Method, system and apparatus for temporarily replacing faulty wind lidar

Abstract

Disclosed in the present application are a method, system and apparatus for temporarily replacing a faulty wind LiDAR. The system comprises: a detection module, used for analyzing output data of a radar, judging, on the basis of the analysis result, whether a LiDAR has had a fault, and if a fault has occurred, outputting logic 0, otherwise, outputting logic 1; a data prediction module, used for predicting five types of meteorological information the wind speed, wind direction, temperature, atmospheric pressure and humidity of wind farm on the basis of historical data and meteorological source data; and a data switching module, used for switching a data source from a simulation system data source to a wind LiDAR apparatus data source when the output logic changes from 0 to 1, switching the data source from the wind LiDAR apparatus data source to the simulation system data source when the logic value changes from 1 to 0, and maintaining an original state when the logic value remains unchanged. The present application can temporarily replace faulty wind LiDARs, thereby reducing the negative effects of wind LiDAR faults on assessment of wind farm output.

Description

Method, system and device for temporarily replacing a faulty laser wind measuring radar

[0001] This application claims priority to the Chinese patent application filed with the China Patent Office on August 22, 2023, with application number 202311058750X and application name “Method, system and device for temporarily replacing a faulty laser wind measurement radar”, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The present application relates to the field of laser wind measurement radars, and in particular to a method, system, and device for temporarily replacing a faulty laser wind measurement radar. Background Art

[0003] Due to the dynamic nature of wind resources, offshore wind farms often install laser wind radars to accurately assess renewable energy output. These radars, like miniature weather stations, provide real-time monitoring of five key meteorological data points: wind speed, wind direction, temperature, air pressure, and humidity. However, as offshore wind farm sites shift from nearshore to deep sea, malfunctions can hinder timely on-site repairs, leading to the loss of all meteorological information collected during the outage and directly impacting wind farm output estimates.

[0004] Summary of the Invention

[0005] The purpose of this application is to provide a method, system and device for temporarily replacing a faulty laser wind measuring radar, aiming to solve the problem of temporarily replacing a faulty laser wind measuring radar and reduce the negative impact of the laser wind measuring radar failure on the wind farm output judgment.

[0006] The present application provides a method for temporarily replacing a faulty laser wind measurement radar, comprising:

[0007] The detection module analyzes the radar's output data and determines whether the laser wind radar has failed based on the analysis results. If a failure occurs, the module outputs logic 0; otherwise, the module outputs logic 1.

[0008] The data prediction module predicts five types of meteorological information of the wind farm, including wind speed, wind direction, temperature, air pressure and humidity, based on historical data and meteorological source data;

[0009] Through the data switching module, when the output logic changes from 0 to 1, the data source is switched from the simulation system data source to the laser wind measurement radar device data source; when the logic value changes from 1 to 0, the data source is switched from the laser wind measurement radar device data source to the simulation system data source; when the logic value does not change, the original state is maintained.

[0010] Based on the above technical content, when a laser wind radar failure is detected, the data prediction module predicts five types of meteorological information of the wind farm based on historical data and meteorological source data, and switches the data source from the laser wind radar device data source to the simulation system data source; the ability to predict meteorological information during the laser wind radar failure can reduce the negative impact of the laser wind radar failure on the wind farm output judgment.

[0011] In one possible implementation, the data prediction module predicts five types of meteorological information of the wind farm, including wind speed, wind direction, temperature, air pressure and humidity, based on historical data and meteorological source data, specifically including: using a conditional autoregressive flow model to model the multidimensional data distribution of high-resolution wind farm meteorological information, and recursively extrapolating data at future moments by sampling the learned distribution.

[0012] Furthermore, the conditional autoregressive flow model is used to model the multidimensional data distribution of high-resolution wind farm meteorological information, which can improve the prediction accuracy of future meteorological information.

[0013] In one possible implementation, the conditional autoregressive flow model includes: a compression layer, a reversible convolution layer, an affine coupling layer, a separation layer, and a periodic fitting layer connected in sequence. The meteorological information is normalized and then input into the compression layer. The compression layer compresses the high-resolution meteorological information and the low-resolution meteorological information and extracts significant features before inputting into the reversible convolution layer. The reversible convolution layer shuffles and reorganizes the information of different dimensions and then inputs into the affine coupling layer. The affine coupling layer obtains the relationship between a certain dimension and other dimensions between wind speed, wind direction, temperature, air pressure, and humidity based on the affine function. The affine coupling layer is used for collaborative prediction of the entire dimension. The separation layer is used to restore the noise latent variables and the guided low-resolution meteorological information. The periodic fitting layer is used to ensure that the generated wind farm meteorological data can capture the real periodic variation law. The condition of the conditional autoregressive flow model is to propagate information forward and backward based on the neighborhood content of the low-resolution meteorological forecast data, thereby obtaining maximum fluidity and effectively guiding the establishment of multi-dimensional high-resolution wind farm meteorological data.

