A mid- to long-term wind farm power generation forecasting method considering multiple influencing factors

Abstract

The present invention relates to the field of wind power generation technology, and specifically to a method for predicting mid- and long-term power generation of a wind farm taking into account multiple influencing factors, including: performing horizontal and vertical predictions on the power generation of a target wind farm in a future period of time to obtain an initial power generation prediction sequence; determining a first influence coefficient according to the degree of influence of future weather on the power generation of photovoltaic power stations, hydropower stations and wind farms in the current region; determining a second influence coefficient according to changes in demand for industrial and commercial electricity consumption in the current region; determining a third influence coefficient according to changes in purchased and sold electricity in the current region; determining a fourth influence coefficient according to the operation and maintenance records of the target wind farm; and correcting the initial power generation prediction sequence according to the first influence coefficient, the second influence coefficient, the third influence coefficient and the fourth influence coefficient to obtain a final power generation prediction sequence for the target wind farm. The present invention can effectively improve the accuracy of mid- and long-term power generation prediction of a wind farm.

Description

Technical Field

[0001] The present invention relates to the technical field of wind power generation, and more particularly to a method for predicting mid- to long-term power generation of a wind farm taking multiple influencing factors into consideration. Background Art

[0002] The medium- and long-term power generation forecast of renewable energy wind farms is of great significance in terms of power grid operation and renewable energy optimization. It assists in formulating dispatch plans for conventional power regulation, building a medium- and long-term power balance system, and ensuring the adequacy of power supply in the power system.

[0003] Wind farm power generation forecasting is widely used in the production and marketing operations of new energy sites. At present, most of the mid- and long-term power generation forecasting methods for new energy wind farms are based on historical power generation data and meteorological data, or on mid- and short-term forecast data, which take fewer factors into consideration. In actual operation, there is a large deviation from the actual situation, which makes it difficult to meet the actual operation requirements. Summary of the invention

[0004] In view of this, the present invention provides a method for predicting the mid- and long-term power generation of a wind farm taking into account multiple influencing factors, which can effectively improve the prediction accuracy of the mid- and long-term power generation of a wind farm.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] A method for predicting mid- to long-term wind farm power generation taking into account multiple influencing factors, characterized by comprising:

[0007] Obtain the current power generation plan, historical power generation and related data of the target wind farm in the area where it is located, make horizontal and vertical predictions of the power generation of the target wind farm in the future, and obtain the initial power generation prediction sequence;

[0008] Determine the first impact coefficient according to the impact of future weather on the power generation of photovoltaic power stations, hydropower stations and wind farms in the current area;

[0009] Determine the second impact coefficient based on the change in demand for industrial and commercial electricity consumption in the current region;

[0010] Determine the third impact coefficient based on the changes in the amount of electricity purchased and sold in the current region;

[0011] Determine the fourth impact coefficient according to the operation and maintenance records of the target wind farm;

[0012] The initial power generation prediction sequence is corrected according to the first influence coefficient, the second influence coefficient, the third influence coefficient and the fourth influence coefficient to obtain a final power generation prediction sequence of the target wind farm.

[0013] Furthermore, the power generation in the same historical period refers to the power generation data of several days before and after the period each year starting from the start of power generation at the target wind farm; the relevant data at least includes temperature, humidity, industrial power consumption, commercial power consumption, purchased power, delivered power, the number of wind turbines shut down for maintenance, surrounding photovoltaic installed capacity and surrounding wind power installed capacity.

[0014] Furthermore, the horizontal prediction and the vertical prediction of the power generation of the target wind farm in the future period are respectively performed to obtain an initial power generation prediction sequence, including:

[0015] Predict the power generation sequence for a period of time in the future according to the time axis of days as the horizontal prediction value;

[0016] According to the year time axis, the power generation at each time point in the future is predicted as the longitudinal prediction value;

[0017] For any time node in the future period of time, the horizontal prediction value is corrected by using the vertical prediction value of the time node to obtain the initial power generation prediction sequence.

