Electric power spot market day-ahead clearing result prediction method

Abstract

The invention discloses an electric power spot market day-ahead clearing result prediction method, and belongs to the technical field of electric power markets. According to the method, data such as electric power spot market disclosure D-day data, spot market historical disclosure data, spot market historical clearing data and customer intention data are mainly used as data sources, and similar Manhattan distance similarity and Euclidean distance similarity calculation methods are adopted to screen most similar data in a database; prediction of the day-ahead electric energy clearing price of the electric power spot market on the D-1 day is achieved through a similar approximation inverse method, prediction of the day-ahead unit output of the user is achieved through combination of the day-ahead electric energy clearing price and the step output quotation of the power generation user using the algorithm, and market clearing trend data reference is provided for the user. The method assists a market subject in obtaining initiative in stepped output quotation and medium-and-long-term contract electric quantity distribution of an electric power spot market, and improves the transaction capability and risk avoiding capability of the market subject.

Description

Technical Field

[0001] The present invention belongs to the technical field of power market, and in particular relates to a method for predicting day-ahead clearing results in a power spot market. Background Art

[0002] The medium- and long-term electricity market has been officially operational since 2015. In September 2023, the National Development and Reform Commission and the National Energy Administration issued a notice on the issuance of the "Basic Rules of the Electricity Spot Market (Trial)" to gradually promote the operation of the electricity spot market. The electricity spot market will be divided into 96 time periods of 15 minutes each day. D-day day-ahead electricity energy transactions will be carried out on D-1, and the transaction objects will be the power generation and consumption from the next natural day to the next working day. Each natural day contains 96 transaction clearing periods (00:15-24:00), and each 15 minutes is a transaction clearing period. The 96 time period day-ahead clearing price curves are formed based on factors such as electricity supply and demand. For example Figure 1 As shown in Figure 1, it is the day-ahead clearing price curve of the electricity spot market. Due to the short time of market formation, there is no mature solution for the day-ahead clearing price curve forecast and clearing power forecast in the electricity spot market. Summary of the invention

[0003] The purpose of the present invention is to provide a method for predicting the day-ahead clearing results of the electricity spot market, which is mainly used in the field of ex ante prediction of electricity spot market transactions, and provides a prediction algorithm for the day-ahead clearing price of electricity energy and the day-ahead unit output of the electricity spot market, to assist the main body of the electricity spot market in estimating the day-ahead clearing price of electricity energy and the day-ahead unit output, and to adjust the supporting transaction strategies. The present invention uses the D-day data disclosed by the electricity spot market, the historical disclosure data of the spot market, the historical clearing data of the spot market, and the customer intention data as data sources to implement the D-1 day to D-day electricity spot market clearing prediction algorithm. The specific technical scheme is: a method for predicting the day-ahead clearing results of the electricity spot market, which is divided into the day-ahead clearing price prediction of electricity energy and the day-ahead unit output prediction, and includes the following steps:

[0004] Step 1: Select the reference day most similar to D-day from the database, and remove the priority power generation load (coal-fired power generation, gas-fired power generation, wind power generation, photovoltaic power generation, and power transmission from outside the region) from the market power load on D-day to obtain the predicted load of coal-fired units in the market on D-day, recorded as P 火-D日 , its data structure:

[0005] P 火-D日 =P D日-市场用电负荷 -P D日-风力发电 -P D日-光伏发电 -P D日-区外送电 -P D日-燃气发电

[0006] =[P 火-D日-1 ,P 火-D日-2,P 火-D日-3 ...P 火-D日-94 ,P 火-D日-95 ,P 火-D日-96 ],

[0007] Similarly, the predicted load P of all coal-fired units in the market on all historical days can be obtained: 火-X日 , respectively, and P 火-D日 Perform comparison screening, the screening includes the following steps:

[0008] S11: First step of screening: Using the Manhattan distance similarity calculation method, the sum of the predicted loads of coal-fired units in the market during the 96 periods of all historical days is compared with the sum of the predicted loads of coal-fired units in the market during the 96 periods of the D day, and the data similarity parameter R is obtained. 1-X日 , its calculation formula is:

[0009]

[0010] Here, the minimum m historical days are taken as alternative reference days.