[0014] In a possible implementation, the affine coupling layer obtains the relationship between a certain dimension and other dimensions of wind speed, wind direction, temperature, air pressure and humidity based on the affine function, specifically including: in the first stage, the input and output of the four dimensions of wind direction o, temperature t, humidity h and air pressure p are fixed unchanged, and the wind speed s dimension is affine transformed. The formula of affine transformation is o, t, h, p = o, t, h, p; s i =w s (o,t,h,p)*s i-1 +w e(o, t, h, p); where the scaling function w_s and the translation function w_e are the neural network models to be learned; in the second stage, the input and output of the four dimensions of wind speed, temperature, humidity and air pressure are fixed, and an affine transformation is performed on the wind direction dimension; in the third stage, the input and output of the four dimensions of wind speed, wind direction, humidity and air pressure are fixed, and an affine transformation is performed on the temperature dimension; in the fourth stage, the input and output of the four dimensions of wind speed, wind direction, temperature and air pressure are fixed, and an affine transformation is performed on the humidity dimension; in the fifth stage, the input and output of the four dimensions of wind speed, wind direction, temperature and air pressure are fixed, and an affine transformation is performed on the air pressure dimension; all affine transformations are reversible processes, so that the interaction relationship between different dimensions is obtained through the cyclic transformation of five stages, and the said interaction relationship is used to promote the collaborative prediction of the entire dimension.

[0015] Furthermore, the affine transformations in the five stages of the affine coupling layer are all reversible processes. The interaction relationship between different dimensions is obtained through cyclic transformation, which can improve the accuracy of the conditional autoregressive flow model in establishing the multidimensional data distribution of high-resolution wind farm meteorological information.

[0016] The present application also provides a system for temporarily replacing a faulty laser wind measurement radar, comprising:

[0017] The detection module is used to analyze the output data of the radar and determine whether the laser wind measurement radar has a fault based on the analysis results. If a fault occurs, the output is logic 0, otherwise it outputs logic 1;

[0018] The data prediction module is used to predict five types of meteorological information of the wind farm, including wind speed, wind direction, temperature, air pressure and humidity, based on historical data and meteorological source data;

[0019] The data switching module is used to switch the data source from the simulation system data source to the laser wind measurement radar device data source when the output logic changes from 0 to 1, and to switch the data source from the laser wind measurement radar device data source to the simulation system data source when the logic value changes from 1 to 0, and to maintain the original state when the logic value does not change.

[0020] In a possible implementation, the data prediction module is specifically configured to: use a conditional autoregressive flow model to model the multidimensional data distribution of high-resolution wind farm meteorological information, and recursively infer data at future moments by sampling the learned distribution.

[0021] In one possible implementation, the conditional autoregressive flow model includes: a compression layer, a reversible convolution layer, an affine coupling layer, a separation layer, and a periodic fitting layer connected in sequence. The meteorological information is normalized and then input into the compression layer. The compression layer compresses the high-resolution meteorological information and the low-resolution meteorological information and extracts significant features before inputting into the reversible convolution layer. The reversible convolution layer shuffles and reorganizes the information of different dimensions and then inputs into the affine coupling layer. The affine coupling layer obtains the relationship between a certain dimension and other dimensions between wind speed, wind direction, temperature, air pressure, and humidity based on the affine function. The affine coupling layer is used for collaborative prediction of the entire dimension. The separation layer is used to restore the noise latent variables and the guided low-resolution meteorological information. The periodic fitting layer is used to ensure that the generated wind farm meteorological data can capture the real periodic variation law. The condition of the conditional autoregressive flow model is to propagate information forward and backward based on the neighborhood content of the low-resolution meteorological forecast data, thereby obtaining maximum fluidity and effectively guiding the establishment of multi-dimensional high-resolution wind farm meteorological data.

