A Carbon-Constrained Load-Side Virtual Power Plant Demand Response Control Method and System
By constructing a two-layer optimization and control model based on carbon emission flows, the problems of carbon responsibility sharing and coordinated carbon reduction in load-side demand response were solved, achieving proactive adjustment on the load side and overall carbon reduction effect.
Patent Information
- Application Number
- CN202611455654.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-18
- Publication Date
- 2026-11-17
AI Technical Summary
Under carbon-constrained scenarios, existing technologies suffer from several shortcomings in load-side demand response, including a lack of systematic design for carbon responsibility sharing mechanisms, limited carbon reduction methods, and an inadequate multi-entity collaborative carbon reduction mechanism involving sources, loads, and storage. Consequently, the overall carbon reduction efficiency is not fully realized.
By introducing carbon emission flow theory, defining a quantitative index of node carbon potential, constructing a two-layer optimization and control model under carbon constraints, calculating the carbon potential of load nodes in real time, generating time-of-use electricity price signals, and collaboratively optimizing electricity consumption plans, the demand response control of flexible load equipment can be realized.
It has achieved a scientific and unified allocation of carbon responsibility on the load side, enhanced the load side's proactive adjustment capabilities, promoted the consumption of new energy sources and improved energy efficiency, and significantly improved overall carbon reduction efficiency.
Smart Images

Figure CN123121693A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of power regulation technology, and specifically relates to a method and system for demand response regulation of a load-side virtual power plant based on carbon constraints. Background Technology
[0002] The power industry faces pressure to reduce emissions, while also needing to address issues such as difficulties in grid integration and consumption of new energy sources and widening peak-valley load differences. Load-side resources have become key to improving system regulation capabilities and carbon reduction efficiency.
[0003] In carbon-constrained scenarios, load-side demand response needs to coordinate carbon reduction with the economical and reliable operation of the system. Current research is in the exploratory stage and faces several key challenges that have not yet been resolved.
[0004] Currently, the main approach is to tap the load-side regulation potential through virtual power plant load-side demand response technology. This involves guiding users to adjust their electricity consumption behavior through price signals or incentives, aggregating distributed adjustable resources such as air conditioners, energy storage, and charging piles, and building a load management system that enables full-domain perception, real-time sharing, and precise control. This allows for real-time and friendly interaction between electricity users and the power grid, balancing user experience with the system's ability to cope with both source and load randomness, and helping the system reduce carbon emissions.
[0005] The existing scheme still has obvious shortcomings: First, the carbon responsibility sharing mechanism lacks a systematic design and there is no scientific and unified sharing standard and implementation framework; second, the carbon reduction means are singular, focusing on load reduction and not exploring multiple paths such as flexible adjustment and energy efficiency improvement; third, the multi-entity collaborative carbon reduction mechanism of source-load-storage is not sound, the linkage between each link is weak, and the overall carbon reduction efficiency has not been fully released. Summary of the Invention
[0006] The purpose of this invention is to provide a method and system for demand response control of a load-side virtual power plant based on carbon constraints, so as to solve the problem of insufficient release of carbon reduction efficiency.
[0007] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a load-side virtual power plant demand response control method based on carbon constraints, comprising: Based on carbon emission flows, the carbon potential of each load node is calculated in real time, whereby the carbon potential represents the equivalent generation-side carbon emission value caused by consuming a unit of electricity at that node. Based on the carbon potential of each load node, a two-layer optimization control model under carbon constraints is constructed. The upper-layer model aims to maximize the overall operating efficiency of the power system and generates time-of-use electricity price signals. The lower-layer model aims to minimize the sum of electricity costs and carbon emission costs for load-side users. The carbon potential of the nodes is used as the input of the two-layer optimization control model, and the output electricity consumption plan is optimized in conjunction with the time-of-use electricity price signals. Using electricity consumption plans as the target for demand response on the load side, demand response control is implemented on the flexible load equipment aggregated by the virtual power plant during preset high carbon emission periods or power shortage periods.
[0008] Furthermore, the real-time calculation of the carbon potential of each load node based on carbon emission flows includes: Carbon potential at power system nodes, in symbol This indicates that a node is defined. The nodal carbon potentials are as follows:
[0009] In the formula: For nodes carbon potential, There are power flow inflow nodes in the connected branches. The set of all branches, Subscript for branch road; branch road The meritorious trend; branch road carbon flux density; branch road carbon flow rate; The nodal carbon potential and the branch carbon flux density have the same dimensions and are numerically equal to the total inflow to the node. The carbon flow density of the branch is a weighted average of the active power flow, which is the equivalent carbon emission value on the generation side caused by a unit of electricity consumption at the node.
[0010] Furthermore, the construction of a carbon-constrained two-layer optimization control model based on the carbon potential of each load node includes: The objective of the first-level objective function is to minimize economic cost.
[0011] In the formula: Where N is the economic cost and N is the total number of generating units; Let be the power generation cost function of the i-th generating unit; For the output of the i-th unit, For network loss costs, The system network loss power, and satisfies = , This is the grid loss electricity price coefficient; The objective of the second-order objective function is to minimize the cost of carbon emissions. The second level aims to minimize carbon emission costs on the load side by adjusting the load curve through demand response to achieve low-carbon optimization. The objective function is as follows:
[0012] In the formula: For carbon trading prices; The scheduling period; For the first Carbon emission intensity of the unit; for Carbon emission reductions achieved through real-time demand response, and real-time load adjustment satisfy . , for The average carbon emission intensity of the system at any given time; for Real-time load adjustment; The average reduction in carbon emissions per unit load adjustment; Demand response constraints: min≤ ≤ max in min、 `max` represents the minimum and maximum adjustable load, respectively. Daily electricity consumption conservation constraint: .
