CN109636515A - A kind of sale of electricity quotient intelligent agent Bidding system and device - Google Patents

A kind of sale of electricity quotient intelligent agent Bidding system and device Download PDF

Info

Publication number
CN109636515A
CN109636515A CN201811448384.8A CN201811448384A CN109636515A CN 109636515 A CN109636515 A CN 109636515A CN 201811448384 A CN201811448384 A CN 201811448384A CN 109636515 A CN109636515 A CN 109636515A
Authority
CN
China
Prior art keywords
bidding
bidding strategy
strategy
bidding strategies
coefficient
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
CN201811448384.8A
Other languages
Chinese (zh)
Inventor
王高琴
张鹏程
魏路平
肖艳炜
王颖
张凯锋
郑亚先
史新红
朱炳铨
郭艳敏
邵平
程海花
龙苏岩
陈爱林
徐骏
吕建虎
叶飞
曾丹
黄春波
杨辰星
冯凯
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
State Grid Zhejiang Electric Power Co Ltd
China Electric Power Research Institute Co Ltd CEPRI
State Grid Shanghai Electric Power Co Ltd
Original Assignee
Southeast University
State Grid Zhejiang Electric Power Co Ltd
China Electric Power Research Institute Co Ltd CEPRI
State Grid Shanghai Electric Power Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Southeast University, State Grid Zhejiang Electric Power Co Ltd, China Electric Power Research Institute Co Ltd CEPRI, State Grid Shanghai Electric Power Co Ltd filed Critical Southeast University
Priority to CN201811448384.8A priority Critical patent/CN109636515A/en
Publication of CN109636515A publication Critical patent/CN109636515A/en
Pending legal-status Critical Current

Links

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
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0611—Request for offers or quotes
    • 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)
  • Economics (AREA)
  • Engineering & Computer Science (AREA)
  • Marketing (AREA)
  • Health & Medical Sciences (AREA)
  • Finance (AREA)
  • Strategic Management (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Accounting & Taxation (AREA)
  • Theoretical Computer Science (AREA)
  • Development Economics (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Primary Health Care (AREA)
  • Tourism & Hospitality (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The present invention relates to a kind of sale of electricity quotient intelligent agent Bidding system and devices, which comprises S1. selects bidding strategies according to the select probability of bidding strategies each in bidding strategies set；S2. according to by the select probability of tendency each bidding strategies of coefficient update of selection bidding strategies corresponding bid income and unselected bidding strategies；S3. Optimal Bidding Strategies are obtained according to the select probability of bidding strategies each after update.Technical solution provided by the invention is able to reflect the decision behavior for concentrating different type sale of electricity company in trade at competitive price, reflects the Decision-making of Bidding behavior of different type sale of electricity quotient, further embodies the decision predisposition of different sale of electricity quotient in real market.

