Method for evaluating uncertainty of online monitoring data of new energy station under complex working conditions

Abstract

The invention discloses a new energy station on-line monitoring data uncertainty evaluation method under a complex working condition, and relates to the technical field of on-line monitoring data, and the method comprises the steps: collecting data information, and generating a sound wave spectrum matrix based on vibration sound wave data; training a first machine learning model for correcting vibration influence wind speed data in real time, a second machine learning model for correcting water vapor influence wind speed data and a third machine learning model for correcting fatigue influence wind speed data, and comprehensively calculating B-class uncertainty and comprehensive real wind speed data; a-class uncertainty is calculated based on the wind speed data; and calculating the comprehensive uncertainty based on the A-class uncertainty and the B-class uncertainty. The method can solve the problem that a traditional method cannot improve the accuracy of data measurement under complex working conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of online monitoring data, and in particular to a method for evaluating uncertainty of online monitoring data of new energy sites under complex working conditions. Background Art

[0002] A new energy station is a collection of all the equipment at the wind farm or solar power station grid connection point that is centrally connected to the power system, including transformers, busbars, lines, converters, energy storage, wind turbines, photovoltaic power generation equipment, reactive power regulation equipment and auxiliary equipment.

[0003] The Chinese patent with authorization announcement number CN113269435B discloses a coupled monitoring and evaluation system for the operating status of a new energy station, constructs a simulation model of a new energy station, determines monitoring indicators for energy consumption of different media, proposes multiple unit consumption indicators, and conducts a comprehensive evaluation of energy usage.

[0004] However, the environment in which new energy stations are located is special, and the monitored data often produces errors that are different from the actual data due to complex working conditions. Therefore, the uncertainty of online monitoring data is particularly important. The above invention fails to provide an uncertainty evaluation method for the data and fails to improve the accuracy of data measurement under complex working conditions.

[0005] In view of this, the present invention proposes an uncertainty evaluation method for online monitoring data of new energy sites under complex working conditions to solve the above problems. Summary of the invention

[0006] In view of the above existing problems, the present invention is proposed.

[0007] Therefore, the present invention provides a method for evaluating uncertainty of online monitoring data of new energy sites under complex working conditions, which can solve the problem that traditional methods cannot improve the accuracy of data measurement under complex working conditions.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for evaluating uncertainty of online monitoring data of new energy stations under complex working conditions, comprising:

[0009] Collect new energy site data, and generate a sound wave spectrum matrix based on vibration sound wave data; train a first machine learning model for real-time correction of vibration-affected wind speed data based on the sound wave spectrum matrix, vibration-affected wind speed data and vibration real wind speed data; train a second machine learning model for real-time correction of water vapor-affected wind speed data based on air humidity data, water vapor-affected wind speed data and water vapor real wind speed data; train a third machine learning model for real-time correction of fatigue-affected wind speed data based on fatigue time, fatigue-affected wind speed data and fatigue real wind speed data; calculate Class A uncertainty based on wind speed data, calculate Class B uncertainty and comprehensive real wind speed data based on the first machine learning model, the second machine learning model and the third machine learning model; calculate comprehensive uncertainty based on Class A uncertainty and Class B uncertainty.

[0010] As a preferred scheme for the uncertainty evaluation method of online monitoring data of new energy stations under complex working conditions described in the present invention, the collected new energy station data includes wind speed data, vibration sound wave data, vibration affected wind speed data, vibration real wind speed data, water vapor affected wind speed data, water vapor real wind speed data, fatigue affected wind speed data, fatigue real wind speed data, air humidity data and fatigue time.

[0011] As a preferred solution of the uncertainty evaluation method of online monitoring data of new energy stations under complex working conditions described in the present invention, wherein: the Class A uncertainty refers to the degree of uncertainty about wind speed under statistical analysis conditions;

[0012] The Class B uncertainty refers to the degree to which non-statistical analysis methods cannot be certain about wind speed;

[0013] There are F wind speed sensors installed on the top of the same wind turbine. A sensor weight QZ is set for each wind speed sensor, and the sensor weight QZ is normalized so that the sensor weight QZ adds up to 1. At the same time, the sensor weight is randomly reset every unit time.

[0014] Wind speed data Where fs i represents the wind speed displayed by the i-th sensor; F is the number of wind speed sensors; QZ i represents the sensor weight of the i-th sensor.

