CN115660841A - Community-level P2P energy transaction bipartite graph matching method considering social influence - Google Patents

Community-level P2P energy transaction bipartite graph matching method considering social influence Download PDF

Info

Publication number
CN115660841A
CN115660841A CN202211193861.7A CN202211193861A CN115660841A CN 115660841 A CN115660841 A CN 115660841A CN 202211193861 A CN202211193861 A CN 202211193861A CN 115660841 A CN115660841 A CN 115660841A
Authority
CN
China
Prior art keywords
community
energy
user
transaction
power
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.)
Pending
Application number
CN202211193861.7A
Other languages
Chinese (zh)
Inventor
夏元兴
徐青山
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.)
Southeast University
Original Assignee
Southeast University
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 Southeast University filed Critical Southeast University
Priority to CN202211193861.7A priority Critical patent/CN115660841A/en
Publication of CN115660841A publication Critical patent/CN115660841A/en
Pending legal-status Critical Current

Links

Images

Landscapes

  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a bipartite graph matching method for community-level P2P energy trading considering social influence, which is used for acquiring user load, electricity price, new energy output range, power generation and utilization preference of users and network parameter data of a power distribution network accessed by a community-level microgrid, and transmitting the data as parameters into an optimization model; analyzing the condition of end-to-end transaction in the community, establishing an internal transaction optimization model of the community, and analyzing the social influence of the transaction in the community; modeling the transaction process inside each community-level microgrid by utilizing multi-subject deep reinforcement learning to obtain an energy transaction model inside each community-level microgrid; matching is carried out among communities, and the total energy exceeding or shortage part in the communities is consumed; and after the energy producers and the energy consumers in each community are traded, outputting the energy shortage/excess part of each community so as to match the communities, solving the optimal matching result by using the Hungarian algorithm in the matching process, and minimizing the weight sum of the matching lines.

