CN113486580A - High-precision numerical modeling method, server and storage medium for in-service wind turbine generator - Google Patents

Abstract

The present application provides a high-precision numerical modeling method, a server and a storage medium for an in-service wind turbine, wherein the method includes the following steps: obtaining sample data: obtaining a calibration value of the material parameters of the wind turbine, sampling within the error interval of the calibration value, Process to obtain material parameter samples; build a simulation model of the wind turbine, simulate and output the node displacement samples corresponding to each material parameter sample; divide the sample data into a training set and a test set; build a neural network, and combine the material parameter samples in the training set As the input of the neural network, use the node displacement samples in the training set as the output of the neural network; train the neural network to obtain optimal weights and bias information; reverse the trained neural network to construct a reverse neural network ; Input the node displacement samples in the test set into the reverse neural network, and output the material parameters of the wind turbine. Through the above steps, the material parameters of the wind turbine can be determined accurately and quickly.

Description

High-precision numerical modeling method, server and storage medium for in-service wind turbines

technical field

The present disclosure generally relates to the technical field of wind turbines, and in particular relates to a high-precision numerical modeling method, a server and a storage medium for an in-service wind turbine.

Background technique

Wind turbines generate electricity through wind, which has the advantages of renewable, good environmental benefits, short infrastructure period, flexible installed capacity, etc. In order to make wind energy a reliable energy source, it is necessary to consider the impact of material parameters on wind turbines;

After a wind turbine has been in service for a period of time, material degradation will occur. At this time, the material parameters of the wind turbine are different from the material parameters calibrated at the factory; and the material parameters are difficult to measure directly, and the wind turbine needs to be disassembled and tested. The disassembly process is complicated and It affects the normal use; while indirect methods are used: for example, gradient iteration, which is easy to fall into local optimum for complex problems; and ant colony algorithm, genetic algorithm, which often requires huge workload and low efficiency.

SUMMARY OF THE INVENTION

In view of the above-mentioned defects or deficiencies in the prior art, it is expected to provide a high-precision numerical modeling method, a server and a storage medium that can accurately and quickly determine the in-service wind turbine.

In the first aspect, the present application provides a high-precision numerical modeling method for an in-service wind turbine, including the following steps:

Obtain sample data: obtain the material parameter calibration value of the wind turbine, sample and process the material parameter sample within the error interval of the calibration value; establish a simulation model of the wind turbine, simulate and output the data of each material parameter sample. Corresponding node displacement samples; dividing the sample data into a training set and a test set;

Building a neural network, using the material parameter samples in the training set as the input of the neural network, and using the node displacement samples in the training set as the output of the neural network;

Train the neural network to obtain optimal weights and bias information;

Reverse the trained neural network to construct a reverse neural network;

The node displacement samples in the test set are input into the reverse neural network, and the material parameters of the wind turbine are output.

According to the technical solutions provided in the embodiments of the present application, the material parameters include elastic modulus, shear modulus and Poisson's ratio.

其中，

为第l层神经元的个数，k为设定值；According to the technical solutions provided in the embodiments of the present application, the number of neurons in each layer of the neural network is set,

in,

is the number of neurons in the lth layer, and k is the set value;

Define the loss function, choose the optimizer, and form the neural network.

According to the technical solutions provided by the embodiments of the present application, the neural network is specifically:

为激活函数；

表示第l层的第i个神经元与第l-1层的第j个神经元相连接的权值；

第l层第i个神经元的偏置。in,

is the activation function;

represents the weight of the connection between the i-th neuron in the l-th layer and the j-th neuron in the l-1 layer;

Bias of the ith neuron in the lth layer.

According to the technical solutions provided by the embodiments of the present application, the reverse neural network is specifically:

为所述反向神经网络的权值矩阵，

为所述反向神经网络的偏置矩阵。in,

is the weight matrix of the inverse neural network,

is the bias matrix of the inverse neural network.

According to the technical solutions provided in the embodiments of the present application, the following steps are further included after outputting the material parameters of the wind turbine:

Comparing the material parameters with the material parameter samples in the test set to obtain error data;

Analyze the error data to verify the accuracy of the algorithm.