[0011] S12: The second step of screening: Using the Euclidean distance similarity calculation method, the predicted loads of the coal-fired units in the 96-hour market of the m alternative reference days are calculated in turn with the predicted loads of the coal-fired units in the 96-hour market of the D day, and the data similarity parameter R is obtained. 2-X日 , its calculation formula is:

[0012]

[0013] The smallest alternative reference date is taken as the reference date, recorded as C day.

[0014] Step 2: Derive the tiered quotation set of coal-fired units in the whole market on D day, use the tiered quotation set of coal-fired units in the whole market on the most similar reference day C day to replace D day, participate in the day-ahead electricity energy clearing price forecast, and convert the coal-fired unit forecast load P on C day into 火-C日 、C day-ahead electricity energy clearing priceJ 日前-C日 Form the ladder quotation set B of all coal-fired units in the market on day C 火-C日 ; The database calls the predicted load P of the coal-fired unit on day C 火-C日 :

[0015] P 火-C日 =[P 火-C日-1 ,P 火-C日-2 ,P 火-C日-3 ...P 火-C日-94 ,P 火-C日-95 ,P 火-C日-96 ],

[0016] The clearing price of electricity energy on day C is recorded as J 日前-C日 :

[0017] J 日前-C日 =[J 日前-C日-1 ,J 日前-C日-2 ,J 日前-C日-3 ……J 日前-C日-94 ,J 日前-C日-95 ,J 日前-C日-96 ],

[0018] Combining two variable arrays, namely the ladder quotation set of coal-fired units in the whole market, denoted as B 火-C日 , its data structure:

[0019] B 火-C日 =[(P 火-C日-1 ,J 日前-C日-1 ),(P 火-C日-2 ,J 日前-C日-2 ),(P 火-C日-3 ,J 日前-C日-3 ),……

[0020] ,(P 火-C日-94 ,J 日前-C日-94 ),(P 火-C日-95 ,J 日前-C日-95 ),(P 火-C日-96 ,J 日前-C日-96 )],

[0021] The array variable C is the daily coal-fired unit predicted load P 火-C日 Arrange in ascending order from left to right and follow the J 日前-C日 Sort in ascending order;

[0022] Step 3: Prediction of the electricity clearing price on D day: P 火-D日 As B 火-C日 Load margin, get the day-ahead electricity energy clearing price J 日前-D日 , its data structure:

[0023] J 日前-D日 =[J 日前-D日-1 ,J 日前-D日-2 ,J 日前-D日-3 ……J 日前-D日-94 ,J 日前-D日-95 ,J 日前-D日-96 ].

[0024] The day-ahead unit output forecast includes the following steps:

[0025] S1: Use the day-ahead electricity energy clearing price J obtained in step 3 above 日前-D日 ;

[0026] S2: Get the step output quotation on D day. According to the customer's local unit output quotation on D day, the step output quotation can be obtained and recorded as B 本地-D日 , B 本地-D日It is a three-variable array, and its data structure is [load start point, load end point, quotation]:

[0027] B 本地-D日 =[(S1,E1,J1),(S2,E2,J2)……(S9,E9,J9),(S 10 ,E 10 ,J 10 )];

[0028] The array variables load start point S and load end point E are arranged in ascending order from left to right according to market rules;

[0029] S3: The unit output forecast on D day, the electricity energy clearing price J on D day 日前-D日 96 sets of data are sequentially 本地-D日 Compare and form the unit output P on D day 本地-D日 , its data structure is as follows:

[0030] P 本地-D日 =[P 本地-D日-1 ,P 本地-D日-2 ,P 本地-D日-3 ...P 本地-D日-94 ,P 本地-D日-95 ,P 本地-D日-96 ].