Description

Bipartite graph matching method considering social influence for community-level P2P energy transaction
Technical Field
The invention belongs to the technical field of P2P (peer-to-peer) transaction in an electric power market, and particularly relates to a bipartite graph matching method for community-level P2P energy transaction considering social influence.
Background
In recent years, with the development of distributed energy technology, end users in electric power systems have become more active. Many users who have installed photovoltaic cell panel, battery energy storage system and electric automobile can sell unnecessary electric power for the system according to the price of electricity of surfing the net. Therefore, the concept of energy producers and consumers is now commonly introduced to describe such active end users. The local P2P energy trading market between the producers and the consumers is one of energy trading scenes which are increasingly important in the fields of power distribution networks and micro-grids in recent years, but the producers and the consumers are often distributed in different community-level micro-grids, each community has excess power and shortage, the mutual influence of the electricity utilization behaviors of users in the same community is not considered in the existing P2P trading, the trading matching difficulty among the communities is high, and the optimal matching result is difficult to solve in polynomial time by a traditional algebraic trading matching model.
Disclosure of Invention
In order to solve the above mentioned shortcomings in the background art, the present invention provides a bipartite graph matching method for community-level P2P energy trading in consideration of social influence.
The purpose of the invention can be realized by the following technical scheme: a bipartite graph matching method for community-level P2P energy trading considering social influence comprises the following steps:
acquiring use data, wherein the use data comprises user load, electricity price, new energy output range, power generation and utilization preference of users and network parameter data of a power distribution network accessed by a community-level microgrid;
analyzing the condition of end-to-end transaction in the community, establishing an internal transaction optimization model of the community, transmitting the obtained use data serving as parameters into the internal transaction optimization model of the community, and analyzing the social influence of the transaction in the community;
modeling the transaction process inside each community-level microgrid by utilizing multi-subject deep reinforcement learning to obtain an energy transaction model inside each community-level microgrid;
after an energy transaction model inside each community-level microgrid is obtained, matching is carried out among communities to consume the part of total energy exceeding or shortage inside the communities, and the matching problem between energy buyers and energy sellers is described in a bipartite graph mode to maximize the benefit of an energy seller of a matching result and minimize the electricity purchasing cost after the energy buyer is matched;
and after the energy producers and the energy consumers in each community are traded, outputting the energy shortage/excess part of each community so as to match each community, solving the optimal matching result by using the Hungarian algorithm in the matching process, and simultaneously minimizing the weight sum of the matching lines so as to minimize the network passing cost in the trading process.
Preferably, the user load comprises year-round load data of the user, and the data collection interval of the user load is 15 minutes at the minimum.
Preferably, the electricity prices include national uniform peak-valley average three-hour electricity prices, and internet access electricity prices of the end users; the new energy output range comprises the upper and lower power generation limits of photovoltaic power generation and wind power generation renewable energy; the power generation and utilization preference of the user comprises a cost function of power generation of the user, a utility function of power utilization and the social state of the user; the network parameters of the power distribution network accessed by the community-level microgrid comprise line parameters of the whole power distribution network, fluctuation ranges of voltages of all nodes and a grid structure of the power distribution network.
Preferably, the process of establishing the intra-community transaction optimization model comprises the following steps:
the internal community transaction optimization model is established as follows:
Figure BDA0003870070700000021
in the formula (I), the compound is shown in the specification,
Figure BDA0003870070700000022
for the payload of user i at time t,
Figure BDA0003870070700000023
i.e. the utility that user i gets to consume the payload at time t,
Figure BDA0003870070700000024
for the electricity purchasing quantity of the P2P trading market in the community,
Figure BDA0003870070700000025
the E-purchase price of the user i at the P2P at the time t, xi t The method comprises the steps of collecting decision variables of the whole community-level market, namely net load and the electricity purchasing quantity of users in a P2P trading market in a community;
schedulable resources for each energy producer and consumer in the community include reducible load, interruptible load, transferable load and distributed energy storage, modeled as follows:
Figure BDA0003870070700000031
Figure BDA0003870070700000032
Figure BDA0003870070700000033
Figure BDA0003870070700000034
Figure BDA0003870070700000035
Figure BDA0003870070700000036
Figure BDA0003870070700000037
Figure BDA0003870070700000038
in the formula (I), the compound is shown in the specification,
Figure BDA0003870070700000039
representing the set of users i in community n,
Figure BDA00038700707000000310
for the run-time of the entire market,
Figure BDA00038700707000000311
in order to load the total amount of resources for flexibility,
Figure BDA00038700707000000312
respectively, user i can reduce load, interrupt load and transfer load at time t,
Figure BDA00038700707000000313
the amount of load over the total time period required for transferable loads,
Figure BDA00038700707000000314
for the rigid load of user i at time t,
Figure BDA00038700707000000315
for the payload of user i at time t,
Figure BDA00038700707000000316
charging power for user i at time t, P i cha The rated charging power for the stored energy of user i,
Figure BDA00038700707000000317
for the charging state of the stored energy of user i,
Figure BDA00038700707000000318
discharge power, P, for user i at time t i dis The rated discharge power of the stored energy for user i,
Figure BDA00038700707000000319
is the discharge state of the stored energy of user i,
Figure BDA00038700707000000320
for the energy storage capacity of the user i at the time t,
Figure BDA00038700707000000321
charging and discharging coefficients of the stored energy of the user i are respectively, delta t is the charging and discharging time step length of the stored energy,
Figure BDA00038700707000000322