[0015] As a preferred solution of the uncertainty evaluation method of online monitoring data of new energy sites under complex working conditions described in the present invention, wherein: the acoustic wave spectrum matrix includes extracting the spectrum by performing short-time Fourier transform on the vibration acoustic wave data;

[0016] The spectrum is represented as a sound wave spectrum matrix; each row and each column in the sound wave spectrum matrix represents a different frequency, and each element represents the amplitude value of the corresponding frequency.

[0017] As a preferred solution of the uncertainty evaluation method of online monitoring data of new energy sites under complex working conditions described in the present invention, wherein: the first machine learning model includes taking the vibration real wind speed data as a set of sound wave spectrum matrices and labels of vibration-affected wind speed data;

[0018] A set of sound wave spectrum matrices, vibration-affected wind speed data, and vibration-actual wind speed data are used as inputs of a first machine learning model, and the first machine learning model uses the corrected vibration-affected wind speed data as output; a real-time set of sound wave spectrum matrices and vibration-affected wind speed data corresponding to the vibration-actual wind speed data are used as prediction targets, and minimizing the loss function value of the first machine learning model is used as a training target; and training is stopped when the loss function value of the first machine learning model is less than or equal to a preset first target loss value;

[0019] The first machine learning model loss function is a mean square error; the mean square error is calculated by The training model is trained with minimization as the goal. In the loss function, mse is the loss function value, i is the group number of the acoustic spectrum matrix and the wind speed data affected by vibration; u is the group number of the acoustic spectrum matrix and the wind speed data affected by vibration; y i is the label corresponding to the i-th group of sound wave spectrum matrix and vibration-affected wind speed data, is the label for real-time correction of the i-th group of acoustic spectrum matrix and vibration-affected wind speed data;

[0020] In actual operation of the wind speed sensor, the wind speed data FS acquired in real time is used as input vibration-affected wind speed data.

[0021] As a preferred solution of the uncertainty evaluation method of online monitoring data of new energy sites under complex working conditions described in the present invention, wherein: the second machine learning model includes taking the real wind speed data of water vapor as a set of labels of air humidity data and wind speed data affected by water vapor;

[0022] A set of air humidity data, water vapor affected wind speed data and water vapor real wind speed data are used as inputs of the second machine learning model, and the second machine learning model uses the corrected water vapor affected wind speed data as output; a real-time set of air humidity data and water vapor affected wind speed data corresponding to the water vapor real wind speed data are used as prediction targets, and minimizing the loss function value of the second machine learning model is used as the training target; training is stopped when the loss function value of the second machine learning model is less than or equal to the preset second target loss value.

[0023] As a preferred solution of the uncertainty evaluation method of online monitoring data of new energy stations under complex working conditions described in the present invention, wherein: the third machine learning model includes taking fatigue real wind speed data as a set of labels of fatigue time and fatigue-affected wind speed data;

[0024] A set of fatigue time, fatigue-affected wind speed data and fatigue-actual wind speed data are used as inputs of a third machine learning model, and the third machine learning model uses the corrected fatigue-affected wind speed data as output; a real-time set of fatigue time and fatigue-affected wind speed data corresponding to the fatigue-actual wind speed data are used as prediction targets, and minimizing the loss function value of the third machine learning model is used as a training target; training is stopped when the loss function value of the third machine learning model is less than or equal to a preset third target loss value;

[0025] The first machine learning model, the second machine learning model and the third machine learning model are recurrent neural network models.

[0026] As a preferred solution of the uncertainty evaluation method of online monitoring data of new energy stations under complex working conditions described in the present invention, wherein: the calculation of comprehensive uncertainty includes:

[0027] Type A uncertainty Where n represents the number of consecutive measurements, FS i represents the wind speed data measured for the i-th time;

[0028] The first machine learning model is constructed with the sound wave spectrum matrix X1 and the vibration-affected wind speed data X2 as independent variables, and the dependent variable is the function G(X1, X2) of the vibration real wind speed data;

[0029] Type B uncertainty of vibration and acoustic wave data In the formula It means to find the function G(X1, X2) with respect to X i The partial derivative of ; k represents the number of independent variables of function G; The independent variable is X i The uncertainty of

[0030] Type B uncertainty U of air humidity data B2 and the Type B uncertainty U about fatigue time B3 The calculation method is the same;

[0031] Will U A , U B1 , U B2 with U B3 The square mean of is taken as the comprehensive uncertainty;

[0032] The average value of the vibration real wind speed data, water vapor real wind speed data and fatigue real wind speed data output in real time by the first machine learning model, the second machine learning model and the third machine learning model is taken as the comprehensive real wind speed data.

