Optimal decision-making method and device for offshore wind power participating in medium- and long-term bilateral negotiation transactions

Abstract

The present invention discloses an optimization decision method and device for offshore wind power participating in medium- and long-term bilateral negotiation transactions, the method comprising S1, obtaining the best matching power load combination by a time series similarity measurement method; S2, on the basis of the best matching power load combination, considering the user's demand response resource reserve capacity and the generation and consumption curve prediction error, constructing a medium- and long-term transaction decision model based on two-stage distributed robust optimization; S3, solving the medium- and long-term transaction decision model based on two-stage distributed robust optimization, and generating offshore wind power transaction decision results. The present invention provides technical support for offshore wind power to find users with optimal characteristic matching for medium- and long-term bilateral negotiation transactions, and at the same time provides a method basis for how to formulate medium- and long-term custom decomposition curves, reducing the losses caused by not considering the generation and consumption curve characteristics of both parties for medium- and long-term transactions, and effectively promoting offshore wind power to participate in market-oriented transactions and consumption.

Description

Technical Field

[0001] The present invention relates to renewable energy technology, and in particular to a method and device for optimizing decision-making for offshore wind power participating in mid- and long-term bilateral negotiation transactions. Background Art

[0002] Offshore wind power has the characteristics of rich resources and high power generation hours. It is the field with the greatest potential for large-scale development in renewable energy and is also a frontier field for the development of new energy. However, due to the inherent randomness, volatility and anti-peak characteristics of offshore wind power, it is easy to suffer large losses when participating in market competition due to discrepancies between actual output and actual output. In this context, how to better realize offshore wind power consumption through market regulation, thereby reducing wind abandonment and reducing transaction risks, has put forward new requirements for the existing offshore wind power market transaction mechanism.

[0003] In the current electricity market transaction mode in my country, medium- and long-term contract transactions occupy an important position, play the role of "ballast stone" in the market, and provide stable electricity supply and expected benefits for market players. There have been many studies at home and abroad on the participation of wind power in medium- and long-term market transactions, but most of the studies are mainly aimed at the centralized bidding transaction mode, and bilateral negotiation transactions represented by large users' direct purchase of electricity are also one of the main transaction methods in the medium- and long-term electricity market. Taking the United Kingdom as an example, more than 90% of its total market electricity is completed in the form of contracts through bilateral negotiation transactions, but there are few studies on the participation of wind power in medium- and long-term bilateral negotiation transactions in China. On the other hand, at present, the medium- and long-term bilateral contracts signed by offshore wind power companies and users are mainly carried out according to the process of decomposition from total transaction electricity to contract electricity. Market players usually agree on custom decomposition curves based on historical electricity load curves and future actual needs, or formulate decomposition rules based on the characteristics of grid unified load regulation. These processes have not taken into account the output characteristics of the offshore wind farm and the actual matching degree of the generation and consumption curves of both parties to the contract, which is easy to lead to greater dynamic transaction risks.

[0004] With the continuous improvement of wind power prediction accuracy and the increasingly important role played by domestic load integrators in the system power balance, offshore wind power has more options for participating in medium- and long-term bilateral negotiation transactions. Especially for large users involved in power curve transactions with large electricity consumption, the common megawatt-level energy storage equipment is actually difficult to meet the needs. This requires power generators to trade with large-scale users and adjust the power load curve to match the power generation characteristics by realizing load combination. Summary of the invention

[0005] The purpose of the present invention is to overcome the deficiencies of the above-mentioned prior art and to provide a method and device for optimizing decision-making for offshore wind power participating in medium- and long-term bilateral negotiation transactions, so as to provide technical support for finding users with the best characteristics matching for offshore wind power to conduct medium- and long-term bilateral negotiation transactions.