the lower limit and the upper limit of the energy storage capacity of the user i.
Preferably, in addition to the operating constraints of the energy producers and consumers' distributed power supplies, the intra-community transactions also need to satisfy the following power balance constraints:
Figure BDA0003870070700000041
Figure BDA0003870070700000042
in the formula (I), the compound is shown in the specification,
Figure BDA0003870070700000043
for the purchased electric power of the nth community to the upper level grid at time t,
Figure BDA0003870070700000044
photovoltaic power generation power of user i at time t，
Figure BDA0003870070700000045
Transacting electric energy for P2P of user i at time t when
Figure BDA0003870070700000046
Positive indicates that user i buys power at time t when
Figure BDA0003870070700000047
A negative indicates that user i sells at time t.
Preferably, the process of obtaining an energy transaction model inside each community-level microgrid comprises:
modeling an end-to-end transaction process within each community as a Markov process, wherein the Markov observations are set as follows:
Figure BDA0003870070700000048
Figure BDA0003870070700000049
Figure BDA00038700707000000410
in the formula (I), the compound is shown in the specification,
Figure BDA00038700707000000411
which is an observation set on the power generation side, t represents an index of time,
Figure BDA00038700707000000412
representing the P2P transaction electricity price of user i at time t,
Figure BDA00038700707000000413
representing the photovoltaic power generation power of the user i at the moment t,
Figure BDA00038700707000000414
a value representing the stiffness load of the user i,
Figure BDA00038700707000000415
indicating the amount of energy stored by user i at time t,
Figure BDA00038700707000000416
for a set of social network observations within a community,
Figure BDA00038700707000000417
respectively represents the operation difficulty, the communication degree, the information depth, the social influence degree and the influence of the friends of the transaction behavior, the higher the numerical value is, the greater the influence of the item on the transaction result is,
Figure BDA00038700707000000418
is an observation set of the whole market;
the Markov action set is built up as follows:
Figure BDA0003870070700000051
Figure BDA0003870070700000052
Figure BDA0003870070700000053
Figure BDA0003870070700000054
Figure BDA0003870070700000055
Figure BDA0003870070700000056
in the formula (I), the compound is shown in the specification,
Figure BDA0003870070700000057
in order to be able to reduce the upper limit of the load,
Figure BDA0003870070700000058
an upper limit for the interruptible load is,
Figure BDA0003870070700000059
is the upper limit of the transferable load,
Figure BDA00038700707000000510
is a Markov motion value;
after the markov action set and the markov observation set are obtained, the state transition process of each individual can be updated according to the action, and the updated state transition is as follows:
Figure BDA00038700707000000511
Figure BDA00038700707000000512
Figure BDA00038700707000000513
Figure BDA00038700707000000514
Figure BDA00038700707000000515
preferably, after obtaining the state transition, defining a reward function corresponding to the state transition as follows:
Figure BDA00038700707000000516
Figure BDA00038700707000000517
Figure BDA00038700707000000518
Figure BDA0003870070700000061
Figure BDA0003870070700000062
in the formula, r i cost,t ,r i uti,t ,r i psy,t ,r i DER,t Respectively represent the electricity selling income and the electricity purchasing utility,
formula (II)
Figure BDA0003870070700000063
Psychological rewards and usage and maintenance costs of distributed power sources that are consistent with the ability of energy producers and consumers in the same community to maintain behavior,
Figure BDA0003870070700000064
meaning that the net load is averaged,
Figure BDA0003870070700000065
and respectively representing the maintenance cost of photovoltaic and distributed energy storage, and obtaining a return function value by taking opposite numbers.
Preferably, the process of matching between communities comprises:
the community P2P matching problem is written as follows:
Figure BDA0003870070700000066
Figure BDA0003870070700000067
Figure BDA0003870070700000068
Figure BDA0003870070700000069
Figure BDA00038700707000000610
Figure BDA00038700707000000611
Figure BDA00038700707000000612
Figure BDA00038700707000000613
in the formula (I), the compound is shown in the specification,
Figure BDA00038700707000000614
for the price community n sells electricity to community m,
Figure BDA00038700707000000615
selling electricity to the community m for the community n,
Figure BDA00038700707000000616
active loss, p, for the entire distribution network transaction t Is one by oneA vector of the contribution power of the community,
Figure BDA00038700707000000617
for the lagrange dual variables corresponding to active balance,
Figure BDA00038700707000000618
for reactive losses of the entire distribution network transaction, q t A vector composed of the reactive power output of each community,
Figure BDA0003870070700000071
for Lagrange dual variables corresponding to reactive balancing, p min,t ,p max,t Respectively the active output lower limit and the active output upper limit of each community,
Figure BDA0003870070700000072
for corresponding Lagrangian dual variables, q min,t ,q max,t Respectively the lower limit and the upper limit of reactive power output of each community,
Figure BDA0003870070700000073
for the corresponding Lagrangian dual variable, v min,t ,v max,t Respectively corresponding to the lower limit and the upper limit of the voltage amplitude for each node,
Figure BDA0003870070700000074
for the corresponding Lagrangian dual variable, | s start,t || 2 Is the square of the complex power amplitude at the beginning of the branch, | s end,t || 2 Is the square of the complex power amplitude at the end of the branch, (S) t ) 2 The upper limit of the complex power amplitude of the whole branch circuit is defined;
in order to minimize the influence of transactions on a power distribution network when P2P transactions are matched among communities, the P2P matching problem of the communities is linearized, and then a Lagrangian function is solved as follows:
Figure BDA0003870070700000075
Figure BDA0003870070700000076
Figure BDA0003870070700000077
Figure BDA0003870070700000078
Figure BDA0003870070700000079
Figure BDA00038700707000000710
in the formula, pi t Representing the weight of the transaction for community n to community m,
Figure BDA00038700707000000711
respectively, representing the effects of energy, loss, node voltage and line blocking on the network.
Preferably, an apparatus comprises:
one or more processors;
a memory for storing one or more programs;
when executed by one or more of the processors, cause the one or more processors to implement a bipartite graph matching method for community-level P2P energy trading that takes into account social influences, as described above.
Preferably, a storage medium containing computer executable instructions for performing a bipartite graph matching method for community-level P2P energy trading taking social influence into account as described above when executed by a computer processor.
