Multi-region decentralized scheduling method based on demand flexibility aggregation

Abstract

The invention relates to the technical field of power system optimization scheduling, in particular to a multi-region decentralized scheduling method based on demand flexibility aggregation, which comprises the following steps: (1) grouping and clustering transferable loads and interruptible loads to generate a flexible load aggregation model; (2) constructing an optimal scheduling model of the multi-region power system based on the flexible load aggregation model; and (3) decomposing each load cluster according to the optimal scheduling model of the multi-region power system, and ensuring that the decomposed load scheduling plan conforms to the user preference and system constraint of each load. According to the multi-region decentralized scheduling method based on demand flexibility aggregation, the scale of the scheduling problem is reduced through load aggregation, it is ensured that the scheduling result conforms to user preference and system constraints through load decomposition, the calculation complexity can be effectively reduced, and the renewable energy consumption capacity and economical efficiency of the system are improved.

Description

Technical Field

[0001] The present invention relates to the field of power system optimization, and in particular to a multi-region decentralized scheduling method based on demand flexibility aggregation. Background Art

[0002] With the large-scale access of renewable energy (such as wind and solar energy), the operation of the power system faces new challenges. Although the traditional centralized dispatching method is simple, it has problems with data privacy and high computational complexity in multi-regional power systems. In addition, the participation of flexible loads (such as transferable loads and interruptible loads) provides new dispatching flexibility for the power system, but due to the large number and dispersion of loads, direct participation in system dispatch will lead to a sharp increase in computational complexity. Although existing load aggregation methods can reduce the scale of the dispatching problem, they often cannot effectively maintain user preferences and load characteristics when decomposing the load, resulting in inaccurate dispatching results.

[0003] Therefore, there is an urgent need for an optimal dispatching method that can effectively aggregate and decompose loads in multi-regional power systems while maintaining user preferences and system constraints. Summary of the Invention

[0004] The main objective of the present invention is to provide a multi-region decentralized scheduling method based on demand flexibility aggregation, which can achieve efficient scheduling of multi-region power systems by reducing computational complexity and reducing coordination requirements between regions.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A multi-region decentralized scheduling method based on demand flexibility aggregation is characterized by comprising the following steps:

[0007] (1) Grouping and clustering transferable loads (SLs) and interruptible loads (DLs) to generate load clusters, and extracting representative parameters of each load cluster to generate a flexible load aggregation model;

[0008] (2) Based on the flexible load aggregation model, an optimal dispatch model for the multi-regional power system is constructed, with the goal of minimizing the system's operating costs, including distributed generation costs, renewable energy curtailment costs, and load transfer costs;

[0009] (3) Decompose each load cluster according to the optimized dispatch model of the multi-region power system to ensure that the decomposed load dispatch plan meets the user preferences and system constraints of each load. Specifically, step (1) includes the following steps:

[0010] (1-1) Transferable load model;

[0011] The shiftable loads (SLs) considered are independent task loads that cannot be interrupted after they start running, but their running time can be adjusted within a certain available time range. For an individual load i, its operating characteristics can be described by the following parameters:

[0012] The power demand of the load at time t can be expressed as:

[0013]

[0014] in, Represents the load power demand at time t: P rated,i Indicates the rated power of the load; u i,t Indicates the switch state at time t, with a value of {0,1}; T start,i ,T stop,i Respectively represent the start and stop time of the load.

[0015] The energy demand of the load at time t satisfies the following constraints:

[0016]

[0017] in, It represents the total running time (hours) of load i, which must be completed within the available time window.

[0018] To ensure that the load running time meets the requirements, the model also needs to meet the following minimum continuous running time constraints:

[0019]

[0020] This constraint ensures that if load i starts at time t (i.e. u i,t -u i,t-1 ), it must run continuously for at least Hour.

[0021] For the aggregation of transferable loads, the upper and lower power constraints must be met:

[0022]

[0023] in, represents the equivalent power demand (kW) of load cluster c at time t. and They represent the minimum and maximum power limits of cluster c, respectively, generated by the clustering algorithm; represents the available time window of cluster c.

[0024] The aggregation model is established based on the above single load model, and its aggregate power model is:

[0025]

[0026] in, is the total power of cluster c at time t; Ω represents the load set included

[0027] To support scheduling optimization, the aggregated load model is redefined as an equivalent model:

[0028]

[0029] in, and are the upper and lower limits of the power of the aggregated load at time t; T start,c and T stop,c are the start time and end time of the cluster respectively.

