Optimal configuration method for wind power energy storage hybrid system
By constructing a physical-data hybrid prediction model and performing rolling correction, the prediction error problem of wind power energy storage hybrid system under complex operating conditions is solved, and the prediction accuracy and system reliability are improved.
Patent Information
- Application Number
- CN202510735300.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-04
- Publication Date
- 2025-07-18
AI Technical Summary
The data prediction model of the wind power energy storage hybrid system under complex operating conditions lacks physical constraints, resulting in large prediction errors and difficulty in improving prediction accuracy and system reliability.
Build a physical prediction model, combine meteorological historical data, establish a physical-data hybrid prediction model, correct the prediction error through a rolling correction mechanism, and optimize the configuration of the wind power energy storage hybrid system.
It reduces prediction error, improves prediction accuracy and system reliability, and realizes high-precision optimized configuration of wind power energy storage hybrid system.
Smart Images
Figure CN120341855A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of wind power generation, and particularly to an optimization configuration method for a wind power energy storage hybrid system. Background Art
[0002] With the continuous growth of global energy demand and the increasing emphasis on environmental protection, the development and utilization of renewable energy have become the key to sustainable development. As a clean and renewable energy source, wind energy has great potential and has become an important part of sustainable development globally. However, the instability and volatility of wind energy pose challenges to the reliability and stability of wind power systems, making the research and application of wind power energy storage hybrid systems particularly important.
[0003] In a wind power energy storage hybrid system, improving the prediction accuracy of wind power output and optimizing the energy storage control strategy are the core challenges for enhancing the system's reliability and economy. In current related technologies, wind power prediction mainly relies on statistical models driven by historical data (such as ARIMA, LSTM, etc.). Although these models can capture time series characteristics, they ignore the essential influence of the physical characteristics of wind turbines (such as power curves and aerodynamic responses) on the output characteristics. Especially under complex working conditions (such as turbulence and yaw error), data-driven models are prone to "black box" distortion due to the lack of physical constraints, resulting in the standard deviation of prediction errors exceeding 15%.
[0004] Therefore, there is an urgent need for an optimization configuration method for a wind power energy storage hybrid system to more effectively address the prediction error problem. Summary of the Invention
[0005] To solve the above problems, the present invention provides an optimization configuration method for a wind power energy storage hybrid system, which solves the problem that data prediction models are prone to distortion due to the lack of physical constraints under complex working conditions, reduces prediction errors, and improves prediction accuracy.
[0006] The present invention provides an optimization configuration method for a wind power energy storage hybrid system, including the following steps:
[0007] S1. Obtain historical meteorological data, historical operation data of the wind turbine, and the wind turbine characteristic curves; the historical meteorological data includes: wind speed, wind direction, temperature, and air pressure; the historical operation data of the wind turbine includes: power, pitch angle, and rotational speed; the wind turbine characteristic curves include: wind pressure - air volume curve, power - air volume curve, efficiency - air volume curve, and power - wind speed curve;
[0008] S2. Based on the historical operation data of the wind turbine and the wind turbine characteristic curves, combine with air density to construct a physical prediction model;
[0009] S3. According to the physical prediction model and the historical meteorological data, construct a physical - data hybrid prediction model;
[0010] S4. Calculate the predicted power according to the meteorological historical data, the historical operation data of the wind turbine, and the physical-data hybrid prediction model.
[0011] S5. Calculate the prediction error according to the predicted power and the historical operation data of the wind turbine, and identify the error distribution type of the prediction error to obtain the identification result; the error distribution types include: Gaussian distribution and non-Gaussian distribution.
[0012] S6. Based on the physical prediction model and the identification result, according to the meteorological historical data, through a rolling correction mechanism, correct the prediction error to realize the optimal configuration of the wind power energy storage hybrid system.
[0013] Preferably, the expression of the physical prediction model in S2 is:
[0014]
[0015] where P is the predicted power, ρ is the air density, π is the pi; R is the radius of the wind turbine blade, C p is the power coefficient, and v is the wind speed.
[0016] Preferably, in S3, according to the physical prediction model and the meteorological historical data, a physical-data hybrid prediction model is constructed, specifically including:
[0017] Input the meteorological historical data into the physical prediction model to obtain the physical predicted power.
[0018] Calculate the historical error according to the physical predicted power and the historical power.
[0019] Construct a training set and a test set based on the meteorological historical data, the physical predicted power, and the historical error.
[0020] Adopt a meta-learning algorithm to train and test the physical prediction model based on the training set and the test set. Until the trained physical prediction model meets the set requirements, use the trained physical prediction model as the physical-data hybrid prediction model.