According to the technical solutions provided by the embodiments of the present application, the method for establishing the simulation model of the wind turbine is specifically:

Drawing a three-dimensional component, correspondingly inputting the material parameter samples into the data of the three-dimensional component, and assembling all the three-dimensional components to form the simulation model.

According to the technical solutions provided by the embodiments of the present application, the method for simulating and outputting the node displacement samples corresponding to each of the material parameter samples is as follows:

Perform dynamic analysis and static analysis on the simulation model;

Determine loads and boundary conditions;

Draw the mesh and output the nodal displacement samples.

Second aspect The present application provides a server, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the above when executing the computer program. Describe the steps of a high-precision numerical modeling method for in-service wind turbines.

The third aspect of the present application provides a computer-readable storage medium, the computer-readable storage medium has a computer program, and when the computer program is executed by a processor, implements the steps of the above-mentioned high-precision numerical modeling method for an in-service wind turbine .

The beneficial effects of the present application are: in the process of acquiring the sample data, the material parameter samples are obtained by sampling and processing within the error interval of the calibration value, and the node displacement samples are obtained by the simulation model; the sampling processing and the simulation processing are used to improve the accuracy of the algorithm; By using the material parameter samples as input and the node displacement samples as output, a neural network is constructed and trained, so that the optimal weights and bias information can be obtained; the reversed neural network can be obtained by inverting the trained neural network. Neural network; at this time, the node displacement samples in the test set are input into the reverse neural network, and the material parameters of the wind turbine can be output.

In the above method, the determination of the material parameters of the wind turbine is realized, and the optimal weight and bias information can be obtained by sampling processing, simulation processing and constructing a neural network, which improves the calculation accuracy; it is similar to genetic algorithm, ant colony algorithm, etc. Compared with the reverse calculation method, the calculation amount is small and the running speed is fast, which improves the calculation efficiency of the material parameters of the wind turbine.

Description of drawings

Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:

Fig. 1 is a flowchart of a high-precision numerical modeling method for an in-service wind turbine provided by the application;

FIG. 2 is a flowchart of algorithm verification after step S500 shown in FIG. 1 .

3 is a schematic structural diagram of the neural network in step S200 shown in FIG. 1;

4 is a schematic structural diagram of an inverse neural network in step S400 shown in FIG. 1;

FIG. 5 is a server provided by the present application.

Detailed ways

The present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the related invention, but not to limit the invention. In addition, it should be noted that, for the convenience of description, only the parts related to the invention are shown in the drawings.

It should be noted that the embodiments in the present application and the features of the embodiments may be combined with each other in the case of no conflict. The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

Example 1

Please refer to FIG. 1 for the flow chart of the high-precision numerical modeling method for in-service wind turbines provided for this application; it includes the following steps:

S100: Obtaining sample data: obtaining the material parameter calibration value of the wind turbine, sampling and processing within the error interval of the calibration value to obtain a material parameter sample; establishing a simulation model of the wind turbine, simulating and outputting each material parameter The node displacement sample corresponding to the sample; the sample data is divided into a training set and a test set;

Specifically, the method of sampling within the error interval of the calibration value adopts Latin hypercube sampling, and the error interval is 90%-110% of the calibration value; specifically, normalize the data after sampling A sample of the material parameters is obtained.

Preferably, the material parameter samples include elastic modulus samples, shear modulus samples and Poisson's ratio samples.

S200: Build a neural network, use the material parameter samples in the training set as the input of the neural network, and use the node displacement samples in the training set as the output of the neural network;

S300: Train the neural network to obtain optimal weights and bias information;

Specifically, the method for training the neural network and obtaining the optimal weight and bias information is as follows: selecting 0.01 or 0.001 for the initial learning rate of the neural network; after convergence, obtaining the optimal weight and bias information;

S400: Reverse the trained neural network to construct a reverse neural network;

S500: Input the node displacement samples in the test set into the inverse neural network, and output the material parameters of the wind turbine.

