Offshore wind power plant multi-unit probability interval prediction aggregation method, system and equipment based on fast Fourier transform and medium

Abstract

The invention belongs to the technical field of offshore wind plant multi-unit power prediction, and discloses an offshore wind plant multi-unit probability interval prediction aggregation method, system and device based on fast Fourier transform and a medium so as to solve the technical defects of a traditional method. The method comprises the steps that a power prediction probability density function of each wind turbine generator set is acquired and discretization processing is carried out on the power prediction probability density function; converting each discrete probability density sequence from a time domain to a frequency domain by using fast Fourier transform; multiplication is carried out on the frequency domain representations of all the wind turbine generators in the frequency domain to obtain an aggregated frequency domain result; inverse fast Fourier transform is carried out on the aggregated frequency domain result, and a multi-unit aggregation probability density function of the time domain is obtained; calculating a cumulative distribution function based on the multi-unit aggregation probability density function, and performing normalization processing to obtain a normalized cumulative distribution function; and obtaining a preset confidence level, and determining a confidence interval of the overall power prediction of the wind power plant through the normalized cumulative distribution function.

Description

Technical Field

[0001] The present invention belongs to the technical field of multi - machine power prediction for offshore wind farms, and specifically relates to a method, system, device, and medium for aggregating multi - unit probability interval prediction of offshore wind farms based on the fast Fourier transform. Background Art

[0002] Offshore wind farms usually consist of multiple wind turbines operating in coordination. Due to the differences in the geographical locations of the turbines, sea conditions, and meteorological environments, the power outputs of each turbine exhibit complex probability distribution characteristics. This multi - machine coordinated operation mode requires accurate prediction of the overall power output of the wind farm to ensure the stable dispatching of the power system. However, traditional single - machine prediction methods cannot effectively integrate the probability distribution information of multiple turbines, resulting in insufficient accuracy of the overall power prediction of the wind farm.

[0003] Currently, the existing methods for aggregating multi - unit power prediction in offshore wind farms mainly include interval prediction based on the dependence relationship of historical data, the quantile regression model of the kernel extreme learning machine, and the interval prediction method combining the LUBE theory and the GRU neural network. Although these methods can handle the uncertainty of power prediction to a certain extent, they still have obvious defects: firstly, they only focus on single - moment prediction or simple data aggregation and do not fully consider the differences in the probability distributions of multiple turbines; secondly, the computational complexity is high, making it difficult to meet the real - time requirements; thirdly, they lack adaptability to complex sea conditions and the correlation between turbines, resulting in insufficient prediction interval coverage and accuracy. Summary of the Invention

[0004] Based on the above - mentioned disadvantages and deficiencies in the prior art, one of the objectives of the present invention is to at least solve one or more of the above - mentioned problems in the prior art. In other words, one of the objectives of the present invention is to provide a method, system, device, and medium for aggregating multi - unit probability interval prediction of offshore wind farms based on the fast Fourier transform that meet one or more of the foregoing requirements, so as to achieve the objectives of innovating the multi - machine probability distribution aggregation method, enhancing the universality of offshore wind power scenarios, improving the accuracy and practicality of power prediction intervals, increasing the computational efficiency and response speed, and improving the multi - machine coordinated analysis function.

[0005] To achieve the above - mentioned invention objectives, the present invention adopts the following technical solutions: In the first aspect, the present invention provides a method for aggregating multi - unit probability interval prediction of offshore wind farms based on the fast Fourier transform, including the steps of: S1. Obtain the power prediction probability density function of each wind turbine in the offshore wind farm, and perform discretization processing respectively to obtain the corresponding discrete probability density sequence; S2. Use the fast Fourier transform to convert each discrete probability density sequence from the time domain to the frequency domain to obtain the corresponding frequency - domain representation; S3. Perform a multiplication operation on the frequency-domain representations of all wind turbines in the frequency domain to obtain an aggregated frequency-domain result; S4. Perform an inverse fast Fourier transform on the aggregated frequency-domain result to obtain a multi-turbine aggregated probability density function in the time domain; S5. Calculate a cumulative distribution function based on the multi-turbine aggregated probability density function and perform a normalization process to obtain a normalized cumulative distribution function; S6. Obtain a preset confidence level and determine a confidence interval for the overall power prediction of the wind farm through the normalized cumulative distribution function.