[0013] Furthermore, the first-order objective function constraints are as follows: Power balance constraints:
[0014] In the formula: This represents the original load power. Unit output constraints: P g,i,min ≤P g,i ≤P g,i,max P g,i,min ,P g,i,max The first Minimum and maximum output of the unit; Unit ramp-up constraints: -R g,i,down ≤P g,i,t - P g,i,t-1 ≤R g,i,up In the formula: R g,i,down, R g,i,up These represent the unit's downward and upward ramp rates, respectively. t This refers to the scheduling time.
[0015] Furthermore, the upper-level model aims to maximize the overall operational efficiency of the power system and generates time-of-use pricing signals; the lower-level model aims to minimize the sum of electricity costs and carbon emission costs for load-side users, and outputs electricity consumption plans through coordinated optimization of time-of-use pricing signals and node carbon potential, including: The upper layer aims to maximize the overall benefits of the system by setting time-of-use electricity prices and carbon price signals; the lower layer aims to minimize the sum of users' electricity costs and carbon emission costs by optimizing electricity consumption plans; and through the synergistic effect of time-of-use electricity prices and node carbon potential, the responsibility for carbon emissions from the power supply side is transferred to the load side. The time-of-use pricing optimization expression of the upper-level model on the system side is as follows:
[0016] In the formula: The scheduling period; This represents the total number of generating units. For the first Taiwanese crew The power generation cost function at any given time; For the first Taiwanese crew Efforts made at all times; M This represents the number of load nodes. for Time-of-use pricing; for node The load power is adjusted continuously, and the constraints of the upper-level model are the same as those mentioned above. Low-carbon response optimization on the load side of the lower-level model: With the goal of achieving optimal low carbon emissions on the lower load side, the model expression is as follows:
[0017] In the formula: For carbon trading prices; for node Carbon potential at all times; No. Nodes The carbon cost at any given moment.
[0018] Furthermore, the constraints for the upper-level and lower-level models are as follows: Point load upper and lower limit constraints: P D,n,min ≤PD,n,t ≤P D,n,max P D,n,min 、P D,n,max These represent the upper and lower limits of the n-node load and the nodal carbon potential correlation constraints, respectively.
[0019] For the first Carbon emission intensity of the Taiwanese unit For the unit For nodes The power allocation coefficient satisfies .
[0020] Furthermore, the constructed carbon-constrained bi-layer optimization control model employs a two-stage coordinated optimization strategy involving the day-ahead and intraday phases: In the aforementioned day-ahead phase, based on the forecast data of new energy output and load demand, and guided by low-carbon electricity prices, a basic demand response plan for the load side is formulated. During the intraday phase, based on real-time data of renewable energy output and load demand, and using ultra-carbon incentives as a correction signal, the demand response baseline plan is adjusted on a rolling basis, including: Demand response plans are developed based on forecast data, with the goal of minimizing the sum of users' electricity costs and carbon costs.
[0021] In the formula: It is the sum of the user's electricity cost and carbon cost. for Low-carbon electricity prices as of [date]; for Predicted load adjustment plan for the current time period; for Predicted carbon emission intensity up to the current date; Adjustable load adjustment constraints: ≤
[0022] for The load adjustment amount can be shifted at any time. The maximum load that can be moved; Load adjustment boundary constraints: PD,min≤ ≤PD,max PD,min,PD,max are respectively Node load upper and lower limits; Based on real-time data-driven rolling adjustments, and with the goal of minimizing ultra-carbon incentive costs, we address fluctuations in new energy sources and load:
[0023] In the formula: To reduce carbon incentive costs, for Intraday ultra-carbon incentive price; This represents the real-time load adjustment within the day. Quickly adjust resource response capability constraints: ≤ , To quickly adjust the maximum response capacity of resources; Boundary constraints after intraday load correction: PD,min≤ ≤PD,max Daily electricity consumption conservation constraint:
[0024] Adjustment range constraints: | |≤δ δ is the adjustment coefficient, which controls the intraday adjustment range.
[0025] Furthermore, the method of using electricity consumption plans as the demand response target on the load side, and performing demand response regulation on the flexible load equipment aggregated by the virtual power plant during preset high carbon emission periods or power shortage periods, includes: Before the demand response begins, the flexible load device is pre-adjusted to store cold or heat in the flexible load device or its medium. During demand response periods, the operating power of the flexible load equipment is reduced, and the stored cold or heat is released to meet load demands. After the demand response ends, the original operating parameters of the flexible load equipment are restored.
[0026] Secondly, the present invention provides a load-side virtual power plant demand response control system based on carbon constraints, comprising: The carbon potential acquisition module is used to calculate the carbon potential of each load node in real time based on the carbon emission flow. The carbon potential represents the equivalent generation-side carbon emission value caused by consuming a unit of electricity at that node. The model building module is used to construct a two-layer optimization control model under carbon constraints based on the carbon potential of each load node. The upper-layer model aims to maximize the overall operating efficiency of the power system and generates time-of-use electricity price signals. The lower-layer model aims to minimize the sum of electricity costs and carbon emission costs for load-side users and optimizes electricity consumption plans. The carbon emission costs include the carbon potential of the node and the carbon trading price. The control module is used to take the electricity consumption plan as the demand response target on the load side. During the preset high carbon emission period or power shortage period, it performs demand response on the aggregated flexible load, stores energy during the low carbon period before the response, releases the stored energy and reduces the real-time power during the response, and resumes normal operation after the response.
[0027] Thirdly, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the carbon-constrained load-side virtual power plant demand response control method.
[0028] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the carbon-constrained load-side virtual power plant demand response control method.
[0029] Compared with the prior art, the present invention has the following technical effects: This application introduces carbon emission flow theory, defines a quantitative index of nodal carbon potential, and constructs a carbon emission flow tracking and responsibility allocation model from the generation side to the load side. This model can calculate the generation-side carbon emissions corresponding to each unit of electricity consumed by each load node in real time and accurately. This provides a scientific, unified, and executable measurement framework for load-side participation in the carbon market and bearing carbon costs, achieving clear traceability and fair allocation of carbon responsibility in the power network.
