CN116415740A - Two-stage robust optimization method for virtual power plant based on saddle uncertainty - Google Patents

Two-stage robust optimization method for virtual power plant based on saddle uncertainty Download PDF

Info

Publication number
CN116415740A
CN116415740A CN202310663465.4A CN202310663465A CN116415740A CN 116415740 A CN116415740 A CN 116415740A CN 202310663465 A CN202310663465 A CN 202310663465A CN 116415740 A CN116415740 A CN 116415740A
Authority
CN
China
Prior art keywords
uncertainty
power plant
virtual power
model
load
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202310663465.4A
Other languages
Chinese (zh)
Other versions
CN116415740B (en
Inventor
李付林
黄红辉
王珂
李毓
叶宏
季克勤
侯健生
张波
程朝阳
马骏达
徐耀辉
贺燕
郭创新
徐敏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jinhua Bada Group Co ltd
Jinhua Power Supply Co of State Grid Zhejiang Electric Power Co Ltd
Original Assignee
Jinhua Bada Group Co ltd
Jinhua Power Supply Co of State Grid Zhejiang Electric Power Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jinhua Bada Group Co ltd, Jinhua Power Supply Co of State Grid Zhejiang Electric Power Co Ltd filed Critical Jinhua Bada Group Co ltd
Priority to CN202310663465.4A priority Critical patent/CN116415740B/en
Publication of CN116415740A publication Critical patent/CN116415740A/en
Application granted granted Critical
Publication of CN116415740B publication Critical patent/CN116415740B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06315—Needs-based resource requirements planning or analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/04—Constraint-based CAD
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2113/00—Details relating to the application field
    • G06F2113/04—Power grid distribution networks
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Health & Medical Sciences (AREA)
  • Game Theory and Decision Science (AREA)
  • Operations Research (AREA)
  • Development Economics (AREA)
  • Water Supply & Treatment (AREA)
  • Public Health (AREA)
  • Educational Administration (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

The disclosure provides a two-stage robust optimization method of a virtual power plant based on saddle uncertainty, which acts on a source network load storage linkage environment, and comprises the following steps: obtaining a conventional model of a virtual power plant, correcting an endogenous uncertainty reference distribution by using a local discrete super , introducing a relative entropy and a box-type polyhedral uncertainty set to optimize the theory model, and generating an endogenous-exogenous uncertainty model; and solving a two-stage robust optimization model by using a column and constraint generation algorithm and combining a strong dual theorem and a linearization theory. And the economical efficiency and the low carbon coordinated operation of the cold-hot electricity virtual power plant are realized. Compared with a common robust optimization model, the two-stage robust optimization model is more flexible, and the optimization performance can be improved while the robustness is maintained.

