A demand response method and system based on dynamic scheduling of power reduction

By obtaining and analyzing the contextual information of environmental characteristics and subsidy prices, calculating consumer power reduction potential, forming a subset of consumers and determining the optimal subsidy price, the shortcomings of dynamic scheduling of consumers' power reduction in the prior art are solved, and the implementation of Dynamic scheduling and supply and demand balance of power systems.

CN114329904BActive Publication Date: 2024-11-15NANJING UNIV OF POSTS & TELECOMM
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202111462701.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-12-02
Publication Date
2024-11-15
Estimated Expiration
2041-12-02

AI Technical Summary

Technical Problem

The existing technology fails to effectively consider the dynamic changes in the potential of subsidy prices on consumers' power reduction, and the lack of learning context information, resulting in insufficient demand response solutions for consumers' power reduction dynamic dispatch.

Method used

By obtaining the context feature vector and consumer preference vector estimates of environmental features and alternative subsidy prices, calculate the upper confidence interval of the estimated consumer power reduction potential, form a subset of consumers, and establish power operations with the goal of cost minimization Business cost model, dynamically determine the optimal subsidy price and consumers participating in demand response.

Benefits of technology

It has achieved a more accurate description of consumers' potential for power reduction, avoided traditional forced power outage measures, promoted the supply and demand balance of power systems, and overcome the disadvantages of assuming that consumers' potential for power reduction is fixed.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114329904B_ABST
    Figure CN114329904B_ABST
Patent Text Reader

Abstract

The present invention discloses a demand response method and system based on dynamic scheduling of power reduction, which belongs to the field of power system, and comprises: obtaining a context feature vector including context features of environmental features and alternative subsidy prices and an estimated value of a consumer's preference vector for the context feature; calculating an estimated value of the consumer's power reduction potential according to the context feature vector and the estimated value of the preference vector, thereby calculating an upper confidence interval of the estimated value of the consumer's power reduction potential; arranging the upper confidence interval in descending order, and forming a consumer subset with consumers corresponding to the first K digits; solving a pre-established power operator cost model with the goal of minimizing cost, and obtaining an optimal subsidy price among alternative subsidy prices, and all consumers in the consumer subset corresponding to the optimal subsidy price are consumers participating in demand response; taking into account the dynamic changes of the estimated value of the consumer's power reduction potential, and solving the technical problem of dynamic scheduling of consumer power reduction.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a demand response method and system based on dynamic dispatching of power reduction, belonging to the field of power systems. Background Art

[0002] With the development of communication technology and intelligent monitoring instruments, demand response based on flexible two-way interaction between power operators and consumers has become a key technology for smart grid reform. Traditionally, when power system capacity is short or reliability is threatened, rigid measures such as forced power outages are often taken, which seriously affect the normal production and life of consumers. In previous studies, the fact that consumers' potential to reduce electricity consumption changes dynamically with respect to contextual information composed of different environmental factors and subsidy prices is often ignored. These studies generally assume that consumers' potential to reduce electricity consumption is fixed and lack the dynamic impact of learning contextual information on consumers' potential to reduce electricity consumption. On this basis, no clear demand response scheme based on dynamic scheduling of demand-side electricity reduction based on electricity reduction has been proposed.

[0003] In view of this, it is an important challenge to learn the impact of environmental factors and subsidy prices on consumers' potential to reduce electricity consumption and to achieve dynamic scheduling of demand-side electricity reduction based on this. Summary of the invention

[0004] The purpose of the present invention is to provide a demand response method and system based on dynamic scheduling of power reduction, which takes into account the dynamic changes of the estimated value of the consumer's power reduction potential. In continuous iteration, the preference vector estimate gradually converges to the real preference vector, thereby obtaining a more accurate characterization of the consumer's power reduction potential, overcoming the drawback of the prior art that lacks consideration of contextual information and assumes that the consumer's power reduction potential is fixed, and solves the technical problem of dynamic scheduling of consumer power reduction.

[0005] To achieve the above objectives, the present invention is implemented by adopting the following technical solutions:

[0006] In a first aspect, the present invention provides a demand response method based on dynamic scheduling of power reduction, comprising:

[0007] Obtaining a context feature vector including context features of environmental features and alternative subsidy prices and an estimated value of a consumer's preference vector for the context features;

[0008] Calculating the estimated value of the consumer's power reduction potential based on the context feature vector and the preference vector estimate, thereby calculating the upper confidence interval of the estimated value of the consumer's power reduction potential;

[0009] Arrange the upper confidence intervals in descending order, and form a consumer subset with consumers corresponding to the first K positions;

[0010] The pre-established power operator cost model is solved with the goal of minimizing cost to obtain the optimal subsidy price among the alternative subsidy prices. All consumers in the consumer subset corresponding to the optimal subsidy price are consumers participating in demand response.