[0022] In a possible implementation, the affine coupling layer obtains the relationship between a certain dimension and other dimensions of wind speed, wind direction, temperature, air pressure and humidity based on the affine function, specifically including: in the first stage, the input and output of the four dimensions of wind direction o, temperature t, humidity h and air pressure p are fixed unchanged, and the wind speed s dimension is affine transformed. The formula of affine transformation is o, t, h, p = o, t, h, p; s i =w s (o,t,h,p)*s i-1 +w e (o, t, h, p); where the scaling function w_s and the translation function w_e are the neural network models to be learned; in the second stage, the input and output of the four dimensions of wind speed, temperature, humidity and air pressure are fixed, and an affine transformation is performed on the wind direction dimension; in the third stage, the input and output of the four dimensions of wind speed, wind direction, humidity and air pressure are fixed, and an affine transformation is performed on the temperature dimension; in the fourth stage, the input and output of the four dimensions of wind speed, wind direction, temperature and air pressure are fixed, and an affine transformation is performed on the humidity dimension; in the fifth stage, the input and output of the four dimensions of wind speed, wind direction, temperature and air pressure are fixed, and an affine transformation is performed on the air pressure dimension; all affine transformations are reversible processes, so that the interaction relationship between different dimensions is obtained through the cyclic transformation of five stages, and the said interaction relationship is used to promote the collaborative prediction of the entire dimension.

[0023] An embodiment of the present application also provides a device for temporarily replacing a faulty laser wind measurement radar, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps of the above method when executed by the processor.

[0024] An embodiment of the present application further provides a computer-readable storage medium, on which a program for implementing information transmission is stored, and when the program is executed by a processor, the steps of the above method are implemented.

[0025] An embodiment of the present application further provides a computer program product, including a computer program, which implements the steps of the above method when executed by a processor.

[0026] By adopting the embodiments of the present application, a faulty laser wind measuring radar can be temporarily replaced, meteorological information can be predicted during the laser wind measuring radar failure, and the negative impact of the laser wind measuring radar failure on the wind farm output judgment can be reduced.

[0027] The above description is only an overview of the technical solution of this application. In order to more clearly understand the technical means of this application, it is implemented in accordance with the contents of the specification, and in order to make the above and other purposes, features and advantages of this application more obvious and easy to understand, the specific implementation methods of this application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the specific implementation methods or the description of the prior art. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0029] FIG1 is a flow chart of a method for temporarily replacing a faulty laser wind measuring radar according to an embodiment of the present application;

[0030] FIG2 is a schematic diagram of a system for temporarily replacing a faulty laser wind measurement radar according to an embodiment of the present application;

[0031] FIG3 is a schematic diagram of a device for temporarily replacing a faulty laser wind measurement radar according to an embodiment of the present application. DETAILED DESCRIPTION

[0032] The following will clearly and completely describe the technical solutions of this application in conjunction with the embodiments. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0033] Method Example

[0034] According to an embodiment of the present application, a method for temporarily replacing a faulty laser wind measuring radar is provided. FIG1 is a flow chart of the method for temporarily replacing a faulty laser wind measuring radar according to an embodiment of the present application. As shown in FIG1 , the method specifically includes:

[0035] The detection module analyzes the radar's output data and determines whether the laser wind radar has failed based on the analysis results. If a failure occurs, the module outputs logic 0; otherwise, the module outputs logic 1.

[0036] The data prediction module predicts five types of meteorological information of the wind farm, including wind speed, wind direction, temperature, air pressure and humidity, based on historical data and meteorological source data;

[0037] Through the data switching module, when the output logic changes from 0 to 1, the data source is switched from the simulation system data source to the laser wind measurement radar device data source; when the logic value changes from 1 to 0, the data source is switched from the laser wind measurement radar device data source to the simulation system data source; when the logic value does not change, the original state is maintained.

[0038] The data prediction module predicts five types of meteorological information of wind farms, including wind speed, wind direction, temperature, air pressure and humidity, based on historical data and meteorological source data. Specifically, it uses the conditional autoregressive flow model to model the multidimensional data distribution of high-resolution wind farm meteorological information, and recursively infers data at future moments by sampling the learned distribution.