The invention has the beneficial effects that:
the invention analyzes the social relationship which determines the energy utilization decision of each energy producer and consumer in each community, creates a multi-subject deep reinforcement learning model based on the social relationship, and accurately delineates the transaction process while ensuring the normal transaction of each user in the community. And considering the line parameters in the network in the process of matching optimization, and optimizing and matching the even and Hungarian algorithms by using Lagrange, the influence of the P2P energy use matching between the social intervals on the power distribution network is fully considered.
Drawings
In order to more clearly illustrate the embodiments or technical solutions in the prior art of the present invention, the drawings used in the description of the embodiments or prior art will be briefly described below, and it is obvious for those skilled in the art that other drawings can be obtained based on these drawings without creative efforts.
FIG. 1 is a schematic diagram of a three-layer structure including communities, energy producers and consumers, and power distribution network operators in a P2P trading market according to the present invention;
FIG. 2 is a diagram of a modeling architecture for the cyber-physical-social system according to the present invention;
FIG. 3 is a schematic diagram of how each community-level microgrid forms a buyer and a seller according to the shortage and excess of internal energy in the embodiment of the invention;
FIG. 4 is a flow chart of a coupling algorithm of multi-agent deep reinforcement learning and Hungarian algorithm according to an embodiment of the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1, a bipartite graph matching method for community-level P2P energy trading considering social influence includes the following steps:
(1) Acquiring user load, electricity price, new energy output range, power generation and utilization preference of users and network parameter data of a power distribution network accessed by a community-level microgrid, and transmitting the collected data serving as parameters into an optimization model;
further, the user load data comprises the load data of the user all year round, and the data acquisition interval is minimum 15 minutes;
further, the electricity price information comprises national uniform peak-valley average three-hour electricity prices and internet access electricity prices of terminal users;
further, the new energy output range comprises the upper and lower power generation limits of renewable energy sources such as photovoltaic power generation and wind power generation;
furthermore, the power generation and utilization preference of the user comprises a cost function of power generation of the user, a utility function of power utilization and the social state of the user;
furthermore, the network parameters of the power distribution network comprise line parameters of the whole power distribution network, the fluctuation range of the voltage of each node and the grid structure of the power distribution network.
(2) Analyzing the end-to-end transaction condition in the community, establishing an internal transaction optimization model of the community, and analyzing the social influence of the transaction in the community;
(21) The optimization goal of the intra-community optimization model is to maximize the welfare of the producers and consumers. Thus, the model for each community can be established as follows:
Figure BDA0003870070700000101
in the formula (I), the compound is shown in the specification,
Figure BDA0003870070700000102
for the payload of user i at time t,
Figure BDA0003870070700000103
i.e. the utility that user i gets by consuming so much load at time t,
Figure BDA0003870070700000104
for the user to purchase the electricity quantity on the P2P trading market in the community,
Figure BDA0003870070700000105
the P2P electricity purchase price, xi, of the user i at the time t t A set of decision variables for the entire community-level market.
(22) Schedulable resources of each energy producer and consumer in the community mainly include reducible load, interruptible load, transferable load and distributed energy storage, and can be modeled as follows:
Figure BDA0003870070700000106
Figure BDA0003870070700000107
Figure BDA0003870070700000108
Figure BDA0003870070700000109
Figure BDA00038700707000001010
Figure BDA00038700707000001011
Figure BDA00038700707000001012
Figure BDA00038700707000001013
in the formula (I), the compound is shown in the specification,
Figure BDA00038700707000001014
representing the set of users i in community n,
Figure BDA00038700707000001015
for the run-time of the entire market,
Figure BDA00038700707000001016
in order to load the total amount of resources for flexibility,
Figure BDA00038700707000001017
respectively, user i's reducible load, interruptible load, transferable load at time t,
Figure BDA00038700707000001018
the amount of load over the total time period required for transferable loads,
Figure BDA00038700707000001019
for the rigid load of user i at time t,
Figure BDA00038700707000001020
for the payload of user i at time t,
Figure BDA00038700707000001021
charging power for user i at time t, P i cha The rated charging power for the stored energy of user i,
Figure BDA00038700707000001022
for the charging state of the stored energy of user i,
Figure BDA00038700707000001023
discharge power, P, for user i at time t i dis The rated discharge power of the stored energy for user i,
Figure BDA00038700707000001024
is the discharge state of the stored energy of user i,
Figure BDA00038700707000001025
for the energy storage capacity of the user i at the time t,
Figure BDA0003870070700000111
the charging and discharging coefficients of the stored energy of the user i are respectively, delta t is the charging and discharging time step length of the stored energy,
Figure BDA0003870070700000112
the lower limit and the upper limit of the energy storage capacity of the user i.
(23) In addition to distributed power supply operating constraints for each of the victims, intra-community transactions also need to satisfy the following power balance constraints:
Figure BDA0003870070700000113
Figure BDA0003870070700000114
in the formula (I), the compound is shown in the specification,
Figure BDA0003870070700000115
for the purchased electric power of the nth community to the upper level grid at time t,
Figure BDA0003870070700000116
for the photovoltaic power generation power of the user i at the time t,
Figure BDA0003870070700000117
and (3) trading electric energy for the P2P of the user i at the moment t, wherein the positive state indicates that the user i buys electricity at the moment t, and the negative state indicates that the user i sells electricity at the moment t.
(3) As shown in fig. 2, in order to consider social influence among energy users in each community-level microgrid, the method adopts multi-agent deep reinforcement learning to model the transaction process in each microgrid, and models the whole community-level microgrid as an information-physics-social system.