[0006] To achieve the above object, the technical solution of the present invention is:

[0007] In a first aspect, the present invention provides a method for optimizing decision-making for offshore wind power participating in medium- and long-term bilateral negotiation transactions, comprising:

[0008] S1. Obtain the optimal matching power load combination through the time series similarity measurement method;

[0009] S2. Based on the optimal matching power load combination, considering the user's demand response resource reserve capacity and the power generation and consumption curve prediction error, a medium- and long-term transaction decision model based on two-stage distributed robust optimization is constructed;

[0010] S3. Solve the medium- and long-term transaction decision model based on the two-stage distributed robust optimization to generate offshore wind power transaction decision results.

[0011] In a second aspect, the present invention provides an optimized decision-making device for offshore wind power participating in medium- and long-term bilateral negotiation transactions, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described above when executing the computer program.

[0012] In a third aspect, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described above are implemented.

[0013] Compared with the prior art, the present invention has the following beneficial effects:

[0014] The present invention proposes an optimization decision-making method for offshore wind power participating in medium- and long-term bilateral negotiated transactions of electricity, which takes into account the similarity of the timing of generation and consumption curves. The method provides technical support for offshore wind power to find users with optimal characteristics matching for medium- and long-term bilateral negotiated transactions. At the same time, it provides a methodological basis for how to formulate medium- and long-term customized decomposition curves, thereby reducing the losses caused by medium- and long-term transactions without considering the characteristics of the generation and consumption curves of both parties, and effectively promoting the participation of offshore wind power in market-oriented transactions and consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A flow chart of a method for optimizing decision-making for offshore wind power participating in medium- and long-term bilateral negotiation transactions provided in Example 1 of the present invention;

[0016] Figure 2 Schematic diagram of Fast-DTW search space changes;

[0017] Figure 3 A schematic diagram of the composition of an optimized decision-making device for offshore wind power participating in medium- and long-term bilateral negotiation transactions provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0018] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0019] Embodiment 1:

[0020] See also Figure 1 As shown, the method for optimizing decision-making for offshore wind power participating in medium- and long-term bilateral negotiation transactions provided in this embodiment mainly includes the following steps:

[0021] S1. Obtain the optimal matching power load combination through the time series similarity measurement method.

[0022] Thus, in this step, the best match between offshore wind power and load is achieved through the time series similarity measurement method, and the optimal transaction load combination is found for the target offshore wind power.

[0023] S2. Based on the optimal matching power load combination, considering the demand response resource reserve capacity and the prediction error of the power generation and consumption curve, a medium- and long-term trading decision model based on two-stage distributed robust optimization is constructed;

[0024] Therefore, in this step, by constructing a medium- and long-term transaction decision-making model based on two-stage distributed robust optimization, on the basis of the optimal matching electricity load combination, the demand response resource reserve and the prediction error of the power generation and consumption curve are considered to further improve the risk resistance of medium- and long-term bilateral negotiated transactions of offshore wind power, and reduce the medium- and long-term transaction losses of offshore wind power companies due to output fluctuations and prediction errors.

[0025] S3. Solve the medium- and long-term transaction decision model based on the two-stage distributed robust optimization to generate offshore wind power transaction decision results.

[0026] It can be seen that the optimization decision-making method for offshore wind power participating in medium- and long-term bilateral negotiated transactions of electricity that considers the timing similarity of generation and consumption curves proposed in this embodiment provides technical support for offshore wind power to find users with optimal characteristics matching for medium- and long-term bilateral negotiated transactions. At the same time, it provides a methodological basis for how to formulate medium- and long-term customized decomposition curves, reduces losses caused by medium- and long-term transactions without considering the characteristics of the generation and consumption curves of both parties, and effectively promotes offshore wind power to participate in market-oriented transactions and consumption.

[0027] In a specific embodiment, the above step S1 includes the following sub-steps:

[0028] S11. Obtain a data sample set D, wherein the data sample set D is formed by all electricity load combination curves and offshore wind power generation curves; that is, the data sample set D is formed by offshore wind power producers and load integrators who have trading intentions providing corresponding curve data to market operators.