Description

Two-stage robust optimization method for virtual power plant based on saddle uncertainty
Technical Field
The disclosure relates to the technical field of virtual power plant optimization, in particular to a two-stage robust optimization method of a virtual power plant based on saddle uncertainty.
Background
The virtual power plant (Virtual Power Plant, VPP) integrates a plurality of distributed energy systems (such as solar energy, wind energy, energy storage and the like) and energy resources such as traditional generator sets, load-dissipating and the like through means such as the internet and smart grid technology and the like to form a virtual power production, scheduling and transaction system. The virtual power plant can be regarded as an integrated management system of energy, can furthest improve the utilization efficiency of energy, optimize the operation and the dispatch of system, reduce the power production cost, and provide more convenient, flexible and diversified energy service for end users. The method has the advantages of reducing wind and light abandoning, ensuring the stability of the power grid, improving the flexibility of the system, reducing the influence of distributed energy on the power grid, and being an important mode for promoting carbon emission reduction and ensuring the stability of the power grid.
However, most of the existing researches are aimed at virtual power plant model construction, and the uncertainty of the distributed power supply output, the internet power price and the energy consumption load is reduced, so that the effectiveness of the virtual power plant optimization scheduling is influenced by neglecting the uncertainty. In uncertainty analysis, classification and modeling of uncertainty in the absence of a priori information directly relates to the accuracy of the virtual power plant optimization method. In the virtual power plant optimization problem, the robust optimization method replaces the exact probability distribution of random variables with an uncertain set, and the scheduling scheme of the system under the worst scene is obtained through an optimization means, so that the system meets the requirements of actual engineering more. For example, in the prior art, the publication number CN115422728A, entitled "virtual power plant optimization control system based on robust optimization of random planning" is applied to the chinese invention, which discloses a virtual power plant optimization control system based on robust optimization of random planning, including a virtual power plant optimization control module for safe operation of an economic and power grid, a virtual power plant random planning optimization control module for random renewable energy, and a virtual power plant adaptive robust optimization control module for random renewable energy, where the virtual power plant optimization control module for safe operation of the economic and power grid includes a virtual power plant economic dispatch model and a virtual power plant safety dispatch model. The virtual power plant optimization control system based on the robust optimization of the random programming disclosed by the invention considers the coordinated operation control mode among distributed power sources, energy storage and demand side users in an area under the condition of random renewable energy sources, and has the characteristics of presenting stable power output to a large power grid under the support of intelligent coordinated regulation and decision, so that a new path is opened up for safe and efficient utilization of new energy power.
Because the utilization of wind and light energy in the power grid system is mature, related application scenes are also researched, and the system can realize expected effects and optimization effects in the virtual power plant which fully utilizes wind and light to generate electricity. However, in some environments with more complex using modes of new energy, for example, after new energy such as source side biogas cogeneration distributed energy is adopted, new parameters are introduced, and damage is caused to an existing wind-light stabilizing system, so that a model designed in the prior art is difficult to match with an actual environment, and reasonable optimization cannot be made.
Disclosure of Invention
In order to solve at least one of the above technical problems, further accurately optimize a virtual power plant in a complex environment, the present disclosure provides a two-stage robust optimization method of the virtual power plant based on saddle uncertainty, acting on a source network load storage linkage environment, the method includes:
s1: acquiring a conventional model of the virtual power plant, establishing a theory model on the basis of the conventional model, and finding exogenous uncertainty and endogenous uncertainty based on the theory model;
s2: correcting the endogenous uncertainty reference distribution by using local discrete super , introducing relative entropy and a box-type polyhedral uncertainty set to optimize the theory model, and generating an endogenous-exogenous uncertainty model of the virtual power plant;
s3: describing a source side biogas cogeneration distributed energy source by using an in-out uncertainty model of a virtual power plant, wherein load can be transferred on a load side, and the energy conversion relation of electric-heat-cold multifunctional coupling equipment between load equipment and electric heat energy storage on a storage side can be reduced, so that a flexible load demand response adjustment link is represented;
s4: taking the economical efficiency and the carbon emission reduction target of the virtual power plant as convergence, and establishing a two-stage robust optimization model of the virtual power plant based on theory uncertainty risk;
s5: and solving a two-stage robust optimization model of the virtual power plant by using a column and constraint generation algorithm and combining a strong dual theorem and a linearization theory.
And the economical efficiency and the low carbon coordinated operation of the cold-hot electricity virtual power plant are realized.
Preferably, the theory discrete super and discrete sub specifications are introduced in the step S1 when creating the saddle model. in the random process, super (super martingal) is defined as: is provided with
Figure SMS_3
Is->
Figure SMS_12
，/>
Figure SMS_13
For two random sequences, for arbitrary +.>
Figure SMS_4
There is->
Figure SMS_6
，/>
Figure SMS_7
Is->
Figure SMS_8
And>
Figure SMS_1
then call->
Figure SMS_9
About->
Figure SMS_10
Super , abbreviated as->
Figure SMS_11
Super , or upper ; if->
Figure SMS_2
Then->
Figure SMS_5
Is sub (sub martingal)Or lower .
The relation between wind power and energy storage decisions and electricity price distribution in the planning problem is complex, and explicit expression of probability models of the wind power and energy storage decisions and electricity price distribution cannot be established under the condition of non-priori. Inspired by the concepts of super and sub , the concepts of local discrete super and local discrete sub are proposed for the endogenous uncertainty, and local discrete super/sub further resolve the endogenous and exogenous uncertainties in the power system planning problem from the aspect of stochastic processes, and a mathematical expression form of the endogenous-exogenous uncertainties is given.
Preferably, the endogenous uncertainty is defined by theory, and the local discrete super is used for describing the expected electricity price reduction process caused by near zero marginal cost under the current capacity decision of source side equipment of the virtual power plant, namely the electricity price is the endogenous uncertainty of the virtual power plant. Modeling of the electricity price, an endogenous uncertainty, is converted into a probability distribution correction reference distribution using a local discrete super , the uncertainty being described by means of relative entropy.
Preferably, the exogenous uncertainty comprises source side wind power equipment and methane cogeneration equipment; modeling the uncertainty of wind power through relative entropy; the biogas equipment adopts box type uncertainty set description, and random uncertainty of the cold-hot electric load is described by using polyhedron uncertainty set. Because the uncertainty of the traditional box type uncertainty set is uncontrollable, the result is over-conservative, and great economy is sacrificed, and the polygon uncertainty set is added with an adjustable uncertainty coefficient based on the box type uncertainty set, so that the conservation can be controlled by changing the uncertainty coefficient. The prior art is therefore generally described using a set of polyhedral uncertainties. The random uncertainty including the cold-hot electric load described in the invention is also described by using a polyhedron uncertainty set. The applicant has found that biogas plants are not suitable for using a polyhedral uncertainty set description. Because the biogas yield prediction scale is generally larger than the robust optimization scheduling strategy time unit, if the description is carried out by adopting a polyhedral uncertainty set, a larger error can appear after calculation, and the error can not be eliminated through the optimization of a later model, and the problem can be solved by modeling the biogas yield uncertainty by selecting a box type uncertainty set.
Preferably, in the step S3, the process adopts constant temperature fermentation, the biomass raw material amount produced every day is approximately regarded as a constant value, and the influence of the concentration of biomass waste liquid is ignored, so that the theoretical yield of the daily biogas is obtained.
Preferably, the flexible load demand response adjusting link in the step S3 comprises flexible cold-hot electricity adjustable loads, and the flexible cold-hot electricity adjustable loads can be divided into transferable loads and load reduction according to the adjusting mode. Wherein TL is capable of time shifting within the allowable transition time range; RL has higher flexibility and can relieve the system energy pressure by reducing its own power or interrupting operation in a time range that allows curtailment;
the energy storage equipment ensures that the energy storage charge state is consistent at the end of a planning period and does not exceed the capacity limit value of the equipment;
