CN117390368A

CN117390368A - Lightning probability calculation method, device and equipment for wind turbine and storage medium

Info

Publication number: CN117390368A
Application number: CN202311672843.1A
Authority: CN
Inventors: 朱琳; 王松; 吴智泉; 贾世迎; 吴文韬; 贾启彤
Original assignee: Yunnan Power Investment Green Energy Technology Co ltd
Current assignee: Yunnan Power Investment Green Energy Technology Co ltd
Priority date: 2023-12-07
Filing date: 2023-12-07
Publication date: 2024-01-12
Anticipated expiration: 2043-12-07
Also published as: CN117390368B

Abstract

The method utilizes the physical characteristics of lightning stroke, is based on the charge density of downlink pilot and the initial physical mechanism of uplink pilot, provides a lightning receiving probability calculation method when a blade is in a lightning receiving interval, enables the lightning receiving probability of the blade to be realized and quantized, and meanwhile, a worker can dynamically adjust the position of a lightning receiving point in the lightning receiving interval according to the lightning receiving probability obtained by the current calculation of the blade so as to update the lightning receiving probability of the adjusted lightning receiving point in real time, and can position the mounting position of the lightning receiving device according to the updated lightning receiving probability without experience judgment. And the method also carries out machine learning on the charge density obtained by lightning meteorological data so as to reduce errors and simultaneously provide a prediction function.

Description

Lightning probability calculation method, device and equipment for wind turbine and storage medium

Technical Field

The application relates to the technical field of electric digital data processing, in particular to a lightning probability calculation method, device and equipment of a wind turbine and a storage medium.

Background

The wind power generation is to convert the kinetic energy of wind into electric energy, the wind energy is a clean and pollution-free renewable energy source, the wind power generation drives the windmill blades to rotate through wind power, and the rotating speed is improved through a speed increaser to promote the generator to generate electricity, and the wind power generation does not need to use fuel, does not generate radiation or air pollution, and is a renewable new energy source. Wind power generation mainly converts wind energy into mechanical work through a wind motor (wind driven generator), and the mechanical work drives a rotor to rotate, and finally, alternating current is output.

In order to ensure good operation efficiency of the wind turbine, the wind turbine is usually required to be built in a relatively flat area so as to avoid influence on wind speed caused by buildings, mountains, forests and the like. Because the length of the sleeve and the blade of the wind motor is longer, the blade of the wind motor is positioned at a high position, and is easy to be struck by lightning in thunderstorm weather.

Lightning protection for wind motors is usually achieved by installing a lightning receptor on a blade, and then acquiring the hardware condition of the blade through on-site judgment in daily inspection and maintenance by staff. When the blade is struck by lightning in a small-amplitude to medium-amplitude manner, cracks and bulge formed by the blade are not obvious, the cracks and bulge are difficult to find in the inspection process, the small cracks and bulge are not treated in time, and further damages such as ponding and insolation cracking are easy to cause, so that the safe operation of the wind turbine generator is influenced; after the lightning strikes the lightning receptor, the current generated by the lightning strike is conducted to the grounding end through the lead wire connected with the lightning receptor, so that the blade is effectively protected from being damaged by the lightning strike.

At present, the installation of the lightning receptor is still in an experience stage, the installation principle of the lightning receptor is to install the lightning receptor with corresponding number according to different lengths of the blades, and the lightning strike characteristics and probability of the fan blades are not clear, so that the lightning receptor installed by experience cannot effectively reduce the lightning strike probability of the blades, the protection pertinence of the lightning receptor is not strong, and the lightning protection effect is not good.

Disclosure of Invention

The main object of the application is to provide a lightning receiving probability calculation method, device, equipment and storage medium of a wind turbine, so as to solve the problem that in the prior art, lightning receiving probability of a blade cannot be effectively reduced by a lightning receiving device, and the lightning receiving device is poor in protection pertinence and lightning protection effect.

In order to achieve the above object, the present application provides the following technical solutions:

the utility model provides a lightning probability calculation method of wind-powered electricity generation machine, wind-powered electricity generation machine includes a tower section of thick bamboo that is fixed in ground and install in three blade on the tower section of thick bamboo, every blade respectively through same rotation axis with tower section of thick bamboo rotates to be connected and forms vertical rotation face, lightning probability calculation method includes:

acquiring the shortest vertical distance and the longest vertical distance between a lightning cloud layer of an area where the wind turbine is located and all blades;

Defining the traversed length from the shortest vertical distance to the longest vertical distance as a lightning receiving interval;

defining a downlink pilot charge equation according to the same flash point in the flash receiving interval;

acquiring a plurality of historical lightning meteorological data of the lightning cloud layer, substituting the historical lightning meteorological data into the downlink pilot charge equation, and performing machine learning on all calculation results to obtain a charge density prediction model;

acquiring real-time lightning meteorological data of the lightning cloud layer and substituting the real-time lightning meteorological data into the charge density prediction model to obtain a charge density prediction value;

acquiring a part of the rotating surface, which coincides with the flash receiving section, as a flash receiving area, and defining a plurality of flash receiving points of the flash receiving area according to a preset strategy;

defining an uplink pilot length equation based on the current flash point, substituting a charge density predicted value into the uplink pilot length equation, and calculating to obtain the uplink pilot length of the current flash point;

calculating the current flash receiving area of the current flash receiving point based on the uplink pilot length of the current flash receiving point, and calculating the current flash receiving times of the current flash receiving point according to the current flash receiving area of the current flash receiving point;

and adding the flash receiving times of all the flash receiving points and calculating arithmetic average based on the current flash receiving points to obtain the flash receiving probability of the current flash receiving points.

