty coefficient for the optimization adjustment of mobile energy storage and converter configuration; represents the active power reduction of load node i at time k′+t′; represents the decision value of the lth power grid data at time k′+t′ in the optimization model; ΔP l (k′+t′) represents the output correction value of the lth power grid data at time k′+t′; P 0,l (k′+t′-1) represents the initial output value of the lth power grid data at time k′+t′-1.

[0021] A joint optimization scheduling system, comprising:

[0022] An acquisition module is configured to acquire post-disaster grid data of a power system's distribution network and road fault information in a transportation network corresponding to the distribution network; the grid data includes dispatchable resource configuration information, photovoltaic forecast information at a first time scale, and photovoltaic forecast information at a second time scale; the period of the first time scale is greater than the period of the second time scale; the photovoltaic forecast information includes photovoltaic output and load demand; the dispatchable resource configuration information includes mobile energy storage, the number of emergency repair personnel, and converter configuration;

[0023] Model building module, used to build optimization scheduling model;

[0024] The optimization scheduling model includes: an optimization model and a correction model; the optimization model includes: a first objective function and a first constraint condition; the correction model includes: a second objective function and a second constraint condition;

[0025] The first objective function is constructed based on the photovoltaic forecast information at the first time scale and the dispatchable resource configuration information, with the goal of minimizing the total cost of the power system; the first constraint conditions include: dispatchable resource configuration information constraint, distribution network topology constraint, and distribution network flow constraint;

[0026] The second objective function is constructed based on the photovoltaic forecast information at the second time scale and the dispatchable resource configuration information, with the goal of minimizing power outage losses, minimizing the correction amount of mobile energy storage, and minimizing the correction amount of converter configuration; the second constraint conditions include: dispatchable resource configuration information constraint, distribution network topology constraint, and distribution network power flow constraint;

[0027] A first solving module is configured to solve the first objective function according to the first constraint condition to obtain a first solution value; the first solution value is the corresponding mobile energy storage output and the transmission power of the converter configuration when the total cost of the power system is minimized;

[0028] a second solving module, configured to input the first solved value into the correction model, solve the second objective function with the second constraint condition, and obtain a first solved correction value;

[0029] A determination module is used to adjust the grid data of the power system using the first solution correction value to obtain a scheduling plan; the scheduling plan is used to allocate photovoltaic output and dispatchable resource configuration information to restore power supply after a disaster.

[0030] A device includes a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to perform the above-mentioned joint optimization scheduling method.

[0031] A medium stores a computer program, which implements the above-mentioned joint optimization scheduling method when executed by a processor.

[0032] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0033] The present invention provides a joint optimization scheduling method, system, device and medium, which constructs a first objective function based on photovoltaic prediction information and dispatchable resource configuration information at a first time scale, constructs a second objective function based on photovoltaic prediction information and dispatchable resource configuration information at a second time scale, then solves the first objective function based on a first constraint condition, inputs the obtained first solution value into a correction model, and then solves the second objective function based on the second constraint condition to determine the scheduling scheme; the present invention solves the problem that a single time scale is difficult to coordinate by adopting the first time scale and the second time scale, and performs joint scheduling based on the two dimensions of dispatchable resource configuration information and photovoltaic prediction information, which can improve the scheduling process of post-disaster power supply restoration. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0035] Figure 1 A flowchart of the joint optimization scheduling method provided by an embodiment of the present invention;

[0036] Figure 2 This is a structural diagram of the joint optimization scheduling system provided by an embodiment of the present invention.

[0037] Explanation of symbols:

[0038] Acquisition module-1, model building module-2, first solution module-3, second solution module-4, determination module-5. DETAILED DESCRIPTION

[0039] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only 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 making creative efforts are within the scope of protection of the present invention.

[0040] The present invention integrates the two dimensions of object and time, and considers the uncertain output of distributed generation (DG) such as wind and solar power, and jointly optimizes the scheduling method with mobile emergency resources and intelligent soft open points (SOPs). This can further improve the power supply restoration effect and reduce the losses caused by large-scale power outages in the distribution network.

[0041] The purpose of the present invention is to provide a joint optimization scheduling method, system, device and medium, which can improve the scheduling process of post-disaster power supply restoration by jointly scheduling schedulable resource configuration information and photovoltaic forecast information.

