Hierarchical distributed cooperative scheduling method and system for smart rod cluster, and computer program product

Abstract

The invention discloses a layered and distributed cooperative scheduling method and system for a smart rod cluster and a computer program product, and belongs to the technical field of power system energy optimization scheduling, and the method comprises the steps: building a day-ahead scheduling optimization model that smart rod cluster aggregation resources participate in an electric energy market and a peak regulation auxiliary service market at the same time; according to the day-ahead scheduling optimization model, proposing a layered and distributed cooperative scheduling method based on an alternating direction multiplier method scaling form; based on a hierarchical distributed collaborative scheduling method, respectively solving an upper-layer smart rod cluster operator management platform sub-problem and a lower-layer smart rod cluster sub-problem; and after the sub-problems of the upper layer and the lower layer are solved, an aggregation power allocation method considering the flexible resource feasible region of the multi-type smart rod cluster is proposed, and the aggregation power of the smart rod cluster is allocated to each smart rod. Based on the alternating direction multiplier method, the scheduling problem of a large-scale resource cluster can be efficiently solved, higher economy is achieved, and privacy of various resource users is better protected.

Description

Technical Field

[0001] The present application relates to the field of energy optimization scheduling in power systems, and particularly to a hierarchical distributed collaborative scheduling method, system, and computer program product for a smart pole cluster. Background Art

[0002] With the rapid development of smart city construction, as a new type of urban infrastructure, 5G smart poles are driving profound changes in urban management with their characteristics of multi-functional integration, intelligence, and standardization. Based on traditional street lamp poles, 5G smart poles integrate diverse functions such as intelligent lighting, mobile communication, electric vehicle charging, environmental monitoring, traffic management, and information interaction, achieving "multiple poles in one" and effectively optimizing the utilization of public space resources.

[0003] As an important carrier of new infrastructure, the smart energy system of 5G smart poles can not only utilize on-pole energy storage units to achieve peak-valley electricity price arbitrage, but also aggregate various adjustable resources to participate in multi-level power market transactions and build a new flexible and interactive power grid operation mode. However, existing research mainly focuses on the scheduling optimization of single-pole energy storage units and generally adopts an "edge-cloud collaboration" architecture, that is, the operation data of each smart pole needs to be uploaded to the central cloud platform for centralized processing. This architecture faces the problem of combinatorial explosion caused by the access of large-scale pole towers in practical applications, and centralized cloud computing also brings decision-making delay, communication congestion caused by the transmission of massive terminal data, and privacy protection issues between clusters. Summary of the Invention

[0004] The present application aims to provide a hierarchical distributed collaborative scheduling method, system, and computer program product for a smart pole cluster based on the alternating direction method of multipliers, which can efficiently handle the scheduling problem of large-scale resource clusters, realize the participation of large-scale smart pole resource clusters in power grid peak shaving economic scheduling, and protect user privacy.

[0005] To achieve the above object, the technical solution of the present application is:

[0006] A hierarchical distributed collaborative scheduling method for a smart pole cluster includes:

[0007] Establish a day-ahead scheduling optimization model for a smart pole cluster to aggregate resources and participate in the electricity energy market and the peak shaving ancillary service market simultaneously;

[0008] Propose a hierarchical distributed collaborative scheduling method based on the scaled form of the alternating direction method of multipliers according to the day-ahead scheduling optimization model;

[0009] Based on the hierarchical distributed collaborative scheduling method, solve the upper-layer smart pole cluster operator management platform sub-problem and the lower-layer smart pole cluster sub-problem respectively;

[0010] After solving the sub-problems of the upper-layer intelligent pole cluster operator management platform and the lower-layer intelligent pole cluster sub-problems, an aggregated power allocation method considering the flexible resource feasible region of multiple types of intelligent pole clusters is proposed, and the aggregated power of the intelligent pole cluster is allocated to each intelligent pole.

[0011] Optionally, the method for establishing a day-ahead scheduling optimization model for the aggregated resources of the intelligent pole cluster to participate in both the electricity energy market and the peak shaving ancillary service market includes: constructing an economic operation objective function and constraint conditions for the intelligent pole cluster considering the electricity energy market and the peak shaving ancillary service market.

[0012] Optionally, the economic operation objective function of the intelligent pole cluster considering the electricity energy market and the peak shaving ancillary service market is expressed as follows:

[0013]

[0014] Wherein, represents the unit electricity price of the intelligent pole cluster i participating in the day-ahead electricity energy market transaction at time t; represents the day-ahead energy market transaction power of the intelligent pole cluster i at time t; represents the compensation price for the upward peak shaving provided by the intelligent pole cluster i to the power grid in the peak shaving ancillary service market at time t, which is given by the clearing of the ancillary service market; represents the compensation price for the downward peak shaving provided by the intelligent pole cluster i to the power grid in the peak shaving ancillary service market at time t, which is given by the clearing of the peak shaving ancillary service market; represents the upward peak shaving power provided by the intelligent pole cluster i in the peak shaving ancillary service market at time t; represents the downward peak shaving power provided by the intelligent pole cluster i in the peak shaving ancillary service market at time t; represents the upward peak shaving cost of the intelligent pole cluster i at time t; represents the downward peak shaving cost of the intelligent pole cluster i at time t; Δt represents the minimum scheduling period; I represents the number of intelligent pole clusters i; T represents the scheduling period of the intelligent pole cluster i;

[0015] Optionally, the constraint conditions include: peak shaving demand balance constraint, electricity energy market transaction power and energy constraint, upward and downward peak shaving power constraint, and energy change constraint considering upward peak shaving and peak shaving power;

