CN109636056B - Multi-energy microgrid decentralized optimization scheduling method based on multi-agent technology - Google Patents

Multi-energy microgrid decentralized optimization scheduling method based on multi-agent technology Download PDF

Info

Publication number
CN109636056B
CN109636056B CN201811579886.4A CN201811579886A CN109636056B CN 109636056 B CN109636056 B CN 109636056B CN 201811579886 A CN201811579886 A CN 201811579886A CN 109636056 B CN109636056 B CN 109636056B
Authority
CN
China
Prior art keywords
energy
power
agent
internet
strategy
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.)
Active
Application number
CN201811579886.4A
Other languages
Chinese (zh)
Other versions
CN109636056A (en
Inventor
张有兵
徐向志
王国烽
杨宇
卢俊杰
翁国庆
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.)
Zhejiang University of Technology ZJUT
Original Assignee
Zhejiang University of Technology ZJUT
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 Zhejiang University of Technology ZJUT filed Critical Zhejiang University of Technology ZJUT
Priority to CN201811579886.4A priority Critical patent/CN109636056B/en
Publication of CN109636056A publication Critical patent/CN109636056A/en
Application granted granted Critical
Publication of CN109636056B publication Critical patent/CN109636056B/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
    • 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

Landscapes

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

Abstract

A multi-energy micro-grid optimized operation method based on multi-agent considers the uncertainty of the operation of an energy Internet system, provides a regional energy Internet optimized operation control strategy, carries out discretization processing on continuous time, models multiple energy devices in the energy Internet, controls the minimized operation cost based on model prediction in each energy local area network, introduces a non-cooperative game, establishes a novel electricity price mechanism aiming at regional energy interaction, and achieves Nash balance through iterative computation. The optimized operation control strategy can effectively reduce the net load fluctuation rate of the system, reduce the economic cost of operation and improve the reliability and the economy of the energy Internet system.