[0030] This application proposes a load flexibility adjustment strategy based on dynamic carbon potential signals. By storing energy during periods of low carbon potential and reducing operating power and releasing energy during demand response periods of high carbon potential, it achieves optimized load shifting over time. While ensuring minimal impact on user comfort, it reduces peak carbon emissions, promotes the integration of renewable energy, and unlocks the deep potential for energy efficiency improvement.
[0031] This application constructs a two-layer optimization and control model with node carbon potential as the key coupling parameter. The upper-layer model aims to maximize overall efficiency by coordinating the optimization of time-of-use electricity prices and carbon price signals. After receiving electricity price and carbon potential signals, the lower-layer model autonomously optimizes electricity consumption plans to minimize the sum of electricity costs and carbon costs. The load side transforms from a passive recipient of instructions to an active regulator, while the virtual power plant acts as an aggregator and coordinator, integrating distributed resources to respond to system demands. This establishes a coordinated chain of source-load-storage, systematically improving overall carbon reduction efficiency. Attached Figure Description
[0032] Figure 1 is a schematic diagram of the carbon flow of the adjacent branch of node n in this invention.
[0033] Figure 2This is a typical daily carbon emission factor curve of the present invention.
[0034] Figure 3 This is a diagram showing the air conditioning power load during the pre-response preparation stage of this invention.
[0035] Figure 4 This is a diagram showing the air conditioning power load during the response phase of this invention.
[0036] Figure 5 This is a graph showing the air conditioning power load during the response termination phase of this invention.
[0037] Figure 6 This is a graph showing the change in carbon emissions during the response process of this invention.
[0038] Figure 7 This is a load curve diagram of the flexible response adjustment of an air conditioner on a certain day according to the present invention.
[0039] Figure 8 This is a flowchart of the present invention. Detailed Implementation
[0040] The present invention will be further described below with reference to the accompanying drawings: Example 1, please refer to Figure 8 This invention provides a load-side virtual power plant demand response control method based on carbon constraints, comprising: S1, based on carbon emission flow, calculate the carbon potential of each load node in real time, whereby the carbon potential represents the equivalent generation-side carbon emission value caused by consuming a unit of electricity at that node. The carbon potential of each load node is calculated in real time. Carbon emission flow is the theoretical basis, viewing electricity flow as the carrier of carbon emissions within the power grid; node carbon potential is a quantitative indicator, defined as the equivalent generation-side carbon emission value resulting from consuming a unit of electricity at that node. Relying on the energy management system of the power dispatch system, the topology of the power grid, active power flow data of each branch, and the real-time carbon emission intensity of each generator are acquired in real time. According to the formula:
[0041] In the formula: For nodes carbon potential, There are power flow inflow nodes in the connected branches. The set of all branches, Subscript for branch road; branch road The meritorious trend; branch road carbon flux density; branch road carbon flow rate; Calculate the inflow node The carbon potential of a node is calculated in real time by summing the carbon flow rates of all its branches and dividing by the sum of the total active power flowing into that node. This calculation can be embedded into the carbon flow calculation module of the power grid dispatching platform for rolling updates at set intervals.
[0042] By precisely tracing carbon emissions from the power generation side to each electricity consumption node according to the power flow path, each load is given a dynamic and quantifiable carbon identity card (i.e., carbon potential). This enables the scientific measurement of the previously difficult-to-define and indirect responsibility for carbon emissions from electricity consumption, providing a unique, fair, and executable data foundation for incorporating carbon costs into user-side decision-making, and achieving transparent and precise allocation of carbon responsibility from the power generation side to the load side.
[0043] S2, based on the carbon potential of each load node, constructs a two-layer optimization control model under carbon constraints. The upper-layer model aims to maximize the overall operating efficiency of the power system and generates time-of-use electricity price signals. The lower-layer model aims to minimize the sum of electricity costs and carbon emission costs for load-side users and outputs electricity consumption plans through the coordinated optimization of time-of-use electricity price signals and the carbon potential of nodes. The upper layer is the system-side optimization model, aiming to minimize the difference between the total grid generation cost and electricity sales revenue, i.e., maximizing overall operational efficiency. The decision variable is the time-of-use (TOU) electricity price signal. The lower layer is the load-side (virtual power plant aggregator) optimization model, aiming to minimize the sum of the total electricity cost and carbon emission cost for a single user or aggregator, thus optimizing the electricity consumption plan. The system operator (upper layer) solves an optimization problem involving power balance and unit constraints based on predicted load, renewable energy output, and unit parameters, deriving and publishing the TOU electricity price and predicted nodal carbon potential for the next cycle. After receiving these price signals, the virtual power plant aggregator platform (lower layer) solves its own electricity cost minimization model based on the electricity consumption characteristics, adjustability potential, and carbon costs of the users it aggregates, forming the optimal electricity consumption plan or load adjustment scheme.
[0044] By leveraging nodal carbon potential, carbon emission constraints on the power supply side are transformed into economic signals directly perceptible to the load side, establishing a dual incentive / constraint mechanism of electricity price + carbon price. This incentivizes the load side to proactively shift its electricity consumption behavior from periods of high electricity price and high carbon potential to periods of low electricity price and low carbon potential in order to minimize total costs. This achieves collaborative optimization between power sources and loads, with carbon as the link and economics as the driving force, transforming the load side from a passive recipient of instructions to an active participant in system regulation and carbon reduction.
[0045] S3 uses the electricity consumption plan as the demand response target on the load side, and performs demand response control on the flexible load equipment aggregated by the virtual power plant during preset high carbon emission periods or power shortage periods.