[0018] Furthermore, a weighted sum is performed on the horizontal prediction value and the vertical prediction value at any time node in a future period of time to obtain the initial power generation prediction sequence.

[0019] Furthermore, when making horizontal and vertical predictions of the power generation of a wind farm in the future, it also includes:

[0020] The long short-term memory network (LSTM) is used to perform horizontal and vertical predictions on the target wind farm's power generation for a period of time in the future, and the horizontal prediction value is corrected by the vertical prediction value to obtain the first initial power generation prediction sequence;

[0021] The autoregressive moving average model ARMA is used to make horizontal and vertical predictions of the power generation of the target wind farm in the future, and the horizontal prediction value is corrected by the vertical prediction value to obtain the second initial power generation prediction sequence;

[0022] The values ​​corresponding to each time node in the first initial power generation prediction sequence and the second initial power generation prediction sequence are averaged to obtain the initial power generation prediction sequence.

[0023] Furthermore, the first influence coefficient is determined according to the influence of future weather on the power generation of photovoltaic power stations, hydropower stations and wind farms in the current area, including:

[0024] The influence of weather on the power generation of photovoltaic power stations, hydropower stations and wind farms is recorded as , and ;

[0025] When the weather is forecast to be sunny, , and The values ​​of are all less than 1.0 and increase successively;

[0026] When rain is forecast, , and The values ​​of are all greater than 1.0;

[0027] When the weather is forecast to be cloudy, , and The values ​​of are all greater than 1.0;

[0028] Will , and The product of the three is taken as the first influence coefficient The value of is expressed as .

[0029] Furthermore, the calculation formula of the second influence coefficient is:

[0030]

[0031] in, The influence coefficient of the change in the demand for industrial and commercial electricity consumption in the current region on the medium- and long-term power generation of the wind farm, that is, the second influence coefficient; It indicates the industrial electricity demand in the current region one year ago; It indicates the commercial electricity demand in the current area one year ago; Indicates the industrial electricity demand in the current area during the current period; Indicates the commercial electricity demand in the current area during the current period.

[0032] Furthermore, the calculation formula of the third influence coefficient is:

[0033]

[0034] in, The influence coefficient of the change of the purchased electricity and sold electricity in the current region on the medium- and long-term power generation of the wind farm, that is, the third influence coefficient; Indicates the amount of electricity purchased in the current region one year ago; Indicates the amount of electricity sold in the current area one year ago; Indicates the amount of electricity purchased in the current area during the current period; It indicates the amount of electricity sold in the current area during the current period. When the amount of electricity purchased or sold increases, it is beneficial to power generation. When the amount of electricity purchased or sold decreases, it is not beneficial to power generation.

[0035] Furthermore, according to the operation and maintenance records of the target wind farm, the number of wind turbines shut down for maintenance is counted. and the number of wind turbines with insufficient generating capacity ,according to and The fourth influence coefficient is calculated using the following formula:

[0036]

[0037]

[0038]

[0039] in, represents the influence coefficient of the target wind farm operation and maintenance record on the medium- and long-term power generation of the wind farm, that is, the fourth influence coefficient; The coefficient indicating the impact of the number of wind turbines shut down for maintenance in the target wind farm on the medium- and long-term power generation of the wind farm; It indicates the influence coefficient of the number of wind turbines with insufficient power generation capacity in the target wind farm on the medium- and long-term power generation of the wind farm. Insufficient power generation capacity means that the maximum power actually generated by the wind turbine is less than 60% of the rated power. Indicates the number of wind turbines currently shut down for maintenance in the target wind farm; Indicates the number of wind turbines in the target wind farm that currently have insufficient power generation capacity; Indicates the total number of wind turbines currently in the target wind farm.

