A global strain reconstruction method and system for wind turbine composite blades

Abstract

The present invention discloses a method and system for reconstructing the global strain of composite wind turbine blades. The method is performed in the following steps: Step S1, Step S2, Step S3, Step S4, Step S5, and Step S6. The present invention also relates to a system for reconstructing the global strain of composite wind turbine blades. The method and system proposed in the present invention can reconstruct and visualize the global strain of composite wind turbine blades, providing an efficient and feasible solution for real-time monitoring of global strain in blades.

Description

Technical Field

[0001] The present invention belongs to the field of wind power technology and relates to a strain reconstruction method and system, in particular to a global strain reconstruction method and system for a wind turbine composite material blade. Background Art

[0002] Blades are key components in wind turbines for generating wind energy. Large, flexible wind turbine blades undergo elastic deformation under the influence of inertia, centrifugal forces, and aerodynamic loads, and this combined effect can even lead to blade damage. Global, real-time monitoring of blade strain during actual wind turbine operation, providing comprehensive insight into the specific location and magnitude of early blade damage, is crucial for ensuring wind power system safety.

[0003] Existing wind turbine blade monitoring usually installs acceleration sensors, resistance strain sensors, etc. at the root of the blade, which cannot obtain the global strain state of the blade; if electrical sensors are set at the middle and tip of the blade, it will increase costs and pose safety risks.

[0004] How to globally monitor and visualize blade strain under the actual operation of wind turbines is an urgent problem that needs to be solved. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for global strain reconstruction of composite blades of wind turbines, thereby realizing global real-time monitoring and visualization of blade strain under the actual operation state of the wind turbine.

[0006] To achieve the above object, the present invention provides a method for global strain reconstruction of a wind turbine composite blade, the method comprising the following steps:

[0007] Establish a blade finite element model that is consistent with the structural characteristic parameters of the actual wind turbine composite blade;

[0008] Determine u blade root strain test positions, where u is an integer greater than 2;

[0009] Establish blade load inversion model;

[0010] The establishment of the blade load inversion model comprises the following steps:

[0011] Set n combinations of operating parameters, where n is an integer greater than 2. The operating parameters include wind speed, wind direction, and turbulence intensity. Based on the blade element momentum theory, divide the blade into m blade elements, where m is an integer greater than 2. Obtain the lift L of the jth blade element under the i-th operating parameter combination. ij , resistance D ij , torque M ij , i=1,2,…n, j=1,2,…m;

[0012] Apply the lift force L in the blade finite element model ij , resistance D ij , torque M ij , obtain the strain ε of the qth blade root strain test position under the i-th working condition parameter combination iq , q=1,2,…u, obtain the blade global strain data set E under the n operating parameter combinations;

[0013] With the lift L ij , resistance D ij , torque M ij and the strain ε iq Establish load inversion training sample data set T;

[0014] Based on the load inversion training sample data set T, the strain ε iq As input, take the lift force L ij , resistance D ij , torque M ij As output, the blade load inversion model is obtained based on the radial basis function neural network with output feedback;

[0015] Performing a strain dynamic test on the actual wind turbine composite blade to obtain strain time series spectra at the u blade root strain test positions, and obtaining an equivalent load time series spectrum of the actual wind turbine composite blade based on the blade load inversion model;

[0016] Establish a reduced-order prediction model for blade load-strain field;

[0017] Taking the equivalent load time series spectrum of the actual wind turbine composite blade as input, a blade global strain cloud diagram is obtained according to the blade load-strain field reduced-order prediction model.

[0018] The steps of establishing a blade finite element model consistent with the structural characteristic parameters of an actual wind turbine composite blade include:

[0019] According to the size, load, material, process level and installation method of the actual wind turbine composite blade, the initial layup parameters of the finite element blade are determined and the initial blade finite element model is established;

[0020] Obtaining actual wind turbine composite blade structural characteristic parameters through the initial blade finite element model; the actual wind turbine composite blade structural characteristic parameters include stiffness distribution characteristics, mass characteristics and modal parameters;

[0021] Selecting the initial layup parameters of the finite element blade as optimization design variables, taking the consistency of the structural characteristic parameters of the finite element blade with the structural characteristic parameters of the actual wind turbine composite blade as the optimization goal, and combining the optimization algorithm to solve and obtain the optimized layup parameters of the finite element blade;

[0022] According to the optimized ply parameters of the finite element blade, the initial blade finite element model is adjusted and modified to obtain an optimized blade finite element model, and the optimized blade finite element model is verified to obtain a blade finite element model that is consistent with the structural characteristic parameters of an actual wind turbine composite blade.