[0029] S12, using a curve numerical similarity matching model based on K-Medoids clustering on the data sample set D to obtain a power consumption curve set U;

[0030] S13, using a curve shape similarity matching model based on Fast-DTW measurement for the power consumption curve set U, to obtain a Fast-DTW distance set L between each power consumption load combination and the offshore wind power generation curve;

[0031] S14. Let λ=min L=min(l1,l2,...,l n ), the power load combination corresponding to λ is the optimal matching power load combination U b .

[0032] In this way, the above sub-steps achieve the optimal matching of generation and consumption curves considering time series similarity based on K-Medoids clustering and Fast-DTW measurement. Compared with the use of DTW alone, the time required to match the appropriate user combination for offshore wind power is greatly reduced. At the same time, the morphological similarity and numerical similarity of the generation and consumption curves are taken into account, which has better matching performance and matching efficiency, meets the actual application needs, and makes up for the shortcomings of using Euclidean distance or dynamic time warping distance alone to match the generation and consumption curves of medium- and long-term bilateral negotiated transactions.

[0033] The numerical similarity of time series curves is usually measured by Euclidean distance. The smaller the Euclidean distance, the more similar the two curves are. For two given and equal-length time series X = (x1, x2, ..., x t ) and Y=(y1,y2,...,y t ), the Euclidean distance calculation expression is as follows:

[0034]

[0035] Where D(X,Y) represents the Euclidean distance between time series X and Y; T represents the length of the time series.

[0036] Since there are power load combinations that are very dissimilar to the offshore wind power generation curve when performing curve matching, in order to improve the clustering effect, this embodiment adopts the K-Medoids algorithm to find a power load combination set with a high numerical similarity to the offshore wind power generation curve. The above step S12 includes the following sub-steps:

[0037] S121, randomly select k data samples from the data sample set D as particle O j (j=1,2,...,k);

[0038] S122, randomly select a non-particle sample O random ; Calculate the exchange object Orandom and O1 as reference point, repeat the operation in step S121 to generate a new set of clusters, calculate the objective function S, and if S<0, random Exchange with O1 and keep the new cluster, otherwise, keep the original center point and cluster; repeat this step until the k center points do not change;

[0039] S123, the power load combination of the same cluster as the offshore wind power generation curve forms a power load combination set U = (u1, u2, ..., u n ), where u n Indicates different combinations of power loads.

[0040] K-Medoids clustering also involves the selection of the number of clusters k. This embodiment uses the sum of squared errors (SSE) as the clustering evaluation index, calculates the sum of the squares of the distances from all data points to the centroid of the cluster to which they belong, and determines the optimal number of clusters k by the elbow method. The calculation expression of SSE is as follows:

[0041]

[0042] Among them, C i is the i-th cluster, p is C i The sample points in m i It is C i The center of mass (C i The mean of all samples in the cluster), SSE is the clustering error of all samples, which represents the quality of clustering effect.

[0043] In a specific embodiment, the curve morphology similarity matching model of the Fast-DTW metric is based on the DTW algorithm, and Fast-DTW is implemented by limiting the search space. DTW achieves the bending of the time axis by extending and shortening the time series, thereby achieving the measurement of the morphological similarity of two time series. For two given time series X = (x1, x2, ..., x n ) and Y=(y1,y2,…,y m ), construct an n×m distance matrix D n×m ,, the matrix element calculation formula is as follows:

[0044]

[0045] The purpose of DTW is to find an optimal curved path W = (w1, w2, ..., w k ), satisfying the point pair w1 = (1,1) to w k =(n,m), there is a minimum cumulative distance:

[0046]

[0047] Where w s It represents the position of the path element in the distance matrix, and usually needs to satisfy the boundary, continuity and monotonicity. To solve the above formula, the cumulative matrix A needs to be constructed through dynamic programming method. n×m :

[0048]

[0049] Where i∈[1,n], j∈[1,m]. Then the dynamic time warping distance of time series X and Y is DTW(X,Y)=A(n,m).