As a further improvement of the present application, obtaining the shortest vertical distance and the longest vertical distance between the lightning cloud layer and all blades in the area where the wind turbine is located, includes:

when one blade is positioned at the top of the rotating surface and is in a vertical state, acquiring the minimum distance between the blade and the lightning cloud layer and defining the minimum vertical distance;

when one blade is positioned at the bottom of the rotating surface and is in a vertical state, the minimum distance between the other two blades and the lightning cloud layer is obtained and defined as the longest vertical distance.

As a further refinement of the present application, defining a downstream pilot charge equation from the same flash point in the flash interval includes:

defining the downstream pilot charge equation according to equation (1):

（1）；

wherein,the charge density of a pilot current column between the current flash point and the lightning cloud layer;The current flash point is the distance from the pilot flow column;A current peak value for the lightning cloud layer;The ground height of the lightning cloud layer is;The ground height is the current access point;；；；；。

As a further improvement of the present application, a plurality of historical lightning meteorological data of the lightning cloud layer are obtained and substituted into the downlink pilot charge equation, and machine learning is performed on all calculation results to obtain a charge density prediction model, including:

Acquiring a current peak value and a ground height of historical lightning meteorological data, substituting the current peak value and the ground height into the downlink pilot charge equation, and calculating to obtain charge density of the historical lightning meteorological data;

integrating the charge density of all lightning meteorological data and storing the charge density as a preset data set;

normalizing the preset data set to obtain a normalized data set;

dividing the normalized data set into a training set, a verification set and a test set according to a preset proportion;

defining a topological relation of a neural network model, wherein the topological relation comprises an input layer, an implicit layer and an output layer which are sequentially connected through signals;

outputting the training set to the input layer, carrying out iteration for preset times through the neural network model, and respectively acquiring root mean square errors of the training set and the current training result based on each iteration;

and acquiring the minimum value in all root mean square errors, and acquiring a training result corresponding to the minimum value as the charge density prediction model.

As a further improvement of the present application, the topological relation is characterized by formula (2):

（2）；

wherein,modeling the neural network;Is the +.>A plurality of input nodes, each input node corresponding to a set of training data of the training set,/- >Is the +.>The input node is to the ++>Preset weights of the input nodes;For connecting to the +.>A threshold of the input nodes;is a transfer function, and->；The upper corner mark (2) in the method is a second layer, namely an hidden layer;The upper corner marks (1, 2) of (a) are from the first layer to the second layer, namely from the input layer to the hidden layer.

As a further improvement of the present application, defining an uplink pilot length equation based on a current access point, and substituting a charge density predicted value into the uplink pilot length equation, and calculating to obtain an uplink pilot length of the current access point, including:

defining the upstream pilot length equation according to equation (3):

（3）；

wherein,the +.>Upstream pilot length of secondary corona;A pilot increment for each corona;For the conversion of the electricity required per unit length of the lead, +.>；Is the charge density prediction value;for the pilotThe volume of the flow column;Is an environmental factor->；Is->The electric field strength of the secondary corona;Is->Pilot head potential of secondary corona.

As a further improvement of the present application, calculating a current flash receiving area of the current flash receiving point based on an uplink pilot length of the current flash receiving point, and calculating a flash receiving number of the current flash receiving point according to the current flash receiving area of the current flash receiving point, includes:

In iteration (3)And based on each iteration determine +.>Whether the preset threshold value is larger than or equal to the preset threshold value;

if yes, the threshold value is greater than or equal to the preset threshold valueDefining a hit radius of a current access flash point;

calculating the current flash area of the flash point according to equation (4):

（4）；

wherein,a hit radius for the current access point;For the polarity constant of the lightning cloud layer, when the lightning cloud layer is positive in polarity +.>When the thunder cloud layer is negative in polarity +.>；

Calculating the number of flash times of the current flash point according to the formula (5):

（5）；

wherein,the positive polarity flash number based on the current preset natural period is used for the current flash point;The average flash receiving times of a plurality of preset natural periods;The probability distribution of the current peak value when the lightning cloud layer is positive in polarity is obtained;The negative polarity flash receiving times based on the current preset natural period are used for the current flash receiving point;And the probability distribution of the current peak value when the lightning cloud layer is negative in polarity is obtained.

In order to achieve the above purpose, the present application further provides the following technical solutions:

a lightning probability calculation device of a wind turbine, the lightning probability calculation device of a wind turbine being applied to the lightning probability calculation method of a wind turbine as described above, the lightning probability calculation device of a wind turbine comprising:

The vertical distance acquisition module is used for acquiring the shortest vertical distance and the longest vertical distance between the lightning cloud layer of the area where the wind turbine is located and all blades;

the flash receiving interval definition module is used for defining the traversed length from the shortest vertical distance to the longest vertical distance as a flash receiving interval;

the downlink pilot charge equation definition module is used for defining a downlink pilot charge equation according to the same flash point in the flash receiving section;

the charge density prediction model acquisition module is used for acquiring a plurality of historical lightning meteorological data of the lightning cloud layer, substituting the historical lightning meteorological data into the downlink pilot charge equation, and performing machine learning on all calculation results to obtain a charge density prediction model;

the charge density prediction value acquisition module is used for acquiring real-time lightning meteorological data of the lightning cloud layer and substituting the real-time lightning meteorological data into the charge density prediction model to obtain a charge density prediction value;

the flash point receiving defining module is used for acquiring a part of the rotating surface, which coincides with the flash point receiving section, as a flash point receiving area and defining a plurality of flash points of the flash point receiving area according to a preset strategy;

the uplink pilot length calculation module is used for defining an uplink pilot length equation based on the current flash point, substituting the charge density predicted value into the uplink pilot length equation, and calculating to obtain the uplink pilot length of the current flash point;

The flash receiving frequency calculation module is used for calculating the flash receiving area of the current flash receiving point based on the uplink pilot length of the current flash receiving point, and calculating the flash receiving frequency of the current flash receiving point according to the flash receiving area of the current flash receiving point;

and the lightning probability calculation module is used for summing the lightning times of all the lightning points and calculating arithmetic average based on the current lightning points to obtain the lightning probability of the current lightning points.

an electronic device comprising a processor, and a memory coupled to the processor, the memory storing program instructions executable by the processor; and the processor realizes the lightning probability calculation method of the wind turbine when executing the program instructions stored in the memory.

a storage medium having stored therein program instructions which, when executed by a processor, implement a lightning probability calculation method enabling the wind turbine as described above.