Description

Multi-energy microgrid decentralized optimization scheduling method based on multi-agent technology
Technical Field
The invention belongs to the field of multi-energy microgrid energy operation optimization, and particularly relates to a decentralized optimization scheduling method for a multi-energy microgrid based on a multi-agent technology.
Background
The production, transportation, conversion of processing and use of fossil energy sources worldwide has caused serious pollution and destruction to the ecological environment, and poses serious threat to human survival. A third industrial revolution represented by a new energy technology and an internet technology is being started, and the construction of the Energy Internet (EI) can promote the industrial technology upgrading and the structure adjustment of the energy industry in China. The energy internet is a flat 'source-network-charge-storage' system and comprises a plurality of energy local area network units, independent power generation units, independent power utilization units, independent energy storage units and the like. Because the owners, task requirements and scheduling targets of the energy internet subunits are different, and the energy internet subunits have the ability and driving force for pursuing self benefit maximization, the diversification of load forms in the energy internet and the uncertainty of Renewable Energy (RES) are provided, the traditional centralized optimization scheduling mode is difficult to be applied to the optimization operation of the energy internet system, and how to deal with the problem which is urgently to be solved in the energy management and optimization operation of the energy internet.
Multi-agent systems (MAS) have great potential in distributed control and management, and currently attract much attention in the field of energy internet. The deep integration of the internet and the traditional energy can improve the networking proportion of renewable energy, realize the diversification of energy supply modes, promote the optimization of an energy structure, realize the flow of energy resources as required, promote the resource saving, efficiently utilize, reduce the total energy consumption and reduce the pollution emission. Along with the development of the energy internet, the equipment coupling relationship and the energy structure in the energy internet system are more complex, and on the other hand, the relationship between the load demand side and the energy internet is more flexible and diversified, so that the operation and management difficulty of the energy internet is greatly increased. From the viewpoint of optimizing operation, ensuring safe and efficient transmission of information flow in the energy production and transmission process is the key for controlling the production operation of the multi-energy system.
The method is characterized in that factors such as different controllable devices and diversified operation modes of the energy Internet are considered, the influence of individuals in the regional energy Internet on optimization strategies is researched, the flexibility of the energy conversion link in the network is emphasized, the overall description and the typical feature extraction of the multi-energy Internet system structure are considered, and the method is a hotspot and difficulty point aiming at energy Internet modeling at present. In an electric power trading market formed by an energy internet system, an energy local area network with high autonomy has stronger subjectivity and intelligence in behavior, so how to accurately model the game behavior of the energy local area network on the basis of considering stability and economy so as to realize the optimal benefit of an energy local area network individual and the coordination control of the whole system is a key direction for developing deep research aiming at the energy internet system in the future.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention provides a decentralized optimization scheduling method for a multi-energy microgrid based on a multi-agent technology, which is applied to energy interconnection and sharing among regional energy internets, wherein the regional energy internets comprise a primary energy source side, various energy conversion devices, energy storage devices, intelligent loads and the like. The primary energy source side (distributed energy, natural gas, coal and the like) passes through the energy conversion equipment, so that the cold, heat and electric load requirements of the user side are met, the consumption capacity of the multi-energy micro-grid on new energy can be effectively improved, the economy of the regional energy Internet is improved, and the influence of uncertainty on the operation of the multi-region energy Internet is weakened.
In order to achieve the purpose, the technical scheme of the invention is as follows:
a decentralized multi-energy microgrid optimization operation method based on multiple agents comprises the following steps:
s1: initializing a system, considering a discrete time model, setting the optimal time to be 24h, carrying out discretization treatment, equally dividing into T time intervals, wherein k belongs to {1, 2.., T } for any kth time, and the duration of the kth time interval is delta T;
s2: defining a multi-Agent (MAS) system, wherein the MAS system comprises a load management Agent, an energy measurement Agent and an in-network electricity price Agent, and the MAS system comprises the following steps:
(1) the load management Agent is an intelligent Agent aiming at the load flow direction and the use condition of a demand side, predicts the use of multi-energy loads of a user side, acquires the use condition of the loads in an area in time and makes a reasonable decision to distribute the load flow direction;
(2) energy measurement Agent: an intelligent Agent corresponding to a primary energy side distributed energy power generation and combined cooling heating and power supply system monitors various load demands of users in real time, a regional energy measurement Agent receives output prediction of distributed energy, monitors various micro-source output conditions, and adjusts power generation/heating/refrigerating states of an energy storage and energy supply device in real time to meet energy demands of regional energy internet user sides;
(3) the power price Agent in the network: the system is an intelligent agent aiming at an energy interaction electricity price mechanism between regional energy source internets; aiming at information interaction and energy interaction among different regions, local information transmitted by a local energy measuring Agent is held, and an electricity price mechanism among multi-region energy sources is established based on a non-cooperative dynamic game model;
s3: adopting backward subtraction to establish an RESs output and user basic load prediction model;
based on typical scenariosThe uncertainty of load and renewable energy power generation is simulated by a reduction representation method, various scenes are generated by adopting Monte Carlo simulation, and random scenes are generated by adopting a distributed sampling method to simulate the fluctuation condition of a predicted value in practice according to a wind speed predicted value and a solar radiation angle predicted value; determining probability distribution of prediction errors according to historical data to obtain random distribution errors, converting RES random variables into output power according to an output characteristic curve, and expressing the distributed output prediction errors by adopting normal probability distribution; the predicted value of the new energy output in the future T time period is represented by a time sequence, and an output scene is set