[0033] A computer device includes a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the processor implements the steps of a method for evaluating uncertainty of online monitoring data of new energy sites under complex working conditions.

[0034] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of a method for evaluating uncertainty of online monitoring data of new energy sites under complex working conditions are implemented.

[0035] The beneficial effects of the present invention are: the error is reduced as much as possible through random algorithms, and the authenticity of the data is comprehensively considered by multiple machine learning models to correct the data. Based on artificial intelligence empowerment, a comprehensive uncertainty evaluation method is creatively proposed, which makes up for the absence of new energy sites in data uncertainty evaluation, obtains more accurate sensor data, and promotes the effective use of new energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0037] Figure 1 A schematic flow chart of a method for evaluating uncertainty in online monitoring data of new energy stations under complex working conditions provided by an embodiment of the present invention.

[0038] Figure 2 A schematic diagram of the electronic device structure of a method for evaluating uncertainty in online monitoring data of new energy stations under complex working conditions provided by an embodiment of the present invention.

[0039] Figure 3 A schematic diagram of the structure of a computer-readable storage medium for a method for evaluating uncertainty in online monitoring data of a new energy station under complex working conditions provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0040] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0041] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0042] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0043] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0044] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0045] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0046] Example 1, reference Figure 1 , which is the first embodiment of the present invention, provides a method for evaluating uncertainty of online monitoring data of new energy sites under complex working conditions, including:

[0047] Collect wind speed data;

[0048] Wind speed data is acquired in real time by a wind speed sensor installed on top of the wind turbine; the wind speed sensor includes but is not limited to an ultrasonic wind speed sensor and a thermal sensor;

[0049] The ultrasonic wind speed sensor measures wind speed by using the ultrasonic time difference method. When sound waves propagate in the air, the speed of sound propagation in the air will be superimposed on the airflow speed in the wind direction. Therefore, if the propagation direction of the ultrasonic wave is the same as the wind direction, its speed will increase; conversely, if the propagation direction of the ultrasonic wave is opposite to the wind direction, its speed will decrease. Therefore, the speed of ultrasonic wave propagation in the air can correspond to the wind speed function to obtain the wind speed.

[0050] The thermal wind speed sensor uses a hot wire (tungsten wire or platinum wire) or a hot film (a thin film made of platinum or chromium) as a probe. The probe is exposed to the air being measured and connected to a Wheatstone bridge. The flow rate of the air in the measured cross-section is detected through the balance relationship of the resistance or current of the Wheatstone bridge, thereby obtaining the wind speed.

[0051] The importance of collecting wind speed in new energy fields is very high. Wind energy is a clean and renewable energy source. Wind speed is one of the important indicators for evaluating wind energy resources. Collecting wind speed data can help evaluate the potential and availability of wind energy resources. By monitoring and recording wind speed, we can understand the wind energy characteristics in different locations and time periods. Wind speed data can also be used to predict future wind energy production and provide basic data for the operation and maintenance of wind farms.

[0052] There are F wind speed sensors installed on the top of the same wind turbine. A random sensor weight QZ is set for each wind speed sensor. The sensor weight QZ is normalized so that the sum of the sensor weight QZ is 1. At the same time, the sensor weight is updated once per unit time and randomly reset.

[0053] Exemplary code for obtaining sensor weights using Python includes:

[0054] import random

[0055] #F represents the number of sensors, assuming there are 5 sensors at this time

[0056] F=5

[0057] #Generate F random numbers

[0058] weights=[random.uniform(0,1)for_in range(F)]

[0059] #Calculate the sum of random numbers

[0060] total_weight = sum(weights)

[0061] # Divide each random number by the sum to get the normalized weight for each sensor

[0062] normalized_weights=[w / total_weight forw inweights]

[0063] # Output the weight of each sensor

[0064] print("Sensor weights:",normalized_weights)

[0065] An example output is:

[0066] Sensor Weight:

[0067] [0.14013943785704522,0.3571129770321358,0.19900639010916443,0.22451617795898507,0.07922501704266936]