[0031] As a further description of the aforementioned scheme, the database stores historical day-ahead electricity energy clearing data, including information disclosed before day-ahead clearing: market electricity load, coal-fired power generation, gas-fired power generation, wind power generation, photovoltaic power generation, out-of-region power transmission, and day-ahead electricity energy clearing price.

[0032] In step 3, the electricity spot market rules require that a day be divided into X periods. Only the third paragraph formula needs to be adjusted, which does not affect the validity of this algorithm. After adjustment, it is:

[0033] When P 火-D日-1 >P 火-C日-X When J 日前-D日-1 =J 日前-C日-X .

[0034] In step 3, the first predicted load P of the coal-fired unit in the market on day D is 火-D日-1 Take the first electricity clearing price J as an example to describe the D-day. 日前-D日-1 Derivation algorithm:

[0035] When P 火-D日-1 ≤P 火-C日-1 When J 日前-D日-1 =J 日前-C日-1

[0036] When P 火-C日-n <P 火-D日-1 ≤P火-C日-n+1 When J 日前-D日-1 =J 日前-C日-n+1

[0037] When P 火-D日-1 >P 火-C日-96 When J 日前-D日-1 =J 日前-C日-96 .

[0038] Furthermore, in S3, if the number of quotation segments is Y, only the third segment formula needs to be adjusted without affecting the validity of the algorithm. After adjustment, it is

[0039] When J 日前-D日-1 ≥J Y When P 本地-D日-1 =E Y .

[0040] Compared with the prior art, the present invention discloses an algorithm for predicting the day-ahead electricity energy clearing price and the day-ahead unit output in the electricity spot market. The algorithm is applied to the quotation strategy and medium- and long-term contract formulation of the electricity spot market entities (power generation enterprises, power sales companies). For market entities, predicting the next day's market clearing situation is very necessary for the strategy formulation, revenue forecasting, and risk control of the electricity spot market. The "similar approximation inverse method" adopted by the present invention realizes the prediction of key data of "day-ahead electricity energy clearing price" and "day-ahead unit output", provides data reference for the market clearing trend, assists market entities in taking the initiative in the step output quotation and medium- and long-term contract power allocation in the electricity spot market, and enhances the trading ability and risk hedging ability of market entities. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a schematic diagram of the day-ahead clearing price curve for the electricity spot market;

[0042] Figure 2 This is a schematic diagram of the tiered output quotation for local units;

[0043] Figure 3 This is a schematic diagram of the tiered quotation collection of coal-fired units in the entire market today;

[0044] Figure 4 This is a schematic diagram of the formation principle of the day-ahead electricity energy clearing price;

[0045] Figure 5 This is a schematic diagram of the principle of the unit output formation on the previous day;

[0046] Figure 6 This is the flow chart of the day-ahead electricity energy clearing price;

[0047] Figure 7 A schematic diagram of the derivation of the tiered quotation set for coal-fired units in the entire market;

[0048] Figure 8 This is the flow chart for predicting the unit output curve on the day before.

[0049] Main terms explanation:

[0050] Electricity Spot Market: A general term for electricity trading activities that are carried out on the spot trading platform on the day before and within a shorter period of time, from the next day, intraday to real-time dispatch.

[0051] Medium and Long-term Transaction: A transaction for the delivery of power products or services within a certain period in the future, including transactions in different time dimensions such as several years, years, months, weeks, and days. Medium and long-term transaction contracts include physical contracts and financial contracts.

[0052] Day-ahead Market: An electric energy market that determines the unit combination status and power generation plan on the operating day (D-day) one day in advance (D-1).

[0053] Market Clearing: The electricity market determines transaction volume and price through competition according to market rules.

[0054] Market Settlement: The calculation and transfer of funds related to market members' participation in the electricity market within a specified period based on transaction results and relevant market rules. DETAILED DESCRIPTION

[0055] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. The preferred embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thoroughly and comprehensively understood.