Description

A kind of sale of electricity quotient intelligent agent Bidding system and device
Technical field
The present invention relates to electricity market fields, and in particular to a kind of sale of electricity quotient intelligent agent Bidding system and device.
Background technique
With the propulsion of power market reform, on the basis of Generation Side competition, the competition of sale of electricity side is also being graduallyed relax control, greatly Amount sale of electricity company participates in market in succession.As new main market players, the bid of sale of electricity company bids behavior to entire electric power Risk management, market mode design, the trading rules formulation in market etc. propose some new challenges, and future will also more show It writes.Carry out power market simulation research, it is necessary first to the report for being how sale of electricity quotient complexity in effectively simulation market of solution Valence decision behavior, establish the sale of electricity Shang dynasty reason offer decision-making models, embody sale of electricity quotient bid behavior for market operation process shadow Ring effect.
In recent years, the behavior simulation of bidding for being concentrated mainly on generation side market member based on the modeling technique of agency is ground Study carefully, by market member being modeled as having the computer intelligence of certain learning decision ability act on behalf of, for particular market, according to Market rules construct market member bidding decision simulation model, by emulation experiment, assess market operating status, examine market rule Reasonability then.In comparison, it is less to act on behalf of tactics research for the behavior of bidding of sales market member, and it is specific to be based primarily upon certain Market mode study the bidding decision method based on forecasted electricity market price and Monte Carlo randomized optimization process, there are a large amount of hypothesis Premise, and for market member bid the factors such as target, market mode variation adaptability it is not strong.
Summary of the invention
In view of the deficiencies of the prior art, the purpose of the invention is to describe the sale of electricity quotient of differentiation target in real market Bid behavior is able to reflect and concentrates in trade at competitive price the present invention provides a kind of sale of electricity quotient intelligent agent Bidding system and device The decision behavior of different type sale of electricity company, reflects the Decision-making of Bidding behavior of different type sale of electricity quotient, and simulates as far as possible Actual market.
The purpose of the present invention is adopt the following technical solutions realization:
A kind of sale of electricity quotient intelligent agent Bidding system, it is improved in that the described method includes:
S1. bidding strategies are selected according to the select probability of bidding strategies each in bidding strategies set；
S2. according to each by the tendency coefficient update of selection bidding strategies corresponding bid income and unselected bidding strategies The select probability of bidding strategies；
S3. Optimal Bidding Strategies are obtained according to the select probability of bidding strategies each after update.
Preferably, in the step S1, the initial selected probability of each bidding strategies in bidding strategies set, as the following formula really It is fixed:
In above formula, p1It (s) is the initial selected probability of s-th of bidding strategies in bidding strategies set, s ∈ [1, M], M are Bidding strategies sum.
Preferably, the step S1, comprising:
According to the select probability of each bidding strategies in the bidding strategies set, using roulette algorithm from bidding strategies collection Bidding strategies are selected in conjunction.
Preferably, the step S2, comprising:
The corresponding receipts of bidding of the bidding strategies are determined according to by the corresponding e-commerce operation target value of selling of selection bidding strategies Benefit；
It is updated according to the income of bidding by the tendency coefficient of selection bidding strategies；
The tendency coefficient of unselected bidding strategies is updated according to the forgetting factor of unselected bidding strategies；
According to the select probability of the corresponding bidding strategies of tendency coefficient update of each bidding strategies.
Further, described to determine the bidding strategies according to by the corresponding e-commerce operation target value of selling of selection bidding strategies Corresponding income of bidding, comprising:
It determines as the following formula and determines that the bidding strategies are corresponding according to by the corresponding e-commerce operation target of selling of selection bidding strategies Income of bidding:
In above formula, R is the corresponding income of bidding of the bidding strategies, αnTo be described by the corresponding sale of electricity of selection bidding strategies Quotient n-th runs target value, θnFor the weight for being runed target value by the corresponding sale of electricity quotient n-th of selection bidding strategies, βnIt is described The conversion coefficient of target value, n ∈ [1,9] are runed by the corresponding sale of electricity quotient n-th of selection bidding strategies.
Specifically, determining that the corresponding sale of electricity quotient first of the bidding strategies runs target value α as the following formula1:
α1=max [(psell*qsell)-(pclear*qclear)]
The second operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula2:
α2=max [qload*(pset-pclear)]
The third operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula3:
α3=max (qclear*pclear)
The 4th operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula4:
α4=max [(qsell*psell)-Δpenalty]
The 5th operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula5:
α5=max [(qsell*psell)-Δpenalty']
The 6th operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula6:
α6=maxqclear
The 7th operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula7:
α7=max (qclear-Δpenalty)
The 8th operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula8:
α8=max (qclear-Δpenalty')
The 9th operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula9:
α9=max (qclear-qclear')
In above formula, psellAnd qsellThe respectively price and electricity of the sale of electricity contract of sale of electricity Shang Yuqi user signing, pclear For the corresponding cleaing price of the bidding strategies, qclearFor the corresponding acceptance of the bid electricity of the bidding strategies, psetFor listed power price, qloadFor load prediction electricity, ΔpenaltyPenalty term when for practical loss of capital more than the receptible loss of capital amount of money, Δpenalty' be It not can guarantee the penalty term when condition got a profit substantially, qclear' be rival conclusion of the business electricity；