[0068] Wind speed data Where fs i represents the wind speed displayed by the i-th sensor; F is the number of wind speed sensors; QZ i represents the sensor weight of the i-th sensor;

[0069] Assigning random sensor weights to sensors to obtain wind speed data will make the wind speed more realistic. Wind farms are generally built at high altitudes. Under the special climatic environment at high altitudes, wind turbines face harsh natural environmental conditions such as low pressure, high radiation, and frequent thunderstorms. Sensor materials are prone to aging and measurement errors. Improving the randomness of the sensor is conducive to avoiding the obstruction of the overall measurement caused by the transmission failure of a certain sensor. At the same time, updating the sensor weight per unit time further improves the degree of randomness, making the measured wind speed data more accurate and effective.

[0070] Collect vibroacoustic data;

[0071] The vibration acoustic wave data is collected in real time by a high-quality microphone in a laboratory environment, where the wind speed sensor is installed on a vibration table to simulate the vibration that occurs when the wind speed sensor is actually working on a wind turbine.

[0072] A vibration table is a device used to simulate and test vibrations. It consists of a platform and one or more exciters (motors or hydraulic cylinders) that can generate vibrations of various frequencies and amplitudes, so that the test object on the platform produces vibrations similar to those in the real environment.

[0073] Collect wind speed data affected by vibration, real wind speed data affected by vibration, wind speed data affected by water vapor, real wind speed data affected by water vapor, wind speed data affected by fatigue, and real wind speed data affected by fatigue;

[0074] The vibration-affected wind speed data refers to the real-time wind speed displayed by the wind speed sensor in a laboratory environment, where the wind speed sensor is installed on a vibration table and in a vibration environment.

[0075] The real wind speed data of vibration is the real-time wind speed displayed by the stable wind speed sensor in the laboratory environment when the same wind source is not installed on the vibration table;

[0076] Collecting air humidity data, wherein the air humidity data is acquired in real time by a humidity sensor;

[0077] The water vapor affects the wind speed data in the laboratory environment. The wind speed sensor displays the real-time wind speed in the laboratory environment with the air humidity adjusted experimentally.

[0078] The real wind speed data of water vapor is the real-time wind speed displayed by the wind speed sensor with the same wind source in a dry environment in a laboratory environment as a comparison with the wind speed data affected by water vapor;

[0079] The fatigue-affected wind speed data are wind speeds collected in real time using wind speed sensors of different durations in a laboratory environment;

[0080] Fatigue real wind speed data is the real-time wind speed collected by a new wind speed sensor in a laboratory environment against fatigue-affected wind speed data;

[0081] Collect fatigue time, which is the number of days between the wind speed sensor production date and the current date;

[0082] Obtained by mathematical operation of the production date of the production package and the current date;

[0083] The vibration sound wave data, air humidity data and fatigue time are also synchronously acquired when the actual wind speed sensor is working.

[0084] Fatigue is the process in which an object accumulates damage and deformation under continuous stress; wind speed sensor components may be damaged and deformed after long-term use. The longer the fatigue time, the lower the accuracy of the wind speed sensor may be.

[0085] By studying the effects of vibration, water vapor humidity, and fatigue time on wind speed sensors in a laboratory environment: In practical applications, wind speed sensors may be affected by structural vibrations, such as wind vibration effects on buildings or vibrations on wind turbine towers, which may interfere with the measurement results of wind speed sensors. By evaluating the impact of vibration on wind speed sensors, it is helpful to improve the accuracy of wind speed measurements. Humidity is a measure of the moisture content in the air. Excessive humidity can cause corrosion of sensor components and lead to measurement errors. By evaluating the impact of humidity on wind speed sensors, corrections and corrections can be made to improve the accuracy of wind speed measurements. By evaluating the impact of fatigue time on wind speed sensors, the aging defects of sensor parts can be partially offset and the accuracy of wind speed data can be improved.

[0086] Generate an acoustic wave spectrum matrix based on the vibration acoustic wave data;

[0087] The process of generating the acoustic wave spectrum matrix includes:

[0088] Perform short-time Fourier transform on the vibration sound wave data to extract the spectrum;

[0089] Short-time Fourier transform can convert sound signals from time domain to frequency domain and generate spectrum for further processing by computer.