[0056] The technical solution of this patent is further described in detail below in conjunction with specific implementation methods.

[0057] The principle of the day-ahead electricity clearing price formation in the electricity spot market:

[0058] There are many types of power sources in the electricity spot market: coal-fired power generation, gas-fired power generation, wind power generation, photovoltaic power generation, and inter-regional power transmission. The pricing units that actually participate in the formation of the clearing price are coal-fired power generation. On D-1, all the centrally dispatched coal-fired units will make tiered quotations for different power generation capacities according to their own conditions, forming a tiered output quotation curve with up to 10 sections, as shown below: Figure 2The market organizer collects the tiered output quotations of all the centrally dispatched coal-fired units to form a tiered quotation set of all the coal-fired units in the market on D day, as shown in the figure below. Figure 3 As shown in the figure, the market coal-fired unit load in each trading clearing period (every 15 minutes) on the D day is used as the margin of the X-axis of the ladder quotation set, and the corresponding Y value is the day-ahead electricity clearing price, forming a 96-period day-ahead electricity clearing price curve.

[0059] Explanation: The power grid requires real-time supply and demand balance, so not all tiered quotations can be cleared. Generally, the power grid has positive reserve (the maximum power generation capacity is greater than the market electricity load), so the market electricity load is used as the clearing margin to meet the balance of power generation supply and demand.

[0060] Since the electricity energy clearing price before D-day is formed on D-1, the market electricity load cannot be real. Therefore, the market organizer will estimate the predicted load of coal-fired units in the 96 periods of D-day as the X-axis clearing margin of the step quotation set.

[0061] by Figure 4 As an example, take the clearing period of 07:00-07:15 on D day as an example. At this time, the predicted load of coal-fired units in the market is 70,000MW. Taking this as the aggregate margin of the tiered quotation of coal-fired units, the clearing price of electricity energy at this time is 600 yuan / MWh. By analogy, taking the predicted load of coal-fired units in the market during the 96 period as the clearing margin, we can get the following: Figure 4 The electricity energy clearing price curve for the 96-period period is shown.

[0062] The principle of the day-ahead unit output formation in the electricity spot market:

[0063] The day-ahead unit output refers to the unit output curve formed based on the quotation of a single unit.

[0064] The day-ahead electricity clearing price for each trading clearing period (every 15 minutes) on D-day obtained above is used as the margin of the Y-axis of the step output quotation curve of each coal-fired unit. The corresponding X value is the day-ahead unit output, forming a 96-period day-ahead unit output curve.

[0065] by Figure 5 Take this as an example to illustrate: Figure 5 is the step output quotation curve of the unit on D day, from Figure 4 Given the clearing price curve of the day-ahead electricity energy in the 96-hour period on D-day, take the clearing period of 07:00-07:15 on D-day as an example. The day-ahead electricity energy clearing price at this time is 600 yuan / MWh. This is used as the margin of the step output quotation curve, and the day-ahead unit output at this time is 750MW. By analogy, taking the day-ahead electricity energy clearing price in the 96-hour period as the clearing margin, we can get the following: Figure 5 The unit output curve for the 96-day period is shown.

[0066] In order to have a more thorough and comprehensive understanding of the disclosure of the present invention, its principles are further explained below.

[0067] Due to the principle of day-ahead electricity clearing price formation in the electricity spot market, the "tiered quotation set of coal-fired units in the entire market" and "forecast load of coal-fired units in the market during the 96-hour period on D-day" used are non-public data. Therefore, there is a lack of data sources to derive the day-ahead electricity clearing price and day-ahead unit output according to the above principles.

[0068] This scheme adopts the "similar approximation inverse method" to derive the above two non-public data, so as to predict the day-ahead electricity energy clearing price on D day, and then derive the day-ahead unit output.

[0069] The following will describe the prediction methods of the day-ahead electricity energy clearing price and the day-ahead unit output respectively.