[0011] Combined with the first aspect, further, in the first round of demand response, the preference vector estimate is obtained by the following method:

[0012]

[0013] A i =I 4×4

[0014] b i =0 4×1

[0015] in, is the estimated value of the preference vector in the first round of demand response, A i is the initialization matrix of consumer i, I 4×4 is the 4-dimensional identity matrix, b i is the initialization vector of consumer i, 0 4×1 is a 4-dimensional 0 vector.

[0016] In combination with the first aspect, further, in the hth round of demand response, the context feature vector is obtained by the following method:

[0017] x h =(e h ,c h )

[0018] Among them, x h is the context feature vector, e h is the environmental feature vector, c h The alternative subsidy price for each unit of electricity cut.

[0019] In combination with the first aspect, further, the environmental feature vector includes information on temperature, date, and whether it is a working day.

[0020] Combined with the first aspect, further, in the h+1th round of demand response, the preference vector estimate is obtained by the following method:

[0021]

[0022]

[0023]

[0024] in, is the estimated value of the preference vector of consumer i in the h+1th round of demand response, A i,h+1 is the matrix of consumer i in the h+1th round of demand response, A i,h is the matrix of consumer i in the h-th round of demand response, is the context feature vector corresponding to the optimal subsidy price, T represents the transpose of the vector, and b i,h+1 is the vector of consumer i in the h+1th round of demand response, b i,h is the vector of consumer i in the hth round of demand response, Δd i,h is the actual power reduction of consumer i in the hth round of demand response.

[0025] In combination with the first aspect, further, the estimated value of the consumer's power reduction potential is calculated according to the context feature vector and the preference vector estimate:

[0026]

[0027] in, is the estimated potential value of power reduction of consumer i in the hth round of demand response, (e h ,c h ) T is the transpose of the context feature vector, is the estimated value of the preference vector of consumer i in the h-th round of demand response;

[0028] The upper confidence interval of the estimated consumer power reduction potential is calculated as:

[0029]

[0030] Among them, U i,h is the upper confidence interval, is the estimated potential value of power reduction of consumer i, represents the uncertainty of the estimated value of the power reduction potential of consumer i, and σ is the trade-off parameter.

[0031] In combination with the first aspect, further, the power operator cost model is established by the following method: the power operator cost is modeled as

[0032] Where S(G) = α(G) 2 +β(G)+γ, α, β, γ are known cost parameters for power operators to purchase electricity from power stations, is the amount of electricity required on the demand side, For the consumer subset, is the estimated potential value of power reduction of consumer i in the hth round of demand response, c h The alternative subsidy price for each unit of electricity cut.

[0033] In a second aspect, the present invention further provides a demand response system based on dynamic scheduling of power reduction, comprising:

[0034] Acquisition module: used for acquiring context feature vectors of context features including environmental features and alternative subsidy prices and estimated values ​​of consumer preference vectors for the context features;

[0035] An upper confidence interval calculation module: used to calculate an estimated value of the consumer's power reduction potential based on the context feature vector and the preference vector estimate, thereby calculating an upper confidence interval of the consumer's power reduction potential estimate;

[0036] Consumer subset acquisition module: used to sort the upper confidence interval in descending order, and form a consumer subset with consumers corresponding to the first K positions;

[0037] Demand response module: used to establish a cost model for power operators, solve the model with the goal of minimizing costs, and obtain the optimal subsidy price among the alternative subsidy prices. All consumers in the consumer subset corresponding to the optimal subsidy price are consumers participating in demand response.