[0042] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0043] Example 1

[0044] like Figure 1 As shown, an embodiment of the present invention provides a joint optimization scheduling method, which includes:

[0045] Step 100: Obtain post-disaster grid data for the power system's distribution network and road fault information in the transportation network corresponding to the distribution network; the grid data includes dispatchable resource configuration information, photovoltaic forecast information at a first time scale, and photovoltaic forecast information at a second time scale. The period of the first time scale is greater than the period of the second time scale. For example, day-ahead photovoltaic forecast information and short-term photovoltaic forecast information for 24 hours after the disaster. Photovoltaic forecast information includes photovoltaic output and load demand; road fault information includes faulty lines and repair time; specifically, road fault information may also include distribution network topology information, a collection of impassable roads in the transportation network, length information for each road section, and traffic density information. Dispatched resource configuration information includes mobile energy storage, the number of repair personnel, and converter configuration. Converter configuration may be the configuration of intelligent soft switches within the network.

[0046] In other words, by collecting post-disaster road fault information, photovoltaic output forecast information, and dispatchable resource configuration information, the traffic network topology laid along the distribution network is abstracted and the time required for mobile emergency resources to travel at different times and on different road sections is further calculated.

[0047] Specifically, the time required for mobile emergency resources to travel is estimated based on the length of different sections of the traffic network and the traffic density of different sections at different times. This operation serves as a preparatory work for the subsequent specific emergency dispatch model.

[0048] Then, the transportation network topology is constructed and the travel time of mobile emergency resources is estimated.

[0049] Assuming the traffic network is laid out along the power distribution network, intersections are equivalent to nodes, and road sections are equivalent to lines containing distance information, resulting in an abstract traffic network topology. Based on the improved Greenshields model density-speed relationship formula, the travel speed of mobile emergency resources at different times is obtained:

[0050]

[0051] Among them, dt represents the traffic density at time t; d max with d min Respectively represent the maximum and minimum values ​​of traffic density; v(d t ) indicates that the vehicle is at d t Driving speed under traffic density, v free With v min They represent the vehicle's speed in free state and in blocked state respectively, and a and b are both constants.

[0052] The estimated travel time is calculated based on the ratio of distance to speed, and the travel time is corrected for the impact of road damage on vehicle driving conditions. Then, the estimated travel time T required for vehicle e to travel between any nodes i and j is obtained through weighted summation. e,i,j .

[0053]

[0054]

[0055] Among them, l i-j represents the length of the road section between adjacent nodes, t i-j For vehicles in d t The travel time under the traffic density. i-j is regarded as the weight of the road segment ij at time t, at which time nodes i and j are adjacent nodes; Indicates rounding up to an integer. M is an arbitrarily large constant. T e,i,j is the estimated travel time of vehicle e between nodes i and j; C(i,j) represents the set of road sections that the vehicle passes through from node i to node j. In this case, nodes i and j are not necessarily adjacent nodes.

[0056] Step 200: Construct an optimization scheduling model; wherein the optimization scheduling model includes: an optimization model and a correction model; the optimization model includes: a first objective function and a first constraint condition; the correction model includes: a second objective function and a second constraint condition.

[0057] The first objective function is constructed based on the photovoltaic forecast information and dispatchable resource configuration information of the first time scale, with the goal of minimizing the total cost of the power system; the first constraint conditions include: dispatchable resource configuration information constraints, distribution network topology constraints and distribution network flow constraints.

[0058] The second objective function is constructed based on the photovoltaic forecast information and dispatchable resource configuration information at the second time scale, with the goals of minimizing power outage losses, minimizing the correction amount of mobile energy storage, and minimizing the correction amount of converter configuration; the second constraint conditions include: dispatchable resource configuration information constraints, distribution network topology constraints, and distribution network flow constraints.

[0059] Specifically, the optimization model is based on long timescales, such as optimizing continuously for 24 hours at one-hour intervals. The model uses the day-ahead forecast of post-disaster PV output (the forecast on the first timescale) as input, constructing a robust model for the joint optimization of mobile emergency resources and SOPs that accounts for uncertainty. The optimization results serve as a benchmark to guide model refinement. A two-stage robust optimization model is established within the optimization model to address the uncertainty of PV output. The model's decision-making using integer variables in the first stage and continuous variables in the second stage addresses the problem of low solution efficiency caused by the large number of integer variables in the model.

[0060] The modified model is based on a short time scale, for example, it can be executed every 15 minutes and optimized continuously for 2 hours at 15-minute intervals.

[0061] Taking into account that the predicted value of photovoltaic output on the first time scale and the predicted value on the second time scale in the same period may be relatively close, in order to reduce the calculation time and prevent unnecessary frequent small adjustments to the mobile energy storage and SOP, before executing the correction model each time, it will be judged whether the indicator of the change of the two photovoltaic prediction values ​​reaches the preset start value. If so, the correction model will be started, and its correction value will be obtained based on the correction amount of the mobile energy storage output and SOP transmission power in the next 15 minutes.