[0070] Day-ahead electricity clearing price forecast: based on Figure 4 It can be seen that the core data that forms the day-ahead electricity energy clearing price consists of two parts:

[0071] 1) Forecasted load of coal-fired units in the market during the 96-hour period on D day

[0072] 2) Tiered quotation collection of coal-fired units in the whole market

[0073] like Figure 5 As shown, the day-ahead electricity energy clearing price prediction process is divided into four steps. The following describes the two data derivation methods and the day-ahead clearing price prediction step by step.

[0074] Step 1: Screening for the reference date most similar to D-day

[0075] The algorithm database will record all historical day-ahead electricity energy clearing data, including day-ahead information disclosure before clearing (market electricity load, coal-fired power generation, gas-fired power generation, wind power generation, photovoltaic power generation, and out-of-region power transmission) and day-ahead electricity energy clearing prices.

[0076] By removing the priority power generation load (coal-fired power generation, gas-fired power generation, wind power generation, photovoltaic power generation, and power transmission from outside the region) from the market power load on D day, the predicted load of coal-fired units in the market on D day is obtained, which is recorded as P 火-D日 , its data structure:

[0077] P 火-D日 =P D日-市场用电负荷 -P D日-风力发电 -P D日-光伏发电 -P D日-区外送电 -P D日-燃气发电

[0078] =[P 火-D日-1 ,P 火-D日-2 ,P火-D日-3 ...P 火-D日-94 ,P 火-D日-95 ,P 火-D日-96 ].

[0079] Similarly, the predicted load P of all coal-fired units in the market on all historical days can be obtained: 火-X日 , respectively, and P 火-D日 Perform comparative screening.

[0080] The first step of screening: Using the Manhattan distance similarity calculation method, the predicted load of coal-fired units in the market during the 96 periods of all historical days is compared with the predicted load of coal-fired units in the market during the 96 periods of the D day, and the data similarity parameter R is obtained. 1-X日 , its calculation formula is:

[0081] Here, the minimum m historical days are taken as alternative reference days.

[0082] The second step of screening: using the Euclidean distance similarity calculation method, the predicted load of the coal-fired units in the 96-hour market of the alternative reference day and the predicted load of the coal-fired units in the 96-hour market of the D-day are calculated in turn, and the data similarity parameter R is obtained. 2-X日 , its calculation formula is:

[0083]

[0084] The smallest alternative reference date is taken as the reference date, recorded as C day.

[0085] Step 2: Collection of tiered quotations of coal-fired units in the entire market on D day

[0086] Since the tiered quotation set of coal-fired units in the entire market is non-public data, and the D-day has not been cleared, the disclosed data is incomplete. Therefore, the tiered quotation set of coal-fired units in the entire market on the most similar reference day C is used to replace D-day to participate in the day-ahead electricity energy clearing price forecast. Figure 7 As shown in the figure, the predicted load P of coal-fired unit on day C is shown in the figure. 火-C日 、C day-ahead electricity energy clearing priceJ 日前-C日 Combined to form the ladder quotation set B of all coal-fired units in the market on day C 火-C日 ,

[0087] The algorithm database calls the predicted load P of the coal-fired unit on day C 火-C日 :

[0088] P 火-C日 =[P 火-C日-1 ,P 火-C日-2 ,P 火-C日-3 ...P 火-C日-94 ,P 火-C日-95 ,P 火-C日-96],

[0089] The day-ahead electricity energy clearing price (denoted as J 日前-C日 ):

[0090] J 日前-C日 =[J 日前-C日-1 ,J 日前-C日-2 ,J 日前-C日-3 ……J 日前-C日-94 ,J 日前-C日-95 ,J 日前-C日-96 ].

