Virtual power plant participation auxiliary service market strategy and system considering demand response

Abstract

The invention discloses a virtual power plant participation auxiliary service market strategy and system considering demand response, and the method comprises the steps: building a power equation of a photovoltaic system, and carrying out the control through a photovoltaic grid-connected converter; establishing a flexible load model considering demand response; establishing a virtual power plant participation auxiliary service market double-layer model considering demand response; and converting the double-layer model into a single-layer model by using a KKT condition, and completing solution. Through optimization control of the grid-connected converter of the photovoltaic system in the virtual power plant, the grid-connected converter can participate in a reactive auxiliary service market to obtain voltage regulation benefits, the voltage stability of the system in an extreme weather scene is ensured, and the benefit channel of the virtual power plant is expanded. Meanwhile, a real-time price compensation mechanism is adopted, the enthusiasm and flexibility of flexible load participating in response are promoted, and compared with a single price compensation mechanism, peak load shifting can be effectively achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric power systems, and in particular relates to a strategy and system for a virtual power plant to participate in ancillary service markets taking demand response into consideration. Background Art

[0002] At present, the ancillary services provided by third-party entities such as virtual power plants are mainly concentrated in the two major areas of peak load regulation and demand response, and their profit model mainly relies on demand response incentives and ancillary service compensation. As an occasional trading mechanism, demand response is triggered when the balance of supply and demand in the power grid faces challenges, but the uncertainty of its trading frequency limits its potential as the main profit channel for virtual power plants. At present, the mechanism for virtual power plants to participate in the electricity market is still imperfect, and it is difficult for aggregators to obtain stable income from the electricity market. Especially under extreme climatic conditions, voltage instability may occur at the terminal nodes of the distribution network. The unified dispatch of reactive resources in the power system can keep the system voltage within a reasonable range. Therefore, it is necessary to propose a new reactive compensation mechanism around the construction of the reactive ancillary service market to obtain benefits. Summary of the invention

[0003] In response to the above problems, the present invention provides a strategy and system for virtual power plants to participate in ancillary service markets that takes demand response into consideration, thereby ensuring voltage stability of the power system and improving reliability.

[0004] The technical solution of the present invention is: a virtual power plant participating in ancillary service market strategy considering demand response, comprising the following steps:

[0005] S1. Establish the power equation of the photovoltaic system and control it through the photovoltaic grid-connected converter;

[0006] S2. On the basis of step S1, a flexible load model considering demand response is established;

[0007] S3, based on step S2, establish a two-layer model of virtual power plants participating in the ancillary service market considering demand response;

[0008] S4, use KKT conditions to transform the double-layer model into a single-layer model and complete the solution.

[0009] In step S1, the power equation of the photovoltaic system includes:

[0010] The power equation of a photovoltaic system operating in constant power factor mode is:

[0011]

[0012] In the formula, and are the photovoltaic active and reactive power injection of the node, is the maximum active power output of photovoltaic system, tanθ is the tangent value of the power factor of photovoltaic system, K and T are the set of all nodes and the set of time points respectively;

[0013] The power equation of the photovoltaic system operating in the grid voltage support mode is:

[0014]

[0015] In the formula, is the capacity of the PV system.

[0016] In step S2, the flexible load includes a transferable load and a reducible load.

[0017] Flexible load model considering demand response, including:

[0018] The constraints are:

[0019]

[0020] In the formula, are the actual transferable load and the curtailable load, respectively. and are the reduction amount of the load that can be reduced and the net transfer amount of the load that can be transferred, and are the initial load that can be cut and the initial load that can be transferred, respectively;

[0021] The cost of flexible load considering demand response is:

[0022]

[0023] In the formula, C cut and C shift are the costs of load reduction and load shifting, respectively. is the unit electricity price of the flexible load, τ and ν are the unsatisfactory cost coefficients of the flexible load, and They are the compensation prices that virtual power plants provide to curtailable load and transferable load respectively.

[0024] In step S3, the two-layer model includes an upper model and a lower model, wherein the goal of the upper model is to maximize the comprehensive benefit of the virtual power plant, and the goal of the lower model is to minimize the electricity cost of the flexible load.