Wherein, penalty term Δ when practical loss of capital is more than the receptible loss of capital amount of money is determined as the following formulapenalty:
Penalty term Δ when not can guarantee the condition got a profit substantially is determined as the following formulapenalty':
In above formula, δpenaltyIt is penalty factor, πlossFor the receptible loss of capital amount of money.
Further, the income of bidding according to is updated by the tendency coefficient of selection bidding strategies, comprising:
It is updated as the following formula by the tendency coefficient of selection bidding strategies:
qt+1(m)=(1-r) qt(m)+(1-e)R
In above formula, qt+1(m) for, by the tendency coefficient of selection bidding strategies, r is in the t+1 times iteration bidding strategies set Forgetting factor, e are empirical parameter, qtIt (m) is the tendency coefficient of m-th of bidding strategies in the t times iteration bidding strategies set, t ∈ [1, T], T are iteration total degree, and m ∈ [1, M], M are bidding strategies sum.
Further, the forgetting factor according to unselected bidding strategies updates the tendency of unselected bidding strategies Coefficient, comprising:
The tendency coefficient of unselected bidding strategies is updated as the following formula:
In above formula, x ∈ [1, M] and x ≠ m, m ∈ [1, M], M are bidding strategies sum, and m is the bidding strategies selected, x For non-selected bidding strategies,；qt+1It (x) is the tendency system of x-th of bidding strategies in the t+1 times iteration bidding strategies set Number, qtIt (x) is the tendency coefficient of x-th of bidding strategies in the t times iteration bidding strategies set.
Further, the select probability of the corresponding bidding strategies of tendency coefficient update according to each bidding strategies, comprising:
The select probability p of s-th of bidding strategies in the t+1 times iteration bidding strategies set is determined as the following formulat+1(s):
In above formula, s ∈ [1, M], M are bidding strategies sum；qt+1It (s) is s in the t+1 times iteration bidding strategies set The tendency coefficient of a bidding strategies, c are cooling ratio；
Wherein, cooling ratio c is determined as the following formula:
In above formula, qtIt (s) is the tendency coefficient of m-th of bidding strategies in the t times iteration bidding strategies set, ε is greater than 0 Real number.
Preferably, the step S3, comprising:
If there are the select probabilities of bidding strategies to be greater than 0.99 in the bidding strategies set, which is optimal Bidding strategies；Otherwise, the step S1 is returned.
A kind of sale of electricity quotient intelligent agent bid device, it is improved in that described device includes:
Selecting unit, for selecting bidding strategies according to the select probability of bidding strategies each in bidding strategies set；
Updating unit, for according to by the tendency of selection bidding strategies corresponding bid income and unselected bidding strategies The select probability of each bidding strategies of coefficient update；
Acquiring unit, for obtaining Optimal Bidding Strategies according to the select probability of bidding strategies each after update.
Compared with the immediate prior art, the invention has the benefit that
Technical solution provided by the invention, it is competing by being selected according to the select probability of bidding strategies each in bidding strategies set Valence strategy is respectively bidded according to by the tendency coefficient update of selection bidding strategies corresponding bid income and unselected bidding strategies The select probability of strategy obtains Optimal Bidding Strategies according to the select probability of bidding strategies each after update, it is competing to be able to reflect concentration The decision behavior of different type sale of electricity company, reflects the Decision-making of Bidding behavior of different type sale of electricity quotient, further body in valence transaction The decision predisposition of different sale of electricity quotient in real market is showed.
Detailed description of the invention
Fig. 1 is a kind of flow chart of sale of electricity quotient intelligent agent Bidding system in the embodiment of the present invention；
Fig. 2 is a kind of structural schematic diagram of sale of electricity quotient intelligent agent bid device in the embodiment of the present invention.
Specific embodiment
Specific embodiments of the present invention will be described in further detail with reference to the accompanying drawing.
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art All other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
The present invention provides a kind of sale of electricity quotient intelligent agent Bidding systems, as shown in Figure 1, which comprises
101. selecting bidding strategies according to the select probability of bidding strategies each in bidding strategies set；
102. according to by the tendency coefficient update of selection bidding strategies corresponding bid income and unselected bidding strategies The select probability of each bidding strategies；
103. obtaining Optimal Bidding Strategies according to the select probability of bidding strategies each after update.
Further, in the step 101, the initial selected probability of each bidding strategies in bidding strategies set, as the following formula It determines:
In above formula, p1It (s) is the initial selected probability of s-th of bidding strategies in bidding strategies set, s ∈ [1, M], M are Bidding strategies sum.
Further, the step 101, comprising:
According to the select probability of each bidding strategies in the bidding strategies set, using roulette algorithm from bidding strategies collection Bidding strategies are selected in conjunction.
Further, the step 102, comprising:
The corresponding receipts of bidding of the bidding strategies are determined according to by the corresponding e-commerce operation target value of selling of selection bidding strategies Benefit；
It is updated according to the income of bidding by the tendency coefficient of selection bidding strategies；
The tendency coefficient of unselected bidding strategies is updated according to the forgetting factor of unselected bidding strategies；
According to the select probability of the corresponding bidding strategies of tendency coefficient update of each bidding strategies.
Specifically, described determine the bidding strategies pair according to by the corresponding e-commerce operation target value of selling of selection bidding strategies The income of bidding answered, comprising:
It determines as the following formula and determines that the bidding strategies are corresponding according to by the corresponding e-commerce operation target of selling of selection bidding strategies Income of bidding:
In above formula, R is the corresponding income of bidding of the bidding strategies, αnTo be described by the corresponding sale of electricity of selection bidding strategies Quotient n-th runs target value, θnFor the weight for being runed target value by the corresponding sale of electricity quotient n-th of selection bidding strategies, βnIt is described The conversion coefficient of target value, n ∈ [1,9] are runed by the corresponding sale of electricity quotient n-th of selection bidding strategies.