Figure GDA0002986449900000031
Is the output value of scene i at time T, and scene omegaiHas an occurrence probability of PiThe minimum probability distance between the scene set before reduction and the final retained scene subset is expressed as follows:
Figure GDA0002986449900000032
in the formula, α represents a set of scenes that are finally deleted after the scene reduction, and the number of scenes is 3000. Initialization reservation set S ═ ω0，ω1，ωi,ωsAdding another scene with the smallest probability distance to the actual abandon set, wherein the scene omega closest to the abandon set is changedlProbability is represented as p (ω'l)＝p(ωl)+p(θk) Until the number of scenes in the abandon set reaches the requirement;
applying a maximum power point tracking (MTTP) method to a new energy output system to enable the new energy output system to work at a maximum power point, wherein based on a prediction result, active output power and basic load prediction of RES are shown as follows:
Figure GDA0002986449900000033
Figure GDA0002986449900000034
Figure GDA0002986449900000035
the total output of the distributed power supply is as follows:
Pres,i＝Ppv,i+Pw,i (5)
s4: constructing a multi-agent system, and calculating the electricity price of the current k time period through an electricity price mechanism;
the energy storage system model is characterized as follows:
Figure GDA0002986449900000036
in the formula (I), the compound is shown in the specification,
Figure GDA0002986449900000037
for the initial state of charge of the energy storage system,
Figure GDA0002986449900000038
the charge capacity is expected for the energy storage system,
Figure GDA0002986449900000039
in order to be the battery capacity of the energy storage system,
Figure GDA00029864499000000313
for the charging power of the regional energy internet i energy storage system,
Figure GDA00029864499000000314
the discharging power of an energy storage system is the regional energy Internet i;
assuming that all energy storage systems have the same lithium ion battery pack and that the charge/discharge power over a single period is considered constant, therefore, the model and constraints for the energy storage system cells are established as follows:
Figure GDA00029864499000000310
in the formula (I), the compound is shown in the specification,
Figure GDA00029864499000000311
respectively representing the SOC states at time t +1 and t,
Figure GDA00029864499000000312
representing the energy storage battery power at time t, Mi,Bi,Ci and DiRespectively representing a system matrix, an input matrix, an output matrix and a feedforward matrix;
Figure GDA0002986449900000041
Figure GDA0002986449900000042
in the formula (I), the compound is shown in the specification,
Figure GDA0002986449900000043
respectively representing the charging and discharging power, eta, of the energy storage system at time tchAnd ηdchRespectively represent charge/discharge efficiencies;
the gas turbine generator set is used for an energy internet system, has high efficiency, fully utilizes natural gas energy, has small environmental pollution, and outputs to a gas turbine as follows:
Figure GDA0002986449900000044
in the formula
Figure GDA0002986449900000045
Generating power of an energy internet i gas turbine at a time t;
Figure GDA0002986449900000046
the maximum generated power of the gas turbine;
Figure GDA0002986449900000047
recovering power for waste heat of the energy Internet i gas turbine at the time t; etacAnd ηrThe power generation efficiency and the waste heat recovery efficiency of the gas turbine are obtained; lambda [ alpha ]gtAs gas consumption rate, λgasTaking 9.7kWh/m as the heat value of natural gas3；
The method comprises the following steps that electric power is used as a transaction core, and in an energy internet market, an intra-network electricity price Agent participating in bidding aims to obtain maximum benefit through a rational bidding strategy;
the optimization problem of the power price Agent in the power grid is expressed as follows:
Figure GDA0002986449900000048
in the formula, P is the optimization target of the power price Agent in the network in the rolling time domain, namely the power price in the network is the interaction power price rb，rsRespectively setting an internal electricity purchasing price and an internal electricity selling price in a day; a is1And a2The power balance reference electricity price respectively corresponds to the electricity price when net loads of the energy internet in the electricity selling region and the electricity purchasing energy internet are zero; in addition, in the formula
Figure GDA0002986449900000049
The expression of the four variables is as follows:
Figure GDA0002986449900000051
in the formula, PLoadAdjusting load, U, for energy Internet igridThe power is the interaction power of the energy Internet i;
s5: constructing a power balance model of the energy local area network to ensure the balance of supply and demand in the energy local area network;
constructing a power balance model of an energy local area network, wherein each component in the energy local area network at a power supply side and a demand side obtains an electric power balance model of the ith energy local area network:
Figure GDA0002986449900000052
wherein the content of the first and second substances,
Figure GDA0002986449900000053
the interaction power between the energy local area network i and the energy local area network in the network is obtained;
Figure GDA0002986449900000054
is transmission line power constraint;
s6: based on model prediction control, minimizing the operation cost of a single energy local area network, and repeating the steps S2-S5;
the total running cost of the single energy local area network is minimized, and the optimization problem of the single energy local area network in a rolling time domain is expressed as a quadratic programming problem:
Figure GDA0002986449900000055
in the formula: tau is the optimized rolling time domain length; u. ofgrid，Qs，QHX，QACEnergy interaction power, energy storage residual capacity, waste heat of a gas turbine for heating and waste heat of a gas turbine for refrigerating are respectively adopted; A. b, C, D are respectively the flexible constraint coefficients of the residual capacity of stored energy, the power generation power of the gas turbine, the waste heat of the gas turbine for heating and the waste heat of the gas turbine for cooling;
s7: the intra-network electricity price Agent receives scheduling information of the load management Agent and the energy measuring Agent, maximizes the income of each participant through a non-cooperative dynamic game model, determines the electricity price per se in real time to obtain a strategy set, and comprises the following processes:
s71: establishing a game model, establishing a mixed integer model according to an optimization problem of the energy Internet in a rolling time domain, and describing the game model:
participant set N+In (1)Each participant comprises a distributed energy source and an energy storage system;
strategy: for any i ∈ N+The number of the n-th bit lines, during the k period,
Figure GDA0002986449900000061