Working principle: In the above process of obtaining sample data, the material parameter samples are obtained by sampling and processing within the error interval of the calibration value, and the node displacement samples are obtained through the simulation model; the sampling processing and simulation processing are used to improve the accuracy of the algorithm;

By using the material parameter samples as input and the node displacement samples as output, a neural network is constructed and trained, so that the optimal weights and bias information can be obtained; the reversed neural network can be obtained by inverting the trained neural network. Neural network; at this time, the node displacement samples in the test set are input into the reverse neural network, and the material parameters of the wind turbine can be output.

In the above method, the determination of the material parameters of the wind turbine is realized, and the optimal weight and bias information can be obtained by sampling processing, simulation processing and constructing a neural network, which improves the calculation accuracy; it is similar to genetic algorithm, ant colony algorithm, etc. Compared with the reverse calculation method, the calculation amount is small and the running speed is fast, which improves the calculation efficiency of the material parameters of the wind turbine.

Preferably, the material parameters include elastic modulus, shear modulus and Poisson's ratio.

Preferably, the method for building the neural network is as follows:

其中，

为第l层神经元的个数，k为设定值；Set the number of neurons in each layer of the neural network,

in,

is the number of neurons in the lth layer, and k is the set value;

Define the loss function, choose the optimizer, and form the neural network.

The specific loss can use the mean square error:

Wherein, MSE (Mean Square Error) is the average value of the sum of squares of the difference between the estimated value and the actual value, the estimated value is the value output by the neural network, and the actual value is the output value in the training sample.

Specifically, the optimizer can use Adam;

Specifically, k is a set constant. As shown in the above formula and FIG. 3 , the neural network is a forward horn network, that is, the number of neurons in each layer in the neural network is an arithmetic progression, which ensures the accuracy of the neural network.

Since the weights and biases of the trained forward speaker network reflect the relationship between the variables in the inverse problem, the trained forward speaker network is reversed, and the weights and biases of the reverse speaker network are obtained by using explicit formulas. After setting, the reverse neural network can be obtained, as shown in Figure 4.

Preferably, the neural network is specifically:

为激活函数；

表示第l层的第i个神经元与第l-1层的第j个神经元相连接的权值；

第l层第i个神经元的偏置。in,

is the activation function;

represents the weight of the connection between the i-th neuron in the l-th layer and the j-th neuron in the l-1 layer;

Bias of the ith neuron in the lth layer.

Preferably, the reverse neural network is specifically:

为所述反向神经网络的权值矩阵，

为所述反向神经网络的偏置矩阵。in,

is the weight matrix of the inverse neural network,

is the bias matrix of the inverse neural network.

In order to facilitate the description of the technical solutions provided in this application, material parameters (elastic modulus, shear modulus and Poisson's ratio) are taken as an example.

S100: Obtain sample data: obtain the calibration values of elastic modulus, shear modulus and Poisson's ratio, respectively, sample and normalize them within the error interval (90%-110% of the calibration value) to obtain material parameter samples ; Establish a simulation model of the wind turbine, simulate and output the elastic modulus, shear modulus and node displacement samples corresponding to Poisson's ratio; divide the sample data into a training set and a test set.

S200: Build a neural network, use the material parameter samples in the training set as the input of the neural network, and use the node displacement samples in the training set as the output of the neural network:

Its matrix form is:

其与第二层的第i个神经元的值

的关系式可表示为：The three material parameters are input into the neural network, that is, the values of all neurons in the input layer (the first layer) can be expressed as

its value with the ith neuron of the second layer

The relation can be expressed as:

Through the above formula, the value of each neuron in the second layer can be obtained, and by analogy, the value of each neuron in the output layer can be obtained, that is, the nodes corresponding to the elastic modulus, shear modulus and Poisson’s ratio displacement value.

S300: Train the neural network to obtain optimal weights and bias information;

S400: Reverse the trained neural network to construct a reverse neural network;

Reverse the trained neural network:

Expand the above formula in parentheses and use the display formula to express as:

和偏置矩阵

进而得到反向神经网络。The weight matrix of the reverse neural network can be obtained by the above formula

and the bias matrix

And then get the reverse neural network.

S500: Input the node displacement samples in the test set into the inverse neural network, and output the material parameters (elastic modulus, shear modulus, Poisson's ratio) of the wind turbine.