[0023] The steps of establishing the blade load-strain field reduced-order prediction model include:

[0024] The data of the blade global strain data set E is written into a column matrix and combined in chronological order to obtain a strain snapshot matrix A;

[0025] The strain snapshot matrix A is modally decomposed by using the intrinsic orthogonal decomposition method to obtain the strain snapshot matrix B of the dominant mode;

[0026] Obtaining, according to the blade load inversion model, equivalent load time series spectra of the n operating parameter combinations;

[0027] Taking the equivalent load time series spectra of the n working condition parameter combinations as input and the strain snapshot matrix B as output, a training database Z is constructed;

[0028] A radial basis function neural network is adopted to construct a blade load-strain field reduced-order prediction model based on the training database Z.

[0029] The present invention also provides a global strain reconstruction system for a wind turbine composite blade, the system comprising:

[0030] Module M1: used to establish a blade finite element model that is consistent with the structural characteristic parameters of the actual wind turbine composite blade;

[0031] Module M2: used to determine u blade root strain test positions, where u is an integer greater than 2;

[0032] Module M3: used to establish blade load inversion model;

[0033] The module M3 includes the following modules:

[0034] Module M31: used to set n combinations of operating parameters, where n is an integer greater than 2. The operating parameters include wind speed, wind direction, and turbulence intensity. Based on the blade element momentum theory, the blade is divided into m blade elements, where m is an integer greater than 2. The lift L of the jth blade element under the i-th operating parameter combination is obtained.ij , resistance D ij , torque M ij , i=1,2,…n, j=1,2,…m;

[0035] Module M32: used to apply the lift force L in the blade finite element model ij , resistance D ij , torque M ij , obtain the strain ε of the qth blade root strain test position under the i-th working condition parameter combination iq , q=1,2,…u, obtain the blade global strain data set E under the n operating parameter combinations;

[0036] Module M33: used to lift the ij , resistance D ij , torque M ij and the strain ε iq Establish load inversion training sample data set T;

[0037] Module M34: used for inverting the training sample data set T based on the load, with the strain ε iq As input, take the lift force L ij , resistance D ij , torque M ij As output, the blade load inversion model is obtained based on the radial basis function neural network with output feedback;

[0038] Module M4: for performing a strain dynamic test on the actual wind turbine composite blade to obtain a strain time series spectrum of the u blade root strain test positions, and based on the blade load inversion model, obtain an equivalent load time series spectrum of the actual wind turbine composite blade;

[0039] Module M5: used to establish a blade load-strain field reduced-order prediction model;

[0040] Module M6: used to obtain a blade global strain cloud diagram based on the blade load-strain field reduced-order prediction model using the equivalent load time series spectrum of the actual wind turbine composite blade as input.

[0041] The module M1 includes the following modules:

[0042] Module M11: used to determine the initial layup parameters of the finite element blade and establish the initial blade finite element model based on the size, load, material, process level, and installation method of the actual wind turbine composite blade;

[0043] Module M12: used to obtain the structural characteristic parameters of the actual wind turbine composite blade using the initial blade finite element model; the structural characteristic parameters of the actual wind turbine composite blade include stiffness distribution characteristics, mass characteristics and modal parameters;

[0044] Module M13: for selecting the initial layup parameters of the finite element blade as optimization design variables, taking the consistency of the structural characteristic parameters of the finite element blade with the structural characteristic parameters of the actual wind turbine composite blade as the optimization goal, and combining the optimization algorithm to solve and obtain the optimized layup parameters of the finite element blade;

[0045] Module M14: used to adjust and modify the initial blade finite element model according to the optimized ply parameters of the finite element blade to obtain an optimized blade finite element model, verify the optimized blade finite element model, and obtain a blade finite element model that is consistent with the structural characteristic parameters of the actual wind turbine composite blade.