According to the method, the shortest vertical distance and the longest vertical distance between the lightning cloud layer of the area where the wind motor is located and all blades are obtained; defining the traversed length from the shortest vertical distance to the longest vertical distance as a lightning receiving interval; defining a downlink pilot charge equation according to the same flash point in the flash interval; acquiring a plurality of historical lightning meteorological data of a lightning cloud layer, substituting the historical lightning meteorological data into a downlink pilot charge equation, and performing machine learning on all calculation results to obtain a charge density prediction model; acquiring real-time lightning meteorological data of a lightning cloud layer and substituting the real-time lightning meteorological data into a charge density prediction model to obtain a charge density prediction value; acquiring a part of the rotating surface, which coincides with the flash receiving section, as a flash receiving area, and defining a plurality of flash receiving points of the flash receiving area according to a preset strategy; defining an uplink pilot length equation based on the current flash point, substituting the charge density predicted value into the uplink pilot length equation, and calculating to obtain the uplink pilot length of the current flash point; calculating the current flash receiving area of the current flash receiving point based on the uplink pilot length of the current flash receiving point, and calculating the current flash receiving times of the current flash receiving point according to the current flash receiving area of the current flash receiving point; and adding the flash receiving times of all the flash receiving points and calculating arithmetic average based on the current flash receiving points to obtain the flash receiving probability of the current flash receiving points. According to the lightning receiving probability calculation method based on the charge density of the downlink pilot and the initial physical mechanism of the uplink pilot, the lightning receiving probability calculation method for the blade in the lightning receiving interval is provided, so that the lightning receiving probability of the blade is enabled to be achieved and quantized, meanwhile, a worker can dynamically adjust the position of the lightning receiving point in the lightning receiving interval according to the lightning receiving probability calculated by the blade at the current time, so that the lightning receiving probability of the adjusted lightning receiving point is updated in real time, and the worker can position the mounting position of the lightning receiver according to the updated lightning receiving probability without experience judgment. And the method also carries out machine learning on the charge density obtained by lightning meteorological data so as to reduce errors and simultaneously provide a prediction function.

Drawings

FIG. 1 is a schematic flow chart of steps of an embodiment of a method for calculating a lightning probability of a wind turbine according to the present application;

FIG. 2 is a schematic diagram of a functional module of an embodiment of a lightning probability calculation device for a wind turbine according to the present disclosure;

FIG. 3 is a schematic structural diagram of one embodiment of an electronic device of the present application;

FIG. 4 is a schematic diagram illustrating the structure of one embodiment of a storage medium of the present application.

Detailed Description

The following description of the technical solutions in the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.

The terms "first," "second," "third," and the like in this application are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "first," "second," and "third" may explicitly or implicitly include at least one such feature. In the description of the present application, the meaning of "plurality" means at least two, for example, two, three, etc., unless specifically defined otherwise. All directional indications (such as up, down, left, right, front, back … …) in the embodiments of the present application are merely used to explain the relative positional relationship, movement, etc. between the components in a particular gesture (as shown in the drawings), and if the particular gesture changes, the directional indication changes accordingly. Furthermore, the terms "comprise" and "have," as well as any variations thereof, are intended to cover a non-exclusive inclusion. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those listed steps or elements but may include other steps or elements not listed or inherent to such process, method, article, or apparatus.

Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the present application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments.

As shown in fig. 1, the present embodiment provides an embodiment of a lightning probability calculation method of a wind turbine, in the present embodiment, the wind turbine includes a tower fixed on the ground and three blades mounted on the tower, each blade is rotatably connected with the tower through a same rotation axis to form a vertical rotation surface, and the lightning probability calculation method includes the following steps:

and S1, acquiring the shortest vertical distance and the longest vertical distance between the lightning cloud layer of the area where the wind motor is located and all blades.

Preferably, when the three blades are inverted Y-shaped, the shortest vertical distance is the shortest distance between the tip of the uppermost vertical blade and the lightning cloud layer; when the three blades are in a positive Y shape, the two uppermost blades and the horizontal plane respectively have angles of 30 degrees and 150 degrees, and at the moment, the blade tips of the two uppermost blades are positioned on the same horizontal plane, so that the shortest distance between the blade tip of one of the two blades and the lightning cloud layer is the shortest distance.

Step S2, defining the traversed length from the shortest vertical distance to the longest vertical distance as a lightning receiving interval.

It is understood that the lightning receiving zone is the shortest distance traversed by the tip of one of the blades sweeping through a sixth arc, which is located at the positive tip of the rotating surface.

Step S3, defining a downlink pilot charge equation according to the same flash point in the flash receiving section.

It can be understood that the flash point is the installation position of the flash receiver, and the same flash point in the flash receiving interval can randomly define the position (can be understood as the initial flash point position), and find the specific position through subsequent calculation and optimization.

Preferably, under lightning conditions, the electric field in the vicinity of the ground object is subjected to a live thundercloud in combination with a descending leader. The height of the thundercloud from the ground is usually 2km to 10km, an individual thundercloud usually has two to three charge centers, the uppermost layer of the thundercloud gathers a large amount of positive charges, the lower part of the thundercloud is negatively charged, one or more weak positive charge areas are sometimes arranged at the bottom, and field intensity with a certain magnitude is generated near the ground, and the absolute value of the amplitude is about 10kV/m to 20kV/m.

And S4, acquiring a plurality of historical lightning meteorological data of the lightning cloud layer, substituting the historical lightning meteorological data into a downlink pilot charge equation, and performing machine learning on all calculation results to obtain a charge density prediction model.