[0039] The conditional autoregressive flow model includes: a compression layer, a reversible convolution layer, an affine coupling layer, a separation layer and a periodic fitting layer connected in sequence. The meteorological information is normalized and then input into the compression layer. The compression layer compresses the high-resolution meteorological information and the low-resolution meteorological information and extracts significant features and then inputs into the reversible convolution layer. The reversible convolution layer shuffles and reorganizes the information of different dimensions and then inputs into the affine coupling layer. The affine coupling layer obtains the relationship between a certain dimension and other dimensions between wind speed, wind direction, temperature, air pressure and humidity based on the affine function. The affine coupling layer is used for collaborative prediction of the entire dimension. The separation layer is used to restore the noise latent variables and the guided low-resolution meteorological information. The periodic fitting layer is used to ensure that the generated wind farm meteorological data can capture the real periodic variation law. The condition of the conditional autoregressive flow model is to propagate information forward and backward based on the neighborhood content of the low-resolution meteorological forecast data, so as to obtain the maximum fluidity and effectively guide the establishment of multi-dimensional high-resolution wind farm meteorological data.

[0040] The affine coupling layer obtains the relationship between one dimension and other dimensions of wind speed, wind direction, temperature, air pressure and humidity based on the affine function. Specifically, in the first stage, the input and output of the four dimensions of wind direction o, temperature t, humidity h and air pressure p are fixed, and the wind speed s dimension is affine transformed. The formula of affine transformation is o, t, h, p = o, t, h, p; s i =w s (o,t,h,p)*s i-1 +w e(o, t, h, p); the scaling function w_s and the translation function w_e are the neural network models to be learned; in the second stage, the input and output of the four dimensions of wind speed, temperature, humidity and air pressure are fixed, and the wind direction dimension is subjected to an affine transformation; in the third stage, the input and output of the four dimensions of wind speed, wind direction, humidity and air pressure are fixed, and the temperature dimension is subjected to an affine transformation; in the fourth stage, the input and output of the four dimensions of wind speed, wind direction, temperature and air pressure are fixed, and the humidity dimension is subjected to an affine transformation; in the fifth stage, the input and output of the four dimensions of wind speed, wind direction, temperature and humidity are fixed, and the pressure dimension is subjected to an affine transformation; all affine transformations are reversible processes, so that the interaction relationship between different dimensions is obtained through the cyclic transformation of the five stages, and the interaction relationship is used to promote the collaborative prediction of the entire dimension.

[0041] The specific implementation methods are as follows:

[0042] The system mainly consists of three modules: laser wind radar fault detection module, data simulation calculation and prediction module, and laser wind radar data transmission channel switching module.

[0043] Laser wind radar fault detection module: By judging the variability of the radar's output data before and after, if multiple data such as air pressure, wind speed, wind direction, and temperature do not change within 10 minutes, it can be determined that the laser wind radar has failed and output a logical value of 0; otherwise, it outputs a logical value of 1.

[0044] The data switching module changes the data source according to the logic value output result. When the logic value changes from 0 to 1, the data source is switched from the simulation system data source to the laser wind measurement radar device data source. When the logic value changes from 1 to 0, the data source is switched from the laser wind measurement radar device data source to the simulation system data source. When the logic value does not change, the original state is maintained.

[0045] The data simulation calculation and prediction module needs to simulate five types of meteorological data of the wind farm in the future, including wind speed, wind direction, temperature, air pressure, and humidity, based on historical data and meteorological source forecast data. These five types of data are closely correlated, but their mutual dependence is difficult to describe through explicit mathematical formulas. At the same time, although the multidimensional space composed of the historical data series of the five types of indicators contains huge prior knowledge, the complex multidimensional distribution cannot be described by empirical distribution expressions.

[0046] The current highest precision resolution scale of the meteorological source for weather forecasts is 25 kilometers, and the content contained is relatively rough. The layout range of wind turbines in wind farms is generally between 3 kilometers and 5 kilometers, which requires very detailed meteorological change information. In order to reconstruct the real data variability of the wind farm from the existing rough meteorological source information, the actual meteorological data variables of the wind farm can be assumed to be composed of two parts: low-frequency meteorological source data variables and high-frequency components. Due to the randomness of meteorological data changes, and the information that needs to be learned consists of five dimensions: wind speed, wind direction, temperature, air pressure, and humidity, the embodiment of the present application does not directly learn the function mapping relationship between low-frequency meteorological source data and high-frequency wind farm data, but uses a conditional autoregressive flow model to model the multidimensional data distribution of high-frequency wind farms, and recursively infers the data at future moments step by step by sampling the learned distribution.