[0091] Combining two variable arrays, namely the ladder quotation set of coal-fired units in the whole market, denoted as B 火-C日 , its data structure:

[0092] B 火-C日 =[(P 火-C日-1 ,J 日前-C日-1 ),(P 火-C日-2 ,J 日前-C日-2 ),(P 火-C日-3 ,J 日前-C日-3 ),……

[0093] ,(P 火-C日-94 ,J 日前-C日-94 ),(P 火-C日-95 ,J 日前-C日-95 ),(P 火-C日-96 ,J 日前-C日-96 )].

[0094] The array variable C is the daily coal-fired unit predicted load P 火-C日 Arrange in ascending order from left to right and follow the J 日前-C日 Sort in ascending order.

[0095] Step 3: Get the forecast load of coal-fired units on D-day:

[0096] In step 1 above, D-day - coal-fired unit predicted load P 火-D日 Already calculated and ready for use.

[0097] Step 4: Prediction of the electricity energy clearing price on D day:

[0098] Referring to the principle of the formation of the day-ahead electricity clearing price in the electricity spot market, P 火-D日 As B 火-C日 Load margin, get the day-ahead electricity energy clearing price J 日前-D日 , its data structure:

[0099] J 日前-D日 =[J 日前-D日-1 ,J 日前-D日-2 ,J 日前-D日-3 ……J 日前-D日-94,J 日前-D日-95 ,J 日前-D日-96 ].

[0100] Take the first forecast load P of the coal-fired unit on the D day market 火-D日-1 Take the first electricity clearing price J as an example to describe the D-day. 日前-D日-1 Derivation algorithm:

[0101] When P 火-D日-1 ≤P 火-C日-1 When J 日前-D日-1 =J 日前-C日-1

[0102] When P 火-C日-n <P 火-D日-1 ≤P 火-C日-n+1 When J 日前-D日-1 =J 日前-C日-n+1

[0103] When P 火-D日-1 >P 火-C日-96 When J 日前-D日-1 =J 日前-C日-96 .

[0104] This algorithm takes the current electricity spot market rules that divide a day into 96 periods as an example. If the market rules require a day to be divided into X periods, only the third paragraph formula needs to be adjusted, which does not affect the effectiveness of this algorithm. After adjustment, it is:

[0105] When P 火-D日-1 >P 火-C日-X When J 日前-D日-1 =J 日前-C日-X .

[0106] The unit output forecast for the day before:

[0107] like Figure 8 As shown in the figure, the day-ahead unit output prediction process consists of three steps, which will be described step by step below.

[0108] Step 1: Obtain the electricity clearing price before D day

[0109] The above has calculated the day-ahead electricity energy clearing price J on D 日前-D日 , which can be obtained directly.

[0110] Step 2: Get the D-day tiered output quotation

[0111] According to the quotation of the local unit output of customer D, Figure 2 As shown, the step output quotation can be recorded as B 本地-D日 , B 本地-D日 It is a three-variable array, and its data structure is [load start point, load end point, quotation].

[0112] B本地-D日 =[(S1,E1,J1),(S2,E2,J2)……(S9,E9,J9),(S 10 ,E 10 ,J 10 )].

[0113] The array variables load starting point S and load end point E are arranged in ascending order from left to right according to market rules.

[0114] Step 3: Unit output forecast before D day

[0115] The day-ahead electricity energy clearing price J 日前-D日 96 sets of data are sequentially 本地-D日 Compare and form the unit output P on D day 本地-D日 , its data structure is as follows:

[0116] P 本地-D日 =[P 本地-D日-1 ,P 本地-D日-2 ,P 本地-D日-3 ...P 本地-D日-94 ,P 本地-D日-95 ,P 本地-D日-96 ].

[0117] The first clearing price J of the electricity energy on the D day 日前-D日-1 For example, describe the output P of the first local unit before D day. 本地-D日-1 The derivation algorithm is as follows:

[0118] When J 日前-D日-1 <J1, P 本地-D日-1 =0

[0119] When J n ≤J 日前-D日-1 <J n+1 When P 本地-D日-1 =E n

[0120] When J 日前-D日-1 ≥J 10 When P 本地-D日-1 =E 10 .