Preferably, the machine learning of the present embodiment employs a bp neural network.

And S5, acquiring real-time lightning meteorological data of the lightning cloud layer and substituting the real-time lightning meteorological data into a charge density prediction model to obtain a charge density prediction value.

Step S6, a part of the rotation surface, which coincides with the flash receiving section, is obtained as a flash receiving area, and a plurality of flash receiving points of the flash receiving area are defined according to a preset strategy.

Preferably, the preset strategy can be set empirically, and then the optimal position is calculated in an iterative manner, or the positions of the flash points can be set directly and randomly, and then the iterative calculation is performed.

Preferably, the portion of the rotation surface coinciding with the lightning receiving zone is in the shape of an arc having a certain thickness, the arc segment of the arc corresponds to the above-mentioned sixth circle, the certain thickness of the arc corresponds to the thickness of the rotation surface, and the thickness of the rotation surface corresponds to the thickness of the area swept by the blade when it rotates.

And S7, defining an uplink pilot length equation based on the current access point, substituting the charge density predicted value into the uplink pilot length equation, and calculating to obtain the uplink pilot length of the current access point.

Step S8, calculating the current flash receiving area of the current flash receiving point based on the uplink pilot length of the current flash receiving point, and calculating the current flash receiving times of the current flash receiving point according to the current flash receiving area of the current flash receiving point.

And S9, adding the flash receiving times of all the flash receiving points, and calculating arithmetic average based on the current flash receiving points to obtain the flash receiving probability of the current flash receiving points.

Preferably, the arithmetic average is directly based on the current flash point's number of flash times being the total number of flash times for all flash points.

Preferably, the different positions of the flash points affect different lightning probability, and the embodiment can find the optimal position of the flash points through a global optimization algorithm so as to make the probability of the flash points highest, thereby protecting the blade from lightning.

Specifically, the global optimizing algorithm finds the optimal position of the flash point by the following steps:

(1) Initializing a particle swarm: a number of particles are randomly generated and a position and velocity are randomly assigned to each particle.

(2) Calculating a fitness function: and calculating the fitness function value of each particle according to the lightning probability of the lightning point.

(3) Updating the individual optimal position: for each particle, its historical optimal position is recorded.

(4) Updating the group optimal position: and selecting the position with the optimal fitness function value from the historical optimal positions of all particles as the group optimal position.

(5) Update speed and location: the speed and position of each particle are updated based on the current position, speed and historical optimal position.

(6) Judging a stopping condition: if the stopping condition is met, outputting the optimal position of the current group; otherwise, returning to the step (2).

Further, the step S1 specifically includes the following steps:

in step S11, when one of the blades is located at the top of the rotation surface and is in a vertical state, the minimum distance between the blade and the lightning cloud layer is obtained and defined as the shortest vertical distance.

In step S12, when one of the blades is located at the bottom of the rotation surface and is in a vertical state, the minimum distance between the other two blades and the lightning cloud layer is obtained and defined as the longest vertical distance.

Preferably, the shortest vertical distance is the midpoint of the arc section of the arc, and the longest vertical distance is the two ends of the arc.

Preferably, the measurement of cloud height is a mature prior art, such as:

(1) And sounding the balloon.

The sounding balloon carries a GNSS receiver for positioning and height setting (or height setting is assisted by a barometer), and then a hygrometer and a camera are used for judging whether the sounding balloon enters a cloud layer, so that the cloud height can be judged.

(2) And (5) laser ranging.

The cloud layer reflects laser, and the laser distance meter (which is not daily and needs long distance and high power) can be used for measuring the height of the cloud (namely the distance from the ground)

And the measurement of the cloud layer height can be directly obtained from a meteorological bureau without obtaining difficulty, and the shortest vertical distance and the longest vertical distance can be obtained through simple addition and subtraction after the cloud layer height is obtained.

Further, the step S3 specifically includes the following steps:

step S31, defining a downstream pilot charge equation according to equation (1):

（1）。

wherein,the charge density of the pilot current column between the current flash point and the lightning cloud layer;The current distance between the flash point and the pilot flow column;The current peak value of the lightning cloud layer;Is the ground height of the thunder cloud layer;The ground height is the current access point;；；；；。

Further, the step S4 specifically includes the following steps:

and S41, acquiring a current peak value and a ground height of historical lightning meteorological data, substituting the current peak value and the ground height into a downlink pilot charge equation, and calculating to obtain the charge density of the historical lightning meteorological data.

Step S42, integrating the charge density of all lightning meteorological data and storing the charge density as a preset data set.

Step S43, carrying out normalization processing on the preset data set to obtain a normalized data set.

Preferably, the normalization method of zero-mean normalization (Z-score normalization) is preferred in this embodiment, and this method gives the mean (mean) and standard deviation (standard deviation) of the raw data to normalize the data, and the processed data conforms to the standard normal distribution, that is, the mean is 0 and the standard deviation is 1. For the normalization method, in this embodiment, batch normalization (Batch Normalization) may be used, compared with simple normalization when training is performed on the previous neural network, only normalization is performed on the input layer data, but no normalization is performed on the middle layer, although normalization is performed on the data set of the input node, the data distribution of the input data after matrix multiplication is more likely to be changed greatly, and as the number of network layers of the hidden layer is deepened continuously, the change of the data distribution will be larger and larger, so that the normalization is performed on the middle layer of the neural network by batch normalization, and the training effect is better.

And S44, dividing the normalized data set into a training set, a verification set and a test set according to a preset proportion.

Preferably, the preset ratio is generally 70%:15%: the proportion of 15% divides the image data into a training set, a verification set and a sample set, namely 70% of the data is the training set, 15% of the data is the verification set and 15% of the data is the sample set.

Step S45, defining a topological relation of the neural network model, wherein the topological relation comprises an input layer, an implicit layer and an output layer which are connected in sequence in a signal mode.