[0046] The optimized power consumption plan output from step S2 is transformed into specific, executable flexible load control actions, triggered during periods of high carbon emissions or power shortages. Its core is time-shifted flexible adjustment based on load characteristics, rather than simple disconnection. Upon receiving the demand response target instruction generated by S2, the virtual power plant control platform sends instructions to the user's air conditioning system when carbon potential is low, appropriately reducing the chilled water outlet temperature setpoint to store cold water in the building's water system and main unit in advance. During the formal response period (high carbon potential period), the platform instructs the air conditioning unit to significantly reduce power or increase the outlet water temperature, relying on the stored cooling capacity to maintain a relatively stable indoor temperature. After the response ends, the system automatically resumes normal operation.
[0047] Through flexible regulation of pre-storage and release of energy, the load curve is optimized over time without reducing the total electricity consumption and service quality of users. It can reduce the net load and carbon emissions of the power grid during peak and high-carbon periods, and promote the consumption of renewable energy during off-peak periods.
[0048] Example 2: This invention provides a load-side virtual power plant demand response control method based on carbon constraints, comprising: The practical implementation of carbon constraints in demand response relies primarily on two key technologies: precise simulation and prediction of electricity and carbon emissions, and a carbon emission intensity assessment index system. These two technologies work together to form a complete technological chain from carbon emission monitoring and accounting to regulation, providing a systematic methodological basis and data foundation for the coordinated optimization of electricity and carbon emissions in demand response.
[0049] In power systems, different power generation technologies have different carbon emission characteristics, but the power flow and carbon emissions generated by various power plants are not different. Therefore, carbon emission flows have a simpler and more flexible application space in power systems, and it is also easier to construct a theoretical framework for carbon emission flows in power systems.
[0050] Carbon emission flows in a power system are essentially labeling each branch of the power flow with a carbon emission tag. Because of the dependency between carbon emission flows and power flow, it can be considered that in a power system, carbon emission flows originate from power plants (power plant nodes), enter the power system along with the power output fed into the grid, flow through the grid following the power flow, and ultimately flow into the consumer terminals (load nodes) on the user side. Carbon emissions can be simplified as carbon being transferred from electricity to users for consumption via carbon emission flows.
[0051] The carbon emission flow of the power system is a fundamental concept, and correspondingly, several new concepts and indicators will emerge, which form the theoretical basis for the analysis of carbon emission flows of the power system.
[0052] Branch carbon flow rate is the most fundamental physical quantity describing carbon flow, characterizing the magnitude of carbon flow in a branch, and is represented by the symbol [symbol missing]. Branch carbon flow is defined as the cumulative amount of carbon emissions corresponding to the carbon flow that travels through a branch with the tidal current within a given time period. The unit of carbon flow is the same as that of carbon emissions. Numerically, branch carbon flow is equal to the amount of carbon emissions generated during the power generation process to generate and maintain the tidal current of that branch within a given time period.
[0053] Branch carbon flow rate, defined as the amount of carbon flowing through a branch per unit time with the tidal current, is denoted by the symbol [symbol missing]. This means that it is numerically equal to the derivative of the branch carbon flow rate with respect to time:
[0054] The unit of branch carbon flow rate is generally . Obviously, branch carbon flow rate is the integral value of branch carbon flow rate over a given time period.
[0055] Carbon emission flows in power systems are dependent on power flow, therefore it is necessary to study carbon emission flows in conjunction with power flow. Considering that carbon emissions in power systems are mainly related to active power flow, to characterize the combined characteristics of the two, the ratio of carbon flow rate to active power flow in any branch of the power system is defined as the carbon flow density of that branch, denoted by the symbol [symbol missing]. express:
[0056] Based on the definitions of branch carbon flow and branch carbon flow density, it is easy to see that the unit of branch carbon flow density is the same as the unit of carbon emission intensity on the generation side, generally 1. Thus, in the power plant outgoing lines, the branch carbon flow density is equal to the carbon emission intensity of the generator unit, while in the lines entering the load terminal, the branch carbon flow density is equal to the carbon emission value on the generation side caused by the unit electricity consumption transmitted in the branch, which has a clear physical meaning.
[0057] Since both branch carbon flow rate and active power flow describe instantaneous values, branch carbon flow density varies with power flow in the system over time. For ease of description, the average branch carbon flow density over a given time period is further defined. Abbreviated as average carbon flux density:
[0058] The dimensions of the average branch carbon flow density are the same as those of the branch carbon flow density. Compared to the branch carbon flow density, the average branch carbon flow density is calculated from the cumulative amount, which is the ratio of carbon flow to electrical charge passing through the branch over a period of time.
[0059] Branch carbon flow density is used to describe the relationship between power flow and carbon flow on branches in a power system. Since power generation and consumption links in a power system exist in the form of nodes, it is necessary to define a physical quantity that describes the carbon emission intensity of nodes by carbon emission flow.
[0060] Therefore, the carbon potential at power system nodes is defined, using the symbol... This indicates that a node is defined. The nodal carbon potentials are as follows:
[0061] In the formula: There are power flow inflow nodes in the connected branches. The set of all branches, Subscript for branch road.
[0062] The nodal carbon potential and the branch carbon flux density have the same dimensions and are numerically equal to the total inflow to the node. The carbon flow density of the branch is a weighted average of the active power flow. The physical meaning of nodal carbon potential is: the equivalent carbon emissions on the generation side caused by a unit of electricity consumption at that node. For a power plant node, its nodal carbon potential is equal to the real-time carbon emission intensity of the power plant.
[0063] Next, we analyze the relationship between the node's carbon potential and the carbon flux density of its adjacent branches. Assume a node in a power system... Its adjacent branch road, such as Figure 1 As shown.
