Energy storage and load aggregation method and system based on comprehensive virtual generator and battery model

Abstract

The invention belongs to the technical field of virtual power plants, and relates to an energy storage and load aggregation method and system based on a comprehensive virtual generator and a battery model. The method comprises the following steps: aggregating multi-energy equipment in a plurality of distributed source network load storage energy systems according to a constructed aggregation model of a comprehensive virtual generator and a virtual battery to obtain an aggregate; for undetermined parameters in the aggregation model, performing parameter solution according to the constructed parameter robust optimization solution model; the parameter robust optimization solution model is divided into two-stage constraints, and a second-stage constraint condition is established based on a distributed source network load storage refined mathematical model; and carrying out approximate linear summation on the aggregation model according to a Minkowski approximate linear summation method, solving to obtain an adjustable power domain of an aggregate, and realizing resource aggregation of the multi-distributed source network load storage energy system. The aggregation processing and solving method optimizes the resource configuration in the multi-distributed source network load storage energy system, and improves the economy and flexibility of the system.

Description

Technical Field

[0001] The present invention belongs to the technical field of virtual power plants, and in particular relates to a method and system for energy storage and load aggregation based on a comprehensive virtual generator and battery model. Background Art

[0002] Virtual power plants play an important role between the power grid management system and distributed energy. They need to aggregate the operating characteristics of distributed energy instead of directly reporting the detailed physical models of all distributed energy to the superior power grid. This is mainly due to two considerations: First, to simplify communication and optimization problem solving: if the detailed physical models of all distributed energy are reported directly to the superior power grid, model information may be missing or wrong during the communication process, and the global optimization problem of the power grid is difficult to solve due to the large scale of variables. By aggregating the operating characteristics of distributed energy, the communication process and the solution of optimization problems can be greatly simplified, and the efficiency and stability of the system can be improved. Second, to protect commercial secrets and maintain market competitive advantages: virtual power plants often participate in various transactions in the power market driven by value. In order to maintain an absolute advantage in market competition, they are not willing to disclose the detailed information of their internal models to the power grid. By aggregating the operating characteristics of distributed energy, virtual power plants can protect commercial secrets and maintain their dominant position in market competition. Therefore, aggregating the operating characteristics of distributed energy in virtual power plants is an effective management and optimization method, which helps to simplify system operation and protect commercial secrets, thereby better realizing participation in the power market and optimization of power grid dispatching.

[0003] The technical characteristics of distributed energy are usually described in detail through power inequality constraints, which mainly include power constraints, ramp constraints and capacity constraints. Taking into account the power constraints of all distributed energy, the adjustable power domain of the virtual power plant can be mapped into a high-dimensional convex polyhedron in a geometric space. In order to solve the adjustable power domain of this virtual power plant, a specific high-dimensional convex polyhedron can be used to approximate it from the inside or outside.

[0004] At present, there are two main mathematical models of high-dimensional convex polyhedrons that are suitable and widely used: virtual battery model and virtual generator model. The virtual battery model is highly complex when facing multi-dimensional uncertainties, and because it ignores the dynamic characteristics of the battery, it lacks accuracy. It is suitable for short-term scheduling but difficult to accurately predict long-term behavior; the virtual generator model simplifies the complexity of the system, lacks an accurate description of the physical process, is difficult to handle nonlinear behavior, lacks flexibility, and has a large computational overhead in large-scale applications. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for energy storage and load aggregation based on a comprehensive virtual generator and battery model, which solves the problems of insufficient model accuracy, complexity and inaccurate dynamics.

[0006] The present invention is achieved through the following technical solutions:

[0007] On the one hand, the present invention discloses a method for energy storage and load aggregation based on a comprehensive virtual generator and battery model, comprising the following process:

[0008] According to the constructed aggregation model of comprehensive virtual generators and virtual batteries, multiple energy devices in multiple distributed source-grid-load-storage energy systems are aggregated to obtain an aggregate;

[0009] For the undetermined parameters in the aggregation model, the parameters are solved according to the constructed parameter robust optimization solution model; the parameter robust optimization solution model is divided into two-stage constraints, and the second-stage constraint conditions are established based on the distributed source-grid-load-storage refined mathematical model; the distributed source-grid-load-storage refined mathematical model includes a distributed power source refined mathematical model, a distributed load refined mathematical model, a distributed energy storage refined mathematical model and a distributed network refined mathematical model;

[0010] According to the Minkowski approximate linear summation method, the aggregation model is approximated linearly summed to obtain the adjustable power domain of the aggregate, thereby realizing resource aggregation of multiple distributed source-grid-load-storage energy systems.

[0011] Based on the above technical solution, further, the distributed power generation refined mathematical model is divided into a new energy unit mathematical model and a traditional energy unit mathematical model;

[0012] The mathematical model of the new energy unit is:

[0013]

[0014] Where: R represents the set of new energy generating units in the distributed source-grid-load-storage energy system; T represents the set of scheduling periods in the distributed source-grid-load-storage energy system; P r,t,max P is the maximum output that the new energy generator r can actually generate during period t; r,t is the active power output of the new energy generator r in period t;

[0015] The mathematical model of traditional energy units is:

[0016]

[0017] Where: G represents the set of traditional energy generating units in the distributed source-grid-load-storage energy system; P g,t are respectively the active output of the traditional energy generator g in period t; Pg,max is the upper limit of active output of traditional energy generator g; P g,min is the lower limit of active power output of traditional energy generator g; R g,max is the upper limit of the output adjustment of the traditional energy generator g, R g,min It is the lower limit of the output adjustment of the traditional energy generator set g.

[0018] Based on the above technical solution, further, the distributed load is divided into a shiftable load and a curtailable load according to the type of load response. The shiftable load corresponds to the transfer in time, and the curtailable load corresponds to the reduction in power.