(31) In order to model deep reinforcement learning, the invention models an end-to-end transaction process in each community as a Markov process, wherein the Markov observation set is as follows:
Figure BDA0003870070700000118
Figure BDA0003870070700000119
Figure BDA00038700707000001110
in the formula (I), the compound is shown in the specification,
Figure BDA00038700707000001111
is an observation set on the power generation side, wherein the meaning of each variable is consistent with that of the corresponding variable in (2),
Figure BDA00038700707000001112
for a set of social network observations within a community,
Figure BDA00038700707000001113
respectively represents the operation difficulty, the communication degree, the information depth, the social influence degree and the influence of the friends of the transaction behavior, the higher the numerical value is, the greater the influence of the item on the transaction result is,
Figure BDA00038700707000001114
is the set of observations across the market.
(32) In addition to the observation set, the present invention builds a Markov action set as follows:
Figure BDA00038700707000001115
Figure BDA0003870070700000121
Figure BDA0003870070700000122
Figure BDA0003870070700000123
Figure BDA0003870070700000124
Figure BDA0003870070700000125
in the formula (I), the compound is shown in the specification,
Figure BDA0003870070700000126
in order to be able to reduce the upper limit of the load,
Figure BDA0003870070700000127
is the upper limit of the interruptible load,
Figure BDA0003870070700000128
is the upper limit of the transferable load,
Figure BDA0003870070700000129
is a markov action value.
(33) After the action set and the state set of the Markov process are obtained, the invention can update the state transition process of each individual according to the action, and the updated state transition is as follows:
Figure BDA00038700707000001210
Figure BDA00038700707000001211
Figure BDA00038700707000001212
Figure BDA00038700707000001213
Figure BDA00038700707000001214
the meaning of each variable in the formula (ii) is identical to that of the corresponding variable in (31) and (32).
(34) After obtaining the state transition action, the invention defines the reward function corresponding to the state transition as follows:
Figure BDA00038700707000001215
Figure BDA00038700707000001216
Figure BDA00038700707000001217
Figure BDA00038700707000001218
Figure BDA00038700707000001219
in the formula, r i cost,t ,r i uti,t ,r i psy,t ,r i DER,t Respectively represents the electricity selling income, the electricity purchasing utility, the psychological reward keeping consistent with the behavior of energy producers and consumers in the same community and the use and maintenance cost of the distributed power supply,
Figure BDA0003870070700000131
meaning that the net load is averaged,
Figure BDA0003870070700000132
and respectively representing the maintenance cost of photovoltaic and distributed energy storage, and obtaining a return function value by taking the opposite number.
(4) After an energy transaction model inside each community-level microgrid is obtained, matching can be performed between communities so as to further consume the part of total energy excess or shortage inside the communities.
(41) The community P2P matching problem at the entire distribution network level can be written as follows:
Figure BDA0003870070700000133
Figure BDA0003870070700000134
Figure BDA0003870070700000135
Figure BDA0003870070700000136
Figure BDA0003870070700000137
Figure BDA0003870070700000138
Figure BDA0003870070700000139
Figure BDA00038700707000001310
in the formula (I), the compound is shown in the specification,
Figure BDA00038700707000001311
for the price community n sells electricity to community m,
Figure BDA00038700707000001312
the amount of electricity sold to community m for community n,
Figure BDA00038700707000001313
active loss for the entire distribution network transaction, p t A vector of contribution powers is formed for each community,
Figure BDA00038700707000001314
for lagrange dual variables corresponding to active balancing,
Figure BDA00038700707000001315
for reactive losses of the entire distribution network transaction, q t Is a vector formed by the reactive output power of each community,
Figure BDA00038700707000001316
for lagrange dual variables corresponding to reactive balance, p min,t ,p max,t Respectively the active output lower limit and the active output upper limit of each community,
Figure BDA00038700707000001317
for corresponding Lagrangian dual variables, q min,t ,q max,t Respectively the lower limit and the upper limit of reactive power output of each community,
Figure BDA00038700707000001318
for corresponding Lagrangian dual variables, v min,t ,v max,t Respectively corresponding to the lower limit and the upper limit of the voltage amplitude for each node,
Figure BDA0003870070700000141
for the corresponding Lagrangian dual variable, | s start,t || 2 Is the square of the complex power amplitude at the beginning of the branch, | s end,t || 2 Is the square of the complex power amplitude at the end of the branch, (S) t ) 2 The upper limit of the complex power amplitude of the whole branch.
(42) In order to minimize the influence of the transaction on the power distribution network when the P2P transaction is matched among communities, the method linearizes the original problem, and then solves the Lagrangian function as follows:
Figure BDA0003870070700000142
Figure BDA0003870070700000143
Figure BDA0003870070700000144
Figure BDA0003870070700000145
Figure BDA0003870070700000146
Figure BDA0003870070700000147
in the formula, pi t Representing the weight of the transaction for community n to community m,
Figure BDA0003870070700000148
respectively representing the effects of energy, loss, node voltage and line blocking on the network, the present invention reduces the impact of inter-community energy trading on the distribution network by minimizing the sum of these matching weights while matching.
(5) After the energy producers and consumers in each community trade according to the multi-agent deep learning algorithm in fig. 4, the energy shortage/excess part of each community can be output, so that the communities are matched, and the sum of the weights of the matching lines is minimized while the best matching result is solved by using the hungarian algorithm in the matching process.
Preferably, an apparatus comprises:
one or more processors;
a memory for storing one or more programs;
when executed by one or more of the processors, cause the one or more processors to implement a bipartite graph matching method for community-level P2P energy trading that takes into account social influences, as described above.
Preferably, a storage medium containing computer executable instructions for performing a bipartite graph matching method for community-level P2P energy trading taking social influence into account as described above when executed by a computer processor.
In the description herein, references to the description of "one embodiment," "an example," "a specific example" or the like are intended to 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 do not necessarily 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.
The foregoing shows and describes the general principles, essential features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed.