[0016] The peak shaving demand balance constraint is expressed as follows:

[0017]

[0018] Wherein, P t req represents the upward and downward peak shaving demand at time t; Denote the variable of the total peak shaving response coefficient of all intelligent pole clusters at time t; the range of the peak shaving demand balance constraint is [0, 1];

[0019] The power and energy constraints of the electricity energy market trading are expressed as follows:

[0020]

[0021] Wherein, Denote the electricity energy market trading power of intelligent pole cluster i at time t; Denote the lower limit value of the aggregated power of intelligent pole cluster i at time t; Denote the upper limit value of the aggregated power of intelligent pole cluster i at time t; Denote the upper limit value of the capacity of the virtual battery of intelligent pole cluster i at time t; Denote the lower limit value of the capacity of the virtual battery of intelligent pole cluster i at time t; Denote the remaining capacity of the virtual battery of intelligent pole cluster i at time t after participating in the day-ahead electricity energy market trading; Denote the remaining capacity of the virtual battery of intelligent pole cluster i at time t+1 after participating in the day-ahead electricity energy market trading;

[0022] The up-regulation and down-regulation power constraints are expressed as follows:

[0023]

[0024] Wherein, Denote the up-regulation state variable of intelligent pole cluster i at time t, which is a binary variable; Denote the down-regulation state variable of intelligent pole cluster i at time t, which is a binary variable; Denote the upper limit value of the up-regulation power provided by intelligent pole cluster i participating in the peak shaving ancillary service market at time t; Denote the upper limit value of the down-regulation power provided by intelligent pole cluster i participating in the peak shaving ancillary service market at time t;

[0025] The energy change constraint considering the peak shaving power is expressed as follows:

[0026]

[0027] Wherein, Denote the equivalent remaining capacity of the virtual battery of intelligent pole cluster i considering the electricity energy market trading power, up-regulation power and down-regulation power at time t; Denote the equivalent remaining capacity of the virtual battery of intelligent pole cluster i considering the electricity energy market trading power, up-regulation power and down-regulation power at time t+1.

[0028] Optionally, the hierarchical distributed collaborative scheduling method based on the scaled form of the alternating direction multiplier method proposed according to the day-ahead scheduling optimization model includes:

[0029] Introduce decoupling auxiliary variables to decouple the peak shaving demand balance constraint, which is expressed as:

[0030]

[0031] Where represents the decoupled auxiliary variable; represents the decoupled auxiliary variable;

[0032] According to the decoupled peak shaving demand balance constraint, scale the Lagrange multiplier vector for the augmented Lagrangian function of the day-ahead optimization scheduling model constructed based on the scaled form of the alternating direction multiplier method, and obtain the hierarchical distributed collaborative scheduling method based on the scaled form of the alternating direction multiplier method.

[0033] Optionally, the augmented Lagrangian function of the day-ahead optimization scheduling model constructed based on the scaled form of the alternating direction multiplier method is as follows:

[0034]

[0035] Where both contain the optimization vectors of T optimization scheduling periods; represents the decoupled auxiliary variable; λ i is the Lagrange multiplier vector; ρ is the penalty coefficient of the coupling constraint;

[0036] The scaling of the Lagrange multiplier vector includes: using u i = λ i / ρ to scale the Lagrange multiplier vector λ i ; where u i represents the scaled Lagrange multiplier vector.

[0037] Optionally, the lower-layer smart pole cluster sub-problem includes: the self-consistent scheduling among smart pole clusters, and the objective function is expressed as:

[0038]

[0039] Where are all vectors containing T periods, x i represents the decision variable, represents the remaining capacity of the smart pole cluster i after the upward peak shaving power at time t, represents the remaining capacity of the smart pole cluster i after the downward peak shaving power at time t; represents the minimum value of the objective function of the lower-layer smart pole cluster i sub-problem;

[0040] The sub - problems of the upper - layer intelligent pole cluster operator management platform include: issuing peak - shaving instructions to each intelligent pole cluster, and the objective function is expressed as:

[0041]

[0042] Among them, represents the minimum value of the objective function of the sub - problems of the upper - layer intelligent pole cluster operator management platform.

[0043] Optionally, the aggregated power allocation method considering the flexible resource feasible region of multiple types of intelligent pole clusters includes: Step S41 to Step S49.

[0044] Step S41: Input the initial capacity, power feasible region, and capacity feasible region of the equivalent virtual battery of each intelligent pole cluster, and let the current time t = 1;

[0045] Step S42: Calculate the charging and discharging power distribution coefficients of each intelligent pole.

[0046] Step S43: Calculate the charging and discharging allocated power of each intelligent pole and the remaining capacity of the equivalent virtual battery after the allocated power.

[0047] Step S44: Determine whether the allocated power meets the feasible region. If so, let the current time t = t + 1 and enter Step S45; if not, enter Step S46.

[0048] Step S45: Determine whether the current time t satisfies being greater than the scheduling period T of the intelligent pole cluster. If so, enter Step S49; if not, return to Step S42.

[0049] Step S46: Fix the allocated power of the intelligent pole exceeding the feasible region to the boundary value.

[0050] Step S47: Calculate the charging and discharging power distribution coefficients of other intelligent poles.

[0051] Step S48: Calculate the charging and discharging allocated power of each intelligent pole and the remaining capacity of the equivalent virtual battery after the allocated power, and return to Step S42.

[0052] Step S49: The aggregated power allocation ends.