aggregate the actions for all participants; the strategy adopted is as follows: including distributed energy output, various demand loads, and other policies taken by participants. The optimization operation strategy maximizes each participant i (i belongs to N)+) Is expressed as piLet ρ beiIs a feasible strategy set;
and (4) yield: used to measure the total profit for each participant, maximize the profit for each participant i, denoted as Ui；
Given the policy set A as followsi＝{A1,A2,…AN}, if and only if:
Figure GDA0002986449900000062
wherein A is a set after updating the strategy set, and the strategy vector A is called Nash equilibrium point (NE), and any regional energy Internet can not improve the respective profit by changing the strategy in a single direction;
s72: policy PiFor the ith energy Internet interaction electricity price, the electricity price strategy set of the energy Internet i is AiAnd A isi＝{P|0≤P≤Pmax}，PmaxFor maximum reportable interactive electricity rates, therefore AiIs a compact convex set, and the participant must have the electricity selling strategy P in the game process, so the set AiIs not empty;
certificate Ui(A) If the function is a concave function, S has a pure strategy Nash equilibrium point; to Ui(A) Performing second derivation, wherein the second derivation is as follows:
Figure GDA0002986449900000063
due to rb,λb,PLoad,i(i-1, 2,3, …, N) are all non-negative, i.e. N is not negative
Figure GDA0002986449900000064
Thus U isiThe game is concave, and the non-cooperative game problem S has pure strategy Nash equilibrium points;
s8: after the optimization is completed, a strategy set is obtained, and whether the Nash equilibrium is achieved or not is calculated; and repeating the steps S3-S7, and when the optimal scheme is determined to be obtained, terminating the game to obtain the optimal operation mode of the multi-energy microgrid.
The invention has the beneficial effects that:
1. in the technical scheme of the invention, uncertainty of load and renewable energy power generation is simulated based on a typical scene reduction representation method, various scenes are generated by adopting Monte Carlo simulation, and fluctuation conditions of predicted values in practice are simulated by adopting a distributed sampling method, so that distributed energy prediction is more accurate.
2. The novel electricity price mechanism of the multi-energy micro-grid based on the multi-agent is provided, the electricity consumption behavior of the active load can be effectively guided through the mutual game among the agents, the effect of peak clipping and valley filling is achieved, and therefore the operation pressure of the power grid is reduced.
3. By analyzing the physical characteristics of a single energy local area network, modeling is carried out on various energy devices in the energy Internet, and the minimum operation cost is controlled in each energy local area network based on model prediction, so that the efficient and stable operation of the system is ensured, and the economic cost of the system is reduced.
4. And a non-cooperative game is introduced, and an internal control operation strategy of each energy local area network is formulated, so that the net load fluctuation rate of the system is effectively reduced, the operation economic cost is reduced, and the reliability and the economy of the energy internet system are improved.
Drawings
Fig. 1 is a schematic diagram of an energy internet system.
Fig. 2 is a graph of RES force.
Fig. 3 is a cold and hot load transfer curve.
Fig. 4 is a graph showing a change in electricity prices.
Fig. 5 is a net load graph.
Fig. 6 is a graph of the optimized operation of the energy local area network.
Fig. 7 is a flowchart of a decentralized optimization scheduling method for a multi-energy microgrid based on a multi-agent technology.
Detailed Description
The invention will be further explained with reference to the drawings.
Referring to fig. 1 to 7, a decentralized optimization scheduling method for a multi-energy microgrid based on a multi-agent technology includes the following steps:
s1: initializing a system, considering a discrete time model, setting the optimal time to be 24h, carrying out discretization treatment, equally dividing into T time intervals, wherein k belongs to {1, 2.., T } for any kth time, and the duration of the kth time interval is delta T;
s2: an energy internet MAS system structure is constructed, and three types of agents are defined, namely:
(1) the load management Agent is an intelligent Agent aiming at the load flow direction and the use condition of a demand side, predicts the use of multi-energy loads of a user side, acquires the use condition of the loads in an area in time and makes a reasonable decision to distribute the load flow direction;
(2) energy measurement Agent: an intelligent Agent corresponding to a primary energy side distributed energy power generation and combined cooling heating and power supply system monitors various load demands of users in real time, a regional energy measurement Agent receives output prediction of distributed energy, monitors various micro-source output conditions, and adjusts power generation/heating/refrigerating states of an energy storage and energy supply device in real time to meet energy demands of regional energy internet user sides;
(3) the power price Agent in the network: the system is an intelligent agent aiming at an energy interaction electricity price mechanism between regional energy source internets; aiming at information interaction and energy interaction among different regions, local information transmitted by a local energy measuring Agent is held, and an electricity price mechanism among multi-region energy sources is established based on a non-cooperative dynamic game model;
s3: adopting backward subtraction to establish an RESs output and user basic load prediction model;
simulating uncertainty of load and renewable energy power generation based on a typical scene reduction representation method; and generating various scenes by adopting Monte Carlo simulation, and simulating fluctuation conditions of predicted values in practice by adopting a distributed sampling method according to the wind speed predicted value and the solar radiation angle predicted value to generate random scenes. Determining probability distribution of prediction errors according to historical data to obtain random distribution errors, converting RES random variables into output power according to an output characteristic curve, and expressing the distributed output prediction errors by adopting normal probability distribution; the predicted value of the new energy output in the future T time period is represented by a time sequence, and an output scene is set
Figure GDA0002986449900000081
Is the output value of scene i at time T, and scene omegaiHas an occurrence probability of PiThe minimum probability distance between the scene set before reduction and the final retained scene subset is expressed as follows:
Figure GDA0002986449900000082
in the formula, α represents a set of scenes that are finally deleted after the scene reduction, and the number of scenes is 3000. Initialization reservation set S ═ ω0，ω1，ωi,ωsAdding another scene with the smallest probability distance to the actual abandon set, wherein the scene omega closest to the abandon set is changedlProbability is represented as p (ω'l)＝p(ωl)+p(θk) Until the number of scenes in the abandon set reaches the requirement;