Preferably, the following steps are further included after outputting the material parameters of the wind turbine:

S601: Compare the output material parameters with the material parameter samples in the test set to obtain error data;

S602: Analyze the error data to verify the accuracy of the algorithm.

Preferably, the method for establishing the simulation model of the wind turbine is as follows:

Drawing a three-dimensional component, correspondingly inputting the material parameter samples into the data of the three-dimensional component, and assembling all the three-dimensional components to form the simulation model.

Preferably, the method for simulating and outputting the nodal displacement samples corresponding to each of the material parameter samples is as follows:

Perform dynamic analysis and static analysis on the simulation model;

Determine loads and boundary conditions;

Draw the mesh and output the nodal displacement samples.

Example 2

Please refer to FIG. 5 for a schematic block diagram of a computer system 700 of a server or server provided by this application, including a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor When the computer program is executed, the steps of the above-mentioned high-precision numerical modeling method for an in-service wind turbine are realized.

As shown in FIG. 5, the computer system 700 includes a central processing unit (CPU) 701, which can be loaded into a random access memory (RAM) 703 according to a program stored in a read only memory (ROM) 702 or from a storage portion program to execute various appropriate actions and processes. In the RAM 703, various programs and data necessary for system operation are also stored. The CPU 701 , the ROM 702 , and the RAM 703 are connected to each other through a bus 704 . An input/output (I/O) interface 705 is also connected to bus 704 .

The following components are connected to the I/O interface 705: an input section 706 including a keyboard, a mouse, etc.; an output section including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 708 including a hard disk, etc.; And a communication section 709 including network interface cards such as LAN cards, modems, and the like. The communication section 709 performs communication processing via a network such as the Internet. Drives are also connected to I/O interface 705 as needed. A removable medium 711, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is mounted on the drive 710 as needed so that a computer program read therefrom is installed into the storage section 708 as needed.

In particular, according to an embodiment of the present invention, the process described above with reference to flowchart 1 may be implemented as a computer software program. For example, Embodiment 1 of the present invention includes a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for executing the method shown in the flowchart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication portion, and/or installed from a removable medium. When the computer program is executed by the central processing unit (CPU) 601, the above-described functions defined in the system of the present application are executed.

Example 3

The present application also provides a computer-readable storage medium, the computer-readable storage medium has a computer program, and when the computer program is executed by a processor, the above-mentioned steps of determining a high-precision numerical modeling method for an in-service wind turbine are implemented. .

It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the above two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or a combination of any of the above. More specific examples of computer readable storage media may include, but are not limited to, electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), erasable Programmable read only memory (EPROM or flash memory), fiber optics, portable compact disk read only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing. In the present invention, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present invention, however, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code therein. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device . Program code embodied on a computer readable medium may be transmitted using any suitable medium including, but not limited to, wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code that contains one or more logical functions for implementing the specified functions executable instructions. It should also be noted that, in some alternative implementations, the functions noted in the blocks may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It is also noted that each block of the block diagrams or flowchart illustrations, and combinations of blocks in the block diagrams or flowchart illustrations, can be implemented in special purpose hardware-based systems that perform the specified functions or operations, or can be implemented using A combination of dedicated hardware and computer instructions is implemented.

The units involved in the embodiments of the present invention may be implemented in a software manner, or may be implemented in a hardware manner, and the described units may also be provided in a processor. Among them, the names of these units do not constitute a limitation on the unit itself under certain circumstances. The described unit or module can also be set in the processor, for example, it can be described as: a processor includes an acquisition module, an initialization module, and a data processing module.

The names of these units or modules do not constitute a limitation on the units or modules themselves, for example, the acquisition module can also be described as an "acquisition module for acquiring sample data".

As another aspect, the present application also provides a computer-readable medium. The computer-readable medium may be included in the electronic device described in the above embodiments; it may also exist alone without being assembled into the electronic device. middle. The computer-readable medium carries one or more programs, and when the one or more programs are executed by an electronic device, the electronic device implements the high-precision numerical modeling method for an in-service wind turbine in the above-mentioned embodiment.