[0040] Furthermore, the calculation formula for the final power generation forecast sequence of the target wind farm is:

[0041]

[0042] in, represents the final power generation forecast sequence of the target wind farm; represents the first influence coefficient, represents the second influence coefficient; represents the third influence coefficient; represents the fourth influence coefficient; Represents the initial power generation forecast sequence.

[0043] It can be seen from the above technical solutions that, compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. The present invention performs two prediction methods for the power generation of the wind farm: horizontal prediction and vertical prediction. The two prediction values ​​are combined to correct the power generation prediction value, thereby achieving a preliminary prediction of the power generation and providing a basis for subsequent predictions.

[0045] 2. The present invention comprehensively considers multiple factors such as historical data, weather, changes in demand for industrial and commercial electricity consumption, changes in purchased and sold electricity, and operation and maintenance records, and fully considers the power generation plan of the power dispatching organization, the operation of other new energy power plants, and changes in demand-side electricity consumption, to predict the medium- and long-term power generation of wind farms, thereby improving the accuracy and credibility of medium- and long-term power generation forecasts of wind farms. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0047] Figure 1 A flow chart of a method for predicting medium- and long-term power generation of a wind farm taking into account multiple influencing factors provided by the present invention;

[0048] Figure 2 This is a schematic diagram of the horizontal prediction and vertical prediction provided by the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] like Figure 1 As shown, an embodiment of the present invention discloses a method for predicting medium- and long-term power generation of a wind farm taking into account multiple influencing factors, comprising the following steps:

[0051] S1. Obtain the current power generation plan, historical power generation and related data of the target wind farm in the area where the target wind farm is located, and make horizontal and vertical predictions of the power generation of the target wind farm in the future to obtain an initial power generation prediction sequence;

[0052] S2. Determine the first impact coefficient according to the impact of future weather on the power generation of photovoltaic power stations, hydropower stations and wind farms in the current area;

[0053] S3. Determine the second impact coefficient according to the change in demand for industrial and commercial electricity consumption in the current region;

[0054] S4. Determine the third impact coefficient according to the changes in the amount of electricity purchased and sold in the current region;

[0055] S5. determining a fourth influence coefficient according to the operation and maintenance record of the target wind farm;

[0056] S6. Correct the initial power generation prediction sequence according to the first influence coefficient, the second influence coefficient, the third influence coefficient and the fourth influence coefficient to obtain a final power generation prediction sequence of the target wind farm.

[0057] The above steps are further explained below.

[0058] S1. Obtain the current power generation plan, historical power generation and related data of the target wind farm in the area where the target wind farm is located.

[0059] Among them, the current power generation plan is obtained through the wind farm SCADA system, which refers to the power generation plan curve issued by the power dispatching agency to the wind farm. The power generation plan curve determines the power generation that the wind farm should have.

[0060] The power generation in the same historical period refers to the power generation data of several days before and after the period each year, starting from the start of power generation at the target wind farm; the present invention refers to the first 5 days and the last 40 days of the period; for example, the current date is December 10, 2024, and the same historical period refers to the data from December 5 to January 20 of the following year, starting from the start of power generation at the wind farm (i.e., the data of the first 5 days and the last 40 days).

[0061] The relevant data include at least temperature, humidity, industrial electricity consumption, commercial electricity consumption, purchased electricity, delivered electricity, number of wind turbines shut down for maintenance, surrounding photovoltaic installed capacity and surrounding wind power installed capacity.

[0062] After that, the power generation in the future period is predicted horizontally and vertically to obtain the initial power generation prediction sequence y0, which specifically includes:

[0063] The power generation sequence for a period of time in the future is predicted backward according to the time axis of days as the horizontal prediction value; for example, it is currently December 10, 2024, and the horizontal prediction is to predict the data from December 11 to January 20 of the following year (the next 40 days).

[0064] According to the year time axis, the power generation at each time node in the future period is predicted as the vertical prediction value; for example, the historical data obtained in the same period are from December 11th to January 20th of the next year from 2016 to 2023, then the data from December 11th to January 20th of the next year in 2024 are predicted; for December 10th, 2024, that is, the historical data of each type of data from December 10th, 2016 to 2023. The horizontal and vertical prediction process is as follows Figure 2 shown.