[0030] To support scheduling optimization, the aggregation model is re-adjusted to an equivalent model with the following constraints:

[0031] P min,c ·u c,t ≤P c,t ≤P max,c ·u c,t ,t∈T avail,c (8)

[0032]

[0033] Among them, P min,c and P max,c They represent the minimum and maximum power requirements of the cluster load respectively; u c,t represents the operating status of cluster c at time t. The equivalent model ensures the flexibility of cluster load and provides a simplified load description for subsequent scheduling optimization.

[0034] (1-2) Deferrable Load (DL) Model

[0035] These loads can be interrupted and shifted within a specific available time range, but must meet the maximum available time limit. For example, electric vehicle (EV) charging loads can be interrupted, but must complete charging before the latest departure time. For a given time range t, the demand response (DR) power of load j is expressed as Their arrival and departure times are T arrival,j and T departure,j , a single interruptible load model can be expressed as

[0036]

[0037] Among them, P rated,jrepresents the rated power of load j; [T arrival,j ,T departure,j ] indicates the available time range of the load.

[0038]

[0039] in, represents the total energy demand of load j; Δt represents the time interval.

[0040] For a group of interruptible loads, the aggregated power is the sum of the individual load powers, as shown below:

[0041]

[0042] Where Δt m represents the time slot of minute time scale; P load and P w They represent the forecast load and wind power in minute time scale respectively; Indicates the power fluctuation of RES.

[0043] The aggregation model can be expressed as:

[0044]

[0045] Similar to the previous transferable load, the energy balance constraints that need to be satisfied are:

[0046]

[0047] The aggregated equivalent model has a discrete power range to support load scheduling.

[0048]

[0049] in, Indicates the maximum power of the load after aggregation; [T arrival,d ,T departure,d ] represents the available time range of the aggregate load. It can be adjusted within the maximum power range, but all energy requirements of the load must be completed before the departure time. The lower limit of is 0, because DL load is interruptible and can be delayed to the next time interval.

[0050] (1-3) Construction of flexible load aggregation model;

[0051] The aggregation of flexible loads is divided into two steps. In the first step, the loads are divided according to the similarity of their available time. After the grouping is completed, a characteristic curve is constructed based on the power demand of the load during the working hours. This characteristic curve includes the following attributes: available time, power demand, working time demand, and start and stop time. In the second step, the characteristic curve is clustered using the k-means clustering algorithm to obtain the clustering result c, so that the number of clusters satisfies |c|<|i|. k-means++ is used to select the initial clustering point, thus overcoming the defect of traditional k-means' reliance on the initial center point. The specific steps are as follows:

[0052] 1. Randomly and uniformly select from the consumption vector X i Select the initial cluster center X c ;

[0053] 2. Select the next cluster center according to the following probability distribution:

[0054]

[0055] Among them, D(X i ) represents the data point X i The shortest distance to the current nearest cluster center.

[0056] 3. Repeat step 2 until k cluster centers are selected.

[0057] 4. Assign each data point to the cluster center closest to it and get the Euclidean distance:

[0058]

[0059] Among them, X i is the load data point, C k is the kth cluster center.

[0060] 5. Calculate the new center point C of each cluster k , as the mean of all points in the cluster:

[0061]

[0062] Repeat steps 4 and 5 until any of the following conditions is met: the change in the cluster center position is less than the set threshold; the maximum number of iterations is reached.

[0063] Specifically, step (2) includes the following steps:

[0064] Based on the flexible load aggregation model, an optimal dispatch model for a multi-regional power system is constructed. The goal is to minimize the system's operating costs, including distributed generation costs, renewable energy curtailment costs, and load transfer costs. The specific objective function expression is as follows:

[0065]

[0066] The power generation cost is calculated by optimizing the output of distributed generators (DG) g,t , reducing the fuel consumption and operation and maintenance costs of traditional power generation resources. g is the marginal cost ($ / kWh) of unit g, reflecting its economic priority; the curtailment cost is calculated through the penalty coefficient W curt and V curt ($ / kWh), minimizing curtailment of wind and solar power and Thereby increasing the absorption rate of renewable energy; load regulation costs are reduced through price signals ($ / kWh), quantifying the cost of scheduling deviations between shiftable and interruptible loads DR , encouraging load aggregators to flexibly adjust electricity consumption plans.