Claims (10)

1. A bipartite graph matching method for community-level P2P energy trading taking social influence into account, comprising the steps of:
acquiring use data, wherein the use data comprises user load, electricity price, new energy output range, power generation and utilization preference of users and network parameter data of a power distribution network accessed by a community-level microgrid;
analyzing the end-to-end transaction condition in the community, establishing an internal community transaction optimization model, transmitting the obtained use data serving as parameters into the internal community transaction optimization model, and analyzing the social influence of the transaction in the community;
modeling the transaction process inside each community-level microgrid by using multi-agent deep reinforcement learning to obtain an energy transaction model inside each community-level microgrid;
after an energy transaction model inside each community-level microgrid is obtained, matching is carried out among communities to consume the part of total energy excess or shortage inside the communities, and the matching problem between energy buyers and energy sellers is described in a bipartite graph mode to maximize the income of an energy seller of a matching result and minimize the electricity purchasing cost after an energy buyer is matched;
and after the energy producers and the energy consumers in each community are traded, outputting the energy shortage/excess part of each community so as to match each community, solving the optimal matching result by using the Hungarian algorithm in the matching process, and simultaneously minimizing the weight sum of the matching lines so as to minimize the network passing cost in the trading process.
2. The bipartite graph matching method for community-level P2P energy trading taking social influence into account of claim 1, wherein the user load comprises year-round load data of the user, and the data collection interval of the user load is 15 minutes at a minimum.
3. The bipartite graph matching method for community-level P2P energy trading taking into account social influence as claimed in claim 1, wherein the electricity prices include national uniform peak-valley-average three-hour electricity prices, end-user internet electricity prices; the new energy output range comprises the upper and lower power generation limits of photovoltaic power generation and wind power generation renewable energy; the power generation and utilization preference of the user comprises a cost function of power generation of the user, a utility function of power utilization and the social state of the user; the network parameters of the power distribution network accessed by the community-level microgrid comprise line parameters of the whole power distribution network, fluctuation ranges of voltages of all nodes and a grid structure of the power distribution network.
4. The bipartite graph matching method for community-level P2P energy trading with consideration of social influence according to claim 1, wherein the process of establishing a community internal trading optimization model comprises the steps of:
the internal community transaction optimization model is established as follows:
Figure FDA0003870070690000021
in the formula (I), the compound is shown in the specification,
Figure FDA0003870070690000022
for the payload of user i at time t,
Figure FDA0003870070690000023
i.e. the utility that user i gets to consume the payload at time t,
Figure FDA0003870070690000024
for the electricity purchasing quantity of the P2P trading market in the community,
Figure FDA0003870070690000025
the E-purchase price of the user i at the P2P at the time t, xi t A set of decision variables for the whole community-level market, the net load and the electricity quantity bought by the user in the P2P trading market in the community;
schedulable resources for each energy producer and consumer in the community include reducible load, interruptible load, transferable load and distributed energy storage, modeled as follows:
Figure FDA0003870070690000026
Figure FDA0003870070690000027
Figure FDA0003870070690000028
Figure FDA0003870070690000029
Figure FDA00038700706900000210
Figure FDA00038700706900000211
Figure FDA00038700706900000212
Figure FDA0003870070690000031
in the formula (I), the compound is shown in the specification,
Figure FDA0003870070690000032
representing the set of users i in community n,
Figure FDA0003870070690000033
for the run-time of the entire market,
Figure FDA0003870070690000034
in order to load the total amount of resources for flexibility,
Figure FDA0003870070690000035
respectively, user i can reduce load, interrupt load and transfer load at time t,
Figure FDA0003870070690000036
the amount of load over the total period of time required for the transferable load,
Figure FDA0003870070690000037
for the rigid load of user i at time t,
Figure FDA0003870070690000038
for the payload of user i at time t,
Figure FDA00038700706900000315
charging power for user i at time t, P i cha The rated charging power for the stored energy of user i,
Figure FDA00038700706900000316
charging of stored energy for user iThe status of the mobile station is,
Figure FDA00038700706900000317
discharge power, P, for user i at time t i dis The rated discharge power of the stored energy for user i,
Figure FDA00038700706900000318
is the discharge state of the stored energy of user i,
Figure FDA0003870070690000039
for the energy storage capacity of the user i at the time t,
Figure FDA00038700706900000310
the charging and discharging coefficients of the stored energy of the user i are respectively, delta t is the charging and discharging time step length of the stored energy,
Figure FDA00038700706900000311
the lower limit and the upper limit of the energy storage capacity of the user i.
5. The bipartite graph matching method for community-level P2P energy trading taking social influence into account of claim 4, wherein in addition to the operational constraints of the distributed power sources of energy producers and consumers, the intra-community trading needs to satisfy the following power balance constraints:
Figure FDA00038700706900000312
Figure FDA00038700706900000313
in the formula (I), the compound is shown in the specification,
Figure FDA00038700706900000314
for the nth community's purchased power to the upper grid at time t,
Figure FDA00038700706900000319
for the photovoltaic power generation power of the user i at the time t,
Figure FDA00038700706900000320
transacting electric energy for P2P of user i at time t when
Figure FDA00038700706900000321
Positive indicates that user i buys power at time t when
Figure FDA00038700706900000322
A negative indicates that user i sells at time t.
6. The bipartite graph matching method for community-level P2P energy trading taking social influence into account of claim 1, wherein the process of obtaining the energy trading model inside each community-level microgrid comprises:
modeling an end-to-end transaction process within each community as a Markov process, wherein the Markov observations are set as follows:
Figure FDA0003870070690000041
Figure FDA0003870070690000042
Figure FDA0003870070690000043
in the formula (I), the compound is shown in the specification,
Figure FDA0003870070690000044
is an observation set on the power generation side, and t representsThe index between the two or more of the data blocks,
Figure FDA0003870070690000045
representing the P2P transaction electricity price of user i at time t,
Figure FDA0003870070690000046
representing the photovoltaic power generation power of the user i at the time t,
Figure FDA0003870070690000047
a value representing the stiffness load of the user i,
Figure FDA0003870070690000048
indicating the energy storage capacity of the user i at the time t,
Figure FDA0003870070690000049
for a set of social network observations within a community,
Figure FDA00038700706900000410
respectively represents the operation difficulty, the communication degree, the information depth, the social influence degree and the influence of the friends of the transaction behavior, the higher the numerical value is, the greater the influence on the transaction result is,
Figure FDA00038700706900000411
is an observation set of the whole market;
the Markov action set is built up as follows:
Figure FDA00038700706900000412
Figure FDA00038700706900000413
Figure FDA00038700706900000414
Figure FDA00038700706900000415
Figure FDA00038700706900000416
Figure FDA00038700706900000417
in the formula (I), the compound is shown in the specification,
Figure FDA00038700706900000418
in order to be able to reduce the upper limit of the load,
Figure FDA00038700706900000419
an upper limit for the interruptible load is,
Figure FDA00038700706900000420
is an upper limit of the transferable load,
Figure FDA00038700706900000421
is a Markov action value;
after the markov action set and the markov observation set are obtained, the state transition process of each individual can be updated according to the action, and the updated state transition is as follows:
Figure FDA0003870070690000051
Figure FDA0003870070690000052
Figure FDA0003870070690000053
Figure FDA0003870070690000054
Figure FDA0003870070690000055
7. the bipartite graph matching method for community-level P2P energy trading with social influence taken into account according to claim 6, wherein after obtaining the state transition, a reward function corresponding to the state transition is defined as follows:
Figure FDA0003870070690000056
Figure FDA0003870070690000057
Figure FDA0003870070690000058
Figure FDA0003870070690000059
Figure FDA00038700706900000510
in the formula (I), the compound is shown in the specification,
Figure FDA00038700706900000511
respectively represent the electricity selling income and the electricity purchasing utility,
formula (II)
Figure FDA00038700706900000512
Psychological rewards to keep energy producers and consumers in the same community behaving consistently and the cost of using and maintaining distributed power sources,
Figure FDA00038700706900000513
meaning that the payload is averaged,
Figure FDA00038700706900000514
and respectively representing the maintenance cost of photovoltaic and distributed energy storage, and obtaining a return function value by taking the opposite number.
8. The bipartite graph matching method for community-level P2P energy trading taking social influence into account of claim 1, wherein the matching between communities comprises:
the community P2P matching problem is written as follows:
Figure FDA0003870070690000061
Figure FDA0003870070690000062
Figure FDA0003870070690000063
Figure FDA0003870070690000064
Figure FDA0003870070690000065
Figure FDA0003870070690000066
Figure FDA0003870070690000067
Figure FDA0003870070690000068
in the formula (I), the compound is shown in the specification,
Figure FDA0003870070690000069
the price for community n to sell electricity to community m,
Figure FDA00038700706900000610
the amount of electricity sold to community m for community n,
Figure FDA00038700706900000611
active loss, p, for the entire distribution network transaction t A vector of contribution powers is formed for each community,
Figure FDA00038700706900000612
for lagrange dual variables corresponding to active balancing,
Figure FDA00038700706900000613
reactive loss, q, for entire distribution network transactions t Is a vector formed by the reactive output power of each community,
Figure FDA00038700706900000614
for corresponding reactive power balanceLagrange dual variable, p min,t ,p max,t Respectively the active output lower limit and the active output upper limit of each community,
Figure FDA00038700706900000615
for corresponding Lagrangian dual variables, q min,t ,q max,t Respectively the lower limit and the upper limit of reactive power output of each community,
Figure FDA00038700706900000616
for corresponding Lagrangian dual variables, v min,t ,v max,t Respectively corresponding to the lower limit and the upper limit of the voltage amplitude for each node,
Figure FDA00038700706900000617
for the corresponding Lagrangian dual variable, | | s start,t || 2 Is the square of the complex power amplitude at the beginning of the branch, | s end,t || 2 Is the square of the complex power amplitude at the end of the branch, (S) t ) 2 The upper limit of the complex power amplitude of the whole branch circuit is set;
in order to minimize the influence of transactions on a power distribution network when P2P transactions are matched among communities, the P2P matching problem of the communities is linearized, and then a Lagrangian function is solved as follows:
Figure FDA0003870070690000071
Figure FDA0003870070690000072
Figure FDA0003870070690000073
Figure FDA0003870070690000074
Figure FDA0003870070690000075
Figure FDA0003870070690000076
in the formula, pi t Representing the weight of the transaction for community n to community m,
Figure FDA0003870070690000077
respectively, showing the effects of energy, loss, node voltage and line blocking on the network.
9. An apparatus, comprising:
one or more processors;
a memory for storing one or more programs;
when executed by one or more of the processors, cause the one or more processors to implement a bipartite graph matching method for community-level P2P energy trading with consideration of social influences as claimed in any one of claims 1-8.
10. A storage medium containing computer executable instructions which when executed by a computer processor are adapted to perform a bipartite graph matching method for community-level P2P energy trading taking social influence into account as claimed in any of claims 1-8.
CN202211193861.7A 2022-09-28 2022-09-28 Community-level P2P energy transaction bipartite graph matching method considering social influence Pending CN115660841A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211193861.7A CN115660841A (en) 2022-09-28 2022-09-28 Community-level P2P energy transaction bipartite graph matching method considering social influence