[0053] Optionally, the calculation of the charging and discharging power distribution coefficients of each intelligent pole is expressed as follows:

[0054]

[0055] Among them, represents the discharging power distribution coefficient of the intelligent pole j inside the intelligent pole cluster i at time t, Denote the charging power distribution coefficient of the smart pole j inside the smart pole cluster i at time t; E j,t Denote the remaining capacity of the equivalent virtual battery of the smart pole j at time t; Denote the upper limit value of the feasible region of the equivalent virtual battery of the smart pole j, Denote the lower limit value of the feasible region of the equivalent virtual battery of the smart pole j, and j represents the number of smart poles inside the smart pole cluster;

[0056] The calculation of the charging and discharging distribution power of each smart pole is expressed as follows:

[0057]

[0058] Among them, Denote the discharging power of the smart pole j inside the smart pole cluster i at time t, Denote the charging power of the smart pole j inside the smart pole cluster i at time t; Denote the total trading power of the smart pole cluster i participating in the day-ahead electricity energy market and the peaking auxiliary service market.

[0059] A hierarchical distributed collaborative scheduling system for a smart pole cluster, which executes a hierarchical distributed collaborative scheduling method for a smart pole cluster as described in any one of the above, includes:

[0060] A day-ahead scheduling optimization module, a hierarchical distributed collaborative scheduling module, a sub-problem solving module, and an aggregated power distribution module;

[0061] The day-ahead scheduling optimization module, the hierarchical distributed collaborative scheduling module, the sub-problem solving module, and the aggregated power distribution module are connected in sequence.

[0062] A computer program product, including a computer program, which implements a hierarchical distributed collaborative scheduling method for a smart pole cluster as described in any one of the above when executed by a processor.

[0063] The hierarchical distributed collaborative scheduling method, system, and computer program product for the smart pole cluster proposed in this application are based on the alternating direction multiplier method, and only need to interact very little data information between each smart pole cluster and the smart pole cluster operator management platform, reducing the communication transmission data volume and transmission time, significantly improving the calculation efficiency and response speed, being able to efficiently handle the scheduling problems of large-scale resource clusters, realizing the independent and autonomous scheduling of the aggregated power of each cluster, reducing the risk of cluster privacy leakage, not only having strong numerical robustness and engineering applicability, but also having higher economy. At the same time, the aggregated power distribution method of this application considers the flexible resource feasible regions of multiple types of smart pole clusters and adopts a two-stage correction strategy to reduce the risk of generating time-periodic power over-limit, and can still maintain stable performance in a complex operating environment.

[0064] In order to make the above features and advantages of the application more obvious and understandable, specific embodiments are given below, and detailed descriptions are provided in conjunction with the accompanying drawings as follows. Description of the Drawings

[0065] Figure 1 It is a flowchart of the hierarchical distributed cooperative scheduling method for the intelligent pole cluster provided by this application.

[0066] Figure 2 It is a flowchart of the aggregated power distribution method proposed by this application.

[0067] Figure 3 It is a comparison chart of the convergence curves of the centralized and hierarchical distributed cooperative scheduling methods.

[0068] Figure 4 It is a module diagram of the hierarchical distributed cooperative scheduling system for the intelligent pole cluster provided by this application.

[0069] Figure 5 It is a schematic diagram of the hierarchical distributed cooperative scheduling architecture for the intelligent pole cluster provided by this application. Detailed Embodiments

[0070] To make the objectives and technical solutions of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of this application. Obviously, the described embodiments are some, but not all, of the embodiments of this application. All other embodiments obtained by those of ordinary skill in the art based on the described embodiments of this application without creative efforts shall fall within the scope of protection of this application.

[0071] This application provides a hierarchical distributed cooperative scheduling method for an intelligent pole cluster. Please refer to Figure 1 , Figure 1 It is a flowchart of the hierarchical distributed cooperative scheduling method for the intelligent pole cluster provided by this application, including: Step S1 to Step S4.

[0072] Step S1: Establish a day-ahead scheduling optimization model for the aggregated resources of the intelligent pole cluster to participate in the electricity energy market and the peak shaving ancillary service market simultaneously;

[0073] Step S2: Propose a hierarchical distributed cooperative scheduling method based on the scaled form of the alternating direction method of multipliers according to the day-ahead scheduling optimization model;

[0074] Step S3: Based on the hierarchical distributed cooperative scheduling method, solve the upper-layer intelligent pole cluster operator management platform sub-problem and the lower-layer intelligent pole cluster sub-problem respectively;

[0075] Step S4: After solving the sub-problems of the upper-layer intelligent pole cluster operator management platform and the lower-layer intelligent pole cluster, an aggregated power allocation method considering the flexible resource feasible region of multiple types of intelligent pole clusters is proposed, and the aggregated power of the intelligent pole cluster is allocated to each intelligent pole.

[0076] In an embodiment of the present application, the intelligent pole cluster includes a 5G intelligent pole cluster.

[0077] In step S1, refer to Figure 1 step S1 in, and establish a day-ahead scheduling optimization model for the intelligent pole cluster to participate in the day-ahead electricity energy market and the peak shaving ancillary service market simultaneously.

[0078] As an example, the day-ahead scheduling optimization model includes: the economic operation objective function and constraint conditions of the intelligent pole cluster considering the electricity energy market and the peak shaving ancillary service market.