[0062] The correction model is a short-term rolling optimization model based on model predictive control (MPC). This model uses the short-term forecast of photovoltaic output (i.e., the forecast value at the second time scale) as input, the actual measured values ​​of mobile energy storage and SOP output as initial values, and the corrections to mobile energy storage output and SOP transmission power within a future finite time domain as decision variables. It performs rolling optimization within a finite time domain and mainly includes three parts: model prediction, rolling optimization, and feedback correction. The proposed correction model optimizes the corrections to mobile energy storage output and SOP transmission power based on the output of the optimization model and the more accurate short-term forecast of photovoltaic output. This prevents frequent and significant adjustments to mobile energy storage and SOP during short-term rolling optimization, while maximizing the effectiveness of power restoration.

[0063] Before executing the correction model each time, it is determined whether the change index of the day-ahead and short-term photovoltaic output forecast values ​​reaches the starting value D. start :

[0064]

[0065] in, is the photovoltaic prediction value change index at time k′; F represents the set of photovoltaic access nodes; t′ represents the t′th period within a lower-level scheduling cycle (2 hours), T S Indicates the number of time periods in a lower-level scheduling cycle (2 hours / 15 minutes = 8); is the day-ahead forecast value of photovoltaic output, It is the short-term forecast value of photovoltaic output.

[0066] Set the starting value D in the correction model start , judged every 15 minutes. At the k′ moment of the revised model, if the predicted value change index Greater than the start value D start , then the corresponding scheduling processing of the modified model is carried out; otherwise, the k′+1 time is judged.

[0067] By solving the decision variables through short-term rolling optimization, the mobile energy storage output and SOP transmission power within a limited time domain in the future are predicted. The prediction model is as follows:

[0068]

[0069] Where: P(k′+τ|k′) is the mobile energy storage output and SOP transmission power predicted at time k′ in the future at time k′+τ; P0(k′) is the initial value of the mobile energy storage and SOP output at time k′, which is obtained from actual measurement; ΔP(k′+t′|k′) is the correction value of the mobile energy storage output and SOP transmission power in the future [k′+(t′-1), k′+t′] period obtained by rolling optimization at time k′, which is the optimization decision variable; τ is the sequence number of the period within a lower-level scheduling cycle.

[0070] Step 300: Solve the first objective function according to the first constraint condition to obtain a first solution value; the first solution value is the corresponding mobile energy storage output and the transmission power of the converter configuration when the total cost of the power system is minimized.

[0071] Considering the uncertainty of the day-ahead photovoltaic forecast value, a two-stage robust optimization objective function, namely the first objective function, is established with the goal of minimizing the sum of power outage losses and tie switch operation costs under the worst photovoltaic output scenario.

[0072] Specifically, the expression of the first objective function is:

[0073]

[0074] Where Z is the integer decision variable output by the optimization model; is the photovoltaic output of load node i at time t; Ω T is the time period set; I is the load node set; ω i,c is the importance coefficient of load node i; α i,t represents the load shedding state at load node i at time t; represents the active power reduction of load node i at time t; B is the branch set of the distribution network; ε is the tie switch operation cost coefficient; αi-j,t is the open or closed state of the line (ij) between the load node i and the adjacent load node j at time t; W is the decision variable output by the optimization model.

[0075] in,

[0076]

[0077]

[0078] They represent the active power reduction and reactive power reduction of load node i at time t respectively; and They represent the charging power and discharging power of the kth mobile energy storage vehicle at time t respectively; They represent the active power and reactive power input from the converter at node i to the load node i at time t of SOP respectively.

[0079] is a 0-1 variable indicating whether the faulty node i′ is repaired by maintenance personnel at time t. If the repair is completed, X k,i,t is a 0-1 variable indicating whether the kth mobile energy storage vehicle is at load node i at time t. If so, then X k,i,t =1;X k,m-n,t is a 0-1 variable indicating whether the kth mobile energy storage vehicle is traveling on the road section mn at time t. If so, then X k,m-n,t =1;Y r,i′,t Y is a 0-1 variable indicating whether the repair personnel of group r are located at the fault node i′ at time t; r,i′,，j′,t is a 0-1 variable indicating whether the repair personnel are traveling between fault nodes i′ and j′ at time t.

[0080] Considering the uncertainty of photovoltaic power output, which will increase the error of photovoltaic output forecast value, the polyhedron uncertainty set U is used to characterize the uncertainty of photovoltaic output based on the photovoltaic output forecast value:

[0081]

[0082] Where: are the photovoltaic output fluctuation value and the upper limit of the fluctuation value at node i at time t, respectively. is the day-ahead forecast value of photovoltaic output, i.e. the forecast value of the first time scale. i,t is a random variable with a value between -1 and 1. Γ represents the uncertainty limit of PV output fluctuations, which can control the total amount of fluctuation and adjust the conservativeness of the model. When Γ is 0, the PV output is assumed to be equal to the predicted value, and the total fluctuation is 0. The larger the value of Γ, the more conservative the model.