[0019] The distributed load refinement mathematical model is divided into a mathematical model of shiftable load and a mathematical model of curtailable load;

[0020] The mathematical model of the translatable load is:

[0021]

[0022] Where: D represents the set of responsive loads in the distributed source-grid-load-storage energy system; Ω T+ represents the time period set of load transfer; Ω T- represents the time period set of load removal; P shift+,d,t P represents the load amount that can be moved in by the response load d within the time period t; shift-,d,t P represents the load that can be removed in response to load d within time period t; shift+,d,max and P shift+,d,min They represent the maximum and minimum values ​​of the load that can be moved into the response load d within the time period t; P shift-,d,max and P shift-,d,min They represent the maximum and minimum values ​​of the load that can be removed by the response load d within the time period t respectively;

[0023] The mathematical model of load reduction is:

[0024]

[0025] Where: D represents the set of responsive loads in the distributed source-grid-load-storage energy system; Ω Tcut represents the set of time periods for load reduction; P cut,d,t P represents the load reduction amount of the responsive load d within the time period t; cut,d,max and P cut,d,min They respectively represent the maximum and minimum values ​​of the load that can be reduced by the responding load d in a unit time period.

[0026] Based on the above technical solution, the distributed energy storage refined mathematical model is further:

[0027]

[0028] Where: ES represents the collection of different distributed energy storage devices in the distributed source-grid-load-storage energy system; E es,t , P c,es,t and P dc,es,t represents the state of charge, charging power and discharging power of the energy storage device es in time period t; ζ c,E and dc,E are the charging efficiency and discharging efficiency of the energy storage device es respectively; E es,min 、E es,max Indicates the upper and lower limits of the capacity of the energy storage device es; P c,es,max represents the upper limit of the charging power of the energy storage device es; P c,es,min Represents the lower limit of the charging power of the energy storage device es; P dc,es,max Represents the upper limit of the discharge power of the energy storage device es; P dc,es,min Indicates the lower limit of the discharge power of the energy storage device es.

[0029] Based on the above technical solution, further, the distributed network refined mathematical model only considers the power balance constraint when establishing it; the power balance constraint of the distributed source-grid-load-storage energy system includes the overall power balance constraint of the distributed source-grid-load-storage energy system and the power balance constraint before and after the load demand response;

[0030] The overall power balance constraint of the distributed source-grid-load-storage energy system is:

[0031]

[0032] Where: P d,t is the load demand after the respondable load d performs demand response; P tie,t is the interconnection power between the distributed source-grid-load-storage energy system and the large power grid, if P tie,t is positive, indicating that the distributed source-grid-load-storage energy system purchases electricity from the large power grid. If P tie,t If it is negative, it means that the distributed source-grid-load-storage energy system sells electricity to the large power grid;

[0033] The power balance constraint before and after load demand response is:

[0034]

[0035] Where: P d,t is the load demand after the respondable load d performs demand response; P d0,t is the load demand before the responding load d performs demand response; P shift,d,t The load for shifting demand response; P cut,d,t The load required for demand response reduction; P shift-,d,t is the load shifted out during period t; Pshift+,d,t is the load shifted out during period t.

[0036] Based on the above technical solution, further, the aggregation model of the integrated virtual generator and virtual battery is shown as follows:

[0037]

[0038] Where: Ω is the feasible domain of the interaction power between the aggregate and the large power grid constrained by the aggregation model;

[0039] P Juhe,t is the interaction power between the aggregate and the large power grid in time period t, corresponding to the interaction power P between the interconnection line of the distributed source-grid-load-storage energy system and the large power grid tie,t ;

[0040] is the power of the aggregate interacting with the large power grid in period t; P Juhe,t,max is the upper limit of the power of the aggregate interacting with the large power grid in period t, P Juhe,t,min is the lower power limit of the aggregate interacting with the large grid in period t, P Juhe,t,max and P Juhe,t,min is a time-varying parameter;

[0041] E Juhe,t,max is the upper limit of the capacity of the aggregate in time period t, E Juhe,t,min is the lower limit of the capacity of the aggregate in time period t, E Juhe,t,max and E Juhe,t,min is a time-varying parameter;

[0042] R Juhe,max is the upper limit of the power ramp of the aggregate interacting with the large power grid during each period, R Juhe,min is the lower limit of the power ramp of the aggregate interacting with the large power grid during each period, R Juhe,max and R Juhe,min is a constant parameter.

[0043] Based on the above technical solution, there are many parameters to be determined in the aggregation model of the virtual generator and the virtual battery, including P Juhe,t,max and P Juhe,t,min 、E Juhe,t,max and E Juhe,t,min and R Juhe,max and R Juhe,min ;

[0044] The parameter robust optimization solution model is a two-stage robust optimization model, which is used to solve the above parameters to be determined;

[0045] The objective function of the two-stage robust optimization model is:

[0046]

[0047] Where: x represents the first-stage variable, which is the parameter to be determined for the aggregation model; y represents the second-stage variable, including the output of distributed power sources in the distributed source-grid-load-storage energy system, the demand response of distributed loads, the capacity and charging and discharging power of distributed energy storage equipment, the large-grid interactive power, and the newly introduced non-negative slack variable of the power balance constraint; γ P,min , γ P,max , γ E,min , γ E,max , γ R,min and γ R,max are the weight coefficients of the corresponding parameters respectively; M is a large real number; κ t,+ and κ t,- is the non-negative slack variable for introducing power balance constraints;

[0048] First stage constraints:

[0049]

[0050] Where: P max , P min , R max and R min It represents the initial bounds of the parameters to be determined for the comprehensive aggregation model;

[0051] The constraints of the second stage are the refined mathematical model of distributed source, grid, load and storage, in which the overall power balance constraint of the distributed source, grid, load and storage energy system is replaced by:

[0052]

[0053] The constraints for coupling the first and second stages are the aggregation model of the integrated virtual generator and virtual battery.

[0054] Based on the above technical solution, further, the aggregation model is approximated linearly summed according to the Minkowski approximate linear summation method to solve the adjustable power domain of the aggregate. The specific process is:

[0055] Firstly, the adjustable power domain of a single distributed source-grid-load-storage energy system is mapped to a geometric space to form a high-dimensional convex polyhedron; the vertex set or half-plane set of the high-dimensional convex polyhedron corresponding to each distributed source-grid-load-storage energy system is solved, and the vertex set or half-plane set is used to represent the adjustable power domain of each distributed source-grid-load-storage energy system;

[0056] Then, according to the Minkowski approximate linear summation method, the vertices or half-planes corresponding to the high-dimensional convex polyhedron of the adjustable power domains of all distributed source-grid-load-storage energy systems are arranged in sequence and added accordingly to obtain a new vertex set or half-plane set;

[0057] New vertex sets or half-plane sets are used to represent adjustable power domains for multiple aggregates.