Specifically, determining that the corresponding sale of electricity quotient first of the bidding strategies runs target value α as the following formula1:
α1=max [(psell*qsell)-(pclear*qclear)]
The second operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula2:
α2=max [qload*(pset-pclear)]
The third operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula3:
α3=max (qclear*pclear)
The 4th operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula4:
α4=max [(qsell*psell)-Δpenalty]
The 5th operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula5:
α5=max [(qsell*psell)-Δpenalty']
The 6th operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula6:
α6=maxqclear
The 7th operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula7:
α7=max (qclear-Δpenalty)
The 8th operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula8:
α8=max (qclear-Δpenalty')
The 9th operation target value α of the corresponding sale of electricity quotient of the bidding strategies is determined as the following formula9:
α9=max (qclear-qclear')
In above formula, psellAnd qsellThe respectively price and electricity of the sale of electricity contract of sale of electricity Shang Yuqi user signing, pclear For the corresponding cleaing price of the bidding strategies, qclearFor the corresponding acceptance of the bid electricity of the bidding strategies, psetFor listed power price, qloadFor load prediction electricity, ΔpenaltyPenalty term when for practical loss of capital more than the receptible loss of capital amount of money, Δpenalty' be It not can guarantee the penalty term when condition got a profit substantially, qclear' be rival conclusion of the business electricity；
Wherein, penalty term Δ when practical loss of capital is more than the receptible loss of capital amount of money is determined as the following formulapenalty:
Penalty term Δ when not can guarantee the condition got a profit substantially is determined as the following formulapenalty':
In above formula, δpenaltyIt is penalty factor, πlossFor the receptible loss of capital amount of money.
Specifically, the income of bidding according to is updated by the tendency coefficient of selection bidding strategies, comprising:
It is updated as the following formula by the tendency coefficient of selection bidding strategies:
qt+1(m)=(1-r) qt(m)+(1-e)R
In above formula, qt+1(m) for, by the tendency coefficient of selection bidding strategies, r is in the t+1 times iteration bidding strategies set Forgetting factor, e are empirical parameter, qtIt (m) is the tendency coefficient of m-th of bidding strategies in the t times iteration bidding strategies set, t ∈ [1, T], T are iteration total degree, and m ∈ [1, M], M are bidding strategies sum.
Specifically, the forgetting factor according to unselected bidding strategies updates the tendency system of unselected bidding strategies Number, comprising:
The tendency coefficient of unselected bidding strategies is updated as the following formula:
In above formula, x ∈ [1, M] and x ≠ m, m ∈ [1, M], M are bidding strategies sum, and m is the bidding strategies selected, x For non-selected bidding strategies,；qt+1It (x) is the tendency system of x-th of bidding strategies in the t+1 times iteration bidding strategies set Number, qtIt (x) is the tendency coefficient of x-th of bidding strategies in the t times iteration bidding strategies set.
Wherein, as t=1, qt(m) or qt(x) it is the initial tendency coefficient of each bidding strategies in bidding strategies set, enables Initial tendency coefficient is 6000, and the initial coefficient that is inclined to is an empirical parameter, guarantees that each strategy has positive choosing at the beginning Probability is selected so as to global convergence.
Specifically, the select probability of the corresponding bidding strategies of tendency coefficient update according to each bidding strategies, comprising:
The select probability p of s-th of bidding strategies in the t+1 times iteration bidding strategies set is determined as the following formulat+1(s):
In above formula, s ∈ [1, M], s both can be m, or x, M are bidding strategies sum；qt+1(s) it is the t+1 times The tendency coefficient of s-th of bidding strategies, c are cooling ratio in iteration bidding strategies set；
Wherein, cooling ratio c is determined as the following formula:
In above formula, qtIt (s) is the tendency coefficient of m-th of bidding strategies in the t times iteration bidding strategies set, ε is greater than 0 Real number.
Further, the step 103, comprising:
If there are the select probabilities of bidding strategies to be greater than 0.99 in the bidding strategies set, which is optimal Bidding strategies；Otherwise, the step S101 is returned.
The present invention also provides a kind of sale of electricity quotient intelligent agent bid devices, as shown in Fig. 2, described device includes:
Selecting unit, for selecting bidding strategies according to the select probability of bidding strategies each in bidding strategies set；
Updating unit, for according to by the tendency of selection bidding strategies corresponding bid income and unselected bidding strategies The select probability of each bidding strategies of coefficient update；
Acquiring unit, for obtaining Optimal Bidding Strategies according to the select probability of bidding strategies each after update.
Further, in the selecting unit, the initial selected probability of each bidding strategies in bidding strategies set, as the following formula It determines:
In above formula, p1It (s) is the initial selected probability of s-th of bidding strategies in bidding strategies set, s ∈ [1, M], M are Bidding strategies sum.
Further, the selecting unit, is used for:
According to the select probability of each bidding strategies in the bidding strategies set, using roulette algorithm from bidding strategies collection Bidding strategies are selected in conjunction.
Further, the updating unit, comprising:
Determining module, for determining the bidding strategies according to by the corresponding e-commerce operation target value of selling of selection bidding strategies Corresponding income of bidding；
First update module, for bidding according to, income is updated by the tendency coefficient of selection bidding strategies；
Second update module, for updating unselected bidding strategies according to the forgetting factor of unselected bidding strategies It is inclined to coefficient；
Third update module, for the select probability for being inclined to the corresponding bidding strategies of coefficient update according to each bidding strategies.
Specifically, the determining module, e-commerce operation is sold according to by selection bidding strategies are corresponding for determining as the following formula Target determines the corresponding income of bidding of the bidding strategies:
In above formula, R is the corresponding income of bidding of the bidding strategies, αnTo be described by the corresponding sale of electricity of selection bidding strategies Quotient n-th runs target value, θnFor the weight for being runed target value by the corresponding sale of electricity quotient n-th of selection bidding strategies, βnIt is described The conversion coefficient of target value, n ∈ [1,9] are runed by the corresponding sale of electricity quotient n-th of selection bidding strategies.
Specifically, the determining module, further includes:
First determines submodule, for determining that the corresponding sale of electricity quotient first of the bidding strategies runs target value as the following formula α1:
α1=max [(psell*qsell)-(pclear*qclear)]
Second determines submodule, for determining the second operation target value of the corresponding sale of electricity quotient of the bidding strategies as the following formula α2:
α2=max [qload*(pset-pclear)]
Third determines submodule, for determining that the third of the corresponding sale of electricity quotient of the bidding strategies runs target value as the following formula α3:
α3=max (qclear*pclear)
4th determines submodule, for determining the 4th operation target value of the corresponding sale of electricity quotient of the bidding strategies as the following formula α4:
α4=max [(qsell*psell)-Δpenalty]
5th determines submodule, for determining the 5th operation target value of the corresponding sale of electricity quotient of the bidding strategies as the following formula α5:
α5=max [(qsell*psell)-Δpenalty']
6th determines submodule, for determining the 6th operation target value of the corresponding sale of electricity quotient of the bidding strategies as the following formula α6:
α6=maxqclear
7th determines submodule, for determining the 7th operation target value of the corresponding sale of electricity quotient of the bidding strategies as the following formula α7:
α7=max (qclear-Δpenalty)
8th determines submodule, for determining the 8th operation target value of the corresponding sale of electricity quotient of the bidding strategies as the following formula α8:
α8=max (qclear-Δpenalty')
9th determines submodule, for determining the 9th operation target value of the corresponding sale of electricity quotient of the bidding strategies as the following formula α9:
α9=max (qclear-qclear')
In above formula, psellAnd qsellThe respectively price and electricity of the sale of electricity contract of sale of electricity Shang Yuqi user signing, pclear For the corresponding cleaing price of the bidding strategies, qclearFor the corresponding acceptance of the bid electricity of the bidding strategies, psetFor listed power price, qloadFor load prediction electricity, ΔpenaltyPenalty term when for practical loss of capital more than the receptible loss of capital amount of money, Δpenalty' be It not can guarantee the penalty term when condition got a profit substantially, qclear' be rival conclusion of the business electricity；
Tenth determines submodule, penalty term when for determining that practical loss of capital is more than the receptible loss of capital amount of money as the following formula Δpenalty:
11st determines submodule, for determining penalty term when not can guarantee the condition got a profit substantially as the following formula Δpenalty':
In above formula, δpenaltyIt is penalty factor, πlossFor the receptible loss of capital amount of money.
Specifically, first update module, for being updated as the following formula by the tendency coefficient of selection bidding strategies:
qt+1(m)=(1-r) qt(m)+(1-e)R
In above formula, qt+1(m) for, by the tendency coefficient of selection bidding strategies, r is in the t+1 times iteration bidding strategies set Forgetting factor, e are empirical parameter, qtIt (m) is the tendency coefficient of m-th of bidding strategies in the t times iteration bidding strategies set, t ∈ [1, T], T are iteration total degree, and m ∈ [1, M], M are bidding strategies sum.
Specifically, second update module is used to update the tendency coefficient of unselected bidding strategies as the following formula:
In above formula, x ∈ [1, M] and x ≠ m, m ∈ [1, M], M are bidding strategies sum, and m is the bidding strategies selected, x For non-selected bidding strategies,；qt+1It (x) is the tendency system of x-th of bidding strategies in the t+1 times iteration bidding strategies set Number, qtIt (x) is the tendency coefficient of x-th of bidding strategies in the t times iteration bidding strategies set.
Wherein, as t=1, qt(m) or qt(x) it is the initial tendency coefficient of each bidding strategies in bidding strategies set, enables Initial tendency coefficient is 6000, and the initial coefficient that is inclined to is an empirical parameter, guarantees that each strategy has positive choosing at the beginning Probability is selected so as to global convergence.
Specifically, the third update module for determine in the t+1 times iteration bidding strategies set as the following formula s-th it is competing The select probability p of valence strategyt+1(s):
In above formula, s ∈ [1, M], s both can be m, or x, M are bidding strategies sum；qt+1(s) it is the t+1 times The tendency coefficient of s-th of bidding strategies, c are cooling ratio in iteration bidding strategies set；
The third update module, further includes:
12nd determines submodule, for determining cooling ratio c as the following formula:
In above formula, qtIt (s) is the tendency coefficient of m-th of bidding strategies in the t times iteration bidding strategies set, ε is greater than 0 Real number.
Further, the acquiring unit, if for there are the select probabilities of bidding strategies in the bidding strategies set Greater than 0.99, then the bidding strategies are Optimal Bidding Strategies；Otherwise, selecting unit is returned.
It should be understood by those skilled in the art that, embodiments herein can provide as method, system or computer program Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the application Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the application, which can be used in one or more, The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces The form of product.
The application is referring to method, the process of equipment (system) and computer program product according to the embodiment of the present application Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates, Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one The step of function of being specified in a box or multiple boxes.
Finally it should be noted that: the above embodiments are merely illustrative of the technical scheme of the present invention and are not intended to be limiting thereof, to the greatest extent Invention is explained in detail referring to above-described embodiment for pipe, it should be understood by those ordinary skilled in the art that: still It can be with modifications or equivalent substitutions are made to specific embodiments of the invention, and without departing from any of spirit and scope of the invention Modification or equivalent replacement, should all cover within the scope of the claims of the present invention.