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211193861.7A CN115660841A (en) 2022-09-28 2022-09-28 Community-level P2P energy transaction bipartite graph matching method considering social influence

Publications (1)

Publication Number Publication Date
CN115660841A true CN115660841A (en) 2023-01-31

Family

ID=84985522

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211193861.7A Pending CN115660841A (en) 2022-09-28 2022-09-28 Community-level P2P energy transaction bipartite graph matching method considering social influence

Country Status (1)

Country Link
CN (1) CN115660841A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116934115A (en) * 2023-07-18 2023-10-24 天津大学 Real-time end-to-end energy transaction method considering time-varying virtual energy storage

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116934115A (en) * 2023-07-18 2023-10-24 天津大学 Real-time end-to-end energy transaction method considering time-varying virtual energy storage
CN116934115B (en) * 2023-07-18 2024-04-09 天津大学 Real-time end-to-end energy transaction method considering time-varying virtual energy storage

Similar Documents

Publication Publication Date Title
Zhou et al. Optimal scheduling of virtual power plant with battery degradation cost
Qiu et al. Optimal scheduling of distributed energy resources as a virtual power plant in a transactive energy framework
Qiu et al. Distributed generation and energy storage system planning for a distribution system operator
Luo et al. Short‐term operational planning framework for virtual power plants with high renewable penetrations
Chen et al. Research on day-ahead transactions between multi-microgrid based on cooperative game model
Lahon et al. Energy management of cooperative microgrids with high‐penetration renewables
Lahon et al. Risk‐based coalition of cooperative microgrids in electricity market environment
Zhou et al. Four‐level robust model for a virtual power plant in energy and reserve markets
Li et al. General Nash bargaining based direct P2P energy trading among prosumers under multiple uncertainties
CN111612248A (en) Power distribution network side source-load coordination method and system
Li et al. Co‐optimisation model for the long‐term design and decision making in community level cloud energy storage system
Vakili et al. Interconnected microgrids: Optimal energy scheduling based on a game‐theoretic approach
Sun et al. Bi-level model for integrated energy service providers in joint electricity and carbon P2P market
Nojavan et al. Optimal energy management of compressed air energy storage in day‐ahead and real‐time energy markets
Pan et al. Optimal planning of solar PV and battery storage with energy management systems for Time‐of‐Use and flat electricity tariffs
CN115660841A (en) Community-level P2P energy transaction bipartite graph matching method considering social influence
Peng et al. Review on bidding strategies for renewable energy power producers participating in electricity spot markets
Nagill et al. Feasibility analysis of heterogeneous energy storage technology for cloud energy storage with distributed generation
Zhaoan et al. Power charging management strategy for electric vehicles based on a Stackelberg game
Safari et al. DeepResTrade: A Peer-to-Peer LSTM-Decision Tree-Based Price Prediction and Blockchain-Enhanced Trading System for Renewable Energy Decentralized Markets
Qiu et al. Local integrated energy system operational optimization considering multi‐type uncertainties: A reinforcement learning approach based on improved TD3 algorithm
CN112865101B (en) Linear transaction method considering uncertainty of output of renewable energy
CN116432862A (en) Multi-main-body game optimization method and device for renewable energy micro-grid
CN116402223A (en) Cooperative scheduling method, system and equipment for power distribution network
CN115940274A (en) Optical storage system configuration method, device and medium

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