[0038] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0039] The present invention provides a demand response method and system based on dynamic scheduling of power reduction. The method calculates the estimated value of the consumer's power reduction potential according to the context feature vector and the preference vector estimate, thereby treating the consumer's power reduction as a virtual controllable resource. The context features in the context feature vector include environmental features and alternative subsidy prices, so the obtained estimated value of the power reduction potential changes dynamically with the context features, and the dynamic influence of environmental factors and subsidy prices on the consumer's power reduction potential is learned online; each consumer maintains a preference vector of the degree of preference for the context features, and the preference vector estimate gradually converges to the real preference vector in continuous iterations, thereby obtaining a more accurate characterization of the consumer's power reduction potential; the pre-established power operator cost model is solved with the goal of minimizing cost, and the optimal subsidy price and consumers participating in the demand response are dynamically determined by optimizing modeling and solving, thereby overcoming the disadvantage of the prior art that the context information is not considered and the consumer's power reduction potential is assumed to be fixed, and the technical problem of dynamic scheduling of consumer (demand side) power reduction is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a flow chart of a demand response method based on dynamic scheduling of power reduction provided by an embodiment of the invention;

[0041] Figure 2 It is a schematic diagram of an implementation scenario provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and are not intended to limit the protection scope of the present invention.

[0043] Example 1

[0044] like Figure 1 As shown, an embodiment of the present invention provides a demand response method based on dynamic scheduling of power reduction, including:

[0045] S1. Obtain a context feature vector including context features of environmental features and alternative subsidy prices and an estimated value of a consumer's preference vector for the context features.

[0046] The total number of rounds H of demand response is preset, the number of consumers K selected from all N consumers to participate in demand response is set each time, and K≤N, and the trade-off parameter σ is set.

[0047] Context features include environmental features and alternative subsidy prices. Environmental features include temperature, date, and whether it is a working day. The above environmental factors are expressed as an environmental feature vector, denoted as e h The above alternative subsidy price is the alternative subsidy price for each unit of electricity reduction, denoted as c h , c min ≤c h ≤c max , c min is the minimum value of the alternative subsidy price, c max is the maximum value of the alternative subsidy prices.

[0048] The environmental feature vector and the alternative subsidy price together form a 4-dimensional context feature vector. In the h-th round of demand response, the context feature vector is obtained by the following method: h =(e h ,c h ).

[0049] Each consumer i maintains a preference vector for the degree of preference for contextual features, denoted as θ i , i=1,…,N.

[0050] In the first round of demand response, the preference vector estimate is obtained by the following method:

[0051]

[0052] A i =I 4×4

[0053] b i =04×1

[0054] in, is the estimated value of the preference vector in the first round of demand response, A i is the initialization matrix of consumer i, I 4×4 is the 4-dimensional identity matrix, b i is the initialization vector of consumer i, 0 4×1 is a 4-dimensional 0 vector.

[0055] The estimated value of the preference vector in the h+1 round of demand response is obtained by the following method: the actual power reduction of consumer i in the previous round of demand response is obtained, and the matrix and vector of each consumer participating in the previous round of demand response are updated accordingly. The estimated value of the preference vector in the h+1 round of demand response is then calculated by the above matrix and vector:

[0056]

[0057]

[0058]

[0059] in, is the estimated value of the preference vector of consumer i in the h+1th round of demand response, A i,h+1 is the matrix of consumer i in the h+1th round of demand response, A i,h is the matrix of consumer i in the h-th round of demand response, is the context feature vector corresponding to the optimal subsidy price, T represents the transpose of the vector, and b i,h+1 is the vector of consumer i in the h+1th round of demand response, b i,h is the vector of consumer i in the hth round of demand response, Δd i,h is the actual power reduction of consumer i in the hth round of demand response.

[0060] S2. Calculate the estimated value of the consumer's power reduction potential based on the context feature vector and the estimated value of the preference vector, thereby calculating an upper confidence interval of the estimated value of the consumer's power reduction potential.

[0061] In round h of demand response:

[0062]

[0063] in, is the estimated potential value of power reduction of consumer i in the hth round of demand response, (e h ,c h ) T is the transpose of the context feature vector, is the estimated value of the preference vector of consumer i in the h-th round of demand response.

[0064] In order to obtain the long-term overall benefits of demand response, the utilization and exploration of consumers' electricity reduction potential are weighed during the online learning process. Utilization means selecting consumers with large estimated electricity reduction potential under known information to participate in demand response, and exploration means selecting consumers with large uncertainty in estimated electricity reduction potential to participate in demand response. The method of the present invention weighs utilization and exploration by calculating the upper confidence interval of the estimated electricity reduction potential of consumers.

[0065]

[0066] Among them, U i,h For the upper confidence interval, the first term As a utilization term, it represents the estimated value of the power reduction potential of consumer i. In the second term, As an exploration term, represents the uncertainty of the estimated value of the power reduction potential of consumer i, and σ is a trade-off parameter used to weigh the parameters between utilization and exploration.

