Wind turbine generator equipment state evaluation method and system

CN118940941APending Publication Date: 2024-11-12HUANENG NEW ENERGY CO LTD SHANXI BRANCH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410975099.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-19
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The equipment status monitoring of traditional wind turbine units relies on regular physical inspections, is inefficient and cannot predict potential failures, and cannot achieve accurate equipment status evaluation and prediction.

Method used

By building an LSTM-MLP model based on historical equipment operating status data, training and optimization of wind turbine equipment status evaluation model is carried out to achieve real-time monitoring and fault prediction.

Benefits of technology

It improves the operating efficiency of wind turbines, reduces maintenance costs, and extends the service life of the equipment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118940941A_ABST
    Figure CN118940941A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of wind driven generators, in particular to a wind turbine generator equipment state evaluation method and system, and the method comprises the steps: obtaining historical equipment operation state data of a wind turbine generator; dividing historical equipment operation state data of the wind turbine generator into a training set and a test set; constructing a wind turbine generator equipment state evaluation model, dividing the training set into time series data and static data, processing the time series data through an LSTM layer in the wind turbine generator equipment state evaluation model, processing the static data through an MLP layer in the wind turbine generator equipment state evaluation model, and merging an output result of the LSTM layer and an output result of the MLP layer to obtain a wind turbine generator equipment state evaluation result. And the merging result is evaluated through the test set until the training of the wind turbine generator equipment state evaluation model is completed. And executing the real-time equipment state evaluation of the wind turbine generator based on the trained wind turbine generator equipment state evaluation model.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbines, and in particular to a method and system for evaluating the equipment status of a wind turbine set. Background Art

[0002] With the vigorous development of sustainable energy, the size and complexity of wind turbines are also increasing. In order to ensure the efficient and reliable operation of wind turbines, it is particularly important to accurately evaluate their equipment status. At present, the equipment status monitoring and evaluation of traditional wind turbines mainly rely on regular physical inspections by staff and preset maintenance schedules, which is not only inefficient but also unable to predict potential failures. Based on this, in response to the above problems, we designed a method and system for evaluating the equipment status of wind turbines. Summary of the invention

[0003] The purpose of the present invention is to provide a method and system for evaluating the equipment status of a wind turbine generator set, which constructs a wind turbine generator set equipment status evaluation model through historical equipment operating status data and optimizes the model during the training process, thereby not only realizing real-time monitoring and prediction of the equipment status of the wind turbine generator set, but also effectively improving the operating efficiency of the wind turbine generator set, reducing maintenance costs, and extending the service life of the equipment.

[0004] The embodiments of the present invention are implemented by the following technical solutions:

[0005] A method for evaluating the status of a wind turbine generator system, the method comprising the following steps:

[0006] Obtain historical equipment operating status data of wind turbines;

[0007] Divide the historical equipment operation status data of the wind turbine into a training set and a test set;

[0008] Construct a wind turbine equipment status assessment model, divide the training set into time series data and static data, process the time series data through the LSTM layer in the wind turbine equipment status assessment model, process the static data through the MLP layer in the wind turbine equipment status assessment model, merge the LSTM layer output results and the MLP layer output results, and evaluate the merged results through the test set until the training of the wind turbine equipment status assessment model is completed;

[0009] The real-time equipment status of the wind turbine is evaluated based on the trained wind turbine equipment status evaluation model.

[0010] Optionally, the wind turbine equipment status assessment model is composed of a combination of an LSTM model and an MLP model, the LSTM model constitutes the LSTM layer, and the MLP model constitutes the MLP layer.

[0011] Optionally, the LSTM layer is constructed by a forget gate, an input gate, an update cell state, and an output gate, and its calculation formula is as follows:

[0012] f t =σ(W f ·[h t-1 ,x t ]+b f )

[0013] i t =σ(W i ·[h t-1 ,x t ]+b i )

[0014]

[0015] o t =σ(W o ·[h t-1 ,x t ]+b o )

[0016] h t =o t *tanh(C t )

[0017] Among them, f t is the activation vector of the forget gate, i t is the activation vector of the input gate, is the candidate vector of the cell state, C t is the cell state at the current time step, o t is the activation vector of the output gate, h t is the output of the current time step, W and b represent the weight matrix and bias vector respectively, σ represents the sigmoid function, tanh represents the hyperbolic tangent function, [h t-1 ,x t ] indicates the connection between the output of the previous time step and the current input.

[0018] Optionally, the calculation formula of the MLP layer is:

[0019] a (l) =f(W (l) ·a (l-1) +b (l) )

[0020] Among them, a (l) is the activation output of layer l, W (l) and b (l) They represent the weight and bias of the lth layer respectively, and f is the activation function.

