Wind power generation day power supply guarantee capability prediction method and system and medium

Abstract

The invention discloses a wind power generation intraday power supply guarantee capability prediction method, and discloses a system with the wind power generation intraday power supply guarantee capability prediction method and a medium, and the wind power generation intraday power supply guarantee capability prediction method comprises the following steps: taking actually measured wind speed, wind direction and other meteorological data as input; the intra-day guaranteed power generation capacity of the wind power plant can be accurately predicted; through test data comparison, the guaranteed electric quantity in the prediction result can be increased by 10-20% compared with the traditional prediction, and the accuracy of wind power generation power prediction and the reliability of the guaranteed electric quantity are remarkably improved. The method can more accurately evaluate the power output of the wind power plant under different meteorological conditions, especially has higher reliability and practicability when guaranteeing electric power supply, assists stable operation of an electric power system and reasonable development of wind power resources, and has remarkable engineering application value and popularization potential.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy power prediction, and in particular to a method and system for predicting the power supply guarantee capability of wind power generation within a day. Background Art

[0002] The spatial distribution characteristics of wind power resources and the temporal fluctuation of wind power output directly affect the operation and dispatch of wind farms and the smooth operation of power grids. In order to support the efficient development and stable operation of wind power resources, accurate prediction of wind farm power output has become a necessary foundation. The current wind power prediction model is mainly based on meteorological data and wind turbine parameters, but under the requirement of ensuring reliable power supply, traditional models often cannot effectively predict power supply capacity, resulting in a large deviation between the prediction results and the actual situation, affecting the stable operation and guarantee capability of wind power in the power grid. Summary of the invention

[0003] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes a method for predicting the daily power supply guarantee capacity of wind power generation, which can use the measured wind speed, wind direction and other meteorological data as input to accurately predict the daily guaranteed power generation capacity of the wind farm; at the same time, through the comparison of test data, it is found that the proportion of guaranteed electricity in the prediction results can be increased by 10-20% compared with the traditional prediction, which significantly improves the accuracy of wind power generation prediction and the reliability of guaranteed electricity.

[0004] The present invention also provides a system and a medium having the above-mentioned method for predicting the power supply guarantee capability of wind power generation within a day.

[0005] According to the first aspect of the present invention, the method for predicting the power supply guarantee capability of wind power generation within a day is characterized by comprising the following steps:

[0006] Constructing a loss function, and determining the observation error between the assurance capability result and the actual observation value based on the loss function, and obtaining the observation deviation between the predicted power and the actual observation value;

[0007] Setting a supply guarantee coefficient, and using the supply guarantee coefficient to adjust the observed deviation to a corrected deviation;

[0008] Performing a truncation operation on the elements in the corrected deviation to restrict the elements to a specified range to obtain a restricted corrected deviation;

[0009] Determining a value of a penalty term based on the restricted corrected deviation and a preset penalty weight;

[0010] Apply different loss terms according to the relationship between the predicted value and the actual observed value, and calculate the average loss according to the loss values ​​of the different loss terms applied to different samples;

[0011] The protection capability of the power system is estimated based on the average loss.

[0012] The method for predicting the power supply guarantee capability of wind power generation within a day according to the embodiment of the present invention has at least the following beneficial effects:

[0013] This method not only takes into account the meteorological factors in the traditional power forecasting model, but also focuses on introducing a penalty mechanism for positive deviations, that is, strengthening the constraints on the overestimated part of the power forecast in the model to reduce the problem of insufficient power supply caused by over-forecasting. In addition, the present invention comprehensively considers the guarantee capability of the power system through a game model that maximizes the proportion of guaranteed power supply, so that the prediction results can better serve the grid dispatching and operation needs. The power output of wind farms under different meteorological conditions can be evaluated more accurately, especially when ensuring power supply, it has higher reliability and practicality, helps the stable operation of the power system and the rational development of wind power resources, and has significant engineering application value and promotion potential.

[0014] According to some embodiments of the present invention, the loss function E MSE It can be expressed as:

[0015]

[0016] Where: y true_i is the actual observed value of the ith data point; y pred_i is the predicted value of the ith data point; n is the number of samples.