[0006] As a preferred solution, the discretization process described in step S1 is specifically: Uniformly select sampling points within the value range of the power value, and the probability density values corresponding to each sampling point form a discrete probability density sequence, where N is a positive integer.

[0007] As a preferred solution, the formula for the fast Fourier transform described in step S2 is: , In the formula, is the frequency-domain representation, is the discrete probability density sequence, is the length of the discrete probability density sequence, , , .

[0008] As a preferred solution, the multiplication operation described in step S3 is specifically: Based on the convolution theorem, perform point-by-point multiplication on the frequency-domain representations of each wind turbine. B

[0009] As a preferred solution, step S5 includes: Perform a discrete integration on the multi-turbine aggregated probability density function to obtain a cumulative distribution function; Perform a normalization process on the cumulative distribution function to make its value range [0, 1]; The normalization process includes dividing the cumulative distribution function by its maximum value.

[0010] As a preferred solution, step S6 includes: Obtain a preset confidence level ; According to the confidence level calculate the two-sided quantiles and ; Obtain the two-sided quantiles and from the cumulative distribution function through linear interpolationThe corresponding power value and ; Use the power value and as the lower and upper limits of the confidence interval to obtain the confidence interval .

[0011] As a preferred solution, the value of the preset confidence level is 90%, 95% or 99%.

[0012] In a second aspect, the present invention provides a multi-unit probability interval prediction aggregation system for an offshore wind farm based on fast Fourier transform, which is used to implement the multi-unit probability interval prediction aggregation method for an offshore wind farm as described in the first aspect, including: A data acquisition module, a discretization processing module, an FFT conversion module, a frequency-domain aggregation module, and a confidence interval calculation module that are connected in sequence. The FFT conversion module includes a conversion unit and an inverse conversion unit; The data acquisition module is used to acquire the power prediction probability density function of each wind turbine in the offshore wind farm; The discretization processing module is used to perform discretization processing on the power prediction probability density function to obtain a discrete probability density sequence; The conversion unit is used to convert the discrete probability density sequence from the time domain to the frequency domain; The frequency-domain aggregation module is used to perform a multiplication operation in the frequency domain to obtain an aggregation result; The inverse conversion unit is used to convert the aggregation result back to the time domain; The confidence interval calculation module is used to calculate the cumulative distribution function and determine the confidence interval.

[0013] In a third aspect, the present invention provides an electronic device, which includes a memory, a processor, and a computer program. When the computer program is executed by the processor, it implements the multi-unit probability interval prediction aggregation method for an offshore wind farm as described in the first aspect.

[0014] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by the processor, it implements the multi-unit probability interval prediction aggregation method for an offshore wind farm as described in the first aspect.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. In the data processing method, the present invention creatively applies the fast Fourier transform to convert the power prediction probability density function of each unit from the time domain to the frequency domain, realizes efficient aggregation through multiplication operations in the frequency domain, and then transforms back to the time domain through inverse transformation. This method breaks through the limitations of traditional simple aggregation methods such as direct addition or weighted average, and provides a new solution to the problem of multi-unit probability distribution aggregation.

[0016] 2. Considering the special environment of offshore wind farms, the present invention fully takes into account the power output uncertainty and probability distribution complexity caused by various factors such as sea conditions and terrain affecting the units in different regions of offshore wind farms. This can effectively aggregate probability distribution curves of different forms (dispersed or concentrated), greatly enhancing the universality of this method in various offshore wind farm scenarios, and effectively overcoming the problem of insufficient adaptability of existing technologies.

[0017] 3. In terms of the practicality of prediction results, the present invention calculates the cumulative distribution function of the aggregated probability density function and accurately determines the power probability interval according to the set confidence level. This provides a credible value range for wind farm power prediction, meets the requirements of power system dispatching and operation decision-making for accurate prediction intervals, and further improves the practicality of prediction results.

[0018] 4. In terms of computing efficiency, the present invention uses the FFT to reduce the computational complexity from O(N 2 ) to O(NlogN), significantly improving the efficiency of multi-unit probability distribution aggregation and subsequent calculations. This enables the system to quickly respond to changes in real-time data of wind farms and update power interval predictions in a timely manner, significantly surpassing traditional methods in terms of computing speed.

[0019] 5. In terms of multi-unit coordinated operation analysis, the present invention focuses on the multi-unit coordinated operation scenario of offshore wind farms, effectively integrates the power output characteristics of each unit that are interrelated and independent, realizes the accurate aggregation of multi-unit probability distributions, provides strong support for the analysis and prediction of the overall power characteristics of wind farms, and makes up for the shortcomings of existing technologies in multi-unit coordinated analysis.