[0083] Step 400: Input the first solution value into the correction model, solve the second objective function with the second constraint condition, and obtain the first solution correction value.

[0084] Step 500: Using the first solved correction value to adjust the grid data of the power system to obtain a dispatching plan; the dispatching plan is used to allocate photovoltaic output and dispatchable resource configuration information to restore power supply after the disaster.

[0085] The correction model is based on the optimization model, but does not schedule mobile emergency resources within the transportation network or restructure the network. The lower-level short-term rolling optimization model, or correction model, uses the mobile energy storage and SOP output values ​​determined by the upper-level scheduling, or optimization, model as a reference. Using a 15-minute interval as the objective function, the correction model aims to minimize power outage losses and minimize the correction of mobile energy storage and SOP compared to the upper-level optimization results. This approach prevents frequent and significant adjustments to mobile energy storage and SOP during the short-term rolling optimization process, while maximizing the effectiveness of power restoration.

[0086] Specifically, the expression of the second objective function is:

[0087]

[0088] Among them, Y is the decision variable output by the correction model; T S is the number of time periods in a cycle; k′ is the time sequence number in the scheduling process; t′ is a time in the cycle; i is the load node; I is the load node set; ω i,c is the importance coefficient of node i; Θ is the grid data set; l is the number of the grid data; κ is the penalty coefficient for the optimization adjustment of mobile energy storage and converter configuration; represents the active power reduction of load node i at time k′+t′; represents the decision value of the lth power grid data at time k′+t′ in the optimization model; ΔP l (k′+t′) represents the output correction value of the lth power grid data at time k′+t′; P 0,l (k′+t′-1) represents the initial output value of the lth power grid data at time k′+t′-1.

[0089]

[0090] are the charging and discharging power correction values ​​of mobile energy storage respectively; They are the correction amounts for the active and reactive power transmitted by SOP respectively.

[0091] Among them, the dispatchable resource configuration information constraints include: mobile energy storage dispatch constraints, mobile energy storage operation constraints, mobile energy storage energy constraints, mobile energy storage state of charge upper and lower limit constraints, mobile energy storage charging and discharging constraints, emergency repair personnel dispatch constraints, fault repair situation constraints, and converter configuration operation constraints.

[0092] Specifically, based on the mobility characteristics of mobile energy storage in the transportation network, mobile energy storage scheduling constraints are constructed.

[0093] Continuity constraint for mobile energy storage. When the mobile energy storage enters a node or is already at the node at time t (corresponding to the right side of the following formula), the mobile energy storage can choose to leave the node or continue to stay at the node at time t+1 (corresponding to the left side of the following formula):

[0094]

[0095] Where: represents the set of all possible destinations of mobile energy storage starting from node j; X represents the set of all possible starting points of the mobile energy storage with node j as the end point. k,i,j,t is a 0-1 variable indicating whether the mobile energy storage k is traveling between nodes i and j at time t. If so, the value is 1; X k,i,t is a 0-1 variable indicating whether the mobile energy storage k is located at node i at time t. If so, the value is 1.

[0096] The same mobile energy storage can only be in two states at a single moment: driving or staying at a certain node:

[0097]

[0098] During the travel between nodes and pre-operation preparation period, the mobile energy storage is considered to have not reached the destination, that is, charging and discharging operations cannot be performed:

[0099]

[0100] Where: t σ It represents the operation preparation time after the mobile energy storage reaches the destination, which is a certain constant.

[0101] The same mobile energy storage can only stay at one node at a time:

[0102]

[0103] Regarding the operating constraints of mobile energy storage, the details are as follows:

[0104] Mobile energy storage charging and discharging state constraints: Mobile energy storage can only be charged and discharged after reaching the destination node, and the charging and discharging states are mutually exclusive.

[0105]

[0106] In the formula: and are 0-1 variables representing the charging and discharging states of the $k$-th mobile energy storage vehicle at node $i$ at time $t$. If in the charging state, then If in the discharging state, then

[0107] Mobile energy storage energy constraint:

[0108]

[0109] In the formula: $E$ k,t is the battery capacity of the $k$-th mobile energy storage vehicle at time $t$; and respectively represent the charging and discharging powers of the $k$-th mobile energy storage vehicle at time $t$; $\eta$ ch and $\eta$ dch respectively represent the charging and discharging efficiencies of the mobile energy storage vehicle; $\Delta t$ is the time interval.