[0067] S3. Arrange the upper confidence intervals in descending order, and form a consumer subset with consumers corresponding to the first K positions.

[0068] The upper confidence interval U of the estimated value of the power reduction potential of consumers i,h Sort in descending order and put the top K consumers in the set In the hth round of demand response, when the alternative price is c h When , the consumers corresponding to the top K in the upper confidence interval of the estimated value of the power reduction potential of consumers constitute the consumer subset,

[0069] S4. Solve the pre-established power operator cost model with the goal of minimizing cost to obtain the optimal subsidy price among the alternative subsidy prices. All consumers in the consumer subset corresponding to the optimal subsidy price are consumers participating in demand response.

[0070] Obtain the amount of electricity required by the demand side (consumers) Build a cost model for power operators.

[0071] like Figure 2 As shown in the figure, in order to meet the electricity demand side, on the one hand, the power operator purchases G units of electricity from the power station and pays the power station an amount S(G); on the other hand, the power operator provides subsidies (optimal subsidy price ) Reward consumers to reduce electricity consumption; Combining the above two aspects, the power operator cost is modeled as

[0072] in, Purchase from power stations for power operators The cost of a unit of electricity is Then S(G)=α(G) 2 +β(G)+γ, α, β, γ are known cost parameters for power operators to purchase electricity from power stations, is the amount of electricity required on the demand side, For the consumer subset, is the estimated potential value of power reduction of consumer i in the hth round of demand response, c h is the alternative subsidy price for each unit of electricity reduction, the second term Indicates that the subsidized price c h Rewarding consumers for cutting The cost per unit of electricity.

[0073] Taking minimizing the cost of power operators as the optimization goal, solve the power operator cost model and obtain the optimal subsidy price among the alternative subsidy prices. The subset of consumers corresponding to the optimal subsidy price All consumers in the are consumers participating in demand response.

[0074] gather It means that when the optimal subsidy price is When , the upper confidence interval of the estimated value of power reduction potential is ranked in the top K consumer subsets,

[0075] After the h-th round of demand response is completed, the actual power reduction of consumer i in the h-th round of demand response is obtained, and the matrix and vector of consumers participating in each h-th round of demand response are updated accordingly. The estimated value of the preference vector in the h+1th round of demand response is calculated through the above matrix and vector:

[0076]

[0077]

[0078]

[0079] in, is the estimated value of the preference vector of consumer i in the h+1th round of demand response, A i,h+1 is the matrix of consumer i in the h+1th round of demand response, A i,h is the matrix of consumer i in the h-th round of demand response, is the context feature vector corresponding to the optimal subsidy price, T represents the transpose of the vector, and b i,h+1 is the vector of consumer i in the h+1th round of demand response, b i,h is the vector of consumer i in the hth round of demand response, Δd i,h is the actual power reduction of consumer i in the hth round of demand response.

[0080] Repeat the above steps S1 to S4 until the preset H times of demand response are completed.

[0081] Demand response is when the effectiveness and reliability of the power system are threatened, and consumers change their inherent electricity consumption behavior patterns after receiving subsidy notices from power operators to induce power reduction or signals of rising electricity prices, thereby ensuring the stability of the power grid.

[0082] The embodiment of the present invention provides a demand response method based on dynamic scheduling of power reduction. In the demand response based on dynamic scheduling of power reduction on the demand side, the power reduction on the demand side is regarded as a virtual controllable resource. By learning the power reduction potential of demand-side consumers, consumers with power reduction potential are dynamically selected to participate in the demand response and subsidy prices are formulated in real time to reward consumers for power reduction behavior, thereby realizing dynamic scheduling of power reduction on the demand side, thereby avoiding various adverse effects brought about by traditional forced power outage measures and promoting the balance of supply and demand of the power system; the power reduction potential estimate of the consumer is calculated according to the context feature vector and the preference vector estimate, and the context feature The contextual features in the vector include environmental features and alternative subsidy prices, so the estimated value of the power reduction potential changes dynamically with the contextual features, and the dynamic impact of environmental factors and subsidy prices on consumers' power reduction potential is learned online; each consumer maintains a preference vector for the degree of preference for contextual features, and the estimated value of the preference vector gradually converges to the real preference vector in continuous iterations, thereby obtaining a more accurate characterization of consumers' power reduction potential; a power operator cost model is established, and the model is solved with the goal of minimizing cost to obtain the optimal subsidy price and the corresponding consumers participating in demand response, solving the technical problem of dynamic scheduling of consumer (demand side) power reduction.