[0020] Further or more detailed beneficial effects will be described in combination with specific embodiments in the detailed implementation manners. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings without creative efforts based on these drawings.

[0022] Figure 1It is a schematic flowchart of the multi - unit probability interval prediction aggregation method for an offshore wind farm according to an embodiment of the present invention.

[0023] Figure 2 It is a schematic structural diagram of the multi - unit probability interval prediction aggregation system for an offshore wind farm according to an embodiment of the present invention.

[0024] Figure 3 It is a structural diagram of the electronic device provided by an embodiment of the present invention.

[0025] Figure 4 It is a result diagram of the comparative test with a 90% confidence interval according to Embodiment 5 of the present invention.

[0026] Figure 5 It is a result diagram of the comparative test with a 60% confidence interval according to Embodiment 5 of the present invention.

[0027] Reference numerals in the attached drawings: 300, electronic device; 301, processor; 302, communication bus; 303, user interface; 304, network interface; 305, memory. Detailed implementation manners

[0028] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention.

[0029] In the following description, multiple embodiments of the present invention are provided. Different embodiments can be replaced or combined. Therefore, the present invention can also be considered to include all possible combinations of the same and / or different embodiments described. Thus, if one embodiment includes features A, B, and C, and another embodiment includes features B and D, then the present invention should also be considered to include embodiments containing one or more of all other possible combinations of A, B, C, and D, even though such an embodiment may not be explicitly described in the following content.

[0030] The following description provides examples and does not limit the scope, applicability, or examples set forth in the claims. Changes can be made to the functions and arrangements of the described elements without departing from the scope of the content of the present invention. Each example can appropriately omit, substitute, or add various processes or components. For example, the described method can be executed in a different order from the described order, and various steps can be added, omitted, or combined. In addition, the features described in some examples can be combined into other examples.

[0031] To facilitate a better understanding of the embodiments of the present invention, before explaining the detailed implementation manners of the present invention in detail, its application scenarios will be described first.

[0032] The offshore wind farm multi-unit probability interval prediction aggregation method described in the embodiments of this specification is applied to the power prediction and operation management process of the offshore wind farm. In these scenarios, the application of the offshore wind farm multi-unit probability interval prediction aggregation method aims to innovate the multi-machine power distribution aggregation means to break through traditional limitations, enhance the adaptability to the output characteristics of units in different regions under complex sea conditions and terrain, improve the accuracy and credibility of power prediction intervals to support power system decision-making, optimize computing efficiency to quickly respond to real-time data and update prediction results, and improve the multi-machine collaborative analysis function to fully support the overall operation characteristics analysis and prediction of the wind farm.

[0033] The following is a brief explanation of the coordinated operation of multiple units, sea conditions, sea-land wind conversion, fast Fourier transform, discrete Fourier transform, probability density function, convolution theorem, frequency domain aggregation, time domain signal, cumulative distribution function, confidence interval, interpolation method, prediction interval coverage, standardized average width, and coverage width criterion involved in multiple embodiments of this specification: Multi-unit coordinated operation refers to the interconnected, collaborative power generation of multiple wind turbines within a wind farm, requiring consideration of both the independence and correlation of power output between the turbines. This paper effectively integrates the probability distribution of multi-unit power forecasts through frequency-domain aggregation, improving overall forecast accuracy in coordinated operation scenarios and addressing the shortcomings of traditional methods in multi-unit coordinated analysis.

[0034] Sea state factors refer to marine environmental parameters that affect the operation of offshore wind farms, including wave height, current velocity, and sea ice cover. This method adapts to the uncertainty of turbine power output under different sea conditions through a probability distribution aggregation mechanism, enhancing the method's applicability in complex marine environments.

[0035] Sea-land wind shift is a phenomenon in coastal areas where wind direction alternates between day and night due to thermal differences between land and sea. This phenomenon significantly impacts the power fluctuations of offshore wind turbines. This paper uses frequency-domain aggregation of probability density functions to effectively integrate the power distribution characteristics of turbines affected by sea-land wind, improving the accuracy of prediction results.

[0036] The Fast Fourier Transform (FFT) is an algorithm for efficiently calculating the Discrete Fourier Transform (DFT). It achieves rapid conversion between the time and frequency domains by decomposing a time-domain signal into a superposition of different frequency components. In this paper, the FFT is used to convert the probability density function of each wind turbine power forecast from the time domain to the frequency domain. Frequency-domain multiplication allows for efficient aggregation of the probability distributions of multiple turbines, significantly reducing computational complexity.