[0067] Accordingly, the objective function needs to satisfy the following constraints:

[0068] 1. Network constraints include node power balance constraints and line power flow constraints. The specific expressions are as follows:

[0069]

[0070] Among them, the power generation and consumption of each node n need to be balanced in real time, including distributed generation P g,t , renewable energy output and Fixed load P d,n,t , and through line P nm,t This constraint ensures the real-time stability of the system operation; the power flow model uses DC power flow approximation, and the line power P nm,t The node phase angle difference θ n,t -θ m,t and line susceptance B nm Determine,simplify the computational complexity; It is the maximum allowable phase angle difference of line n→m, which prevents line overload or voltage instability.

[0071] 2. Regional autonomy and coordination constraints ensure that each region independently optimizes local resources. The specific expression is as follows:

[0072]

[0073] Among them, the power P from area a to b ab,t With reverse power P ba,t Conservation relations must be met to avoid redundant energy calculations; the switching power must not exceed the rated capacity of the line. Prevent equipment damage; total power purchase from the main grid in all regions Need to be lower than the mainnet supply limit Ensure the security of mainnet operation

[0074] Specifically, the step (3) includes the following steps:

[0075] According to the optimal dispatch model of the multi-regional power system, each load cluster is decomposed to ensure that the decomposed load dispatch plan meets the user preferences and system constraints of each load. The specific objective function is as follows:

[0076]

[0077] The decomposition model is constructed by minimizing the reference plan With actual load The absolute deviation δ c,t , ensuring the feasibility of aggregate scheduling; each load i must be completed within the available window hours of operation; if the load starts at time tt, it needs to run continuously for at least hours to avoid damage to equipment due to frequent starting and stopping.

[0078] The beneficial effects of the present invention are as follows: the multi-region decentralized scheduling method based on demand flexibility aggregation proposed in the present invention reduces the scale of the scheduling problem through load aggregation, and ensures that the scheduling results meet user preferences and system constraints through load decomposition, which can effectively reduce the computational complexity and improve the system's renewable energy absorption capacity and economy. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 It is a flow chart of the overall method of the present invention. DETAILED DESCRIPTION

[0080] The present invention provides a multi-region decentralized scheduling method based on demand flexibility aggregation, and each step is described in detail below.

[0081] Step 1: Group and cluster the transferable loads (SLs) and interruptible loads (DLs) to generate load clusters, and extract the representative parameters of each load cluster to generate a flexible load aggregation model;

[0082] (1-1) Transferable load model;

[0083] The shiftable loads (SLs) considered are independent task loads that cannot be interrupted after they start running, but their running time can be adjusted within a certain available time range. For an individual load i, its operating characteristics can be described by the following parameters:

[0084] The power demand of the load at time t can be expressed as:

[0085]

[0086] in, Represents the load power demand at time t: P rated,i Indicates the rated power of the load; u i,t Indicates the switch state at time t, with a value of {0,1}; T start,i ,T stop,i Respectively represent the start and stop time of the load.

[0087] The energy demand of the load at time t satisfies the following constraints:

[0088]

[0089] in, It represents the total running time (hours) of load i, which must be completed within the available time window.

[0090] To ensure that the load running time meets the requirements, the model also needs to meet the following minimum continuous running time constraints:

[0091]

[0092] This constraint ensures that if load i starts at time t (i.e. u i,t -u i,t-1 ), it must run continuously for at least Hour.

[0093] For the aggregation of transferable loads, the upper and lower power constraints must be met:

[0094]

[0095] in, represents the equivalent power demand (kW) of load cluster c at time t. and They represent the minimum and maximum power limits of cluster c, respectively, generated by the clustering algorithm; represents the available time window of cluster c.

[0096] The aggregation model is established based on the above single load model, and its aggregate power model is:

[0097]

[0098] in, is the total power of cluster c at time t; Ω represents the load set included

[0099] To support scheduling optimization, the aggregated load model is redefined as an equivalent model:

[0100]

[0101] in, and are the upper and lower limits of the power of the aggregated load at time t; T start,c and T stop,c are the start time and end time of the cluster respectively.

[0102] To support scheduling optimization, the aggregation model is re-adjusted to an equivalent model with the following constraints:

[0103] P min,c ·u c,t ≤P c,t ≤P max,c ·u c,t ,t∈T avail,c (8)

[0104]

[0105] Among them, P min,c and P max,c They represent the minimum and maximum power requirements of the cluster load respectively; u c,t represents the operating status of cluster c at time t. The equivalent model ensures the flexibility of cluster load and provides a simplified load description for subsequent scheduling optimization.