[0110] Mobile energy storage state of charge upper and lower limit constraints:

[0111]

[0112] In the formula: $s_{SOC}^{max}$ max and $s_{SOC}^{min}$ min respectively represent the maximum and minimum values of the state of charge of the mobile energy storage vehicle; $E$ c is the rated capacity of the mobile energy storage.

[0113] Mobile energy storage charging and discharging constraints:

[0114]

[0115]

[0116] In the formula: and <x respectively represent the charging and discharging powers of the $k$-th mobile energy storage vehicle at time $t$; and respectively represent the upper limits of the charging and discharging powers of the $k$-th mobile energy storage vehicle.

[0117] Regarding the dispatching constraints of emergency repair personnel, specifically as follows:

[0118] Continuity constraint for emergency repair personnel driving. When an emergency repair person drives into a certain node at time $t$ or is already at this node (corresponding to the right side of the following formula), this emergency repair team can only choose to drive away from this node or continue to stay at this node at time $t +$ 1 (corresponding to the left side of the following formula):

[0119]

[0120] Where: i′ is the fault node in the network, I fault is the set of all faulty nodes in the network, r represents the rth group of emergency repair personnel; Y r,i′,t Y is a 0-1 variable indicating whether the repair personnel of group r are located at the fault node i′ at time t; r,i′,j′,t is a 0-1 variable indicating whether the repair personnel are traveling between fault nodes i′ and j′ at time t.

[0121] The same group of repair personnel can only be in two states at a single moment: driving or staying at a certain node to carry out repairs:

[0122]

[0123] If the repair personnel are driving between two nodes or have not completed the repair at the previous node, they cannot appear at the next node:

[0124]

[0125] Where: It represents the time required for the rth group of repair personnel to repair fault i′, which is a certain constant.

[0126] The same group of repair workers can only stay at one fault node at a time:

[0127]

[0128] The constraints on fault repair are as follows:

[0129] If the time required to repair fault i′ is not reached, fault i′ is considered not repaired:

[0130]

[0131] Where: is a 0-1 variable indicating whether the fault node i′ has been repaired. If the repair is completed,

[0132] Assume that the failed node will not fail again after repair is completed:

[0133]

[0134] It is assumed that the line can be closed only when the faults at both ends of the faulty line are repaired:

[0135]

[0136] Where: B faultis the set of fault branches, (i′-j′) represents the fault line between adjacent fault nodes i′ and jv; α i-j,t is a 0-1 variable indicating whether the line (ij) between adjacent nodes is open or closed at time t. If the line (ij) is closed, α i-j,t =1.

[0137] Any fault can only be repaired once:

[0138]

[0139] Regarding the intelligent soft switch SOP operating constraints:

[0140] The intelligent soft openpoint (SOP) considered in the embodiment of the present invention is a double-terminal back-to-back voltage-sourced converter (B2B-VSC). As a fully controlled power electronic device, the SOP has the advantages of continuously regulating power and providing reactive voltage support. In the event of an accident, the SOP usually operates in the VdcQ-Vf control mode, that is, the energized normal side operates in the VdcQ mode, with voltage and frequency support provided by a power supply with self-starting capability; the power failure side operates in the Vf control mode. At this time, the fault-side VSC can provide voltage and frequency support and input active and reactive power to the fault-side power grid, further improving the power supply recovery effect and reducing the losses caused by large-scale power outages in the distribution network.

[0141] Active power balance constraints of intelligent soft switching:

[0142]

[0143]

[0144] Where: represents the active power injected by SOP into node i at time t; represents the reactive power injected by SOP to node i at time t; H represents the set of SOP access nodes; is the active power loss of the VSC on side i; is the active power loss coefficient of the VSC on side i. SOP reactive power output and access capacity constraints:

[0145]

[0146]

[0147] Where: are the upper and lower limits of reactive power injected by SOP to node i respectively; Represents the capacity of the SOP installed between nodes i and j.

[0148] For the nonlinear relationship in the SOP operation constraint above, it can be converted into a second-order cone constraint by the second-order cone relaxation method:

[0149]

[0150] The network topology constraints are as follows:

[0151] Island division constraint: The subgrid where photovoltaic or mobile energy storage is used for power supply at node i is called an island.

[0152] Determine whether each node in the network has a power supply, that is, determine the source node:

[0153]

[0154] Φ j =1,j∈F

[0155] Where: Φ j Indicates whether node j has power supply, Φ j =1 indicates that node j is connected to mobile energy storage or photovoltaic power source, and j is called the source node; G is the set of nodes that mobile energy storage can access; F is the set of photovoltaic access nodes.