[0121] This algorithm takes the example that the power spot market rules allow local units to bid for a maximum of 10 output segments. If the maximum number of bid segments is Y, only the third segment formula needs to be adjusted, which does not affect the effectiveness of this algorithm. After adjustment, it is

[0122] When J 日前-D日-1 ≥J Y When P 本地-D日-1 =E Y .

[0123] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0124] The above implementation modes are only used to illustrate the present invention, but not to limit the present invention. Ordinary technicians in the relevant technical field can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, all equivalent technical solutions also belong to the scope of the present invention. The patent protection scope of the present invention should be defined by the claims.

Claims

1. A method for predicting the day-ahead clearing results of the electricity spot market, characterized by: The following steps are involved: Step 1: Select the reference day that is most similar to D-day from the database, and remove the priority power generation load from the market power load on D-day to obtain the predicted load of coal-fired units in the market on D-day, recorded as P 火-D日 , its data structure: P 火-D日 =P D日-市场用电负荷 -P D日-风力发电 -P D日-光伏发电 -P D日-区外送电 -P D日-燃气发电 =[P 火-D日-1 ,P 火-D日-2 ,P 火-D日-3 ……P 火-D日-94 ,P 火-D日-95 ,P 火-D日-96 ], Similarly, the predicted load P of all coal-fired units in the market on all historical days can be obtained: 火-X日 , respectively, and P 火-D日 Perform comparison screening, the screening includes the following steps: S11: First step of screening: Using the Manhattan distance similarity calculation method, the sum of the predicted loads of coal-fired units in the market during the 96 periods of all historical days is compared with the sum of the predicted loads of coal-fired units in the market during the 96 periods of the D day, and the data similarity parameter R is obtained. 1-X日 , its calculation formula is: Take the m historical days with the smallest calculation results as the candidate reference days; S12: The second step of screening: Using the Euclidean distance similarity calculation method, the predicted load of the coal-fired units in the market during the 96 periods of the alternative reference day and the predicted load of the coal-fired units in the market during the 96 periods of the D day are calculated in turn by the Euclidean distance similarity, and the data similarity parameter R is obtained. 2-X日 , its calculation formula is: Take the minimum R 2-X日 The historical day is the reference day, recorded as C day: Step 2: Derive the tiered quotation set of coal-fired units in the whole market on day D, use the tiered quotation set of coal-fired units in the whole market on the most similar reference day C to replace day D, participate in the day-ahead electricity energy clearing price forecast, and convert the forecast load P of coal-fired units on day C into 火-C日 、C day-ahead electricity energy clearing priceJ 日前-C日 Form the ladder quotation set B of all coal-fired units in the market on day C 火-C日 ; The database calls the predicted load P of the coal-fired unit on day C 火-C日 : P 火-C日 =[P 火-C日-1 ,P 火-C日-2 ,P 火-C日-3 ……P 火-C日-94 ,P 火-C日-95 ,P 火-C日-96 ], The clearing price of electricity energy on day C is recorded as J 日前-C日 : I 日前-C日 =[J 日前-C日-1 ,J 日前-C日-2 ,J 日前-C日-3 ……I 日前-C日-94 ,J 日前-C日-95 ,J 日前-C日-96 ], Combining two variable arrays, namely the ladder quotation set of coal-fired units in the whole market, denoted as B 火-C日 , its data structure: B 火-C日 =[(P 火-C日-1 ,J 日前-C日-1 ),(P 火-C日-2 ,J 日前-C日-2 ),(P 火-C日-3 ,J 日前-C日-3 ),……,(P 火-C日-94 ,J 日前-C日-94 ),(P 火-C日-95 ,J 日前-C日-95 ),(P 火-C日-96 ,J 日前-C日-96 )], The array variable C is the daily coal-fired unit predicted load P 火-C日 Arrange in ascending order from left to right and follow the J 日前-C日 Sort in ascending order; Step 3: Prediction of the electricity energy clearing price on D day: P 火-D日 As B 火-C日 Load margin, get the day-ahead electricity energy clearing price J 日前-D日 , its data structure: I 日前-D日 =[J 日前-D日-1 ,J 日前-D日-2 ,J 日前-D日-3 ……I 日前-D日-94 ,J 日前-D日-95 ,J 日前-D日-96 ] 2. A method for predicting day-ahead clearing results in the electricity spot market according to claim 1, characterized in that: S1: Use the day-ahead electricity energy clearing price J obtained in step 3 on D 日前-D日 ; S2: Get the step output quotation on D day. According to the customer's local unit output quotation on D day, the step output quotation can be obtained and recorded as B 本地-D日 , B 本地-D日 It is a three-variable array, and its data structure is [load start point, load end point, quotation]: <h2 style=";text-align:left;direction:ltr">B<h2 style=";text-align:left;direction:ltr"> 本地-D日 <h2 style=";text-align:left;direction:ltr"> ([(S1,E1,J1),(S2,E2,J2)……(S9,E9,J9),(S<h2 style=";text-align:left;direction:ltr"> 10 <h2 style=";text-align:left;direction:ltr"> ,E<h2 style=";text-align:left;direction:ltr"> 10 <h2 style=";text-align:left;direction:ltr"> ,J<h2 style=";text-align:left;direction:ltr"> 10 <h2 style=";text-align:left;direction:ltr"> )]; The array variables load start point S and load end point E are arranged in ascending order from left to right according to market rules; S3: The unit output forecast on D day, the electricity energy clearing price J on D day 日前-D日 96 sets of data are sequentially 本地-D日 Compare and form the unit output P on D day 本地-D日 , its data structure is as follows: P 本地-D日 =[P 本地-D日-1 ,P 本地-D日-2 ,P 本地-D日-3 ……P 本地-D日-94 ,P 本地-D日-95 ,P 本地-D日-96 ]。 3. A method for predicting day-ahead clearing results in the electricity spot market according to claim 1, characterized in that: The database stores the day-ahead electricity energy clearing data of historical days, including the information disclosed before the day-ahead clearing: market electricity load, coal-fired power generation, gas-fired power generation, wind power generation, photovoltaic power generation, power transmission outside the region, and the day-ahead electricity energy clearing price.