[0021] Preferably, in S5, according to the predicted power and the historical operation data of the wind turbine, calculate the prediction error, and identify the error distribution type of the prediction error to obtain the identification result. The specific content includes:
[0022] Calculate the prediction error according to the predicted power and the historical power.
[0023] Judge whether the error distribution type of the prediction error is a Gaussian distribution according to the distribution of the prediction error.
[0024] Preferably, the expression of the prediction error is:
[0025] E = P actual - P pred;
[0026] Among them, E is the prediction error, P actual is the historical power, P pred is the predicted power.
[0027] Preferably, according to the distribution of the prediction error, determine whether the distribution type of the prediction error is a Gaussian distribution. The specific content includes:
[0028] If the distribution of the prediction error is a normal distribution, then the distribution type of the prediction error is a Gaussian distribution;
[0029] If the distribution of the prediction error shows skewness or multimodal distribution, then the distribution type of the prediction error is a non-Gaussian distribution.
[0030] Preferably, in S6, based on the physical prediction model and the recognition result, according to the meteorological historical data, through a rolling correction mechanism, correct the prediction error to achieve the optimal configuration of the wind power energy storage hybrid system. The specific content includes:
[0031] If the error distribution type of the prediction error is a Gaussian distribution, then calculate the prediction deviation based on the prediction error;
[0032] Through the rolling correction mechanism, based on the meteorological historical data, the physical predicted power and the historical error, correct the prediction deviation to achieve the optimal configuration of the wind power energy storage hybrid system.
[0033] The present invention provides an optimal configuration system for a wind power energy storage hybrid system, including:
[0034] A data acquisition module for acquiring meteorological historical data, historical operation data of the fan, and fan characteristic curves; the meteorological historical data includes: wind speed, wind direction, temperature, and air pressure; the historical operation data of the fan includes: power, pitch angle, and rotational speed; the fan characteristic curves include: wind pressure - air volume curve, power - air volume curve, efficiency - air volume curve, and power - wind speed curve;
[0035] A first model construction module for constructing a physical prediction model based on the historical operation data of the fan and the fan characteristic curves in combination with air density;
[0036] A second model construction module for constructing a physical - data hybrid prediction model according to the physical prediction model and the meteorological historical data;
[0037] A predicted power calculation module for calculating the predicted power according to the meteorological historical data, the historical operation data of the fan, and the physical - data hybrid prediction model;
[0038] An error analysis module is used to calculate a prediction error based on predicted power and historical wind turbine operation data, and identify the type of error distribution of the prediction error to obtain an identification result; the types of error distribution include: Gaussian distribution and non-Gaussian distribution.
[0039] A correction module is used to correct the prediction error based on a physical prediction model and the identification result, according to historical meteorological data, through a rolling correction mechanism, to achieve an optimal configuration of a wind power energy storage hybrid system.
[0040] The present invention provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and when the processor calls the computer program in the memory, the content of the method for optimizing the configuration of the wind power energy storage hybrid system as described above is implemented.
[0041] The present invention provides a storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are loaded and executed by a processor, the content of the method for optimizing the configuration of the wind power energy storage hybrid system as described above is implemented.
[0042] In summary, for the method for optimizing the configuration of a wind power energy storage hybrid system of the present invention, by constructing a physical prediction model, combining historical meteorological data, constructing a physical-data hybrid prediction model, obtaining predicted power, and performing prediction error correction, the problem that the data prediction model is prone to distortion due to lack of physical constraints under complex working conditions is solved, the prediction error is reduced, and the prediction accuracy is improved.
[0043] The technical method of the present invention will be further described in detail below with reference to the drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] Figure 1 is a schematic flowchart of a method for optimizing the configuration of a wind power energy storage hybrid system of the present invention;
[0045] Figure 2 is a schematic diagram of the modules of a system for optimizing the configuration of a wind power energy storage hybrid system of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0046] The technical method of the present invention will be further described below with reference to the drawings and embodiments. It should be noted that: unless otherwise specifically stated, the relative arrangements, numerical expressions and values of the components and steps described in these embodiments do not limit the scope of the present application.
[0047] The following description of at least one exemplary embodiment is merely illustrative in nature and in no way serves as a limitation of the present application, its application or use.
[0048] Technologies, systems, and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, the technologies, systems, and devices should be regarded as part of the specification.
[0049] In all examples shown and discussed herein, any specific values should be construed as merely exemplary and not as a limitation. Thus, other examples of the exemplary embodiments may have different values.