in order to better analyze the system performance, a maximum power point tracking (MTTP) method is applied to a new energy output system, so that the new energy output system operates at a maximum power point. Based on the prediction results, the active output power and base load prediction of the RES is shown as:
Figure GDA0002986449900000083
Figure GDA0002986449900000084
Figure GDA0002986449900000085
the total output of the distributed power supply is as follows:
Pres,i＝Ppv,i+Pw,i (5)
s4: constructing a multi-agent system, and calculating the electricity price of the current k time period through an electricity price mechanism;
the energy storage system model is characterized as follows:
Figure GDA0002986449900000086
in the formula (I), the compound is shown in the specification,
Figure GDA0002986449900000087
for the initial state of charge of the energy storage system,
Figure GDA0002986449900000088
the charge capacity is expected for the energy storage system,
Figure GDA0002986449900000089
in order to be the battery capacity of the energy storage system,
Figure GDA00029864499000000810
for the charging power of the regional energy internet i energy storage system,
Figure GDA00029864499000000811
the discharging power of an energy storage system is the regional energy Internet i;
assuming that all energy storage systems have the same lithium ion battery pack and that the charge/discharge power over a single period is considered constant, therefore, the model and constraints for the energy storage system cells are established as follows:
Figure GDA0002986449900000091
in the formula (I), the compound is shown in the specification,
Figure GDA0002986449900000092
respectively representing the SOC states at time t +1 and t,
Figure GDA00029864499000000911
representing the energy storage battery power at time t, Mi,Bi,Ci and DiRespectively representing a system matrix, an input matrix, an output matrix and a feedforward matrix;
Figure GDA0002986449900000093
Figure GDA0002986449900000094
in the formula (I), the compound is shown in the specification,
Figure GDA0002986449900000095
respectively representing the charging and discharging power, eta, of the energy storage system at time tchAnd ηdchRespectively represent charge/discharge efficiencies;
the gas turbine generator set is used for an energy internet system, has high efficiency, fully utilizes natural gas energy, has small environmental pollution, and outputs to a gas turbine as follows:
Figure GDA0002986449900000096
in the formula
Figure GDA0002986449900000097
Generating power of an energy internet i gas turbine at a time t;
Figure GDA0002986449900000098
the maximum generated power of the gas turbine;
Figure GDA0002986449900000099
recovering power for waste heat of the energy Internet i gas turbine at the time t; etacAnd ηrThe power generation efficiency and the waste heat recovery efficiency of the gas turbine are obtained; lambda [ alpha ]gtAs gas consumption rate, λgasTaking 9.7kWh/m as the heat value of natural gas3；
The method comprises the following steps that electric power is used as a transaction core, and in an energy internet market, an intra-network electricity price Agent participating in bidding aims to obtain maximum benefit through a rational bidding strategy;
the optimization problem of the power price Agent in the power grid is expressed as follows:
Figure GDA00029864499000000910
in the formula, P is the optimization target of the power price Agent in the network in the rolling time domain, namely the power price in the network is the interaction power price rb，rsRespectively setting an internal electricity purchasing price and an internal electricity selling price in a day; a is1And a2The power balance reference electricity price respectively corresponds to the electricity price when net loads of the energy internet in the electricity selling region and the electricity purchasing energy internet are zero; in addition, in the formula
Figure GDA0002986449900000101
The expression of the four variables is as follows:
Figure GDA0002986449900000102
in the formula, PLoadAdjusting load, U, for energy Internet igridThe power is the interaction power of the energy Internet i;
s5: constructing a power balance model of the energy local area network to ensure the balance of supply and demand in the energy local area network, wherein the process is as follows:
constructing a power balance model of an energy local area network, wherein the power balance model of the ith energy local area network can be obtained by constructing each component of the energy local area network at a power supply side and a demand side:
Figure GDA0002986449900000103
wherein the content of the first and second substances,
Figure GDA0002986449900000104
the interaction power between the energy local area network i and the energy local area network in the network is obtained;
Figure GDA0002986449900000105
is transmission line power constraint;
s6: based on model prediction control, minimizing the operation cost of a single energy local area network, and repeating the processes of the steps S2-S5 as follows:
the total running cost of the single energy local area network is minimized, and the optimization problem of the single energy local area network in the rolling time domain can be expressed as a quadratic programming problem:
Figure GDA0002986449900000106
in the formula: tau is the optimized rolling time domain length; u. ofgrid，Qs，QHX，QACEnergy interaction power, energy storage residual capacity, waste heat of a gas turbine for heating and waste heat of a gas turbine for refrigerating are respectively adopted; A. b, C, D are respectively the flexible constraint coefficients of the residual capacity of stored energy, the power generation power of the gas turbine, the waste heat of the gas turbine for heating and the waste heat of the gas turbine for cooling;
s7: the intra-network electricity price Agent receives scheduling information of the load management Agent and the energy measuring Agent, maximizes the income of each participant through a non-cooperative dynamic game model, and determines the electricity price of the participant in real time to obtain a strategy set;
further, the step S7 is as follows:
s71: establishing a game model, establishing a mixed integer model according to an optimization problem of the energy Internet in a rolling time domain, and describing the game model as follows:
the participants are in-network electricity price agents in the set N +, and each participant comprises a distributed energy source and an energy storage system;
strategy: for any i ∈ N+The number of the n-th bit lines, during the k period,
Figure GDA0002986449900000111
aggregate the actions for all participants; the strategy adopted is as follows: including distributed energy output, various demand loads, and other policies taken by participants. The optimization operation strategy maximizes each participant i (i belongs to N)+) Is expressed as piLet ρ beiIs a feasible strategy set;
and (4) yield: used to measure the total profit for each participant, maximize the profit for each participant i, denoted as Ui；
Given the policy set A as followsi＝{A1,A2,…AN}, if and only if:
Figure GDA0002986449900000112
wherein A is a set after updating the strategy set, and the strategy vector A is called Nash equilibrium point (NE), and any regional energy Internet can not improve the respective profit by changing the strategy in a single direction;
s72: policy PiFor the ith energy Internet interaction electricity price, the electricity price strategy set of the energy Internet i is AiAnd A isi＝{P|0≤P≤Pmax}，PmaxFor maximum reportable interactive electricity rates, therefore AiIs a compact convex set, and the participant must have the electricity selling strategy P in the game process, so the set AiIs not empty;