Step S46, the training set is output to the input layer, iteration is carried out for preset times through the neural network model, and root mean square errors of the verification set and the current training result are respectively obtained based on each iteration.

Preferably, the root mean square error is. Wherein (1)>Is root mean square error>For the number of training sets, +.>Is->True value of the individual training set, +.>Is->Training results after the training of the training sets are completed.

It should be noted that the root mean square error formula is only used for principle illustration, and the letter symbols and meanings of the root mean square error formula are not mutually communicated with other formulas in the embodiment.

In step S47, the minimum value of all the root mean square errors is obtained, and the training result corresponding to the minimum value is obtained as the charge density prediction model.

Further, the topological relation is characterized by the formula (2):

（2）。

wherein,is a neural network model;Is the->Each input node corresponds to one group of training data of the training set, ++>Is the->Input node to hidden layer +.>Preset weights of the input nodes;To connect to the +.>A threshold of the input nodes;Is a transfer function, and。

the sign meaning of the transfer function is not communicated with other places.

Preferably, the number in brackets of the symbol corner mark in formula (2) of this embodiment is the number of layers, e.gThe upper corner mark (2) in (a) is a second layer, namely an implicit layer, < + >>The upper corner marks (1, 2) of (a) are from the first layer to the second layer, namely from the input layer to the hidden layer.

Preferably, training a model to train a neural network typically requires providing a large amount of data, i.e., a data set; the data sets are generally divided into three categories, namely training set (training set), validation set (validation set) and test set (test set) as described above.

Wherein, one epoch (increase in number) is a process equal to one training with all samples in the training set, which means one forward propagation (forward pass) and one backward propagation (back pass); when the number of samples (i.e., training sets) of one epoch is too large, excessive time may be consumed for performing one training, and it is not necessary to use all data of the training set for each training, the whole training set needs to be divided into a plurality of small blocks, that is, a plurality of batches for performing the training; one epoch is made up of one or more latches, which are part of a training set, with only a portion of the data being used for each training process, i.e., one latch, and one iteration being used for training one latch.

Preferably, the neural network training specifically comprises a Perceptron (Perceptron) composed of two layers of neurons, an input layer receiving external input signals and transmitting to an output layer, wherein the output layer is M-P neurons, and a sensor is arrangedAs a step function, and given a training data set, weight +.>(=1, 2,..n), and training threshold +.>Can be obtained by learning->It can be understood that a weight corresponding to a fixed value with a fixed input of-1, 0 +.>。

It should be noted that the step function is not interconnected with the symbolic meaning of the other formulas in the embodiment, and the step function is only schematically illustrated and does not participate in the calculation of the other formulas.

Preferably, the number of times of training the neural network in this embodiment may be set to 500 times.

Preferably, the learning rate of 1 st to 250 th epochs may be set to 0.01, the learning rate of 251 st to 325 th epochs may be set to 0.001, and the learning rate of 326 th to 1000 th epochs may be set to 0.0001.

It can be understood that the neural network training of this embodiment mainly includes the following ideas:

(1) initializing weight and bias items in a network, and initializing parameter values (the weight of an output unit, the bias items, the weight of a hidden unit and the bias items are all parameters of a model) to obtain the output value of each layer of elements for activating forward propagation, thereby obtaining the value of a loss function.

(2) And activating forward propagation to obtain the output value of each layer and the expected value of the loss function of each layer.

(3) According to the loss function, calculating an error term of the output unit and an error term of the hidden unit, calculating various errors, calculating a gradient of a parameter with respect to the loss function or calculating a partial derivative according to a calculus chain law. Solving partial derivatives for vectors or matrixes in the composite function, wherein the editing derivatives of the internal functions of the composite function are always multiplied left; for scalar bias derivative in the composite function, the derivative of the internal function of the composite function can be multiplied left or right.

(4) The weights and bias terms in the neural network are updated.

(5) Repeating the steps (2) - (4) until the loss function is smaller than a preset threshold or the iteration times are used up, and outputting the parameters at the moment to obtain the current optimal parameters.

Further, the step S7 specifically includes the following steps:

step S71, defining an uplink pilot length equation according to equation (3):

（3）。/>

wherein,the +.>Upstream pilot length of secondary corona;A pilot increment for each corona;For the conversion of the electricity required per unit length of the lead, +.>；Is a charge density prediction value;Is the volume of the pilot flow column;Is an environmental factor->(coulombs per volt per meter); / >Is->The electric field strength of the secondary corona;Is->Pilot head potential of secondary corona.

Preferably, the upstream pilot length of the first coronaHighly positive dependence on the flash point, +.>The length values of (2) are linearly distributed between 5m and 200m and increase with increasing height of the flash point. A minimum value of 5m may be reached when the flash point height is at an extremely low value (e.g., near 0m to 5 m), regardless; when the height of the flash point is 100m, < >>25m; when the height of the flash point is 200m, < >>60m; when the height of the flash point is 300m, < >>95m;60m; when the height of the flash point is 400m, < >>115m. Adopting a wind turbine with a blade length of 42.2m and a tower height of 100m, and neglecting the size of a rotating shaft, wherein the height interval of the wind turbine during rotation is 121.1m to 142.2m, and the wind turbine is +.>The length interval of (2) is 35m to 40m.

Preferably, the pilot head potentialThe absolute value of (2) increases with the increase of the upstream pilot length, and also shows positive correlation, the pilot head potential of the first corona is the initial value +.>Correspond to->In->25m>15MV (million volts); at->60m>-18.5MV; at->95m>-26.2MV; at->115m>Is-40.5 MV. Adopts- >The length of (2) is 35m to 40m, then +.>The absolute value of the potential of (a) is 15.5MV to 16.5MV.

Specifically, the starting end of the upstream pilot length is the current access point.