[0079] Construct the economic operation objective function of the intelligent pole cluster considering the electricity energy market and the peak shaving ancillary service market, which is expressed by the formula:

[0080]

[0081] Among them, represents the unit electricity price of the 5G intelligent pole cluster i participating in the day-ahead electricity energy market transaction at time t; represents the day-ahead energy market transaction power of the 5G intelligent pole cluster i at time t; represents the compensation price for upward peak shaving provided by the 5G intelligent pole cluster i to the power grid in the peak shaving ancillary service market at time t, which is given by the clearing of the ancillary service market; represents the compensation price for downward peak shaving provided by the 5G intelligent pole cluster i to the power grid in the peak shaving ancillary service market at time t, which is given by the clearing of the peak shaving ancillary service market; represents the upward peak shaving power provided by the 5G intelligent pole cluster i in the peak shaving ancillary service market at time t; represents the downward peak shaving power provided by the 5G intelligent pole cluster i in the peak shaving ancillary service market at time t; represents the upward peak shaving cost of the 5G intelligent pole cluster i at time t; represents the downward peak shaving cost of the 5G intelligent pole cluster i at time t; Δt represents the minimum scheduling period; I represents the number of 5G intelligent pole clusters i; T represents the scheduling period of the 5G intelligent pole cluster i.

[0082] Among them, the upward peak shaving cost and the downward peak shaving cost of the 5G intelligent pole cluster i at time t are expressed as follows:

[0083]

[0084]

[0085] Among them, c ES represents the battery loss cost per unit capacity; λ e represents the electricity cost per unit; c com represents the user compensation cost per unit power; a1 represents the battery loss cost proportion coefficient and the user compensation cost proportion coefficient; a2 represents the battery loss cost proportion coefficient and the user compensation cost proportion coefficient;

[0086] The battery loss cost c per unit capacity ES is expressed as follows:

[0087]

[0088] Among them, c b represents the fixed energy storage cost; L(D ref ) represents the battery cycle life in terms of the number of cycles; the battery cycle life is different at different discharge depths, D ref represents the reference discharge depth of the battery, E v represents the rated capacity of the battery, and μ represents the charge-discharge efficiency of the battery.

[0089] Furthermore, the constraint conditions for establishing the economic operation objective function are included: peak shaving demand balance constraint, power and energy constraints in the electricity energy market trading, up-regulation and down-regulation power constraints, and energy change constraints considering up-regulation power and peak shaving power;

[0090] The peak shaving demand balance constraint is expressed as follows:

[0091]

[0092] Among them, P t req represents the up and down peak shaving demand at time t; represents the total peak shaving response coefficient variable of all 5G smart pole clusters at time t; the range of the peak shaving demand balance constraint is [0, 1].

[0093] The power and energy constraints in the electricity energy market trading are expressed as follows:

[0094]

[0095] Among them, represents the electricity energy market trading power of the 5G smart pole cluster i at time t; represents the lower limit value of the aggregated power of the 5G smart pole cluster i at time t; represents the upper limit value of the aggregated power of the 5G smart pole cluster i at time t; Represents the upper limit value of the capacity of the virtual battery of the 5G intelligent pole cluster i at time t; Represents the lower limit value of the capacity of the virtual battery of the 5G intelligent pole cluster i at time t; Represents the remaining capacity of the virtual battery of the 5G intelligent pole cluster i at time t after participating in the day-ahead electric energy market transaction; Represents the remaining capacity of the virtual battery of the 5G intelligent pole cluster i at time t+1 after participating in the day-ahead electric energy market transaction.

[0096] The up-regulation and down-regulation power constraints are expressed as follows:

[0097]

[0098] Among them, Represents the up-regulation state variable of the 5G intelligent pole cluster i at time t, which is a binary variable; Represents the down-regulation state variable of the 5G intelligent pole cluster i at time t, which is a binary variable; Represents the upper limit value of the up-regulation power provided by the 5G intelligent pole cluster i in the regulation ancillary service market at time t; Represents the upper limit value of the down-regulation power provided by the 5G intelligent pole cluster i in the regulation ancillary service market at time t.

[0099] The energy change constraint considering the regulation power is expressed as follows:

[0100]

[0101] Among them, Represents the equivalent remaining capacity considering the power of the virtual battery of the 5G intelligent pole cluster i participating in the electric energy market transaction, up-regulation power and down-regulation power at time t; Represents the equivalent remaining capacity considering the power of the virtual battery of the 5G intelligent pole cluster i participating in the electric energy market transaction, up-regulation power and down-regulation power at time t+1.

[0102] In step S2, please refer to Figure 1 Step S2 in, and a hierarchical distributed cooperative scheduling method based on the scaled form of the alternating direction multiplier method is proposed according to the day-ahead scheduling optimization model.

[0103] As an example, introduce the decoupled auxiliary variables and The peak regulation demand balance constraint proposed in step S1 is decoupled and transformed into:

[0104]

[0105] Among them, Represents the decoupled auxiliary variable; Represents a decoupled auxiliary variable.

[0106] Furthermore, introduce the coupling constraint between the 5G smart pole cluster i and the 5G smart pole cluster operator management platform, which is expressed as follows:

[0107]

[0108] Furthermore, the augmented Lagrangian function of the day-ahead optimal scheduling model constructed based on the scaled form of the alternating direction method of multipliers is as follows:

[0109]

[0110] Wherein, Both contain the optimization vectors of T optimization scheduling periods; Represents a decoupled auxiliary variable; λ i Is the Lagrange multiplier vector; ρ is the penalty coefficient of the coupling constraint.