[0021] Optionally, the merging result is evaluated by the test set, which specifically relies on calculating the loss function of the test set and the training set, calculating the gradient of the wind turbine equipment state assessment model parameters according to the loss function, and iterating the wind turbine equipment state assessment model parameters by the Adam algorithm. The loss function calculation formula is:

[0022]

[0023] Among them, L is the loss function, y is the loss value, is the predicted value, N is the number of samples, M is the number of categories, y ij is the indicator variable for whether the i-th sample belongs to the j-th category, The model predicts the probability that the i-th sample belongs to the j-th class.

[0024] Optionally, the wind turbine equipment state assessment model parameters are iterated by the Adam algorithm, and the specific process is as follows:

[0025] Initialize the parameters of the wind turbine equipment state assessment model, where the parameters of the wind turbine equipment state assessment model include: weights and biases of the LSTM and MLP layers, first-order moment estimates of the gradient, second-order moment estimates of the gradient, and the number of iterations. The first-order moment estimates of the gradient and the second-order moment estimates of the gradient are initialized to 0;

[0026] In each iteration, the gradient of the parameters of the wind turbine equipment status assessment model is calculated through the loss function;

[0027] The first-order moment estimate and the second-order moment estimate of the gradient are updated by the following formula:

[0028] m t =β1m t-1 +(1-β1)g t

[0029]

[0030] Among them, m t is the first-order moment estimate of the gradient, v t is the second-order moment estimate of the gradient, g t is the gradient, β1 and β2 are the exponential decay rates of the first-order moment estimate and the second-order moment estimate respectively;

[0031] Perform bias correction on the first-order and second-order moment estimates of the gradient:

[0032]

[0033] in, is the bias-corrected first-order moment estimate, is the bias-corrected second-order moment estimate, are the exponential decay rates of the first-order moment estimate and the second-order moment estimate after t iterations, respectively;

[0034] Update wind turbine equipment condition assessment model parameters:

[0035]

[0036] Among them, θ is the wind turbine equipment condition assessment model parameter, θ t is the parameter of the wind turbine equipment status assessment model after the tth iteration, α is the learning rate, and ò is a constant.

[0037] A wind turbine equipment status assessment system, comprising:

[0038] A data acquisition module, which acquires historical equipment operation status data of wind turbines;

[0039] A data processing module divides the historical equipment operation status data of the wind turbine into a training set and a test set;

[0040] The model training module builds a wind turbine equipment status assessment model, divides the training set into time series data and static data, processes the time series data through the LSTM layer in the wind turbine equipment status assessment model, processes the static data through the MLP layer in the wind turbine equipment status assessment model, merges the output results of the LSTM layer and the output results of the MLP layer, and evaluates the merged results through the test set until the training of the wind turbine equipment status assessment model is completed;

[0041] The state assessment module performs the assessment of the real-time equipment state of the wind turbine based on the trained wind turbine equipment state assessment model.

[0042] The technical solution of the embodiment of the present invention has at least the following advantages and beneficial effects:

[0043] The embodiment of the present invention constructs a wind turbine equipment status assessment model through historical equipment operating status data and optimizes the model during the training process, which not only realizes real-time monitoring and prediction of the wind turbine equipment status, but also effectively improves the operating efficiency of the wind turbine, reduces maintenance costs, and extends the service life of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 A schematic flow chart of a method for evaluating the status of a wind turbine generator system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0046] Example 1

[0047] like Figure 1 As shown, a method for evaluating the status of a wind turbine equipment is provided, the method comprising the steps of:

[0048] Obtain historical equipment operating status data of wind turbines;

[0049] Divide the historical equipment operation status data of the wind turbine into a training set and a test set;

[0050] Construct a wind turbine equipment status assessment model, divide the training set into time series data and static data, process the time series data through the LSTM layer in the wind turbine equipment status assessment model, process the static data through the MLP layer in the wind turbine equipment status assessment model, merge the LSTM layer output results and the MLP layer output results, and evaluate the merged results through the test set until the training of the wind turbine equipment status assessment model is completed;

[0051] The real-time equipment status of the wind turbine is evaluated based on the trained wind turbine equipment status evaluation model.

[0052] In this embodiment, the wind turbine equipment status assessment model is composed of a LSTM model and an MLP model. The LSTM model constitutes the LSTM layer, and the MLP model constitutes the MLP layer.