[0050] Unless otherwise defined, technical terms or scientific terms used in the present invention shall have the ordinary meaning as understood by those of ordinary skill in the art to which the present invention pertains.
[0051] The present invention provides an optimization configuration method for a wind power energy storage hybrid system, as Figure 1 shown, which specifically includes:
[0052] Step S1: Obtain historical meteorological data, historical operation data of the wind turbine, and the wind turbine characteristic curve. The historical meteorological data includes: wind speed, wind direction, temperature, and air pressure. The historical operation data of the wind turbine includes: power, pitch angle, and rotational speed. The wind turbine characteristic curve includes: wind pressure - air volume curve, power - air volume curve, efficiency - air volume curve, and power - wind speed curve.
[0053] The wind turbine characteristic curve among them is a curve representing the relationship between the main performance parameters of the ventilator (such as air volume L, wind pressure H, power N, and efficiency η), but the wind turbine characteristic curve is broadly defined as a set of curves covering various performance parameters of the wind turbine, and also includes power - wind speed curve, power coefficient - tip speed ratio curve, torque - rotational speed curve, and thrust coefficient - wind speed curve.
[0054] Step S2: Based on the historical operation data of the wind turbine and the wind turbine characteristic curve, construct a physical prediction model in combination with air density.
[0055] Among them, the expression of the physical prediction model in step S2 is:
[0056]
[0057] Among them, P is the predicted power, ρ is the air density, π is the pi; R is the radius of the wind turbine blade, C p is the power coefficient, and v is the wind speed.
[0058] The physical prediction model constructed in step S2 provides a theoretical support for prediction by introducing the operation mechanism of the wind turbine, the law of energy conversion, and the dynamic characteristics of the system. And in rare weather (such as typhoon, low temperature) or when the wind turbine fails, the physical prediction model can provide conservative prediction values to avoid the data - driven model from outputting unreasonable results due to the lack of training samples.
[0059] Step S3: Construct a physical-data hybrid prediction model based on the physical prediction model and meteorological historical data.
[0060] In the physical-data hybrid prediction model constructed in step S3, the physical model provides a theoretical boundary for the prediction to prevent the data-driven model from outputting unreasonable results due to noise or outliers. The data model can reduce the model complexity and the risk of overfitting. Therefore, by combining the advantages of physical laws and data-driven technologies, the physical-data hybrid prediction model significantly improves the prediction accuracy, control reliability, and strategy interpretability of the system, realizes the closed-loop of high-precision prediction - low-risk control - interpretable decision-making, and is the key technical path to improve the reliability and economy of the wind power energy storage system.
[0061] The specific content of step S3 includes:
[0062] Input the meteorological historical data into the physical prediction model to obtain the physical predicted power.
[0063] Calculate the historical error based on the physical predicted power and the historical power.
[0064] Construct a training set and a test set based on the meteorological historical data, the physical predicted power, and the historical error.
[0065] Adopt a meta-learning algorithm to train and test the physical prediction model based on the training set and the test set. When the trained physical prediction model meets the set requirements, use the trained physical prediction model as the physical-data hybrid prediction model. The meta-learning algorithm in this invention focuses on the parameters and structure of the learning algorithm itself, aiming to improve the performance of the machine learning algorithm by learning how to learn, so as to adapt to different tasks or environments and have good generalization ability. In this invention, the meta-learning algorithm improves the generalization performance of the physical prediction model based on the knowledge learned from multiple tasks through the training set and the test set, making the physical prediction model perform better on new tasks, having higher utilization rate of data and stronger adaptability.
[0066] It can be understood that when constructing the physical-data hybrid prediction model, inputting the historical error can enable the data model to capture the dynamic characteristics, systematic biases, or environmental noises not covered by the physical model, thereby realizing the dynamic correction and adaptive optimization of the prediction results, and significantly improving the prediction accuracy and control reliability of the wind power energy storage.
[0067] Step S4: Calculate the predicted power based on the meteorological historical data, the historical operation data of the wind turbine, and the physical-data hybrid prediction model.
[0068] Step S5: Calculate the prediction error based on the predicted power and the historical operation data of the wind turbine, and identify the error distribution type of the prediction error to obtain the identification result. Among them, the error distribution types include: Gaussian distribution and non-Gaussian distribution.
[0069] The specific content of step S5 includes:
[0070] Calculate the prediction error based on the predicted power and the historical power. Among them, the expression of the prediction error is:
[0071] E = P actual - P pred ;
[0072] Among them, E is the prediction error, P actual is the historical power, and P pred is the predicted power.