[0025] The objective function of the upper model is:

[0026] maxR VPP =R AS +R DR +R AV -C VPP -CE -C PV (1-11)

[0027] In the formula, R VPP is the comprehensive income of the virtual power plant, R AS , R DR , R AV They are the benefits of virtual power plants participating in the peak load ancillary service market, supplying power to flexible loads, and participating in the reactive power ancillary service market, respectively. VPP , C E , C PV They are the operating cost of the virtual power plant, the cost of purchasing electricity from higher authorities, and the penalty cost for abandoning photovoltaic power generation;

[0028] in,

[0029]

[0030] In the formula, λ PS , VF are the compensation coefficients for peak shaving and valley filling, ΔP t PS , ΔP t VF are the peak shaving and valley filling capacity declared by the virtual power plant, λ t E is the unit electricity price, λ V is the profit coefficient of participating in the reactive power auxiliary service market to provide voltage regulation, ΔV k,t is the voltage improvement of node k at time t, P t O is the amount of electricity purchased from the superior unit, λ PV is the light abandonment penalty coefficient, C ES , C DR They are the energy storage operation cost of the virtual power plant and the compensation cost of the flexible load participating in the response;

[0031] in,

[0032] C DR =C cut +C shift (1-19)

[0033] In the formula, λ ES is the loss coefficient of the battery per unit charge and discharge, and are the charging and discharging power of the energy storage system respectively.

[0034] The constraints of the upper model include:

[0035]

[0036]

[0037] In the formula, and are respectively the active and reactive power injection of the generator at node k, and are the active load and reactive load of node k respectively, and are respectively the active power and reactive power flowing on branch ki, and are the active power and reactive power flowing on branch jk, δ(k) and σ(k) are the set of post-order nodes and pre-order nodes of node k, respectively. jk and r jk are the resistance and reactance of branch jk respectively, B is the set of all branches in the system, l jk,t 、u k,t and u j,t are the square of the current amplitude of branch jk, the square of the voltage of node k and node j, is the square of the upper limit of the current amplitude, and are the squares of the lower and upper limits of the node voltage, respectively. and is the state of charge of the energy storage system and its upper limit,

[0038] and are the charge and discharge signs of the energy storage system, η ES is the charging and discharging efficiency of energy storage, and They are the charging power, discharging power and state of charge of the energy storage system at time t-1 respectively.

[0039] The objective function of the lower model is:

[0040] minC DR =C cut +C shift (1-29).

[0041] In step S4, the dual method is used to calculate the KKT conditions of the lower model and combine them into the upper model, and then transform it into a single-layer model and solve it;

[0042] Among them, the KKT condition of the lower model is:

[0043]

[0044] In the formula, and are the Lagrange multipliers respectively.

[0045] It also includes: virtual power plants provide real-time price compensation for flexible loads. When the power supply is at peak consumption, the compensation price increases are used to guide flexible load users to actively reduce or transfer loads; when the power supply is at low consumption, the compensation price decreases are used.

[0046] Considering demand response virtual power plants participating in the ancillary services market system, including:

[0047] A control module is used to establish the power equation of the photovoltaic system and control it through the photovoltaic grid-connected converter;

[0048] Load module, used to establish a flexible load model considering demand response;

[0049] A two-layer module, used to establish a two-layer model of virtual power plants participating in the ancillary service market taking into account demand response;

[0050] The solution module is used to transform the double-layer model into a single-layer model using KKT conditions and complete the solution.

[0051] In operation, the present invention firstly participates in the reactive power auxiliary service market through the optimized control of the photovoltaic grid-connected converter to cope with the problem of voltage exceeding the limit at the terminal node of the distribution network under extreme conditions; secondly, a flexible load model considering demand response and a double-layer model of the virtual power plant participating in the auxiliary service market are formed, and the KKT condition is used to convert the double-layer model into a single-layer model for solution; finally, a real-time price compensation mechanism can be used to promote the enthusiasm and flexibility of the flexible load to participate in the response, and effectively realize peak shaving and valley filling. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is the IEEE 33-node distribution network topology diagram.

[0053] Figure 2 It is the load curve and photovoltaic output curve.

[0054] Figure 3 It is the voltage curve diagram under different scenarios.

[0055] Figure 4 It is a master-slave game structure diagram.

[0056] Figure 5 It is the load curve, photovoltaic output curve, RL and TL prediction value diagram,

[0057] Figure 6 This is the voltage support effect diagram.

[0058] Figure 7 This is the effect diagram of cutting peaks and filling valleys.

[0059] Figure 8This is the demand response result diagram of the single price compensation mechanism.

[0060] Fig. 9 is the compensation mechanism demand response result diagram,

[0061] Fig.10 This is a diagram of the real-time compensation price mechanism.

[0062] Fig.11 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0063] like Fig.11 As shown in the figure, the strategy of participating in the auxiliary service market of virtual power plants considering demand response includes the following steps:

[0064] S1. Establish the power equation of the photovoltaic system and control it through the photovoltaic grid-connected converter; analyze the undervoltage problem of the terminal node of the distribution network in extreme scenarios, and provide reactive power to support voltage stability through the optimization control of the photovoltaic grid-connected converter. The virtual power plant can participate in the reactive auxiliary service market to obtain corresponding benefits.

[0065] S2. On the basis of step S1, a flexible load model considering demand response is established;

[0066] S3, based on step S2, establish a two-layer model of virtual power plants participating in the ancillary service market considering demand response;

[0067] S4, using KKT conditions to transform the double-layer model into a single-layer model and complete the solution. Through the solution, the values ​​of relevant variables and objective functions are obtained, so that the virtual power plant can reliably participate in the auxiliary service market.

[0068] The present invention aims to explore the participation of virtual power plants in the reactive power auxiliary service market, especially under extreme climatic conditions. In view of the voltage instability problem that may occur at the terminal nodes of the distribution network, the control strategy of photovoltaic grid-connected converters is optimized to provide reactive power support to maintain the voltage stability of the terminal nodes, thereby obtaining voltage regulation benefits.

[0069] The present invention explores the optimal strategy of virtual power plants in participating in the peak load regulation and reactive power auxiliary service markets to expand their profit channels. By constructing a two-layer optimization model and compensation mechanism that considers demand response, the response potential of adjustable resources is improved and the enthusiasm of users to participate in demand response is enhanced.

[0070] In step S1, the undervoltage problem of the terminal node of the distribution network in the extreme scenario is analyzed, and reactive power is provided to support voltage stability through the optimization control of the photovoltaic grid-connected converter, including:

[0071] The main operating modes of the photovoltaic system are: constant power factor operating mode and grid voltage support mode. The power equation of the photovoltaic system operating in the constant power factor mode is:

[0072]

[0073] In the formula, and are the photovoltaic active and reactive power injection of the node, is the maximum active power output of photovoltaic system, tanθ is the tangent value of the power factor of photovoltaic system, K and T are the set of all nodes and the set of time points respectively.

[0074] PV node k should satisfy formulas (1-1) and (1-2) at time t.

[0075] Since the ratio of active power to reactive power output by the PV system is constant in the constant power factor operation mode, when extreme weather conditions such as solar eclipses and sudden rains occur, the active power and reactive power output by the PV system are approximately 0. Active power imbalance can be dealt with through primary frequency regulation, demand response and energy storage systems. Reactive power imbalance is more difficult to deal with and may cause undervoltage problems at the end nodes of the system. Calculations such as Figure 1 The node voltage distribution diagram of the distribution network system is shown to illustrate the undervoltage problem under extreme weather conditions.

[0076] The system load level and the maximum photovoltaic output are as follows: Figure 2 As shown in the figure, the peak load of the system occurs at around 14:00. At this time, the photovoltaic output curve under normal circumstances is shown as the green dotted line in the figure below. If extreme weather conditions such as solar eclipses and heavy rains occur, the photovoltaic output will drop sharply. Figure 2 As shown by the blue curve, the voltage distribution diagram of node 33 of the computing system is as follows Figure 3 Shown by the blue dashed line.

[0077] Under extreme weather conditions, the photovoltaic system operating in constant power factor mode cannot provide sufficient reactive power support, resulting in undervoltage problems at the end nodes of the system, with the voltage lower than 0.95pu, affecting the stability of the system voltage.

[0078] By optimizing the control of the photovoltaic grid-connected converter, the photovoltaic system can switch between the constant power factor mode and the grid voltage support mode to provide more reactive power support. The power equation of the photovoltaic system operating in the grid voltage support mode is:

[0079]

[0080] In the formula, is the capacity of the PV system. Formulas (1-3) and (1-4) ensure that in extreme weather conditions, the PV system operating in the grid voltage support mode can provide more power support to ensure voltage stability, such as Figure 3 As shown by the green dotted line. Therefore, the virtual power plant can participate in the reactive power auxiliary service market and obtain corresponding benefits through the optimized control of the photovoltaic grid-connected converter.

[0081] In step S2, a flexible load model considering demand response is established, including:

[0082] Flexible loads can be divided into transferable loads (TL) and reducible loads (RL). The main constraints are:

[0083]

[0084] In the formula, are the actual transferable load and the curtailable load, respectively. and are the reduction amount of the load that can be reduced and the net transfer amount of the load that can be transferred, and are the initial load that can be cut and the initial load that can be transferred, respectively;

[0085] The cost of flexible load considering demand response is:

[0086]

[0087] In the formula, C cut and C shift are the costs of load reduction and load shifting, respectively. is the unit electricity price of the flexible load, τ and ν are the unsatisfactory cost coefficients of the flexible load (specifically, the quadratic coefficient and the linear coefficient), and They are the compensation prices that virtual power plants provide to curtailable load and transferable load respectively.

[0088] (1-5)-(1-10) are flexible load models, (1-5)-(1-8) describe power (as a constraint), and (1-9) and (1-10) describe cost (as an objective function).

[0089] In step S3, a two-layer model of virtual power plants participating in the ancillary service market considering demand response is established, wherein the upper layer model maximizes the comprehensive benefits of virtual power plants, and the lower layer model minimizes the electricity cost of flexible loads.

[0090] Among them, the objective function of the upper model is:

[0091] maxRVPP =R AS +R DR +R AV -C VPP -C E -C PV (1-11)

[0092] In the formula, R VPP is the comprehensive income of the virtual power plant, R AS , R DR , R AV They are the benefits of virtual power plants participating in the peak load ancillary service market, supplying power to flexible loads, and participating in the reactive power ancillary service market, respectively. VPP , C E , C PV They are the operating cost of the virtual power plant, the cost of purchasing electricity from superiors and the penalty cost for abandoning photovoltaic power generation.

[0093]

[0094] In the formula, λ PS , VF are the compensation coefficients for peak shaving and valley filling, ΔP t PS , ΔP t VF They are the peak shaving and valley filling capacity declared by the virtual power plant. is the unit electricity price, λ V is the profit coefficient of participating in the reactive power auxiliary service market to provide voltage regulation, ΔV k,t is the voltage improvement of node k at time t, P t O is the amount of electricity purchased from the superior unit, λ PV is the light abandonment penalty coefficient, C ES , C DR They are the energy storage operation cost of the virtual power plant and the compensation cost of the flexible load participating in the response, as shown below:

[0095]

[0096] C DR =C cut +C shift (1-19)

[0097] In the formula, λ ES is the loss coefficient of the battery per unit charge and discharge, and are the charging and discharging power of the energy storage system respectively.

[0098] The main constraints are:

[0099]

[0100]

[0101] In the formula, and are respectively the active and reactive power injection of the generator at node k, and are the active load and reactive load of node k respectively, and are respectively the active power and reactive power flowing on branch ki, and are the active power and reactive power flowing on branch jk, δ(k) and σ(k) are the set of post-order nodes and pre-order nodes of node k, respectively. jk and r jk are the resistance and reactance of branch jk respectively, B is the set of all branches in the system, l jk,t 、u k,t and u j,t are the square of the current amplitude of branch jk, the square of the voltage of node k and node j, is the square of the upper limit of the current amplitude, and are the squares of the lower and upper limits of the node voltage, respectively. and is the state of charge of the energy storage system and its upper limit, and are the charge and discharge signs of the energy storage system, η ES is the charging and discharging efficiency of energy storage, and They are the charging power, discharging power and state of charge of the energy storage system at time t-1 respectively.

[0102] Formulas (1-20) and (1-21) are the node active power and reactive power balance equations respectively, (1-22) and (1-23) are second-order cone relaxation power flow models, (1-24) and (1-25) are voltage and current safety constraints, and (1-26)-(1-28) are energy storage state related constraints.

[0103] The objective function of the lower model is:

[0104] minC DR =C cut +C shift (1-29)

[0105] The main constraints are formulas (1-5)-(1-10), (1-19).

[0106] The upper model and the lower model are formed as follows Figure 4 The game structure shown in Figure 1 is shown in Figure 1. The upper model determines the optimal flexible load compensation price based on the interests of the virtual power plant itself; the lower model determines the amount of flexible load participation response based on the flexible load compensation price. The nodes of the upper and lower models are solved until an equilibrium state is reached.

[0107] Use KKT conditions to transform the double-layer model into a single-layer model to complete the solution, including:

[0108] Commonly used methods for solving double-layer optimization models include: alternating multiplier method, dual method and genetic algorithm, etc. The present invention adopts the dual method. Since the lower model is a convex optimization model, the KKT conditions (Karush-Kuhn-Tucker Conditions) of the lower model can be calculated and combined with the upper model to transform the double-layer optimization problem into a single-layer problem for solution.

[0109] The single-layer model can be solved directly using commercial solvers.

[0110] The KKT conditions for the underlying problem are:

[0111]

[0112] In the formula, and is the Lagrange multiplier of formula (1-7), and is the Lagrange multiplier of formula (1-8). Therefore, a single-layer model as shown in formula (1-37) can be formed, the objective function is formula (1-11), and the constraints are formulas (1-1)-(1-10) and formulas (1-12)-(1-28):

[0113]

[0114] It also includes using real-time compensation prices to guide users to adjust their electricity consumption behaviors, promoting the enthusiasm of flexible loads to participate in demand response, and comparing the benefits of different price compensation mechanisms, including:

[0115] Virtual power plants provide real-time price compensation for flexible loads, and based on price signals from the electricity market, guide flexible loads to adjust their electricity consumption through economic incentives, thereby achieving supply and demand balance and optimized operation of the power system. Fig.10 As shown in the figure, when the power supply is tight (such as during peak power consumption periods), the compensation electricity price increases to guide flexible load users to actively reduce or transfer their loads; when the power supply is sufficient (such as during low power consumption periods), the compensation electricity price decreases.

[0116] In such Figure 1 A case analysis is conducted in the system shown in the figure. The system load level and photovoltaic output are as follows Figure 2The time-of-use electricity price is shown in Table 1, and the Gurobi solver is used for optimization.

[0117] Table 1 Time-of-use electricity price

[0118]

[0119]

[0120] The system load level, PV output, load that can be curtailed and load that can be transferred are predicted as follows: Figure 5 The load peaks at around 14:00. Assuming that extreme weather conditions occur between 14:00 and 16:00 in the afternoon, the photovoltaic output drops sharply. The voltage at the end node of the system is calculated as follows: Figure 3 shown.

[0121] like Figure 6 As shown in the figure, when the photovoltaic system is always running in constant power factor mode, the voltage of node 33 is lower than 0.95pu due to insufficient reactive power in the system from 14:00 to 16:00, resulting in undervoltage problem. Through the optimization control of the photovoltaic grid-connected converter, the voltage of node 33 can always be higher than 0.95pu, ensuring the voltage safety of the system. The virtual power plant can provide sufficient reactive power to the system through the optimization control of the photovoltaic system grid-connected converter, thereby participating in the reactive auxiliary service market and obtaining additional benefits.

[0122] Virtual power plants can participate in the peak load regulation market and adjust flexible loads, energy storage and other equipment to reduce peak loads and fill valleys. Figure 7 As shown. During the load peak period of 7:00-9:00, a discharge instruction is sent to the energy storage to reduce the amount of electricity purchased by the virtual power plant from the superior, while increasing the flexible load transfer to achieve peak shaving. During the load valley period of 10:00-13:00, valley filling is achieved through the charging of energy storage and the increase of flexible load. The peak-shaving income obtained by the virtual power plant according to the optimization model proposed by the present invention is 42.63 thousand yuan.

[0123] The compensation price and flexible load response quantity are as follows: Fig. 9 As shown. It can be seen that both RL and TL follow the trend of the compensation price. When the compensation price is high, the curtailable load will follow the response, and the transferable load will transfer out the load. In addition, the time periods for flexible participation in response are mainly 5:00-10:00 and 14:00-20:00. At this time, the load level of the system is relatively high. Therefore, the virtual power plant increases the compensation price to promote the enthusiasm of flexible loads to participate in the response. It should be noted that when the curtailable load does not participate in the response, the corresponding compensation price is not zero, because the total cost of the flexible load also takes into account the dissatisfied cost. At this time, the dissatisfied cost of participating in the response is greater than the compensation cost, so RL chooses not to participate in the response. In addition, as Figure 8As shown in the figure, for the single price compensation mechanism, the flexible load only responds to the single compensation price, making the regulation of the virtual power plant more disordered. The real-time compensation mechanism can effectively expand the enthusiasm and flexibility of the flexible load to participate in the response, and the response volume can quickly follow the changes in the compensation price.

[0124] Table 2 Comparison of benefits of different compensation mechanisms

[0125]

[0126]

[0127] Compared with the single price compensation mechanism, the real-time compensation price can more effectively mobilize the enthusiasm of flexible loads to participate in the response, increase the revenue of virtual power plants when participating in the peak-shaving market by 42%, and reduce the compensation costs paid by virtual power plants to flexible loads. The single compensation price makes flexible loads lose their adjustment space and reduces the power supply revenue of flexible loads. Since the voltage regulation revenue is mainly affected by the optimization control of photovoltaic grid-connected converters, the two compensation mechanisms have little effect on the revenue obtained by virtual power plants participating in the reactive power compensation market. The optimization model under the compensation mechanism adopted by the present invention can increase the revenue of virtual power plants by 62.02%.

[0128] Considering demand response virtual power plants participating in the ancillary services market system, including:

[0129] A control module is used to establish the power equation of the photovoltaic system and control it through the photovoltaic grid-connected converter;

[0130] Load module, used to establish a flexible load model considering demand response;

[0131] A two-layer module, used to establish a two-layer model of virtual power plants participating in the ancillary service market taking into account demand response;

[0132] The solution module is used to transform the double-layer model into a single-layer model using KKT conditions and complete the solution.

[0133] The present invention optimizes the control of the grid-connected converter of the photovoltaic system inside the virtual power plant, so that it can participate in the reactive auxiliary service market to obtain voltage regulation income, ensure the voltage stability of the system in extreme weather scenarios, and expand the revenue channels of the virtual power plant. At the same time, the present invention adopts a real-time price compensation mechanism to promote the enthusiasm and flexibility of flexible loads to participate in the response, which can effectively achieve peak shaving and valley filling compared to a single price compensation mechanism.

[0134] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be covered by the scope of the claims of the present invention.

Claims

1. Considering the demand response of virtual power plants participating in the ancillary service market strategy, the characteristics are: The following steps are involved: S1. Establish the power equation of the photovoltaic system and control it through the photovoltaic grid-connected converter; S2. On the basis of step S1, a flexible load model considering demand response is established; S3, based on step S2, establish a two-layer model of virtual power plants participating in the ancillary service market considering demand response; S4, use KKT conditions to transform the double-layer model into a single-layer model and complete the solution.

2. The strategy for participating in the ancillary service market of a virtual power plant considering demand response according to claim 1 is characterized in that: In step S1, the power equation of the photovoltaic system includes: The power equation of a photovoltaic system operating in constant power factor mode is: In the formula, and are the photovoltaic active and reactive power injection of the node, is the maximum active power output of photovoltaic system, tanθ is the tangent value of the power factor of photovoltaic system, K and T are the set of all nodes and the set of time points respectively; The power equation of the photovoltaic system operating in the grid voltage support mode is: In the formula, is the capacity of the PV system.

3. The strategy of participating in the ancillary service market of the virtual power plant considering demand response according to claim 1 is characterized in that: In step S2, the flexible load includes a transferable load and a reducible load. Flexible load model considering demand response, including: The constraints are: In the formula, are the actual transferable load and the curtailable load, respectively. and are the reduction of the load that can be reduced and the net transfer of the load that can be transferred, and are the initial load that can be cut and the initial load that can be transferred, respectively; The cost of flexible load considering demand response is: In the formula, C cut and C shift are the costs of load reduction and load shifting, respectively. is the unit electricity price of the flexible load, τ and ν are the unsatisfactory cost coefficients of the flexible load, and They are the compensation prices that virtual power plants provide to curtailable load and transferable load respectively.

4. The strategy for participating in the ancillary service market of a virtual power plant taking into account demand response according to claim 3 is characterized in that: In step S3, the two-layer model includes an upper model and a lower model, wherein the goal of the upper model is to maximize the comprehensive benefit of the virtual power plant, and the goal of the lower model is to minimize the electricity cost of the flexible load.

5. The strategy of participating in the ancillary service market of the virtual power plant considering demand response according to claim 4 is characterized in that: The objective function of the upper model is: maxR VPP =R AS +R DR +R AV -C VPP -C E -C PV (1-11) In the formula, R VPP is the comprehensive income of the virtual power plant, R AS , R DR , R AV are the benefits of virtual power plants participating in the peak load ancillary service market, supplying power to flexible loads, and participating in the reactive power ancillary service market, respectively. VPP , C E , C PV They are the operating cost of the virtual power plant, the cost of purchasing electricity from higher authorities, and the penalty cost for abandoning photovoltaic power generation; in, In the formula, λ PS , VF are the compensation coefficients for peak shaving and valley filling, ΔP t PS , ΔP t VF are the peak shaving and valley filling capacity declared by the virtual power plant, λ t E is the unit electricity price, λ V is the profit coefficient of participating in the reactive power auxiliary service market to provide voltage regulation, ΔV k,t is the voltage improvement of node k at time t, P t O is the amount of electricity purchased from the superior unit, λ PV is the light abandonment penalty coefficient, C ES , C DR They are the energy storage operation cost of the virtual power plant and the compensation cost of the flexible load participating in the response; in, C DR =C cut +C shift (1-19) In the formula, λ ES is the loss coefficient of the battery per unit charge and discharge, and are the charging and discharging power of the energy storage system respectively.

6. The strategy of participating in the ancillary service market of the virtual power plant considering demand response according to claim 5 is characterized in that: The constraints of the upper model include: In the formula, and are respectively the active and reactive power injection of the generator at node k, and are the active load and reactive load of node k respectively, and are respectively the active power and reactive power flowing on branch ki, and are the active power and reactive power flowing on branch jk, δ(k) and σ(k) are the set of post-order nodes and pre-order nodes of node k, respectively. jk and r jk are the resistance and reactance of branch jk respectively, B is the set of all branches in the system, l jk,t 、u k,t and u j,t are the square of the current amplitude of branch jk, the square of the voltage of node k and node j, is the square of the upper limit of the current amplitude, and are the squares of the lower and upper limits of the node voltage, respectively. and is the state of charge of the energy storage system and its upper limit, and are the charge and discharge signs of the energy storage system, η ES is the charging and discharging efficiency of energy storage, and They are the charging power, discharging power and state of charge of the energy storage system at time t-1 respectively.

7. The strategy of participating in the ancillary service market of the virtual power plant considering demand response according to claim 4 is characterized in that: The objective function of the lower model is: minC DR =C cut +C shift (1-29)。 8. The strategy of participating in the ancillary service market of a virtual power plant considering demand response according to claim 1, characterized in that: In step S4, the dual method is used to calculate the KKT conditions of the lower model and combine them into the upper model, and then transform it into a single-layer model and solve it; Among them, the KKT condition of the lower model is: In the formula, and are the Lagrange multipliers respectively.

9. The strategy of participating in the ancillary service market of a virtual power plant considering demand response according to claim 1, characterized in that: It also includes: virtual power plants provide real-time price compensation for flexible loads. When the power supply is at peak consumption, the compensation price increases are used to guide flexible load users to actively reduce or transfer loads; when the power supply is at low consumption, the compensation price decreases are used.

10. Considering the participation of virtual power plants in the ancillary service market system with demand response, the system is characterized by: include: A control module is used to establish the power equation of the photovoltaic system and control it through the photovoltaic grid-connected converter; Load module, used to establish a flexible load model considering demand response; A two-layer module, used to establish a two-layer model of virtual power plants participating in the ancillary service market taking into account demand response; The solution module is used to transform the double-layer model into a single-layer model using KKT conditions and complete the solution.

Images