[0037] The Discrete Fourier Transform (DFT) is a mathematical method for converting a discrete-time sequence into a frequency-domain representation and is a core tool in digital signal processing. In the present invention, the DFT, through its efficient implementation form FFT, converts the power prediction data from the time domain to the frequency domain for aggregation, providing a frequency-domain analysis basis for subsequent processing.

[0038] The Probability Density Function (PDF) is a function that describes the likelihood of a random variable near a certain value, and its integral represents the probability of the variable within a certain interval. In the present invention, the probability density function uses the power prediction value as the abscissa and the probability density value as the ordinate to characterize the uncertainty distribution of the power output of a single wind turbine, and is the basic data form for realizing the aggregation of the multi-machine probability distribution.

[0039] The Convolution Theorem states that the convolution operation in the time domain corresponds to the multiplication operation in the frequency domain, and it is the theoretical basis for using the Fast Fourier Transform for convolution calculation. In the present invention, the aggregation of the multi-machine probability density function is achieved through frequency-domain multiplication, which is equivalent to time-domain convolution but with higher computational efficiency, and is the key theoretical basis for realizing the efficient integration of the multi-machine probability distribution.

[0040] Frequency-domain aggregation refers to the fusion processing of the power prediction data of multiple units in the frequency domain, and the superposition of the probability distribution is achieved through multiplication operations. In the present invention, compared with direct time-domain convolution, frequency-domain aggregation significantly reduces the computational complexity and retains the physical meaning, ensuring that the aggregation result is more consistent with the actual power output characteristics.

[0041] A time-domain signal is a signal representation form with time as the independent variable, such as the measured power output sequence. In the present invention, the time-domain signal is converted into a frequency-domain representation through FFT, which is convenient for subsequent aggregation operations and is the basic data form of the entire method process.

[0042] The Cumulative Distribution Function (CDF) is a function that describes the probability that a random variable is less than or equal to a certain value and is the integral result of the probability density function. In the present invention, the cumulative distribution function is calculated based on the aggregated probability density function and is used to determine the power interval at a specific confidence level, providing a credible value range for wind farm power prediction.

[0043] A confidence interval is an interval estimate that contains an unknown population parameter at a certain confidence level, reflecting the reliability of statistical inference. In the present invention, the credible range of the overall power prediction of an offshore wind farm is determined through the cumulative distribution function, providing a decision-making basis for power dispatching. Its calculation is based on the aggregated probability distribution data, ensuring the accuracy and reliability of the results.

[0044] Interpolation is a method of estimating missing values through known data points. In the present invention, linear interpolation and spline interpolation are used to fill in the missing values of power data, ensuring data integrity and providing an accurate data basis for subsequent probability distribution generation.

[0045] The prediction interval coverage probability (PICP) refers to the proportion of true values falling within the prediction interval, reflecting the reliability of the prediction interval. For example, if the confidence level is set at 90%, a PICP of 0.92 means that the actual value has a 92% probability of being covered by the interval, slightly higher than the expected confidence level.

[0046] The normalized average width (PINAW) is used to measure the proportion of the average width of the prediction interval relative to the overall fluctuation range of the data. The calculation formula is the average value of the interval width divided by the difference between the maximum and minimum values of the data. The smaller this index, the more compact and accurate the prediction interval. For example, a PINAW of 0.15 means that the average interval width accounts for 15% of the total data span.

[0047] The coverage width criterion (CWC) is a comprehensive evaluation index that combines the normalized average width with a coverage rate deviation penalty term. The formula is CWC = PINAW + γe-η(picp - μ). Where γ and η are adjustment parameters, and μ is the target coverage rate. This index optimizes the interval width while ensuring the coverage rate. For example, when the PICP is lower than the target value, the exponential term will increase significantly, prompting the model to adjust parameters to improve the coverage rate.