certificate Ui(A) If the function is a concave function, S has a pure strategy Nash equilibrium point; to Ui(A) Carry out a second derivationThe second derivative is as follows:
Figure GDA0002986449900000113
due to rb,λb,PLoad,i(i-1, 2,3, …, N) are all non-negative, i.e. N is not negative
Figure GDA0002986449900000114
Thus U isiThe game is concave, and the non-cooperative game problem S has pure strategy Nash equilibrium points;
s8: after the optimization is completed, a strategy set is obtained, and whether the Nash equilibrium is achieved or not is calculated; repeating the steps S3-S7, and when the optimal scheme is determined to be obtained, stopping the game to obtain the optimal operation mode of the multi-energy microgrid;
and (4) optimizing result comparison analysis:
in order to be able to visually verify the effect of the strategy provided by the invention, the following 3 modes are simulated:
case 1: and (4) not carrying out an optimization mode, not carrying out power interaction on each energy local area network, and generating equipment generates power at full load.
Case 2: and (4) not considering game optimization, only considering power interaction of the energy local area network.
Case 3: the optimized control operation strategy based on the non-cooperative game is provided.
The scene analysis design is 4 different energy local area network structures, the power supply side of each energy local area network system is composed of a photovoltaic, a fan, a gas turbine, energy storage and other power grids, the demand side is supplied by a basic load and an electric refrigerator, wherein the energy interaction among the energy internets is completed through a single bus, and the energy internets net load after interaction is interacted with the external power grids through the single bus. The rated photovoltaic and wind power output power, the energy storage capacity and the gas turbine capacity of each energy local area network system are shown in the following table 1, and the energy conversion equipment parameters are shown in the following table 2. The power transmission capacity of the bus of the energy local area network is 4000kW, and the maximum energy storage charge and discharge power is 2000 kW.
Item photovoltaic/kW Wind power/kW Energy storage/kWh Gas turbine capacity W
Energy local area network 1 3000 2500 13000 2000
Energy local area network 2 3300 3550 13000 2000
Energy local area network 3 4000 3900 13000 1000
Energy local area network 4 3500 3750 13000 1000
TABLE 1
Figure GDA0002986449900000121
TABLE 2
The wind-solar output of each energy local area network is shown in fig. 2, fig. 2(a) is a photovoltaic output curve of the 4 energy local area networks, and fig. 2(b) is a fan output curve of the 4 energy local area networks. Fig. 3 is a cold and heat load curve.
As shown in fig. 4, interactive electricity prices are determined through mutual gaming between the multi-region energy internet, through multiple iterations, the electricity prices finally tend to a stable value, all participants choose not to change own strategies, and respective benefits are maximized. Each participant obtains an optimal strategy through the game, and the convergence process is repeated to finally achieve Nash balance.
As can be seen from fig. 5, in case 3, compared to cases 1 and 2, the peak-to-valley difference is reduced by 82.44% and 29.22%, the fluctuation rate is reduced by 80.05% and 27.08%, the power difference of the energy internet is reduced, and the stability of the system is improved. In addition, compared with the mode 1, the energy utilization rate of the modes 2 and 3 is obviously improved.
As can be seen from fig. 6, when the new energy output is insufficient, the energy storage system actively outputs power to the gas turbine, and each energy internet actively performs power interaction, thereby stabilizing the system load fluctuation. Wherein the energy storage system and the gas turbine play an important role. When the output of the gas turbine is small, the utilization amount of waste heat is small, at the moment, the heat load is mainly provided by the gas boiler, and the cold load is mainly provided by the electric refrigerator; when the output of the gas turbine is high, the heat load is mainly provided by the gas turbine, and the waste heat is utilized for refrigeration to meet the requirement of the cooling load, so that the overall operation efficiency and the fuel utilization rate of the system are improved.
Figure GDA0002986449900000122
Figure GDA0002986449900000131
TABLE 3
Analysis of the data in table 3 leads to the following conclusions:
1) compared with the non-optimized case 1, the cases 2 and 3 have the advantages that the utilization rate of the distributed energy is obviously improved, the loss of the abandoned wind and light is close to 0, and the photoelectric subsidy is respectively improved by 4.85% and 4.82%. In addition, the charge-discharge loss, the electric energy loss and the operation and maintenance cost of case 3 are slightly increased compared with those of case 2, which shows that the energy storage system is actively scheduled in the optimization process, the economic impact is small, and the optimized scheduling scheme is reliable.
2) In case 3, compared with cases 1 and 2, the total income is respectively improved by 290.06% and 123.31%, which shows that the non-cooperative game rolling optimization process has obvious improvement on the economy of the energy internet.
3) Compared with case 1, the natural gas cost of cases 2 and 3 is respectively increased by 11.01 percent and 4.52 percent, and the average power generation cost of the gas turbine is respectively increased by 19.92 percent and 15.11 percent, which shows that cases 2 and 3 do not depend on the gas turbine but select other modes for supplying heat and cold in the optimization process. Compared with case 2, case 3 has less cost for adding natural gas, which shows that the operation strategy can be adjusted in time by the in-network electricity price Agent, so that multiple energy sources can actively participate in the system operation.
And (4) considering the uncertainty of the operation of the energy Internet system, providing a regional energy Internet optimal operation control strategy, and modeling various energy devices in the energy Internet. The minimum operation cost is controlled in each energy local area network based on model prediction, a non-cooperative game is introduced, a novel electricity price mechanism is established aiming at energy interaction, the game reaches Nash balance through iterative calculation, and cases show that the provided optimization operation control strategy can effectively reduce the net load fluctuation rate of the system, reduce the operation economic cost and improve the reliability and the economy of an energy internet system.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples described in this specification can be combined and combined by one skilled in the art.
While embodiments of the present invention have been shown and described, it is to be understood that the embodiments described herein are merely illustrative of the principles of the invention and that the scope of the invention should not be considered limited to the specific forms set forth herein, but rather by the claims that follow and include equivalent technical means that will occur to those skilled in the art upon consideration of the teachings herein.