[0058] Based on the above technical solution, the high-dimensional convex polyhedron is further characterized in the form of a half-plane, and the compact form of the representation is:

[0059]

[0060] Where: Ω (j) is the adjustable power domain of aggregate j described by the half-plane method; is the power P adjusted by aggregate j in each period within the scheduling period T Juhe,t The column vector formed by E, A (j) is the power constraint coefficient matrix corresponding to aggregate j; is the power constraint parameter vector corresponding to aggregate j.

[0061] On the other hand, the present invention also discloses an energy storage and load aggregation system based on a comprehensive virtual generator and battery model, comprising:

[0062] An aggregation module is used to aggregate multiple energy devices in multiple distributed source-grid-load-storage energy systems according to the constructed aggregation model of comprehensive virtual generators and virtual batteries to obtain an aggregate;

[0063] A solution module is used to solve the parameters of the undetermined parameters in the aggregation model according to the constructed parameter robust optimization solution model; the parameter robust optimization solution model is divided into two-stage constraints, and the second-stage constraint conditions are established based on the distributed source-grid-load-storage refined mathematical model; the distributed source-grid-load-storage refined mathematical model includes a distributed power source refined mathematical model, a distributed load refined mathematical model, a distributed energy storage refined mathematical model and a distributed network refined mathematical model;

[0064] The adjustable power domain calculation module is used to perform approximate linear summation on the aggregation model according to the Minkowski approximate linear summation method, solve the adjustable power domain of the aggregate, and realize resource aggregation of multiple distributed source-grid-load-storage energy systems.

[0065] Compared with the prior art, the present invention has the following beneficial technical effects:

[0066] The present invention discloses a method for energy storage and load aggregation based on a comprehensive virtual generator and battery model. By establishing a refined mathematical model of distributed power sources, distributed loads, distributed energy storage and distributed networks, the present invention provides an accurate representation of the source-grid-load-storage (distributed power generation, power grid, load, energy storage) synergistic interaction of distributed source-grid-load-storage energy systems; at the same time, the aggregation model of the comprehensive virtual generator and virtual battery is established by comprehensively considering power constraints, ramp constraints and capacity constraints, ensuring the efficiency and reliability of the optimized operation of the distributed source-grid-load-storage energy system. This method can fully reflect the complexity and dynamics of the actual operation of the system, and provides strong support for the optimized operation of the distributed source-grid-load-storage energy system.

[0067] The important innovations of the present invention are that the operating characteristics of multiple energy devices in the system are effectively aggregated using an aggregate model of integrated virtual generators and virtual batteries, and that a two-stage robust optimization parameter solution method for the integrated aggregate model is proposed. This aggregation processing and solution method not only optimizes the resource allocation in the multi-distributed source-grid-load-storage energy system, and improves the economy and flexibility of the distributed source-grid-load-storage energy system, but also realizes the effective processing of high-dimensional convex polyhedrons through the application of half-plane representation and the Minkowski sum principle, so that the operating characteristics of multiple energy devices are optimized at the system level, and can adapt to parameter uncertainty and changes in the external environment, thereby enhancing the robustness of the distributed source-grid-load-storage energy system. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 It is a design flow chart of the energy storage and load aggregation method based on the comprehensive virtual generator and battery model in Example 1 of the present invention;

[0069] Figure 2 This is a schematic diagram of Minkowski approximate linear summation in Example 1 of the present invention;

[0070] Figure 3 It is a flow chart of the energy storage and load aggregation method based on the comprehensive virtual generator and battery model in Example 2 of the present invention;

[0071] Figure 4 The schematic diagram of the energy storage and load aggregation system based on the comprehensive virtual generator and battery model of the present invention is shown in FIG. DETAILED DESCRIPTION

[0072] In order to make the purpose, technical solution and advantages of the present invention more clear, the following is further described in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention, that is, the embodiments described are only part of the embodiments of the present invention, not all embodiments.

[0073] The detailed description of the embodiments of the present invention provided in the following drawings is not intended to limit the scope of the invention claimed for protection, but merely represents a selected embodiment of the present invention. Based on the drawings and embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0074] The features and performances of the present invention are further described in detail below in conjunction with the embodiments.

[0075] Example 1

[0076] From a design perspective, the energy storage and load aggregation method based on the integrated virtual generator and battery model of the present invention is described, such as Figure 1 As shown, it includes the following parts:

[0077] (I) Establishing a refined mathematical model for distributed power generation, grid, load and storage

[0078] (1.1) Distributed power generation refined mathematical model

[0079] Distributed power sources refer to various power generation equipment dispersedly deployed in the power system, which can meet the power supply of local or specific needs. These distributed power sources can be new energy or traditional energy, mainly including solar photovoltaic power generation systems, wind power generation systems, small hydropower systems, biomass power generation systems, geothermal power generation systems, micro nuclear power generation systems, etc.

[0080] In terms of mathematical models, new energy usually only constrains the unit output and does not consider the impact of the unit's climbing, while traditional energy not only constrains the unit output but also considers the limitations of the unit's climbing.

[0081] The new energy unit can be modeled as:

[0082]

[0083] Where: R represents the set of new energy generating units in the distributed source-grid-load-storage energy system; T represents the set of scheduling periods in the distributed source-grid-load-storage energy system; P r,t,max is the maximum output that the new energy generator r can actually generate during period t; P r,t It is the active power output of the new energy generator r in period t.

[0084] Traditional energy units can be modeled as:

[0085]

[0086] Where: G represents the set of traditional energy generating units in the distributed source-grid-load-storage energy system; P g,tare respectively the active output of the traditional energy generator g in period t; P g,max is the upper limit of active output of traditional energy generator g; P g,min is the lower limit of active power output of traditional energy generator g; R g,max is the upper limit of the output adjustment of the traditional energy generator g, R g,min It is the lower limit of the output adjustment of the traditional energy generator set g.