[0048] Example 1: As Figure 1 shown, this embodiment provides a multi-unit probability interval prediction aggregation method for an offshore wind farm based on fast Fourier transform, including the steps: S1. Obtain the power prediction probability density function of each wind turbine in the offshore wind farm. Let the power prediction probability density function of each wind turbine be ( , (where the total number of wind turbines is), before performing the FFT operation, the continuous probability density function needs to be discretized and converted into a discrete sequence. S2. Use the fast Fourier transform to transform each discrete probability density sequence from the time domain to the frequency domain to obtain the corresponding frequency domain representation. S3. Perform a multiplication operation on the frequency domain representations of all wind turbines in the frequency domain to obtain the aggregated frequency domain result. S4. Perform an inverse fast Fourier transform (IFFT) on the aggregated frequency domain result to obtain the multi-unit aggregated probability density function in the time domain. S5. Calculate the cumulative distribution function based on the multi-unit aggregated probability density function and perform normalization to obtain the normalized cumulative distribution function. S6. Obtain the preset confidence level and determine the confidence interval for the overall power prediction of the wind farm through the normalized cumulative distribution function.

[0049] In this embodiment, the fast Fourier transform (FFT) is used to transform the probability density function of the power prediction of each wind turbine from the time domain to the frequency domain. In the frequency domain, the aggregation of the frequency domain representations of each unit is achieved through simple multiplication operations, and then it is transformed back to the time domain through the inverse fast Fourier transform (IFFT) to obtain the aggregated probability density function. This process can comprehensively and meticulously integrate the probability distribution information of each unit. For example, for wind turbines at different locations, the power of wind turbines near the coast is affected by sea-land breezes, and the probability distribution may be relatively wide and fluctuating; while the power of turbines in deep water areas is affected by factors such as sea conditions, and the probability distribution is relatively concentrated. Through the FFT and frequency domain aggregation operations in this embodiment, these different forms of probability distributions can be effectively integrated, avoiding the problem of only focusing on the prediction at a single moment and ignoring the aggregation of the probability distributions of multiple units. Due to the accurate aggregation of the probability distributions of multiple units, this embodiment can more realistically reflect the comprehensive characteristics of the power output of the entire wind farm. In an offshore wind farm, the power outputs of each unit are interrelated and independent. Only by fully considering this characteristic can a foundation be laid for accurately predicting the overall power range. Compared with the prior art "Wind Power Interval Prediction Method Based on Quantile Regression of Kernel Extreme Learning Machine", this embodiment no longer stays at the stage of data collection and simple processing, but delves into the core level of probability distribution for integration, so that the aggregated probability distribution can more comprehensively reflect the differences between each unit, thereby providing a more reliable data basis for subsequent power interval prediction. Based on the above effective aggregation and accurate reflection of the overall power characteristics of the wind farm, this embodiment can significantly improve the accuracy of the overall power interval prediction of the wind farm. Accurate power interval prediction is crucial for power system scheduling. In power dispatching, accurate power interval prediction can help the dispatching department more reasonably arrange the power generation plan and avoid power surplus or insufficient supply caused by inaccurate prediction. For example, when the predicted power interval is more accurate, the dispatching department can more precisely coordinate the cooperation between the wind farm and other power generation resources to ensure the balance of power supply and demand and maintain the stable operation of the power grid, which is difficult to achieve with the prior art, thus making up for the defect of limited prediction accuracy caused by insufficient aggregation of the probability distributions of multiple units in "A Wind Power Interval Prediction Method, System and Storage Medium".

[0050] Specifically, this embodiment provides a preferred implementation manner. The discretization process in step S1 is specifically as follows: Uniformly select sampling points within the value range of the power value. The probability density values corresponding to each sampling point form a discrete probability density sequence, where is a positive integer.

[0051] As a preferred solution, the formula for the fast Fourier transform in step S2 is: , In the formula, is the frequency-domain representation, is the discrete probability density sequence, is the length of the discrete probability density sequence, , , .

[0052] It can be understood that the Fast Fourier Transform (FFT), as an efficient algorithm for the Discrete Fourier Transform (DFT), can quickly transform a time-domain signal into the frequency domain. For a discrete sequence x[n] of length N, its DFT is defined as: , where , . The computational complexity of the Discrete Fourier Transform (DFT) is O(N 2 ), and the computational efficiency is low when the data volume is large. The Fast Fourier Transform (FFT) adopted in this embodiment cleverly utilizes the periodicity and symmetry of the rotation factor to decompose and reorganize the calculation process. For example, when is even, the calculation of the DFT can be decomposed into two DFT calculations of length , greatly reducing the amount of calculation and reducing the computational complexity to , significantly improving the computational efficiency.