[0090] Spectrum refers to the representation of a signal in the frequency domain, describing the signal's components and energy distribution at different frequencies. In the fields of audio, wireless communications, signal processing, etc., spectrum is a commonly used tool to analyze and understand the frequency characteristics of a signal.

[0091] The spectrum is represented by the sound wave spectrum matrix; each row and column in the sound wave spectrum matrix represents a different frequency, and each element represents the amplitude value of the corresponding frequency. Since machine learning cannot directly process the spectrum, the spectrum is represented by a matrix. The sound wave spectrum matrix reflects the sound characteristics during vibration, which is conducive to further improving the accuracy of data measurement in the future.

[0092] Accurate wind speed data can help developers optimize the design and performance of wind turbines. Different types of wind turbines respond differently to different wind speed ranges and wind directions. Understanding wind speed characteristics can help select the appropriate generator type and parameter settings, thereby improving power generation efficiency.

[0093] The first machine learning model for real-time correction of vibration-affected wind speed data is trained based on the acoustic spectrum matrix, vibration-affected wind speed data, and vibration-actual wind speed data;

[0094] A second machine learning model is trained based on air humidity data, water vapor effect wind speed data, and water vapor actual wind speed data to correct water vapor effect wind speed data in real time;

[0095] A third machine learning model is trained based on fatigue time, fatigue-affected wind speed data, and fatigue-real wind speed data to correct fatigue-affected wind speed data in real time;

[0096] The training process of the first machine learning model includes:

[0097] The vibration real wind speed data is taken as a set of acoustic wave spectrum matrices and labels of vibration-affected wind speed data;

[0098] A set of sound wave spectrum matrices, vibration-affected wind speed data, and vibration-actual wind speed data are used as inputs of a first machine learning model, and the first machine learning model uses the corrected vibration-affected wind speed data as output; a real-time set of sound wave spectrum matrices and vibration-affected wind speed data corresponding to the vibration-actual wind speed data are used as prediction targets, and minimizing the loss function value of the first machine learning model is used as a training target; and training is stopped when the loss function value of the first machine learning model is less than or equal to a preset first target loss value;

[0099] The first machine learning model loss function is a mean square error; the mean square error is calculated by The model is trained with minimization as the goal. In the loss function, mse is the loss function value, i is the group number of the acoustic spectrum matrix and the wind speed data affected by vibration; u is the group number of the acoustic spectrum matrix and the wind speed data affected by vibration; y i is the label corresponding to the i-th group of sound wave spectrum matrix and vibration-affected wind speed data, It is the label for real-time correction of the i-th group of acoustic spectrum matrix and vibration-affected wind speed data.

[0100] In the actual operation of the wind speed sensor, the real-time wind speed data FS is used as the input vibration-affected wind speed data; the second machine learning model is the same as the third machine learning model.

[0101] The training process of the second machine learning model includes:

[0102] The real wind speed data of water vapor is used as a set of labels of air humidity data and wind speed data affected by water vapor;

[0103] A set of air humidity data, water vapor affected wind speed data and water vapor real wind speed data are used as inputs of the second machine learning model, and the second machine learning model uses the corrected water vapor affected wind speed data as output; a real-time set of air humidity data and water vapor affected wind speed data corresponding to the water vapor real wind speed data are used as prediction targets, and minimizing the loss function value of the second machine learning model is used as the training target; training is stopped when the loss function value of the second machine learning model is less than or equal to the preset second target loss value.

[0104] The training process of the third machine learning model includes:

[0105] The fatigue real wind speed data is used as a set of labels of fatigue time and fatigue-affected wind speed data;

[0106] A set of fatigue time, fatigue-affected wind speed data and fatigue real wind speed data are used as inputs of a third machine learning model, and the third machine learning model uses the corrected fatigue-affected wind speed data as output; a real-time set of fatigue time and fatigue-affected wind speed data corresponding to the real wind speed data are used as prediction targets, and minimizing the loss function value of the third machine learning model is used as the training target; and training is stopped when the loss function value of the third machine learning model is less than or equal to the preset third target loss value.

[0107] The first machine learning model, the second machine learning model and the third machine learning model are recurrent neural network models.