the cold and hot electric equipment in the virtual power plant takes biomass methane and electric energy as primary energy sources and writes a cold and hot electric power balance equation matrix form.
Preferably, in the step S4, in the establishing a two-stage robust optimization model of the virtual power plant based on theory uncertainty risk, the overall optimization objective of the virtual power plant system considers economy and low carbon property, the overall operation cost optimization objective of the virtual power plant is operation and maintenance cost, electricity purchasing cost, start-stop cost and load adjustment cost, and the overall carbon bank cost optimization objective of the virtual power plant is combined to form an objective function; and matrixing the optimization target and the constraint condition.
Preferably, in the step S5, a column and constraint generation algorithm is used for solving, the column and constraint generation algorithm divides a two-stage robust optimization model corresponding to a formula into a main problem and a sub-problem, the decomposed main problem and sub-problem are simplified in form, an uncertain variable set is substituted into the robust equation, after linearization by a large M method, the sub-problem is converted into a mixed integer linear optimization model, a group of initial worst source load data is given at first, and an initial cost upper limit, a initial cost lower limit, a maximum iteration number and a maximum allowable error are defined; substituting the worst source load data into the main problem type to solve, updating the lower bound LC of the function, and outputting the optimized switch variable value; substituting the optimized value of the switch variable into the sub-problem to solve, and updating the worst condition source load data and the function upper bound UC; and calculating a two-stage result error UC-LC, judging whether the error is smaller than an allowable value, if the error is in accordance with the allowable value, exiting the operation, if the error is not in accordance with the allowable value, adding a variable, updating a constraint condition of a main problem and the current iteration number, and carrying out loop solution until the condition is met or the maximum iteration number is reached.
The invention has the beneficial effects that: the invention discloses a two-stage robust optimization method of a virtual power plant based on saddle uncertainty, which is based on a virtual power plant in a source network charge storage linkage scene and on the theory, considers the network side online electricity price endophytic uncertainty and source side wind power output, methane yield, charge side cold-heat-electricity load power and other exogenous uncertainties, fully considers typical equipment such as wind power generation, methane cogeneration units, electric heating, electric refrigeration, transferable loads, load-reducible, electric heating-cold energy storage and the like, establishes a mathematical model thereof, and describes energy conversion processes of different equipment and flexible load demand response adjustment links. According to the method, the economical efficiency and the carbon emission reduction target of the virtual power plant are comprehensively considered, and a two-stage robust optimization model of the virtual power plant is established in theory of uncertainty risk: in the first stage, fine tuning of the strategy is realized by adding some robustness constraints, so that the initial solution is ensured to have certain robustness, and certain optimization performance can be maintained even under the condition of uncertainty change; in the second stage, the objective of the optimization problem is to maximize the optimization performance while maintaining robustness, typically by an objective function with robustness constraints. According to the invention, a column and constraint generation algorithm is adopted, and a strong dual theorem and a linearization theory are combined to solve a two-stage robust optimization model, so that the economical efficiency and the low-carbon coordinated operation of the cold-hot electric virtual power plant are realized. Compared with the prior art, the method has the advantages that compared with a common robust optimization model, the two-stage robust optimization model is more flexible, and the optimization performance can be improved while the robustness is maintained.
Drawings
The accompanying drawings, which are included to provide a further understanding of the disclosure and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the disclosure and together with the description serve to explain the principles of the disclosure.
FIG. 1 is a schematic flow diagram of a two-stage robust optimization method for a virtual power plant based on saddle uncertainty.
Detailed Description
The present disclosure is described in further detail below with reference to the drawings and the embodiments. It is to be understood that the specific embodiments described herein are merely illustrative of the relevant content and not limiting of the present disclosure. It should be further noted that, for convenience of description, only a portion relevant to the present disclosure is shown in the drawings.
In addition, embodiments of the present disclosure and features of the embodiments may be combined with each other without conflict. The present disclosure will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.
The model established by the virtual power plant comprises a thermodynamic model, a chemical model, a control model and the like. These are widely used in the prior art. However, virtual power plants encounter a variety of complex scenarios during use. If the optimization is not carried out, various restrictions and hidden dangers exist in the use process, and the overall efficiency is greater than that of the optimization. There are therefore a number of optimisation methods for virtual power plants in the prior art. However, the initial model can only be optimized for simple scenes, for example, only wind, light and electricity under ideal conditions are accessed, and with the continuous increase of power requirements, energy diversification is a necessary trend. In the last decades, the proportion of renewable energy sources has been increasing because of the importance of environmental protection, as well as the advances in technology. However, unlike conventional energy sources, renewable energy sources have a tendency to "disperse" and also to provide "wave" energy. Because neither wind energy nor light energy is concentrated nor is it a sustainable source of energy. While "renewable energy" is "decentralized" and "fluctuating", this means that future power supplies will be more and more "decentralized" and will be more and more "fluctuating". Therefore, besides the great development of wind, light and water, the energy conversion capacity of the rural biogas cogeneration distributed energy sources and the electric-heat-cold multi-energy coupling equipment between the electric heating and energy storage at the storage side is greatly developed. The source network charge storage linkage environment is a novel power system which realizes information sharing and coordination control between source network charge storages and high-efficiency utilization and optimal configuration of energy sources in the power system through an intelligent technology and an informatization means.
In the source network charge storage linkage environment, the information of each link can be monitored and controlled in real time, so that the optimal configuration and the efficient utilization of energy are realized. Specifically, the source network load storage linkage environment includes the following aspects:
source side control: the energy is intelligently controlled in all links such as collection, transmission, storage and utilization, so that the optimal configuration and the efficient utilization of the energy are realized.
Network side control: the power network is optimized and scheduled by monitoring and controlling the running state and the load condition of the power network in real time.
Load side control: the balance and optimization of the power load are realized by monitoring and controlling the power consumption requirement of the power consumer in real time.
Storage side control: by monitoring and controlling the running state and the load condition of the power storage equipment in real time, the power storage is optimized and scheduled.
Through the coordination control of the links, the source network charge storage linkage environment can realize comprehensive monitoring and management of energy, so that the energy utilization efficiency is improved, the energy consumption and emission are reduced, and sustainable development is realized. The diversified energy sources enable the optimization mode of the virtual power plant in the prior art to not meet the requirements.
Most of the existing virtual power plant scheduling researches aim at exogenous uncertainties such as resources, environments, weather and the like, influence of the endogenous uncertainties on a scheduling scheme is ignored, and particularly influence of a decision process on the uncertainties is easily caused, so that decision results are over-conserved and lack of prospective, and further the problem of energy storage equipment investment redundancy is caused. The endogenous uncertainty concept originates from operational theory and is also expressed as decision-dependent in the context of stochastic planning. The method mainly comprises the steps that a decision result directly influences the future distribution of uncertainty, or the moment of decision action influences the probability distribution of uncertainty. However, the above concepts cannot give a strict mathematical definition, and it is difficult to distinguish the attribute of uncertainty without prior information, which hinders the next planning modeling. There is also a growing concern about in power system related research. In the power supply planning problem, the endogenetic uncertain factor of the electricity price cannot describe the distribution characteristics by using a scene set method, the corresponding electricity price fluctuation range lacks basis, and the provided method cannot guide the planning and scheduling problem of the virtual power plant.