[0047] Assuming the noise distribution of the training and test data is identical, the conditional autoregressive flow model designed in this embodiment is expressed as: P(y_t|x_t-1,x_t,x_t+1). The overall framework of the model is as follows: During the training phase, an encoding process is performed. Guided by a historical low-frequency meteorological source data sequence, high-frequency meteorological information is encoded into noise latent variables that follow a certain distribution through a reversible neural network. During the inference phase, it is transformed into a decoding process. Under the influence of future fuzzy meteorological source data, the original noise latent variables are decoded into detailed future wind farm meteorological content. The parameters of the conditional autoregressive flow model are updated using the maximum likelihood function and gradient descent method. This model can effectively solve the learning and prediction problems of multidimensional data.

[0048] The model of conditional autoregressive flow consists of an activation layer, an affine layer, a compression layer, and a separation layer. First, the data is normalized. The compression layer merges high-resolution meteorological information with low-resolution meteorological information, while the separation layer does the opposite. In view of the internal interaction between the five variables of wind speed, wind direction, temperature, air pressure, and humidity, in order to explore the relationship between each dimension and other dimensions, the embodiment of the present application designs an affine function of dimensional interaction. In the first stage, the input and output of the four dimensions of wind direction o, temperature t, humidity h, and air pressure p are fixed and remain unchanged, and the wind speed s dimension is affine transformed. The formula of affine transformation is o, t, h, p = o, t, h, p; s(i) = w_s(o, t, h, p) * s(i-1) + w_e(o, t, h, p); where the scaling function w_s and the translation function w_e are the neural network models to be learned. In the second stage, the inputs and outputs of wind speed, temperature, humidity, and air pressure are fixed, and an affine transformation is performed on the wind direction dimension. In the third stage, the inputs and outputs of wind speed, wind direction, humidity, and air pressure are fixed, and an affine transformation is performed on the temperature dimension. In the fourth stage, the inputs and outputs of wind speed, wind direction, temperature, and air pressure are fixed, and an affine transformation is performed on the humidity dimension. In the fifth stage, the inputs and outputs of wind speed, wind direction, temperature, and air pressure are fixed, and an affine transformation is performed on the air pressure dimension. All affine transformations are reversible. This five-stage cyclic transformation process yields interactive relationships between different dimensions, which effectively promotes collaborative prediction across all dimensions.

[0049] It's generally assumed that meteorological changes follow the Markov property, meaning the meteorological state at the next moment is determined solely by the current moment's state. In light of the Markov property, the meteorological state at each moment should be closely related to the meteorological state at the previous and next moments, and independent of the state at non-adjacent moments. Therefore, a bidirectional flow structure consisting of a time-lagged stream and a time-advanced stream is introduced to guide the reconstruction of high-resolution wind farm meteorological information. The output features of the time-lagged stream are modeled as: h_i = f_z(x_t, x_t-1, h_i-1), with information propagating in the lagging direction. The output features of the time-advanced stream are modeled as: h_i = f_c(x_t, x_t+1, h_i+1), with information propagating in the leading direction. The output feature values ​​of the two streams are fused and concatenated to form low-resolution meteorological source information for the current and adjacent moments, which serves as the condition for updating high-resolution wind farm meteorological data at the target moment. The network weights of f_c and f_z at each time point are shared, but the weights of the autoregressive stream model are not shared but interconnected. In this way, the information of each high-resolution wind farm output can fully interact with the low-resolution neighborhood meteorological source information.

[0050] Taking into account the daily changing characteristics of the weather over time, generally speaking, the temperature is a cyclical process of gradually rising from low to high and then gradually falling from high to low from the early morning of the first day to the early morning of the second day. The humidity is a cyclical process of gradually falling from high to low and then gradually rising from low to high from the early morning of the first day to the early morning of the second day. Therefore, the historical data in these two dimensions should have a cyclical change trend. However, since these changing trends are mainly affected by sunrise and sunset, and the times of sunrise and sunset change dynamically, with sunrise early and sunset late in summer and sunrise late and sunset early in winter, they are not fixed time periods. For this reason, the embodiment of the present application introduces a variable period fitting layer, sets the time reference vector to t = 12 hours according to sunrise at 6 am and sunset at 6 pm, and constructs 12 different time periods according to the 12 months of the year. This layer approximates the periodic variation of temperature and humidity in one year through the weighted superposition of 12 sine and cosine functions, namely y = Σa_icos(it) + b_isin(it). a and b are calculated through a nonlinear neural network model. The final model can fully learn the periodic variation of the sequence dynamics, thereby guiding the accurate establishment of meteorological information of the wind farm.