Claims (11)

1.一种售电商智能代理投标方法，其特征在于，所述方法包括：1. a method for selling electricity business intelligence agent bidding, it is characterised in that the method comprises: S1.根据竞价策略集合中各竞价策略的选择概率选择竞价策略；S1. Select a bidding strategy according to the selection probability of each bidding strategy in the bidding strategy set; S2.根据被选择竞价策略对应的竞价收益和未被选择竞价策略的倾向系数更新各竞价策略的选择概率；S2. Update the selection probability of each bidding strategy according to the bidding income corresponding to the selected bidding strategy and the propensity coefficient of the unselected bidding strategy; S3.根据更新后各竞价策略的选择概率获取最优竞价策略。S3. Obtain the optimal bidding strategy according to the selection probability of each bidding strategy after the update. 2.如权利要求1所述的方法，其特征在于，所述步骤S1中，竞价策略集合中各竞价策略的初始选择概率，按下式确定：2. The method according to claim 1, wherein, in the step S1, the initial selection probability of each bidding strategy in the bidding strategy set is determined as follows: 上式中，p1(s)为竞价策略集合中第s个竞价策略的初始选择概率，s∈[1,M]，M为竞价策略总数。In the above formula, p 1 (s) is the initial selection probability of the s-th bidding strategy in the bidding strategy set, s∈[1,M], M is the total number of bidding strategies. 3.如权利要求1所述的方法，其特征在于，所述步骤S1，包括：3. The method of claim 1, wherein the step S1 comprises: 根据所述竞价策略集合中各竞价策略的选择概率，利用轮盘赌算法从竞价策略集合中选择竞价策略。According to the selection probability of each bidding strategy in the bidding strategy set, a roulette algorithm is used to select a bidding strategy from the bidding strategy set. 4.如权利要求1所述的方法，其特征在于，所述步骤S2，包括：4. The method of claim 1, wherein the step S2 comprises: 根据被选择竞价策略对应的售电商运营目标值确定所述被选择竞价策略的竞价收益；Determine the bidding revenue of the selected bidding strategy according to the target value of the e-commerce retailer operation corresponding to the selected bidding strategy; 根据所述被选择竞价策略的竞价收益更新被选择竞价策略的倾向系数；Update the propensity coefficient of the selected bidding strategy according to the bidding revenue of the selected bidding strategy; 根据未被选择竞价策略的遗忘因子更新未被选择竞价策略的倾向系数；Update the propensity coefficient of the unselected bidding strategy according to the forgetting factor of the unselected bidding strategy; 根据各竞价策略的倾向系数更新相应竞价策略的选择概率。The selection probability of the corresponding bidding strategy is updated according to the propensity coefficient of each bidding strategy. 5.如权利要求4所述的方法，其特征在于，所述根据被选择竞价策略对应的售电商运营目标值确定所述竞价策略对应的竞价收益，包括：5. The method according to claim 4, wherein determining the bidding revenue corresponding to the bidding strategy according to the target value of the e-commerce sales business operation corresponding to the selected bidding strategy, comprising: 根据被选择竞价策略对应的售电商运营目标，按下式确定所述被选择竞价策略对应的竞价收益：According to the operation target of the e-commerce retailer corresponding to the selected bidding strategy, the bidding revenue corresponding to the selected bidding strategy is determined as follows: 上式中，R为所述竞价策略对应的竞价收益，αn为所述被选择竞价策略对应的售电商第n运营目标值，θn为所述被选择竞价策略对应的售电商第n运营目标值的权重，βn为所述被选择竞价策略对应的售电商第n运营目标值的转换系数，n∈[1,9]。In the above formula, R is the bidding revenue corresponding to the bidding strategy, α n is the nth operation target value of the e-commerce retailer corresponding to the selected bidding strategy, and θ n is the e-commerce retailer's nth corresponding to the selected bidding strategy. The weight of n operation target value, β n is the conversion coefficient of the nth operation target value of the e-commerce retailer corresponding to the selected bidding strategy, n∈[1,9]. 6.如权利要求5所述的方法，其特征在于，按下式确定所述竞价策略对应的售电商第一运营目标值α1：6. The method according to claim 5, wherein the first operation target value α 1 of the electricity seller corresponding to the bidding strategy is determined as follows: α1＝max[(psell*qsell)-(pclear*qclear)]α 1 =max[(p sell *q sell )-(p clear *q clear )] 按下式确定所述竞价策略对应的售电商的第二运营目标值α2：The second operating target value α 2 of the e-commerce retailer corresponding to the bidding strategy is determined as follows: α2＝max[qload*(pset-pclear)]α 2 =max[q load *(p set -p clear )] 按下式确定所述竞价策略对应的售电商的第三运营目标值α3：The third operation target value α 3 of the e-commerce retailer corresponding to the bidding strategy is determined as follows: α3＝max(qclear*pclear)α 3 =max(q clear *p clear ) 按下式确定所述竞价策略对应的售电商的第四运营目标值α4：The fourth operation target value α 4 of the e-commerce retailer corresponding to the bidding strategy is determined as follows: α4＝max[(qsell*psell)-Δpenalty]α 4 =max[(q sell *p sell )-Δ penalty ] 按下式确定所述竞价策略对应的售电商的第五运营目标值α5：The fifth operation target value α 5 of the e-commerce retailer corresponding to the bidding strategy is determined as follows: α5＝max[(qsell*psell)-Δpenalty']α 5 =max[(q sell *p sell )-Δ penalty '] 按下式确定所述竞价策略对应的售电商的第六运营目标值α6：The sixth operation target value α 6 of the e-commerce retailer corresponding to the bidding strategy is determined as follows: α6＝maxqclear α 6 =maxq clear 按下式确定所述竞价策略对应的售电商的第七运营目标值α7：The seventh operational target value α 7 of the e-commerce retailer corresponding to the bidding strategy is determined as follows: α7＝max(qclear-Δpenalty)α 7 =max(q clear -Δ penalty ) 按下式确定所述竞价策略对应的售电商的第八运营目标值α8：The eighth operational target value α 8 of the e-commerce retailer corresponding to the bidding strategy is determined as follows: α8＝max(qclear-Δpenalty')α 8 =max(q clear -Δ penalty ') 按下式确定所述竞价策略对应的售电商的第九运营目标值α9：The ninth operation target value α 9 of the e-commerce retailer corresponding to the bidding strategy is determined as follows: α9＝max(qclear-qclear')α 9 =max(q clear -q clear ') 