[0083] A demand response method based on dynamic scheduling of power reduction provided by an embodiment of the present invention decomposes demand response into online learning and optimization problems. The online learning method studies the dynamic impact of environmental factors and subsidy prices on consumers' potential for power reduction, establishes and solves the power operator's cost model, and realizes demand response based on dynamic scheduling of power reduction.

[0084] In the online learning problem, the present invention learns online the dynamic impact of environmental factors and subsidy prices on consumers' potential to reduce electricity consumption. Each consumer maintains a preference vector for the degree of preference for contextual features. The estimated consumer preference vector gradually converges to the true preference vector during continuous iterations, thereby obtaining a more accurate characterization of consumers' potential to reduce electricity consumption.

[0085] At the same time, in order to obtain the long-term overall benefits of demand response, the utilization and exploration of consumers' electricity reduction potential are weighed in the online learning process. Utilization means selecting consumers with large estimated electricity reduction potential under known information to participate in demand response, and exploration means selecting consumers with large uncertainty in estimated electricity reduction potential to participate in demand response.

[0086] In the optimization problem, considering that power operators purchase electricity from power stations and reward consumers for reducing electricity consumption by providing subsidy prices on the other hand to meet the demand side electricity demand, the present invention models the operating costs of power operators on this basis, and then takes minimizing the operating costs of power operators as the optimization goal to solve for the optimal subsidy price.

[0087] In summary, the present invention obtains a more accurate characterization of consumers' potential for reducing electricity consumption through online learning, and dynamically determines the optimal subsidy price and consumers participating in demand response through optimization modeling, thereby overcoming the traditional drawback of assuming that consumers' potential for reducing electricity consumption is fixed due to the lack of consideration of contextual information, and solves the technical problem of dynamic scheduling of demand-side electricity reduction, thereby avoiding the various adverse effects brought about by traditional rigid measures such as forced power outages when the power system capacity is short.

[0088] Example 2

[0089] An embodiment of the present invention provides a demand response system based on dynamic scheduling of power reduction, including:

[0090] Acquisition module: used for acquiring context feature vectors of context features including environmental features and alternative subsidy prices and estimated values ​​of consumer preference vectors for the context features;

[0091] An upper confidence interval calculation module: used to calculate an estimated value of the consumer's power reduction potential based on the context feature vector and the preference vector estimate, thereby calculating an upper confidence interval of the consumer's power reduction potential estimate;

[0092] Consumer subset acquisition module: used to sort the upper confidence interval in descending order, and form a consumer subset with consumers corresponding to the first K positions;

[0093] Demand response module: used to establish a cost model for power operators, solve the model with the goal of minimizing costs, and obtain the optimal subsidy price among the alternative subsidy prices. All consumers in the consumer subset corresponding to the optimal subsidy price are consumers participating in demand response.

[0094] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0095] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0096] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0097] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0098] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A demand response method based on dynamic scheduling of power reduction, characterized in that: include: Obtaining a context feature vector including context features of environmental features and alternative subsidy prices and an estimated value of a consumer's preference vector for the context features; Calculating the estimated value of the consumer's power reduction potential based on the context feature vector and the preference vector estimate, thereby calculating the upper confidence interval of the estimated value of the consumer's power reduction potential; Arrange the upper confidence intervals in descending order, and form a consumer subset with consumers corresponding to the first K positions; Solve the pre-established power operator cost model with the goal of minimizing cost, and obtain the optimal subsidy price among the alternative subsidy prices. All consumers in the consumer subset corresponding to the optimal subsidy price are consumers participating in demand response. In the hth round of demand response, the context feature vector is obtained by the following method: x h =(e h ,c h ) Among them, x h is the context feature vector, e h is the environmental feature vector, c h is the alternative subsidy price for each unit of electricity cut; In the h+1th round of demand response, the preference vector estimate is obtained by the following method: in, is the estimated value of the preference vector of consumer i in the h+1th round of demand response, A i,h+1 is the matrix of consumer i in the h+1th round of demand response, A i,h is the matrix of consumer i in the h-th round of demand response, is the context feature vector corresponding to the optimal subsidy price, T represents the transpose of the vector, and b i,h+1 is the vector of consumer i in the h+1th round of demand response, b i,h is the vector of consumer i in the hth round of demand response, Δd i,h is the actual power reduction of consumer i in the hth round of demand response; Calculate the estimated value of the consumer's power reduction potential based on the context feature vector and the preference vector estimate: in, is the estimated potential value of power reduction of consumer i in the hth round of demand response, (e h ,c h ) T is the transpose of the context feature vector, is the estimated value of the preference vector of consumer i in the h-th round of demand response; The upper confidence interval of the estimated consumer power reduction potential is calculated as: Among them, U i,h is the upper confidence interval, is the estimated potential value of power reduction of consumer i, represents the uncertainty of the estimated value of the power reduction potential of consumer i, and σ is the trade-off parameter; The power operator cost model is established by the following method: The power operator cost is modeled as Where S(G) = α(G) 2 +β(G)+γ, α, β, γ are known cost parameters for power operators to purchase electricity from power stations, is the amount of electricity required on the demand side, For the consumer subset, is the estimated potential value of power reduction of consumer i in the hth round of demand response, c h The alternative subsidy price for each unit of electricity cut.