[0017] According to some embodiments of the present invention, the formula for obtaining the observed deviation d between the predicted power and the actual observed value can be expressed as:

[0018] d=y pred_i -y true_i

[0019] According to some embodiments of the present invention, the supply guarantee coefficient β can correct part of the negative deviation to a positive deviation, and the expression of the corrected deviation Z is:

[0020] Z=d+β

[0021] According to some embodiments of the present invention, in the step of performing a truncation operation on the elements in the corrected deviation and limiting the elements within a specified range to obtain the restricted corrected deviation, a clamp function is used to perform a truncation operation on the elements in the corrected deviation, that is, elements less than 0 in Z are truncated to 0; that is, negative values ​​are set to 0, which can be specifically expressed as:

[0022]

[0023] According to some embodiments of the present invention, the step of determining the value of the penalty term based on the limited corrected deviation and the preset penalty weight is formulated as follows:

[0024]

[0025] Among them, Q penalty is the penalty term, and α is the penalty weight.

[0026] According to some embodiments of the present invention, the evaluation function of the method is:

[0027]

[0028] Among them, x represents the parameter vectors α, β and γ that need to be optimized; g(x) is the guarantee rate; M is a very large constant used to punish those that do not meet g 1 (x) the solution required;

[0029] The constraints are satisfied:

[0030] g 1 (x)≥0.95

[0031] Parameter range:

[0032] 1≤α≤100

[0033] 1≤β≤100

[0034] 0.1≤γ≤5.0.

[0035] The system for predicting the intraday power supply guarantee capability of wind power generation according to the second aspect of the present invention is characterized by comprising:

[0036] A loss function construction module, capable of constructing a loss function, and based on the loss function, controlling the error between the assurance capability result and the actual observation value, and obtaining the observation deviation between the predicted power and the actual observation value;

[0037] A deviation correction module, capable of setting a supply guarantee coefficient, and using the supply guarantee coefficient to adjust the observed deviation to a corrected deviation;

[0038] A deviation truncation module is capable of performing a truncation operation on the elements in the corrected deviation, limiting the elements within a specified range to obtain a corrected deviation after the limit;

[0039] A penalty term calculation module, capable of determining a value of a penalty term based on the restricted corrected deviation and a preset penalty weight;

[0040] A loss calculation module, capable of applying different loss items according to the relationship between the predicted value and the actual observed value, and calculating the average loss according to the loss values ​​of the different loss items applied to different samples;

[0041] The supply guarantee capability estimation module can estimate the guarantee capability of the power system based on the average loss.

[0042] Furthermore, the loss function E in the loss function construction module MSE It can be expressed as:

[0043]

[0044] Where: y true_i is the actual observed value of the ith data point; y pred_i is the predicted value of the ith data point; n is the number of samples.

[0045] Furthermore, the loss function building module obtains the observed deviation d between the predicted power and the actual observed value, which can be expressed as:

[0046] d=y pred_i -y true_i

[0047] Furthermore, the supply guarantee coefficient β in the deviation correction module can correct part of the negative deviation to a positive deviation, and the expression of the corrected deviation Z is:

[0048] Z=d+β

[0049] Furthermore, in the deviation truncation module, a clamp function is used to perform a truncation operation on the elements in the corrected deviation, which is to truncate the elements in Z that are less than 0 to 0; that is, to set negative values ​​to 0, which can be specifically expressed as:

[0050]

[0051] Furthermore, in the penalty item calculation module, the penalty item calculation formula is:

[0052]

[0053] Among them, Q penalty is the penalty term, and α is the penalty weight.

[0054] Furthermore, the evaluation function of the system is:

[0055]

[0056] Among them, x represents the parameter vectors α, β and γ that need to be optimized; g(x) is the guarantee rate; M is a very large constant used to punish those that do not meet g 1 (x) the solution required;

[0057] The constraints are satisfied:

[0058] g 1 (x)≥0.95

[0059] Parameter range:

[0060] 1≤α≤100

[0061] 1≤β≤100

[0062] 0.1≤γ≤5.0.

[0063] According to the computer-readable storage medium of the third aspect of the embodiment of the present invention, the medium stores computer-executable instructions, and the computer-executable instructions are used to execute the above-mentioned method for predicting the power supply guarantee capability of wind power generation within the day.