Further, the step S8 specifically includes the following steps:

step S81, iterative (3)And based on each iteration determine +.>Whether the threshold value is greater than or equal to a preset threshold value, ifAnd if the threshold is greater than or equal to the preset threshold, executing step S82.

Preferably, the preset threshold value is generally two meters, that is, the critical length of the pilot development is 2m, and when the critical length is greater than or equal to the critical length, the stable pilot initiation is determined.

Step S82, the step of setting the threshold value to be greater than or equal to the preset threshold valueDefined as the hit radius of the current access point.

Step S83, calculating the current flash area of the flash point according to equation (4):

（4）。

wherein,a hit radius for the current access point;Is the polarity constant of the lightning cloud layer, when the lightning cloud layer is positive in polarity +.>When the thunder cloud layer is negative in polarity +.>。

Preferably, the lightning receiving area is generally a horizontal circle.

Step S84, calculating the current flash number of the flash points according to the formula (5):

（5）。

wherein,the positive polarity flash number based on the current preset natural period is used for the current flash point;The average flash receiving times of a plurality of preset natural periods; / >The probability distribution of the current peak value when the lightning cloud layer is positive in polarity;the negative polarity flash receiving times based on the current preset natural period are used for the current flash receiving point;The probability distribution of the current peak value when the lightning cloud layer is negative in polarity.

Illustrating: this example simulates a typical wind motor in an air domain of 1000m by 4000m by MATLAB.

The precondition is that the equivalent physical model of the typical wind motor is 42.2m blade length, 100m tower height and 3 annular array blades with 120 degrees; the air-domain has a field strength of 15 kV/m; the height of the lightning cloud layer is 4000m; the first guide column is a cylindrical structure with the radius of 5 m. The lightning receiving area is arched, the vertical symmetry axis of the lightning receiving area is defined as 90 degrees, and the vertical symmetry axis is vertical to the horizontal plane, so that the starting point of the lightning receiving area is 30 degrees, the height is 0m, the midpoint of the lightning receiving area is 90 degrees, the height is 21.1m (the maximum length of the blade positioned in the lightning receiving area), the ending point of the lightning receiving area is 150 degrees, and the height is 0m;

the single blade lightning receiving probability in the lightning receiving area is obtained through 1000 times of simulation and is shown in the following table 1:

table 1 single blade lightning probability within the lightning receiving area.

From the above table, the blade lightning probability accords with the tip discharge characteristic, and the closer to the blade tip, the greater the lightning probability is, and the curve linear relation is formed.

According to the embodiment, the shortest vertical distance and the longest vertical distance between a lightning cloud layer of the area where the wind motor is located and all blades are obtained; defining the traversed length from the shortest vertical distance to the longest vertical distance as a lightning receiving interval; defining a downlink pilot charge equation according to the same flash point in the flash interval; acquiring a plurality of historical lightning meteorological data of a lightning cloud layer, substituting the historical lightning meteorological data into a downlink pilot charge equation, and performing machine learning on all calculation results to obtain a charge density prediction model; acquiring real-time lightning meteorological data of a lightning cloud layer and substituting the real-time lightning meteorological data into a charge density prediction model to obtain a charge density prediction value; acquiring a part of the rotating surface, which coincides with the flash receiving section, as a flash receiving area, and defining a plurality of flash receiving points of the flash receiving area according to a preset strategy; defining an uplink pilot length equation based on the current flash point, substituting the charge density predicted value into the uplink pilot length equation, and calculating to obtain the uplink pilot length of the current flash point; calculating the current flash receiving area of the current flash receiving point based on the uplink pilot length of the current flash receiving point, and calculating the current flash receiving times of the current flash receiving point according to the current flash receiving area of the current flash receiving point; and adding the flash receiving times of all the flash receiving points and calculating arithmetic average based on the current flash receiving points to obtain the flash receiving probability of the current flash receiving points. According to the lightning stroke probability calculation method, the physical characteristics of lightning strokes are utilized, the lightning stroke probability calculation method when the blade is in the lightning stroke region is provided based on the charge density of the downlink pilot and the initial physical mechanism of the uplink pilot, so that the lightning stroke probability of the blade is enabled to be achieved and quantized, meanwhile, a worker can dynamically adjust the position of the lightning stroke point in the lightning stroke region according to the lightning stroke probability calculated by the blade at the current time, so that the lightning stroke probability of the adjusted lightning stroke point is updated in real time, and the worker can position the mounting position of the lightning stroke receiver according to the updated lightning stroke probability without experience judgment. In addition, the embodiment also carries out machine learning on the charge density obtained by lightning meteorological data so as to reduce errors and simultaneously provide a prediction function. In addition, the embodiment simulates in MATLAB based on the probability calculation method, and simulates lightning probability distribution of a typical wind motor based on different blade postures, descending leading distances and lightning receptor arrangement modes of the lightning receptor areas, and simulation results show that the protection range of the lightning receptor can be changed along with the change of the angle of the blade, and the protection effect of the lightning receptor at the blade tip of the blade is better than that of the lightning receptor at the blade body of the blade.

As shown in fig. 2, this embodiment provides an embodiment of a lightning probability calculation device of a wind turbine, where the lightning probability calculation device is applied to the lightning probability calculation method in the above embodiment, and the device includes a vertical distance acquisition module 1, a lightning interval definition module 2, a downlink pilot charge equation definition module 3, a charge density prediction model acquisition module 4, a charge density prediction value acquisition module 5, a lightning point definition module 6, an uplink pilot length calculation module 7, a lightning number calculation module 8, and a lightning probability calculation module 9 that are electrically connected in sequence.