[0111] Furthermore, based on the scaled form of the alternating direction method of multipliers, use u i = λ i / ρ to scale the Lagrange multiplier vector λ i To obtain a hierarchical distributed cooperative scheduling method based on the scaled form of the alternating direction method of multipliers; wherein, u i Represents the scaled Lagrange multiplier vector.

[0112] In step S3, please refer to Figure 1 In step S3, based on the hierarchical distributed cooperative scheduling method, solve the upper-layer smart pole cluster operator management platform sub-problem and the lower-layer smart pole cluster sub-problem respectively.

[0113] As an example, this application decouples the centralized scheduling problem between the 5G smart pole cluster and the 5G smart pole cluster operator management platform to form an upper-layer smart pole cluster operator management platform and a lower-layer 5G smart pole cluster. The upper-layer smart pole cluster operator management platform coordinates the peak shaving response power uniformly, and the lower-layer 5G smart pole cluster aggregates the power for autonomous scheduling, realizing that each smart pole cluster and the smart pole cluster operator management platform only interact independently through a small amount of data.

[0114] The lower-layer 5G smart pole cluster sub-problem includes: the self-consistent scheduling between 5G smart pole clusters, and the objective function is expressed as:

[0115]

[0116] Wherein, Are all vectors containing T time periods, x i Represents the decision variable, Denote the remaining capacity of the 5G smart pole cluster i after peak shaving power adjustment at time t. Denote the remaining capacity of the 5G smart pole cluster i after valley filling power adjustment at time t. Denote the minimum value of the objective function of the sub-problem of the lower-layer 5G smart pole cluster i.

[0117] Specifically, the constraint conditions of the sub-problem of the lower-layer 5G smart pole cluster include: energy market trading power and energy constraints, peak shaving and valley filling power constraints, and energy change constraints considering peak shaving power.

[0118] The sub-problem of the upper-layer 5G smart pole cluster operator management platform includes: issuing peak shaving instructions to each 5G smart pole cluster, and the objective function is expressed as:

[0119]

[0120] Among them, Denote the objective function of the sub-problem of the upper-layer 5G smart pole cluster operator management platform.

[0121] Specifically, the constraint conditions of the sub-problem of the upper-layer 5G smart pole cluster operator management platform include: peak shaving demand balance constraint.

[0122] In step S4, please refer to Figure 1 the S4 step in

[0123] After solving the sub-problem of the upper-layer smart pole cluster operator management platform and the sub-problem of the lower-layer smart pole cluster, a method for aggregating power distribution considering the flexible resource feasible region of multiple types of smart pole clusters is proposed, and the aggregated power of the smart pole cluster is distributed to each smart pole.

[0124] As an example, after solving the sub-problem of the upper-layer 5G smart pole cluster operator management platform and the sub-problem of the lower-layer 5G smart pole cluster to maximize economic benefits under the condition of meeting the constraints, the aggregated power is distributed within each smart pole cluster and the power is distributed to the single-pole smart poles at the device level.

[0125] Please refer to Figure 2 , Figure 2The flowchart of the aggregation power distribution method proposed in this application includes steps S41 to S49.

[0126] Step S41: Input the initial capacity, power feasible region, and capacity feasible region of the equivalent virtual battery of each smart pole cluster, and set the current time t = 1;

[0127] Step S42: Calculate the charging and discharging power distribution coefficients of each smart pole;

[0128] Step S43: Calculate the charging and discharging distribution power of each smart pole and the remaining capacity of the equivalent virtual battery after the distribution power;

[0129] Step S44: Determine whether the distribution power meets the feasible region. If so, set the current time t = t + 1 and enter Step S45; if not, enter Step S46;

[0130] Step S45: Determine whether the current time t satisfies being greater than the scheduling period T of the smart pole cluster. If so, enter Step S49; if not, return to Step S42;

[0131] Step S46: Fix the distribution power of the smart pole that exceeds the feasible region to the boundary value;

[0132] Step S47: Calculate the charging and discharging power distribution coefficients of other smart poles;

[0133] Step S48: Calculate the charging and discharging distribution power of each smart pole and the remaining capacity of the equivalent virtual battery after the distribution power, and return to Step S42;

[0134] Step S49: The aggregation power distribution ends.

[0135] As an example, calculating the charging and discharging power distribution coefficients of each smart pole is expressed as follows:

[0136]

[0137] Among them, represents the discharging power distribution coefficient of the 5G smart pole j inside the 5G smart pole cluster i at time t, represents the charging power distribution coefficient of the 5G smart pole j inside the 5G smart pole cluster i at time t; E j,t represents the remaining capacity of the equivalent virtual battery of the 5G smart pole j at time t; represents the upper limit value of the feasible region of the equivalent virtual battery of the 5G smart pole j, represents the lower limit value of the feasible region of the equivalent virtual battery of the 5G smart pole j, and j represents the number of 5G smart poles inside the 5G smart pole cluster.

[0138] As an example, calculating the charging and discharging distribution power of each smart pole according to the distribution coefficient is expressed as follows:

[0139]

[0140] Among them, represents the discharge power of the 5G smart pole j inside the 5G smart pole cluster i at time t, represents the charging power of the 5G smart pole j inside the 5G smart pole cluster i at time t; represents the total trading power of the 5G smart pole cluster i participating in the day-ahead electricity energy market and the peaking ancillary service market.