[0064] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0066] Figure 1 A schematic diagram of the steps of a method for predicting power supply assurance capability of wind power generation within a day according to an embodiment of the present invention;

[0067] Figure 2 It is a structural block diagram of a system for predicting the power supply guarantee capability of wind power generation within a day according to an embodiment of the present invention. DETAILED DESCRIPTION

[0068] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0069] In the description of the present invention, it should be understood that descriptions involving orientations, such as up, down, front, back, left, right, etc., and orientations or positional relationships indicated are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0070] In the description of the present invention, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed" etc. are understood as not including the number itself, and "above", "below", "within" etc. are understood as including the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0071] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention based on the specific content of the technical solution.

[0072] Traditional loss functions, such as RMSE and MAE, only focus on reducing the error between the predicted power and the actual observed value during model training, so that the model is more inclined to output prediction results with smaller errors. For the prediction of guarantee capability, the objective function of the neural network model should not only focus on the error between the guarantee capability and the actual observed value, but also be more inclined to make the guarantee capability result output by the model not higher than the actual observed value. Therefore, the design of the custom loss function should meet these two requirements at the same time.

[0073] Embodiment 1

[0074] Based on the limitations of the above-mentioned traditional loss function, an embodiment of the present invention provides a method for predicting the power supply guarantee capability of wind power generation within a day, such as Figure 1 shown.

[0075] First, we still use E MSE The loss function is used to control the error between the guarantee capability results and the actual observed values, so that the model can maximize the output level of the prediction results and ensure that the guarantee capability results are of practical significance.

[0076]

[0077] Where: y true_i is the actual observed value of the ith data point; y pred_i is the predicted value of the ith data point; n is the number of samples.

[0078] Calculates the deviation between the predicted power and the actual observed value.

[0079] d=y pred_i -y true_i (2)

[0080] And a guarantee coefficient β is given. The introduction of the guarantee coefficient is to correct some negative deviations into positive deviations, increase the losses of both, so that the model can identify large losses and tend to output values ​​with smaller losses.

[0081] Z=d+β (3)

[0082] The clamp function is used to truncate the elements in Z and limit the elements to a specified range. Here, the elements in Z that are less than 0 are truncated to 0. That is, negative values ​​are set to 0, and only the part where the predicted value is higher than the actual value is extracted for the subsequent calculation of the penalty term.

[0083]

[0084] Q penalty The penalty term squares the non-negative differences, then averages them, and finally multiplies them by a penalty weight α.

[0085]

[0086] Different loss terms are applied depending on the relationship between the predicted value and the actual observed value. When the predicted power is greater than the actual observed value, E with γ weight is applied. MSE Loss function with a regularization penalty; when the predicted value is less than the actual observed value, only E with a γ weight is applied MSE The loss term is used to reduce the error between the two and bring the curves of the two closer together. Finally, the average loss is calculated based on the loss values ​​of different loss terms applied to different samples.

[0087]

[0088] In order to realize the prediction of wind power guarantee capacity, a multi-objective optimization strategy is introduced to achieve the two goals of model prediction: on the premise of ensuring that the proportion of predicted output lower than the actual observed value is greater than or equal to a certain value, the model can maximize the output level of the predicted results, ensuring that the guarantee capacity results have practical reference value for actual engineering problems.

[0089] In order to ensure the effectiveness and reliability of wind power capacity prediction results for the power system's "power supply guarantee", the optimization goal is to set the proportion of predicted output lower than the actual observed value to be greater than or equal to 95%.

[0090] Two optimization goals:

[0091] 1. Target 1: The ratio of actual power to guaranteed capacity result. We hope that this ratio is greater than or equal to 0.95.

[0092] 2. Objective 2: Predict the period when the guarantee curve is lower than the measured power, and the ratio of the guarantee capacity power to the actual power power. We hope that this ratio is as large as possible.

[0093] In summary, a comprehensive evaluation function is defined to combine these two objectives for optimization. The following is the mathematical model of the comprehensive evaluation index function of the multi-objective optimization strategy f(x):

[0094]

[0095] Where: x represents the parameter vectors α, β and γ to be optimized; g(x) is the guarantee rate; M is a very large constant used to penalize those that do not meet g. 1 (x) Required solution.