[0064] If the sets of branches flowing into and out of this node are respectively and From the first Branch road flows into node and from the first Branch outflow node The active currents are respectively and Due to the non-selectivity of electrical energy (the principle of proportional sharing), every branch in the outflow power flow has a component of every inflow power flow. Let's consider the branches in the outflow power flow... Contains tributaries of the inflow tidal current The component is According to the principle of proportional sharing, we have:
[0065] Assuming the line Inflow node The carbon flow density of the active current is Then the first The carbon flow rate of the active power flow in the outflow branch should be all Middle branch road to branch road The sum of carbon flow rate contributions is expressed as follows:
[0066] Therefore, the branch road carbon flux density for:
[0067] From the above formula, we get:
[0068] eliminate
[0069]
[0070] Given the tidal currents and carbon flow densities (carbon flow) flowing into all nodes, the carbon flow density of the outflowing tidal currents is constant and independent of the branches. Therefore, the carbon flow density of all tidal currents flowing out of a node is equal to the carbon potential of that node.
[0071] For the allocation of grid losses, the grid losses in each branch can be equivalent to the electrical load at the beginning of the line. The source of carbon emissions from grid losses in the system can be obtained by the generator-load carbon flow contribution rate analysis method. Therefore, the matrix required for carbon emission flow calculation can be determined entirely based on the grid structure, generator data, and power flow calculation results. Moreover, the calculation matrix is generated based on the vectors of power flow calculation results and network topology, which can significantly save computational storage space and ensure synchronization between carbon emission flow calculation and system power flow calculation.
[0072] Efficient implementation of load-side demand response under carbon constraints requires the establishment of scientific mathematical models and optimization methods to achieve a balance between system economy and low carbon emissions. Based on differences in system structure, control models are generally divided into two categories: single power systems and integrated energy systems. This paper focuses on single power systems and proposes a load-side control model under carbon constraints. This model achieves economic decoupling and low carbon goals in stages through two-level optimization control: the first level aims to minimize system operating costs by optimizing unit combination and power allocation; the second level aims to minimize load-side carbon emission costs by adjusting the load curve through demand response. To better coordinate the interests of power sources and loads, the model adopts a two-layer optimization structure: the upper layer aims to maximize overall system benefits by setting time-of-use electricity prices and carbon price signals; the lower layer aims to minimize the sum of user electricity costs and carbon emission costs by optimizing electricity consumption schemes. This model can accurately transfer the carbon emission responsibility of the power source side to the load side through carbon potential indicators, transforming the load from a passive recipient to an active regulator.
[0073] Two-level regulation model construction The first stage aims to minimize the economic operating cost of the system by optimizing unit combination and power allocation. The objective function and constraints are as follows: First-order objective function (minimizing economic cost)
[0074] Where: N is the total number of generating units; The generation cost function of the i-th generating unit (usually a quadratic function) = a i P g,i 2 + b i P g,i + c i ,in a i ,b i ,c i (cost coefficient) P g,i For the output of the i-th unit, C loss For network loss costs, P loss The system network loss power, and satisfies C loss = k loss P loss , k loss This is the grid loss electricity price coefficient.
[0075] Constraints: (1) Power balance constraint: In the formula: Original load power (kW); (2) Unit output constraints: P g,i,min ≤P g,i ≤P g,i,max In the formula: P g,i,min ,P g,i,max These are the minimum and maximum output (kW) of the i-th unit, respectively. (3) Unit ramping constraints: -R g,i,down ≤P g,i,t - P g,i,t-1 ≤R g,i,up In the formula: R g,i,down , R g,i,upThe values are the unit's downward and upward ramp rates (kW / h), respectively, and t is the dispatch time.
[0076] Second-order objective function (minimize carbon emission costs) The second level aims to minimize carbon emission costs on the load side by adjusting the load curve through demand response to achieve low-carbon optimization. The objective function is as follows: (2-5) In equation (2-5): kc is the carbon trading price; T is the scheduling period (h); Let be the carbon emission intensity of the i-th unit; The carbon emission reduction achieved by demand response at time t, and the load adjustment at time t. satisfy . , Let t be the average carbon emission intensity of the system at time t; The load adjustment amount at time t (kW, positive indicates reduction, negative indicates transfer).
[0077] Demand response constraints: min≤ ≤ max in min、 max represents the minimum and maximum adjustable load (kW), respectively. Daily electricity consumption conservation constraint: (Consider only load transfer scenarios).
[0078] Dual-layer optimized structure To reconcile the interests of power sources and loads, a two-tiered optimization structure is introduced: the upper tier aims to maximize overall system efficiency by setting time-of-use (TOU) electricity prices and carbon price signals; the lower tier aims to minimize the sum of user electricity costs and carbon emission costs by optimizing electricity consumption plans. Through the coordinated driving force of TOU electricity price kt and node carbon potential φn,t, the carbon emission responsibility of the power source side is accurately transferred to the load side, promoting the load to transform from a passive recipient to an active regulator.
[0079] Upper-level model (system side, time-of-use pricing optimization):
[0080] In the formula: T is the scheduling cycle; N is the total number of generating units; Let be the power generation cost function of the i-th generating unit at time t; Let represent the output (kW) of the i-th unit at time t; M represents the number of load nodes. The time-of-use electricity price at time t; The load power (kW) after adjustment at node n at time t is the same as the constraints of the upper-level model.
[0081] Lower-level model (load side, low-carbon response optimization): With the goal of achieving optimal low carbon emissions on the lower load side, the model expression is as follows: (2-7) In equation (2-7): This refers to the price of carbon trading.
[0082] The carbon potential at node n at time t represents the node's carbon responsibility. The carbon cost at time t of the nth node reflects the transfer of carbon responsibility. Constraints: Point load upper and lower limit constraints: P D,n,min ≤P D,n,t ≤P D,n,max P D,n,min ,P D,n,max These are the upper and lower limits of the load on node n, respectively. Nodal carbon potential correlation constraints:
[0083] Let i be the carbon emission intensity of the i-th unit. Let i be the power allocation coefficient for unit i to node n, satisfying .