For example, the electronic device may implement as shown in Figure 1:

S100: Obtaining sample data: obtaining the material parameter calibration value of the wind turbine, sampling and processing within the error interval of the calibration value to obtain a material parameter sample; establishing a simulation model of the wind turbine, simulating and outputting each material parameter The node displacement sample corresponding to the sample; the sample data is divided into a training set and a test set;

S200: Build a neural network, use the material parameter samples in the training set as the input of the neural network, and use the node displacement samples in the training set as the output of the neural network;

S300: Train the neural network to obtain optimal weights and bias information;

S400: Reverse the trained neural network to construct a reverse neural network;

S500: Input the node displacement samples in the test set into the inverse neural network, and output the material parameters of the wind turbine.

The above description is only a preferred embodiment of the present application and an illustration of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to the technical solution formed by the specific combination of the above-mentioned technical features, and should also cover the above-mentioned technical features without departing from the inventive concept. Other technical solutions formed by any combination of its equivalent features. For example, a technical solution is formed by replacing the above-mentioned features with the technical features disclosed in this application (but not limited to) with similar functions.

Claims (10)

1. A high-precision numerical modeling method for an in-service wind turbine generator is characterized by comprising the following steps: the method comprises the following steps:

acquiring sample data: acquiring a material parameter calibration value of a wind turbine generator, and sampling and processing in an error interval of the calibration value to obtain a material parameter sample; establishing a simulation model of the wind turbine generator, simulating and outputting a node displacement sample corresponding to each material parameter sample; dividing the sample data into a training set and a testing set;

building a neural network, taking the material parameter sample in the training set as the input of the neural network, and taking the node displacement sample in the training set as the output of the neural network;

training the neural network to obtain optimal weight and bias information;

reversing the trained neural network to construct a reverse neural network;

and inputting the node displacement samples concentrated in the test into the reverse neural network, and outputting the material parameters of the wind turbine generator.

2. The in-service wind turbine generator high-precision numerical modeling method according to claim 1, characterized in that: the material parameters include modulus of elasticity, shear modulus, and poisson's ratio.

3. The in-service wind turbine generator high-precision numerical modeling method according to claim 1, characterized in that: the method for building the neural network specifically comprises the following steps:

setting the number of neurons in each layer of the neural network,

wherein,

the number of the first layer neurons is defined, and k is a set value;

defining a loss function, and selecting an optimizer to form the neural network.

4. The in-service wind turbine generator high-precision numerical modeling method according to claim 3, characterized in that: the neural network specifically comprises:

wherein,

is an activation function;

representing the weight value of the ith neuron of the l layer connected with the jth neuron of the l-1 layer;

bias of ith neuron of l layer.

5. The in-service wind turbine generator high-precision numerical modeling method according to claim 4, characterized in that: the reverse neural network specifically comprises:

wherein,

is a weight matrix of the inverse neural network,

is a bias matrix of the inverse neural network.

6. The in-service wind turbine generator high-precision numerical modeling method according to claim 1, characterized in that: the method also comprises the following steps after the material parameters of the wind turbine generator are output:

comparing the material parameter with the material parameter sample in the test set to obtain error data;

and analyzing the error data and verifying the accuracy of the algorithm.

7. The in-service wind turbine generator high-precision numerical modeling method according to claim 1, characterized in that: the method for establishing the simulation model of the wind turbine generator specifically comprises the following steps:

drawing a three-dimensional part, correspondingly inputting the material parameter sample into the data of the three-dimensional part, and assembling all the three-dimensional parts to form the simulation model.

8. The in-service wind turbine generator high-precision numerical modeling method according to claim 1, characterized in that: the method for simulating and outputting the node displacement sample corresponding to each material parameter sample specifically comprises the following steps:

performing dynamic analysis and static analysis on the simulation model;

determining a load and a boundary condition;

and drawing a grid and outputting the node displacement sample.

9. A server comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: the processor, when executing the computer program, implements the steps of the high-precision numerical modeling method for an active wind turbine according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program, the computer-readable storage medium characterized by: the computer program, when being executed by a processor, implements the steps of the method for high-precision numerical modeling of an active wind turbine according to any of claims 1 to 8.

Images