[0141] In an embodiment of the present application, the penalty coefficient ρ in the hierarchical distributed cooperative scheduling method based on the alternating direction multiplier method is taken as 10-5, and the convergence error precision is taken as 10 -4 . The energy storage batteries of the 5G smart poles all adopt lithium iron phosphate cascade batteries, and their inherent cost is taken as 0.4 yuan / Wh. The simulation period is set to 24h, and the simulation step is 1h. In order to verify the feasibility of the hierarchical distributed 5G smart pole cluster cooperative scheduling optimization method provided by the present application, Table 1 gives the objective function results of solving the day-ahead scheduling optimization problem of the 5G smart pole cluster participating in the power grid peaking by using the centralized method and the hierarchical distributed cooperative scheduling method provided by the present application respectively.

[0142] Table 1 Comparison between the centralized and hierarchical distributed methods

[0143]

[0144] It can be calculated from the data in Table 1 that compared with the centralized method, the total objective deviation of the hierarchical distributed cooperative scheduling method provided by the present application is less than 10 -5 , which can be ignored. The deviation of the operating cost objective results of each smart pole cluster is 0.27%, -0.19%, and 0.09% respectively, all of which can meet the deviation range allowed by the actual engineering scheduling. Therefore, the hierarchical distributed cooperative scheduling method provided by the present application has an approximately equivalent solution effect compared with the centralized method, and the proposed method only needs to interact very little data information between each smart pole cluster and the smart pole cluster operator management platform, reducing the communication transmission data volume and transmission time, and can realize the independent autonomous scheduling of the aggregated power of each cluster, reducing the risk of cluster privacy leakage.

[0145] Please refer to Figure 3 , Figure 3 which is a comparison chart of the convergence curves of the centralized and hierarchical distributed cooperative scheduling methods. To verify the convergence characteristics of the hierarchical distributed optimization cooperative scheduling provided by the present application, Figure 3The original-dual residual evolution trajectory and its comparison with the preset convergence criterion are shown through a two-dimensional residual monitoring mechanism. The logarithmic coordinate visualization technology is used to perform cross-order-of-magnitude compression mapping on the residual magnitude, effectively revealing the global convergence dynamic characteristics of the two methods. The convergence process shows significant three-stage characteristics: in the initial oscillation stage (1-20 iterations), the residual amplitude fluctuates by an order of magnitude of 10^0; in the stable decline stage (21-60 iterations), the residual shows an exponential decay trend; in the asymptotic convergence stage (after 61 iterations), it enters the superlinear convergence domain. In particular, the original residual and the dual residual of the hierarchical distributed cooperative scheduling method provided in this application first break through the convergence threshold at the 43rd and 47th iterations respectively, and after 42 optimization oscillations, steady-state convergence is achieved at the 85th iteration (residual fluctuation amplitude < 10^-5), fully verifying the advantages of the hierarchical distributed cooperative scheduling method provided in this application in terms of numerical robustness and engineering applicability.

[0146] This application also provides a hierarchical distributed cooperative scheduling system for a smart pole cluster. Please refer to Figure 4 , Figure 4 which is a module diagram of the hierarchical distributed cooperative scheduling system for the smart pole cluster provided in this application. It executes a hierarchical distributed cooperative scheduling method for a smart pole cluster as described above, including:

[0147] A day-ahead scheduling optimization module 41, which is used to establish a day-ahead scheduling optimization model for the smart pole cluster to aggregate resources and participate in the electricity energy market and the peak shaving ancillary service market simultaneously;

[0148] A hierarchical distributed cooperative scheduling module 42, which is used to propose a hierarchical distributed cooperative scheduling method based on the scaled form of the alternating direction multiplier method according to the day-ahead scheduling optimization model;

[0149] A sub-problem solving module 43, which is used to solve the upper-layer smart pole cluster operator management platform sub-problem and the lower-layer smart pole cluster sub-problem respectively based on the hierarchical distributed cooperative scheduling method;

[0150] An aggregated power allocation module 44, which is used to propose an aggregated power allocation method considering the flexible resource feasible region of multiple types of smart pole clusters after solving the upper-layer smart pole cluster operator management platform sub-problem and the lower-layer smart pole cluster sub-problem, and allocate the aggregated power of the smart pole cluster to each smart pole.

[0151] As an example, the day-ahead scheduling optimization module 41, the hierarchical distributed cooperative scheduling module 42, the sub-problem solving module 43, and the aggregated power allocation module 44 are connected in sequence.

[0152] As an example, please refer to Figure 5 , Figure 5Schematic diagram of the hierarchical distributed cooperative scheduling architecture of the intelligent pole cluster provided by this application. The hierarchical distributed cooperative scheduling architecture of the intelligent pole cluster includes: i intelligent pole clusters, specifically including intelligent pole clusters 511 to 51i, the intelligent pole cluster operator management platform 52, the electric energy market 53, the peak shaving auxiliary service market 54, and the superior power grid 55; the superior power grid 55 is respectively connected to the first ends of the intelligent pole clusters 511 to 51i, and the second ends of the intelligent pole clusters 511 to 51i are respectively connected to the first ends of the intelligent pole cluster operator management platform 52, and the second ends of the intelligent pole cluster operator management platform 52 are respectively connected to the electric energy market 53 and the peak shaving auxiliary service market 54.

[0153] As an example, the superior power grid 55 provides electric energy for the intelligent pole clusters 511 to 51i; the intelligent pole clusters 511 to 51i send the electric energy trading power to the intelligent pole cluster operator management platform 52 and receive the peak shaving instructions and price information sent by the intelligent pole cluster operator management platform 52; the electric energy market 53 sends the electric energy market price to the intelligent pole cluster operator management platform 52 and receives the electric energy trading power sent by the intelligent pole cluster operator management platform 52; the peak shaving auxiliary service market 54 sends the peak shaving compensation price and peak shaving demand information to the intelligent pole cluster operator management platform 52 and receives the response peak shaving power sent by the intelligent pole cluster operator management platform 52.