[0046] The module M5 includes the following modules:

[0047] Module M51: used to compile the data of the blade global strain data set E into a column matrix and combine them in chronological order to obtain a strain snapshot matrix A;

[0048] Module M52: for performing modal decomposition on the strain snapshot matrix A using an intrinsic orthogonal decomposition method to obtain a strain snapshot matrix B of a dominant mode;

[0049] Module M53: obtaining the equivalent load time series spectrum of the n operating parameter combinations according to the blade load inversion model;

[0050] Module M54: used to construct a training database Z using the equivalent load time series spectrum of the n working condition parameter combinations as input and the strain snapshot matrix B as output;

[0051] Module M55: used to construct a blade load-strain field reduced-order prediction model based on the training database Z using a radial basis function neural network.

[0052] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0053] The method of establishing a blade finite element model proposed in the present invention is combined with an optimization algorithm to ensure the consistency of the structural characteristic parameters of the blade finite element model and the actual composite material blade, thereby improving the accuracy of the global stress field numerical simulation.

[0054] The strain test position of the present invention is located at the root of the blade, which avoids the cost and safety problems caused by setting multiple electrical sensors at the middle and tip of the blade.

[0055] The blade load inversion model proposed in the present invention adopts a radial basis function neural network with output feedback, fully utilizes a multi-condition sample data set, and combines the strain time series spectrum of the blade strain test position to achieve efficient and accurate acquisition of the equivalent load time series spectrum of the actual wind turbine composite blade.

[0056] The load-strain field reduction prediction model proposed in the present invention adopts the intrinsic orthogonal decomposition method, which greatly improves the computational efficiency of the neural network model and provides an efficient and feasible solution for global strain reconstruction and visualization of wind turbine composite blades. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0058] Figure 1 The figure is a flow chart of an embodiment of the method and system for global strain reconstruction of a composite material blade of a wind turbine according to the present invention. DETAILED DESCRIPTION

[0059] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0060] The purpose of the present invention is to provide a method and system for global strain reconstruction of a wind turbine composite blade.

[0061] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0062] An embodiment of the global strain reconstruction method for a wind turbine composite blade of the present invention comprises the following steps:

[0063] Step S1: establishing a blade finite element model that is consistent with the structural characteristic parameters of an actual wind turbine composite blade;

[0064] Step S2: determining u blade root strain test positions, where u is an integer greater than 2;

[0065] Step S3: establishing a blade load inversion model;

[0066] The establishment of the blade load inversion model includes the following steps S31 to S34:

[0067] Step S31: Set n combinations of operating parameters, where n is an integer greater than 2. The operating parameters include wind speed, wind direction, and turbulence intensity. Based on the blade element momentum theory, divide the blade into m blade elements, where m is an integer greater than 2. Obtain the lift L of the jth blade element under the i-th operating parameter combination. ij , resistance D ij , torque M ij , i=1,2,…n, j=1,2,…m;

[0068] Step S32: Apply the lift force L to the blade finite element model ij , resistance D ij , torque M ij , obtain the strain ε of the qth blade root strain test position under the i-th working condition parameter combination iq , q=1,2,…u, obtain the blade global strain data set E under the n operating parameter combinations;

[0069] Step S33: Using the lift L ij , resistance D ij , torque M ij and the strain ε iq Establish load inversion training sample data set T;

[0070] Step S34: Based on the load inversion training sample data set T, the strain ε iq As input, take the lift force L ij , resistance D ij , torque M ij As output, the blade load inversion model is obtained based on the radial basis function neural network with output feedback;

[0071] Step S4: performing a strain dynamic test on the actual wind turbine composite blade to obtain strain time series spectra at the u blade root strain test positions, and obtaining an equivalent load time series spectrum of the actual wind turbine composite blade based on the blade load inversion model;

[0072] Step S5: establishing a blade load-strain field reduced-order prediction model;

[0073] Step S6: using the equivalent load time series spectrum of the actual wind turbine composite blade as input, and obtaining a blade global strain cloud diagram according to the blade load-strain field reduced-order prediction model.