Claims (5)

1. A multi-energy microgrid decentralized optimization scheduling method based on a multi-agent technology is characterized by comprising the following steps:
s1: initializing a system, considering a discrete time model, setting the optimal time to be 24h, carrying out discretization treatment, equally dividing into T time intervals, wherein k belongs to {1, 2.., T } for any kth time, and the duration of the kth time interval is delta T;
s2: defining a multi-Agent MAS system, which comprises a load management Agent, an energy measurement Agent and an in-network electricity price Agent;
s3: adopting backward subtraction to establish an RESs output and user basic load prediction model;
s4: acquiring multi-agent system information, and calculating the electricity price of the current k time period through an electricity price mechanism;
s5: constructing a power balance model of the energy local area network to ensure the balance of supply and demand in the energy local area network;
s6: based on model prediction control, minimizing the operation cost of a single energy local area network, and repeating the steps S1-S4;
s7: the intra-network electricity price Agent receives scheduling information of the load management Agent and the energy measuring Agent, maximizes the income of each participant through a non-cooperative dynamic game model, and determines the electricity price of the participant in real time to obtain a strategy set;
s8: after the optimization is completed, a strategy set is obtained, and whether the Nash equilibrium is achieved or not is calculated; repeating the steps S2-S6, and when the optimal scheme is determined to be obtained, stopping the game to obtain the optimal operation mode of the multi-energy microgrid;
in step S3, determining prediction error probability distribution according to historical data to obtain a random distribution error, converting the RES random variable into output power according to an output characteristic curve, and expressing the distributed output prediction error by using normal probability distribution; the predicted value of the new energy output in the future T time period is represented by a time sequence, and an output scene is set
Figure FDA0002986449890000011
Figure FDA0002986449890000012
Is the output value of scene i at time T, and scene omegaiHas an occurrence probability of PiThe minimum probability distance between the scene set before reduction and the final retained scene subset is expressed as follows:
Figure FDA0002986449890000016
where α denotes a scene set finally deleted after scene reduction, the number of scenes is 3000, and the initialization reservation set S ═ ω is set0，ω1，ωi,ωsAdding another scene with the smallest probability distance to the actual abandon set, wherein the scene omega closest to the abandon set is changedlProbability is represented as p (ω'l)＝p(ωl)+p(θk) Until the number of scenes in the abandon set reaches the requirement;
in order to better analyze the system performance, a maximum power point tracking (MTTP) method is applied to a new energy output system, so that the new energy output system operates at a maximum power point, and based on a prediction result, active output power and base load prediction of RES are shown as follows:
Figure FDA0002986449890000013
Figure FDA0002986449890000014
Figure FDA0002986449890000015
the total output of the distributed power supply is as follows:
Pres,i＝Ppv,i+Pw,i (5)；
the step S4 process is as follows:
the energy storage system model is characterized as follows:
Figure FDA0002986449890000021
in the formula (I), the compound is shown in the specification,
Figure FDA0002986449890000022
for the initial state of charge of the energy storage system,
Figure FDA0002986449890000023
the charge capacity is expected for the energy storage system,
Figure FDA0002986449890000024
for the battery capacity of the energy storage system, Pi cCharging power, P, for regional energy Internet i energy storage systemi dThe discharging power of an energy storage system is the regional energy Internet i;
assuming that all energy storage systems have the same lithium ion battery pack and that the charge/discharge power over a single period is considered constant, therefore, the model and constraints for the energy storage system cells are established as follows:
Figure FDA0002986449890000025
in the formula (I), the compound is shown in the specification,
Figure FDA0002986449890000026
representing the SOC state, P, at times t +1 and t, respectivelyi tRepresenting the energy storage battery power at time t, Mi,Bi,CiAnd DiRespectively representing a system matrix, an input matrix, an output matrix and a feedforward matrix;
Figure FDA0002986449890000027
Figure FDA0002986449890000028
in the formula (I), the compound is shown in the specification,
Figure FDA0002986449890000029
respectively representing the charging and discharging power, eta, of the energy storage system at time tchAnd ηdchRespectively represent charge/discharge efficiencies;
the gas turbine generator set is used for an energy internet system, has high efficiency, fully utilizes natural gas energy, has small environmental pollution, and outputs to a gas turbine as follows:
Figure FDA00029864498900000210
in the formula
Figure FDA00029864498900000211
Generating power of an energy internet i gas turbine at a time t;
Figure FDA00029864498900000212
the maximum generated power of the gas turbine;
Figure FDA00029864498900000213
recovering power for waste heat of the energy Internet i gas turbine at the time t; etacAnd ηrThe power generation efficiency and the waste heat recovery efficiency of the gas turbine are obtained; lambda [ alpha ]gtAs gas consumption rate, λgasTaking 9.7kWh/m as the heat value of natural gas3；
The method comprises the following steps that electric power is used as a transaction core, and in an energy internet market, an intra-network electricity price Agent participating in bidding aims to obtain maximum benefit through a rational bidding strategy;
the optimization problem of the power price Agent in the power grid is expressed as follows:
Figure FDA0002986449890000031
Figure FDA0002986449890000032
Figure FDA0002986449890000033
in the formula, P is the optimization target of the power price Agent in the network in the rolling time domain, namely the power price in the network is the interaction power price rb，rsRespectively setting an internal electricity purchasing price and an internal electricity selling price in a day; a is1And a2The power balance reference electricity price respectively corresponds to the electricity price when net loads of the energy internet in the electricity selling region and the electricity purchasing energy internet are zero; in addition, in the formula
Figure FDA0002986449890000034
The expression of the four variables is as follows:
Figure FDA0002986449890000035
in the formula, PLoadAdjusting load, U, for energy Internet igridAnd the power is the interaction power of the energy Internet i.
2. The method as claimed in claim 1, wherein in step S2, an energy internet MAS system structure is constructed, defining three types of agents, namely:
(1) the load management Agent is an intelligent Agent aiming at the load flow direction and the use condition of a demand side, predicts the use of multi-energy loads of a user side, acquires the use condition of the loads in an area in time and makes a reasonable decision to distribute the load flow direction;
(2) energy measurement Agent: an intelligent Agent corresponding to a primary energy side distributed energy power generation and combined cooling heating and power supply system monitors various load demands of users in real time, a regional energy measurement Agent receives output prediction of distributed energy, monitors various micro-source output conditions, and adjusts power generation/heating/refrigerating states of an energy storage and energy supply device in real time to meet energy demands of regional energy internet user sides;
(3) the power price Agent in the network: the system is an intelligent agent aiming at an energy interaction electricity price mechanism between regional energy source internets; aiming at information interaction and energy interaction among different regions, local information transmitted by a local energy measuring Agent is held, and an electricity price mechanism among multi-region energy sources is established based on a non-cooperative dynamic game model.
3. The method as claimed in claim 1, wherein the step S5 is performed by the following steps:
constructing a power balance model of an energy local area network, wherein the power balance model of the ith energy local area network can be obtained by constructing each component of the energy local area network at a power supply side and a demand side:
Figure FDA0002986449890000041
Figure FDA0002986449890000042
wherein the content of the first and second substances,
Figure FDA0002986449890000043
the interaction power between the energy local area network i and the energy local area network in the network is obtained;
Figure FDA0002986449890000044
is a transmission line power constraint.
4. The method as claimed in claim 3, wherein the step S6 is performed by the following steps:
s61: the total running cost of the single energy local area network is minimized, and the optimization problem of the single energy local area network in the rolling time domain can be expressed as a quadratic programming problem:
Figure FDA0002986449890000045
in the formula: tau is the optimized rolling time domain length; u. ofgrid，Qs，QHX，QACEnergy interaction power, energy storage residual capacity, waste heat of a gas turbine for heating and waste heat of a gas turbine for refrigerating are respectively adopted; A. b, C, D are respectively flexible restraint systems of residual capacity of stored energy, power generated by gas turbine, waste heat of gas turbine for heating and waste heat of gas turbine for coolingAnd (4) counting.
5. The method for decentralized optimized dispatching of multi-energy microgrid based on multi-agent technology as claimed in claim 1 or 2, wherein said step S7 procedure is as follows:
s71: establishing a game model, establishing a mixed integer model according to an optimization problem of the energy Internet in a rolling time domain, and describing the game model as follows:
the participants are in-network electricity price agents in the set N +, and each participant comprises a distributed energy source and an energy storage system;
strategy: for any i ∈ N+The number of the n-th bit lines, during the k period,
Figure FDA0002986449890000046
aggregate the actions for all participants; the strategy adopted is as follows: the optimization operation strategy maximizes each participant i (i belongs to N) including distributed energy output, various demand loads and strategies adopted by other participants+) Is expressed as piLet ρ beiIs a feasible strategy set;
and (4) yield: used to measure the total profit for each participant, maximize the profit for each participant i, denoted as Ui；
Given policy set Ai＝{A1,A2,…AN}, if and only if:
Figure FDA0002986449890000051
a is a set after updating the strategy set, called strategy vector A is a Nash equilibrium point, and any regional energy Internet can not improve the respective profit by changing the strategy unilaterally;
s72: policy PiFor the ith energy Internet interaction electricity price, the electricity price strategy set of the energy Internet i is AiAnd A isi＝{P|0≤P≤Pmax}，PmaxFor maximum reportable interactive electricity rates, therefore AiIs a compact convex set, and the participant must have the electricity selling strategy P in the game process, so the set AiIs not empty;
certificate Ui(A) If the function is a concave function, S has a pure strategy Nash equilibrium point; to Ui(A) Performing second derivation, wherein the second derivation is as follows:
Figure FDA0002986449890000052
due to rb,λb,PLoad,i(i-1, 2,3, …, N) are all non-negative, i.e. N is not negative
Figure FDA0002986449890000053
Thus U isiAnd the non-cooperative game problem S has pure strategy Nash equilibrium points.
CN201811579886.4A 2018-12-24 2018-12-24 Multi-energy microgrid decentralized optimization scheduling method based on multi-agent technology Active CN109636056B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811579886.4A CN109636056B (en) 2018-12-24 2018-12-24 Multi-energy microgrid decentralized optimization scheduling method based on multi-agent technology