[0154] This application also provides a computer program product, including a computer program, which when executed by a processor implements a hierarchical distributed cooperative scheduling method for an intelligent pole cluster as described above.

[0155] The hierarchical distributed cooperative scheduling method, system and computer program product of the intelligent pole cluster proposed in this application, based on the alternating direction multiplier method, only need the intelligent pole clusters and the intelligent pole cluster operator management platform to exchange very little data information, reducing the communication transmission data volume and transmission time, significantly improving the calculation efficiency and response speed, being able to efficiently handle the scheduling problems of large-scale resource clusters, realizing the independent autonomous scheduling of the aggregated power of each cluster, reducing the risk of cluster privacy leakage, not only having strong numerical robustness and engineering applicability, but also having higher economy. At the same time, the aggregated power distribution method of this application considers the flexible resource feasible regions of multiple types of intelligent pole clusters and adopts a two-stage correction strategy to reduce the risk of generating time-periodic power over-limit and still maintain stable performance in a complex operating environment.

[0156] Although this application has been disclosed above with examples, it is not intended to limit this application. Any person with ordinary knowledge in the technical field to which this application belongs, without departing from the spirit and scope of this application, can make some modifications and refinements. Therefore, the protection scope of this application shall be subject to that defined by the appended patent application scope.

Claims

1. A hierarchical distributed collaborative scheduling method for a smart pole cluster, characterized in that including establishing a day-ahead scheduling optimization model for the aggregated resources of a smart pole cluster to participate in the electricity energy market and the peaking ancillary service market simultaneously proposing a hierarchical distributed collaborative scheduling method based on the scaled form of the alternating direction method of multipliers according to the day-ahead scheduling optimization model solving the upper-layer sub-problem of the smart pole cluster operator management platform and the lower-layer sub-problem of the smart pole cluster respectively based on the hierarchical distributed collaborative scheduling method after solving the upper-layer sub-problem of the smart pole cluster operator management platform and the lower-layer sub-problem of the smart pole cluster, proposing an aggregated power allocation method considering the flexible resource feasible region of multiple types of smart pole clusters, and allocating the aggregated power of the smart pole cluster to each smart pole 2. The hierarchical distributed cooperative scheduling method for the intelligent pole cluster according to claim 1, wherein The establishment of the day-ahead scheduling optimization model for the aggregated resources of the smart pole cluster to participate in the electricity energy market and the peaking ancillary service market simultaneously includes: constructing an economic operation objective function and constraint conditions for the smart pole cluster considering the electricity energy market and the peaking ancillary service market 3. The hierarchical distributed cooperative scheduling method for the intelligent pole cluster according to claim 2, characterized in that, The economic operation objective function of the smart pole cluster considering the electricity energy market and the peaking ancillary service market is expressed as follows Among them, represents the unit electricity price at which the smart pole cluster i participates in the day-ahead electricity energy market transaction at time t; represents the day-ahead energy market transaction power of the smart pole cluster i at time t; represents the compensation price for upward peak regulation provided by the smart pole cluster i to the power grid in the peak regulation ancillary service market at time t, which is given by the clearing of the ancillary service market; represents the compensation price for downward peak regulation provided by the smart pole cluster i to the power grid in the peak regulation ancillary service market at time t, which is given by the clearing of the peak regulation ancillary service market; represents the upward peak regulation power provided by the smart pole cluster i in the peak regulation ancillary service market at time t; represents the downward peak regulation power provided by the smart pole cluster i in the peak regulation ancillary service market at time t; represents the upward peak regulation cost of the smart pole cluster i at time t; represents the downward peak regulation cost of the smart pole cluster i at time t; Δt represents the minimum scheduling period; I represents the number of smart pole clusters i; T represents the scheduling period of the smart pole cluster i.

4. The hierarchical distributed cooperative scheduling method for the intelligent pole cluster according to claim 2, wherein The constraint conditions include: peaking demand balance constraint, electricity energy market trading power and energy constraint, up-regulation and down-regulation power constraints, and energy change constraint considering up-regulation power The peaking demand balance constraint is expressed as follows Among them, P t req represents the up and down peak shaving demand at time t; represents the total peak shaving response coefficient variable of all intelligent pole clusters at time t; the range of the peak shaving demand balance constraint is [0, 1]; The electricity energy market trading power and energy constraint is expressed as follows Among them, represents the trading power of the smart pole cluster i in the electricity energy market at time t; represents the lower limit value of the aggregated power of the smart pole cluster i at time t; represents the upper limit value of the aggregated power of the smart pole cluster i at time t; represents the upper limit value of the capacity of the virtual battery of the smart pole cluster i at time t; represents the lower limit value of the capacity of the virtual battery of the smart pole cluster i at time t; represents the remaining capacity of the virtual battery of the smart pole cluster i at time t after participating in the day-ahead electricity energy market trading; represents the remaining capacity of the virtual battery of the smart pole cluster i at time t + 1 after participating in the day-ahead electricity energy market trading; The up-regulation and down-regulation power constraints are expressed as follows Among them, represents the upward peaking state variable of the intelligent pole cluster i at time t, which is a binary variable; represents the downward peaking state variable of the intelligent pole cluster i at time t, which is a binary variable; represents the upper limit of the upward peaking power provided by the intelligent pole cluster i in the peaking ancillary service market at time t; represents the upper limit of the downward peaking power provided by the intelligent pole cluster i in the peaking ancillary service market at time t; The energy change constraint considering peaking power is expressed as follows Among them, represents the equivalent remaining capacity of the virtual battery of the smart pole cluster i after participating in the trading power, upward peaking power, and downward peaking power in the electricity energy market at time t; represents the equivalent remaining capacity of the virtual battery of the smart pole cluster i after participating in the trading power, upward peaking power, and downward peaking power in the electricity energy market at time t + 1.