[0053] The LSTM layer is constructed by a forget gate, an input gate, an update cell state, and an output gate, and its calculation formula is as follows:

[0054] f t =σ(W f ·[h t-1 ,x t ]+b f )

[0055] i t =σ(W i ·[h t-1 ,x t ]+b i )

[0056]

[0057] o t =σ(W o ·[h t-1 ,x t ]+b o )

[0058] h t =o t *tanh(C t )

[0059] Among them, f t is the activation vector of the forget gate, i t is the activation vector of the input gate, is the candidate vector of the cell state, C t is the cell state at the current time step, o t is the activation vector of the output gate, h t is the output of the current time step, W and b represent the weight matrix and bias vector respectively, σ represents the sigmoid function, tanh represents the hyperbolic tangent function, [h t-1 ,x t ] indicates the connection between the output of the previous time step and the current input.

[0060] The calculation formula of the MLP layer is:

[0061] a (l) =f(W (l) ·a (l-1) +b (l) )

[0062] Among them, a (l) is the activation output of layer l, W (l) and b (l) They represent the weight and bias of the lth layer respectively, and f is the activation function.

[0063] In this embodiment, the evaluation of the merged result by the test set specifically relies on calculating the loss function of the test set and the training set, calculating the gradient of the wind turbine equipment state assessment model parameters according to the loss function, and iterating the wind turbine equipment state assessment model parameters by the Adam algorithm. The loss function calculation formula is:

[0064]

[0065] Among them, L is the loss function, y is the loss value, is the predicted value, N is the number of samples, M is the number of categories, y ij is the indicator variable for whether the i-th sample belongs to the j-th category, The model predicts the probability that the i-th sample belongs to the j-th class.

[0066] Furthermore, the parameters of the wind turbine equipment status assessment model are iterated by the Adam algorithm, and the specific process is as follows:

[0067] Initialize the parameters of the wind turbine equipment state assessment model, where the parameters of the wind turbine equipment state assessment model include: weights and biases of the LSTM and MLP layers, first-order moment estimates of the gradient, second-order moment estimates of the gradient, and the number of iterations. The first-order moment estimates of the gradient and the second-order moment estimates of the gradient are initialized to 0;

[0068] In each iteration, the gradient of the parameters of the wind turbine equipment status assessment model is calculated through the loss function;

[0069] The first-order moment estimate and the second-order moment estimate of the gradient are updated by the following formula:

[0070] m t =β1m t-1 +(1-β1)g t

[0071]

[0072] Among them, m t is the first-order moment estimate of the gradient, v t is the second-order moment estimate of the gradient, g t is the gradient, β1 and β2 are the exponential decay rates of the first-order moment estimate and the second-order moment estimate, respectively, and their values ​​are close to 0;

[0073] Perform bias correction on the first-order and second-order moment estimates of the gradient:

[0074]

[0075] in, is the bias-corrected first-order moment estimate, is the bias-corrected second-order moment estimate, are the exponential decay rates of the first-order moment estimate and the second-order moment estimate after t iterations, respectively;

[0076] Update wind turbine equipment condition assessment model parameters:

[0077]

[0078] Among them, θ is the wind turbine equipment status assessment model parameter, including all weights and biases that need to be learned through training, θ t is the parameter of the wind turbine equipment status assessment model after the tth iteration, α is the learning rate, and ò is a constant.

[0079] In the specific application of this embodiment, through the above steps, a LSTM-MLP hybrid model capable of evaluating the equipment status of a wind turbine can be trained. In each iteration, the parameters of the model will be continuously updated through the Adam optimization algorithm to minimize the loss function until the performance of the model reaches a predetermined standard or a predetermined number of iterations is completed.

[0080] Example 2

[0081] A wind turbine equipment status assessment system, comprising:

[0082] A data acquisition module, which acquires historical equipment operation status data of wind turbines;

[0083] A data processing module divides the historical equipment operation status data of the wind turbine into a training set and a test set;

[0084] The model training module builds a wind turbine equipment status assessment model, divides the training set into time series data and static data, processes the time series data through the LSTM layer in the wind turbine equipment status assessment model, processes the static data through the MLP layer in the wind turbine equipment status assessment model, merges the output results of the LSTM layer and the output results of the MLP layer, and evaluates the merged results through the test set until the training of the wind turbine equipment status assessment model is completed;

[0085] The state assessment module performs the assessment of the real-time equipment state of the wind turbine based on the trained wind turbine equipment state assessment model.

[0086] The wind turbine equipment status assessment system provided in this embodiment and the wind turbine equipment status assessment method provided in the above embodiment are based on the same inventive concept. For more specific working principles of each module in the embodiment of the present invention, please refer to the above embodiment and will not be repeated in the embodiment of the present invention.