[0073] Judge whether the distribution type of the prediction error is a Gaussian distribution according to the distribution of the prediction error. The specific judgment process includes:
[0074] If the distribution of the prediction error is a normal distribution, the distribution type of the prediction error is a Gaussian distribution.
[0075] If the distribution of the prediction error shows skewness or multimodal distribution, the distribution type of the prediction error is a non-Gaussian distribution.
[0076] It can be understood that by judging the distribution type of the prediction error, the validity of the prediction error can be inferred. If the distribution type of the prediction error is a Gaussian distribution, the optimal estimate of the prediction error can be provided, improving the accuracy of the prediction.
[0077] Step S6: Based on the physical prediction model and the recognition result, according to the meteorological historical data, through a rolling correction mechanism, correct the prediction error to achieve the optimal configuration of the wind power energy storage hybrid system.
[0078] The specific content of step S6 includes:
[0079] If the error distribution type of the prediction error is a Gaussian distribution, calculate the prediction deviation based on the prediction error.
[0080] Through the rolling correction mechanism, based on the meteorological historical data, the physical predicted power and the historical error, correct the prediction deviation to achieve the optimal configuration of the wind power energy storage hybrid system.
[0081] It can be understood that the rolling correction mechanism corrects the prediction error by means of dynamic adjustment and continuous iteration, can correct deviations in a timely manner, avoid error accumulation, correct precisely, avoid blind adjustment, and at the same time accumulate experience in the long term to optimize the long-term performance.
[0082] The present invention provides a system for optimizing the configuration of a wind power energy storage hybrid system, as Figure 2 shown, including:
[0083] A data acquisition module, configured to acquire historical meteorological data, historical operation data of a wind turbine, and a wind turbine characteristic curve. Among them, the historical meteorological data includes: wind speed, wind direction, temperature, and air pressure; the historical operation data of the wind turbine includes: power, pitch angle, and rotational speed; the wind turbine characteristic curve includes: a wind pressure - air volume curve, a power - air volume curve, an efficiency - air volume curve, and a power - wind speed curve;
[0084] A first model construction module, configured to construct a physical prediction model based on the historical operation data of the wind turbine and the wind turbine characteristic curve, in combination with air density.
[0085] A second model construction module, configured to construct a physical - data hybrid prediction model according to the physical prediction model and the historical meteorological data.
[0086] A predicted power calculation module, configured to calculate a predicted power according to the historical meteorological data, the historical operation data of the wind turbine, and the physical - data hybrid prediction model.
[0087] An error analysis module, configured to calculate a prediction error according to the predicted power and the historical operation data of the wind turbine, and identify the error distribution type of the prediction error to obtain an identification result. Among them, the error distribution types include: Gaussian distribution and non - Gaussian distribution.
[0088] A correction module, configured to correct the prediction error based on the physical prediction model and the identification result, according to the historical meteorological data, through a rolling correction mechanism, to achieve an optimal configuration of the wind - power energy storage hybrid system.
[0089] The present invention provides an electronic device, including a memory and a processor. A computer program is stored in the memory. When the processor calls the computer program in the memory, the content of the above - mentioned method for optimizing the configuration of the wind - power energy storage hybrid system is implemented.
[0090] The present invention provides a storage medium, in which computer - executable instructions are stored. When the computer - executable instructions are loaded and executed by a processor, the content of the above - mentioned method for optimizing the configuration of the wind - power energy storage hybrid system is implemented.
[0091] Finally, it should be noted that: the above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that: they can still modify the technical method of the present invention or make equivalent replacements, and these modifications or equivalent replacements cannot make the modified technical method deviate from the spirit and scope of the technical method of the present invention.
Claims
1. An optimization configuration method for a wind power energy storage hybrid system, characterized in that, It includes the following steps: S1. Obtain meteorological historical data, wind turbine historical operation data, and wind turbine characteristic curves; The meteorological historical data includes: wind speed, wind direction, temperature, and air pressure; the wind turbine historical operation data includes: power, pitch angle, and rotational speed; the wind turbine characteristic curves include: wind pressure - air volume curve, power - air volume curve, efficiency - air volume curve, and power - wind speed curve; S2. Based on the wind turbine historical operation data and wind turbine characteristic curves, combined with air density, construct a physical prediction model; S3. According to the physical prediction model and meteorological historical data, construct a physical - data hybrid prediction model; S4. According to the meteorological historical data, wind turbine historical operation data, and physical - data hybrid prediction model, calculate the predicted power; S5. According to the predicted power and wind turbine historical operation data, calculate the prediction error, and identify the error distribution type of the prediction error to obtain the identification result; the error distribution types include: Gaussian distribution and non - Gaussian distribution; S6. Based on the physical prediction model and the identification result, according to the meteorological historical data, through a rolling correction mechanism, correct the prediction error to achieve the optimal configuration of the wind - power energy storage hybrid system.