[0106] (1-2) Deferrable Load (DL) Model

[0107] These loads can be interrupted and shifted within a specific available time range, but must meet the maximum available time limit. For example, electric vehicle (EV) charging loads can be interrupted, but must complete charging before the latest departure time. For a given time range t, the demand response (DR) power of load j is expressed as Their arrival and departure times are T arrival,j and T departure,j , a single interruptible load model can be expressed as

[0108]

[0109] Among them, P rated,j represents the rated power of load j; [T arrival,j ,T departure,j ] indicates the available time range of the load.

[0110]

[0111] in, represents the total energy demand of load j; Δt represents the time interval.

[0112] For a group of interruptible loads, the aggregated power is the sum of the individual load powers, as shown below:

[0113]

[0114] Where Δt m represents the time slot of minute time scale; P load and P w They represent the forecast load and wind power in minute time scale respectively; Indicates the power fluctuation of RES.

[0115] The aggregation model can be expressed as:

[0116]

[0117] Similar to the previous transferable load, the energy balance constraints that need to be satisfied are:

[0118]

[0119] The aggregated equivalent model has a discrete power range to support load scheduling.

[0120]

[0121] in, Indicates the maximum power of the load after aggregation; [T arrival,d ,T departure,d ] represents the available time range of the aggregate load. It can be adjusted within the maximum power range, but all energy requirements of the load must be completed before the departure time. The lower limit of is 0, because DL load is interruptible and can be delayed to the next time interval.

[0122] (1-3) Construction of flexible load aggregation model;

[0123] The aggregation of flexible loads is divided into two steps. In the first step, the loads are divided according to the similarity of their available time. After the grouping is completed, a characteristic curve is constructed based on the power demand of the load during the working hours. This characteristic curve includes the following attributes: available time, power demand, working time demand, and start and stop time. In the second step, the characteristic curve is clustered using the k-means clustering algorithm to obtain the clustering result c, so that the number of clusters satisfies |c|<|i|. k-means++ is used to select the initial clustering point, thus overcoming the defect of traditional k-means' reliance on the initial center point. The specific steps are as follows:

[0124] 1. Randomly and uniformly select from the consumption vector X i Select the initial cluster center X c ;

[0125] 2. Select the next cluster center according to the following probability distribution:

[0126]

[0127] Among them, D(X i ) represents the data point X i The shortest distance to the current nearest cluster center.

[0128] 3. Repeat step 2 until k cluster centers are selected.

[0129] 4. Assign each data point to the cluster center closest to it and get the Euclidean distance:

[0130]

[0131] Among them, X i is the load data point, C k is the kth cluster center.

[0132] 5. Calculate the new center point C of each cluster k , as the mean of all points in the cluster:

[0133]

[0134] Repeat steps 4 and 5 until any of the following conditions is met: the change in the cluster center position is less than the set threshold; the maximum number of iterations is reached.

[0135] Step 2: Based on the flexible load aggregation model, an optimal dispatch model for the multi-regional power system is constructed. The goal is to minimize the system's operating costs, including distributed generation costs, renewable energy curtailment costs, and load shifting costs.

[0136] Based on the flexible load aggregation model, an optimal dispatch model for a multi-regional power system is constructed. The goal is to minimize the system's operating costs, including distributed generation costs, renewable energy curtailment costs, and load transfer costs. The specific objective function expression is as follows:

[0137]

[0138] The power generation cost is calculated by optimizing the output of distributed generators (DG) g,t , reducing the fuel consumption and operation and maintenance costs of traditional power generation resources. gis the marginal cost ($ / kWh) of unit g, reflecting its economic priority; the curtailment cost is calculated through the penalty coefficient W curt and V curt ($ / kWh), minimizing curtailment of wind and solar power and , thereby increasing the absorption rate of renewable energy; load regulation costs are reduced through price signals ($ / kWh), quantifying the cost of scheduling deviations between shiftable and interruptible loads DR , encouraging load aggregators to flexibly adjust electricity consumption plans.