In order to achieve the above object, the present invention provides the following technical solutions, in which EP is abbreviated as electricity price hereinafter, referring to electricity price:
a virtual power plant fuzzy opportunity constraint optimization scheduling method considering uncertainty of source load and user will comprises the following steps:
step one: acquiring a conventional model of the virtual power plant, establishing a theory model on the basis of the conventional model, and finding exogenous uncertainty and endogenous uncertainty based on the theory model;
wherein, based on the specifications of the discrete super and the discrete sub in the theory, the local discrete super and the local discrete sub are proposed to characterize the local probability dependence relationship of the uncertainty variable of the planning problem and the planning decision variable. The invention introduces saddle point theory, simply called saddle theory. The definition of a saddle point refers to a stagnation point (point with a first derivative of 0) that is not a local minimum, called a saddle point. The mathematical meaning is: the gradient (first derivative) value of the objective function at this point is 0, but one direction from the point of change is the maximum point of the function and the other direction is the minimum point of the function.
as a way to describe the random process, adapt to random sequences
Figure SMS_14
, for any->
Figure SMS_15
The method comprises the following steps:
Figure SMS_16
，
in the method, in the process of the invention,
Figure SMS_17
indicating the expected value of the variable in brackets, ">
Figure SMS_18
For two random process variables, < >>
Figure SMS_19
Is->
Figure SMS_20
Is a function of (2).
Assuming that a certain uncertainty in the planning problem is defined as a discrete stochastic process
Figure SMS_21
The corresponding uncertainty variable is +.>
Figure SMS_22
Decision variables in the optimization problem are defined as random discrete processes +.>
Figure SMS_23
Decision variable is +.>
Figure SMS_24
. Random process
Figure SMS_25
And->
Figure SMS_26
Can construct procedure->
Figure SMS_27
：
Figure SMS_28
，
Wherein:
Figure SMS_35
for the time increment in probability space, +.>
Figure SMS_32
Representation->
Figure SMS_37
Time of day variable->
Figure SMS_34
Is expected to be and
Figure SMS_39
everywhere the same->
Figure SMS_36
Representation->
Figure SMS_49
At->
Figure SMS_42
Go up to generate->
Figure SMS_48
Sub-algebra. If the uncertainty variable and the decision variable satisfy the above formula, the idea according to the theory can be considered as +.>
Figure SMS_29
Decision variable +.>
Figure SMS_46
Can be constructed as->
Figure SMS_40
of (a) i.e. "". The decision variable can be "" in relation to the uncertainty variable, representing that the uncertainty variable is exogenous uncertainty, formula (2) can be regarded as +.>
Figure SMS_44
Moment->
Figure SMS_41
Decision variable before time ∈ ->
Figure SMS_43
Uncertainty variable +.>
Figure SMS_33
For decision variables->
Figure SMS_38
Conditional desire and decision variable->
Figure SMS_31
Is irrelevant; further, formula (2) may be considered as desirable with +.>
Figure SMS_47
Decision variable +.>
Figure SMS_30
Is also irrelevant, and the condition expectations are still +.>
Figure SMS_45
。
The relation between wind power and energy storage decisions and electricity price distribution in the planning problem is complex, and explicit expression of probability models of the wind power and energy storage decisions and electricity price distribution cannot be established under the condition of non-priori. Inspired by the concepts of super and sub , the concepts of local discrete super and local discrete sub are proposed for endogenous uncertainty:
Figure SMS_50
，
wherein:
Figure SMS_53
is an uncertain variable->
Figure SMS_56
Corresponding random procedure, < > on >>
Figure SMS_58
Is->
Figure SMS_54
Corresponding random procedure, < > on >>
Figure SMS_55
Is->
Figure SMS_57
Is as much as desired->
Figure SMS_59
Is->
Figure SMS_51
Corresponding random procedure, < > on >>
Figure SMS_62
Is->
Figure SMS_63
Is not limited to the above-described embodiments. />
Figure SMS_64
An upper limit of a running period is planned; formulae (3) and (4) correspond to local discrete super and local discrete sub , which can express +.>
Figure SMS_52
And->
Figure SMS_60
I.e. the current decision quantity has a dependency on the uncertainty variable in the current and future, or the decision result makes the uncertainty variable in the current or future hope ∈>
Figure SMS_61
Will locally decrement or increment. the discriminant describes the dependency of decision timing on this uncertainty more closely than bayesian dependencies. and local discrete super/sub further resolve endogenous and exogenous uncertainties in the power system planning problem from the aspect of stochastic processes, giving a mathematical representation of the endogenous-exogenous uncertainties. Both endogenous and exogenous uncertainties are very important factors in saddles. Saddle points are extreme points of a function, and both endogenous and exogenous uncertainties can be affectedExtremum position and nature of the function. In the optimization problem of the invention, the influence of the endogenous uncertainty and the exogenous uncertainty on the optimization result needs to be considered to determine whether the optimal solution is stable and reliable.
Step two: correcting the endogenous uncertainty reference distribution by using local discrete super , introducing relative entropy and a box-type polyhedral uncertainty set to optimize the theory model, and generating an endogenous-exogenous uncertainty model of the virtual power plant; according to theory model, taking into account the net side net electricity price in-situ uncertainty, source side wind power output, methane yield, load side cold-heat-electricity load power and other in-situ uncertainty, correcting the in-situ uncertainty reference distribution by using local discrete super , and integrating relative entropy, box type and polyhedron uncertainty to describe in-situ uncertainty model. The Relative entropy (Relative entropy) is an important concept in information theory, also called KL divergence (Kullback-Leibler divergence). It measures the "difference" or "distance" between two probability distributions.
In the aspect of endophytic uncertainty, defined by theory, the current source side equipment capacity decision of the virtual power plant can be described by using local discrete super , and the expected electricity price reduction process caused by near zero marginal cost is that the electricity price is the endophytic uncertainty of the virtual power plant. Modeling of this endogenous uncertainty in electricity prices is converted into a probability distribution correction reference distribution using a local discrete super , the uncertainty being described by means of relative entropy:
under ideal conditions, it is assumed that historical electricity price data
Figure SMS_65
Satisfy normal distribution->
Figure SMS_66
Historical price->
Figure SMS_67
Is represented by the discrete process of (a):
Figure SMS_68
，
wherein:
Figure SMS_69
for an empty set of decision sets, it is indicated that there is no decision space before planning has not started. The above equation indicates that the electricity price expectation will not change under passive side device capacity decision conditions. In the prior test, the wind power permeability is increased to lower the electricity price, and the electricity price +_in the beginning of planning can be constructed according to the prior information>
Figure SMS_70
Is greater than %>
Figure SMS_71
：
Figure SMS_72
，
For initial electricity price
Figure SMS_73
The distribution is corrected to obtain a corrected reference distribution->
Figure SMS_74
. According to the big theorem, the local discrete super is expected to converge to the mean value, and the reference distribution correction process can be constructed as follows:
Figure SMS_75
，
wherein:
Figure SMS_76
is the number of samples; />
Figure SMS_77
Correction factors for mean and variance; />
Figure SMS_78
And estimating an increment factor for the wind power permeability. After correction, the reference distribution is ensured to approach the electricity price distribution of the actual source side equipment after operation, and the real reference value of the reference distribution is exerted.
Describing a set of electricity price uncertainty using relative entropy:
Figure SMS_79
，
wherein:
Figure SMS_80
is an uncertainty variable of wind power error; />
Figure SMS_81
Is the sample space of wind power error; />
Figure SMS_82
And
Figure SMS_83
the actual probability density function of the electricity price and the corrected probability density function of the reference electricity price. Whereas the actual wind power distribution +.>
Figure SMS_84
When not measurable, a set of uncertainty in electricity prices for different years can be constructed based on relative entropy:
Figure SMS_85
，
wherein:
Figure SMS_86
is->
Figure SMS_87
A collection of annual electricity price uncertainties; />
Figure SMS_88
For planning period->
Figure SMS_89
Actual distribution of annual electricity prices;
Figure SMS_90
to program a cycle.
A robust opportunity constraint is established for the endogenous uncertainty of electricity price:
Figure SMS_91
/>
Figure SMS_92
，
wherein:
Figure SMS_95
is a price risk factor; />
Figure SMS_102
Is +.>
Figure SMS_103
The individual nodes are->
Figure SMS_94
The electricity price at the moment; />
Figure SMS_98
Figure SMS_99
For electricity price probability, < >>
Figure SMS_101
The probability quantiles; the above-mentioned reference electricity price distribution set is +.>
Figure SMS_93
In the worst case, the future electricity price +.>
Figure SMS_96
Higher than a certain limit value of history synchronization>
Figure SMS_97
The probability of (2) is still greater than the probability threshold
Figure SMS_100
。
Exogenous uncertainty aspect, source side wind power and methane cogeneration equipmentThe method can not influence the current and future output fluctuation, does not change the expectation of future wind power distribution, belongs to the uncertainty of exogenous generation of a virtual power plant, and models the uncertainty of wind power by means of relative entropy. Relative entropy of wind power