[0087] (1.2) Distributed load refinement mathematical model

[0088] Distributed load refers to various electrical equipment and loads distributed in the power system, which are usually scattered in different locations and users, including household and commercial loads such as lighting, air conditioning, and electrical appliances, industrial loads such as motors, electric boilers, and electric refrigeration, agricultural loads such as irrigation, ventilation, heating, and agricultural machinery, transportation loads such as electric vehicles, electric bicycles, and electric buses, and communication equipment loads such as communication base stations and data centers. These different loads can flexibly adjust their electricity consumption behavior and save electricity costs under the guidance of electricity prices or incentives according to their own needs and economic conditions, which is load demand response.

[0089] According to the type of load response, it can be divided into shiftable load and curtailable load, which correspond to the transfer in time and the reduction in electricity respectively.

[0090] Specifically, the translatable load can be moved over time while keeping the total demand unchanged, that is, adjusting the working time while ensuring the completion of the total task volume.

[0091] The curtailable load can be cut off within a limited range within a specific period of time.

[0092] The translatable load can be modeled as:

[0093]

[0094] Where: D represents the set of responsive loads in the distributed source-grid-load-storage energy system; Ω T+ represents the time period set of load transfer; Ω T- represents the time period set of load removal; P shift+,d,t P represents the load amount that can be moved in by the response load d within the time period t; shift-,d,t P represents the load that can be removed in response to load d within time period t; shift+,d,max and P shift+,d,min They represent the maximum and minimum values ​​of the load that can be moved into the response load d within the time period t; P shift-,d,max and P shift-,d,min They respectively represent the maximum and minimum values ​​of the load that can be removed by the response load d within the time period t.

[0095] The curtailable load can be modeled as:

[0096]

[0097] Where: D represents the set of responsive loads in the distributed source-grid-load-storage energy system; Ω Tcut represents the set of time periods for load reduction; P cut,d,t P represents the load reduction amount of the responsive load d within the time period t; cut,d,max and P cut,d,min They respectively represent the maximum and minimum values ​​of the load that can be reduced by the responding load d in a unit time period.

[0098] (1.3) Refined mathematical model of distributed energy storage

[0099] Distributed energy storage refers to various energy storage devices distributed in the power system, which are used to store electrical energy and release it when needed to balance the supply and demand of the power system and improve the stability of the system.

[0100] Common distributed energy storage devices include lithium-ion battery energy storage systems, supercapacitor energy storage systems, compressed air energy storage systems, and electric vehicle battery energy storage systems.

[0101] Distributed energy storage devices can be uniformly modeled as:

[0102]

[0103] Where: ES represents the collection of different distributed energy storage devices in the distributed source-grid-load-storage energy system; E es,t , P c,es,t and P dc,es,t represents the state of charge, charging power and discharging power of the energy storage device es in time period t; ζ c,E and dc,E are the charging efficiency and discharging efficiency of the energy storage device es respectively; E es,min 、E es,max Indicates the upper and lower limits of the capacity of the energy storage device es; P c,es,max represents the upper limit of the charging power of the energy storage device es; P c,es,min Represents the lower limit of the charging power of the energy storage device es; P dc,es,max Represents the upper limit of the discharge power of the energy storage device es; P dc,es,min Indicates the lower limit of the discharge power of the energy storage device es.

[0104] (1.4) Distributed network refined mathematical model

[0105] For microgrids and integrated energy systems in parks where distributed resources are relatively concentrated, such as commercial complexes and small industrial parks, their scale is relatively small compared to traditional transmission and distribution networks, and the scope of service is relatively limited. Therefore, in terms of the network, only power balance constraints are usually considered to simplify the modeling and operation management of microgrids and integrated energy systems in parks where distributed resources are relatively concentrated, and reduce system complexity.

[0106] The power balance constraint of the distributed source-grid-load-storage energy system consists of two parts: one is the overall power balance constraint of the distributed source-grid-load-storage energy system, and the other is the power balance constraint before and after load demand response.

[0107] The overall power balance constraint of the distributed source-grid-load-storage energy system is:

[0108]

[0109] Where: P d,t is the load demand after the respondable load d performs demand response; P tie,t is the interconnection power between the distributed source-grid-load-storage energy system and the large power grid, if P tie,t is positive, indicating that the distributed source-grid-load-storage energy system purchases electricity from the large power grid. If P tie,t A negative value indicates that the distributed source-grid-load-storage energy system sells electricity to the large power grid.

[0110] The power balance constraint before and after load demand response is:

[0111]

[0112] Where: P d,t is the load demand after the respondable load d performs demand response; P d0,t is the load demand before the responding load d performs demand response; P shift,d,t is the load for shifting demand response; P cut,d,t The load required for demand response reduction; P shift-,d,t is the load shifted out during period t; P shift+,d,t is the load shifted out during period t.

[0113] (II) Establishing an aggregation model of comprehensive virtual generators and virtual batteries

[0114] Virtual power plants play an important role between the power grid management system and distributed energy. They need to aggregate the operating characteristics of distributed energy instead of directly reporting the detailed physical models of all distributed energy to the superior power grid. This is mainly due to two considerations: First, to simplify communication and optimization problem solving: if the detailed physical models of all distributed energy are reported directly to the superior power grid, model information may be missing or wrong during the communication process, and the global optimization problem of the power grid is difficult to solve due to the large scale of variables. By aggregating the operating characteristics of distributed energy, the communication process and the solution of optimization problems can be greatly simplified, and the efficiency and stability of the system can be improved. Second, to protect commercial secrets and maintain market competitive advantages: virtual power plants often participate in various transactions in the power market driven by value. In order to maintain an absolute advantage in market competition, they are not willing to disclose the detailed information of their internal models to the power grid. By aggregating the operating characteristics of distributed energy, virtual power plants can protect commercial secrets and maintain their dominant position in market competition. Therefore, aggregating the operating characteristics of distributed energy in virtual power plants is an effective management and optimization method, which helps to simplify system operation and protect commercial secrets, thereby better realizing participation in the power market and optimization of power grid dispatching.

[0115] The technical characteristics of distributed energy are usually described in detail through power inequality constraints, which mainly include three categories: power constraints, ramp constraints, and capacity constraints. Taking into account the power constraints of all distributed energy sources, the adjustable power domain of the virtual power plant can be mapped into a high-dimensional convex polyhedron in a geometric space. In order to solve the adjustable power domain of this virtual power plant, a specific high-dimensional convex polyhedron can be used to approximate it from the inside or outside. At present, there are two main mathematical models of suitable and widely used high-dimensional convex polyhedrons: virtual battery model and virtual generator model.