[0084] High-proportion distributed new energy system strategy optimization With the rapid development of distributed photovoltaic and wind power and other new energy sources in new power distribution networks, the uncertainty of new energy output and load demand has increased. In order to fully tap the response potential of the demand side, a day-ahead and intraday coordination strategy is adopted, which connects low-carbon price signals and ultra-carbon incentive mechanisms. The day-ahead phase focuses on low-carbon guidance to formulate basic plans, while the intraday phase focuses on real-time adjustments to deal with fluctuations, jointly ensuring the low-carbon and economical operation of the system. Low-carbon constraints are planned to be formulated Demand response plans are developed based on forecast data (error ≤ 15%), with the goal of minimizing the sum of user electricity costs and carbon costs.
[0085] In the formula: The low-carbon electricity price is given at time t. The predicted load adjustment plan (kW) for the day before time t; Let t be the predicted carbon emission intensity before time t.
[0086] Constraints: Adjustable load adjustment constraints: ≤
[0087] The amount of load that can be shifted at time t. The maximum load that can be moved; Load adjustment boundary constraints: PD,min≤ ≤PD,max PD, min, PD, max are the upper and lower limits of the load of node n, respectively. Ultra-carbon incentives are subject to real-time adjustments and will be implemented within the day. Based on real-time data-driven rolling adjustments, and with the goal of minimizing ultra-carbon incentive costs, we address fluctuations in new energy sources and load:
[0088] In the formula: The intraday supercarbon incentive price at time t (RMB / kWh); This represents the real-time load adjustment within the day (kW, positive indicates reduction, negative indicates increase).
[0089] Constraints: Quickly adjust resource response capability constraints: ≤ , To quickly adjust the maximum response capacity of resources; Boundary constraints after intraday load correction: PD,min≤ ≤PD,max Daily electricity consumption conservation constraint: (Only considering load transfer scenarios, without changing the total daily electricity consumption).
[0090] Adjustment range constraints: | |≤δ (δ is the adjustment coefficient, usually taken as 0.2~0.5, to control the intraday adjustment range).
[0091] Example 3: This invention connects user-controlled water-cooled air conditioners to a virtual power plant load-side demand response control platform to perform flexible load control of the load-side cooling system during periods of power shortage and high carbon prices. The carbon footprint factors of the main power generation types are shown in Table 1 below: Table 1
[0092] The actual carbon emission factor of the user-side load can be calculated based on the carbon footprint factor of various power generation types according to the actual power generation ratio, yielding the actual carbon emission factor data for the load side. During demand response, load adjustments can be made based on the carbon emission factor during the response period and optimization methods constrained by carbon, ensuring low-carbon operation during the response period. A typical user-side carbon emission factor curve for a region is shown below. Figure 2 As shown.
[0093] Flexible load regulation of air conditioning systems is based on the principle of air conditioning water systems and building cold storage. The air conditioning water system and the building itself have cold and heat storage characteristics. Based on the AI intelligent agent of the load management platform, the control strategy is issued. First, the cooling is enhanced during the low carbon emission factor period to store cold in the water system and the building. During the high carbon emission factor period, the cooling output of the chiller system is reduced to make full use of the cold release of the water system and achieve the purpose of demand response and low carbon operation.
[0094] Flexible adjustment of air conditioning load is divided into three stages: pre-response preparation, during response, and end of response.
[0095] Before responding to demand, the air conditioning system lowers the setpoint of the chilled water outlet temperature. If the existing operating air conditioning units can meet the lowered temperature setting, no additional units need to be started. If the existing operating units cannot meet the demand, the system will automatically start another unit until the outlet water temperature drops to the setpoint. At this point, the air conditioning system stores the cooling capacity in the cooling medium (usually water).
[0096] For daytime and intraday control measures lasting more than 2 hours: a 1-hour pre-cooling of the water system is required during the pre-response preparation phase; for intraday control measures lasting less than 2 hours: a minimum of half an hour of pre-cooling of the water system is required during the pre-response preparation phase for rapid response. The load curve for the pre-response preparation phase of a shopping mall project in Fuzhou City under daytime control is shown below. Figure 3 As shown.
[0097] When the response period begins, the system automatically adjusts the air conditioning system's outlet water temperature to a higher parameter or shuts down the chiller. At this time, the system will adjust the number of units and load according to actual demand through peak-shaving strategies, significantly reducing the load and entering a low-power operation phase. Subsequently, the system gradually releases the cooling capacity previously stored in the medium. During the control process, the system will monitor the temperature and humidity parameters at the building's terminals in real time. If fluctuations in these parameters are detected, the air conditioning system will adjust its operating parameters accordingly to ensure terminal comfort. The load curve during the response period is shown below. Figure 4 As shown.
[0098] After the demand response time ends, the air conditioning system automatically returns to its original operating parameters. After the response ends, due to equipment restart, the user-side load will increase slightly. Simultaneously, because the carbon emission factor tends to increase, carbon emissions will also be higher than during the response. The load curve at the end of the response is shown below. Figure 5 As shown.
[0099] The changes in carbon emissions during the demand response process are not only related to electricity load but also closely related to dynamic carbon emission factors. Based on the dynamic carbon emission factors of Hangzhou's electricity sector, the carbon emission change curve during the response process is shown below. Figure 6 As shown.
[0100] Carbon emissions are closely related to the electricity load on the user side; the higher the electricity consumption, the greater the carbon emissions, and vice versa. Secondly, carbon emission intensity is related to the carbon emission factor of the energy type; the higher the proportion of green energy, the lower the carbon emission factor and the lower the carbon emission intensity, and vice versa. Demand response periods coincide with periods of highest carbon emission intensity, thus reducing load while achieving overall low-carbon operation.