Figure SMS_104
Expressed as:
Figure SMS_105
，
wherein:
Figure SMS_106
is an uncertainty variable of wind power error; />
Figure SMS_107
Is the sample space of wind power error; />
Figure SMS_108
And
Figure SMS_109
is an actual probability density function of wind power error and a reference probability density function of wind power error. The actual wind power distribution is not measurable at present, so as to ensure the actual wind power distribution +.>
Figure SMS_110
And generating a reference distribution->
Figure SMS_111
Is based on the relative entropy to construct a wind power uncertainty set +.>
Figure SMS_112
The following are provided:
Figure SMS_113
，
wherein:
Figure SMS_114
for the distance threshold between the reference distribution and the actual distribution, the chi-square distribution method is adopted to obtain the sampling typical scene number +.>
Figure SMS_115
Then selecting corresponding divergence value:
Figure SMS_116
，
chi-square distribution is a statistical distribution. It describes the statistical relationship between two random variables reflecting the degree of independence of the two random variables. When the random variables are completely independent, the chi-square distribution relation is satisfied between the random variables. The density function of the chi-square distribution contains a degree of freedom parameter which reflects the degree of correlation between random variables, with a larger value indicating a smaller correlation.
Wherein:
Figure SMS_117
the upper quantile is distributed for the chi-square of the degree of freedom of N-1, so that the wind power can be ensured to be not less than +.>
Figure SMS_118
The probability of (2) being included in the uncertainty set +.>
Figure SMS_119
Is a kind of medium.
Describing the exogenous uncertainty of wind power as a robust opportunity constraint:
Figure SMS_120
/>
Figure SMS_121
，
wherein:
Figure SMS_124
for the wind power actual output power, < >>
Figure SMS_125
And->
Figure SMS_127
Nodes +.>
Figure SMS_123
The actual wind curtailment rate and wind curtailment rate limit of the wind farm; />
Figure SMS_126
Figure SMS_128
For the reference wind power distribution set obtained in the foregoing>
Figure SMS_129
In the mean that even in the worst distribution case, the probability that the wind abandoning rate is smaller than a given value is still larger than the probability threshold +.>
Figure SMS_122
。
Other equipment aspects of exogenous uncertainty are that the biogas equipment adopts a box type uncertainty set, and random uncertainty of the cold-hot electric load is described by using a polyhedron uncertainty set:
Figure SMS_130
，
wherein:
Figure SMS_131
daily biogas production and daily biogas production are respectively treated by->
Figure SMS_132
Predicted values of electric, thermal and cold load power at the moment; />
Figure SMS_133
Daily biogas production and daily biogas production are respectively treated by->
Figure SMS_134
Real values of electric, thermal and cold load power at moment; />
Figure SMS_135
Dividing intoRespectively represent methane and load->
Figure SMS_136
Is determined by the maximum deviation rate of (2); />
Figure SMS_137
And the variable is 0-1, and the uncertainty coefficient of the daily output of the methane and the different moments of the cold and hot electric loads is represented.
Step three: and describing a source side biogas cogeneration distributed energy source by using an in-out uncertainty model of the virtual power plant, wherein load can be transferred on a load side, and the energy conversion relation of an electric-heat-cold multifunctional coupling device between load equipment and electric heat energy storage on a storage side can be reduced, so that a flexible load demand response adjustment link is represented.
In the production link of the biogas, the technological process adopts constant-temperature fermentation, the biomass raw material quantity produced every day is approximately regarded as a constant value, and the influence of the concentration of biomass waste liquid is ignored, so that the theoretical yield of the biogas every day is obtained:
Figure SMS_138
，
wherein:
Figure SMS_139
is the daily output of methane; />
Figure SMS_140
The gas yield of biomass raw materials; />
Figure SMS_141
Daily yield of biomass feedstock. The biogas production process is accompanied with the generation of accessory products such as organic fertilizers and the like, so that the production cost can be further reduced. The organic fertilizer production is deduced from the law of conservation of mass as follows:
Figure SMS_142
，
wherein:
Figure SMS_143
is to haveDaily output of machine fertilizer; />
Figure SMS_144
Biomass feedstock density; />
Figure SMS_145
Is the density of methane; />
Figure SMS_146
The content of the organic fertilizer in the residue of the biogas digester.
The biomass energy utilization process also needs to meet the biogas yield constraint:
Figure SMS_147
，
the flexible load demand response adjusting link, the flexible cold and hot electricity adjustable load can be divided into a transferable load (transferable load, TL) and a Reducible Load (RL) according to the adjusting mode. Wherein TL is capable of time shifting within the allowable transition time range; RL has greater flexibility and can relieve system power pressure by reducing its own power or interrupting operation in a time frame that allows curtailment.
In the transferable load, the allowable transfer time period of TL is set as
Figure SMS_148
The method is characterized in that the method meets the constraint conditions of a transfer power range, a minimum continuous operation duration, a transfer electric quantity conservation and the like in the adjustment process:
Figure SMS_149
，
wherein: upper corner mark
Figure SMS_153
Respectively representing electric, thermal, cold load and power types; />
Figure SMS_159
For regulating the later->
Figure SMS_163
Moment transferable load->
Figure SMS_152
Is a power of (2); />
Figure SMS_160
Is a variable 0-1, representing->
Figure SMS_162
Moment transferable load->
Figure SMS_164
The value 0 represents +.>
Figure SMS_150
The time does not participate in regulation, otherwise, the representatives participate in regulation; />
Figure SMS_156
Respectively is the transferable load after regulation->
Figure SMS_158
Minimum and maximum power of (2); />
Figure SMS_161
For transferring load->
Figure SMS_151
Is a minimum transition length of (2); />
Figure SMS_154
To adjust the front->
Figure SMS_155
Moment transferable load->
Figure SMS_157
Is set, is provided.
In the load reduction, the allowable transition period of the RL is set to be
Figure SMS_165
Which should be during the adjustment processMeeting the constraint conditions of reducing power range, continuously reducing transfer range and the like:
Figure SMS_166
，
Figure SMS_169
for regulating the later->
Figure SMS_170
Load can be reduced at the moment->
Figure SMS_173
Is a power of (2); />
Figure SMS_168
Is a variable 0-1, representing->
Figure SMS_172
Moment transferable load->
Figure SMS_174
The value 0 represents +.>
Figure SMS_175
The time participates in regulation; />
Figure SMS_167
Respectively, can reduce load after adjusting>
Figure SMS_171
Minimum and maximum power of (2); />
Figure SMS_176
To cut load->
Figure SMS_177
Minimum, maximum cut-down duration of (c).
Figure SMS_178
Time-of-day electrical/thermal/cold load and flexible loadThe relation of (2) is:
Figure SMS_179
，
the energy storage equipment ensures that the energy storage charge state is consistent at the end of a planning period, and the capacity limit value of the equipment is not exceeded:
Figure SMS_180
，
wherein:
Figure SMS_183
the maximum and minimum values of the stored energy y are respectively; />
Figure SMS_186
Respectively->
Figure SMS_188
Time energy storage->
Figure SMS_182
Charging and discharging power of (a); />
Figure SMS_185
Is a variable 0-1, representing +.>
Figure SMS_189
Time energy storage->
Figure SMS_190
A value of 1 indicates a charge state and a value of 0 indicates a discharge state; />
Figure SMS_181
Respectively store energy->
Figure SMS_184
Maximum and minimum values of charge and discharge energy power; />
Figure SMS_187
Is the energy charging and discharging efficiency.
Cold and hot in virtual power plantsThe electric equipment takes biomass methane and electric energy as primary energy sources, and the matrix form of the cold-heat electric power balance equation is shown as a formula (26). Wherein:
Figure SMS_193
respectively->
Figure SMS_198
Electric, thermal, cold load power at moment;
Figure SMS_199
is->
Figure SMS_192
Biogas consumption at moment CHP; />
Figure SMS_195
The heat value of the methane is; />
Figure SMS_196
Respectively->
Figure SMS_197
The power grid interaction power, the wind power output power and the electric power absorbed by the electric heating and refrigerating equipment at moment; />
Figure SMS_191
The power generation and heating efficiencies of the methane cogeneration CHP are respectively; />
Figure SMS_200
Is the electric heating efficiency; />
Figure SMS_201
Is the electric refrigeration efficiency;
Figure SMS_202
respectively->
Figure SMS_194
And the charging and discharging power of electric, thermal and cold energy storage at any time.
For the sake of more clear expression of the meaning of the coupled energy supply equation, split description is made. Wherein the method comprises the steps of：L Representing a load demand vector;Irepresenting a conversion efficiency matrix;E representing a driving energy vector of the energy supply system;D represents the cold-hot electricity energy storage output power vector,
Figure SMS_203
Figure SMS_204
，
step four: combining a virtual power plant source-network-load-storage model, comprehensively considering the economy and carbon emission reduction targets of the virtual power plant, taking the economy and the carbon emission reduction targets of the virtual power plant as convergence, and establishing a virtual power plant two-stage robust optimization model based on theory uncertainty risk.
The overall operation cost optimization objective of the virtual power plant is as follows:
Figure SMS_205
，
wherein:
Figure SMS_211
the unit cycle operation cost of the ecological agriculture IES; />
Figure SMS_207
For operation and maintenance cost, electricity purchasing cost, start-stop cost and load adjusting cost; />
Figure SMS_224
For the unit->
Figure SMS_208