[0116] However, in order to more comprehensively consider the power constraints, ramp constraints and capacity constraints of distributed energy, the present invention adopts a comprehensive model, namely, an aggregate model of comprehensive virtual generators and virtual batteries. This comprehensive model can more accurately reflect the actual operating conditions of virtual power plants and provide a more reliable decision-making basis for energy management.

[0117] The aggregation model of the integrated virtual generator and virtual battery is as follows:

[0118]

[0119] Where: Ω is the feasible domain of the interaction power between the aggregate and the large power grid constrained by the aggregation model; P Juhe,t is the interaction power between the aggregate and the large power grid in time period t, corresponding to the interaction power P between the interconnection line of the distributed source-grid-load-storage energy system and the large power grid tie,t ; is the power of the aggregate interacting with the large power grid in period t; P Juhe,t,max and P Juhe,t,min is the upper and lower limits of the power of the aggregate interacting with the large power grid in time period t, which is a time-varying parameter; E Juhe,t,max and E Juhe,t,min is the upper and lower limits of the capacity of the aggregate in time period t, which is a time-varying parameter; R Juhe,max and R Juhe,min It is the upper and lower limits of the power ramp of the aggregate interacting with the large power grid during each period, and is a constant parameter.

[0120] The first formula of formula (16) represents the power constraint, the second formula represents the capacity constraint, and the third formula represents the ramp constraint.

[0121] (III) Establishing a robust optimization solution model for aggregation model parameters

[0122] There are many parameters to be determined in the aggregation model of virtual generators and virtual batteries, including the upper and lower limits of the power of the aggregate in the interaction with the large grid at time period t. Juhe,t,max and P Juhe,t,min , the upper and lower limits of the capacity of the aggregate in time period t E Juhe,t,max and E Juhe,t,min And the upper and lower limits of the power of the aggregate to interact with the large power grid during each period R Juhe,max and R Juhe,min .

[0123] The above parameters can be solved by constructing a two-stage robust optimization model. The first-stage variable x is the parameter to be solved of the comprehensive aggregation model. The second-stage variable y is the output of distributed power sources in the distributed source-grid-load-storage energy system, the demand response of distributed loads, the capacity and charging and discharging power of distributed energy storage equipment, the large grid interaction power, and the newly introduced power balance constraint non-negative slack variable κ. t,+ and κ t,- The two-stage robust optimization model aims to minimize the maximum power balance deviation of the comprehensive aggregation model, that is, to find a comprehensive aggregation model that fits the actual power feasible domain of the distributed source-grid-load-storage energy system as closely as possible.

[0124] The objective function of the two-stage robust optimization model is:

[0125]

[0126] Where: x represents the first-stage variable, which is the parameter to be determined in the comprehensive aggregation model; y represents the second-stage variable, including the output of distributed power sources in the distributed source-grid-load-storage energy system, the demand response of distributed loads, the capacity and charging and discharging power of distributed energy storage equipment, the large-grid interactive power, and the newly introduced non-negative slack variable of power balance constraint; γ P,min , γ P,max, γ E,min , γ E,max , γ R,min and γ R,max are the weight coefficients of the corresponding parameters respectively; M is a large real number; κ t,+ and κ t,- is a non-negative slack variable that introduces power balance constraints.

[0127] First stage constraints:

[0128]

[0129] Where: P max , P min , R max and R min Represents the initial bounds of the parameters to be sought for the comprehensive aggregation model.

[0130] The constraints of the second stage are the refined mathematical model of distributed source, grid, load and storage, in which the overall power balance constraint of the distributed source, grid, load and storage energy system should be changed to:

[0131]

[0132] The constraints for coupling the first and second stages are the aggregation model of the integrated virtual generator and virtual battery.

[0133] The two-stage robust optimization model can decompose the original problem into a main problem and sub-problems and solve them alternately through the Column & Constraint Generation (C&CG) algorithm or the Benders decomposition algorithm. A considerable amount of research has been done to explain it, which will not be repeated here.

[0134] (IV) Minkowski approximate linear summation method for comprehensive aggregation model

[0135] According to the Minkowski approximate linear summation method, the aggregate model is approximated linearly to obtain the adjustable power domain of the aggregate. The specific process is as follows:

[0136] Firstly, the adjustable power domain of a single distributed source-grid-load-storage energy system is mapped to a geometric space to form a high-dimensional convex polyhedron; the vertex set or half-plane set of the high-dimensional convex polyhedron corresponding to each distributed source-grid-load-storage energy system is solved, and the vertex set or half-plane set is used to represent the adjustable power domain of each distributed source-grid-load-storage energy system;

[0137] Then, according to the Minkowski approximate linear summation method, the vertices or half-planes corresponding to the high-dimensional convex polyhedron of the adjustable power domains of all distributed source-grid-load-storage energy systems are arranged in sequence and added accordingly to obtain a new vertex set or half-plane set;

[0138] New vertex sets or half-plane sets are used to represent adjustable power domains for multiple aggregates.

[0139] Compared with vertex set representation, half-plane representation is more intuitive and more applicable for the feasible domain representation of distributed source-grid-load-storage energy systems. Therefore, the half-plane form is used to represent high-dimensional convex polyhedrons, and its compact form is:

[0140]

[0141] Where: Ω (j) is the adjustable power domain of the distributed source-grid-load-storage energy system aggregate j described by the half-plane method; is the power P adjusted by the distributed source-grid-load-storage energy system aggregate j in each period of the scheduling cycle T Juhe,t The column vector formed by E, A (j) is the power constraint coefficient matrix corresponding to the distributed source-grid-load-storage energy system aggregate j; is the power constraint parameter vector corresponding to the distributed source-grid-load-storage energy system aggregate j.

[0142] The following uses two distributed source-grid-load-storage energy system aggregates as examples to introduce the Minkowski approximate linear summation method. Figure 2 As shown in the figure, the adjustable power domain of the two distributed source-grid-load-storage energy system aggregates obtained by Minkowski approximate linear summation can be expressed as:

[0143]

[0144] Where: is the Minkowski summation symbol; The adjustable power domain of the aggregate of two distributed source-grid-load-storage energy systems described by the half-plane method; It is a column vector composed of the power adjustment of the two distributed source-grid-load-storage energy system aggregates at each moment within the scheduling period T.