[0108] Uncertainty is a technical term in physics. It means the degree of uncertainty about the measured value due to the existence of measurement error. Conversely, it also indicates the reliability of the result. It is an indicator of the quality of the measurement result. The smaller the uncertainty, the higher the quality, the higher the level, and the higher its use value; the larger the uncertainty, the lower the quality of the measurement result, the lower the level, and the lower its use value.

[0109] Uncertainty refers to the inaccuracy of the measurement result caused by experimental conditions, measuring equipment, human errors, etc. when measuring or estimating a physical quantity. Uncertainty refers to the difference between the measurement result and the true value, and refers to the degree of uncertainty about the measured value.

[0110] The method of evaluating standard uncertainty by statistical analysis of the observation series is called uncertainty type A evaluation; the method of evaluating standard uncertainty by a method different from statistical analysis of the observation series is called uncertainty type B evaluation.

[0111] In the uncertainty assessment of Class A, Class A uncertainty Where n represents the number of consecutive measurements, FS i represents the wind speed data measured for the i-th time;

[0112] It can be understood that, taking the first machine learning model as an example, the completion of the training of the first machine learning model is equivalent to constructing a function G(X1, X2) whose independent variables are the sound wave spectrum matrix X1 and the vibration-affected wind speed data X2, and whose dependent variable is the vibration real wind speed data; the same is true for the second machine learning model and the third machine learning model;

[0113] Type B uncertainty of vibration and acoustic wave data In the formula It means to find the function G(X1, X2) with respect to X iThe partial derivative of ; k represents the number of independent variables of function G; The independent variable is X i Uncertainty; when the independent variable is X1, Set to 1; when the independent variable is X2, Set to the absolute value of the difference between the vibration real wind speed data and the vibration affected wind speed data.

[0114] Type B uncertainty U of air humidity data B2 and the Type B uncertainty U about fatigue time B3 The calculation method is the same;

[0115] Will U A , U B1 , U B2 with U B3 The square mean of is taken as the comprehensive uncertainty;

[0116] The average value of the vibration real wind speed data, water vapor real wind speed data and fatigue real wind speed data output in real time by the first machine learning model, the second machine learning model and the third machine learning model is taken as the comprehensive real wind speed data.

[0117] The comprehensive real wind speed data comprehensively considers the different errors caused by various influencing factors under complex working conditions, minimizes the errors caused by sensors under complex working conditions, and helps workers use more accurate data for analysis and effectively utilize wind power energy.

[0118] Example 1 reduces errors as much as possible through random algorithms, and uses multiple machine learning models to comprehensively consider the authenticity of data, corrects the data, and creatively proposes a comprehensive uncertainty evaluation method based on artificial intelligence empowerment, which makes up for the absence of new energy sites in data uncertainty evaluation, obtains more accurate sensor data, and promotes the effective use of new energy.

[0119] Example 2, please refer to Figure 2 As shown, according to another aspect of the present application, an electronic device 500 for evaluating uncertainty of online monitoring data of new energy stations under complex working conditions is also provided. The electronic device 500 may include one or more processors and one or more memories. The memory stores a computer readable code, and when the computer readable code is executed by one or more processors, the above-mentioned method for evaluating uncertainty of online monitoring data of new energy stations under complex working conditions can be executed.

[0120] The method or system according to the embodiment of the present application can also be used by Figure 2 The electronic device architecture shown in FIG. Figure 2As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as the ROM 503 or the hard disk 507, may store the uncertainty evaluation method for online monitoring data of new energy sites under complex working conditions provided in this application. The uncertainty evaluation method for online monitoring data of new energy sites under complex working conditions may, for example, include: collecting wind speed data, vibration sound wave data, vibration affected wind speed data, vibration real wind speed data, water vapor affected wind speed data, water vapor real wind speed data, fatigue affected wind speed data, fatigue real wind speed data, air humidity data and fatigue time; generating a sound wave spectrum matrix based on the vibration sound wave data; training a first machine learning model for real-time correction of vibration affected wind speed data based on the sound wave spectrum matrix, vibration affected wind speed data and vibration real wind speed data; training a second machine learning model for real-time correction of water vapor affected wind speed data based on air humidity data, water vapor affected wind speed data and water vapor real wind speed data; training a third machine learning model for real-time correction of fatigue affected wind speed data based on fatigue time, fatigue affected wind speed data and fatigue real wind speed data; calculating Class A uncertainty based on wind speed data; calculating Class B uncertainty and comprehensive real wind speed data based on the first machine learning model, the second machine learning model and the third machine learning model; calculating comprehensive uncertainty based on Class A uncertainty and Class B uncertainty.

