Virtual power plant unit output control method considering improved demand response

Abstract

The invention relates to a virtual power plant unit output control method considering improved demand response, and the method comprises the steps: S1, collecting the output of each unit and the historical data of each influence factor, and carrying out the prediction based on the historical data, and obtaining the output of each unit and the first prediction data of each influence factor in the next day; s2, the predicted output of each unit in the next day and first prediction data of each influence factor are substituted into a pre-configured day-ahead optimization model, and in the day-ahead optimization model, the lowest demand response cost is obtained based on the consumption surplus value, and the carbon emission cost is obtained based on the carbon intensity; s3, the day-ahead optimization model is used for solving to obtain an output time sequence of each unit in the next day, and each element in the time sequence corresponds to each time period of the next day; and S4, on the basis of the second prediction data of each influence factor, combining the obtained output time sequence of each unit, and performing intra-day optimization to obtain the output of each unit. Compared with the prior art, the method has the advantages that the influence of the demand side resource response on the system scheduling result is effectively reflected through the change trend of the electric quantity transaction volume, so that the economy and the safety are improved.

Description

Technical Field

[0001] The present invention relates to the field of power dispatching, and in particular to a method for controlling the output of virtual power plant units taking into account improved demand response. Background Art

[0002] The proportion of installed capacity of new energy sources has increased rapidly, new grid-connected entities such as virtual power plants have emerged widely, and the construction of new power systems has made significant progress, further promoting the realization of the "dual carbon" goals. For new power systems with a gradually increasing proportion of new energy sources, the coupling of power abandonment and power shortages has made the problem of power balance more complicated. In new power systems, the uncertainty on the source side has increased and the balancing ability has decreased, and it is becoming increasingly difficult to maintain the balance between supply and demand by relying solely on source-side control measures. Based on this, the present invention uses load demand response to provide regulation resources to the system, which has become an effective means to alleviate the structural power shortage of the system and achieve power balance.

[0003] A VPP is an electricity operation model that utilizes advanced digital and intelligent technologies to aggregate distributed generation (DGs) and adjustable loads, enabling them to collaboratively participate in system operations and market transactions. DR is a measure within VPPs to guide users in optimizing their electricity consumption. Compared to traditional dispatch methods, it offers faster demand-side load response and is low-carbon. During the demand response process, user behavior exhibits a degree of uncertainty. Some studies have explored the impact of price elasticity curve uncertainty on both the demand and supply sides. Other studies have constructed demand response uncertainty models in which the response probability distribution function varies with incentive prices, examining the impact of incentive levels on parameters such as demand response volume and response volatility. These approaches consider the impact of economic factors on the willingness to respond to uncertain responses within the VPP model but fail to account for the vulnerability of users to various uncertainties, such as psychological factors, during actual user participation. Therefore, it is necessary to comprehensively consider user behavior under the combined influence of economic, low-carbon, and consumer psychology factors, and develop an improved demand response uncertainty model for demand response resources under different incentive levels. This approach aims to maintain the stability and security of VPPs while ensuring their economic operation.

[0004] Regarding VPP operational structure and optimization, researchers at home and abroad are currently constructing VPP cluster architectures that incorporate distributed renewable energy, energy storage, controllable loads, and more, and proposing a two-stage VPP scheduling model optimized for multiple time scales. Furthermore, some studies have constructed VPP scheduling models that incorporate DR, gas turbines, wind power, and photovoltaic power generation, and participate in carbon emissions trading. These studies primarily focus on increasing incentives through internal VPP electricity price subsidies, but have not yet considered the incentives for demand-side resource participation in DR, resulting in an inability to balance economic efficiency and safety. Summary of the Invention

[0005] The purpose of the present invention is to provide a virtual power plant unit output control method taking into account improved demand response in order to solve the defects of the above-mentioned prior art.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] A virtual power plant unit output control method considering improved demand response includes:

[0008] Step S1: collecting historical data of the output of each unit and each influencing factor, and predicting the output of each unit and each influencing factor for the next day based on the historical data;

[0009] Step S2: Substitute the predicted output of each unit for the next day and the first predicted data of each influencing factor into a pre-configured day-ahead optimization model, wherein the optimization objective of the day-ahead optimization model is to minimize the sum of the operation and maintenance costs of wind power, photovoltaic power, and energy storage, the operation cost of the gas turbine, the number of intervals for purchasing and selling electricity costs, the number of intervals for carbon emission costs, and the demand response cost, wherein the minimum demand response cost is obtained based on the consumer surplus value, and the carbon emission cost is obtained based on the carbon intensity;

[0010] Step S3: using the day-ahead optimization model to solve and obtain the output time series of each unit for the next day, wherein each element in the time series corresponds to each time period of the next day;

[0011] Step S4: Based on the second prediction data of each influencing factor and in combination with the obtained output time series of each unit, intraday optimization is performed to obtain the output of each unit.

[0012] The demand response cost is:

[0013]

[0014] Where: C DR,t is the demand response cost, N T is the number of moments in the time period T, N load is the total number of nodes, π i,t is the consumption surplus value of electricity consumption at node i at time t, is the compensation electricity price of node i at time t, is the load that the i-th node is willing to participate in demand response at time t.

[0015] The load that the i-th user is willing to participate in demand response is:

[0016]

[0017] Where: P load_i0 is the demand load before implementing time-of-use electricity pricing, λ PVis the load response transfer rate during peak and valley periods, λ PF is the load response transfer rate between peak and normal periods, λ FV is the load response transfer rate during the off-peak period, is the average load during peak hours, The average load during normal periods.

[0018] The carbon emission costs are specifically:

[0019]

[0020] Where: C ce,T is the cost of carbon emissions, is the unit carbon price, λ E is the carbon price base price, E Gc,i (T) is the net CO2 emission of the i-th gas turbine in time period T, E Gb,i (T) is the carbon quota of the i-th gas turbine in time period T, is the comprehensive carbon intensity, λ Gi is the carbon emission benchmark per unit of electricity, P Gi (t) is the net CO2 emission of the i-th gas turbine at time t.

[0021] The comprehensive carbon intensity is specifically:

[0022]

[0023] in: is the carbon intensity of the generator corresponding to node i, is the output of the generator corresponding to node i, is the carbon intensity of branch k, f t,k is the unit carbon emission of branch k, is a collection of transmission lines.

[0024] The step S4 comprises:

[0025] Step S4-1: Select the first time period of the day as the current proposed intraday optimization period;

[0026] Step S4-2: Obtaining second prediction data of various influencing factors of the current proposed intraday optimization period based on historical data prediction;

[0027] Step S4-3: Based on the obtained output time series of each unit, extract the processed value corresponding to the current proposed intraday optimization period as the initial output value of each unit, and substitute the initial output value of each unit and the second predicted data of each influencing factor into the intraday optimization model, wherein the optimization objective of the intraday optimization model is to minimize the sum of the intraday incentive electricity price demand response cost, the output deviation penalty within the T period of the day, the intraday operation and maintenance costs of wind power, photovoltaic power and energy storage, the intraday gas turbine operation cost, the number of intraday electricity purchase and sales cost intervals, and the number of intraday carbon emission cost intervals;

[0028] Step S4-4: Use the intraday optimization model to solve and obtain the output of each unit in the current quasi-intraday optimization period, and determine whether the traversal of all intraday periods is completed. If so, exit; otherwise, select the next intraday period as the current quasi-intraday optimization period and return to step S4-2.

[0029] The intra-day improved demand response cost is:

[0030]

[0031] in: is the demand response cost of the intraday incentive electricity price type, Time-of-use electricity price, is the load before demand response, As the basic electricity price, To encourage subsidies for electricity prices, is the load after demand response, Δt in Intraday time scale.

[0032] The output deviation penalty within the T period of the day is the sum of the output deviations of all units.

[0033] A device for testing the plug-and-charge function of a charging pile includes a memory, a processor, and a program stored in the memory. When the processor executes the program, the method described above is implemented.

[0034] A storage medium stores a program, which implements the above method when executed.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] This paper utilizes the surplus value of consumption to guide energy users to participate in demand response, and then optimizes VPP operation based on the carbon intensity of different loads. Using the aforementioned method, wind and solar power output and electricity prices are modeled, while the carbon price and demand response are modeled using the surplus value theory. With the goal of minimizing VPP operating costs, a VPP optimal dispatch model is established by comprehensively considering electricity market transactions, carbon trading, and an improved demand response mechanism. A short-timescale rolling optimization algorithm is used to solve the model, innovatively establishing a link between carbon prices and electricity prices. This further determines the demand-side resource response in the VPP system, resolving the issues of energy reserve and electricity price compensation in the VPP optimal dispatch model. Ultimately, the impact of demand-side resource response on system dispatch outcomes is effectively reflected through the changing trends in electricity transaction volume, improving economic efficiency while ensuring the safe operation of the VPP system. The DR guidance form of the surplus value theory leaves room for VPP dispatch and provides guidance for future VPP system optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is the demand response coefficient curve;

[0038] Figure 2 It is the residual value valuation curve;

[0039] Figure 3 Optimize the scheduling system architecture for VPP;

[0040] Figure 4 Schematic diagram of the main steps of the method of the present invention;

[0041] Figure 5 The system scheduling results under scenario 1 and scenario 2 are shown in Figure 1, where (a) is the output diagram of each unit under scenario 1, and (b) is the output diagram of each unit under scenario 2;

[0042] Figure 6 This is a comparison chart of demand-side load response results between scenario 2 and scenario 3;

[0043] Figure 7 This is the output result diagram of the demand-side load before and after participating in demand response;

[0044] Figure 8 The diagrams of the scheduling results of each unit of the VPP system under scenario 3 are as follows: (a) is the gas turbine output comparison diagram, (b) is the energy storage output comparison diagram, and (c) is the VPP purchase and sales electricity comparison diagram;

[0045] Figure 9 The following are the dispatch results of each unit of the VPP system under scenario 4, where (a) is the gas turbine output comparison diagram, (b) is the energy storage output comparison diagram, and (c) is the VPP purchase and sales electricity comparison diagram. DETAILED DESCRIPTION

[0046] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0047] This application aims to establish a VPP optimization scheduling model that takes into account improved demand response using an appropriate step-by-step pricing method. In the VPP system, electricity prices and carbon prices are affected by the supply and demand and quotation relationship, and there will be corresponding fluctuations. Demand response, as a means of user-side load regulation, is easily affected by factors such as user psychology and user participation. Under the combined effect of multiple uncertainties, the VPP coordination and control center needs to coordinate the output of multiple aggregation units in actual operation to maintain safe and stable system operation.

[0048] The application is based on the following ideas:

[0049] (1) Establish a demand response model based on the compensation electricity price method of demand response transfer volume.

[0050] The ratio of the amount of electricity consumed by users shifting from high-price periods to low-price periods after implementing time-of-use electricity prices to the load corresponding to the high-price period is calculated. This ratio is used to represent the user's flexibility in participating in DR. A variable unit subsidy amount is then allocated based on the duration and response level of the user's participation in DR. The basic idea behind demand response transfer volume is as follows:

[0051] 1) Load response transfer rate during peak and valley periods

[0052]

[0053] Δp PV =p P -p V

[0054]

[0055] Among them, P p is the electricity price during peak hours, p v is the electricity price during off-peak hours, a pv is the upper limit of the dead zone of the electricity price difference during peak and valley periods, b pv k is the upper limit of the linear region of the price difference between peak and valley periods, PV is the load transfer rate during peak and valley periods.

[0056] 2) Load response transfer rate during peak and normal periods

[0057]

[0058] Δp PF=p P -p F

[0059]

[0060] Where: p F is the electricity price during the normal power period, k PF is the slope value of the linear region of the price difference between peak and average periods, a pF The difference between peak and flat electricity prices is in the upper limit of the dead zone, b pF The price difference between peak and flat periods is in the upper limit of the linear region.

[0061] 3) Load response transfer rate during off-peak periods

[0062]

[0063] Δp FV =p F -p V

[0064]

[0065] Where: k FV is the slope value of the linear region of electricity price difference during flat and valley periods, a FV b is the upper limit of the dead zone for the price difference during the off-peak period, FV It is the upper limit value of the price difference in the linear region during the off-peak period.

[0066] The demand load trend of the system is defined as:

[0067]

[0068] Where: P load_i0 is the demand load before implementing time-of-use electricity pricing, λ PV is the load response transfer rate during peak and valley periods, λ PF is the load response transfer rate between peak and normal periods, λ FV is the load response transfer rate during the off-peak period, is the average load during peak hours, The average load during normal periods.

[0069] Peak, flat, and valley subsidy amounts should be individually customized, allowing users to selectively develop response plans based on different incentives. Variable unit subsidy amounts are then allocated based on the duration and degree of user participation in DR. The electricity price subsidy amount is defined as:

[0070]

[0071] Where, is the electric load of the user in response to the stimulus of node i in period t; is the initial electricity load of the user in period t; is the compensation price for node i during the time period corresponding to time t. To further incentivize users to participate in DR, the compensation price given to users is proportional to their response power, that is, the more responses, the more compensation.

[0072] (2) The load carbon intensity method is used to process carbon prices and an improved demand response model is established based on the consumption surplus value theory.

[0073] The present invention proposes a VPP model mainly composed of wind power, photovoltaics, energy storage, curtailable loads, and gas turbines. The carbon emission sources in the VPP model are mainly gas turbines and curtailable loads on the user side. The present invention adds the concepts of carbon emission constraints and consumer surplus value to the VPP operation model, purposefully guides the low-carbon economic operation of the VPP, and clarifies the production and emission responsibilities of each carbon emission source. While encouraging users to actively participate in DR, the excessive electricity consumption behavior of users affected by incentives is appropriately restrained. In order to deal with the uncertainty of carbon prices, a carbon emission flow model characterized by a carbon intensity index is introduced to evaluate the emissions of the system. Carbon intensity is defined as:

[0074]

[0075] in: is the carbon intensity of the generator corresponding to node i, is the output of the generator corresponding to node i, is the carbon intensity of branch k, f t,k is the unit carbon emission of branch k, is a collection of transmission lines.

[0076] Extensive research has been conducted on demand-side user behavior. In addressing carbon emissions, the impact of carbon price fluctuations on user participation in DR has been considered. Fluctuating carbon prices can affect the amount of response shifts, and shifts in electricity consumption can negatively impact electricity price compensation. To ensure user participation and comfort, a consumer surplus value function is introduced. Consumer surplus generally refers to the difference between the benefits or satisfaction an individual or consumer actually receives after purchasing a good or service and the price paid. In DR, consumer surplus can be used to express a user's perceived value of their energy consumption after participating in DR, provided their rigid energy needs are met. When calculating costs, taking into account the impact of carbon price fluctuations on users can reduce demand shifts to a certain extent. By introducing the consumer surplus concept into the consumer experience valuation function, under the same carbon price, a user experience valuation is calculated to characterize the user's true level of participation in a low-carbon lifestyle.

[0077] In the present invention, the user's initial demand response coefficient is λ, the incentive price is x, and the demand response coefficient curve based on user psychology is shown in the attached figure. Figure 1 As shown. Figure 1 The shaded area and the line y = λ1 represent the effective user response. When the incentive price is 0, although users have a certain response range [0, λ0], as the incentive price increases, the demand response coefficient increases, and user behavior gradually transitions from increasing load to participating in the demand response and reducing load. When the incentive price is x0, the critical point N is reached, at which point all users participate in demand response and reduce load, and the overall system load does not increase. When the incentive price continues to increase to x1, the system is saturated, and the user response reaches its maximum value. At this point, the uncertainty of user participation in the response can be ignored.

[0078] When the incentive price is x, the upper and lower limit curves of demand response λ min and λ max Satisfies the following expression:

[0079]

[0080] The user experience valuation on the demand side can be represented by the consumption surplus value calculated by the demand-price function in line with consumer psychology, as shown in the attached figure. Figure 2 As shown. Usually, demand is negatively correlated with price, and the common function is q = a·p ε , in general, q is the load; p is the price; a is the demand price constant. is the original load of node i in the period corresponding to time t; is the rigid load of node i in the period corresponding to time t; is the sales price of node i at time t. At this time, if the carbon price is the base price λ E , then, depending on the amount of carbon emissions, the unit carbon price will also change, such as To facilitate the solution, the power load factor of node i at time t is The actual power load factor of node i at time t is

[0081] Specifically, in this application, the consumption surplus value of the user's electricity consumption at node i in the corresponding period at time t is assumed to be:

[0082]

[0083] In order to avoid the decision variables appearing in the real number of the natural logarithm and thus the nonlinear part of the decision variables appearing in the function, some approximation is done in the second step. According to microeconomic theory, the attached Figure 2The blue area in the figure represents the user's consumption valuation. A higher value indicates a stronger energy experience. A high user-side experience valuation indicates that, at low carbon prices, users are generating excessive energy for DR by transferring response volume to obtain subsidies. This lack of incentive to exploit this surplus value results in the system's costs exceeding actual costs, inevitably impacting their energy experience and reducing the experience valuation.

[0084] (3) Establishment of VPP Optimal Scheduling Model Considering Improved Demand Response

[0085] The VPP system studied in this application uses communication technology to aggregate its internal units into a whole, participate in the power system dispatch, and connect to the main grid for energy exchange. On the basis of meeting the energy demand of users and corresponding constraints, the total cost (including wind and solar operation and maintenance costs, gas turbine costs, demand response costs, carbon emission costs and electricity purchase and sales costs) is minimized. The present invention establishes a connection between electricity prices and carbon prices through demand response transfers, and establishes a VPP multi-time scale optimization dispatch model considering improved demand response. The VPP system optimization dispatch architecture is shown in the attached figure. Figure 3 shown.

[0086] A virtual power plant unit output control method considering improved demand response, such as Figure 4 Shown, including:

[0087] Step S1: collecting historical data of the output of each unit and each influencing factor, and predicting the output of each unit and each influencing factor for the next day based on the historical data;

[0088] Step S2: Substitute the predicted output of each unit for the next day and the first predicted data of each influencing factor into a pre-configured day-ahead optimization model, wherein the optimization objective of the day-ahead optimization model is to minimize the sum of the operation and maintenance costs of wind power, photovoltaic power, and energy storage, the operation cost of the gas turbine, the number of intervals for purchasing and selling electricity costs, the number of intervals for carbon emission costs, and the demand response cost. The minimum demand response cost is obtained based on the consumer surplus value, and the carbon emission cost is obtained based on the carbon intensity.

[0089] The demand response cost is:

[0090]

[0091] Where: C DR,t is the demand response cost, N T is the number of moments in the time period T, N load is the total number of nodes, π i,t is the consumption surplus value of electricity consumption at node i at time t, is the compensation electricity price of node i at time t, is the load that the i-th node is willing to participate in demand response at time t.

[0092] The load that the i-th user is willing to participate in demand response is:

[0093]

[0094] Where: P load_i0 is the demand load before implementing time-of-use electricity pricing, λ PV is the load response transfer rate during peak and valley periods, λ PF is the load response transfer rate between peak and normal periods, λ FV is the load response transfer rate during the off-peak period, is the average load during peak hours, The average load during normal periods.

[0095] The specific carbon emission costs are:

[0096]

[0097] Where: C ce,T is the cost of carbon emissions, is the unit carbon price, λ E is the carbon price base price, E Gc,i (T) is the net CO2 emission of the i-th gas turbine in time period T, E Gb,i (T) is the carbon quota of the i-th gas turbine in time period T, is the comprehensive carbon intensity, λ Gi is the carbon emission benchmark per unit of electricity, P Gi (t) is the net CO2 emission of the i-th gas turbine at time t.

[0098] The comprehensive carbon intensity is as follows:

[0099]

[0100] in: is the carbon intensity of the generator corresponding to node i, is the output of the generator corresponding to node i, is the carbon intensity of branch k, f t,k is the unit carbon emission of branch k, is a collection of transmission lines.

[0101] Specifically, the objective function of the day-ahead optimization model is:

[0102] minC vpp =min(C op,T +C th,T +C bs,T +C ce,T +C DR,T )

[0103] Among them, C vpp is the total operating cost of the VPP system during the T period; C op,t is the operation and maintenance cost of wind power, photovoltaic power and energy storage; C th,t is the gas turbine operating cost; C bs,t is the cost interval for electricity purchase and sale; C ce,t is the carbon emission cost interval number; C DR,t is the demand response cost.

[0104] Operation and maintenance costs of wind power, photovoltaic power and energy storage C op,t for:

[0105]

[0106] Among them, t is the specific scheduling period; s w 、s sol 、s sto It is divided into wind turbine operation and maintenance costs, photovoltaic operation and maintenance costs and energy storage battery operation and maintenance costs; P w (t), P sol (t), P dis (t) are wind power output, photovoltaic output and battery discharge value at time t respectively.

[0107] Gas turbine operating cost C th,t for:

[0108] The operating cost of a gas turbine unit consists of fuel cost and start-up and shutdown cost. The fuel cost C fue,t It is often expressed as a quadratic function related to the unit output, and the start-up and shutdown cost C ss,t It is related to the start-up and shutdown status of the gas turbine at time t. If the unit is started or shut down at this moment, it will be included in the start-up and shutdown costs, as shown in the following formula:

[0109]

[0110] Among them, a, b, and c are the coefficient values ​​of the quadratic term, linear term, and constant term of fuel cost, respectively, which are usually determined by mathematical fitting; P Gi (t) is the output value of the i-th gas turbine at time t; C Gsi with C Gti are the startup cost and shutdown cost of the i-th gas turbine respectively; δ Gsi (t) and δ Gti (t) is a Boolean variable, indicating the start and stop status of the i-th gas turbine at time t. When starting, δ Gsi (t) is set to 1, otherwise it is set to 0. Similarly, when the shutdown is Gti (t) is set to 1, otherwise it is set to 0; N u is the number of gas turbines;

[0111] VPP electricity purchase and sales cost C bs,t for:

[0112]

[0113] Among them, C buy (t) and C sell (t) are the purchase price and sales price of electricity of the power grid in the tth period respectively; P buy (t) and P sell (t) are the power purchased and sold by the VPP to the grid at the tth moment (the power purchased is defined as a positive number, and the power sold is defined as a negative number);

[0114] The specific constraints of the day-ahead optimization model are as follows:

[0115] 1) Gas turbine constraints

[0116] Unit output and climbing constraints:

[0117]

[0118] Where, δ Gri (t) is a Boolean variable representing the operating status of the i-th unit at time t, which is set to 1 if it is running and 0 if it is stopped; P Gi_min 、P G are the lower and upper limits of the output of the i-th gas turbine respectively; P Ci is the ramp power of the i-th gas turbine.

[0119] The unit state constraints are shown in the following formula.

[0120]

[0121] 3) VPP power purchase and sales transaction constraints

[0122]

[0123] Where, δ bs (t) is a Boolean variable, indicating the VPP power purchase and sale flag at time t, which is set to 1 when power is purchased and 0 when power is sold; power It is the power exchange limit between VPP and main grid.

[0124] 4) Energy storage constraints

[0125] The battery charge and discharge constraints are shown in the following equation.

[0126]

[0127] Where, P bat (t), P cha (t), Pdis (t) are the battery output, charging capacity and discharging capacity for the period corresponding to time t; δ cha (t), δ dis (t) is a Boolean variable, representing the battery charging and discharging flags at time t. cha (t) is set to 1, otherwise it is 0, when discharging δ dis (t) is set to 1, otherwise it is 0; P cs It is the battery charge and discharge limit.

[0128] The battery state constraints are shown in the following equation.

[0129]

[0130] Where, δ sta (t) is the battery rest flag at time t, which is set to 1 when the battery is neither discharging nor charging, and 0 in other cases.

[0131] Definition of battery state of charge (SOC) at time t:

[0132]

[0133] Where, E bat (t) is the battery capacity at time t, E max The maximum capacity of the battery.

[0134] Battery SOC constraints:

[0135]

[0136] 5) Incentive demand response constraints

[0137] The fluctuation of demand response when the incentive price is x is F(x):

[0138]

[0139] The corresponding constraints are as follows:

[0140]

[0141] The constraints are demand response volatility constraint, response capacity constraint and response coefficient constraint. min With F max are the minimum and maximum fluctuation coefficients respectively; χ represents the maximum value of the VPP system demand response coefficient.

[0142] Step S3: using the day-ahead optimization model to solve and obtain the output time series of each unit for the next day, wherein each element in the time series corresponds to each time period of the next day;

[0143] Step S4: Based on the second prediction data of each influencing factor and in combination with the obtained output time series of each unit, intraday optimization is performed to obtain the output of each unit, including:

[0144] Step S4-1: Select the first time period of the day as the current proposed intraday optimization period;

[0145] Step S4-2: Obtaining second prediction data of various influencing factors of the current proposed intraday optimization period based on historical data prediction;

[0146] Step S4-3: Based on the obtained output time series of each unit, extract the processed value corresponding to the current proposed intraday optimization period as the initial output value of each unit, and substitute the initial output value of each unit and the second predicted data of each influencing factor into the intraday optimization model, wherein the optimization objective of the intraday optimization model is to minimize the sum of the intraday incentive electricity price demand response cost, the output deviation penalty within the T period of the day, the intraday operation and maintenance costs of wind power, photovoltaic power, and energy storage, the intraday gas turbine operation cost, the number of intraday electricity purchase and sales cost intervals, and the number of intraday carbon emission cost intervals;

[0147] The objective function of the intraday optimization model is:

[0148]

[0149] Where, This is the penalty fee for output deviation within the T period of the day; The meaning and expression of the remaining wind power, photovoltaic and energy storage operation and maintenance costs, gas turbine operating costs, VPP power purchase and sales costs and carbon emission costs are the same as today and will not be repeated here.

[0150] The cost of demand response with an intraday incentive electricity price is:

[0151]

[0152] in: is the demand response cost of the intraday incentive electricity price type, Time-of-use electricity price, is the load before demand response, As the basic electricity price, To encourage subsidies for electricity prices, is the load after demand response, Δt in Intraday time scale.

[0153] The output deviation penalty during the T period of the day is the sum of the output deviations of all units. Specifically:

[0154]

[0155] Where, are the deviations of the daily dispatch values ​​of gas turbines, batteries, and grid transaction power compared with the day-ahead plan; Γ GT ,Γ bat ,Γ net are the deviation penalty coefficients for gas turbines, batteries, and electricity purchase and sales transactions, respectively.

[0156] The constraints are as follows:

[0157] The constraints for intraday rolling optimal scheduling are largely similar to those for day-ahead scheduling, except that all day-ahead data is converted to intraday data. These constraints include gas turbine constraints, power balance constraints, VPP power purchase and sales constraints, and energy storage constraints, which are not detailed here. It is important to note that since the near-term scheduling has already determined the gas turbine start and stop schedule, the near-term start and stop schedule will be used for intraday scheduling.

[0158] The start and stop status of the gas turbine within a day must also comply with the status constraints of the day-ahead scheduling.

[0159]

[0160] Where, They represent the start and stop status of the i-th gas turbine at time t within the day and are Boolean variables.

[0161] For this model, the mixed integer nonlinear model is converted to a mixed integer linear model. The YALMIP toolbox of MATLAB is also needed to build the model, and the CPLEX solver can be called to solve it. The solution process has two stages, which are divided into the long-term model solution of the day-ahead and the short-term rolling optimization solution of the day-ahead. The modified day-ahead optimal adjustment result is achieved by using the intraday forecast data. The specific process is shown in the attached figure. Figure 4 shown.

[0162] Step S4-4: Use the intraday optimization model to solve and obtain the output of each unit in the current quasi-intraday optimization period, and determine whether the traversal of all intraday periods is completed. If so, exit; otherwise, select the next intraday period as the current quasi-intraday optimization period and return to step S4-2.

[0163] Generally, influencing factors may include weather, user compliance, electricity purchase and sales prices, carbon prices, etc.

[0164] This example examines a small VPP system in East China, comprising a gas turbine, wind power, photovoltaic power generation, battery energy storage, and user flexible loads. The key parameters are as follows: scheduling period T = 24 hours, optimization time scale Δt = 1 hour. The carbon emission benchmark per unit of electricity is 0.76 kg / kWh; the power exchange limit between the VPP system and the main grid is set at 200 kW; and the operation and maintenance costs for wind power, photovoltaic power generation, and battery energy storage are 0.52 yuan / kW, 0.72 yuan / kW, and 0.5 yuan / kW, respectively.

[0165] In order to compare and analyze the impact of multiple uncertain factors such as the incentive electricity price and the improved demand response mechanism on the VPP optimization scheduling results in this invention, four scenarios are set up for analysis. The specific scenario settings are as follows:

[0166] Scenario 1: Virtual power plant operation and dispatch strategy without considering demand response and consumer psychology.

[0167] Scenario 2: Economic operation and dispatch strategy of virtual power plants after considering incentive compensation.

[0168] Scenario 3: Economic operation and dispatch strategy of virtual power plants after considering incentive compensation and demand response.

[0169] Scenario 4: Virtual power plant operation scheduling strategy considering the tiered pricing mechanism and improved demand response.

[0170] (Proposed by the present invention)

[0171] The optimized scheduling results of scenario 1 are shown in the attached Figure 5 (a). Figure 5As can be seen, during the 1:00-4:00 AM period, wind and photovoltaic resources are relatively scarce, user load demand is low at night, and the electricity purchase price is low during this time. At this time, the VPP primarily meets user load demand through purchased power and battery discharge, based on actual conditions. In Scenario 2, due to the impact of incentive prices, user response load decreases somewhat, and gas turbine output is limited to control carbon costs. During the 6:00-12:00 AM period, wind and photovoltaic resources increase, reaching their maximum, and user load demand also increases significantly. Users shift their load, reducing system output peaks and thus obtaining lower electricity purchase prices. At 5:00, 8:00, and 9:00 AM, due to higher electricity prices, the VPP system sells excess electricity to the power market, generating profits and reducing overall system operating costs. The gas turbine also purchases more carbon emission rights through the tiered carbon price carbon trading mechanism, thereby reducing carbon emission costs. During the 6:00-7:00 AM period, the VPP system meets user load demand by increasing gas turbine output and leveraging increased wind and solar resources. During the 13:00-19:00 period, user load continues to increase, while electricity purchase and sales prices decrease. Therefore, the VPP system purchases electricity and charges batteries. Between 20:00 and midnight, wind power resources reach their maximum, user load also reaches its maximum, and gas turbines maintain minimum output, thereby reducing carbon costs. At this time, the purchase price is high, while the sales price is low. The VPP system meets user demand during peak load periods by purchasing electricity and discharging batteries. A comparative analysis shows that, compared to Scenario 1, in Scenario 2, the impact of incentive prices on user load is significantly reduced at 12:00 and 20:00. Furthermore, user load also exhibits a significant valley-filling response at 4:00 and 16:00.

[0172] In summary, the VPP system can combine the purchase and sale prices of electricity, the amount of load participating in demand response, and the charging and discharging of energy storage to meet load demand while achieving low-price purchases and high-price sales of electricity, thereby profiting from the electricity market. By rationally regulating its internal resources within the dispatch cycle, it ensures the stability of the entire system and improves the economic efficiency of VPP operation.

[0173] Impact of uncertainty in compensation price on dispatch results based on demand response transfer volume: In order to describe the impact of uncertainty in compensation price on dispatch results, scenario 1 and scenario 2 are compared and analyzed. The power transfer of demand-side load in VPP after participating in demand response in scenario 1 and scenario 2 is shown in the attached figure. Figure 6 As shown. By analyzing the user-side load that responds flexibly to changes in electricity prices, it is found that the VPP system will purchase electricity from the electricity market when the local electricity purchase price is low, and sell excess electricity to the electricity market when the electricity sales price is high, thereby reducing the peak-valley difference. Figure 5 and attached Figure 6As can be seen, the curtailable load experienced a minimal decrease before 9:00 AM, but showed a significant response to the main grid electricity price increase after 10:00 AM, resulting in a reduction in VPP electricity sales. At this point, since the cost of reducing load is lower than the main grid electricity price, the system can reduce its carbon emissions by reducing load. The shiftable load shifted some load from 12:00 PM, reducing electricity demand during the peak electricity price period until it recovered at 11:00 PM, when electricity prices began to fall. During this period, the VPP also met its own needs through discharge from its energy storage batteries, reducing electricity purchases and ultimately reducing costs. Overall, electricity price uncertainty primarily impacts VPP electricity purchasing and sales behavior. When factoring in electricity price uncertainty, the VPP system will adjust its electricity purchasing and sales behavior to account for this uncertainty.

[0174] The impact of uncertainty in improved demand response on dispatch results: Take the optimization results in scenarios 3 and 4, as shown in the attached Figure 7 As shown. Analyzing load electricity consumption on an hourly basis shows that, without implementing demand response, electricity demand surges between 12:00 PM and 7:00 PM to 10:00 PM. Once gas turbines are generating electricity, they must purchase carbon emission credits from the carbon market, increasing costs and impacting the environment. Although overall load has been low for extended periods, the available load for user-side demand response is low, resulting in low resource utilization on the demand side. With improved demand response, more users are participating in DR during previously peak demand periods, significantly reducing the peak-to-valley difference in the load curve and significantly improving its volatility. This helps mitigate the impact of renewable energy uncertainty on the system and effectively reduces carbon emissions in a short period of time. Although load demand increased at 12:00 PM, due to higher wind power output and increased energy storage charging in the morning, demand-side user participation in DR was low, resulting in closer curve nodes. From 6:00 PM to midnight, photovoltaic power generation gradually ceased operations, leading to a sharp increase in electricity purchase costs and increased load pressure on the main grid. On the demand side, more users are participating in DR, resulting in a significantly higher DR curve than the original. Simultaneously, energy storage discharges, working with wind power and participating users to maintain load demand. Analysis shows that incorporating improved demand response ensures system operation within a controllable range, increasing system stability. Therefore, demand response uncertainty primarily regulates user loads in the VPP system, shaving peak loads and shifting valleys to maintain system stability.

[0175] Analysis of VPP economic and low-carbon operation under short-time scale conditions within the day considering improved demand response: VPP operation cost under each scenario. Analyze the dispatch results of each unit of the VPP system in scenario 3, as shown in the attached figure. Figure 8As shown, during the day-ahead scheduling phase, the overall trends in gas turbine output, battery charging and discharging, and VPP power purchases and sales remain unchanged during the intraday scheduling phase. In particular, deviations in VPP power purchases and sales and gas turbine output primarily occur during peak and valley output. During the intraday phase, each unit's output undergoes real-time correction and adjustment to meet the overall system operating objectives. During this phase, the overall intraday adjustment of energy storage is greater than that of the gas turbine and power purchases and sales, with the primary adjustment periods occurring at 6:00 AM, 12:00 PM, and 8:00 PM. This indicates that the optimization strategy incorporating demand response reduces gas turbine output, significantly reducing system costs and better aligning with the government's low-carbon operation policies. Since gas turbine output is no longer concentrated during peak demand periods, the optimized overall output distribution is more rational, alleviating system operating pressure and ensuring stable operation.

[0176] Different from Scenario 3, Scenario 4 takes into account the improved demand response, which makes the user load change significantly. Figure 9 As shown in the figure. The volatility of the output curves of each unit in the VPP system has increased significantly. Especially during periods of high electricity prices or peak electricity consumption, the main adjustment periods for resource scheduling are 10:00-14:00 and 20:00-22:00. The compensation price has also been increased according to the increase in the transfer response volume, prompting demand-side user resources to participate more actively in DR. In addition, users tend to autonomously reduce electricity consumption reasonably during peak hours and transfer flexible loads to low-price periods. After understanding the user's electricity consumption behavior, the consumption surplus valuation function is used to make the correction and adjustment of the overall unit output of the VPP system on the day before more reasonable and controllable. While meeting the overall operation goals of the system, the fluctuations of the entire VPP system are further stabilized, realizing the reasonable operation of the VPP system in an economical and low-carbon manner.

[0177] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

Claims

1. A method for controlling the output of a virtual power plant unit considering improved demand response, characterized in that: include: Step S1: collecting historical data of the output of each unit and each influencing factor, and predicting the output of each unit and each influencing factor for the next day based on the historical data; Step S2: Substitute the predicted output of each unit for the next day and the first predicted data of each influencing factor into a pre-configured day-ahead optimization model, wherein the optimization objective of the day-ahead optimization model is to minimize the sum of the operation and maintenance costs of wind power, photovoltaic power, and energy storage, the operation cost of the gas turbine, the number of intervals for purchasing and selling electricity costs, the number of intervals for carbon emission costs, and the demand response cost, wherein the minimum demand response cost is obtained based on the consumer surplus value, and the carbon emission cost is obtained based on the carbon intensity; Step S3: using the day-ahead optimization model to solve and obtain the output time series of each unit for the next day, wherein each element in the time series corresponds to each time period of the next day; Step S4: Based on the second prediction data of each influencing factor and in combination with the obtained output time series of each unit, intraday optimization is performed to obtain the output of each unit.

2. A virtual power plant unit output control method considering improved demand response according to claim 1, characterized in that: The demand response cost is: Where: C DR,t is the demand response cost, N T is the number of moments in the time period T, N load is the total number of nodes, π i,t is the consumption surplus value of electricity consumption at node i at time t, is the compensation electricity price of node i at time t, is the load that the i-th node is willing to participate in demand response at time t.

3. A virtual power plant unit output control method considering improved demand response according to claim 2, characterized in that: The load that the i-th user is willing to participate in demand response is: Where: P load_i0 is the demand load before implementing time-of-use electricity pricing, λ PV is the load response transfer rate during peak and valley periods, λ PF is the load response transfer rate between peak and normal periods, λ FV is the load response transfer rate during the off-peak period, is the average load during peak hours, The average load during normal periods.

4. The method for controlling the output of a virtual power plant unit considering improved demand response according to claim 2, characterized in that: The carbon emission costs are specifically: Where: C ce,T is the cost of carbon emissions, is the unit carbon price, λ E is the carbon price base price, E Gc,i (T) is the net CO2 emission of the i-th gas turbine in time period T, E Gb,i (T) is the carbon quota of the i-th gas turbine in time period T, is the comprehensive carbon intensity, λ Gi is the carbon emission benchmark per unit of electricity, P Gi (t) is the net CO2 emission of the i-th gas turbine at time t.

5. A virtual power plant unit output control method considering improved demand response according to claim 4, characterized in that: The comprehensive carbon intensity is specifically: in: is the carbon intensity of the generator corresponding to node i, is the output of the generator corresponding to node i, is the carbon intensity of branch k, f t,k is the unit carbon emission of branch k, is a collection of transmission lines.

6. The method for controlling the output of a virtual power plant unit considering improved demand response according to claim 2, characterized in that: The step S4 comprises: Step S4-1: Select the first time period of the day as the current proposed intraday optimization period; Step S4-2: Obtaining second prediction data of various influencing factors of the current proposed intraday optimization period based on historical data prediction; Step S4-3: Based on the obtained output time series of each unit, extract the processed value corresponding to the current proposed intraday optimization period as the initial output value of each unit, and substitute the initial output value of each unit and the second predicted data of each influencing factor into the intraday optimization model, wherein the optimization objective of the intraday optimization model is to minimize the sum of the intraday incentive electricity price demand response cost, the output deviation penalty within the T period of the day, the intraday operation and maintenance costs of wind power, photovoltaic power and energy storage, the intraday gas turbine operation cost, the number of intraday electricity purchase and sales cost intervals, and the number of intraday carbon emission cost intervals; Step S4-4: Use the intraday optimization model to solve and obtain the output of each unit in the current quasi-intraday optimization period, and determine whether the traversal of all intraday periods is completed. If so, exit; otherwise, select the next intraday period as the current quasi-intraday optimization period and return to step S4-2.

7. A virtual power plant unit output control method considering improved demand response according to claim 6, characterized in that: The intra-day improved demand response cost is: in: is the demand response cost of the intraday incentive electricity price type, Time-of-use electricity price, is the load before demand response, As the basic electricity price, To encourage subsidies for electricity prices, is the load after demand response, Δt in Intraday time scale.

8. The method for controlling the output of a virtual power plant unit considering improved demand response according to claim 6, characterized in that: The output deviation penalty within the T period of the day is the sum of the output deviations of all units.

9. A charging pile plug-and-charge function test device, comprising a memory, a processor, and a program stored in the memory, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 8 is implemented.

10. A storage medium having a program stored thereon, characterized in that: When the program is executed, the method according to any one of claims 1 to 8 is implemented.

Images