4. The method for predicting the day-ahead clearing results of the electricity spot market according to claim 1, characterized in that: In step 3, if the electricity spot market rules require that a day be divided into X periods, only the third formula needs to be adjusted without affecting the validity of the algorithm. The adjusted formula is: When P 火-D日-1 >P 火-C日-X When J 日前-D日-1 =J 日前-C日-X .

5. The method for predicting the day-ahead clearing results of the electricity spot market according to claim 1, characterized in that: In step 3, the first predicted load P of the coal-fired unit in the market on day D is 火-D日-1 Take the first electricity clearing price J as an example to describe the D-day. 日前-D日-1 Derivation algorithm: When P 火-D日-1 ≤P 火-C日-1 When J 日前-D日-1 =J 日前-C日-1 When P 火-C日-n <P 火-D日-1 ≤P 火-C日-n+1 When J 日前-D日-1 =J 日前-C日-n+1 When P 火-D日-1 >P 火-C日-96 When J 日前-D日-1 =J 日前-C日-96 .

6. A method for predicting day-ahead clearing results in the electricity spot market according to claim 2, characterized in that: In S3, if the number of quotation segments is Y, only the third segment formula needs to be adjusted, which does not affect the validity of this algorithm. After adjustment, it is When J 日前-D日-1 ≥J Y When P 本地-D日-1 =E Y .

7. A method for predicting day-ahead clearing results in the electricity spot market according to claim 1, characterized in that: The priority power generation loads include coal-fired power generation, gas-fired power generation, wind power generation, photovoltaic power generation, and power transmission out of the region.

Images