[0050] In order to speed up the DTW algorithm, the method of limiting the search space is used to implement Fast-DTW, that is, reducing the search space in the distance matrix, such as Figure 2 The shaded area is the actual exploration space, and the blank area is not explored. Therefore, based on the load combination set U in step S123, the Fast-DTW distance L between each load combination and the offshore wind power generation curve can be calculated as (l1, l2, ..., l n ).

[0051] On the basis of the optimal matching electricity load combination, considering the user's demand response resource reserve capacity and the prediction error of the power generation and consumption curve, a medium- and long-term trading decision-making model based on two-stage distributed robust optimization is constructed. The DR resources that adapt to the output characteristics of the offshore wind power in the future are configured for the electricity load combination through the optimization model. Part of the income is anchored by the electricity contract and DR capacity configuration, which reduces the risk of signing medium- and long-term contracts for offshore wind power and improves the performance of offshore wind power companies.

[0052] Specifically, the objective function of the medium- and long-term trading decision model is:

[0053]

[0054] Among them, λ p is the time-based contract electricity price; P is the time period type set, where 1 represents the peak period, 2 represents the normal period, and 3 represents the valley period; N I N is the user DR resource set; T N is the set of callable time periods of DR; T，p The corresponding time periods of peaks, flats and valleys are as follows; is the up and down adjustment amount of demand response resource i at time t; L t is the baseline load L before adjustment at time t t θ p To simulate the spot time-of-use electricity price; Δ tis the difference between wind power output and total load after demand response optimization configuration at time t, Represents the output shortfall when wind power output is less than the total load. Represents the output surplus when wind power output is greater than the total load; α i is the calling cost of demand response i; β i is the reserve capacity price for demand response i; The reserve capacity of demand response i purchased for the contract.

[0055] The general constraints of the objective function of the medium- and long-term trading decision-making model include:

[0056] a) Adjust upper and lower limit constraints

[0057]

[0058] in, for The corresponding 0-1 variable represents whether the demand response resource i is configured and called. The linearized constraint condition can be written as:

[0059]

[0060]

[0061] b) Reserve capacity constraints

[0062]

[0063] Among them, z i,t The linearization variable introduced are the upper and lower limits of the reserve capacity that demand response resource i can provide.

[0064] c) Positive and negative output coupling constraints

[0065]

[0066] d) Call count constraints

[0067]

[0068] Among them, Num i is the upper limit of the number of times demand response resource i can be called within the callable time period.

[0069] The special constraints of the objective function of the medium- and long-term trading decision-making model include:

[0070] a) Transferable load constraints

[0071] Transferable load is defined as load that can be freely transferred within a specified period of time X, where the load can only be reduced before t+K and transferred to the time between t+K and t+X, and the total power consumption before and after the transfer is equal. The corresponding constraints can be written as:

[0072]

[0073] b) Load constraints can be reduced

[0074] The load that can be reduced is defined as the amount of load that can be reduced within a specified period of time. The corresponding constraint can be written as:

[0075]

[0076] Among them, N Ti A collection of call time periods during which load can be reduced.

[0077] (4) Supply and demand matching constraints

[0078]

[0079] Among them, P t is the wind power output at time t; They are 0-1 variable, indicating whether there is a surplus or deficit in output; cons represents an infinite constant, which is 10 here. 8 .