The unit operation and maintenance cost of each unit corresponds to the unit output power; />
Figure SMS_216
Is->
Figure SMS_221
Time unit->
Figure SMS_222
Is a force of the (a); />
Figure SMS_213
For storing energy->
Figure SMS_215
The unit operation cost of (3); />
Figure SMS_206
Is->
Figure SMS_214
The charging/discharging power of the stored energy y at the moment. Wherein, the machine set is->
Figure SMS_210
The method comprises wind power, methane cogeneration, electric heating and electric refrigeration; subscript of
Figure SMS_219
Respectively representing three energy storage types of electricity, heat and cold; />
Figure SMS_212
Is->
Figure SMS_217
The electricity purchase price is carried out at any time; />
Figure SMS_218
The single start-stop cost of the unit z; />
Figure SMS_226
Is a variable which is 0 to 1, representation->
Figure SMS_220
Time unit->
Figure SMS_223
Is in the working state of->
Figure SMS_209
Is->
Figure SMS_225
Time TL adjusts the compensation amount:
Figure SMS_227
，
wherein the method comprises the steps of
Figure SMS_228
For transferring load->
Figure SMS_229
The amount of subsidy per unit power; />
Figure SMS_230
Is->
Figure SMS_231
Time RL adjusts the compensation amount:
Figure SMS_232
，
wherein the method comprises the steps of
Figure SMS_233
To adjust the front->
Figure SMS_234
Load can be reduced at the moment->
Figure SMS_235
A power; />
Figure SMS_236
To cut load->
Figure SMS_237
To subsidize the amount per unit power.
The overall operation cost optimization objective of the virtual power plant is as follows:
Figure SMS_238
，
wherein:
Figure SMS_239
the daily treatment cost of the carbon emission of the system is realized; />
Figure SMS_240
Penalty amount per CO2 emission; />
Figure SMS_241
And the carbon emission factors are used for respectively supplying power to the marsh gas and the power grid. />
The overall optimization objective of the virtual power plant system considers economy and low carbon performance, and an objective function is as follows:
Figure SMS_242
，
matrixing the optimization target and the constraint condition, and writing a robust optimization model:
Figure SMS_243
，
wherein:x、y0-1 variable and continuous variable column vectors, respectively; s isSRespectively as variablesxCorresponding coefficient vectors and matrixes, wherein s is a unit start-stop cost coefficient,SCalculating a matrix for the start-stop state;mas a variableyCorresponding coefficient vectors, meaning power-related cost coefficients;A、B、C、D、F、G is a coefficient matrix;a、b、c、d、uas a vector of coefficients,uis a source load predictor coefficient. Constraint 1 in formula (32) corresponds to formulas (20) and (23); constraint 2 corresponds to condition 2 of formula (26), formula (21), formula (18) and formula (25); constraint 3 corresponds to condition 1 of formula (10), formula (14), formula (24), and formula (25); constraint 4 corresponds to conditions 3 and 4 of formula (15), formula (19), formula (22) and formula (25); constraint 5 corresponds.
Step five: and solving a two-stage robust optimization model by using a column and constraint generation algorithm and combining a strong dual theorem and a linearization theory, so as to realize the economical efficiency and low-carbon coordinated operation of the cold-hot electric virtual power plant. The column and constraint generation algorithm (Column and Constraint Generation Algorithm, CCG) is an algorithm that builds a linear programming model step by step. It approximates the optimal solution by incrementally generating new columns (equivalent to variables) and rows (equivalent to constraints) rather than initially specifying the complete model. This makes it more efficient in processing highly sparse but structured models. CCG algorithms typically start with a relatively small-scale model and solve their linear programming relaxation problem to get a viable solution. It is then checked whether the feasible solution satisfies the optimality condition. If so, the solution is the optimal solution, the algorithm ends, otherwise, new variables or constraints need to be added incrementally. The most "valuable" new variables or constraints are generated and added to the model. The basis for the judgment of the "value" is the extent to which it can improve the model objective function. And solving the relaxation problem of the new linear programming model again to obtain a new feasible solution. Repeating the steps until the optimal solution is obtained. Compared with directly establishing a complete model with larger scale, the CCG algorithm has the main advantages that:
1. avoiding the processing of large-scale models at first and being more efficient. With the progress of the solving process, the model scale gradually increases, but only a small amount of change is made on the basis of the current model each time, so that the solution is relatively easy.
2. It is more intelligent to automatically decide when new variables or constraints are needed and which ones need to be added. This can efficiently approach the optimal solution.
3. Since only a small number of variables or constraints are added at a time, the planning model and underlying data structure may be updated incrementally as implemented. This also improves algorithm efficiency. The method provides a more efficient method for solving the ultra-large scale linear programming model. This is also why the present invention chooses to employ this algorithm.
And (3) dividing the two-stage robust optimization model corresponding to the formula (32) into a main problem and a sub-problem by adopting a CCG algorithm to solve the main problem and the sub-problem. The reduced form of the decomposed main problem is:
Figure SMS_244
Figure SMS_245
，
sub-problem reduction form:
Figure SMS_246
，
wherein:
Figure SMS_247
the function value of the corresponding sub-problem is the other cost except the start-stop cost of the virtual power plant; SX means
Figure SMS_248
The process vector formed by the difference value between the start-stop state of the time unit and the start-stop state at the previous time has three conditions of 0, 1 and 1;
Figure SMS_249
the values of the optimization variable and the uncertain variable in the 2 nd stage after the iteration are respectively taken; />
Figure SMS_250
Is the maximum allowed number of iterations.
In a sub-problem of this,μ、ν、ω、πoptimizing variables for sub-problemsyThe dual variables of the relevant constraints are,
Figure SMS_251
the source load error vector accords with the definition of the worst condition when the new energy output of the virtual power plant is less and the load is greater, and the source load errors are respectively
Figure SMS_252
；/>
Figure SMS_253
A variable of 0-1, which represents the uncertainty coefficient of the source load at different moments; />
Figure SMS_254
For the introduced continuous variable, for equivalent uncertainty coefficient +.>
Figure SMS_255
With dual variables->
Figure SMS_256
Is a product of (2); m is a sufficiently large positive number; />
Figure SMS_257
The uncertainty of the photovoltaic and the cold-hot electrical load periods are respectively represented.
In the first stage main problem, the variable is a set start-stop 0-1 variable x, and other constraints are iteratively input by the sub-problem except for the constraint related to x in the 1 st line in the formula (32). In the second-stage sub-problem, the variables are all continuous variables y, and the 2 nd, 3 rd, 4 th and 5 th constraint except the start and stop of the unit in the constraint formula (32).
And substituting the uncertain variable set u into a robust equation, linearizing by a large M method, converting the sub-problem into a mixed integer linear optimization model, and solving by using a CCG algorithm. The large M method is a numerical method for solving a linear program. The nonlinear constraint condition of the linear programming problem is linearized by introducing a large M parameter into the constraint condition, so that the linear programming method such as SIMPLEX is used for solving. The solving process is as follows:
firstly, a group of initial worst source load data is given, and an initial cost upper limit, a initial cost lower limit, a maximum iteration number and a maximum allowable error are defined; substituting the worst source load data into the main problem type to solve, updating the lower bound LC of the function, and outputting the optimized switch variable value; substituting the optimized value of the switch variable into the sub-problem to solve, and updating the worst condition source load data and the function upper bound UC; calculating a result error UC-LC of two stages, and judging whether the error is smaller than an allowable value
Figure SMS_258
If yes, the operation is stopped, if not, the variable is added +.>
Figure SMS_259
And updating the constraint condition of the main problem and the current iteration number to carry out loop solving until the condition is met or the maximum iteration number is reached. By such a wayThe two-stage robust optimization model is more flexible, and can improve the optimization performance while maintaining the robustness.
In the description of the present specification, reference to the terms "one embodiment/manner," "some embodiments/manner," "example," "specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment/manner or example is included in at least one embodiment/manner or example of the present application. In this specification, the schematic representations of the above terms are not necessarily for the same embodiment/manner or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments/modes or examples. Furthermore, the various embodiments/modes or examples described in this specification and the features of the various embodiments/modes or examples can be combined and combined by persons skilled in the art without contradiction.
Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In the description of the present application, the meaning of "plurality" is at least two, such as two, three, etc., unless explicitly defined otherwise.
It will be appreciated by those skilled in the art that the above-described embodiments are merely for clarity of illustration of the disclosure, and are not intended to limit the scope of the disclosure. Other variations or modifications will be apparent to persons skilled in the art from the foregoing disclosure, and such variations or modifications are intended to be within the scope of the present disclosure.