上式中，psell和qsell分别为售电商与其用户签订的售电合同的价格与电量，pclear为所述竞价策略对应的出清价格，qclear为所述竞价策略对应的中标电量，pset为目录电价，qload为负荷预测电量，Δpenalty为实际亏本超过能接受的亏本金额时的惩罚项，Δpenalty'为未能保证基本盈利的条件时的惩罚项，qclear'为竞争对手的成交电量；In the above formula, p sell and q sell are the price and electricity volume of the electricity sales contract signed by the e-commerce company and its users, respectively, p clear is the clearing price corresponding to the bidding strategy, and q clear is the bid-winning electricity volume corresponding to the bidding strategy. , p set is the catalog electricity price, q load is the load forecast electricity, Δ penalty is the penalty item when the actual loss exceeds the acceptable amount of loss, Δ penalty ' is the penalty item when the basic profit cannot be guaranteed, q clear ' is Competitor's transaction volume; 其中，按下式确定实际亏本超过能接受的亏本金额时的惩罚项Δpenalty：Among them, the penalty term Δpenalty when the actual loss exceeds the acceptable loss amount is determined as follows: 按下式确定未能保证基本盈利的条件时的惩罚项Δpenalty'：The penalty term Δpenalty ' when the condition of failing to guarantee basic profit is determined as follows: 上式中，δpenalty是惩罚因子，πloss为能接受的亏本金额。In the above formula, δ penalty is the penalty factor, and π loss is the acceptable loss amount. 7.如权利要求4所述的方法，其特征在于，所述根据所述竞价收益更新被选择竞价策略的倾向系数，包括：7. The method of claim 4, wherein the updating the propensity coefficient of the selected bidding strategy according to the bidding revenue comprises: 按下式更新被选择竞价策略的倾向系数：Update the propensity coefficient of the selected bidding strategy as follows: qt+1(m)＝(1-r)qt(m)+(1-e)Rq t+1 (m)=(1-r)q t (m)+(1-e)R 上式中，qt+1(m)为第t+1次迭代竞价策略集合中被选择竞价策略的倾向系数，r为遗忘因子，e为经验参数，qt(m)为第t次迭代竞价策略集合中第m个竞价策略的倾向系数，t∈[1,T]，T为迭代总次数，m∈[1,M]，M为竞价策略总数。In the above formula, q t+1 (m) is the propensity coefficient of the bidding strategy selected in the t+1 th iteration bidding strategy set, r is the forgetting factor, e is the empirical parameter, and q t (m) is the t th iteration. The propensity coefficient of the mth bidding strategy in the bidding strategy set, t∈[1,T], T is the total number of iterations, m∈[1,M], M is the total number of bidding strategies. 8.如权利要求4所述的方法，其特征在于，所述根据未被选择竞价策略的遗忘因子更新未被选择竞价策略的倾向系数，包括：8. The method of claim 4, wherein the updating the propensity coefficient of the unselected bidding strategy according to the forgetting factor of the unselected bidding strategy, comprising: 按下式更新未被选择竞价策略的倾向系数：Update the propensity coefficient for the unselected bidding strategy as follows: 上式中，x∈[1,M]且x≠m，m∈[1,M]，m为被选择的竞价策略，x为未被选择的竞价策略，M为竞价策略总数；qt+1(x)为第t+1次迭代竞价策略集合中第x个竞价策略的倾向系数，qt(x)为第t次迭代竞价策略集合中第x个竞价策略的倾向系数。In the above formula, x∈[1,M] and x≠m, m∈[1,M], m is the selected bidding strategy, x is the unselected bidding strategy, M is the total number of bidding strategies; q t+ 1 (x) is the propensity coefficient of the x-th bidding strategy in the t+1-th iteration bidding strategy set, and q t (x) is the propensity coefficient of the x-th bidding strategy in the t-th iteration bidding strategy set. 9.如权利要求4所述的方法，其特征在于，所述根据各竞价策略的倾向系数更新相应竞价策略的选择概率，包括：9. The method of claim 4, wherein the updating the selection probability of the corresponding bidding strategy according to the propensity coefficient of each bidding strategy comprises: 按下式确定第t+1次迭代竞价策略集合中第s个竞价策略的选择概率pt+1(s)：The selection probability p t+1 (s) of the s-th bidding strategy in the t+1-th iteration bidding strategy set is determined as follows: 上式中，s∈[1,M]，M为竞价策略总数；qt+1(s)为第t+1次迭代竞价策略集合中第s个竞价策略的倾向系数，c为冷却系数；In the above formula, s∈[1,M], M is the total number of bidding strategies; q t+1 (s) is the propensity coefficient of the s-th bidding strategy in the t+1-th iteration bidding strategy set, and c is the cooling coefficient; 其中，按下式确定冷却系数c：Among them, the cooling coefficient c is determined as follows: 上式中，qt(s)为第t次迭代竞价策略集合中第m个竞价策略的倾向系数，ε为大于0的实数。In the above formula, q t (s) is the propensity coefficient of the m-th bidding strategy in the t-th iteration bidding strategy set, and ε is a real number greater than 0. 10.如权利要求1所述的方法，其特征在于，所述步骤S3，包括：10. The method of claim 1, wherein the step S3 comprises: 若所述竞价策略集合中存在竞价策略的选择概率大于0.99，则该竞价策略为最优竞价策略；否则，返回所述步骤S1。If the selection probability of a bidding strategy in the bidding strategy set is greater than 0.99, the bidding strategy is the optimal bidding strategy; otherwise, return to step S1. 11.一种售电商智能代理投标装置，其特征在于，所述装置包括：11. An intelligent proxy bidding device for e-commerce sales, characterized in that the device comprises: 选择单元，用于根据竞价策略集合中各竞价策略的选择概率选择竞价策略；The selection unit is used to select the bidding strategy according to the selection probability of each bidding strategy in the bidding strategy set; 更新单元，用于根据被选择竞价策略对应的竞价收益和未被选择竞价策略的倾向系数更新各竞价策略的选择概率；an update unit, configured to update the selection probability of each bidding strategy according to the bidding income corresponding to the selected bidding strategy and the propensity coefficient of the unselected bidding strategy; 获取单元，用于根据更新后各竞价策略的选择概率获取最优竞价策略。The obtaining unit is used to obtain the optimal bidding strategy according to the selection probability of each bidding strategy after updating.
CN201811448384.8A 2018-11-30 2018-11-30 A kind of sale of electricity quotient intelligent agent Bidding system and device Pending CN109636515A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811448384.8A CN109636515A (en) 2018-11-30 2018-11-30 A kind of sale of electricity quotient intelligent agent Bidding system and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811448384.8A CN109636515A (en) 2018-11-30 2018-11-30 A kind of sale of electricity quotient intelligent agent Bidding system and device