[0080] In order to deal with the prediction error of power generation and consumption curve and improve the robustness of the decision-making model, this embodiment proposes a medium- and long-term offshore wind power transaction decision model based on two-stage distributed robust optimization. The first stage is the user DR resource reserve capacity purchase decision stage, which determines the optimal DR capacity configuration for medium- and long-term offshore wind power transactions; the second stage is the simulation operation stage, which uses the spot time-of-use electricity price as a penalty for insufficient wind power output to calculate the expected benefits of wind power suppliers in medium- and long-term transactions. The overall model is modified as follows:

[0081]

[0082] Among them, s is a typical scenario; P s is the probability of typical scene distribution; C inv The purchase cost of DR resource reserve capacity for users participating in medium- and long-term offshore wind power transactions; is the total expenditure of offshore wind power participating in medium- and long-term transactions in the simulated operation phase under the sth scenario; is the total revenue of offshore wind power participating in medium- and long-term transactions in the simulated operation phase under the sth scenario. And,

[0083]

[0084]

[0085]

[0086] For convenience, the first-stage decision variable is represented by x, which is the reserve capacity of demand response i purchased under the contract; the second-stage decision variable is represented by y s It indicates that the operating variables including the adjustment amount and calling status of each user's demand response resources are included. The medium- and long-term transaction decision model of offshore wind power based on two-stage distributed robust optimization is expressed as follows:

[0087]

[0088] stCx≤c (21)

[0089] Fx≥Hy s , (twenty two)

[0090] Ky s =0, (twenty three)

[0091] Gy s ≤g, (twenty four)

[0092] Among them, Ω p P s Ax represents the purchase cost of DR resource reserve capacity of users participating in medium- and long-term transactions of offshore wind power; By s represents the total revenue of offshore wind power participating in the medium and long term in the simulated operation stage under the sth scenario; formula (21) represents the reserve capacity constraint (10); formula (22) represents the upper and lower limit constraints of regulation (11); formula (23) represents the equality constraints in the positive and negative output coupling constraint (11), the transferable load constraint (13), and the supply and demand matching constraint (15); formula (24) represents the call number constraint (12), the curtailable load constraint (14) and the inequality constraints in formulas (11), (13), and (15).

[0093] Considering the forecast error of the emission curve, M actual scenarios are generated by the scenario generation method to simulate the historical data situation, and the clustering screening is satisfied. M1, M2, ...M Ns A finite discrete scenario can construct a theoretical distribution of uncertain variables in Since the initial distribution obtained through historical data is inconsistent with the actual distribution, it is necessary to establish a confidence set under different θ values, as shown in the following formula:

[0094]

[0095] Among them, Ω p For p s The feasible domain of θ1 and θ ∞ are the probability allowable deviation values ​​under the corresponding constraints of 1-norm and ∞-norm, respectively, and satisfy the following equation

[0096]

[0097] The model is divided into the main problem (MP) and the subproblem (SP) and solved iteratively using the column and constraint generation algorithm (CCG). The main problem objective is shown in equation (27), which solves the optimal solution that satisfies the constraints under a known probability distribution and provides a lower bound:

[0098] MP:

[0099]

[0100] Where m is the number of iterations.

[0101] After solving the main problem and obtaining the first-stage variable x*, the subproblem looks for the worst probability distribution under the simulation run, returns to the main problem, and provides an upper bound for the optimal solution:

[0102] SP:

[0103] The inner second-stage variables in formula (29) are independent of the outer discrete scenario probabilities, so the inner minimum problem can be solved first and then the outer problem.

[0104] Thus, by adopting this method, offshore wind power transaction decision results can be generated, including: the optimal matching transaction user combination U b , medium and long term contract curve l b and a plan for purchasing spare capacity for DR resources.

[0105] In summary, the offshore wind power participation in medium- and long-term bilateral negotiation transaction optimization decision-making method provided in this embodiment has the following technical advantages compared with the prior art:

[0106] (1) In view of the shortcomings of using Euclidean distance or dynamic time warping distance alone to match the generation and consumption curves of medium- and long-term bilateral negotiated transactions, an optimal matching method for generation and consumption curves based on K-Medoids clustering and Fast-DTW measurement considering time series similarity is proposed. Compared with using DTW alone, this method greatly reduces the time required to match the appropriate user combination for offshore wind power, while taking into account the morphological similarity and numerical similarity of the generation and consumption curves. It has better matching performance and matching efficiency and meets the actual application needs.