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811579886.4A CN109636056B (en) 2018-12-24 2018-12-24 Multi-energy microgrid decentralized optimization scheduling method based on multi-agent technology

Publications (2)

Publication Number Publication Date
CN109636056A CN109636056A (en) 2019-04-16
CN109636056B true CN109636056B (en) 2021-08-03

Family

ID=66076731

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811579886.4A Active CN109636056B (en) 2018-12-24 2018-12-24 Multi-energy microgrid decentralized optimization scheduling method based on multi-agent technology

Country Status (1)

Country Link
CN (1) CN109636056B (en)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110867902B (en) * 2019-10-15 2023-05-16 东北大学 Micro-grid distributed power supply decentration optimization operation method based on power generation prediction
CN110728406B (en) * 2019-10-15 2022-07-29 南京邮电大学 Multi-agent power generation optimal scheduling method based on reinforcement learning
CN110989339A (en) * 2019-11-01 2020-04-10 浙江工业大学 Modeling method for multi-source power system by using capacitive energy storage technology
CN111404148B (en) * 2020-03-23 2021-06-15 华翔翔能科技股份有限公司 Wind, light and methane multi-microgrid system and peer-to-peer energy and communication transaction method thereof
CN112651112B (en) * 2020-12-17 2023-07-11 湖南大学 Collaborative decision-making method, system and equipment for electric energy transaction and system operation of internet micro-grid
CN112784439A (en) * 2021-02-09 2021-05-11 清华大学 Energy internet planning method and device based on discretization model
CN113128036B (en) * 2021-04-08 2022-08-30 国网山东省电力公司电力科学研究院 Multi-main-body energy system interactive operation coordination control method and system
CN113935540A (en) * 2021-10-28 2022-01-14 昆明电力交易中心有限责任公司 Cooperative game-based interactive energy optimization method for comprehensive energy park
CN114611772B (en) * 2022-02-24 2024-04-19 华南理工大学 Multi-agent reinforcement learning-based multi-microgrid system collaborative optimization method

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105977991B (en) * 2016-05-10 2018-07-10 浙江工业大学 A kind of self microgrid Optimal Configuration Method for considering price type demand response
CN106230020B (en) * 2016-08-11 2018-10-19 浙江工业大学 The electric vehicle interactive response control method of distributed generation resource consumption is considered under a kind of micro-capacitance sensor
CN106655243B (en) * 2016-08-12 2019-07-30 浙江工业大学 The automatic demand response method of electric car for the purpose of maintaining the micro-capacitance sensor equilibrium of supply and demand
CN106410861B (en) * 2016-11-04 2018-09-18 浙江工业大学 A kind of micro-capacitance sensor optimization operation real-time control method based on schedulable ability
CN107545325B (en) * 2017-08-21 2021-02-26 浙江工业大学 Multi-microgrid interconnection operation optimization method based on game theory
CN107895971B (en) * 2017-11-28 2021-05-07 国网山东省电力公司德州供电公司 Regional energy Internet scheduling method based on stochastic programming and model predictive control
CN108376999B (en) * 2018-04-02 2020-08-18 浙江工业大学 Multi-microgrid fault management method considering uncertainty of island operation time

Also Published As

Publication number Publication date
CN109636056A (en) 2019-04-16

Similar Documents

Publication Publication Date Title
CN109636056B (en) Multi-energy microgrid decentralized optimization scheduling method based on multi-agent technology
Hou et al. Multi-objective economic dispatch of a microgrid considering electric vehicle and transferable load
Xin-gang et al. Economic-environmental dispatch of microgrid based on improved quantum particle swarm optimization
Roslan et al. Scheduling controller for microgrids energy management system using optimization algorithm in achieving cost saving and emission reduction
Zhao et al. An MAS based energy management system for a stand-alone microgrid at high altitude
CN106849190B (en) A kind of microgrid real-time scheduling method of providing multiple forms of energy to complement each other based on Rollout algorithm
Moghaddam et al. Multi-operation management of a typical micro-grids using Particle Swarm Optimization: A comparative study
Hannan et al. Binary particle swarm optimization for scheduling MG integrated virtual power plant toward energy saving
Motevasel et al. Multi-objective energy management of CHP (combined heat and power)-based micro-grid
Brekken et al. Optimal energy storage sizing and control for wind power applications
Daneshvar et al. Energy exchange control in multiple microgrids with transactive energy management
CN106451550B (en) A kind of micro-grid connection Optimization Scheduling based on improvement subgradient population
Wu et al. A multi-agent-based energy-coordination control system for grid-connected large-scale wind–photovoltaic energy storage power-generation units
CN111934360B (en) Virtual power plant-energy storage system energy collaborative optimization regulation and control method based on model predictive control
CN109861302B (en) Master-slave game-based energy internet day-ahead optimization control method
CN112583017A (en) Hybrid micro-grid energy distribution method and system considering energy storage operation constraint
CN115099725B (en) Energy management and distribution method and system for comprehensive energy system
Yin et al. Relaxed deep generative adversarial networks for real-time economic smart generation dispatch and control of integrated energy systems
Niknam et al. Probabilistic model of polymer exchange fuel cell power plants for hydrogen, thermal and electrical energy management
CN110889581A (en) Electric vehicle-participated transformer area optimal scheduling method and system
Yoldas et al. Optimal control of microgrids with multi-stage mixed-integer nonlinear programming guided $ Q $-learning algorithm
CN112311017A (en) Optimal collaborative scheduling method for virtual power plant and main network
CN115241923A (en) Micro-grid multi-objective optimization configuration method based on snake optimization algorithm
CN115114854A (en) Two-stage self-organizing optimization aggregation method and system for distributed resources of virtual power plant
Yunshou et al. Cooperative operation framework for a wind-solar-CCHP multi-energy system based on Nash bargaining solution

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