[0053] Specifically, this embodiment provides a preferred implementation manner, and the multiplication operation in step S3 is specifically: Based on the convolution theorem, perform point-by-point multiplication on the frequency-domain representations of each wind turbine. Let the aggregated frequency-domain result be , then . Through the product operation in the frequency domain in this embodiment, the aggregation effect of the probability density functions of each unit is achieved, and compared with directly performing convolution calculation in the time domain, the computational complexity is greatly reduced.

[0054] Specifically, this embodiment provides a preferred implementation manner, and step S5 includes: performing discrete integration on the multi-unit aggregated probability density function to obtain the cumulative distribution function; performing normalization processing on the cumulative distribution function so that its value range is [0, 1]; the normalization processing includes dividing the cumulative distribution function by its maximum value. More specifically, the cumulative distribution function is obtained by cumulatively summing the probability density values, that is , where is the interval of the power value during discretization. For example, if the value range of the power value during discretization is , and a total of N sampling points are selected, then , the cumulative summation process starts from m=0, and gradually accumulates the product of the probability density value of each discrete point and the interval to obtain the cumulative probability of the corresponding power value. In order to ensure the correctness of the cumulative distribution function, it is normalized to satisfy By Divide by The sum over the entire domain gives the normalized cumulative distribution function .set up The normalized cumulative distribution function is ,The normalization operation ensures that the cumulative distribution function can accurately reflect the ,probability distribution of the power value in the entire value range, and its ,value range is always between [0,1].

[0055] Specifically, this embodiment provides a preferred implementation method, step S6 includes: obtaining a preset confidence level ; According to the confidence level Calculate two-sided quantiles and ; Obtain two-sided quantiles from the cumulative distribution function by linear interpolation and Corresponding power value and The power value and As the lower and upper bounds of the confidence interval, we get the confidence interval More specifically, The linear interpolation formula for the target probability value is , the corresponding power value and The lower and upper limits of the confidence interval are By determining the confidence interval, a certain credible value range can be provided for wind farm power forecasting, providing an important reference for power system scheduling and operation decision-making.

[0056] Specifically, this embodiment provides a preferred implementation in which the preset confidence level is set to 90%, 95%, or 99%. It will be understood that the confidence level represents the probability that the determined confidence interval will contain the true value across multiple repeated experiments. For example, a 90% confidence level means that across a large number of repeated experiments, approximately 90% of the confidence intervals will contain the true wind farm power value.

[0057] Example 2: like Figure 2 As shown, this embodiment provides a system for probabilistic interval prediction aggregation of multiple units in an offshore wind farm based on fast Fourier transform, which is used to implement the method for probabilistic interval prediction aggregation of multiple units in an offshore wind farm as described in Example 1, including: A data acquisition module, a discretization processing module, an FFT conversion module, a frequency domain aggregation module, and a confidence interval calculation module connected in sequence. The FFT conversion module includes a conversion unit and an inverse conversion unit; The data acquisition module is configured to acquire the power prediction probability density function of each wind turbine in an offshore wind farm; The discretization processing module is configured to perform discretization processing on the power prediction probability density function to obtain a discrete probability density sequence; The conversion unit is configured to convert the discrete probability density sequence from the time domain to the frequency domain; The frequency domain aggregation module is configured to perform a multiplication operation in the frequency domain to obtain an aggregation result; The inverse conversion unit is configured to convert the aggregation result back to the time domain; The confidence interval calculation module is configured to calculate the cumulative distribution function and determine the confidence interval.

[0058] Embodiment III: As Figure 3 shown, this embodiment provides an electronic device, which may include: at least one processor, at least one network interface, a user interface, a memory, and at least one communication bus.

[0059] Among them, the communication bus can be used to realize the connection and communication of the above-mentioned various components.

[0060] Among them, the user interface may include buttons, and the optional user interface may further include a standard wired interface and a wireless interface.

[0061] Among them, the network interface may but is not limited to including a Bluetooth module, an NFC module, a Wi-Fi module, etc.

[0062] Among them, the processor may include one or more processing cores. The processor uses various interfaces and lines to connect various parts within the entire electronic device, and by running or executing instructions, programs, code sets, or instruction sets stored in the memory, and by calling data stored in the memory, it executes various functions of the electronic device and processes data. Optionally, the processor may be implemented in at least one of the hardware forms of DSP, FPGA, and PLA. The processor may integrate one or several combinations of CPU, GPU, and modem, etc. Among them, the CPU mainly processes the operating system, user interface, and application programs, etc.; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communication. It can be understood that the above-mentioned modem may not be integrated into the processor and may be implemented separately by a single chip.