The vertical distance acquisition module 1 is used for acquiring the shortest vertical distance and the longest vertical distance between a lightning cloud layer of an area where the wind motor is located and all blades; the lightning receiving interval definition module 2 is used for defining the length traversed from the shortest vertical distance to the longest vertical distance as a lightning receiving interval; the downstream pilot charge equation definition module 3 is configured to define a downstream pilot charge equation according to the same flash point in the flash interval; the charge density prediction model acquisition module 4 is used for acquiring a plurality of historical lightning meteorological data of the lightning cloud layer, substituting the historical lightning meteorological data into a downlink pilot charge equation, and performing machine learning on all calculation results to obtain a charge density prediction model; the charge density prediction value acquisition module 5 is used for acquiring real-time lightning meteorological data of the lightning cloud layer and substituting the real-time lightning meteorological data into a charge density prediction model to obtain a charge density prediction value; the flash point definition module 6 is used for acquiring the overlapping part of the rotating surface and the flash zone as a flash zone and defining a plurality of flash points of the flash zone according to a preset strategy; the uplink pilot length calculation module 7 is used for defining an uplink pilot length equation based on the current contact point, substituting the charge density predicted value into the uplink pilot length equation, and calculating to obtain the uplink pilot length of the current contact point; the flash number calculation module 8 is configured to calculate a flash area of the current flash point based on an uplink pilot length of the current flash point, and calculate the flash number of the current flash point according to the flash area of the current flash point; the lightning probability calculation module 9 is configured to sum the lightning times of all the lightning points and calculate an arithmetic average based on the current lightning point to obtain the lightning probability of the current lightning point.

Further, the vertical distance acquisition module comprises a first vertical distance acquisition sub-module and a second vertical distance acquisition sub-module which are electrically connected in sequence; the second vertical distance acquisition sub-module is electrically connected with the lightning receiving interval definition module.

The first vertical distance acquisition submodule is used for acquiring the minimum distance between one blade and the lightning cloud layer and defining the minimum vertical distance when the blade is positioned at the top of the rotating surface and is in a vertical state; the second vertical distance obtaining submodule is used for obtaining the minimum distance between the other two blades and the lightning cloud layer and defining the minimum vertical distance when one of the blades is positioned at the bottom of the rotating surface and is in a vertical state.

Further, the downstream pilot charge equation defining module is specifically configured to define a downstream pilot charge equation according to equation (1):

（1）。

wherein,the charge density of the pilot current column between the current flash point and the lightning cloud layer;The current flash point is the distance from the pilot flow column;The current peak value of the lightning cloud layer;Is the ground height of the thunder cloud layer;the ground height is the current access point;；；；；。

Further, the charge density prediction model acquisition module comprises a first charge density prediction model acquisition sub-module, a second charge density prediction model acquisition sub-module, a third charge density prediction model acquisition sub-module, a fourth charge density prediction model acquisition sub-module, a fifth charge density prediction model acquisition sub-module, a sixth charge density prediction model acquisition sub-module and a seventh charge density prediction model acquisition sub-module which are electrically connected in sequence; the first charge density prediction model acquisition submodule is electrically connected with the downlink pilot charge equation definition module, and the seventh charge density prediction model acquisition submodule is electrically connected with the charge density prediction value acquisition module.

The first charge density prediction model acquisition submodule is used for acquiring a current peak value and a ground-to-ground height of historical lightning meteorological data and substituting the current peak value and the ground-to-ground height into a downlink pilot charge equation to calculate and obtain the charge density of the historical lightning meteorological data; the second charge density prediction model acquisition submodule is used for integrating charge densities of all lightning meteorological data and storing the charge densities as a preset data set; the third charge density prediction model acquisition submodule is used for carrying out normalization processing on a preset data set to obtain a normalized data set; the fourth charge density prediction model acquisition submodule is used for dividing the normalized data set into a training set, a verification set and a test set according to a preset proportion; the fifth charge density prediction model acquisition submodule is used for defining a topological relation of the neural network model, wherein the topological relation comprises an input layer, an implicit layer and an output layer which are connected in sequence in a signal manner; the sixth charge density prediction model obtaining submodule is used for outputting a training set to an input layer, carrying out iteration for preset times through a neural network model, and respectively obtaining root mean square errors of a verification set and a current training result based on each iteration; the seventh charge density prediction model obtaining sub-module is configured to obtain a minimum value of all root mean square errors, and obtain a training result corresponding to the minimum value as the charge density prediction model.

Further, the fifth charge density prediction model acquisition submodule is configured to carry a topological relation of the neural network model represented by the expression (2):

（2）。

wherein,is a neural network model;Is the->Each input node corresponds to one group of training data of the training set, ++>Is the->Input node to hidden layer +.>Preset weights of the input nodes;To connect to the +.>A threshold of the input nodes;Is a transfer function, and；the upper corner mark (2) in (a) is a second layer, namely an implicit layer, < + >>The upper corner marks (1, 2) of (a) are from the first layer to the second layer, namely from the input layer to the hidden layer.

Further, the upstream pilot length calculation module is specifically configured to define an upstream pilot length equation according to equation (3):

（3）。

wherein,the +.>Upstream pilot length of secondary corona;A pilot increment for each corona;For the conversion of the electricity required per unit length of the lead, +.>；Is a charge density prediction value;Is the volume of the pilot flow column;Is an environmental factor->；Is->The electric field strength of the secondary corona;Is->Pilot head potential of secondary corona.

Further, the flash receiving frequency calculation module comprises a first flash receiving frequency calculation sub-module, a second flash receiving frequency calculation sub-module, a third flash receiving frequency calculation sub-module, a fourth flash receiving frequency calculation sub-module and a fifth flash receiving frequency calculation sub-module which are electrically connected in sequence; the first receiving time computing sub-module is electrically connected with the uplink pilot length computing module, and the fifth receiving time computing sub-module is electrically connected with the receiving probability computing module.

Wherein the first lightning number calculation sub-module is used for iterating the uplink pilot length calculation module in (3)And based on each iteration determine +.>Whether the threshold value is larger than or equal to a preset threshold value.

The second lightning number calculation sub-module is used for ifIf the number is greater than or equal to the preset threshold value, the number is greater than or equal to the preset threshold value +.>Defined as the hit radius of the current access point.