[0121] Furthermore, the electronic device 500 may also include a user interface 508. Figure 2 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 2 One or more components of an electronic device are shown.

[0122] Example 3, please refer to Figure 3 As shown, a computer-readable storage medium 600 according to an embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by the processor, the uncertainty evaluation method for online monitoring data of new energy sites under complex working conditions according to the embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0123] In addition, according to the implementation mode of the present application, the process described with reference to the flowchart above can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium, the non-transitory machine-readable storage medium stores machine-readable instructions, the machine-readable instructions can be executed by a processor to execute instructions corresponding to the method steps provided in the present application, such as: collecting wind speed data, vibration sound wave data, vibration-affected wind speed data, vibration-real wind speed data, water vapor-affected wind speed data, water vapor-real wind speed data, fatigue-affected wind speed data, fatigue-real wind speed data, air humidity data and fatigue time; generating a sound wave spectrum matrix based on the vibration sound wave data; generating a sound wave spectrum matrix based on the sound wave spectrum matrix, vibration-affected wind speed data and vibration-real wind speed data A first machine learning model is trained to correct the wind speed data affected by vibration in real time; a second machine learning model is trained to correct the wind speed data affected by water vapor in real time based on air humidity data, wind speed data affected by water vapor and real wind speed data of water vapor; a third machine learning model is trained to correct the wind speed data affected by fatigue in real time based on fatigue time, wind speed data affected by fatigue and real wind speed data of fatigue; Class A uncertainty is calculated based on wind speed data; Class B uncertainty and comprehensive real wind speed data are calculated based on the first machine learning model, the second machine learning model and the third machine learning model; and comprehensive uncertainty is calculated based on Class A uncertainty and Class B uncertainty.

[0124] When the computer program is executed by a central processing unit (CPU), the above functions defined in the method of the present application are performed.

[0125] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments 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 or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through a wired network or a wireless network. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0126] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0127] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0128] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a method for evaluating the uncertainty of online monitoring data of new energy sites under complex working conditions. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0129] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0130] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0131] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0132] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. The uncertainty evaluation method of online monitoring data of new energy stations under complex working conditions is characterized by: include, Collect new energy station data and generate an acoustic spectrum matrix based on vibration acoustic wave data; The first machine learning model for real-time correction of vibration-affected wind speed data is trained based on the acoustic spectrum matrix, vibration-affected wind speed data, and vibration-actual wind speed data; A second machine learning model is trained based on air humidity data, water vapor effect wind speed data, and water vapor actual wind speed data to correct water vapor effect wind speed data in real time; Based on fatigue time, fatigue-affected wind speed data and fatigue-actual wind speed data, a third machine learning model is trained to correct fatigue-affected wind speed data in real time, Class A uncertainty is calculated based on wind speed data, and Class B uncertainty and comprehensive actual wind speed data are calculated based on the first machine learning model, the second machine learning model and the third machine learning model; Calculate the combined uncertainty based on Type A uncertainty and Type B uncertainty.

2. The uncertainty evaluation method for online monitoring data of new energy stations under complex working conditions as claimed in claim 1 is characterized by: The collected new energy station data include wind speed data, vibration sound wave data, vibration affected wind speed data, vibration real wind speed data, water vapor affected wind speed data, water vapor real wind speed data, fatigue affected wind speed data, fatigue real wind speed data, air humidity data and fatigue time.

3. The uncertainty evaluation method for online monitoring data of new energy stations under complex working conditions as claimed in claim 2 is characterized by: The Class A uncertainty refers to the degree of uncertainty about wind speed under statistical analysis conditions; The Class B uncertainty refers to the degree to which non-statistical analysis methods cannot be certain about wind speed; There are F wind speed sensors installed on the top of the same wind turbine. A sensor weight QZ is set for each wind speed sensor, and the sensor weight QZ is normalized so that the sensor weight QZ adds up to 1. At the same time, the sensor weight is randomly reset every unit time. Wind speed data Where fs i represents the wind speed displayed by the i-th sensor; F is the number of wind speed sensors; QZ i represents the sensor weight of the i-th sensor.