[0096] Constraints:

[0097] g 1 (x)≥0.95 (8)

[0098] Parameter range:

[0099] 1≤α≤100 (9)

[0100] 1≤β≤100 (10)

[0101] 0.1≤γ≤5.0 (11)

[0102] The embodiment of the second aspect of the present invention provides a system 20 for predicting the power supply guarantee capability of wind power generation within a day. Figure 2 As shown, including:

[0103] The loss function construction module 201 is capable of constructing a loss function, and determining the observation error between the guarantee capability result and the actual observation value based on the loss function, and obtaining the observation deviation between the predicted power and the actual observation value;

[0104] The deviation correction module 202 is capable of setting a supply guarantee coefficient and using the supply guarantee coefficient to adjust the observed deviation to a corrected deviation;

[0105] The deviation truncation module 203 can perform a truncation operation on the elements in the corrected deviation to limit the elements within a specified range to obtain a corrected deviation after the limit;

[0106] A penalty term calculation module 204, capable of determining a value of a penalty term based on the restricted corrected deviation and a preset penalty weight;

[0107] The loss calculation module 205 can apply different loss items according to the relationship between the predicted value and the actual observed value, and calculate the average loss according to the loss values ​​of the different loss items applied to different samples;

[0108] The guarantee capability estimation module 206 can estimate the guarantee capability of the power system based on the average loss.

[0109] Furthermore, there is a 95% probability that the actual power is greater than the guaranteed capacity, and the ratio of the guaranteed capacity power to the actual power power is maximized.

[0110] Furthermore, the loss function E MSE It can be expressed as:

[0111]

[0112] Where: y true_i is the actual observed value of the ith data point; y pred_i is the predicted value of the ith data point; n is the number of samples.

[0113] Furthermore, the observed deviation d between the predicted power and the actual observed value is obtained, and its formula can be expressed as:

[0114] d=y pred_i -y true_i

[0115] Furthermore, the supply guarantee coefficient β can correct some negative deviations into positive deviations. The expression of the corrected deviation Z is:

[0116] Z=d+β

[0117] Furthermore, in the step of performing a truncation operation on the elements in the corrected deviation and limiting the elements within a specified range to obtain the restricted corrected deviation, a clamp function is used to perform a truncation operation on the elements in the corrected deviation, that is, elements less than 0 in Z are truncated to 0; that is, negative values ​​are set to 0, which can be specifically expressed as:

[0118]

[0119] Furthermore, the step of determining the value of the penalty term based on the limited corrected deviation and the preset penalty weight is as follows:

[0120]

[0121] Among them, Q penalty is the penalty term, and α is the penalty weight.

[0122] Furthermore, the evaluation function of the system is:

[0123]

[0124] Among them, x represents the parameter vectors α, β and γ that need to be optimized; g(x) is the guarantee rate; M is a very large constant used to punish those that do not meet g 1 (x) the solution required;

[0125] The constraints are satisfied:

[0126] g 1 (x)≥0.95

[0127] Parameter range:

[0128] 1≤α≤100

[0129] 1≤β≤100

[0130] 0.1≤γ≤5.0.

[0131] Another aspect of the present invention provides a computer-readable storage medium storing computer-executable instructions for executing the above-mentioned Figure 1 The method for predicting the power supply guarantee capability of wind power generation within a day is shown.

[0132] The device embodiments described above are only illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0133] It will be appreciated by those skilled in the art that all or some of the steps and systems in the disclosed method above may be implemented as software, firmware, hardware and appropriate combinations thereof. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or a non-transitory medium) and a communication medium (or a temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that may be used to store desired information and may be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0134] The above is a specific description of the preferred implementation of the present application, but the present application is not limited to the above-mentioned implementation mode. Technical personnel familiar with the field can also make various equivalent deformations or substitutions without violating the spirit of the present application. These equivalent deformations or substitutions are all included in the scope defined by the claims of the present application.

Claims

1. A method for predicting the power supply guarantee capability of wind power generation within a day, characterized in that: The following steps are involved: Constructing a loss function, and determining the observation error between the assurance capability result and the actual observation value based on the loss function, and obtaining the observation deviation between the predicted power and the actual observation value; Setting a supply guarantee coefficient, and using the supply guarantee coefficient to adjust the observed deviation to a corrected deviation; Performing a truncation operation on the elements in the corrected deviation to restrict the elements to a specified range to obtain a restricted corrected deviation; Determining a value of a penalty term based on the restricted corrected deviation and a preset penalty weight; Apply different loss terms according to the relationship between the predicted value and the actual observed value, and calculate the average loss according to the loss values ​​of the different loss terms applied to different samples; The protection capability of the power system is estimated based on the average loss.