Claims (8)

1. The two-stage robust optimization method of the virtual power plant based on saddle uncertainty is characterized by acting on a source network load storage linkage environment, and comprises the following steps:
s1: acquiring a conventional model of the virtual power plant, establishing a theory model on the basis of the conventional model, and finding exogenous uncertainty and endogenous uncertainty based on the theory model;
s2: correcting the endogenous uncertainty reference distribution by using local discrete super , introducing relative entropy and a box-type polyhedral uncertainty set to optimize the theory model, and generating an endogenous-exogenous uncertainty model of the virtual power plant;
s3: describing a source side biogas cogeneration distributed energy source by using an in-out uncertainty model of a virtual power plant, wherein load can be transferred on a load side, and the energy conversion relation of electric-heat-cold multifunctional coupling equipment between load equipment and electric heat energy storage on a storage side can be reduced, so that a flexible load demand response adjustment link is represented;
s4: taking the economical efficiency and the carbon emission reduction target of the virtual power plant as convergence, and establishing a two-stage robust optimization model of the virtual power plant based on theory uncertainty risk;
s5: and solving a two-stage robust optimization model of the virtual power plant by using a column and constraint generation algorithm and combining a strong dual theorem and a linearization theory.
2. The saddle uncertainty-based two-stage robust optimization method for a virtual power plant according to claim 1, wherein the step S1 introduces -based discrete super and discrete sub specifications when creating the saddle model.
3. The saddle uncertainty-based two-stage robust optimization method for a virtual power plant according to claim 1, wherein the endogenous uncertainty is defined by theory, and the expected decrease of electricity price due to near zero marginal cost, namely the electricity price is the endogenous uncertainty of the virtual power plant under the current capacity decision of source side equipment of the virtual power plant is described by using local discrete super .
4. The saddle uncertainty-based two-stage robust optimization method for a virtual power plant of claim 1, wherein the exogenous uncertainty comprises a source side wind power plant and a biogas cogeneration plant; modeling the uncertainty of wind power through relative entropy; the biogas equipment adopts box type uncertainty set description, and random uncertainty of the cold-hot electric load is described by using polyhedron uncertainty set.
5. The saddle uncertainty-based two-stage robust optimization method for a virtual power plant according to claim 1 or 4, wherein in the biogas production step in step S3, constant-temperature fermentation is adopted in the process, the biomass raw material amount produced every day is used as a constant value, and the influence of the concentration of biomass waste liquid is ignored, so that the theoretical yield of daily biogas is obtained.
6. The saddle uncertainty-based two-stage robust optimization method for a virtual power plant according to claim 1, wherein the flexible load demand response adjustment link in step S3 includes flexible cold-hot electricity adjustable loads, and is divided into transferable loads and reducible loads according to adjustment modes thereof.
7. The saddle uncertainty-based two-stage robust optimization method for a virtual power plant according to claim 1, wherein in the two-stage robust optimization model for a virtual power plant based on uncertainty risk established in step S4, the overall optimization objective of a virtual power plant system considers economy and low carbon, the overall operation cost optimization objective of the virtual power plant is operation and maintenance cost, electricity purchasing cost, start-stop cost and load adjustment cost, and the overall carbon bank cost optimization objective of the virtual power plant is combined to form an objective function; and matrixing the optimization target and the constraint condition.
8. The two-stage robust optimization method of the virtual power plant based on saddle uncertainty as claimed in claim 1, wherein in the step S5, a column and constraint generation algorithm is used for solving, the column and constraint generation algorithm divides a two-stage robust optimization model corresponding to a formula into a main problem and a sub problem, the decomposed main problem and the sub problem are simplified in form, an uncertain variable set is substituted into a robust equation, after linearization by a large M method, the sub problem is converted into a mixed integer linear optimization model, a group of initial worst source load data is given first, and an initial cost upper limit and a lower limit, a maximum iteration number and a maximum allowable error are defined; substituting the worst source load data into the main problem type to solve, updating the lower bound LC of the function, and outputting the optimized switch variable value; substituting the optimized value of the switch variable into the sub-problem to solve, and updating the worst condition source load data and the function upper bound UC; and calculating a two-stage result error UC-LC, judging whether the error is smaller than an allowable value, if the error is in accordance with the allowable value, exiting the operation, if the error is not in accordance with the allowable value, adding a variable, updating a constraint condition of a main problem and the current iteration number, and carrying out loop solution until the condition is met or the maximum iteration number is reached.
CN202310663465.4A 2023-06-06 2023-06-06 Two-stage robust optimization method for virtual power plant based on saddle uncertainty Active CN116415740B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310663465.4A CN116415740B (en) 2023-06-06 2023-06-06 Two-stage robust optimization method for virtual power plant based on saddle uncertainty