[0107] (2) Construct a medium- and long-term transaction decision model based on two-stage distributed robust optimization. On the basis of the optimal matching power load combination, the demand response resource reserve and the power generation and consumption curve prediction error are considered to further improve the risk resistance of medium- and long-term bilateral negotiated transactions of offshore wind power and reduce the medium- and long-term transaction losses of offshore wind power companies caused by output fluctuations and prediction errors.

[0108] Embodiment 2:

[0109] See also Figure 3 As shown, the offshore wind power participation in medium- and long-term bilateral negotiation transaction optimization decision-making device provided in this embodiment includes a processor 31, a memory 32, and a computer program 33 stored in the memory 32 and executable on the processor 31, such as an offshore wind power participation in medium- and long-term bilateral negotiation transaction optimization decision-making program. When the processor 31 executes the computer program 33, the steps of the above-mentioned embodiment 1 are implemented, such as Figure 1 Steps shown.

[0110] Exemplarily, the computer program 33 may be divided into one or more modules / units, which are stored in the memory 32 and executed by the processor 31 to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program 33 in the offshore wind power participation medium- and long-term bilateral negotiation transaction optimization decision-making device.

[0111] The processor 31 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0112] The memory 32 may be an internal storage unit of the medium- and long-term bilateral negotiation transaction optimization decision-making device for offshore wind power, such as a hard disk or memory of the medium- and long-term bilateral negotiation transaction optimization decision-making device for offshore wind power. The memory 32 may also be an external storage device of the medium- and long-term bilateral negotiation transaction optimization decision-making device for offshore wind power, such as a plug-in hard disk, a smart memory card (SmartMedia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the medium- and long-term bilateral negotiation transaction optimization decision-making device for offshore wind power. Further, the memory 32 may also include both an internal storage unit and an external storage device of the medium- and long-term bilateral negotiation transaction optimization decision-making device for offshore wind power. The memory 32 is used to store the computer program and other programs and data required by the medium- and long-term bilateral negotiation transaction optimization decision-making device for offshore wind power. The memory 32 may also be used to temporarily store data that has been output or is to be output.

[0113] Embodiment 3:

[0114] This embodiment provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the method described in Embodiment 1 are implemented.

[0115] The computer-readable medium shown can be any device that can contain, store, communicate, propagate or transmit a program for use with an instruction execution system, device or apparatus or in conjunction with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection portion with one or more wirings (electronic devices), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and editable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium can even be paper or other suitable media on which the program can be printed, such as by optically scanning the paper or other medium, then editing, interpreting or processing in other suitable ways as necessary to obtain the program electronically, and then storing it in a computer memory.

[0116] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable ordinary technicians in the field to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent changes or modifications made based on the essence of the content of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for optimizing decision-making for offshore wind power participating in medium- and long-term bilateral negotiation transactions, characterized in that: include: S1. Obtain the optimal matching power load combination through the time series similarity measurement method; S2. Based on the optimal matching power load combination, considering the user's demand response resource reserve capacity and the power generation and consumption curve prediction error, a medium- and long-term transaction decision model based on two-stage distributed robust optimization is constructed; S3, solving the medium- and long-term transaction decision model based on two-stage distributed robust optimization to generate offshore wind power transaction decision results; The step S1 comprises: S11, obtaining a data sample set D, wherein the data sample set D is formed by all power load combination curves and offshore wind power generation curves; S12, using a curve numerical similarity matching model based on K-Medoids clustering on the data sample set D to obtain a power consumption curve set U; S13, using a curve shape similarity matching model based on Fast-DTW measurement for the power consumption curve set U, to obtain a Fast-DTW distance set L between each power consumption load combination and the offshore wind power generation curve; S14, let λ = minL = min(l1,l2,...,l n ), the power load combination corresponding to λ is the optimal matching power load combination U b ; The objective function of the medium- and long-term trading decision model based on two-stage distributed robust optimization is: Among them, λ p is the time-based contract electricity price; P is the time period type set, where 1 represents the peak period, 2 represents the normal period, and 3 represents the valley period; N I N is the user DR resource set; T N is the callable time period set of DR; T，p The corresponding time periods of peaks, flats and valleys are as follows; is the up and down adjustment amount of demand response resource i at time t; L t is the baseline load L before adjustment at time t t θ p To simulate the spot time-of-use electricity price; Δ t is the difference between wind power output and total load after demand response optimization configuration at time t, Represents the output shortfall when wind power output is less than the total load. Represents the output surplus when wind power output is greater than the total load; α i is the calling cost of demand response i; β i is the reserve capacity price for demand response i; The reserve capacity of demand response i purchased for the contract.