[0065] For any time node in the future, the horizontal prediction value is corrected by using the vertical prediction value of the time node to obtain the initial power generation prediction sequence. The specific correction method can be a weighted summation method, that is, the horizontal prediction value and the vertical prediction value of any time node in the future are weighted and summed to obtain the initial power generation prediction sequence; the proportion of the vertical prediction value is 80%, and the proportion of the horizontal prediction value is 20%. The weight value can be adjusted according to the actual situation.

[0066] To further ensure the accuracy of the prediction results, the present invention can use multiple models to perform horizontal prediction and vertical prediction respectively, average the multiple results, and use the average value as the initial power generation prediction sequence. .

[0067] Specifically, the long short-term memory network LSTM is used to perform horizontal and vertical predictions on the power generation of the target wind farm in the future, and the horizontal prediction value is corrected by the vertical prediction value to obtain the first initial power generation prediction sequence ;

[0068] The autoregressive moving average model ARMA is used to make horizontal and vertical predictions of the power generation of the target wind farm in the future, and the horizontal prediction value is corrected by the vertical prediction value to obtain the second initial power generation prediction sequence. ;

[0069] The values ​​corresponding to each time node in the first initial power generation forecast sequence and the second initial power generation forecast sequence are averaged as the initial power generation forecast sequence .

[0070] S2. Determine the first impact coefficient based on the impact of future weather on the power generation of photovoltaic power stations, hydropower stations and wind farms in the current area. The current area refers to the area under the jurisdiction of the same power dispatching organization as the wind farm, including:

[0071] The influence of weather on the power generation of photovoltaic power stations, hydropower stations and wind farms is recorded as , and ;

[0072] When the weather is forecast to be sunny, , and The values ​​of are all less than 1.0 and increase successively; in this embodiment, The value of is 0.97. The value of is 0.98, The value of is 0.99.

[0073] When the weather is predicted to be rainy, it is not conducive to photovoltaic power generation. , and The values ​​of are all greater than 1.0; in this embodiment, The value of is 1.2, The value of is 1.03, The value of is 1.05.

[0074] When the weather is predicted to be cloudy, it is not conducive to photovoltaic power generation. , and The values ​​of are all greater than 1.0; The value of is 1.12, The value of is 1.01, The value of is 1.08.

[0075] Will , and The product of the three is taken as the first influence coefficient The value of is expressed as .

[0076] S3. Determine the second impact coefficient based on the change in demand for industrial and commercial electricity in the current region. , the specific calculation formula is:

[0077]

[0078] in, The influence coefficient of the change in the demand for industrial and commercial electricity consumption in the current region on the medium- and long-term power generation of the wind farm, that is, the second influence coefficient; It indicates the industrial electricity demand in the current region one year ago; It indicates the commercial electricity demand in the current area one year ago; Indicates the industrial electricity demand in the current area during the current period; Indicates the commercial electricity demand in the current area during the current period.

[0079] S4. Determine the third impact coefficient based on the changes in the amount of electricity purchased and sold in the current region , the specific calculation formula is:

[0080]

[0081] in, The influence coefficient of the change of the purchased electricity and sold electricity in the current region on the medium- and long-term power generation of the wind farm, that is, the third influence coefficient; Indicates the amount of electricity purchased in the current region one year ago; Indicates the amount of electricity sold in the current area one year ago; Indicates the amount of electricity purchased in the current area during the current period; It indicates the amount of electricity sold in the current area during the current period. When the amount of electricity purchased or sold increases, it is beneficial to power generation. When the amount of electricity purchased or sold decreases, it is not beneficial to power generation.