2. The method according to claim 1, characterized in that The loss function E MSE It can be expressed as: Where: y true_i is the actual observed value of the ith data point; y pred_i is the predicted value of the ith data point; n is the number of samples.

3. The method according to claim 2, characterized in that The formula for obtaining the observed deviation d between the predicted power and the actual observed value can be expressed as: d=y pred_i -y true_i 。 4. The method according to claim 3, characterized in that The supply guarantee coefficient β can correct part of the negative deviation to a positive deviation. The expression of the corrected deviation Z is: Z=d+β.

5. The method according to claim 4, characterized in that In the step of performing a truncation operation on the elements in the corrected deviation and limiting the elements within a specified range to obtain the corrected deviation after the limit, a clamp function is used to perform a truncation operation on the elements in the corrected deviation, that is, elements in Z that are less than 0 are truncated to 0; that is, negative values ​​are set to 0, which can be specifically expressed as:

6. The method according to claim 5, characterized in that The step of determining the value of the penalty term based on the limited corrected deviation and the preset penalty weight is formulated as follows: Among them, Q penalty is the penalty term, and α is the penalty weight.

7. The method according to claim 1, characterized in that The evaluation function of the method is: Where x represents the parameter vectors α, β and γ to be optimized; g(x) is the guarantee rate; M is a very large constant used to penalize solutions that do not meet the requirements of g1(x); The constraints are satisfied: g1(x)≥0.95 Parameter range: 1≤α≤100 1≤β≤100 0.1≤γ≤5.

0.

8. A system for predicting the power supply guarantee capability of wind power generation within a day, characterized in that: include: A loss function construction module, capable of constructing a loss function, and determining an observation error between a guarantee capability result and an actual observation value based on the loss function, and obtaining an observation deviation between the predicted power and the actual observation value; A deviation correction module, capable of setting a supply guarantee coefficient, and using the supply guarantee coefficient to adjust the observed deviation to a corrected deviation; A deviation truncation module is capable of performing a truncation operation on the elements in the corrected deviation, limiting the elements within a specified range to obtain a corrected deviation after the limit; A penalty term calculation module, capable of determining a value of a penalty term based on the restricted corrected deviation and a preset penalty weight; A loss calculation module, capable of applying different loss items according to the relationship between the predicted value and the actual observed value, and calculating the average loss according to the loss values ​​of the different loss items applied to different samples; The supply guarantee capability estimation module can estimate the guarantee capability of the power system based on the average loss.

9. The system according to claim 8, characterized in that The loss function E in the loss function building module MSE It can be expressed as: Where: y true_i is the actual observed value of the ith data point; y pred_i is the predicted value of the ith data point; n is the number of samples.

10. The system according to claim 9, characterized in that The loss function building module obtains the observed deviation d between the predicted power and the actual observed value, and its formula can be expressed as: d=y pred_i -y true_i 。 11. The system according to claim 10, characterized in that The supply guarantee coefficient β in the deviation correction module can correct part of the negative deviation to a positive deviation. The expression of the corrected deviation Z is: Z=d+β.

12. The system according to claim 11, characterized in that In the deviation truncation module, the clamp function is used to perform a truncation operation on the elements in the corrected deviation, which is to truncate the elements in Z that are less than 0 to 0; that is, to set negative values ​​to 0, which can be specifically expressed as:

13. The system according to claim 12, characterized in that In the penalty item calculation module, the penalty item calculation formula is: Among them, Q penalty is the penalty term, and α is the penalty weight.

14. The system according to claim 8, characterized in that The evaluation function of the system is: Where x represents the parameter vectors α, β and γ to be optimized; g(x) is the guarantee rate; M is a very large constant used to penalize solutions that do not meet the requirements of g1(x); The constraints are satisfied: g1(x)≥0.95 Parameter range: 1≤α≤100 1≤β≤100 0.1≤γ≤5.

0.

15. A computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are used to execute the method according to any one of claims 1 to 8.

Images