5. The hierarchical distributed cooperative scheduling method for the intelligent pole cluster according to claim 1, characterized in that, The proposal of the hierarchical distributed collaborative scheduling method based on the scaled form of the alternating direction method of multipliers according to the day-ahead scheduling optimization model includes introducing a decoupling auxiliary variable to decouple the peaking demand balance constraint, expressed as Among them, represents a decoupled auxiliary variable; represents a decoupled auxiliary variable; According to the decoupled peaking demand balance constraint, scaling the Lagrange multiplier vector for the augmented Lagrangian function of the day-ahead optimization scheduling model constructed based on the scaled form of the alternating direction method of multipliers, to obtain the hierarchical distributed collaborative scheduling method based on the scaled form of the alternating direction method of multipliers 6. The hierarchical distributed cooperative scheduling method for the intelligent pole cluster according to claim 5, wherein, The augmented Lagrangian function of the day-ahead optimization scheduling model constructed based on the scaled form of the alternating direction method of multipliers is as follows Among them, both contain the optimization vectors of T optimized scheduling periods; represents the decoupled auxiliary variable; λ i is the Lagrange multiplier vector; ρ is the penalty coefficient of the coupling constraint; The scaling of the Lagrange multiplier vector includes: using u i = λ i / ρ to scale the Lagrange multiplier vector λ i ; where u i represents the scaled Lagrange multiplier vector.

7. The hierarchical distributed cooperative scheduling method for the intelligent pole cluster according to claim 1, characterized in that The lower-layer sub-problem of the smart pole cluster includes: self-consistent scheduling among smart pole clusters, and the objective function is expressed as Among them, are all vectors containing T time periods, and x i represents the decision variable, represents the remaining capacity of the smart pole cluster i after peak shaving power adjustment at time t, represents the remaining capacity of the smart pole cluster i after valley filling power adjustment at time t; represents the minimum value of the objective function of the sub-problem of the lower-layer smart pole cluster i; The upper-layer sub-problem of the smart pole cluster operator management platform includes: issuing peaking instructions to each smart pole cluster, and the objective function is expressed as Among them, represents the minimum value of the objective function of the sub-problem of the upper-layer intelligent pole cluster operator management platform.

8. The hierarchical distributed cooperative scheduling method of the intelligent pole cluster according to claim 1, characterized in that, The aggregated power allocation method considering the flexible resource feasible region of multiple types of smart pole clusters includes: steps S41 to S49 Step S41: Input the initial capacity, power feasible region, and capacity feasible region of the equivalent virtual battery of each smart pole cluster, and set the current time t = 1 Step S42: Calculate the charging and discharging power distribution coefficients of each smart pole Step S43: Calculate the charging and discharging power allocated to each smart pole and the remaining capacity of the equivalent virtual battery after the allocated power Step S44: Determine whether the allocated power meets the feasible region. If so, set the current time t = t + 1 and go to Step S45; if not, go to Step S46 Step S45: Determine whether the current time t is greater than the scheduling period T of the smart pole cluster. If so, proceed to Step S49; if not, return to Step S42; Step S46: The power allocated to the smart poles beyond the feasible region is fixed at the boundary value; Step S47: Calculate the charge and discharge power distribution coefficients of the other smart poles; Step S48: Calculate the charge and discharge allocated power of each smart pole and the remaining capacity of the equivalent virtual battery after the allocated power, and return to Step S42; Step S49: The aggregated power distribution ends.

9. The hierarchical distributed collaborative scheduling method for the intelligent pole cluster according to claim 8, wherein, The calculation of the charge and discharge power distribution coefficients of each smart pole is expressed as follows: Among them, represents the discharge power distribution coefficient of the smart pole j inside the smart pole cluster i at time t, represents the charging power distribution coefficient of the smart pole j inside the smart pole cluster i at time t; E j,t represents the remaining capacity of the equivalent virtual battery of the smart pole j at time t; represents the upper limit value of the feasible region of the equivalent virtual battery of the smart pole j, represents the lower limit value of the feasible region of the equivalent virtual battery of the smart pole j, and j represents the number of smart poles inside the smart pole cluster; The calculation of the charge and discharge allocated power of each smart pole is expressed as follows: Among them, represents the discharge power of the smart pole j inside the smart pole cluster i at time t, represents the charging power of the smart pole j inside the smart pole cluster i at time t; represents the total trading power of the smart pole cluster i participating in the day-ahead electricity energy market and the peaking ancillary service market.

10. A hierarchical distributed collaborative scheduling system for a smart pole cluster, which executes a hierarchical distributed collaborative scheduling method for a smart pole cluster as described in any one of claims 1-9, characterized in that, It includes: A day-ahead scheduling optimization module, a hierarchical distributed cooperative scheduling module, a sub-problem solving module, and an aggregated power distribution module; The day-ahead scheduling optimization module, the hierarchical distributed cooperative scheduling module, the sub-problem solving module, and the aggregated power distribution module are connected in sequence.

11. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements a hierarchical distributed cooperative scheduling method for a smart pole cluster as described in any one of claims 1-9.

Images