[0063] Among them, the memory may include RAM and may also include ROM. Optionally, the memory includes a non-transitory computer-readable medium. The memory can be used to store instructions, programs, codes, code sets or instruction sets. The memory may include a program storage area and a data storage area. Among them, the program storage area can store instructions for implementing the operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-mentioned method embodiments, etc.; the data storage area can store the data involved in the above-mentioned method embodiments. Optionally, the memory can also be at least one storage device located far from the aforementioned processor. The memory as a computer storage medium may include an operating system, a network communication module, a user interface module, and a prediction aggregation application program. The processor can be used to call the prediction aggregation application program stored in the memory and execute the steps of the multi-unit probability interval prediction aggregation method for an offshore wind farm mentioned in the foregoing embodiments.

[0064] Embodiment 4: This embodiment provides a computer-readable storage medium, in which instructions are stored. When they run on a computer or a processor, the computer or the processor is caused to execute one or more of the steps in the Figure 1 embodiments shown above. If the respective component modules of the above electronic device are implemented in the form of software functional units and sold or used as independent products, they can be stored in the computer-readable storage medium.

[0065] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of this specification are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired manner (such as coaxial cable, optical fiber, Digital Subscriber Line (DSL)) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more integrated available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., Digital Versatile Disc (DVD)), or a semiconductor medium (e.g., Solid State Disk (SSD)), etc.

[0066] Those of ordinary skill in the art can understand that all or part of the processes in implementing the method in the above Embodiment 1 can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above various methods. The aforementioned storage medium includes: various media such as ROM, RAM, magnetic disk, or optical disc that can store program codes. Without conflict, the technical features in this embodiment and the implementation solutions can be combined arbitrarily.

[0067] Embodiment Five: To verify the effectiveness of the method for aggregating probability interval predictions of multiple wind turbines in an offshore wind farm based on fast Fourier transform described in this specification, in this embodiment, a comparative experiment is set up based on the actual application scenario of the method for aggregating probability interval predictions of multiple wind turbines in the offshore wind farm. Among them, the control group uses a traditional prediction method, while the verification group applies the prediction aggregation method described in the present invention. Through this setting in this embodiment, the prediction results of the control group and the verification group are obtained and compared and analyzed with the true values. The prediction results of the 90% confidence interval of the wind farm power are as Figure 4As shown in the figure, in this embodiment, three indicators, namely the prediction interval coverage probability (PICP), the normalized average width (PINAW), and the coverage width criterion (CWC), are used to evaluate the interval prediction results. The evaluation results are shown in Table 1. Since the PICP of the traditional power interval prediction does not reach 90%, the CWC of the traditional power interval prediction is much higher than that of the aggregated power prediction. It can be seen that the aggregated power interval prediction method is more reliable than the traditional power interval prediction. Compared with the traditional power interval prediction, the PINAW of the aggregated power interval prediction is reduced by 2.0%, indicating that the prediction interval has higher accuracy.

[0068] Table 1:

[0069] The prediction results of the 60% confidence interval of the wind farm power are as Figure 5 shown. In this embodiment, three indicators, namely the prediction interval coverage probability (PICP), the normalized average width (PINAW), and the coverage width criterion (CWC), are used to evaluate the interval prediction results. The evaluation results are shown in Table 2. The PICPs of both methods meet the standards. Compared with the traditional power interval prediction, the PINAW of the aggregated power interval prediction is reduced by 38.1%, indicating that the prediction interval has higher accuracy. Generally speaking, the CWC of the aggregated power interval prediction is smaller and the prediction effect is better.

[0070] Table 2:

[0071] Based on the above, this embodiment verifies the effectiveness of a method for aggregating probability interval predictions of multiple wind turbines in an offshore wind farm based on the fast Fourier transform described in this specification.

[0072] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0073] In the above embodiments, the descriptions of the respective embodiments have their own emphases. For parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0074] The above are only exemplary embodiments of the present invention and should not be used to limit the scope of the present invention. That is, any equivalent changes and modifications made in accordance with the teachings of the present invention still fall within the scope covered by the present invention. Those skilled in the art will readily conceive of other embodiments of the present invention after considering the specification and practicing the disclosure herein. The present invention is intended to cover any variations, uses, or adaptations of the present invention, which follow the general principles of the present invention and include common general knowledge or conventional technical means in the technical field not recorded in the present invention. The specification and examples are only regarded as exemplary, and the scope and spirit of the present invention are defined by the claims.