[0051] During normal periods when radar data is not faulty, a reverse isolation device is used to transmit the data to the data center for training without prediction. When a radar failure occurs, only prediction is performed without training.

[0052] This embodiment of the application proposes learning the mapping relationship between the probability distribution of low-resolution meteorological source data, the distribution of noise latent variables, and the probability distribution of high-resolution laser wind radar meteorological data. This mapping relationship is modeled using a reversible conditional autoregressive flow model. The autoregressive flow uses a reversible multi-dimensional alternating dependency affine transformation. The low-resolution flow information is propagated in two different directions, leading and lagging, to fully explore the continuity and variability of features. At the same time, the periodic variation of sunrise and sunset is also learned through trigonometric transformation.

[0053] From the description of the above embodiments, it can be seen that in the embodiments of the present application, when a laser wind measuring radar failure is detected, five types of meteorological information of the wind farm are predicted based on historical data and meteorological source data through the data prediction module, and the data source is switched from the laser wind measuring radar device data source to the simulation system data source; the meteorological information can be predicted during the laser wind measuring radar failure, which can reduce the negative impact of the laser wind measuring radar failure on the wind farm output judgment.

[0054] System or device embodiment 1

[0055] According to an embodiment of the present application, a system for temporarily replacing a faulty laser wind measurement radar is provided. FIG2 is a schematic diagram of the system for temporarily replacing a faulty laser wind measurement radar according to an embodiment of the present application. As shown in FIG2 , the system specifically includes:

[0056] The detection module is used to analyze the output data of the radar and determine whether the lidar has a fault based on the analysis results. If a fault occurs, it outputs a logic 0, otherwise it outputs a logic 1;

[0057] The data prediction module is used to predict five types of meteorological information of the wind farm, including wind speed, wind direction, temperature, air pressure and humidity, based on historical data and meteorological source data;

[0058] The data switching module is used to switch the data source from the simulation system data source to the laser wind measurement radar device data source when the output logic changes from 0 to 1, and to switch the data source from the laser wind measurement radar device data source to the simulation system data source when the logic value changes from 1 to 0, and to maintain the original state when the logic value does not change.

[0059] The data prediction module is specifically used to: use the conditional autoregressive flow model to model the multidimensional data distribution of high-resolution wind farm meteorological information, and recursively infer future data by sampling the learned distribution.

[0060] The conditional autoregressive flow model includes: a compression layer, a reversible convolution layer, an affine coupling layer, a separation layer and a periodic fitting layer connected in sequence. The meteorological information is normalized and then input into the compression layer. The compression layer compresses the high-resolution meteorological information and the low-resolution meteorological information and extracts significant features and then inputs into the reversible convolution layer. The reversible convolution layer shuffles and reorganizes the information of different dimensions and then inputs into the affine coupling layer. The affine coupling layer obtains the relationship between a certain dimension and other dimensions between wind speed, wind direction, temperature, air pressure and humidity based on the affine function. The affine coupling layer is used for collaborative prediction of the entire dimension. The separation layer is used to restore the noise latent variables and the guided low-resolution meteorological information. The periodic fitting layer is used to ensure that the generated wind farm meteorological data can capture the real periodic variation law. The condition of the conditional autoregressive flow model is to propagate information forward and backward based on the neighborhood content of the low-resolution meteorological forecast data, so as to obtain the maximum fluidity and effectively guide the establishment of multi-dimensional high-resolution wind farm meteorological data.

[0061] The affine coupling layer obtains the relationship between one dimension and other dimensions of wind speed, wind direction, temperature, air pressure and humidity based on the affine function. Specifically, in the first stage, the input and output of the four dimensions of wind direction o, temperature t, humidity h and air pressure p are fixed, and the wind speed s dimension is affine transformed. The formula of affine transformation is o, t, h, p = o, t, h, p; s i =w s (o,t,h,p)*s i-1 +w e(o, t, h, p); the scaling function w_s and the translation function w_e are the neural network models to be learned; in the second stage, the input and output of the four dimensions of wind speed, temperature, humidity and air pressure are fixed, and the wind direction dimension is subjected to an affine transformation; in the third stage, the input and output of the four dimensions of wind speed, wind direction, humidity and air pressure are fixed, and the temperature dimension is subjected to an affine transformation; in the fourth stage, the input and output of the four dimensions of wind speed, wind direction, temperature and air pressure are fixed, and the humidity dimension is subjected to an affine transformation; in the fifth stage, the input and output of the four dimensions of wind speed, wind direction, temperature and air pressure are fixed, and the pressure dimension is subjected to an affine transformation; all affine transformations are reversible processes, so that the interaction relationship between different dimensions is obtained through the cyclic transformation of the five stages, and the interaction relationship is used to promote the coordinated prediction of the entire dimension.