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310663465.4A CN116415740B (en) 2023-06-06 2023-06-06 Two-stage robust optimization method for virtual power plant based on saddle uncertainty

Publications (2)

Publication Number Publication Date
CN116415740A true CN116415740A (en) 2023-07-11
CN116415740B CN116415740B (en) 2023-09-12

Family

ID=87056337

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310663465.4A Active CN116415740B (en) 2023-06-06 2023-06-06 Two-stage robust optimization method for virtual power plant based on saddle uncertainty

Country Status (1)

Country Link
CN (1) CN116415740B (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20190023791A (en) * 2017-08-30 2019-03-08 인천대학교 산학협력단 System and Method of Simplified Robust Optimal Operation of Microgrids by Band of Wirtual Equivalent Load Variation Considering the Uncertainty of Renewable Generation and Loads
CN110390467A (en) * 2019-06-25 2019-10-29 河海大学 A kind of random ADAPTIVE ROBUST Optimization Scheduling of virtual plant distinguished based on key scenes
CN111817292A (en) * 2020-06-28 2020-10-23 国网青海省电力公司 Wind-solar energy storage robust configuration method and device for power system
CN112084629A (en) * 2020-08-11 2020-12-15 清华大学 Multi-energy virtual power plant polymerization method based on two-stage robust optimization
CN113507119A (en) * 2021-07-14 2021-10-15 华北电力大学 Power distribution system operation method for promoting renewable energy consumption based on electric-thermal game
CN114140022A (en) * 2021-12-10 2022-03-04 国网山西省电力公司电力科学研究院 Multi-virtual power plant distributed dynamic economic dispatching method and system
CN114139780A (en) * 2021-11-16 2022-03-04 国网山西省电力公司电力科学研究院 Coordinated optimization method and system for virtual power plant and power distribution network containing distributed power supply
CN115912507A (en) * 2022-12-08 2023-04-04 长沙理工大学 Green village and town power distribution network area cooperative autonomous method with biomass energy participating in peak regulation

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20190023791A (en) * 2017-08-30 2019-03-08 인천대학교 산학협력단 System and Method of Simplified Robust Optimal Operation of Microgrids by Band of Wirtual Equivalent Load Variation Considering the Uncertainty of Renewable Generation and Loads
CN110390467A (en) * 2019-06-25 2019-10-29 河海大学 A kind of random ADAPTIVE ROBUST Optimization Scheduling of virtual plant distinguished based on key scenes
CN111817292A (en) * 2020-06-28 2020-10-23 国网青海省电力公司 Wind-solar energy storage robust configuration method and device for power system
CN112084629A (en) * 2020-08-11 2020-12-15 清华大学 Multi-energy virtual power plant polymerization method based on two-stage robust optimization
CN113507119A (en) * 2021-07-14 2021-10-15 华北电力大学 Power distribution system operation method for promoting renewable energy consumption based on electric-thermal game
CN114139780A (en) * 2021-11-16 2022-03-04 国网山西省电力公司电力科学研究院 Coordinated optimization method and system for virtual power plant and power distribution network containing distributed power supply
CN114140022A (en) * 2021-12-10 2022-03-04 国网山西省电力公司电力科学研究院 Multi-virtual power plant distributed dynamic economic dispatching method and system
CN115912507A (en) * 2022-12-08 2023-04-04 长沙理工大学 Green village and town power distribution network area cooperative autonomous method with biomass energy participating in peak regulation

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
王骞等: "考虑内–外生双重不确定性的风储***联合规划方法", 《中国电机工程学报》, vol. 43, no. 1, pages 169 - 180 *

Also Published As

Publication number Publication date
CN116415740B (en) 2023-09-12

Similar Documents

Publication Publication Date Title
Li et al. A hybrid optimization-based scheduling strategy for combined cooling, heating, and power system with thermal energy storage
CN110555595B (en) Biogas-wind-light all-renewable energy system based on energy hub and method thereof
CN112465181A (en) Two-stage optimization scheduling method supporting source-network-load-storage multi-element ubiquitous coordination
Uchman et al. The analysis of dynamic operation of power-to-SNG system with hydrogen generator powered with renewable energy, hydrogen storage and methanation unit
CN111539572B (en) Optimal planning method for optical biogas micro energy network
He et al. The many-objective optimal design of renewable energy cogeneration system
Afzali et al. Urban community energy systems design under uncertainty for specified levels of carbon dioxide emissions
CN113988714B (en) Multi-uncertainty-based dynamic planning method, equipment and medium for park comprehensive energy system
CN111293682A (en) Multi-microgrid energy management method based on cooperative model predictive control
CN115860413A (en) Grid-connected combined cooling heating and power micro-grid system economic scheduling method based on load demand response and double-layer adjustable robust optimization
Li et al. Intraday multi-objective hierarchical coordinated operation of a multi-energy system
Han et al. Multi-stage distributionally robust optimization for hybrid energy storage in regional integrated energy system considering robustness and nonanticipativity
Ji et al. Optimal schedule of solid electric thermal storage considering consumer behavior characteristics in combined electricity and heat networks
CN117081143A (en) Method for promoting coordination and optimization operation of park comprehensive energy system for distributed photovoltaic on-site digestion
Yuan et al. A distributed multi-objective optimization method for scheduling of integrated electricity and hydrogen systems
Zhou et al. Optimal integration of renewable energy in refinery hydrogen management systems: Energy storage and direct utilization
CN111555362B (en) Optimal regulation and control method and device for full-renewable energy source thermoelectric storage coupling system
CN116415740B (en) Two-stage robust optimization method for virtual power plant based on saddle uncertainty
Yang Multi‐objective optimization of integrated gas–electricity energy system based on improved multi‐object cuckoo algorithm
Li et al. Stochastic-robust planning optimization method based on tracking-economy extreme scenario tradeoff for CCHP multi-energy system
Fu et al. Off-design characteristics of energy conversion equipment in integrated energy systems
CN111082442B (en) Energy storage capacity optimal configuration method based on improved FPA
CN112685879B (en) Multi-objective optimization method for regional electric heating interconnection energy system
CN115187129A (en) Micro-energy network optimal scheduling method and system considering heat supply network mass flow regulation
Xing et al. Multi-energy simulation and optimal scheduling strategy based on digital twin

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant