The bidding method for energy storage power stations to participate in the frequency regulation auxiliary service market

Abstract

The present invention discloses a bidding method for energy storage power stations to participate in the frequency regulation auxiliary service market, the method comprising the following steps: step 1, using a particle swarm algorithm to simulate the day-ahead quotation strategy of the energy storage power station; step 2, substituting the quotation of the energy storage power station into the day-ahead pre-clearing model of the frequency regulation auxiliary service market to obtain the day-ahead market clearing electricity price and the winning bid capacity; step 3, substituting the day-ahead market clearing electricity price and the winning bid capacity of the energy storage power station into the energy storage power station revenue model to find the day-ahead optimal quotation strategy; step 4, introducing the energy storage dynamic capacity constraint factor in the intraday real-time clearing stage to adjust the initial quotation. This method can make day-ahead quotations for the characteristics that the command signal in the frequency regulation auxiliary service market cannot be estimated in advance, and use the price level to reflect the charging and discharging willingness of the energy storage equipment under different charge states.

Description

Technical Field

[0001] The present invention belongs to the field of power market bidding and relates to a bidding method for energy storage power stations to participate in a frequency regulation auxiliary service market. Background Art

[0002] With the proposal of the dual carbon goals, my country is vigorously promoting the transformation of a high proportion of renewable energy structure. As the main energy carrier, the power system is also transforming towards a new power system dominated by new energy. However, large-scale wind power and photovoltaic power generation have brought new challenges to the frequency stability of the power system: wind power and photovoltaic power generation are characterized by volatility and randomness. Large-scale new energy power generation grid connection leads to a mismatch between the output of the power generation side and the power demand on the load side, thereby exacerbating the power imbalance between the source and the load, and seriously threatening the safe operation of the power grid; at the same time, the proportion of traditional generators has decreased year by year, which will lead to a weakening of the system's frequency regulation capability. Therefore, driven by the transformation of the energy structure, how to suppress the frequency fluctuation of the power grid has become a current research hotspot.

[0003] With the development of energy storage technology and the improvement of the openness of the power market, my country has clarified the main position of new energy storage such as electrochemical energy storage and flywheel energy storage in grid connection, and vigorously encouraged new energy storage to participate in the auxiliary service market. Compared with traditional thermal power units, energy storage has the characteristics of rapid response, precise regulation, flexible control, two-way regulation, and high energy conversion efficiency, and can be used as a high-quality frequency regulation resource. The participation of energy storage power stations in frequency regulation can not only improve the response efficiency of frequency regulation resources, but also reduce the frequency regulation cost of the power system, which has positive significance for maintaining the relative stability of the power grid frequency and promoting the reform of the power auxiliary service market. Therefore, exploring the bidding strategy of energy storage power stations participating in the frequency regulation auxiliary service market has important reference value for energy storage power stations to enhance their market competitiveness in future new power systems.

[0004] The market-oriented operation of auxiliary services in my country is being piloted in some regions, and the profit model of energy storage power stations in the frequency regulation auxiliary service market is still in the exploratory stage. At present, there have been some studies on the bidding strategies of power generators in the power market, but the command signals in the frequency regulation auxiliary service market are difficult to predict in advance, and energy storage power stations are constrained by the state of charge during charging and discharging. Few studies have analyzed the particularity of energy storage power stations participating in the frequency regulation auxiliary service market. Summary of the invention

[0005] In view of the above shortcomings, the present invention provides a bidding method for energy storage power stations to participate in the frequency regulation auxiliary service market. The method can make day-ahead quotations based on the characteristic that command signals in the frequency regulation auxiliary service market cannot be predicted in advance, and use the high and low prices to reflect the charging and discharging willingness of energy storage equipment under different charge states.

[0006] The objective of the present invention is achieved through the following technical solutions:

[0007] A bidding method for energy storage power stations to participate in the frequency regulation auxiliary service market includes the following steps:

[0008] Step 1: Use particle swarm algorithm to simulate the day-ahead bidding strategy of energy storage power stations

[0009] Step 1: Initialize the particle swarm's position, velocity, inertia factor, learning factor, maximum number of iterations, algorithm termination error and other parameters;

[0010] Step 1 and 2: The particles adjust their own speed according to the inertia factor and the learning factor, thereby changing their positions. The position set of the particle group represents the day-ahead bidding strategy set of the energy storage power station.

[0011] Step 2: Substitute the energy storage power station quotation into the day-ahead pre-clearing model of the frequency regulation auxiliary service market to obtain the day-ahead market clearing electricity price and winning capacity

[0012] Step 21: Taking the minimum total cost of power system frequency regulation as the goal, considering the frequency regulation capacity constraints and mileage constraints of each energy storage power station, a day-ahead pre-clearing model of the frequency regulation auxiliary service market is constructed based on the market unified price settlement mechanism, where:

[0013] The objective function and constraints of the day-ahead clearing model for the frequency regulation ancillary service market are:

[0014]

[0015] Where F is the system frequency regulation cost, ρ c is the frequency regulation capacity compensation stipulated by the market, R c,i is the frequency regulation capacity of the energy storage unit i, m,i is the frequency regulation mileage price declared by energy storage unit i, k sm is the historical frequency regulation mileage coefficient of the power system, k m,i is the historical mileage call coefficient of energy storage unit i, R cd is the frequency regulation capacity demand in the next day’s dispatch cycle, R c0,i is the frequency regulation capacity declared by energy storage unit i;

[0016] Step 22: Substitute each group of bidding strategies generated by the particle swarm algorithm into the day-ahead pre-clearing model of the frequency regulation auxiliary service market to obtain the day-ahead clearing price of the market and the winning capacity of the energy storage power station under different bidding strategies;

[0017] Step 3: Substitute the day-ahead market clearing electricity price and the winning bid capacity of the energy storage power station into the energy storage power station revenue model to find the day-ahead optimal bidding strategy

[0018] Step 31: Taking the maximum frequency regulation benefit of energy storage power stations as the goal, considering the declared capacity constraints and declared mileage price constraints, a profit model for energy storage power stations participating in frequency regulation is constructed, where:

[0019] The objective function and constraints of the energy storage power station revenue model are:

[0020] maxE i =ρ c R c,i +λ m k m,i R c,i -C i (R c,i );

[0021]

[0022]

[0023]

[0024] In the formula, E i is the frequency regulation benefit of energy storage unit i, λ m is the frequency modulation mileage clearing price, C i (R c,i ) represents the frequency regulation cost of the energy storage power station as a function of its bid capacity, ρ mmin is the lower limit of the frequency regulation mileage quotation stipulated by the market, ρ mmax P is the upper limit of the frequency regulation mileage quotation stipulated by the market, N,i is the rated power of energy storage unit i, λ i is the marginal cost of energy storage station i, β i is the optimization parameter of the power generator, C i (t), C inv,i (t), C life,i (t), C loss,i (t) are the frequency regulation cost, investment depreciation cost, life loss cost, and power consumption cost of energy storage unit i at time t, respectively. p is a constant between 0.8 and 2.1, c fr,i is the unit power cost of energy storage unit i, E N,i is the rated capacity of energy storage unit i, η i d is the discharge efficiency of energy storage unit i, ρ e is the on-grid electricity price, N 0,i is the equivalent cycle number of energy storage unit i under 100% DOD, ΔT is the duration of the frequency modulation control cycle;

[0025] Step 32: Substitute the day-ahead market clearing electricity price and the winning capacity of the energy storage power station obtained by optimizing the day-ahead pre-clearing model of the frequency regulation auxiliary service market into the revenue model of the energy storage power station participating in frequency regulation. The frequency regulation revenue of the energy storage power station is used as the fitness function of the particle swarm algorithm. Each particle continuously optimizes its own position and speed to iteratively search for the best result, thereby determining the day-ahead optimal bidding strategy of the energy storage power station.

[0026] Step 4: Introduce the energy storage dynamic capacity constraint factor to adjust the initial quotation during the intraday real-time clearing stage

[0027] The first 5 minutes of the frequency regulation control cycle is used as the intraday real-time clearing stage of the frequency regulation auxiliary service market. In order to prevent the energy storage power station from being unable to continue to participate in the frequency regulation market due to excessive or low power consumption, the energy storage dynamic capacity constraint factor m is introduced in the intraday real-time clearing stage. SOC Adjustments to initial mileage quote:

[0028] ρ m,i ′=ρ m,i (1+m SOC );

[0029] In the formula, ρ m,i ′ is the mileage price adjusted in the pre-clearing phase, where:

[0030] m SOC The mathematical expression for charging in an energy storage power station is:

[0031]

[0032] m SOC The mathematical expression for discharging in an energy storage power station is:

[0033]

[0034] In the formula, SOC is the state of charge calculated by the energy storage power station based on the real-time power. low is the lower limit of the SOC confidence interval of the energy storage device, SOC high It is the upper limit of the SOC confidence interval of the energy storage device.

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

[0036] (1) The present invention establishes a two-layer optimization bidding model for energy storage power stations to participate in the frequency regulation auxiliary service market, defines the lower layer as a day-ahead pre-clearing model for the frequency regulation auxiliary service market, and defines the upper layer as a revenue model for energy storage power stations. The particle swarm algorithm is used to solve the two-layer optimization model, so that the energy storage power stations can fully consider the impact of the pre-clearing of the frequency regulation market on their revenue during the day-ahead quotation stage, thereby prompting them to make rational quotations and improve their market competitiveness.

[0037] (2) The present invention takes into account the fact that the command signal in the frequency regulation auxiliary service market cannot be predicted in advance during the day-ahead quotation stage, and rationally simplifies the energy storage frequency regulation cost to avoid the energy storage power station underestimating its own cost and causing losses;

[0038] (3) The present invention introduces a dynamic capacity constraint factor in the intraday real-time clearing stage in view of the state of charge constraint characteristics of energy storage equipment, and reflects the charging and discharging willingness of energy storage power stations under different states of charge in the declared price, thereby reducing the phenomenon that energy storage power stations are unable to output power as agreed due to their own excessively high or low power levels, which has a positive effect on suppressing the short-term monopoly caused by excessive market power of energy storage power stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is an overall flow chart of the bidding method of the present invention;

[0040] Figure 2 It is a schematic diagram of the bidding stage of the present invention;

[0041] Figure 3 It is a schematic diagram of a double-layer optimization model in the day-ahead pre-clearing stage of the present invention;

[0042] Figure 4 It is a flow chart of the particle swarm algorithm of the present invention for solving a double-layer optimization model;

[0043] Figure 5 is the frequency regulation benefit of the energy storage power station A in Example 1 of the present invention as β A Change curve;

[0044] Figure 6 is a SOC variation trend diagram of each energy storage power station in scenario 1 in embodiment 2 of the present invention;

[0045] Figure 7 is a SOC variation trend diagram of each energy storage power station in scenario 2 in embodiment 2 of the present invention;

[0046] Figure 8 4 is a SOC variation trend diagram of the lead-carbon battery C in Example 2 of the present invention. DETAILED DESCRIPTION

[0047] The technical solution of the present invention is further described below in conjunction with the accompanying drawings, but is not limited thereto. Any modification or equivalent replacement of the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention should be included in the protection scope of the present invention.

[0048] The present invention provides a bidding method for energy storage power stations to participate in the frequency regulation auxiliary service market, such as Figure 2As shown in the figure, this method divides the energy storage power station's participation in the frequency regulation auxiliary service market into the day-ahead pre-clearing stage and the intraday real-time clearing stage according to the time scale, where the intraday real-time clearing stage is defined as the first 5 minutes of the frequency regulation control cycle, so that the energy storage equipment can adjust its willingness to participate in the up and down frequency regulation market in time according to its own SOC during the intraday real-time clearing stage. Figure 1 As shown, the specific steps of the method are as follows:

[0049] Step 1: Use particle swarm algorithm to simulate the day-ahead bidding strategy of energy storage power stations

[0050] Step 1: Initialize the particle swarm's position, velocity, inertia factor, learning factor, maximum number of iterations, algorithm termination error and other parameters;

[0051] Step 1 and 2: The particles adjust their own speed according to the inertia factor and the learning factor, thereby changing their positions. The position set of the particle group represents the day-ahead bidding strategy set of the energy storage power station. The fitness function is used to iteratively optimize the bidding strategy set until the maximum number of iterations is reached or the algorithm termination error is met, wherein: the initial strategy set is randomly generated, and then optimized according to the fitness function.

[0052] Step 2: Substitute the energy storage power station quotation into the day-ahead pre-clearing model of the frequency regulation auxiliary service market to obtain the day-ahead market clearing electricity price and winning capacity

[0053] With the goal of minimizing the total cost of power system frequency regulation, considering the frequency regulation capacity constraints and mileage constraints of each energy storage power station, a day-ahead pre-clearing model for the frequency regulation ancillary service market is constructed based on the market unified price settlement mechanism:

[0054]

[0055] Where F is the system frequency regulation cost, ρ c is the frequency regulation capacity compensation stipulated by the market, R c,i is the frequency regulation capacity of the energy storage unit i, m,i is the frequency regulation mileage price declared by energy storage unit i, k sm is the historical frequency regulation mileage coefficient of the power system, k m,i is the historical mileage call coefficient of energy storage unit i, R cd is the frequency regulation capacity demand in the next day’s dispatch cycle, R c0,i It is the frequency regulation capacity declared by energy storage unit i.

[0056] The bidding strategies generated by the particle swarm algorithm are substituted into the day-ahead pre-clearing model of the frequency regulation ancillary service market, and the day-ahead clearing prices and the winning bid capacities of energy storage power stations under different bidding strategies are obtained.

[0057] Step 3: Substitute the day-ahead market clearing electricity price and the winning bid capacity of the energy storage power station into the energy storage power station revenue model to find the day-ahead optimal bidding strategy

[0058] The frequency regulation cost of energy storage power stations is divided into capacity cost and mileage cost. The capacity cost includes the depreciation cost of energy storage investment, and the mileage cost includes the energy storage life loss cost and the power loss cost. The calculation formula is:

[0059] C i (t) = C inv,i (t)+C life,i (t)+C loss,i (t) (2);

[0060]

[0061]

[0062]

[0063] In the formula, C i (t), C inv,i (t), C life,i (t), C loss,i (t) are the frequency regulation cost, investment depreciation cost, life loss cost, and power consumption cost of energy storage unit i at time t, respectively. cap,i is the unit capacity cost of energy storage unit i, E N,i is the rated capacity of energy storage unit i, s is the maximum number of frequency modulation calls in a day, r is the discount rate, T float,i is the floating charge life of energy storage unit i, P i c (t) is the charging power of energy storage unit i at time t, P i d (t) is the discharge power of energy storage unit i at time t, is the charging efficiency of energy storage unit i, is the discharge efficiency of energy storage unit i, k p is a constant usually between 0.8 and 2.1, c fr,i is the unit power cost of energy storage unit i, P N,i is the rated power of energy storage unit i, N 0,i is the equivalent cycle number of energy storage unit i under 100% DOD, ρ e For the on-grid electricity price.

[0064] In a market that does not distinguish between up and down frequency regulation, energy storage power stations cannot predict the frequency regulation direction of a certain instruction cycle the next day during the day-ahead bidding stage. According to formula (4), the life depreciation cost of energy storage participating in up frequency regulation (discharging) is greater than that of down frequency regulation (charging). Therefore, in order to avoid losses caused by underestimating the cost of energy storage, it is assumed that energy storage participates in up frequency regulation during the control period, so formula (4) is simplified to:

[0065]

[0066] For formula (5), it can be found that when the winning bid frequency regulation capacity value is the same, the power loss cost is greater when energy storage participates in frequency regulation (discharge), so (5) can also be simplified as:

[0067]

[0068] Since the frequency modulation command signal is frequent and unpredictable, and generally has a large variance, in order to minimize the underestimation of the energy storage frequency modulation cost and reduce the possibility of energy storage frequency modulation at a loss, the energy storage unit i corresponding to the maximum frequency modulation demand within the control period is selected to win the bid for the frequency modulation capacity R c,i To estimate the frequency regulation cost. Therefore, when declaring on the previous day, the calculation formula for the estimated frequency regulation cost of the energy storage power station in each trading period can be summarized as follows:

[0069]

[0070] According to formula (8), the marginal cost of energy storage power station participating in frequency regulation can be calculated:

[0071]

[0072] From formula (9), we can see that the frequency regulation capacity cost of energy storage is independent of its bid capacity and is a fixed value, which is consistent with the fact that most of the current frequency regulation market mechanisms in my country compensate frequency regulation capacity at a fixed price. The frequency regulation mileage cost is related to its bid capacity R c,i Therefore, if the energy storage power station quotes at marginal cost, the mileage price it declares is:

[0073]

[0074] However, in the actual electricity market, power generators are very likely to bid prices that deviate from marginal costs due to profit, that is, to try to find the optimal supply function ρ m,i (R c,i ):

[0075]

[0076] In the formula, β i This is the optimization parameter of the power generator. Each energy storage power station adjusts β ivalue to maximize its own interests.

[0077] With the goal of maximizing the frequency regulation benefits of energy storage power stations, considering the declared capacity constraints and declared mileage price constraints, a profit model for energy storage power stations participating in frequency regulation is constructed:

[0078] maxE i =ρ c R c,i +λ m k m,i R c,i -C i (R c,i ) (12);

[0079]

[0080] In the formula, E i is the frequency regulation benefit of energy storage unit i, λ m is the frequency modulation mileage clearing price, ρ mmin is the lower limit of the frequency regulation mileage quotation stipulated by the market, ρ mmax The upper limit of frequency regulation mileage quotation stipulated by the market.

[0081] like Figure 3 As shown in the figure, a two-layer optimization model is constructed. In the upper model, each energy storage power station continuously optimizes its bidding strategy based on the clearing results of the frequency regulation auxiliary service market and the published market information, so as to maximize its own interests. In the lower model, the dispatching and trading center, as the decision maker, is responsible for announcing market information such as frequency regulation demand and settlement rules in the previous day. Then, based on the price and capacity declared by the upper-layer frequency regulation resources, the minimum system frequency regulation cost is taken as the market clearing target, so as to realize the centralized clearing of the frequency regulation auxiliary service market.

[0082] The day-ahead market clearing electricity price and the winning capacity of the energy storage power station obtained by optimizing the day-ahead pre-clearing model of the frequency regulation auxiliary service market are substituted into the energy storage power station revenue model, and the frequency regulation revenue of the energy storage power station is used as the fitness function of the particle swarm algorithm. Each particle continuously optimizes its own position and speed to iteratively search for the best result, thereby determining the day-ahead optimal bidding strategy of the energy storage power station. Figure 4 The specific solution process is as follows:

[0083] The particle swarm algorithm is used to generate the initial bidding strategy set, and each bidding strategy is substituted into the lower-level frequency regulation market clearing model to obtain the market clearing electricity price and the winning frequency regulation capacity of the energy storage power station. Then, it is substituted into the upper-level energy storage frequency regulation profit model to calculate the profit under each bidding strategy, and the bidding strategy with the highest profit is selected as the optimal solution. The individual optimal value and group optimal value of the particle swarm are updated until the convergence criterion is met, and the optimal bidding strategy of the energy storage power station is output.

[0084] Step 4: Introduce the energy storage dynamic capacity constraint factor to adjust the initial quotation during the intraday real-time clearing stage

[0085] The first 5 minutes of the frequency regulation control cycle is used as the intraday real-time clearing stage of the frequency regulation auxiliary service market. In order to prevent the energy storage power station from being unable to continue to participate in the frequency regulation market due to excessive or low power consumption, the energy storage dynamic capacity constraint factor is introduced in the intraday real-time clearing stage to adjust the initial mileage quotation. At this time, the energy storage power station can obtain its own real-time SOC, and the frequency regulation demand forecast in the real-time clearing stage is more accurate than the day-ahead pre-clearing. Therefore, the energy storage dynamic capacity constraint factor m can be reasonably introduced in the intraday real-time clearing stage. SOC Adjustments to initial mileage quote:

[0086] m SOC The mathematical expression for charging in an energy storage power station is:

[0087]

[0088] m SOC The mathematical expression for discharging in an energy storage power station is:

[0089]

[0090] In the formula, SOC low The lower limit of the SOC confidence interval of the energy storage device is 40% in the present invention; high The upper limit of the SOC confidence interval of the energy storage device is 60% in the present invention.

[0091] In the frequency modulation control cycle, the predicted frequency modulation demand of each instruction cycle is summed. If the value is greater than 0, it means that the frequency modulation demand in this control cycle is large, corresponding to the energy storage discharge, m SOC According to formula (15), if the value is less than 0, it means that the frequency regulation demand is large during the control period, corresponding to energy storage charging, m SOC The value is obtained according to formula (14).

[0092] Energy storage power stations introduce m in the real-time clearing stage of the day SOC Adjust the initial mileage quote the day before:

[0093] ρ m,i ′=ρ m,i (1+m SOC ) (16);

[0094] In the formula, ρ m,i ′ is the mileage price adjusted in the pre-clearing stage.

[0095] Embodiment 1:

[0096] The parameters of three different specifications of energy storage power stations and the information of a frequency regulation control cycle of the next day announced by the grid dispatching agency refer to Table 1:

[0097] Table 1 Energy storage power station parameters and public information on the next day's grid frequency regulation

[0098]

[0099] When optimizing the bidding strategy of energy storage power station A alone, it is assumed that energy storage power station A does not take an aggressive bidding approach to bid a price lower than its marginal cost, and energy storage power stations B and C both bid their supply functions according to the marginal cost curve. Under this condition, the bidding strategy parameter β of energy storage power station A is optimized. A , the optimal bidding strategy and expected maximum profit of energy storage power station A are listed in Table 2:

[0100] Table 2 Optimal bidding strategy and expected maximum revenue of energy storage power station A

[0101]

[0102] The frequency regulation mileage quotations and winning bid capacities of various energy storage power stations under this bidding strategy are listed in Table 3:

[0103] Table 3 Frequency regulation mileage quotation and winning capacity of each energy storage power station

[0104]

[0105] In order to verify the correctness of the solution of the two-level optimization problem based on the particle swarm algorithm, the price strategy parameter β of the energy storage power station A is changed. A , it can be observed that its frequency modulation gain increases with β A Changes such as Figure 4 shown. Figure 5 This indicates that if energy storage station A declares a mileage price higher than its own peripheral cost, its frequency regulation revenue will increase with β A The optimal bidding strategy of energy storage power station A is to declare the marginal cost function. This method can obtain the maximum benefit based on conservative bidding.

[0106] The same method is used to adjust the bidding strategy parameters β of energy storage power stations B and C respectively. B , β C After optimization, we can get the results shown in Table 4. The results show that under low risk aversion, the optimal bidding strategy for energy storage power stations is to declare their own peripheral costs. Blindly bidding too high will lead to a reduction in the winning frequency regulation capacity, thereby affecting its frequency regulation revenue.

[0107] Table 4 Optimal bidding strategies for energy storage power stations B and C under low risk aversion

[0108]

[0109] Embodiment 2:

[0110] In order to verify the effectiveness of considering dynamic capacity constraint factors for each energy storage power station in the real-time clearing stage, two different scenarios are designed below:

[0111] Scenario 1: Without considering the SOC constraint of the energy storage power station, the dynamic capacity constraint factor is not introduced in the real-time clearing stage to adjust the frequency regulation mileage quotation of each power station, and the profit margin of the energy storage power station is fixed at 0.5, that is, the adjusted mileage price is:

[0112] ρ m,i ′=ρ m,i (1+f i ) (17);

[0113] In the formula, f i In order to determine the profit margin of each energy storage power station in the real-time clearing stage, 0.5 is taken in this example.

[0114] Scenario 2: Considering the SOC constraint of the energy storage power station, a dynamic capacity constraint factor is introduced in the real-time clearing stage, and the frequency regulation mileage quotation of each power station is adjusted according to equations (14), (15), and (16).

[0115] The basic parameters of the five energy storage power stations and the information published by the power grid are shown in Table 5. The 24-hour RegD (Dynamic Regulation Signal) signal of the PJM frequency regulation market in the United States in May 2014 was selected as the simulated frequency regulation instruction, and the proposed energy storage dynamic capacity constraint factor m was SOC The value of the RegD signal multiplied by the winning frequency regulation capacity of each energy storage power station in each 15-minute frequency regulation control cycle is the real-time output of the energy storage power station in each instruction cycle.

[0116] Since each energy storage power station bids at the same time in the actual power market, it is impossible to predict the bidding strategies of other competitors in advance. Therefore, according to the analysis of Example 1, it can be seen that the most reliable bidding strategy for each energy storage power station at this time is to declare its own marginal cost curve as its supply function, thereby determining its own frequency regulation mileage quotation and declared capacity. Therefore, in this example, the five energy storage power stations all use their own marginal cost as the declared supply function.

[0117] Table 5 Basic parameters of energy storage power stations and information published by the power grid

[0118]

[0119]

[0120] Under two different scenarios, the dynamic change trends of SOC of each energy storage power station within 24 hours are as follows: Figure 6 and Figure 7 In order to more clearly observe the role of the energy storage dynamic capacity constraint factor, the SOC states of the lead-carbon battery C in two different scenarios are compared, as shown in Figure 2. Figure 8 As shown, the results show that considering the dynamic capacity constraint factor of energy storage can extend the time that energy storage power stations participate in the frequency regulation ancillary service market.

[0121] The number of times each energy storage power station exits the frequency regulation auxiliary service market due to state of charge constraints under two different scenarios is listed in Table 6.

[0122] Table 6 The number of times each energy storage power station exits the frequency regulation market in different scenarios

[0123]

[0124] In the market mechanism that does not consider the dynamic capacity factor of energy storage, the duration and number of times that each energy storage power station exits the frequency regulation market are significantly longer than the market mechanism that considers the dynamic capacity factor. The SOC of energy storage equipment frequently approaches its lower limit, which not only causes loss to the life of the energy storage, but also is very likely to lead to fines due to its own power failure to achieve the winning bid. Therefore, the energy storage power station bidding strategy based on the dynamic capacity constraint factor can extend the service life and participation time of the energy storage power station in the frequency regulation market and avoid unnecessary fines; for the power system, it can enable the energy storage power station to provide more reliable frequency regulation services and ensure a more stable frequency of the power system.

[0125] The total grid frequency regulation costs in the frequency regulation ancillary service market within one day under two different scenarios are listed in Table 7.

[0126] Table 7 Total cost of 24h grid frequency regulation under different scenarios

[0127]

[0128] The results show that the introduction of energy storage dynamic capacity constraint factors can, to a certain extent, suppress the phenomenon of energy storage power stations manipulating the market due to excessive market power, thereby reducing the frequency regulation costs of the power grid and facilitating the standardized reform of the power market.

Claims

1. A bidding method for energy storage power stations to participate in the frequency regulation auxiliary service market, characterized in that The method comprises the following steps: Step 1: Use the particle swarm algorithm to simulate the day-ahead bidding strategy of the energy storage power station: Step 1: Initialize the position, velocity, inertia factor, learning factor, maximum number of iterations, and algorithm termination error parameters of the particle swarm; Step 1 and 2: The particles adjust their own speed according to the inertia factor and the learning factor, thereby changing their positions. The position set of the particle group represents the day-ahead bidding strategy set of the energy storage power station. Step 2: Substitute the energy storage power station quotation into the day-ahead pre-clearing model of the frequency regulation auxiliary service market to obtain the day-ahead market clearing electricity price and winning bid capacity: Step 21: Taking the minimum total cost of power system frequency regulation as the goal, considering the winning frequency regulation capacity constraints and mileage constraints of each energy storage power station, a day-ahead pre-clearing model of the frequency regulation ancillary service market is constructed based on the market unified price settlement mechanism. The objective function and constraints of the day-ahead pre-clearing model of the frequency regulation ancillary service market are: Where F is the system frequency regulation cost, ρ c is the frequency regulation capacity compensation stipulated by the market, R c,i is the frequency regulation capacity of the energy storage unit i, m,i is the frequency regulation mileage price declared by energy storage unit i, k sm is the historical frequency regulation mileage coefficient of the power system, k m,i is the historical mileage call coefficient of energy storage unit i, R cd is the frequency regulation capacity demand in the next day’s dispatch cycle, R c0,i is the frequency regulation capacity declared by energy storage unit i; Step 22: Substitute each group of bidding strategies generated by the particle swarm algorithm into the day-ahead pre-clearing model of the frequency regulation auxiliary service market to obtain the day-ahead clearing electricity price and the winning capacity of the energy storage power station under different bidding strategies; Step 3: Substitute the day-ahead market clearing electricity price and the winning capacity of the energy storage power station into the energy storage power station revenue model to find the day-ahead optimal bidding strategy: Step 31: Taking the maximum frequency regulation benefit of energy storage power stations as the goal, considering the declared capacity constraint and the declared mileage price constraint, a revenue model for energy storage power stations participating in frequency regulation is constructed. The objective function and constraint conditions of the revenue model are: maxE i =ρ c R c,i +λ m k m,i R c,i -C i (R c,i ); In the formula, E i is the frequency regulation benefit of energy storage unit i, λ m is the frequency modulation mileage clearing price, C i (R c,i ) represents the frequency regulation cost of the energy storage power station as a function of its bid capacity, ρ mmin is the lower limit of the frequency regulation mileage quotation stipulated by the market, ρ mmax P is the upper limit of the frequency regulation mileage quotation stipulated by the market, N,i is the rated power of energy storage unit i, λ i is the marginal cost of energy storage station i, β i is the optimization parameter of the power generator, C i (t), C inv,i (t), C life,i (t), C loss,i (t) are the frequency regulation cost, investment depreciation cost, life loss cost, and power consumption cost of energy storage unit i at time t, respectively. p is a constant between 0.8 and 2.1, c fr,i is the unit power cost of energy storage unit i, E N,i is the rated capacity of energy storage unit i, is the discharge efficiency of energy storage unit i, ρ e is the on-grid electricity price, N 0,i is the equivalent cycle number of energy storage unit i under 100% DOD, ΔT is the duration of the frequency modulation control cycle; Step 32: Substitute the day-ahead market clearing electricity price and the winning capacity of the energy storage power station obtained by optimizing the day-ahead pre-clearing model of the frequency regulation auxiliary service market into the revenue model of the energy storage power station participating in frequency regulation. The frequency regulation revenue of the energy storage power station is used as the fitness function of the particle swarm algorithm. Each particle continuously optimizes its own position and speed to iteratively search for the best result, thereby determining the day-ahead optimal bidding strategy of the energy storage power station. Step 4: Introduce the energy storage dynamic capacity constraint factor in the intraday real-time clearing stage to adjust the initial mileage quotation: The first 5 minutes of the frequency regulation control cycle is used as the intraday real-time clearing stage of the frequency regulation auxiliary service market. In order to prevent the energy storage power station from being unable to continue to participate in the frequency regulation market due to excessive or low power consumption, the energy storage dynamic capacity constraint factor m is introduced in the intraday real-time clearing stage. SOC Adjustments to initial mileage quote: r m,i ′=ρ m,i (1+m SOC ); In the formula, ρ m,i ' is the mileage price adjusted in the pre-clearing stage; The m SOC The mathematical expression for charging in an energy storage power station is: The m SOC The mathematical expression when the energy storage power station discharges is: In the formula, SOC is the state of charge calculated by the energy storage power station based on the real-time power. low is the lower limit of the SOC confidence interval of the energy storage device, SOC low =0.4, SOC high is the upper limit of the SOC confidence interval of the energy storage device, SOC high =0.

6.

2. The bidding method for energy storage power station participating in the frequency regulation auxiliary service market according to claim 1 is characterized in that The C i (t), C inv,i (t), C life,i (t), C loss,i (t) The following conditions are met: C i (t)=C inv,i (t)+C life,i (t)+C loss,i (t) In the formula, c cap,i is the unit capacity cost of energy storage unit i, s is the maximum number of frequency modulation calls in a day, r is the discount rate, T float,i is the floating charge life of energy storage unit i, P N,i is the rated power of energy storage unit i.

Images