[0139] Accordingly, the objective function needs to satisfy the following constraints:

[0140] 1. Network constraints include node power balance constraints and line power flow constraints. The specific expressions are as follows:

[0141]

[0142] Among them, the power generation and consumption of each node n need to be balanced in real time, including distributed generation P g,t , renewable energy output and Fixed load P d,n,t , and through line P nm,t This constraint ensures the real-time stability of the system operation; the power flow model uses DC power flow approximation, and the line power P nm,t The node phase angle difference θ n,t -θ m,t and line susceptance B nm Determine,simplify the computational complexity; It is the maximum allowable phase angle difference of line n→m, which prevents line overload or voltage instability.

[0143] 2. Regional autonomy and coordination constraints ensure that each region independently optimizes local resources. The specific expression is as follows:

[0144]

[0145] Among them, the power P from area a to b ab,t With reverse power P ba,t Conservation relations must be met to avoid redundant energy calculations; the switching power must not exceed the rated capacity of the line. , to prevent equipment damage; the total amount of electricity purchased from the main grid in all regions Need to be lower than the mainnet supply limit Ensure the security of mainnet operation

[0146] Step 3: Decompose each load cluster according to the optimal dispatch model of the multi-regional power system to ensure that the decomposed load dispatch plan meets the user preferences and system constraints of each load.

[0147] According to the optimal dispatch model of the multi-regional power system, each load cluster is decomposed to ensure that the decomposed load dispatch plan meets the user preferences and system constraints of each load. The specific objective function is as follows:

[0148]

[0149] The decomposition model is constructed by minimizing the reference plan With actual load The absolute deviation δ c,t , ensuring the feasibility of aggregate scheduling; each load i must be completed within the available window hours of operation; if the load starts at time tt, it needs to run continuously for at least hours to avoid damage to equipment due to frequent starting and stopping.

[0150] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be included therein. Any reference sign in a claim should not be construed as limiting the claim to which it relates.

[0151] In addition, it should be understood that although this specification is described in terms of implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A multi-region decentralized scheduling method based on demand flexibility aggregation, characterized in that: The following steps are involved: (1) Grouping and clustering transferable and interruptible loads to generate load clusters, and extracting representative parameters of each load cluster to generate a flexible load aggregation model; (2) Based on the flexible load aggregation model, an optimal dispatch model for the multi-regional power system is constructed, with the goal of minimizing the system's operating costs, including distributed generation costs, renewable energy curtailment costs, and load transfer costs; (3) Decomposing each load cluster according to the optimal dispatching model of the multi-regional power system so that the decomposed load dispatching plan meets the user preference and system constraints of each load.

2. A multi-region decentralized scheduling method based on demand flexibility aggregation according to claim 1, characterized in that: The step (1) is carried out as follows: (1-1) Transferable load model; The transferable loads are independent task loads that cannot be interrupted after starting to run, but can adjust their running time within a certain available time range. For an individual load i, its running characteristics are described by the following parameters: The power demand of the load at time t is expressed as: in, Represents the load power demand at time t: P rated,i Indicates the rated power of the load; u i,t Indicates the switch state at time t, with a value of {0,1}; T start,i ,T stop,i Respectively represent the start and stop time of the load. The energy demand of the load at time t satisfies the following constraints: in, represents the total running time of load i, completed within the available time window; The model satisfies the following minimum continuous runtime constraints: Constraint (3) ensures that if load i starts at time t, that is, u i,t -u i,t-1 , it must run continuously for at least Hour; For the aggregation of transferable loads, the upper and lower power constraints are met: in, represents the equivalent power demand of load cluster c at time t, P and They represent the minimum and maximum power limits of cluster c, respectively, generated by the clustering algorithm; represents the available time window of cluster c; The aggregation model is established based on the above single load model, and its aggregate power model is: in, is the total power of cluster c at time t; Ω represents the set of included loads; The aggregated load model is redefined as an equivalent model to support dispatch optimization: in, and are the upper and lower limits of the power of the aggregated load at time t; T start，c and T stop，c are the start and end times of clustering respectively; The aggregation model is re-adjusted to an equivalent model to support scheduling optimization, with the following constraints: P min,c ·in c,t ≤P c,t ≤P max,c ·in c,t ,t∈T avail,c (8) Among them, P min,c and P max,c They represent the minimum and maximum power requirements of the cluster load respectively; u c,t represents the running status of cluster c at time t; (1-2) Interruptible load model Interruptible loads can be interrupted and shifted within a set available time range. For a given time range t, the demand response DR power of load j is expressed as Their arrival and departure times are T arrival,j and T departure,j , a single interruptible load model is expressed as Among them, P rated,j represents the rated power of load j; [T arrival,j ,T departure,j ] represents the available time range of the load; in, represents the total energy demand of load j; Δt represents the time interval; For a group of interruptible loads, the aggregated power is the sum of the individual load powers, as shown below: Where Δt m represents the time slot of minute time scale; P load and P w They represent the forecast load and wind power in minute time scale respectively; Indicates the power fluctuation of RES; The aggregation model is represented as: The energy balance constraints satisfied are: The aggregated equivalent model has a discrete power range to support load scheduling; in, Indicates the maximum power of the load after aggregation; [T arrival,d ,T departure,d ] represents the available time range of the aggregate load, Able to adjust within the maximum power range, but all energy requirements of the load must be met before the departure time, The lower limit of is 0; (1-3) Construction of flexible load aggregation model; Aggregation of flexible loads involves two steps. First, loads are grouped based on their availability. After grouping, characteristic curves are constructed based on the power requirements of the loads during their working hours. The characteristic curves include the following attributes: availability, power requirements, working time requirements, and start and stop times. Second, the characteristic curves are clustered using the k-means clustering algorithm to obtain clustering results c, such that the number of clusters satisfies |c| < |i|. k-means++ is used to select the initial clustering points. The specific steps are as follows:

1. Randomly and uniformly select from the consumption vector X i Select the initial cluster center X c ; 2. Select the next cluster center according to the following probability distribution: Among them, D(X i ) represents the data point X i The shortest distance to the current nearest cluster center; 3. Repeat step 2 until k cluster centers are selected; 4. Assign each data point to the cluster center closest to it and get the Euclidean distance: Among them, X i is the load data point, C k is the kth cluster center; 5. Calculate the new center point C of each cluster k , as the mean of all points in the cluster: Repeat steps 4 and 5 until any of the following conditions is met: the change in the cluster center position is less than the set threshold; the maximum number of iterations is reached.

3. A multi-region decentralized scheduling method based on demand flexibility aggregation according to claim 2, characterized in that: The step (2) is carried out as follows: Based on the flexible load aggregation model, an optimal dispatch model for a multi-regional power system is constructed. The goal is to minimize the system's operating costs, including distributed generation costs, renewable energy curtailment costs, and load transfer costs. The specific objective function expression is as follows: Among them, the power generation cost is calculated by optimizing the output P of the distributed generator DG. g,t , reducing the fuel consumption and operation and maintenance costs of traditional power generation resources; among them, Π g is the marginal cost of unit g, reflecting its economic priority; the curtailment cost is calculated through the penalty coefficient W curt and V curt , minimizing the amount of curtailed wind and photovoltaic power and Thereby increasing the absorption rate of renewable energy; load regulation costs are reduced through price signals Quantify the cost of scheduling deviations between shiftable and interruptible loads DR , encouraging load aggregators to flexibly adjust electricity consumption plans; The objective function must satisfy the following constraints:

1. Network constraints include node power balance constraints and line power flow constraints. The specific expressions are as follows: Among them, the power generation and consumption of each node n need to be balanced in real time, including distributed generation P g,t , renewable energy output and Fixed load P d,n,t , and through line P nm,t The power exchange; the power flow model uses DC power flow approximation, the line power P nm,t The node phase angle difference θ n,t -θ m,t and line susceptance B nm Determine,simplify the computational complexity; The maximum permissible phase angle difference of line n→m is used to prevent line overload or voltage instability.

2. Regional autonomy and coordination constraints ensure that each region independently optimizes local resources. The specific expression is as follows: Among them, the power P from area a to b ab,t With reverse power P ba,t Satisfy conservation relations; the exchange power shall not exceed the rated capacity of the line Total electricity purchased from the main network in all regions Below the mainnet supply cap 4. A multi-region decentralized scheduling method based on demand flexibility aggregation according to claim 3, characterized in that: The step (3) is carried out as follows: According to the optimal dispatch model of the multi-regional power system, each load cluster is decomposed to ensure that the decomposed load dispatch plan meets the user preferences and system constraints of each load. The specific objective function is as follows: The decomposition model is constructed by minimizing the reference plan With actual load The absolute deviation δ c,t , ensuring the feasibility of aggregate scheduling; each load i is completed within the available window hours of operation; if the load starts at time tt, it needs to run continuously for at least Hour.

Images