[0082] S5. Determine the fourth impact coefficient based on the operation and maintenance records of the target wind farm , including:

[0083] According to the operation and maintenance records of the target wind farm, the number of wind turbines shut down for maintenance is counted and the number of wind turbines with insufficient generating capacity ,according to and The fourth influence coefficient is calculated using the following formula:

[0084]

[0085]

[0086]

[0087] in, represents the influence coefficient of the target wind farm operation and maintenance record on the medium- and long-term power generation of the wind farm, that is, the fourth influence coefficient; The coefficient indicating the impact of the number of wind turbines shut down for maintenance in the target wind farm on the medium- and long-term power generation of the wind farm; It indicates the influence coefficient of the number of wind turbines with insufficient power generation capacity in the target wind farm on the medium- and long-term power generation of the wind farm. Insufficient power generation capacity means that the maximum power actually generated by the wind turbine is less than 60% of the rated power. Indicates the number of wind turbines currently shut down for maintenance in the target wind farm; Indicates the number of wind turbines in the target wind farm that currently have insufficient power generation capacity; Indicates the total number of wind turbines currently in the target wind farm.

[0088] S5. Taking the above influencing factors into consideration, the final power generation forecast sequence of the target wind farm is calculated. The specific calculation formula is:

[0089]

[0090] in, represents the final power generation forecast sequence of the target wind farm; represents the first influence coefficient, represents the second influence coefficient; represents the third influence coefficient; represents the fourth influence coefficient; Represents the initial power generation forecast sequence.

[0091] The present invention improves the credibility of the medium- and long-term power generation forecast of the wind farm by integrating multiple factors such as historical data, weather, changes in demand for industrial and commercial electricity consumption, changes in purchased and sold electricity, and operation and maintenance records.

[0092] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0093] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting mid- to long-term wind farm power generation taking into account multiple influencing factors, characterized in that: include: Obtain the current power generation plan, historical power generation and related data of the target wind farm in the area where the target wind farm is located, perform horizontal and vertical predictions on the power generation of the target wind farm in the future period, and obtain an initial power generation prediction sequence; the related data at least include temperature, humidity, industrial power consumption, commercial power consumption, purchased power, delivered power, the number of wind turbines shut down for maintenance, surrounding photovoltaic installed capacity and surrounding wind power installed capacity; Determine the first impact coefficient according to the impact of future weather on the power generation of photovoltaic power stations, hydropower stations and wind farms in the current area; Determine the second impact coefficient based on the change in demand for industrial and commercial electricity consumption in the current region; Determine the third impact coefficient based on the changes in the amount of electricity purchased and sold in the current region; Determine the fourth impact coefficient according to the operation and maintenance records of the target wind farm; Correcting the initial power generation prediction sequence according to the first influence coefficient, the second influence coefficient, the third influence coefficient and the fourth influence coefficient to obtain a final power generation prediction sequence of the target wind farm; The process of determining the first influence coefficient includes: The influence of weather on the power generation of photovoltaic power stations, hydropower stations and wind farms is recorded as , and ; When the weather is forecast to be sunny, , and The values ​​of are all less than 1.0 and increase successively; When rain is forecast, , and The values ​​of are all greater than 1.0; When the weather is forecast to be cloudy, , and The values ​​of are all greater than 1.0; Will , and The product of the three is taken as the first influence coefficient The value of is expressed as .

2. The method for predicting mid- to long-term wind farm power generation considering multiple influencing factors according to claim 1, characterized in that: The power generation in the same historical period refers to the power generation data of several days before and after the period in each year, starting from the start of power generation of the target wind farm.

3. The method for predicting mid- to long-term wind farm power generation considering multiple influencing factors according to claim 1, characterized in that: The horizontal prediction and the vertical prediction of the power generation of the target wind farm in the future period are respectively performed to obtain the initial power generation prediction sequence, including: Predict the power generation sequence for a period of time in the future according to the time axis of days as the horizontal prediction value; According to the year time axis, the power generation at each time point in the future is predicted as the longitudinal prediction value; For any time node in the future period of time, the horizontal prediction value is corrected by using the vertical prediction value of the time node to obtain the initial power generation prediction sequence.