4. The uncertainty evaluation method for online monitoring data of new energy stations under complex working conditions as claimed in claim 3 is characterized by: The acoustic wave spectrum matrix includes: performing short-time Fourier transform on the vibration acoustic wave data to extract the spectrum; The spectrum is represented as a sound wave spectrum matrix; each row and each column in the sound wave spectrum matrix represents a different frequency, and each element represents the amplitude value of the corresponding frequency.

5. The uncertainty evaluation method for online monitoring data of new energy stations under complex working conditions as claimed in claim 4 is characterized by: The first machine learning model includes taking the vibration real wind speed data as a set of acoustic spectrum matrices and labels of vibration-affected wind speed data; A set of sound wave spectrum matrices, vibration-affected wind speed data, and vibration-actual wind speed data are used as inputs of a first machine learning model, and the first machine learning model uses the corrected vibration-affected wind speed data as output; a real-time set of sound wave spectrum matrices and vibration-affected wind speed data corresponding to the vibration-actual wind speed data are used as prediction targets, and minimizing the loss function value of the first machine learning model is used as a training target; and training is stopped when the loss function value of the first machine learning model is less than or equal to a preset first target loss value; The first machine learning model loss function is a mean square error; the mean square error is calculated by The training model is trained with minimization as the goal. In the loss function, mse is the loss function value, i is the group number of the acoustic spectrum matrix and the wind speed data affected by vibration; u is the group number of the acoustic spectrum matrix and the wind speed data affected by vibration; y i is the label corresponding to the i-th group of sound wave spectrum matrix and vibration-affected wind speed data, is the label for real-time correction of the i-th group of acoustic spectrum matrix and vibration-affected wind speed data; In actual operation of the wind speed sensor, the wind speed data FS acquired in real time is used as input vibration-affected wind speed data.

6. The uncertainty evaluation method for online monitoring data of new energy stations under complex working conditions as claimed in claim 5 is characterized by: The second machine learning model includes using the water vapor real wind speed data as a set of air humidity data and water vapor impact wind speed data labels; A set of air humidity data, water vapor affected wind speed data and water vapor real wind speed data are used as inputs of a second machine learning model, and the second machine learning model uses the corrected water vapor affected wind speed data as output; a set of real-time air humidity data and water vapor affected wind speed data corresponding to the real water vapor wind speed data are used as prediction targets, and minimizing the loss function value of the second machine learning model is used as a training target; Stop training when the loss function value of the second machine learning model is less than or equal to the preset second target loss value.

7. The uncertainty evaluation method for online monitoring data of new energy stations under complex working conditions as claimed in claim 6 is characterized by: The third machine learning model includes using fatigue real wind speed data as a set of labels of fatigue time and fatigue-affected wind speed data; A set of fatigue time, fatigue-affected wind speed data and fatigue-actual wind speed data are used as inputs of a third machine learning model, wherein the third machine learning model uses corrected fatigue-affected wind speed data as output; A set of real-time fatigue time and fatigue-affected wind speed data corresponding to fatigue real wind speed data is used as the prediction target, and minimizing the loss function value of the third machine learning model is used as the training target; Stop training when the loss function value of the third machine learning model is less than or equal to the preset third target loss value; The first machine learning model, the second machine learning model and the third machine learning model are recurrent neural network models.

8. The uncertainty evaluation method for online monitoring data of new energy stations under complex working conditions as claimed in claim 7 is characterized by: The calculated combined uncertainty includes, Type A uncertainty Where n represents the number of consecutive measurements, FS i represents the wind speed data measured for the i-th time; The first machine learning model is constructed with the sound wave spectrum matrix X1 and the vibration-affected wind speed data X2 as independent variables, and the dependent variable is the function G(X1, X2) of the vibration real wind speed data; Type B uncertainty of vibration and acoustic wave data In the formula It means to find the function G(X1, X2) with respect to X i The partial derivative of; k represents the number of independent variables of function G; U Xi The independent variable is X i The uncertainty of Type B uncertainty U of air humidity data B2 and the Type B uncertainty U about fatigue time B3 The calculation method is the same; Will U A , U B1 , U B2 with U B3 The square mean of is taken as the comprehensive uncertainty; The average value of the vibration real wind speed data, water vapor real wind speed data and fatigue real wind speed data output in real time by the first machine learning model, the second machine learning model and the third machine learning model is taken as the comprehensive real wind speed data.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images