[0101] Intelligent control is achieved through the above methods. Based on the operating characteristics of the water-cooled air conditioning equipment and the actual user needs, a customized automatic control strategy is adopted to perform fine-grained control of the user's equipment. The total baseline load and measured load curve changes of the control task are shown below. Figure 7 As shown, the average pressure reduction during the control response period was 68.8%, reducing carbon dioxide emissions by approximately 13.1%.
[0102] Practice has shown that the application effects of load-side demand response under carbon constraints are mainly reflected in three aspects: First, the load reduction effect is significant, and the overall carbon emissions can be significantly reduced. After adopting the remote control strategy of the new load management platform, the load reduction at the control point is about 30%-70%, while the carbon emissions during the response period decrease by 55%-80%.
[0103] Secondly, the impact on room temperature was minimal. During both control tasks, the average indoor temperature change for users was between 0℃ and 0.5℃, resulting in minimal impact on users' overall comfort during the control period; users were essentially unaware of the changes.
[0104] In terms of economics, demand response reduces investment in power generation facilities built to handle small peak loads, thus improving overall social welfare. Statistics show that 3%-5% of my country's peak load is concentrated within 50 hours. If these loads were entirely met by building fossil fuel power units, an investment of hundreds of billions of yuan would be required; however, through demand-side resource regulation, the investment scale can be significantly reduced.
[0105] In another embodiment of the present invention, a carbon-constrained load-side virtual power plant demand response control system is provided, which can be used to implement the above-mentioned carbon-constrained load-side virtual power plant demand response control method. Specifically, the system includes: The carbon potential acquisition module is used to calculate the carbon potential of each load node in real time based on the carbon emission flow. The carbon potential represents the equivalent generation-side carbon emission value caused by consuming a unit of electricity at that node. The model building module is used to construct a two-layer optimization control model under carbon constraints based on the carbon potential of each load node. The upper-layer model aims to maximize the overall operating efficiency of the power system and generates time-of-use electricity price signals. The lower-layer model aims to minimize the sum of electricity costs and carbon emission costs for load-side users and optimizes electricity consumption plans. The carbon emission costs include the carbon potential of the node and the carbon trading price. The control module is used to take the electricity consumption plan as the demand response target on the load side. During the preset high carbon emission period or power shortage period, it performs demand response on the aggregated flexible load, stores energy during the low carbon period before the response, releases the stored energy and reduces the real-time power during the response, and resumes normal operation after the response.
[0106] The module division in this embodiment of the invention is illustrative and represents only one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional modules in the various embodiments of the invention can be integrated into a single processor, exist as separate physical entities, or be integrated into a single module. The integrated modules described above can be implemented in hardware or as software functional modules.
[0107] In another embodiment of the present invention, a computer device is provided, comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the computer storage medium to achieve a corresponding method flow or corresponding function. The processor described in this embodiment of the present invention can be used for the operation of a load-side virtual power plant demand response control method based on carbon constraints.
[0108] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the load-side virtual power plant demand response control method based on carbon constraints in the above embodiments.
[0109] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0110] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0111] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0112] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0113] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.
Claims
1. A load-side virtual power plant demand response control method based on carbon constraints, characterized in that, include: Based on carbon emission flows, the carbon potential of each load node is calculated in real time, whereby the carbon potential represents the equivalent generation-side carbon emission value caused by consuming a unit of electricity at that node. Based on the carbon potential of each load node, a two-layer optimization control model under carbon constraints is constructed. The two-layer optimization control model includes an upper-layer model and a lower-layer model. The upper-layer model aims to maximize the overall operating efficiency of the power system and generates time-of-use electricity price signals. The lower-layer model aims to minimize the sum of electricity costs and carbon emission costs for load-side users. It uses the carbon potential of the nodes as the input of the two-layer optimization control model and combines the time-of-use electricity price signals to collaboratively optimize the output electricity consumption plan. Using electricity consumption plans as the target for demand response on the load side, demand response control is implemented on the flexible load equipment aggregated by the virtual power plant during preset high carbon emission periods or power shortage periods.
2. The load-side virtual power plant demand response control method based on carbon constraints according to claim 1, characterized in that, The method of calculating the carbon potential of each load node in real time based on carbon emission flows includes: Carbon potential at power system nodes, in symbol This indicates that a node is defined. The nodal carbon potentials are as follows: In the formula: For nodes carbon potential, There are power flow inflow nodes in the connected branches. The set of all branches, Subscript for branch road; branch road The meritorious trend; branch road carbon flux density; branch road carbon flow rate; The carbon potential of a node and the carbon flux density of its branches have the same dimensions and are numerically equal to the total carbon flux density flowing into the node. The carbon flow density of the branch is a weighted average of the active power flow, which is the equivalent carbon emission value on the generation side caused by a unit of electricity consumption at the node.
3. The load-side virtual power plant demand response control method based on carbon constraints according to claim 1, characterized in that, The carbon potential of each load node is used to construct a two-layer optimization control model under carbon constraints, including: The objective of the first-level objective function is to minimize economic cost, specifically: In the formula: Where N is the economic cost and N is the total number of generating units; Let be the power generation cost function of the i-th generating unit; For the output of the i-th unit, For network loss costs, The system network loss power, and satisfies = , This is the grid loss electricity price coefficient; The objective of the second-order objective function is to minimize carbon emission costs, specifically: With the goal of minimizing carbon emission costs on the load side, low-carbon optimization is achieved by adjusting the load curve through demand response. The objective function is as follows: In the formula: For carbon trading prices; The scheduling period; For the first Carbon emission intensity of the unit; for Carbon emission reductions achieved through real-time demand response, and real-time load adjustment satisfy . , for The average carbon emission intensity of the system at any given time; for Real-time load adjustment; The average reduction in carbon emissions per unit load adjustment; Demand response constraints: min≤ ≤ max in min、 `max` represents the minimum and maximum adjustable load, respectively. Daily electricity consumption conservation constraint: 。 4. The load-side virtual power plant demand response control method based on carbon constraints according to claim 3, characterized in that, The constraints of the first-level objective function include: Power balance constraints: In the formula: This represents the original load power. Unit output constraints: P g,i,min ≤P g,i ≤P g,i,max P g,i,min ,P g,i,max The first Minimum and maximum output of the unit; Unit ramp-up constraints: -R g,i,down ≤P g,i,t - P g,i,t-1 ≤R g,i,up In the formula: R g,i,down , R g,i,up These represent the unit's downward and upward ramp rates, respectively. t This refers to the scheduling time.