2. The optimization configuration method of a wind power energy storage hybrid system according to claim 1, characterized in that, The expression of the physical prediction model in S2 is: Among them, P is the predicted power, ρ is the air density, π is the pi; R is the radius of the fan blade, C p is the power coefficient, and v is the wind speed.
3. The optimization configuration method of a wind power energy storage hybrid system according to claim 1, characterized in that In S3, according to the physical prediction model and meteorological historical data, constructing a physical - data hybrid prediction model specifically includes: Input the meteorological historical data into the physical prediction model to obtain the physical predicted power; Calculate the historical error according to the physical predicted power and historical power; Based on the meteorological historical data, physical predicted power, and historical error, construct a training set and a test set; Adopt a meta - learning algorithm to train and test the physical prediction model based on the training set and the test set. Until the trained physical prediction model meets the set requirements, use the trained physical prediction model as the physical - data hybrid prediction model.
4. A method for optimizing the configuration of a wind power energy storage hybrid system according to claim 1, characterized in that In S5, according to the predicted power and wind turbine historical operation data, calculate the prediction error, and identify the error distribution type of the prediction error to obtain the identification result. The specific content includes: Calculate the prediction error according to the predicted power and historical power; According to the distribution of the prediction error, judge whether the distribution type of the prediction error is a Gaussian distribution.
5. The optimization configuration method of a wind power energy storage hybrid system according to claim 4, characterized in that The expression of the prediction error is: E = P actual -P pred ; Among them, E is the prediction error, P actual is the historical power, P pred is the predicted power.
6. The optimization configuration method of a wind power energy storage hybrid system according to claim 4, characterized in that According to the distribution of the prediction error, judge whether the distribution type of the prediction error is a Gaussian distribution. The specific content includes: If the distribution of the prediction error is a normal distribution, then the distribution type of the prediction error is a Gaussian distribution; If the distribution of the prediction error shows a skewed or multimodal distribution, then the distribution type of the prediction error is a non - Gaussian distribution.
7. The optimization configuration method of a wind power energy storage hybrid system according to claim 3, characterized in that In S6, based on the physical prediction model and the identification result, according to the meteorological historical data, through a rolling correction mechanism, correct the prediction error to achieve the optimal configuration of the wind - power energy storage hybrid system. The specific content includes: If the error distribution type of the prediction error is a Gaussian distribution, then calculate the prediction deviation based on the prediction error; Through the rolling correction mechanism, based on the meteorological historical data, physical predicted power, and historical error, correct the prediction deviation to achieve the optimal configuration of the wind - power energy storage hybrid system.
8. An optimized configuration system for a wind power energy storage hybrid system, characterized in that, It includes: A data acquisition module for obtaining meteorological historical data, wind turbine historical operation data, and wind turbine characteristic curves; The meteorological historical data includes: wind speed, wind direction, temperature, and air pressure; the fan historical operation data includes: power, pitch angle, and rotational speed; the fan characteristic curves include: wind pressure - air volume curve, power - air volume curve, efficiency - air volume curve, and power - wind speed curve; The first model construction module is used to construct a physical prediction model based on the fan historical operation data and the fan characteristic curves, in combination with air density; The second model construction module is used to construct a physical - data hybrid prediction model according to the physical prediction model and the meteorological historical data; The predicted power calculation module is used to calculate the predicted power according to the meteorological historical data, the fan historical operation data, and the physical - data hybrid prediction model; The error analysis module is used to calculate the prediction error according to the predicted power and the fan historical operation data, and identify the error distribution type of the prediction error to obtain an identification result; the error distribution types include: Gaussian distribution and non - Gaussian distribution; The correction module is used to correct the prediction error based on the physical prediction model and the identification result, according to the meteorological historical data, through a rolling correction mechanism, to achieve the optimal configuration of the wind - power energy storage hybrid system.
9. An electronic device, characterized in that, It includes a memory and a processor. A computer program is stored in the memory. When the processor calls the computer program in the memory, the content of the method for optimizing the configuration of the wind - power energy storage hybrid system according to any one of claims 1 to 7 is implemented.
10. A storage medium, characterized in that, Computer - executable instructions are stored in the storage medium. When the computer - executable instructions are loaded and executed by the processor, the content of the method for optimizing the configuration of the wind - power energy storage hybrid system according to any one of claims 1 to 7 is implemented.