Virtual power plant resource aggregation model considering multiple time scales and cross nodes

Abstract

The invention belongs to the technical field of virtual power plant resource aggregation, and particularly relates to a virtual power plant resource aggregation model considering multiple time scales and cross nodes. Comprising the following steps: S1, establishing a net load settlement model; s2, establishing an aggregator operation strategy model; s3, establishing a market subject operation strategy behavior model; s4, establishing a lower-layer clearing model, and determining network constraints; in order to solve the problem that a current market mechanism is still based on a traditional metering settlement mode and does not actively adapt to the requirements of an emerging market subject, the invention proposes an aggregation and operation strategy of an aggregator in consideration of a net load metering settlement method, and constructs a power market settlement mode based on the net load. Comparing and analyzing the operation strategy and the operation effect of the virtual power plant under the net load settlement mode and the incentive mechanism; and comparing and analyzing the influence of the quotation strategies of other market subjects under the finite rationality on the operation strategy of the virtual power plant after the market is further developed.

Description

Technical Field

[0001] The present invention belongs to the technical field of virtual power plant resource aggregation, and in particular relates to a virtual power plant resource aggregation model that considers multiple time scales and cross-nodes. Background Art

[0002] At present, China's installed capacity of new energy is increasing rapidly. The research problem has gradually changed from how to promote the increase of installed capacity of new energy to how to promote the consumption of new energy and promote the economically sustainable operation of new energy. The transformation of the new power system is imminent, and the relevant market mechanism also needs to be further improved. The advancement of China's power market construction and the innovation of production technology and operation methods have given rise to the concept of new market entities. The construction and development of China's power market has promoted the emergence of new distribution and sales of electricity business and promoted the process of emerging market entities participating in the market. However, the current supporting market mechanism is not yet perfect, and the new advantages of emerging market entities have not been fully explored.

[0003] Current research focuses on two aspects: one is the design of mechanisms suitable for emerging market players, including fixed rate incentive mechanisms, performance-based incentive methods, and dynamic incentive mechanisms. The other is the analysis of operating strategies of emerging market players. Most of the above discussions revolve around the technical characteristics of emerging market players, focusing on utilizing the differences in production costs, output characteristics, supply and demand situations, etc. of heterogeneous resources, and reducing the uncertainty of output and price of a single resource in the production process by coupling the raw material market or production process, achieving control over production risks and production costs, or providing value-added services through information collection to obtain additional benefits. In this process, new market players play a role in promoting the optimal allocation of resources across markets, time and space.

[0004] However, most of the current discussions start from the technical characteristics, and the analysis of the operational characteristics brought by the structure of the virtual power plant itself is not in-depth enough, and the structural advantages of the virtual power plant have not been fully explored. In the scenario of building a virtual power plant, existing research has failed to fully utilize the structural advantages of the virtual power plant. The mechanism design and operation strategy of incentivizing virtual power plants are the key to improving the flexibility and stability of the power system. Future research should further explore more effective incentive mechanisms and operation strategies to cope with the changing power market demand and technical challenges. Summary of the invention

[0005] The purpose of the present invention is to overcome the deficiencies of the prior art and to provide a virtual power plant resource aggregation model that considers multiple time scales and cross-nodes.

[0006] The object of the present invention is achieved by: a virtual power plant resource aggregation model considering multiple time scales and across nodes, comprising the following steps:

[0007] S1. Establish a net load settlement model;

[0008] S2. Establish aggregator operation strategy model;

[0009] S3. Establish a market player operational strategy behavior model;

[0010] S4. Establish a lower-level clearing model and determine network constraints.

[0011] Furthermore, the step S1 of establishing a net load settlement model includes:

[0012] At the aggregation node in a certain period of time, the fee paid by the aggregator to the aggregated resources is

[0013] R dis,t =P g,t ×Q g,t -P c,t ×Q c,t (1)

[0014] Among them, R dis,t P represents the fee paid by the aggregator to the aggregated resources at time t; g,t is the power generation of the virtual power plant in time period t; Q g,t is the node marginal price in time period t; P c,t is the power consumption of the virtual power plant in time period t; Q c,t is the regional price in time period t;

[0015] The fee paid by the market operator to the aggregator is

[0016]

[0017] Among them, R con,t The fees paid by the market operator to the aggregator, from which the aggregator earns the difference

[0018]

[0019] Furthermore, the step S2 of establishing the aggregator operation strategy model includes:

[0020] By predicting the next-day output of aggregated renewable energy, the user's electricity consumption behavior is adjusted to balance the load demand entering the spot market with the aggregated renewable energy as much as possible, and to reduce the net load in each period as much as possible.

[0021] min|P t -α×load t -load de,t +load in,t | (4)

[0022] Among them, P tis the power generation of the aggregated resource at time t, α is the total load rate of a specific node at time t, load t is the total load of the node at time t, load de,t is the demand response reduction at time node t, load in,t is the increase in demand response at time node t; in terms of quotation strategy, considering that the current aggregator is small in scale, it accepts spot market prices and does not quote volume; settlement is made based on the actual output of each time period; the aggregator and the power grid use the net load method. When transmitting electricity to the outside during a certain period of time, settlement is made based on the electricity price of the node where the aggregated power source is located; when purchasing electricity from the power grid during a certain period of time, settlement is made based on the average electricity price of users.

[0023] Furthermore, the step S3 of establishing the market subject operation strategy behavior model includes:

[0024] Modeling the bidding behavior of thermal power units

[0025] When analyzing the quotes at a certain point in time, thermal power units cannot judge the specific congestion situation of the system at that time and can only consider the quote strategy based on the situation of that period in the historical data and their own operating conditions;

[0026] r t =λ t ×P t -(a×P t 2 +b×P t +c) (5)

[0027] Among them, r t is the revenue of the thermal power unit at time t, λ t is the price of aggregated resources at time t, a, b, c are the cost coefficients of thermal power units; the direction of improvement of each segment’s quotation is the direction of the derivative of the total revenue for that segment’s quotation. For simplicity, when the derivative is greater than 0, the quotation is increased, and the step size is determined by the derivative size, i.e.

[0028]

[0029] Among them, λ 0,t is the initial declared price at time t; is the declared price of the kth section of the thermal power unit at time t; ε is the increased quotation step; the predicted node electricity price formula is as follows;

[0030]

[0031]

[0032] in, c is the starting declared price of the kth section of the thermal power unit at time t;k Predict the segmentation coefficient for thermal power bidding; is the output of the kth segment at time t; p k is the total output at time t; the predicted electricity price is a linear combination of the quotations of thermal power units, and its coefficient c k It includes two parts. One reflects the supply and demand situation, which is determined by the ratio of the actual output of the quoted volume in this section to the available capacity of the quoted volume in this section. The function of this part is to exclude the electricity volume in the quoted section that has not been traded because the quoted volume is higher than the final marginal node price. The other part simulates the probability of becoming the marginal price, which is expressed by the degree of closeness between the quoted volume in this section and the marginal electricity price after the last clearing. Taking into account the situation that the quoted volume may be much higher than the actual clearing price, the denominator is taken as the larger of the two prices to ensure that the fraction is always between [0,1]. The closer the historical quoted volume is to the historical clearing price, the larger the absolute value of this part is. At the same time, the square of this part is taken instead of the linear value to further increase the degree of differentiation.

[0033] The unit's bid will not only affect the clearing price, but also the clearing quantity. The unit's forecast clearing quantity formula is as follows:

[0034]

[0035] Among them, P min is the minimum output of the thermal power units that have been put into operation; d i,t (i=1,2,3) is the unit predicted clearing power section coefficient; the power part coefficient d i,t It is related to the electricity price. When the bid price for this section is less than the clearing price, the electricity volume for this section is traded. i,t =1, when the bid price for this section is greater than the clearing price, the electricity volume for this section will not be traded, d i,t =0, when the bid price is equal to the clearing price, d i,t is a value between [0,1], indicating that the electricity quantity in this section is the marginal bidding section; the above function is approximately fitted as a piecewise function, and its derivative is calculated. The derivative is not 0 only when it is close to the marginal electricity price, and it is approximately a coefficient that is approximately 1 when the bid is close to the marginal electricity price;

[0036]

[0037] Considering that this part of the derivative is related to whether the quotation is close to the clearing price, the coefficient c is used k The structure of the second half is measured, and the derivative of this part is linearized to obtain

[0038]

[0039] Thus, we can get the derivative of the predicted single-period revenue of thermal power units with respect to the quotations of each period;

[0040]

[0041] The derivative for the first quote is

[0042]

[0043] The derivatives of the second and third quotations are

[0044]

[0045] The aggregator decides whether to continue to increase the quotation level based on the derivative of the three-segment quotations in each period until the quotation cap is reached.

[0046] Furthermore, the step S4 establishes a lower-level clearing model and determines network constraints including:

[0047] The lower model is the system clearing model, which aims to maximize social benefits while minimizing wind and solar power curtailment. Without considering load quotation, maximizing social benefits is equivalent to minimizing the cost of purchasing various types of power sources.

[0048]

[0049] Among them, F is the total cost of purchasing electricity for the system; Provide quotation for each section of thermal power unit; Provide quotes for each section of the energy storage device; Provide quotes for each section of the wind turbine; Clear the electricity consumption of thermal power units in sections; Clearing electricity for energy storage devices in stages; Clear the electricity consumption of wind turbines in sections; To abandon the amount of wind and solar power; is the load loss per unit time; puni1 is the unit penalty for wind and solar power abandonment; puni2 is the unit penalty for load loss;

[0050] 1) Output constraints of thermal power units

[0051] Output constraint is

[0052]

[0053] in, It is the upper and lower limits of the output of a single quotation section of a thermal power unit;

[0054] Constraints on segmented reporting

[0055]

[0056] in, is the total output of thermal power unit gi in period t;

[0057] Total output constraint

[0058]

[0059] in, The upper and lower limits of the total output of the thermal power unit at time t;

[0060] Hill climbing constraint

[0061]

[0062] Among them, R gi is the maximum climbing amount of the thermal power unit;

[0063] 2) The output constraint of the energy storage device is

[0064]

[0065] in, It is a 0-1 variable, indicating that the device cannot be charged and discharged at the same time; is the total discharge capacity of the energy storage device at time t; is the total charge of the energy storage device at time t; P e is the amount of charge or discharge allowed by the energy storage device at time t;

[0066] The capacity constraint is

[0067]

[0068] Se×SOC min ≤E t ≤Se×SOC max (26)

[0069] Among them, E t is the capacity of the energy storage device at time t; Se0 is the initial capacity of the energy storage device; SOC min ,SOC max The upper and lower limits of the state of charge of the energy storage device;

[0070] 3) Wind power output constraint is

[0071]

[0072] in, is the total output of the wind turbine at time t; pw t is the total output of wind resources at time t;

[0073] The wind and solar curtailment constraints are

[0074]

[0075] 4) Load loss constraint is

[0076]

[0077] in, is the total load of node i at time t;

[0078] 5) Network constraints

[0079]

[0080] in, is the total output of the photovoltaic unit at time t, is the total output of lighting resources at time t; The power injected into the system node at time t; M g M is the node location matrix of the thermal power unit connected to the power grid; b M is the node location matrix of the energy storage device connected to the power grid; w M is the node location matrix of wind turbines connected to the power grid; s is the node position matrix of the photovoltaic unit connected to the power grid; B is the impedance matrix of the network; θ is the node phase angle matrix; θ i ,θ j is the phase angle between node i and node j; is the active power transmitted between two nodes; are the upper and lower limits of the line transmission power. Since this model involves energy storage devices, it is currently difficult to find the duality of the lower layer model and convert it into a single layer problem. Therefore, an iterative approach is used to gradually approach the optimal solution.

[0081] The beneficial effects of the present invention are as follows: a virtual power plant resource aggregation model considering multiple time scales and across nodes of the present invention includes steps S1, establishing a net load settlement model; S2, establishing an aggregator operation strategy model; S3, establishing a market entity operation strategy behavior model; S4, establishing a lower-level clearing model and determining network constraints; through the above steps, the present invention aims to address the problem that the current market mechanism is still based on the traditional metering and settlement method and has not actively adapted to the needs of emerging market entities. It proposes to consider the aggregation and operation strategy of aggregators under the net load metering and settlement method, construct a settlement method based on net load in the electricity market, and compare and analyze the operation strategies and operation effects of virtual power plants under the net load settlement method and incentive mechanism; compare and analyze the impact of the bidding strategies of other market entities under the limited rationality on the operation strategies of virtual power plants after further development of the market. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0083] Figure 1 This is a diagram of the resource settlement framework under the net load settlement method.

[0084] Figure 2 The following is a diagram of VPP resource aggregation under two different incentive mechanisms.

[0085] Figure 3 This is a diagram of VPP resource call status under different incentive mechanisms.

[0086] Figure 4 This is a comparison chart of marginal electricity prices for four scenario nodes. DETAILED DESCRIPTION

[0087] The present invention will be further described below in conjunction with the accompanying drawings.

[0088] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0089] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0090] A virtual power plant resource aggregation model considering multiple time scales and cross-nodes of the present invention comprises the following steps:

[0091] S1. Establish a net load settlement model;

[0092] S2. Establish aggregator operation strategy model;

[0093] S3. Establish a market player operational strategy behavior model;

[0094] S4. Establish a lower-level clearing model and determine network constraints.

[0095] like Figure 1As shown in the figure, under the net load settlement method, the electricity market, virtual power plants and resources form a two-level resource settlement framework; within the virtual power plant, the metering and settlement of power generation and consumption between the aggregator and the aggregated resources still follow the traditional method, and each sets up a metering instrument at the access point to measure the actual power generation and consumption; and one or more metering nodes need to be determined between the electricity market and the virtual power plant. This node is the node where the aggregated resources of the virtual power plant are actually connected to the power grid. The only difference is that the online power of the virtual power plant in each period is measured according to the actual online power of the node, that is, the value obtained by subtracting the power consumption from the power generation. When the value is positive, the virtual power plant supplies electricity to the grid, and the price is calculated according to the marginal electricity price of the node; when the value is negative, the virtual power plant receives electricity, and the price is calculated according to the regional user node price.

[0096] Furthermore, the step S1 of establishing a net load settlement model includes:

[0097] At the aggregation node in a certain period of time, the fee paid by the aggregator to the aggregated resources is

[0098] R dis,t =P g,t ×Q g,t -P c,t ×Q c,t (1)

[0099] Among them, R dis,t P represents the fee paid by the aggregator to the aggregated resources at time t; g,t is the power generation of the virtual power plant in time period t; Q g,t is the node marginal price in time period t; P c,t is the power consumption of the virtual power plant in time period t; Q c,t is the regional price in time period t;

[0100] The fee paid by the market operator to the aggregator is

[0101]

[0102] Among them, R con,t The fees paid by the market operator to the aggregator, from which the aggregator earns the difference

[0103]

[0104] Furthermore, the step S2 of establishing the aggregator operation strategy model includes:

[0105] After aggregating renewable resources and loads, the aggregator invests in transforming terminal equipment and building an operation platform to enable the load to have a certain regulation capability. By predicting the output of the aggregated renewable energy the next day, the user's electricity consumption behavior is adjusted to balance the load demand entering the spot market with the aggregated renewable energy as much as possible, and to reduce the net load in each period as much as possible.

[0106] min|P t -α×load t -load de,t +load in,t | (4)

[0107] Among them, P t is the power generation of the aggregated resource at time t, α is the total load rate of a specific node at time t, and loadd t is the total load of the node at time t, load de,t is the demand response reduction at time node t, load in,t is the increase in demand response at time node t; in terms of quotation strategy, considering that the current aggregator is small in scale, it accepts spot market prices and does not quote volume; settlement is made based on the actual output of each time period; the aggregator and the power grid use the net load method. When transmitting electricity to the outside during a certain period of time, settlement is made based on the electricity price of the node where the aggregated power source is located; when purchasing electricity from the power grid during a certain period of time, settlement is made based on the average electricity price of users.

[0108] Furthermore, the step S3 of establishing the market subject operation strategy behavior model includes:

[0109] Considering the continuous increase in installed capacity of renewable energy in the future, the increasing number of regulatory resources required by the system, the further active spot market, and more flexible price fluctuations, flexibly adjustable resources such as thermal power units can take advantage of their own advantages and raise spot prices in some periods. This behavior will lead to an overall increase in user electricity prices. At this time, the aggregator aggregates renewable resources at the power node, reduces the price of the overall power generation node of the virtual power plant, aggregates loads of the same scale, and tries to maintain the balance of power generation and consumption through load management.

[0110] Modeling the bidding behavior of thermal power units

[0111] When analyzing the quotes at a certain point in time, thermal power units cannot judge the specific congestion situation of the system at that time and can only consider the quote strategy based on the situation of that period in the historical data and their own operating conditions;

[0112] r t =λ t ×P t -(a×P t 2 +b×Pt +c) (5)

[0113] Among them, r t is the revenue of the thermal power unit at time t, λ t is the price of aggregated resources at time t, a, b, c are the cost coefficients of thermal power units; the direction of improvement of each segment’s quotation is the direction of the derivative of the total revenue for that segment’s quotation. For simplicity, when the derivative is greater than 0, the quotation is increased, and the step size is determined by the derivative size, i.e.

[0114]

[0115] Among them, λ 0,t is the initial declared price at time t; is the price reported by the kth section of the thermal power unit at time t; ε is the increase in the quotation step; since the unit only knows its own output and quotation, it can only predict the clearing result of a certain period in the future based on its current situation and historical data, so its prediction is modeled. The formula for the predicted node electricity price is as follows;

[0116]

[0117] in, c is the starting declared price of the kth section of the thermal power unit at time t; k Predict the segmentation coefficient for thermal power bidding; is the output of the kth segment at time t; p k is the total output at time t; the predicted electricity price is a linear combination of the quotations of thermal power units, and its coefficient c k It includes two parts. One reflects the supply and demand situation, which is determined by the ratio of the actual output of the quoted volume in this section to the available capacity of the quoted volume in this section. The function of this part is to exclude the electricity volume in the quoted section that has not been traded because the quoted volume is higher than the final marginal node price. The other part simulates the probability of becoming the marginal price, which is expressed by the degree of closeness between the quoted volume in this section and the marginal electricity price after the last clearing. Taking into account the situation that the quoted volume may be much higher than the actual clearing price, the denominator is taken as the larger of the two prices to ensure that the fraction is always between [0,1]. The closer the historical quoted volume is to the historical clearing price, the larger the absolute value of this part is. At the same time, the square of this part is taken instead of the linear value to further increase the degree of differentiation.

[0118] The unit's bid will not only affect the clearing price, but also the clearing quantity. The unit's forecast clearing quantity formula is as follows:

[0119]

[0120] Among them, P min is the minimum output of the thermal power units that have been put into operation; d i,t(i=1,2,3) is the unit predicted clearing power segment coefficient; without considering the unit start and stop, the output lower limit does not change with the quotation and is not within the adjustable range, so this part is considered as a constant and removed. The remaining part is a linear function of each price and quantity segment. The coefficient d of the power part i,t It is related to the electricity price. When the bid price for this section is less than the clearing price, the electricity volume for this section is traded. i,t =1, when the bid price for this section is greater than the clearing price, the electricity volume for this section will not be traded, d i,t =0, when the bid price is equal to the clearing price, d i,t is a value between [0,1], indicating that the electricity quantity in this section is the marginal bidding section; the above function is approximately fitted as a piecewise function, and its derivative is calculated. The derivative is not 0 only when it is close to the marginal electricity price, and it is approximately a coefficient that is approximately 1 when the bid is close to the marginal electricity price;

[0121]

[0122] Considering that this part of the derivative is related to whether the quotation is close to the clearing price, the coefficient c is used k The structure of the second half is measured, and the derivative of this part is linearized to obtain

[0123]

[0124] Thus, we can get the derivative of the predicted single-period revenue of thermal power units with respect to the quotations of each period;

[0125]

[0126] The derivative for the first quote is

[0127]

[0128] The derivatives of the second and third quotations are

[0129]

[0130] The aggregator decides whether to continue to increase the quotation level based on the derivative of the three-segment quotations in each period until the quotation cap is reached.

[0131] Furthermore, the step S4 establishes a lower-level clearing model and determines network constraints including:

[0132] The lower model is the system clearing model, which aims to maximize social benefits while minimizing wind and solar power curtailment. Without considering load quotation, maximizing social benefits is equivalent to minimizing the cost of purchasing various types of power sources.

[0133]

[0134] Among them, F is the total cost of purchasing electricity for the system; Provide quotation for each section of thermal power unit; Provide quotes for each section of the energy storage device; Provide quotes for each section of the wind turbine; Clear the electricity consumption of thermal power units in sections; Clearing electricity for energy storage devices in stages; Clear the electricity consumption of wind turbines in sections; To abandon the amount of wind and solar power; is the load loss per unit time; puni1 is the unit penalty for wind and solar power abandonment; puni2 is the unit penalty for load loss;

[0135] 1) Output constraints of thermal power units

[0136] Output constraint is

[0137]

[0138] in, It is the upper and lower limits of the output of a single quotation section of a thermal power unit;

[0139] Constraints on segmented reporting

[0140]

[0141] in, is the total output of thermal power unit gi in period t;

[0142] Total output constraint

[0143]

[0144] in, The upper and lower limits of the total output of the thermal power unit at time t;

[0145] Hill climbing constraint

[0146]

[0147] Among them, R gi is the maximum climbing amount of the thermal power unit;

[0148] 2) The output constraint of the energy storage device is

[0149]

[0150] in, It is a 0-1 variable, indicating that the device cannot be charged and discharged at the same time; is the total discharge capacity of the energy storage device at time t; is the total charge of the energy storage device at time t; P e is the amount of charge or discharge allowed by the energy storage device at time t;

[0151] The capacity constraint is

[0152]

[0153] Se×SOC min ≤E t ≤Se×SOC max (26)

[0154] Among them, E t is the capacity of the energy storage device at time t; Se0 is the initial capacity of the energy storage device; SOC min ,SOC max The upper and lower limits of the state of charge of the energy storage device;

[0155] 3) Wind power output constraint is

[0156]

[0157] in, is the total output of the wind turbine at time t; pw t is the total output of wind resources at time t;

[0158] The wind and solar curtailment constraints are

[0159]

[0160] 4) Load loss constraint is

[0161]

[0162] in, is the total load of node i at time t;

[0163] 5) Network constraints

[0164]

[0165] in, is the total output of the photovoltaic unit at time t, is the total output of lighting resources at time t; The power injected into the system node at time t; M g M is the node location matrix of the thermal power unit connected to the power grid; b M is the node location matrix of the energy storage device connected to the power grid; w M is the node location matrix of wind turbines connected to the power grid; s is the node position matrix of the photovoltaic unit connected to the power grid; B is the impedance matrix of the network; θ is the node phase angle matrix; θ i ,θ j is the phase angle between node i and node j; is the active power transmitted between two nodes; are the upper and lower limits of the line transmission power. Since this model involves energy storage devices, it is currently difficult to find the duality of the lower layer model and convert it into a single layer problem. Therefore, an iterative approach is used to gradually approach the optimal solution.

[0166] The model of the present invention is a two-layer optimization model. The upper layer includes two major market players. One is the traditional thermal power units, which use their market advantages to make strategic quotations under the premise of limited rationality to obtain the ultimate maximum benefit; the other is the virtual power plant, which obtains higher benefits through two-stage resource aggregation and optimized scheduling. The first stage is based on the estimation of spot prices, looking for more advantageous nodes and optimizing the types and scales of aggregated resources. The second stage is to schedule resources in the day-ahead market to obtain greater benefits.

[0167] The lower model is cleared with the goal of minimizing the system operating cost. Since this model involves energy storage devices, it is currently difficult to find the duality of the lower model and convert it into a single-layer problem. Therefore, an iterative approach is used to gradually approach the optimal solution.

[0168] The following is an example of a specific scenario, using a 6-node network to simulate a regional power grid, with transmission capacity restrictions between different nodes. Nodes 1 and 2 are connected to thermal power units, node 3 is connected to a centralized energy storage device, node 4 is connected to a centralized wind turbine, and node 5 is connected to a distributed photovoltaic unit. Each node is connected to a load, and each has a different load curve.

[0169] The wind power, photovoltaic power and load curves of each node are fitted using historical data to obtain the probability density of output and load at each time point, and the expected output and load curves under normal operating conditions are obtained.

[0170] Four scenarios are set: Scenario 1 does not consider the strategic bids of thermal power units and the resource aggregation and optimal scheduling of virtual power plants; Scenario 2 only considers the strategic bids of thermal power units and does not consider the resource aggregation and optimal scheduling of virtual power plants; Scenario 3 considers the strategic bids of thermal power units, and virtual power plants aggregate and optimize resources under the incentive mechanism of fixed rates; Scenario 4 considers the strategic bids of thermal power units, and virtual power plants aggregate and optimize resources under the net load settlement method. Compare the composition and benefits of virtual power plants and the electricity prices at system clearing nodes in each scenario.

[0171] (1) Virtual power plant operation and revenue analysis

[0172] The resource aggregation and calling of the virtual power plant under the two incentive mechanisms are calculated in stages.

[0173] In the first stage, the virtual power plant optimizes the type and scale of resource aggregation based on the predicted electricity price and incentive mechanism. In the second stage, after considering the volatility of wind power and load forecasts in the spot market stage, the aggregated resources are optimized and dispatched to achieve the goal of maximizing benefits. The final calculation results are shown in Table 1-2 and Figure 2 shown.

[0174] Table 1 VPP aggregation resources in scenario 3

[0175] Scene 3 Node 1 Node 2 Node 3 Node 4 Node 5 Node 6 power supply 0 0.1 / 0.0652 / / load 0 0 0 0 0 0.5158

[0176] Table 2: Scenario 4 VPP aggregation resources

[0177] Scene 4 Node 1 Node 2 Node 3 Node 4 Node 5 Node 6 power supply 0.0593 0.1 / 0.0652 / / load 0 0.0183 0.1572 0 0.1146 0.1298

[0178] Comparing Table 1 with Table 2, under different incentive mechanisms, the resource aggregation strategies of virtual power plants vary greatly. When the scale of aggregated wind power resources is certain, the load resources aggregated by the virtual power plant in scenario 3 are concentrated at node 6, while the load resources aggregated by the virtual power plant in scenario 4 are dispersed across various nodes. Figure 3 The left and right figures in the first row show the expected aggregation and calling effects of virtual power plants. The left figure shows that under the incentive of fixed rates, virtual power plants can always maintain a balance between power supply and demand, with a net load of zero. The right figure shows that under the net load settlement mechanism, virtual power plants maintain a balance between power supply and demand during system congestion periods, and the normal operation of aggregated resources is not affected in other periods. The second row shows the actual response of virtual power plant aggregated resources during the spot market period. The left figure shows that although the response can obtain fixed income and the virtual power plant aggregates adjustable resources, the user load and thermal power units have no motivation to change the original operation arrangement to maintain the overall balance of power supply and demand. The right figure shows that when the system is blocked and the electricity prices of each node are different, the virtual power plant will actively call the aggregated thermal power units, plus the previously aggregated wind power resources, to keep the load consistent as much as possible, and maintain the virtual power plant net load at zero. As the scale and number of virtual power plants in this scenario gradually increase, they will help cope with the fluctuations of renewable energy in the region and gradually form regional self-sufficiency.

[0179] In terms of revenue, since the aggregated resources of the virtual power plant in scenario 3 did not respond to the call, the adjustment revenue was 0. In contrast, in scenario 4, the aggregated resources of the virtual power plant responded to the call, and the adjustment revenue for the whole day was 3,789 yuan.

[0180] (2) Electricity price at system clearing node

[0181] In the four scenarios, the system is congested during some periods, and the marginal electricity prices of different nodes are different. Figure 4As shown in the figure, after considering the bidding strategy of thermal power units in scenario 2, the overall electricity price level of the system has increased. The electricity price levels of scenario 2 and scenario 3 are relatively similar. After considering the net load settlement method, the different periods of node marginal electricity prices in scenario 4 have decreased, and the electricity price level in the corresponding period has decreased to a certain extent, indicating that the resource scheduling optimization of the virtual power plant under this incentive mechanism has a certain stabilizing effect on the spot price.

[0182] In summary, a virtual power plant resource aggregation model considering multiple time scales and across nodes of the present invention includes steps S1, establishing a net load settlement model; S2, establishing an aggregator operation strategy model; S3, establishing a market entity operation strategy behavior model; S4, establishing a lower-level clearing model, and determining network constraints; through the above steps, the present invention aims to address the problem that the current market mechanism is still based on the traditional metering and settlement method and has not actively adapted to the needs of emerging market entities. It proposes to consider the aggregation and operation strategy of aggregators under the net load metering and settlement method, construct a settlement method based on net load in the power market, and compare and analyze the operation strategies and operation effects of virtual power plants under the net load settlement method and incentive mechanism; and compare and analyze the impact of the quotation strategies of other market entities under the limited rationality on the operation strategies of virtual power plants after further development of the market.

[0183] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. A virtual power plant resource aggregation model considering multiple time scales and across nodes, characterized in that: The following steps are involved: S1. Establish a net load settlement model; S2. Establish aggregator operation strategy model; S3. Establish a market player operational strategy behavior model; S4. Establish a lower-level clearing model and determine network constraints.

2. A virtual power plant resource aggregation model considering multiple time scales and across nodes as claimed in claim 1, characterized in that: The step S1 of establishing a net load settlement model includes: At the aggregation node in a certain period of time, the fee paid by the aggregator to the aggregated resources is R dis,t =P g，t ×Q g,t -P c,t ×Q c，t (1) Among them, R dis,t P represents the fee paid by the aggregator to the aggregated resources at time t; g,t is the power generation of the virtual power plant in time period t; Q g,t is the node marginal price in time period t; P c,t is the power consumption of the virtual power plant in time period t; Q c,t is the regional price in time period t; The fee paid by the market operator to the aggregator is Among them, R con,t The fees paid by the market operator to the aggregator, from which the aggregator earns the difference 3. A virtual power plant resource aggregation model considering multiple time scales and across nodes as claimed in claim 1, characterized in that: The step S2 of establishing the aggregator operation strategy model includes: By predicting the next-day output of aggregated renewable energy, the user's electricity consumption behavior is adjusted to balance the load demand entering the spot market with the aggregated renewable energy as much as possible, and to reduce the net load in each period as much as possible. min|P t -α×load t -load de，t +load in，t | (4) Among them, P t is the power generation of the aggregated resource at time t, α is the total load rate of a specific node at time t, load t is the total load of the node at time t, load de,t is the demand response reduction at time node t, load in,t is the increase in demand response at time node t; the aggregator and the power grid use the net load method. When supplying electricity to the outside during a period of time, the settlement is based on the electricity price of the node where the aggregated power source is located. When purchasing electricity from the power grid during a period of time, the settlement is based on the average electricity price of users.

4. A virtual power plant resource aggregation model considering multiple time scales and across nodes as claimed in claim 1, characterized in that: The step S3 of establishing the market subject operation strategy behavior model includes: Modeling the bidding behavior of thermal power units When analyzing the quotes at a certain point in time, thermal power units cannot judge the specific congestion situation of the system at that time and can only consider the quote strategy based on the situation of that period in the historical data and their own operating conditions; r t =λ t ×P t -(a×P t 2 +b×P t +c) (5) Among them, r t is the revenue of the thermal power unit at time t, λ t is the price of aggregated resources at time t, a, b, c are the cost coefficients of thermal power units; the direction of improvement of each segment’s quotation is the direction of the derivative of the total revenue for that segment’s quotation. For simplicity, when the derivative is greater than 0, the quotation is increased, and the step size is determined by the derivative size, i.e. Among them, λ 0,t is the initial declared price at time t; is the declared price of the kth section of the thermal power unit at time t; ε is the increased quotation step; the predicted node electricity price formula is as follows; in, c is the starting declared price of the kth section of the thermal power unit at time t; k Predict the segmentation coefficient for thermal power bidding; is the output of the kth segment at time t; p k is the total output at time t; the predicted electricity price is a linear combination of the quotations of thermal power units, and its coefficient c k It includes two parts. One part reflects the supply and demand situation, which is determined by the ratio of the actual output of the quoted quantity in this section to the available capacity of the quoted quantity in this section. The function of this part is to exclude the electricity quantity in the quoted section that has not been traded because the quoted quantity is higher than the final marginal node price. The other part simulates the probability of becoming the marginal price, which is expressed by the degree of closeness between the quoted quantity in this section and the marginal electricity price after the last clearing. The formula for predicting the clearing power of the unit is as follows; Among them, P min is the minimum output of the thermal power units that have been put into operation; d i,t (i=1,2,3) is the unit predicted clearing power section coefficient; the power part coefficient d i,t It is related to the electricity price. When the bid price for this section is less than the clearing price, the electricity volume for this section is traded. i,t =1, when the bid price for this section is greater than the clearing price, the electricity volume for this section will not be traded, d i,t =0, when the bid price is equal to the clearing price, d i,t is a value between [0,1], indicating that the electricity quantity in this section is the marginal bidding section; the above function is approximately fitted as a piecewise function, and its derivative is calculated. The derivative is not 0 only when it is close to the marginal electricity price, and it is approximately a coefficient that is approximately 1 when the bid is close to the marginal electricity price; Considering that this part of the derivative is related to whether the quotation is close to the clearing price, the coefficient c is used k The structure of the second half is measured, and the derivative of this part is linearized to obtain Thus, we can get the derivative of the predicted single-period revenue of thermal power units with respect to the quotations of each period; The derivative for the first quote is The derivatives of the second and third quotations are The aggregator decides whether to continue to increase the quotation level based on the derivative of the three-segment quotations in each period until the quotation cap is reached.

5. A virtual power plant resource aggregation model considering multiple time scales and across nodes as claimed in claim 1, characterized in that: The step S4 establishes a lower-level clearing model and determines network constraints including: The lower model is the system clearing model, which aims to maximize social benefits while minimizing wind and solar power curtailment. Without considering load quotation, maximizing social benefits is equivalent to minimizing the cost of purchasing various types of power sources. Among them, F is the total cost of purchasing electricity for the system; Provide quotation for each section of thermal power unit; Provide quotes for each section of the energy storage device; Provide quotes for each section of the wind turbine; Clear the electricity consumption of thermal power units in sections; Clearing electricity for energy storage devices in stages; Clear electricity for wind turbines in sections; To abandon the amount of wind and solar power; is the amount of load loss per unit time; puni1 is the unit penalty for wind and solar power abandonment; puni2 is the unit penalty for load loss; 1) Output constraints of thermal power units Output constraint is in, It is the upper and lower limits of the output of a single quotation section of a thermal power unit; Constraints on segmented reporting in, is the total output of thermal power unit gi in period t; Total output constraint in, The upper and lower limits of the total output of the thermal power unit at time t; Hill climbing constraint Among them, R gi is the maximum climbing amount of the thermal power unit; 2) The output constraint of the energy storage device is in, It is a 0-1 variable, indicating that the device cannot be charged and discharged at the same time; is the total discharge capacity of the energy storage device at time t; is the total charge of the energy storage device at time t; P e is the amount of charge or discharge allowed by the energy storage device at time t; The capacity constraint is Se×SOC min ≤E t ≤Se×SOC max (26) Among them, E t is the capacity of the energy storage device at time t; Se0 is the initial capacity of the energy storage device; SOC min ,SOC max The upper and lower limits of the state of charge of the energy storage device; 3) Wind power output constraint is in, is the total output of the wind turbine at time t; pw t is the total output of wind resources at time t; The wind and solar curtailment constraints are 4) Load loss constraint is in, is the total load of node i at time t; 5) Network constraints in, is the total output of the photovoltaic unit at time t, is the total output of lighting resources at time t; The power injected into the system node at time t; M g M is the node location matrix of the thermal power unit connected to the power grid; b M is the node location matrix of the energy storage device connected to the power grid; w M is the node location matrix of wind turbines connected to the power grid; s is the node position matrix of the photovoltaic unit connected to the power grid; B is the impedance matrix of the network; θ is the node phase angle matrix; θ i ,θ j is the phase angle between node i and node j; is the active power transmitted between two nodes; are the upper and lower limits of the line transmission power. Since this model involves energy storage devices, it is currently difficult to find the duality of the lower layer model and convert it into a single layer problem. Therefore, an iterative approach is used to gradually approach the optimal solution.

Images