[0062] The embodiment of the present application is a system embodiment corresponding to the above-mentioned method embodiment. The specific operations of each module can be understood by referring to the description of the method embodiment, and will not be repeated here.

[0063] Device Example 1

[0064] An embodiment of the present application provides a device for temporarily replacing a faulty laser wind measurement radar, as shown in Figure 3, including: a memory 30, a processor 32, and a computer program stored in the memory 30 and executable on the processor 32. When the computer program is executed by the processor, the steps in the above method embodiment are implemented.

[0065] Device Example 2

[0066] An embodiment of the present application provides a computer-readable storage medium, on which a program for implementing information transmission is stored. When the program is executed by the processor 32, the steps in the above method embodiment are implemented.

[0067] Device Example 3

[0068] An embodiment of the present application provides a computer program product, including a computer program. When the computer program is executed by the processor 32, the steps in the above method embodiment are implemented.

[0069] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. These modifications or replacements of the technical solutions of the embodiments of the present application do not cause the essence of the corresponding technical solutions to deviate from the scope of the present solutions.

Claims

1. A system for temporarily replacing a faulty laser wind radar, characterized in that: include: The detection module is used to analyze the output data of the radar and determine whether the laser radar has a fault based on the analysis results. If a fault occurs, the output is logic 0, otherwise the output is logic 1; Data prediction module, used to predict five types of meteorological information of wind farms, including wind speed, wind direction, temperature, air pressure and humidity, based on historical data and meteorological source data; The data switching module is used to switch the data source from the simulation system data source to the laser wind measurement radar device data source when the output logic changes from 0 to 1, and to switch the data source from the laser wind measurement radar device data source to the simulation system data source when the logic value changes from 1 to 0, and to maintain the original state when the logic value does not change.

2. The system according to claim 1, characterized in that The data prediction module is specifically used to: use a conditional autoregressive flow model to model the multidimensional data distribution of high-resolution wind farm meteorological information, and recursively infer data at future moments by sampling and learning the distribution.

3. The system according to claim 2, characterized in that The conditional autoregressive flow model includes: a compression layer, a reversible convolution layer, an affine coupling layer, a separation layer and a periodic fitting layer connected in sequence. The meteorological information is normalized and then input into the compression layer. The compression layer compresses the high-resolution meteorological information and the low-resolution meteorological information and extracts significant features and then inputs into the reversible convolution layer. The reversible convolution layer shuffles and reorganizes the information of different dimensions and then inputs into the affine coupling layer. The affine coupling layer obtains the relationship between a certain dimension and other dimensions among wind speed, wind direction, temperature, air pressure and humidity based on the affine function. The affine coupling layer is used for the coordinated prediction of the entire dimension. The separation layer is used to restore the noise latent variables and the guided low-resolution meteorological information. The periodic fitting layer is used to ensure that the generated wind farm meteorological data can capture the real periodic variation law. The condition of the conditional autoregressive flow model is to propagate information forward and backward based on the neighborhood content of the low-resolution meteorological forecast data, so as to obtain the maximum fluidity and effectively guide the establishment of multi-dimensional high-resolution wind farm meteorological data.

4. The system according to claim 3, characterized in that The affine coupling layer obtains the relationship between a certain dimension and other dimensions of wind speed, wind direction, temperature, air pressure and humidity based on the affine function, specifically including: in the first stage, the input and output of the four dimensions of wind direction o, temperature t, humidity h and air pressure p are fixed, and the wind speed s dimension is affine transformed. The formula of affine transformation is o, t, h, p = o, t, h, p; s i =w s (o,t,h,p)*s i-1 +w e (o, t, h, p); where the scaling function w_s and the translation function w_e are the neural network models to be learned; in the second stage, the input and output of the four dimensions of wind speed, temperature, humidity and air pressure are fixed, and an affine transformation is performed on the wind direction dimension; in the third stage, the input and output of the four dimensions of wind speed, wind direction, humidity and air pressure are fixed, and an affine transformation is performed on the temperature dimension; in the fourth stage, the input and output of the four dimensions of wind speed, wind direction, temperature and air pressure are fixed, and an affine transformation is performed on the humidity dimension; in the fifth stage, the input and output of the four dimensions of wind speed, wind direction, temperature and humidity are fixed, and an affine transformation is performed on the air pressure dimension; all affine transformations are reversible processes, so that the interaction relationship between different dimensions is obtained through the cyclic transformation of five stages, and the interaction relationship is used to promote the collaborative prediction of the entire dimension.