[0145] Example 2

[0146] like Figure 3 As shown, from the application perspective, the present invention discloses a method for energy storage and load aggregation based on a comprehensive virtual generator and battery model, including the following processes:

[0147] According to the constructed aggregation model of comprehensive virtual generators and virtual batteries, multiple energy devices in multiple distributed source-grid-load-storage energy systems are aggregated to obtain an aggregate;

[0148] For the undetermined parameters in the aggregation model, the parameters are solved according to the constructed parameter robust optimization solution model; the parameter robust optimization solution model is divided into two-stage constraints, and the second-stage constraint conditions are established based on the distributed source-grid-load-storage refined mathematical model; the distributed source-grid-load-storage refined mathematical model includes a distributed power source refined mathematical model, a distributed load refined mathematical model, a distributed energy storage refined mathematical model and a distributed network refined mathematical model;

[0149] According to the Minkowski approximate linear summation method, the aggregation model is approximated linearly summed to obtain the adjustable power domain of the aggregate. Based on the adjustable power domain, the control strategy is specified to control the distributed source-grid-load-storage energy system to achieve resource aggregation of multiple distributed source-grid-load-storage energy systems.

[0150] Specifically, the corresponding control strategy is formulated based on the analysis of the adjustable power domain and the actual needs of the system. For example, if the current grid load is large and the power supply needs to be increased, this can be achieved by increasing the output power of distributed power sources, releasing the energy of distributed energy storage devices, or adjusting the power consumption of adjustable loads. The control strategy should comprehensively consider multiple factors such as energy cost, equipment life, and power supply reliability.

[0151] According to the formulated control strategy, energy equipment is precisely controlled through electronic controllers and other equipment. For distributed power sources, their output power setting value can be adjusted; for distributed energy storage equipment, their charging and discharging power can be controlled; for distributed loads, their power consumption can be adjusted through intelligent control devices. In the process of implementing control actions, the response of the system should be monitored in real time to ensure that the control effect meets expectations.

[0152] Example 3

[0153] like Figure 4 As shown, the present invention discloses an energy storage and load aggregation system based on a comprehensive virtual generator and battery model, comprising:

[0154] An aggregation module is used to aggregate multiple energy devices in multiple distributed source-grid-load-storage energy systems according to the constructed aggregation model of comprehensive virtual generators and virtual batteries to obtain an aggregate;

[0155] A solution module is used to solve the parameters of the undetermined parameters in the aggregation model according to the constructed parameter robust optimization solution model; the parameter robust optimization solution model is divided into two-stage constraints, and the second-stage constraint conditions are established based on the distributed source-grid-load-storage refined mathematical model; the distributed source-grid-load-storage refined mathematical model includes a distributed power source refined mathematical model, a distributed load refined mathematical model, a distributed energy storage refined mathematical model and a distributed network refined mathematical model;

[0156] The adjustable power domain calculation module is used to perform approximate linear summation on the aggregation model according to the Minkowski approximate linear summation method, solve the adjustable power domain of the aggregate, formulate a control strategy based on the adjustable power domain, control the distributed source-grid-load-storage energy system, and realize resource aggregation of multiple distributed source-grid-load-storage energy systems.

[0157] The operations of the flow chart in the present invention may be implemented out of order, or may be implemented in reverse order or simultaneously. In addition, one or more other operations may be added to the flow chart. One or more operations may be removed from the flow chart.

[0158] Through the description of the above implementation modes, those skilled in the art can clearly understand that the present disclosure can be implemented by means of software plus necessary general hardware, and of course can also be implemented by special hardware including special integrated circuits, special CPUs, special memories, special components, etc. In general, all functions performed by computer programs can be easily implemented by corresponding hardware, and the specific hardware structures used to implement the same function can also be various, such as analog circuits, digital circuits or special circuits, etc. However, for the present disclosure, software program implementation is a better implementation mode in most cases.

[0159] In summary, the method of the present invention aims at the characteristics of the coordinated interaction of distributed source-grid-load-storage energy systems, and respectively establishes refined mathematical models of distributed power sources, distributed loads, distributed energy storage and distributed networks; in order to comprehensively consider the power constraints, climbing constraints and capacity constraints of the distributed source-grid-load-storage energy systems, an aggregation model of comprehensive virtual generators and virtual batteries is used to aggregate the operating characteristics of multiple energy devices in the system; for many undetermined parameters in the comprehensive aggregation model, a two-stage robust optimization parameter solution method of the comprehensive aggregation model is proposed to solve the parameters; according to the Minkowski sum principle, different comprehensive aggregation models are approximated linearly summed to obtain an adjustable power domain of the aggregate, and a control strategy is formulated based on the adjustable power domain to control the distributed source-grid-load-storage energy system, so as to realize resource aggregation of multiple distributed source-grid-load-storage energy systems.

[0160] The main application scenario targeted by the present invention is the virtual power plant, which is one of the important technologies with great prospects and application potential in the current energy industry. The core goal of the virtual power plant is to achieve integrated management and optimized scheduling of distributed energy resources (such as photovoltaics, wind energy, energy storage equipment, loads, etc.) by integrating these distributed energy resources. Based on the above needs, the present invention proposes a method for energy storage and load aggregation based on a comprehensive virtual generator and battery model, which can effectively solve these practical industrial problems. Specifically, the method in the invention can aggregate and optimize resources for multiple distributed energy systems while taking uncertainty into account by establishing a comprehensive aggregation model and a robust optimization solution model.

[0161] In actual industrial operations, the power grid dispatching system needs to optimize the control and dispatch of multiple distributed energy devices. The comprehensive virtual generator and virtual battery model proposed in the present invention, through aggregation processing, converts the operating characteristics of many distributed energy sources into a simpler and controllable dispatching model, which greatly simplifies the complexity of dispatching. For example, through the refined mathematical models of new energy and traditional energy generators, the power grid dispatching center can monitor and control the output power of these devices in real time, thereby ensuring the stability of power supply; in addition, the energy storage device model in the present invention can help the dispatching center flexibly arrange the charging and discharging strategies of energy storage devices according to the real-time load conditions of the power grid, balance the supply and demand of electricity, and avoid instability of the power grid; the refined modeling of distributed loads, especially the shiftable load and reducible load models, can enable the power dispatching system to reduce the load pressure of the power grid and improve the stability of the system by reducing or delaying the power consumption time of specific loads during peak power demand periods.