Publications (1)

Publication Number Publication Date
CN109636515A true CN109636515A (en) 2019-04-16

Family

ID=66069975

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811448384.8A Pending CN109636515A (en) 2018-11-30 2018-11-30 A kind of sale of electricity quotient intelligent agent Bidding system and device

Country Status (1)

Country Link
CN (1) CN109636515A (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030233315A1 (en) * 2002-02-26 2003-12-18 Byde Andrew Robert Bidding in multiple on-line auctions
CN107644370A (en) * 2017-09-29 2018-01-30 中国电力科学研究院 Price competing method and system are brought in a kind of self-reinforcing study together

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030233315A1 (en) * 2002-02-26 2003-12-18 Byde Andrew Robert Bidding in multiple on-line auctions
CN107644370A (en) * 2017-09-29 2018-01-30 中国电力科学研究院 Price competing method and system are brought in a kind of self-reinforcing study together

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
王野平等: ""智能代理模拟在电力市场中的运用"", 《华南理工大学学报（自然科学版）》 *

Similar Documents

Publication Publication Date Title
Gilli et al. A data-driven optimization heuristic for downside risk minimization
Bichler et al. A computational analysis of linear price iterative combinatorial auction formats
CN112651770B (en) Optimization method and system for load declaration of electricity sellers in electricity spot market
Gimenez-Funes et al. Designing bidding strategies for trading agents in electronic auctions
Mikkilä et al. Empirical deep hedging
Ghaderi et al. Behavioral simulation and optimization of generation companies in electricity markets by fuzzy cognitive map
Banal-Estanol et al. Behavioural simulations in spot electricity markets
An et al. Deep reinforcement learning for quantitative trading: Challenges and opportunities
Jia et al. A deep reinforcement learning bidding algorithm on electricity market
CN111861648B (en) Price negotiating strategy model learning method based on simulation training
Pippas et al. The evolution of reinforcement learning in quantitative finance
Nanduri et al. Economic impact assessment and operational decision making in emission and transmission constrained electricity markets
Chandlekar et al. Multi-unit double auctions: equilibrium analysis and bidding strategy using DDPG in smart-grids
Mascioli et al. A Financial Market Simulation Environment for Trading Agents Using Deep Reinforcement Learning
CN109636515A (en) A kind of sale of electricity quotient intelligent agent Bidding system and device
CN116611499A (en) Method and apparatus for training reinforcement learning system for automatic bidding
Alaei et al. Anticipatory random walks: A dynamic model of crowdfunding
Pinto et al. Multi-Agent Negotiation for Coalition Formation and Management in Electricity
Lin et al. Investigating the robustness and generalizability of deep reinforcement learning based optimal trade execution systems
Pardo et al. A Modular Framework for Reinforcement Learning Optimal Execution
Delage et al. Multi-Agent based simulation of FOREX exchange market
de Meer Pardo et al. A Modular Framework for Reinforcement Learning Optimal Execution.
Lu A Deep Reinforcement Learning-Enabled Portfolio Management System with Quarterly Stock Re-Selection Based on Financial Statements
Zubrii et al. Forecasting real option price model by means of evolutionary and genetic algorithms
Chowdhury et al. SPOT: Strategies for Power Trading in Wholesale Electricity Markets

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20190416

RJ01 Rejection of invention patent application after publication