[0087] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for evaluating the status of a wind turbine generator system, characterized in that: The steps of the method include: Obtain historical equipment operating status data of wind turbines; Divide the historical equipment operation status data of the wind turbine into a training set and a test set; Construct a wind turbine equipment status assessment model, divide the training set into time series data and static data, process the time series data through the LSTM layer in the wind turbine equipment status assessment model, process the static data through the MLP layer in the wind turbine equipment status assessment model, merge the LSTM layer output results and the MLP layer output results, and evaluate the merged results through the test set until the training of the wind turbine equipment status assessment model is completed; The real-time equipment status of the wind turbine is evaluated based on the trained wind turbine equipment status evaluation model.

2. The method for evaluating the status of a wind turbine generator system according to claim 1, characterized in that: The wind turbine equipment status assessment model is composed of a LSTM model and an MLP model. The LSTM model constitutes the LSTM layer, and the MLP model constitutes the MLP layer.

3. The method for evaluating the status of a wind turbine generator system according to claim 2, characterized in that: The LSTM layer is constructed by a forget gate, an input gate, an update cell state, and an output gate, and its calculation formula is as follows: f t =σ(W f ·[h t-1 ,x t ]+b f ) i t =σ(W i ·[h t-1 ,x t ]+b i ) the t =σ(W o ·[h t-1 ,x t ]+b o ) ht=o t *tanh(C t ) Among them, f t is the activation vector of the forget gate, i t is the activation vector of the input gate, is the candidate vector of cell state, C t is the cell state at the current time step, o t is the activation vector of the output gate, h t is the output of the current time step, W and b represent the weight matrix and bias vector respectively, σ represents the sigmoid function, tanh represents the hyperbolic tangent function, [h t-1 ,x t ] indicates the connection between the output of the previous time step and the current input.

4. The method for evaluating the status of a wind turbine generator system according to claim 2, characterized in that: The calculation formula of the MLP layer is: a (l) =f(W (l) ·a (l-1) +b (l) ) Among them, a (l) is the activation output of layer l, W (l) and b (l) They represent the weight and bias of the lth layer respectively, and f is the activation function.

5. The method for evaluating the status of a wind turbine equipment according to claim 1, characterized in that: The evaluation of the merged result by the test set specifically relies on calculating the loss function of the test set and the training set, calculating the gradient of the wind turbine equipment state evaluation model parameters according to the loss function, and iterating the wind turbine equipment state evaluation model parameters by the Adam algorithm. The loss function calculation formula is: Among them, L is the loss function, y is the loss value, is the predicted value, N is the number of samples, M is the number of categories, y ij is the indicator variable for whether the i-th sample belongs to the j-th category, The model predicts the probability that the i-th sample belongs to the j-th class.

6. The method for evaluating the status of a wind turbine generator system according to claim 5, characterized in that: The Adam algorithm is used to iterate the parameters of the wind turbine equipment status assessment model, and the specific process is as follows: Initialize the parameters of the wind turbine equipment state assessment model, where the parameters of the wind turbine equipment state assessment model include: weights and biases of the LSTM and MLP layers, first-order moment estimates of the gradient, second-order moment estimates of the gradient, and the number of iterations. The first-order moment estimates of the gradient and the second-order moment estimates of the gradient are initialized to 0; In each iteration, the gradient of the parameters of the wind turbine equipment status assessment model is calculated through the loss function; The first-order moment estimate and the second-order moment estimate of the gradient are updated by the following formula: m t =β1m t-1 +(1-β1)g t Among them, m t is the first-order moment estimate of the gradient, v t is the second-order moment estimate of the gradient, g t is the gradient, β1 and β2 are the exponential decay rates of the first-order moment estimate and the second-order moment estimate respectively; Perform bias correction on the first-order and second-order moment estimates of the gradient: in, is the bias-corrected first-order moment estimate, is the bias-corrected second-order moment estimate, are the exponential decay rates of the first-order moment estimate and the second-order moment estimate after t iterations, respectively; Update the parameters of the wind turbine equipment condition assessment model: Among them, θ is the wind turbine equipment condition assessment model parameter, θ t is the parameter of the wind turbine equipment status assessment model after the tth iteration, α is the learning rate, and ò is a constant.

7. A wind turbine equipment status assessment system, characterized in that: include: A data acquisition module is used to acquire historical equipment operation status data of wind turbines; A data processing module divides the historical equipment operation status data of the wind turbine into a training set and a test set; The model training module builds a wind turbine equipment status assessment model, divides the training set into time series data and static data, processes the time series data through the LSTM layer in the wind turbine equipment status assessment model, processes the static data through the MLP layer in the wind turbine equipment status assessment model, merges the output results of the LSTM layer and the output results of the MLP layer, and evaluates the merged results through the test set until the training of the wind turbine equipment status assessment model is completed; The state assessment module performs the assessment of the real-time equipment state of the wind turbine based on the trained wind turbine equipment state assessment model.