[0162] In the actual industrial operation of virtual power plants, the operating environment and external conditions of the equipment (such as weather, load demand, and market electricity prices) have great uncertainty. The two-stage robust optimization solution model of the present invention can effectively solve these common fluctuation problems in actual industrial operations by introducing uncertain variables: the constraints of the first stage impose specific power constraints on the output power of distributed energy equipment and the charging and discharging power of energy storage equipment. These constraints are dynamically adjusted according to the actual needs of the power grid. The constraints of the second stage are established based on the actual mathematical model of distributed source, network, load and storage, and can handle complex power trading and scheduling problems in different power market environments. Through the two-stage optimization solution method, the power grid dispatching system can ensure the overall stability and economy of the power grid by adjusting the power output, charging and discharging strategies, and load response behaviors of each distributed energy device in actual industrial scenarios.

[0163] In view of the uncertain factors such as external environmental fluctuations and equipment failures in actual industrial applications, the robust design proposed in this invention has strong industrial applicability. Through the model and algorithm of this invention, the virtual power plant can still maintain the stable operation of the system when facing these practical problems, and ensure the normal operation of various distributed energy systems in different environments. This has great practical significance for the current large-scale grid connection of new energy and the improvement of the intelligent management level of the power grid.

[0164] In summary, the present invention is closely related to actual industrial activities, especially in virtual power plant technology, optimization management of distributed energy systems, power dispatching and other actual industrial applications. Through the aggregation processing of comprehensive virtual generator and battery models, as well as robust optimization solution and Minkowski approximate linear summation method, the present invention solves many technical problems in distributed energy dispatching and optimization in actual industrial operations, and provides effective technical support for achieving more efficient energy management.

[0165] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for energy storage and load aggregation based on a comprehensive virtual generator and battery model, characterized in that: The process includes: According to the constructed aggregation model of comprehensive virtual generators and virtual batteries, multiple energy devices in multiple distributed source-grid-load-storage energy systems are aggregated to obtain an aggregate; For the undetermined parameters in the aggregation model, the parameters are solved according to the constructed parameter robust optimization solution model; the parameter robust optimization solution model is divided into two-stage constraints, and the second-stage constraint conditions are established based on the distributed source-grid-load-storage refined mathematical model; the distributed source-grid-load-storage refined mathematical model includes a distributed power source refined mathematical model, a distributed load refined mathematical model, a distributed energy storage refined mathematical model and a distributed network refined mathematical model; According to the Minkowski approximate linear summation method, the aggregation model is approximated linearly summed to obtain the adjustable power domain of the aggregate. Based on the adjustable power domain, a control strategy is formulated to control the distributed source-grid-load-storage energy system to achieve resource aggregation of multiple distributed source-grid-load-storage energy systems.

2. The energy storage and load aggregation method based on the integrated virtual generator and battery model according to claim 1 is characterized in that: The refined mathematical model of distributed power generation is divided into the mathematical model of new energy units and the mathematical model of traditional energy units; The mathematical model of the new energy unit is: Where: R represents the set of new energy generating units in the distributed source-grid-load-storage energy system; T represents the set of scheduling periods in the distributed source-grid-load-storage energy system; P r,t,max P is the maximum output that the new energy generator r can actually generate during period t; r,t is the active power output of the new energy generator r in period t; The mathematical model of traditional energy units is: Where: G represents the set of traditional energy generating units in the distributed source-grid-load-storage energy system; P g,t are respectively the active output of the traditional energy generator g in period t; P g,max is the upper limit of active output of traditional energy generator g; P g,min is the lower limit of active power output of traditional energy generator g; R g,max is the upper limit of the output adjustment of the traditional energy generator g, R g,min It is the lower limit of the output adjustment of the traditional energy generator set g.

3. The energy storage and load aggregation method based on the integrated virtual generator and battery model according to claim 1 is characterized in that: Distributed loads are divided into shiftable loads and curtailable loads according to the type of load response. Shiftable loads correspond to time transfers, and curtailable loads correspond to power reductions. The distributed load refinement mathematical model is divided into a mathematical model of shiftable load and a mathematical model of curtailable load; The mathematical model of the translatable load is: Where: D represents the set of responsive loads in the distributed source-grid-load-storage energy system; Ω T+ represents the time period set of load transfer; Ω T- represents the time period set of load removal; P shift+,d,t P represents the load amount that can be moved in by the response load d within the time period t; shift-,d,t P represents the load that can be removed in response to load d within time period t; shift+,d,max and P shift+,d,min They represent the maximum and minimum values ​​of the load that can be moved into the response load d within the time period t; P shift-,d,max and P shift-,d,min They represent the maximum and minimum values ​​of the load that can be removed by the response load d within the time period t respectively; The mathematical model of load reduction is: Where: D represents the set of responsive loads in the distributed source-grid-load-storage energy system; Ω Tcut represents the set of time periods for load reduction; P cut,d,t P represents the load reduction amount of the responsive load d within the time period t; cut,d,max and P cut,d,min They respectively represent the maximum and minimum values ​​of the load that can be reduced by the responding load d in a unit time period.

4. The energy storage and load aggregation method based on the integrated virtual generator and battery model according to claim 1 is characterized in that: The refined mathematical model of distributed energy storage is: Where: ES represents the collection of different distributed energy storage devices in the distributed source-grid-load-storage energy system; E es,t , P c,es,t and P dc,es,t represents the state of charge, charging power and discharging power of the energy storage device es in time period t; ζ c,E and dc,E are the charging efficiency and discharging efficiency of the energy storage device es respectively; E es,min 、E es,max Indicates the upper and lower limits of the capacity of the energy storage device es; P c,es,max Indicates the upper limit of the charging power of the energy storage device es; P c,es,min Represents the lower limit of the charging power of the energy storage device es; P dc,es,max Indicates the upper limit of the discharge power of the energy storage device es; P dc,es,min Indicates the lower limit of the discharge power of the energy storage device es.