Claims

1. A method for aggregating probabilistic interval prediction of multiple wind turbines in an offshore wind farm based on fast Fourier transform, characterized in that, Including the steps: S1. Obtain the power prediction probability density function of each wind turbine in the offshore wind farm, and perform discretization processing respectively to obtain the corresponding discrete probability density sequence; S2. Use the fast Fourier transform to convert each discrete probability density sequence from the time domain to the frequency domain to obtain the corresponding frequency domain representation; S3. Perform multiplication operations on the frequency domain representations of all wind turbines in the frequency domain to obtain the aggregated frequency domain result; S4. Perform the inverse fast Fourier transform on the aggregated frequency domain result to obtain the multi-turbine aggregated probability density function in the time domain; S5. Calculate the cumulative distribution function based on the multi-turbine aggregated probability density function, and perform normalization processing to obtain the normalized cumulative distribution function; S6. Obtain the preset confidence level, and determine the confidence interval of the overall power prediction of the wind farm through the normalized cumulative distribution function.

2. A method for aggregating probability interval prediction of multiple wind turbines in an offshore wind farm based on fast Fourier transform according to claim 1, characterized in that The discretization processing described in step S1 is specifically: Uniformly select within the range of power values sampling points, and the probability density values corresponding to each sampling point form a discrete probability density sequence, where is a positive integer.

3. A multi-unit probability interval prediction aggregation method for an offshore wind farm based on fast Fourier transform according to claim 2, characterized in that, The formula for the fast Fourier transform described in step S2 is: ， In the formula, is the frequency-domain representation, is the discrete probability density sequence, is the length of the discrete probability density sequence, , , .

4. A method for aggregating probabilistic interval predictions of multiple wind turbines in an offshore wind farm based on fast Fourier transform according to claim 3, characterized in that The multiplication operation described in step S3 is specifically: Based on the convolution theorem, perform point-by-point multiplication on the frequency domain representations of each wind turbine.

5. A method for aggregating probability interval predictions of multiple units in an offshore wind farm based on fast Fourier transform according to claim 4, characterized in that, Step S5 includes: Perform discrete integration on the multi-turbine aggregated probability density function to obtain the cumulative distribution function; Perform normalization processing on the cumulative distribution function so that its value range is [0, 1]; The normalization processing includes dividing the cumulative distribution function by its maximum value.

6. A method for aggregating probabilistic interval prediction of multiple wind turbines in an offshore wind farm based on fast Fourier transform according to claim 5, characterized in that, Step S6 includes: Obtain a preset confidence level ; According to the confidence level Calculate the two-sided quantile and ; Obtaining two-sided quantiles from the cumulative distribution function through linear interpolation and the corresponding power values and ; Take the power value and as the lower and upper limits of the confidence interval, and obtain the confidence interval .

7. A multi-turbine probability interval prediction aggregation method for an offshore wind farm based on the fast Fourier transform according to claim 6, characterized in that: The value of the preset confidence level is 90%, 95% or 99%.

8. A multi-unit probability interval prediction aggregation system for an offshore wind farm based on fast Fourier transform, characterized in that, Used to implement the multi-turbine probability interval prediction aggregation method for an offshore wind farm as described in any one of claims 1 to 7, including: A data acquisition module, a discretization processing module, an FFT conversion module, a frequency domain aggregation module, and a confidence interval calculation module connected in sequence. The FFT conversion module includes a conversion unit and an inverse conversion unit: The data acquisition module is used to obtain the power prediction probability density function of each wind turbine in the offshore wind farm; The discretization processing module is used to perform discretization processing on the power prediction probability density function to obtain a discrete probability density sequence; The conversion unit is used to convert the discrete probability density sequence from the time domain to the frequency domain; The frequency domain aggregation module is used to perform multiplication operations in the frequency domain to obtain an aggregation result; The inverse conversion unit is used to convert the aggregation result back to the time domain; The confidence interval calculation module is used to calculate the cumulative distribution function and determine the confidence interval.

9. A computer device, the computer device comprising a memory, a processor, and a computer program, characterized in that, The computer program, when executed by a processor, implements the multi-turbine probability interval prediction aggregation method for an offshore wind farm as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program, when executed by a processor, implements the multi-turbine probability interval prediction aggregation method for an offshore wind farm as described in any one of claims 1 to 7.

Images