[0156] If any node i has a power supply, it must belong to an island, otherwise it may not:

[0157]

[0158] Where: φ i,s,t is the island partition variable. When node i belongs to island s at time t, φ i,s,t =1, denoted as i∈s; M is a large constant.

[0159] Determine that when the parent node of a node belongs to an island, its child nodes can belong to the island:

[0160]

[0161] Where: j-i,t is a 0-1 variable describing the parent-child relationship between adjacent nodes. If j is the parent node of i, χ j-i,t =1;α i-j,t is a 0-1 variable indicating whether the line (ij) between adjacent nodes is open or closed at time t. If the line (ij) is closed, α i-j,t =1.

[0162] Hub-and-spoke network topology constraints:

[0163] Ensure that any pair of adjacent nodes ij has only one parent-child relationship:

[0164] χ i-j,t +χ j-i,t =α i-j,t ;

[0165] Except for the source node, each node has only one parent node at any time:

[0166] ∑ i∈δ(j) χ i-j,t =1,j∈{Φ j =0};

[0167] The source node has no parent node:

[0168] ∑ i∈δ(j) χ i-j,t =0,j∈{Φ j =1};

[0169] Regarding the power flow constraints of the distribution network, the details are as follows:

[0170] Considering the safe operation of the distribution network, the DistFlow equation is used to constrain the power flow. The following constraints must be satisfied at all times, so the subscript t is ignored.

[0171]

[0172]

[0173]

[0174]

[0175]

[0176]

[0177]

[0178] Where: R i-j 、X i-j are branch resistance and branch reactance respectively; are the square of the branch current amplitude and the square of the node voltage amplitude respectively; They are the upper and lower limit constraints of the square of the voltage amplitude and the upper limit constraint of the square of the current amplitude.

[0179] For the above power flow constraints, there is a nonlinear relationship between voltage, current and power in the last equation. The second-order cone relaxation method is used to convert it into a second-order cone constraint:

[0180]

[0181] The constraints of the revised model are the mobile energy storage operation constraints, the mobile soft switch operation constraints, and the distribution network power flow constraints. They are similar to the constraints in the upper-level optimization model and will not be detailed here.

[0182] Through the above short-term rolling optimization, the mobile energy storage output and SOP transmission power correction value vector ΔP(k′+t′|k′) at the future k′+t′ can be obtained at the k′ time, then T S The control sequence matrix composed of the time period correction value vector is:

[0183] [ΔP(k′+1|k′),ΔP(k′+2|k′),...,ΔP(k′+T s |k′)]

[0184] In order to prevent the control process from deviating from the ideal state, the short-term rolling optimization at time k′ only executes the first vector in the control sequence matrix, and obtains the mobile energy storage output and SOP transmission power at time k′+1:

[0185] P(k′+1|k′)=P0(k′)+ΔP(k′+1|k′)

[0186] Considering that the short-term prediction value of photovoltaic output is highly accurate, there is still a certain degree of uncertainty. Therefore, it is impossible to guarantee that the predicted value is the same as the actual output. Therefore, it is also impossible to guarantee that after the optimization scheduling of the previous period, the actual output value of mobile energy storage and SOP is the same as the scheduling value. Therefore, it is necessary to introduce a feedback correction link. At time k'+1, the actual measured output value of mobile energy storage and SOP is used as the initial value of the short-term rolling optimization at time k'+1, forming a closed-loop control:

[0187] P0(k′+1)=P real (k′+1)

[0188] Where: P0(k′+1) represents the initial output value of mobile energy storage and SOP at time k′+1; P real (k′+1) represents the measured value of mobile energy storage and SOP output at time k′+1 obtained through optimal scheduling at time k′.

[0189] The column and constraints generation (C&CG) algorithm can be used to solve two-stage robust optimization models. Specifically, the two-stage robust optimization problem is decomposed into a main problem and subproblems. The main problem solves the first-stage decision variables, while the subproblems determine the compensation variables and uncertainty variables. Cut sets are then returned to the main problem to update the constraints. The C&CG algorithm transforms the inner min problem of the two-layer maxmin subproblem structure into a max problem through duality theory, resulting in a single-layer max problem that is easier to solve. The main problem and subproblems are then solved iteratively until the upper and lower bounds converge and the accuracy-based convergence condition is met.

[0190] The short-term rolling optimization model of the modified model based on model predictive control belongs to a second-order cone programming problem and can be solved using the commercial solver GUROBI.