5. A method for temporarily replacing a faulty laser wind measuring radar, characterized in that: include: The detection module analyzes the output data of the radar and determines whether the laser radar is faulty based on the analysis results. If a fault occurs, the output is logic 0, otherwise the output is logic 1; The data prediction module predicts five types of meteorological information of the wind farm, including wind speed, wind direction, temperature, air pressure and humidity, based on historical data and meteorological source data; Through the data switching module, when the output logic changes from 0 to 1, the data source is switched from the simulation system data source to the laser wind and thunder measurement data source. The data source of the device is switched from the laser wind radar device data source to the simulation system data source when the logic value changes from 1 to 0, and the original state is maintained when the logic value does not change.

6. The method according to claim 5, characterized in that The data prediction module predicts five types of meteorological information of wind farms, including wind speed, wind direction, temperature, air pressure and humidity, based on historical data and meteorological source data. Specifically, the conditional autoregressive flow model is used to model the multidimensional data distribution of high-resolution wind farm meteorological information, and the data at future times is recursively inferred by sampling the learned distribution.

7. The method according to claim 6, characterized in that The conditional autoregressive flow model includes: a compression layer, a reversible convolution layer, an affine coupling layer, a separation layer and a periodic fitting layer connected in sequence. The meteorological information is normalized and then input into the compression layer. The compression layer compresses the high-resolution meteorological information and the low-resolution meteorological information and extracts significant features and then inputs into the reversible convolution layer. The reversible convolution layer shuffles and reorganizes the information of different dimensions and then inputs into the affine coupling layer. The affine coupling layer obtains the relationship between a certain dimension and other dimensions among wind speed, wind direction, temperature, air pressure and humidity based on the affine function. The affine coupling layer is used for the coordinated prediction of the entire dimension. The separation layer is used to restore the noise latent variables and the guided low-resolution meteorological information. The periodic fitting layer is used to ensure that the generated wind farm meteorological data can capture the real periodic variation law. The condition of the conditional autoregressive flow model is to propagate information forward and backward based on the neighborhood content of the low-resolution meteorological forecast data, so as to obtain the maximum fluidity and effectively guide the establishment of multi-dimensional high-resolution wind farm meteorological data.

8. The method according to claim 7, characterized in that The affine coupling layer obtains the relationship between a certain dimension and other dimensions of wind speed, wind direction, temperature, air pressure and humidity based on the affine function, specifically including: in the first stage, the input and output of the four dimensions of wind direction o, temperature t, humidity h and air pressure p are fixed, and the wind speed s dimension is affine transformed. The formula of affine transformation is o, t, h, p = o, t, h, p; s i =w s (o,t,h,p)*s i-1 +w e (o, t, h, p); where the scaling function w_s and the translation function w_e are the neural network models to be learned; in the second stage, the input and output of the four dimensions of wind speed, temperature, humidity and air pressure are fixed, and an affine transformation is performed on the wind direction dimension; in the third stage, the input and output of the four dimensions of wind speed, wind direction, humidity and air pressure are fixed, and an affine transformation is performed on the temperature dimension; in the fourth stage, the input and output of the four dimensions of wind speed, wind direction, temperature and air pressure are fixed, and an affine transformation is performed on the humidity dimension; in the fifth stage, the input and output of the four dimensions of wind speed, wind direction, temperature and humidity are fixed, and an affine transformation is performed on the air pressure dimension; all affine transformations are reversible processes, so that the interaction relationship between different dimensions is obtained through the cyclic transformation of five stages, and the interaction relationship is used to promote the collaborative prediction of the entire dimension.

9. A device for temporarily replacing a faulty laser wind measuring radar, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the steps of the method for temporarily replacing a faulty laser wind measuring radar according to any one of claims 5 to 8 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores an information transmission implementation program, and when the program is executed by a processor, the steps of the method for temporarily replacing a faulty laser wind measuring radar according to any one of claims 5 to 8 are implemented.

11. A computer program product, characterized in that The method comprises a computer program, which, when executed by a processor, implements the steps of the method for temporarily replacing a faulty laser wind measuring radar according to any one of claims 5 to 8.

Images