2. The method for optimizing decision-making for offshore wind power participating in medium- and long-term bilateral negotiation transactions according to claim 1, characterized in that: The step S12 comprises: S121, randomly select k data samples from the data sample set D as particle O j (j=1,2,...,k); S122, randomly select a non-particle sample O random ; Calculate the exchange object O random and O1 as reference point, repeat the operation in step S121 to generate a new set of clusters, calculate the objective function S, and if S<0, random Exchange with O1 and keep the new cluster, otherwise, keep the original center point and cluster; repeat this step until the k center points do not change; S123, the power load combination of the same cluster as the offshore wind power generation curve forms a power load combination set U = (u1, u2, ..., u n ), where u n Indicates different combinations of power loads.

3. The method for optimizing decision-making for offshore wind power participating in medium- and long-term bilateral negotiation transactions as claimed in claim 2, characterized in that: The intra-cluster squared error and SSE are used as clustering evaluation indicators. The sum of the squares of the distances from all data points to the centroid of the cluster to which they belong is calculated. The optimal number of clusters k is determined by the elbow method. The calculation expression of SSE is as follows: Among them, C i is the i-th cluster, p is C i The sample points in m i It is C i The centroid of , SSE is the clustering error of all samples, which represents the quality of clustering effect.

4. The method for optimizing decision-making for offshore wind power participating in medium- and long-term bilateral negotiation transactions according to claim 2, characterized in that: The curve morphology similarity matching model measured by Fast-DTW is based on the DTW algorithm, and Fast-DTW is implemented by restricting the search space.

5. The method for optimizing decision-making for offshore wind power participating in medium- and long-term bilateral negotiation transactions as claimed in claim 4, characterized in that: The DTW algorithm includes: For two given time series X=(x1,x2,...,x n ) and Y=(y1,y2,...,y m ), construct an n×m distance matrix D n×m , the matrix element calculation formula is as follows: The purpose of DTW is to find an optimal curved path W = (w1, w2, ..., w k ), satisfying the point pair w1 = (1,1) to w k =(n,m), there is a minimum cumulative distance: Where w s Represents the position of the path element in the distance matrix, and usually needs to satisfy the boundary, continuity and monotonicity. To solve the above formula, the cumulative matrix A needs to be constructed through dynamic programming method. n×m : Where i∈[1,n], j∈[1,m], the dynamic time warping distance of time series X and Y is DTW(X,Y)=A(n,m).

6. The method for optimizing decision-making for offshore wind power participating in medium- and long-term bilateral negotiation transactions according to claim 1, characterized in that: The first stage of the medium- and long-term transaction decision model based on two-stage distributed robust optimization is the user's DR resource spare capacity purchase decision stage, which determines the optimal DR capacity configuration for medium- and long-term offshore wind power transactions; the second stage is the simulation operation stage, which uses the spot time-of-use electricity price as a penalty for insufficient wind power output to calculate the expected profit of the wind power supplier in medium- and long-term transactions.

7. An offshore wind power participation in medium- and long-term bilateral negotiation transaction optimization decision-making device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

8. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images