5. The energy storage and load aggregation method based on the integrated virtual generator and battery model according to claim 1 is characterized in that: The distributed network refined mathematical model only considers the power balance constraint when establishing it; the power balance constraint of the distributed source-grid-load-storage energy system includes the overall power balance constraint of the distributed source-grid-load-storage energy system and the power balance constraint before and after the load demand response; The overall power balance constraint of the distributed source-grid-load-storage energy system is: Where: P d,t is the load demand after the respondable load d performs demand response; P tie,t is the interconnection power between the distributed source-grid-load-storage energy system and the large power grid, if P tie,t is positive, indicating that the distributed source-grid-load-storage energy system purchases electricity from the large power grid. If P tie,t If it is negative, it means that the distributed source-grid-load-storage energy system sells electricity to the large power grid; The power balance constraint before and after load demand response is: Where: P d,t is the load demand after the respondable load d performs demand response; P d0,t is the load demand before the responding load d performs demand response; P shift,d,t The load for shifting demand response; P cut,d,t The load required for demand response reduction; P shift-,d,t is the load shifted out during period t; P shift+,d,t is the load shifted out during period t.

6. The energy storage and load aggregation method based on the integrated virtual generator and battery model according to claim 1 is characterized in that: The aggregation model of the integrated virtual generator and virtual battery is shown as follows: Where: Ω is the feasible domain of the interaction power between the aggregate and the large power grid constrained by the aggregation model; P Juhe,t is the interaction power between the aggregate and the large power grid in time period t, corresponding to the interaction power P between the interconnection line of the distributed source-grid-load-storage energy system and the large power grid tie,t ; is the power of the aggregate interacting with the large power grid in period t; P Juhe,t,max is the upper limit of the power of the aggregate interacting with the large power grid in period t, P Juhe,t,min is the lower power limit of the aggregate interacting with the large grid in period t, P Juhe,t,max and P Juhe,t,min is a time-varying parameter; E Juhe,t,max is the upper limit of the capacity of the aggregate in time period t, E Juhe,t,min is the lower limit of the capacity of the aggregate in time period t, E Juhe,t,max and E Juhe,t,min is a time-varying parameter; R Juhe,max is the upper limit of the power ramp of the aggregate interacting with the large power grid during each period, R Juhe,min is the lower limit of the power ramp of the aggregate interacting with the large power grid during each period, R Juhe,max and R Juhe,min is a constant parameter.

7. The energy storage and load aggregation method based on the integrated virtual generator and battery model according to claim 6 is characterized in that: There are several parameters to be determined in the aggregation model of virtual generators and virtual batteries, including P Juhe,t,max and P Juhe,t,min 、E Juhe,t,max and E Juhe,t,min and R Juhe,max and R Juhe,min ; The parameter robust optimization solution model is a two-stage robust optimization model, which is used to solve the above parameters to be determined; The objective function of the two-stage robust optimization model is: Where: x represents the first-stage variable, which is the parameter to be determined for the aggregation model; y represents the second-stage variable, including the output of distributed power sources in the distributed source-grid-load-storage energy system, the demand response of distributed loads, the capacity and charging and discharging power of distributed energy storage equipment, the large-grid interactive power, and the newly introduced non-negative slack variable of the power balance constraint; γ P,min , γ P,max , γ E,min , γ E,max , γ R,min and γ R,max are the weight coefficients of the corresponding parameters respectively; M is a positive number; κ t,+ and κ t,- is the non-negative slack variable for introducing power balance constraints; First stage constraints: Where: P max , P min , R max and R min It represents the initial bounds of the parameters to be determined for the comprehensive aggregation model; The constraints of the second stage are the refined mathematical model of distributed source, grid, load and storage, in which the overall power balance constraint of the distributed source, grid, load and storage energy system is replaced by: The constraints for coupling the first and second stages are the aggregation model of the integrated virtual generator and virtual battery.

8. The energy storage and load aggregation method based on the integrated virtual generator and battery model according to claim 1 is characterized in that: The aggregate model is approximated linearly by the Minkowski approximate linear summation method to obtain the adjustable power domain of the aggregate. The specific process is as follows: Firstly, the adjustable power domain of a single distributed source-grid-load-storage energy system is mapped to a geometric space to form a high-dimensional convex polyhedron; the vertex set or half-plane set of the high-dimensional convex polyhedron corresponding to each distributed source-grid-load-storage energy system is solved, and the vertex set or half-plane set is used to represent the adjustable power domain of each distributed source-grid-load-storage energy system; Then, according to the Minkowski approximate linear summation method, the vertices or half-planes corresponding to the high-dimensional convex polyhedron of the adjustable power domains of all distributed source-grid-load-storage energy systems are arranged in sequence and added accordingly to obtain a new vertex set or half-plane set; New vertex sets or half-plane sets are used to represent adjustable power domains for multiple aggregates.

9. The energy storage and load aggregation method based on the integrated virtual generator and battery model according to claim 8 is characterized in that: The high-dimensional convex polyhedron is represented in the form of a half-plane, and the compact form of the representation is: Where: Ω (j) is the adjustable power domain of aggregate j described by the half-plane method; is the power P adjusted by aggregate j in each period within the scheduling period T Juhe,t The column vector formed by E, A (j) is the power constraint coefficient matrix corresponding to aggregate j; is the power constraint parameter vector corresponding to aggregate j.

10. An energy storage and load aggregation system based on an integrated virtual generator and battery model, characterized in that: include: An aggregation module is used to aggregate multiple energy devices in multiple distributed source-grid-load-storage energy systems according to the constructed aggregation model of comprehensive virtual generators and virtual batteries to obtain an aggregate; A solution module is used to solve the parameters of the undetermined parameters in the aggregation model according to the constructed parameter robust optimization solution model; the parameter robust optimization solution model is divided into two-stage constraints, and the second-stage constraint conditions are established based on the distributed source-grid-load-storage refined mathematical model; the distributed source-grid-load-storage refined mathematical model includes a distributed power source refined mathematical model, a distributed load refined mathematical model, a distributed energy storage refined mathematical model and a distributed network refined mathematical model; An adjustable power domain calculation module is used to perform approximate linear summation on the aggregate model according to the Minkowski approximate linear summation method to obtain an adjustable power domain of the aggregate; Based on the adjustable power domain, a control strategy is formulated to control the distributed source-grid-load-storage energy system and realize resource aggregation of multiple distributed source-grid-load-storage energy systems.

Images