[0191] Example 2

[0192] like Figure 2 As shown, an embodiment of the present invention provides a joint optimization scheduling system, which includes: an acquisition module 1, a model building module 2, a first solution module 3, a second solution module 4 and a determination module 5.

[0193] Acquisition module 1 is used to obtain the power distribution network data of the power system after the disaster and the road fault information in the transportation network corresponding to the distribution network; the power distribution network data includes: dispatchable resource configuration information, photovoltaic forecast information of the first time scale and photovoltaic forecast information of the second time scale; the period of the first time scale is greater than the period of the second time scale; the photovoltaic forecast information includes: photovoltaic output and load demand; the dispatchable resource configuration information includes: mobile energy storage, the number of emergency repair personnel and converter configuration.

[0194] Model building module 2 is used to build an optimization scheduling model.

[0195] The optimization scheduling model includes: an optimization model and a correction model; the optimization model includes: a first objective function and a first constraint condition; the correction model includes: a second objective function and a second constraint condition.

[0196] The first objective function is constructed based on the photovoltaic forecast information and dispatchable resource configuration information of the first time scale, with the goal of minimizing the total cost of the power system; the first constraint conditions include: dispatchable resource configuration information constraints, distribution network topology constraints and distribution network flow constraints.

[0197] The second objective function is constructed based on the photovoltaic forecast information and dispatchable resource configuration information at the second time scale, with the goals of minimizing power outage losses, minimizing the correction amount of mobile energy storage, and minimizing the correction amount of converter configuration; the second constraint conditions include: dispatchable resource configuration information constraints, distribution network topology constraints, and distribution network flow constraints.

[0198] The first solving module 3 is used to solve the first objective function according to the first constraint condition to obtain a first solution value; the first solution value is the corresponding mobile energy storage output and the transmission power of the converter configuration when the total cost of the power system is minimized.

[0199] The second solving module 4 is used to input the first solution value into the correction model, solve the second objective function with the second constraint condition, and obtain the first solution correction value.

[0200] Determination module 5 is used to adjust the grid data of the power system using the first solution correction value to obtain a scheduling plan; the scheduling plan is used to allocate photovoltaic output and dispatchable resource configuration information to restore power supply after the disaster.

[0201] Example 3

[0202] An embodiment of the present invention provides a device including a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the joint optimization scheduling method in embodiment 1.

[0203] As an optional implementation, the electronic device may be a server.

[0204] In one embodiment, the present invention further provides a medium storing a computer program, which implements the joint optimization scheduling method in embodiment 1 when executed by a processor.

[0205] The benefits of the present invention are:

[0206] 1. The post-disaster emergency dispatch process of the distribution network is divided into optimization and correction according to different time scales. Among them, the optimization model establishes a two-stage robust optimization model for the vehicle scheduling and operation scheduling problems of emergency resources on a long time scale. This makes the proposed dispatching method more capable of coping with the uncertainty of distributed generation (DG) output, and solves the problem of low solution efficiency caused by the large number of integer variables in the optimization model. In addition, the correction model establishes a rolling optimization model on a short time scale based on the model predictive control (MPC) theory, which better solves the problem of decreased accuracy of DG output prediction on a long time scale.

[0207] 2. The smart soft switch (SOP) scheduling resource configuration information is added as the scheduling object to the distribution network post-disaster emergency scheduling model, which fully utilizes the SOP's ability to quickly adjust power and provide voltage and reactive power support during the distribution network power supply restoration process. The proposed scheduling method can further improve the power supply restoration effect and reduce the losses caused by large-scale power outages in the distribution network.

[0208] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0209] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A joint optimization scheduling method, characterized in that: The method comprises: Obtaining post-disaster grid data of the power system's distribution network and road fault information in the transportation network corresponding to the distribution network; the grid data includes: dispatchable resource configuration information, photovoltaic forecast information at a first time scale, and photovoltaic forecast information at a second time scale; the period of the first time scale is greater than the period of the second time scale; the photovoltaic forecast information includes: photovoltaic output and load demand; the dispatchable resource configuration information includes: mobile energy storage, number of emergency repair personnel, and converter configuration; Build an optimized scheduling model; The optimization scheduling model includes: an optimization model and a correction model; the optimization model includes: a first objective function and a first constraint condition; the correction model includes: a second objective function and a second constraint condition; The first objective function is constructed based on the photovoltaic forecast information at the first time scale and the dispatchable resource configuration information, with the goal of minimizing the total cost of the power system; the first constraint conditions include: dispatchable resource configuration information constraint, distribution network topology constraint, and distribution network flow constraint; The second objective function is constructed based on the photovoltaic forecast information at the second time scale and the dispatchable resource configuration information, with the goal of minimizing power outage losses, minimizing the correction amount of mobile energy storage, and minimizing the correction amount of converter configuration; the second constraint conditions include: dispatchable resource configuration information constraint, distribution network topology constraint, and distribution network power flow constraint; Solving the first objective function according to the first constraint condition to obtain a first solution value; the first solution value is the corresponding mobile energy storage output and the transmission power of the converter configuration when the total cost of the power system is minimized; Inputting the first solution value into the correction model, solving the second objective function with the second constraint condition, and obtaining a first solution correction value; The first solved correction value is used to adjust the grid data of the power system to obtain a scheduling plan; the scheduling plan is used to allocate photovoltaic output and dispatchable resource configuration information to restore power supply after the disaster.