4. The method for predicting mid- to long-term wind farm power generation considering multiple influencing factors according to claim 3 is characterized in that: The initial power generation prediction sequence is obtained by performing weighted summation on the horizontal prediction value and the vertical prediction value at any time node in the future period of time.

5. The method for predicting mid- to long-term wind farm power generation considering multiple influencing factors according to claim 3, characterized in that: When making horizontal and vertical forecasts of wind farm power generation in the future, it also includes: The long short-term memory network (LSTM) is used to perform horizontal and vertical predictions on the target wind farm's power generation for a period of time in the future, and the horizontal prediction value is corrected by the vertical prediction value to obtain the first initial power generation prediction sequence; The autoregressive moving average model ARMA is used to make horizontal and vertical predictions of the power generation of the target wind farm in the future, and the horizontal prediction value is corrected by the vertical prediction value to obtain the second initial power generation prediction sequence; The values ​​corresponding to each time node in the first initial power generation prediction sequence and the second initial power generation prediction sequence are averaged to obtain the initial power generation prediction sequence.

6. The method for predicting mid- to long-term wind farm power generation considering multiple influencing factors according to claim 1, characterized in that: The calculation formula of the second influence coefficient is: ； in, The influence coefficient of the change in the demand for industrial and commercial electricity consumption in the current region on the medium- and long-term power generation of the wind farm, that is, the second influence coefficient; It indicates the industrial electricity demand in the current region one year ago; It represents the commercial electricity demand in the current area one year ago; Indicates the industrial electricity demand in the current area during the current period; Indicates the commercial electricity demand in the current area during the current period.

7. The method for predicting mid- to long-term wind farm power generation considering multiple influencing factors according to claim 1, characterized in that: The calculation formula of the third influence coefficient is: ； in, The influence coefficient of the change of the purchased electricity and sold electricity in the current region on the medium- and long-term power generation of the wind farm, that is, the third influence coefficient; Indicates the amount of electricity purchased in the current region one year ago; Indicates the amount of electricity sold in the current area one year ago; Indicates the amount of electricity purchased in the current area during the current period; It indicates the amount of electricity sold in the current area during the current period. When the amount of electricity purchased or sold increases, it is beneficial to power generation. When the amount of electricity purchased or sold decreases, it is not beneficial to power generation.

8. The method for predicting mid- to long-term wind farm power generation considering multiple influencing factors according to claim 1, characterized in that: According to the operation and maintenance records of the target wind farm, the number of wind turbines shut down for maintenance is counted and the number of wind turbines with insufficient generating capacity ,according to and The fourth influence coefficient is calculated using the following formula: ； ； ； in, represents the influence coefficient of the target wind farm operation and maintenance record on the medium- and long-term power generation of the wind farm, that is, the fourth influence coefficient; The coefficient indicating the impact of the number of wind turbines shut down for maintenance in the target wind farm on the medium- and long-term power generation of the wind farm; It indicates the influence coefficient of the number of wind turbines with insufficient power generation capacity in the target wind farm on the medium- and long-term power generation of the wind farm. Insufficient power generation capacity means that the maximum power actually generated by the wind turbine is less than 60% of the rated power. Indicates the number of wind turbines currently shut down for maintenance in the target wind farm; Indicates the number of wind turbines in the target wind farm that currently have insufficient power generation capacity; Indicates the total number of wind turbines currently in the target wind farm.

9. The method for predicting mid- to long-term wind farm power generation considering multiple influencing factors according to claim 1, characterized in that: The calculation formula for the final power generation forecast sequence of the target wind farm is: ； in, represents the final power generation forecast sequence of the target wind farm; represents the first influence coefficient, represents the second influence coefficient; represents the third influence coefficient; represents the fourth influence coefficient; Represents the initial power generation forecast sequence.

Images