5. The load-side virtual power plant demand response control method based on carbon constraints according to claim 1, characterized in that, The upper-level model generates time-of-use electricity price signals with the goal of maximizing the overall operational efficiency of the power system. The lower-level model aims to minimize the sum of electricity costs and carbon emission costs for load-side users. It uses time-of-use pricing signals and node carbon potential to collaboratively optimize the output electricity consumption plan, including: The upper-level model aims to maximize the overall benefits of the system and sets time-of-use electricity and carbon price signals; The lower-level model aims to minimize the sum of user electricity costs and carbon emission costs by optimizing electricity consumption plans; through time-of-use pricing and node carbon potential, it transfers the responsibility for carbon emissions from the power supply side to the load side. The time-of-use pricing optimization expression of the upper-level model on the system side is as follows: In the formula: The scheduling period; This represents the total number of generating units. For the first Taiwanese crew The power generation cost function at any given time; For the first Taiwanese crew Efforts made at all times; M This represents the number of load nodes. for Time-of-use pricing; for node The load power is adjusted continuously, and the constraints of the upper-level model are the same as those mentioned above. Low-carbon response optimization on the load side of the lower-level model: With the goal of achieving optimal low carbon emissions on the lower load side, the model expression is as follows: In the formula: For carbon trading prices; for node Carbon potential at all times; No. Nodes The carbon cost at any given moment.
6. The load-side virtual power plant demand response control method based on carbon constraints according to claim 5, characterized in that, Constraints on the upper and lower level models: Point load upper and lower limit constraints: P D,n,min ≤P D,n,t ≤P D,n,max P D,n,min 、P D,n,max These are the upper and lower limits of the n-node load and the nodal carbon potential correlation constraints, respectively: For the first Carbon emission intensity of the Taiwanese unit For the unit For nodes The power allocation coefficient satisfies .
7. The load-side virtual power plant demand response control method based on carbon constraints according to claim 6, characterized in that, The proposed two-layer optimization control model under carbon constraints employs a two-stage coordinated optimization strategy involving the day-ahead and intraday phases. In the aforementioned day-ahead phase, based on the forecast data of new energy output and load demand, and guided by low-carbon electricity prices, a basic demand response plan for the load side is formulated. During the intraday phase, based on real-time data of renewable energy output and load demand, and using ultra-carbon incentives as a correction signal, the demand response baseline plan is adjusted on a rolling basis, including: Demand response plans are developed based on forecast data, with the goal of minimizing the sum of users' electricity costs and carbon costs. In the formula: It is the sum of the user's electricity cost and carbon cost. for Low-carbon electricity prices as of [date]; for Predicted load adjustment plan for the current time period; for Predicted carbon emission intensity up to the current date; Adjustable load adjustment constraints: ≤ for The load adjustment amount can be shifted at any time. The maximum load that can be moved; Load adjustment boundary constraints: PD,min≤ ≤PD,max PD,min,PD,max are respectively Node load upper and lower limits; Based on real-time data-driven rolling adjustments, and with the goal of minimizing ultra-carbon incentive costs, we address fluctuations in new energy sources and load: In the formula: To reduce carbon incentive costs, for Intraday ultra-carbon incentive price; This represents the real-time load adjustment within the day. Quickly adjust resource response capability constraints: ≤ , To quickly adjust the maximum response capacity of resources; Boundary constraints after intraday load correction: PD,min≤ ≤PD,max Daily electricity consumption conservation constraint: Adjustment range constraints: | |≤δ δ is the adjustment coefficient, which controls the intraday adjustment range.
8. The load-side virtual power plant demand response control method based on carbon constraints according to claim 1, characterized in that, The method of using electricity consumption plans as the demand response target on the load side, and performing demand response regulation on the flexible load equipment aggregated by the virtual power plant during preset high carbon emission periods or power shortage periods, includes: Before the demand response begins, the flexible load device is pre-adjusted to store cold or heat in the flexible load device or its medium. During demand response periods, the operating power of the flexible load equipment is reduced, and the stored cold or heat is released to meet load demands. After the demand response ends, the original operating parameters of the flexible load equipment are restored.
9. A load-side virtual power plant demand response control system based on carbon constraints, characterized in that, include: The carbon potential acquisition module is used to calculate the carbon potential of each load node in real time based on the carbon emission flow. The carbon potential represents the equivalent generation-side carbon emission value caused by consuming a unit of electricity at that node. The model building module is used to construct a two-layer optimization control model under carbon constraints based on the carbon potential of each load node. The upper-layer model aims to maximize the overall operating efficiency of the power system and generates time-of-use electricity price signals. The lower-layer model aims to minimize the sum of electricity costs and carbon emission costs for load-side users and optimizes electricity consumption plans. The carbon emission costs include the carbon potential of the node and the carbon trading price. The control module is used to take the electricity consumption plan as the demand response target on the load side. During the preset high carbon emission period or power shortage period, it performs demand response on the aggregated flexible load, stores energy during the low carbon period before the response, releases the stored energy and reduces the real-time power during the response, and resumes normal operation after the response.
10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the load-side virtual power plant demand response control method based on any one of claims 1 to 8.
11. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the load-side virtual power plant demand response control method based on any one of claims 1 to 8.