The third flash number calculation sub-module is used for calculating the flash area of the current flash point according to the formula (4):

（4）。

wherein,a hit radius for the current access point;Is the polarity constant of the lightning cloud layer, when the lightning cloud layer is positive in polarity +.>When the thunder cloud layer is negativePolarity of->。

The fifth flash number calculation sub-module is configured to calculate the flash number of the current flash point according to equation (5):

（5）。

wherein,the positive polarity flash number based on the current preset natural period is used for the current flash point;The average flash receiving times of a plurality of preset natural periods;The probability distribution of the current peak value when the lightning cloud layer is positive in polarity;the negative polarity flash receiving times based on the current preset natural period are used for the current flash receiving point;The probability distribution of the current peak value when the lightning cloud layer is negative in polarity.

It should be noted that, the present embodiment is an apparatus embodiment based on the foregoing method embodiment, and additional contents such as optimization, expansion, limitation, and illustration of the present embodiment may be referred to the foregoing method embodiment, which is not repeated herein.

Fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application. As shown in fig. 3, the electronic device 10 includes a processor 101 and a memory 102 coupled to the processor 101.

The memory 102 stores program instructions for implementing a lightning probability calculation method for a wind turbine according to any of the above embodiments.

The processor 101 is configured to execute program instructions stored in the memory 102 to perform lightning probability calculation of the wind turbine.

The processor 101 may also be referred to as a CPU (Central Processing Unit ). The processor 101 may be an integrated circuit chip with signal processing capabilities. Processor 101 may also be a general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

Further, fig. 4 is a schematic structural diagram of a storage medium according to an embodiment of the present application, and referring to fig. 4, the storage medium 11 according to an embodiment of the present application stores a program instruction 111 capable of implementing all the methods described above, where the program instruction 111 may be stored in the storage medium in the form of a software product, and includes several instructions for making a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) execute all or part of the steps of the methods described in various embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, an optical disk, or other various media capable of storing program codes, or a terminal device such as a computer, a server, a mobile phone, a tablet, or the like.

In the several embodiments provided in the present application, it should be understood that the disclosed systems, apparatuses, and methods may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of elements is merely a logical functional division, and there may be additional divisions of actual implementation, e.g., multiple elements or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other forms.

In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units. The foregoing is only the embodiments of the present application, and is not intended to limit the scope of the patent application, and all equivalent structures or equivalent processes using the contents of the specification and drawings of the present application or directly or indirectly applied to other related technical fields are included in the scope of the patent protection of the present application.

Claims

1. The utility model provides a lightning probability calculation method of wind-powered electricity generation machine, wind-powered electricity generation machine includes a tower section of thick bamboo that is fixed in ground and install in three blade on the tower section of thick bamboo, every blade respectively through same rotation axis with tower section of thick bamboo rotates to be connected and forms vertical rotation face, its characterized in that, lightning probability calculation method includes:

2. The lightning probability calculation method according to claim 1, wherein obtaining the shortest vertical distance and the longest vertical distance between the lightning cloud layer and all blades in the area where the wind turbine is located comprises:

3. The lightning probability calculation method of claim 1, wherein defining a downstream pilot charge equation from the same lightning point in the lightning interval comprises:

Defining the downstream pilot charge equation according to equation (1):

（1）；

4. The lightning probability calculation method of claim 3, wherein obtaining a plurality of historical lightning meteorological data of the lightning cloud layer and substituting the historical lightning meteorological data into the downlink pilot charge equation, and performing machine learning on all calculation results to obtain a charge density prediction model, comprises:

normalizing the preset data set to obtain a normalized data set;

5. The lightning probability calculation method of claim 4, wherein the topological relation is characterized by the following formula (2):

（2）；

wherein,modeling the neural network;Is the +.>A plurality of input nodes, each input node corresponding to a set of training data of the training set,/->Is the +.>Input nodes to the hidden layerPreset weights of the input nodes;For connecting to the +.>A threshold of the input nodes;is a transfer function, and->；The upper corner mark (2) in (a) is a second layer, namely an implicit layer, < + >>The upper corner marks (1, 2) of (a) are from the first layer to the second layer, namely from the input layer to the hidden layer.

6. The lightning probability calculation method of claim 3, wherein defining an uplink pilot length equation based on a current lightning point and substituting a charge density predicted value into the uplink pilot length equation, calculating an uplink pilot length of the current lightning point, comprises:

Defining the upstream pilot length equation according to equation (3):

（3）；

wherein,the +.>Upstream pilot length of secondary corona;A pilot increment for each corona;for the conversion of the electricity required per unit length of the lead, +.>；Is the charge density prediction value;A volume for the pilot flow column;Is an environmental factor->；Is->The electric field strength of the secondary corona;Is->Secondary electric powerThe pilot head potential of the halo.

7. The method of claim 6, wherein calculating the current flash point's flash area based on the current flash point's upstream pilot length and calculating the current flash point's flash number based on the current flash point's flash area comprises:

（4）；

wherein,a hit radius for the current access point;For the polarity constant of the lightning cloud layer, when the lightning cloud layer is positive in polarity +.>When the thunder cloud layer is negative in polarity +. >；

（5）；

8. A lightning probability calculation device of a wind turbine, the lightning probability calculation device of a wind turbine being applied to the lightning probability calculation method of a wind turbine according to one of claims 1 to 7, characterized in that the lightning probability calculation device of a wind turbine comprises:

9. An electronic device comprising a processor, and a memory coupled to the processor, the memory storing program instructions executable by the processor; the processor, when executing the program instructions stored in the memory, implements the lightning probability calculation method for a wind turbine according to any one of claims 1 to 7.

10. A storage medium having stored therein program instructions which, when executed by a processor, implement a lightning probability calculation method capable of implementing a wind turbine according to any one of claims 1 to 7.