2. The joint optimization scheduling method according to claim 1, characterized in that: The dispatchable resource configuration information constraints include: mobile energy storage dispatch constraints, mobile energy storage operation constraints, mobile energy storage energy constraints, mobile energy storage state of charge upper and lower limit constraints, mobile energy storage charging and discharging constraints, emergency repair personnel dispatch constraints, fault repair situation constraints and converter configuration operation constraints.

3. The joint optimization scheduling method according to claim 1, characterized in that: The expression of the first objective function is: Where Z is the integer decision variable output by the optimization model; is the photovoltaic output of load node i at time t; Ω T is the time period set; I is the load node set; ω i,c is the importance coefficient of load node i; α i,t represents the load shedding state at load node i at time t; represents the active power reduction of load node i at time t; B is the branch set of the distribution network; ε is the tie switch operation cost coefficient; α i-j,t is the open or closed state of the line (ij) between the load node i and the adjacent load node j at time t; W is the decision variable output by the optimization model.

4. The joint optimization scheduling method according to claim 1, characterized in that: The expression of the second objective function is: Among them, Y is the decision variable output by the correction model; T S is the number of time periods in a cycle; k′ is the time sequence number in the scheduling process; t′ is a time in the cycle; i is the load node; I is the load node set; ω i,c is the importance coefficient of node i; Θ is the grid data set; l is the number of the grid data; κ is the penalty coefficient for the optimization adjustment of mobile energy storage and converter configuration; represents the active power reduction of load node i at time k'+t'; represents the decision value of the lth power grid data at time k'+t' in the optimization model; ΔP l (k'+t') represents the output correction value of the lth power grid data at time k'+t'; P 0,l (k'+t'-1) represents the initial output value of the lth power grid data at time k'+t'-1.

5. A joint optimization scheduling system, characterized in that: The system comprises: An acquisition module is configured to acquire post-disaster grid data of a power system's distribution network and road fault information in a transportation network corresponding to the distribution network; the grid data includes dispatchable resource configuration information, photovoltaic forecast information at a first time scale, and photovoltaic forecast information at a second time scale; the period of the first time scale is greater than the period of the second time scale; the photovoltaic forecast information includes photovoltaic output and load demand; the dispatchable resource configuration information includes mobile energy storage, the number of emergency repair personnel, and converter configuration; Model building module, used to build optimization scheduling model; The optimization scheduling model includes: an optimization model and a correction model; the optimization model includes: a first objective function and a first constraint condition; the correction model includes: a second objective function and a second constraint condition; The first objective function is constructed based on the photovoltaic forecast information at the first time scale and the dispatchable resource configuration information, with the goal of minimizing the total cost of the power system; the first constraint conditions include: dispatchable resource configuration information constraint, distribution network topology constraint, and distribution network flow constraint; The second objective function is constructed based on the photovoltaic forecast information at the second time scale and the dispatchable resource configuration information, with the goal of minimizing power outage losses, minimizing the correction amount of mobile energy storage, and minimizing the correction amount of converter configuration; the second constraint conditions include: dispatchable resource configuration information constraint, distribution network topology constraint, and distribution network power flow constraint; A first solving module is configured to solve the first objective function according to the first constraint condition to obtain a first solution value; the first solution value is the corresponding mobile energy storage output and the transmission power of the converter configuration when the total cost of the power system is minimized; a second solving module, configured to input the first solved value into the correction model, solve the second objective function with the second constraint condition, and obtain a first solved correction value; A determination module is used to adjust the grid data of the power system using the first solution correction value to obtain a scheduling plan; the scheduling plan is used to allocate photovoltaic output and dispatchable resource configuration information to restore power supply after a disaster.

6. An electronic device, characterized in that: The electronic device comprises a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the joint optimization scheduling method according to any one of claims 1 to 4.

7. A storage medium, characterized in that: It stores a computer program, which, when executed by a processor, implements the joint optimization scheduling method according to any one of claims 1 to 4.

Images