2. A demand response method based on dynamic scheduling of power reduction according to claim 1, characterized in that: In the first round of demand response, the preference vector estimate is obtained by the following method: A i =I 4×4 b i =0 4×1 in, is the estimated value of the preference vector in the first round of demand response, A i is the initialization matrix of consumer i, I 4×4 is the 4-dimensional identity matrix, b i is the initialization vector of consumer i, 0 4×1 is a 4-dimensional 0 vector.

3. A demand response method based on dynamic scheduling of power reduction according to claim 1, characterized in that: The environmental feature vector includes information on temperature, date, and whether it is a working day.

4. A demand response system based on dynamic scheduling of power reduction, characterized in that: include: Acquisition module: used for acquiring context feature vectors of context features including environmental features and alternative subsidy prices and estimated values ​​of consumer preference vectors for the context features; An upper confidence interval calculation module: used to calculate an estimated value of the consumer's power reduction potential based on the context feature vector and the preference vector estimate, thereby calculating an upper confidence interval of the consumer's power reduction potential estimate; Consumer subset acquisition module: used to sort the upper confidence interval in descending order, and form a consumer subset with consumers corresponding to the first K positions; Demand response module: used to solve the pre-established power operator cost model with the goal of minimizing cost, and obtain the optimal subsidy price among the alternative subsidy prices. All consumers in the consumer subset corresponding to the optimal subsidy price are consumers participating in demand response; In the hth round of demand response, the context feature vector is obtained by the following method: x h =(e h ,c h ) Among them, x h is the context feature vector, e h is the environmental feature vector, c h is the alternative subsidy price for each unit of electricity cut; In the h+1th round of demand response, the preference vector estimate is obtained by the following method: in, is the estimated value of the preference vector of consumer i in the h+1th round of demand response, A i,h+1 is the matrix of consumer i in the h+1th round of demand response, A i,h is the matrix of consumer i in the h-th round of demand response, is the context feature vector corresponding to the optimal subsidy price, T represents the transpose of the vector, and b i,h+1 is the vector of consumer i in the h+1th round of demand response, b i,h is the vector of consumer i in the hth round of demand response, Δd i,h is the actual power reduction of consumer i in the hth round of demand response; Calculate the estimated value of the consumer's power reduction potential based on the context feature vector and the preference vector estimate: in, is the estimated potential value of power reduction of consumer i in the hth round of demand response, (e h ,c h ) T is the transpose of the context feature vector, is the estimated value of the preference vector of consumer i in the h-th round of demand response; The upper confidence interval of the estimated consumer power reduction potential is calculated as: Among them, U i,h is the upper confidence interval, is the estimated potential value of power reduction of consumer i, represents the uncertainty of the estimated value of the power reduction potential of consumer i, and σ is the trade-off parameter; The power operator cost model is established by the following method: The power operator cost is modeled as Where S(G) = α(G) 2 +β(G)+γ, α, β, γ are known cost parameters for power operators to purchase electricity from power stations, is the amount of electricity required on the demand side, For the consumer subset, is the estimated potential value of power reduction of consumer i in the hth round of demand response, c h The alternative subsidy price for each unit of electricity cut.

Citation Information

Patent Citations

  • Industrial park comprehensive demand response scheduling method based on multi-energy cooperation

    CN110649594A

  • Integrated energy system scheduling method and device considering demand response and coupling degree

    CN113128775A

  • System and method of sensitivity-driven pricing service for non-stationary demand management

    US20160189175A1