[0074] The steps of establishing a blade finite element model consistent with the structural characteristic parameters of an actual wind turbine composite blade include:

[0075] Step S11: determining the initial layup parameters of the finite element blade according to the size, load, material, process level, and installation method of the actual wind turbine composite blade, and establishing an initial blade finite element model;

[0076] Step S12: obtaining structural characteristic parameters of an actual wind turbine composite blade through the initial blade finite element model; the structural characteristic parameters of the actual wind turbine composite blade include stiffness distribution characteristics, mass characteristics, and modal parameters;

[0077] Step S13: selecting the initial layup parameters of the finite element blade as optimization design variables, taking the consistency of the structural characteristic parameters of the finite element blade with the structural characteristic parameters of the actual wind turbine composite blade as the optimization goal, and combining an optimization algorithm to solve and obtain the optimized layup parameters of the finite element blade;

[0078] Step S14: Based on the optimized ply parameters of the finite element blade, the initial blade finite element model is adjusted and modified to obtain an optimized blade finite element model, and the optimized blade finite element model is verified to obtain a blade finite element model that is consistent with the structural characteristic parameters of the actual wind turbine composite blade. The steps of establishing the blade load-strain field reduced-order prediction model include:

[0079] Step S51: compiling the data of the blade global strain dataset E into a column matrix and combining them in chronological order to obtain a strain snapshot matrix A;

[0080] Step S52: performing modal decomposition on the strain snapshot matrix A using an intrinsic orthogonal decomposition method to obtain a strain snapshot matrix B of a dominant mode;

[0081] Step S53: obtaining the equivalent load time series spectrum of the n operating parameter combinations according to the blade load inversion model;

[0082] Step S54: constructing a training database Z using the equivalent load time series spectra of the n working condition parameter combinations as input and the strain snapshot matrix B as output;

[0083] Step S55: using a radial basis function neural network and based on the training database Z, constructing a blade load-strain field reduced-order prediction model.

[0084] Based on the same inventive concept, an embodiment of the present invention also provides a global strain reconstruction system for a wind turbine composite blade. Since the principles of these devices in solving problems are similar to a global strain reconstruction method for a wind turbine composite blade, the implementation of these devices can refer to the implementation of the method, and the repeated parts will not be repeated.

[0085] The contents described in the embodiments of this specification are merely an enumeration of the implementation forms of the inventive concept. The scope of protection of the present invention should not be regarded as limited to the specific forms described in the embodiments. The scope of protection of the present invention also extends to equivalent technical means that can be conceived by those skilled in the art based on the inventive concept.

Claims

1. A global strain reconstruction method for a wind turbine composite blade, characterized in that: The following steps are involved: Step S1: establishing a blade finite element model that is consistent with the structural characteristic parameters of an actual wind turbine composite blade; Step S2: determining u blade root strain test positions, where u is an integer greater than 2; Step S3: establishing a blade load inversion model; The establishment of the blade load inversion model includes the following steps S31 to S34: Step S31: Set n combinations of operating parameters, where n is an integer greater than 2. The operating parameters include wind speed, wind direction, and turbulence intensity. Based on the blade element momentum theory, divide the blade into m blade elements, where m is an integer greater than 2. Obtain the lift L of the jth blade element under the i-th operating parameter combination. ij , resistance D ij , torque M ij , i=1,2,…n, j=1,2,…m; Step S32: Apply the lift force L to the blade finite element model ij , resistance D ij , torque M ij , obtain the strain ε of the qth blade root strain test position under the i-th working condition parameter combination iq , q=1,2,…u, obtain the blade global strain data set E under the n operating parameter combinations; Step S33: Using the lift L ij , resistance D ij , torque M ij and the strain ε iq Establish load inversion training sample data set T; Step S34: Based on the load inversion training sample data set T, the strain ε iq As input, take the lift force L ij , resistance D ij , torque M ij As output, the blade load inversion model is obtained based on the radial basis function neural network with output feedback; Step S4: performing a strain dynamic test on the actual wind turbine composite blade to obtain strain time series spectra at the u blade root strain test positions, and obtaining an equivalent load time series spectrum of the actual wind turbine composite blade based on the blade load inversion model; Step S5: establishing a blade load-strain field reduced-order prediction model; Step S6: using the equivalent load time series spectrum of the actual wind turbine composite blade as input, and obtaining a blade global strain cloud diagram according to the blade load-strain field reduced-order prediction model.

2. The global strain reconstruction method for wind turbine composite blades according to claim 1, characterized in that: The steps of establishing a blade finite element model consistent with the structural characteristic parameters of an actual wind turbine composite blade include: Step S11: determining the initial layup parameters of the finite element blade according to the size, load, material, process level, and installation method of the actual wind turbine composite blade, and establishing an initial blade finite element model; Step S12: obtaining structural characteristic parameters of an actual wind turbine composite blade through the initial blade finite element model; the structural characteristic parameters of the actual wind turbine composite blade include stiffness distribution characteristics, mass characteristics, and modal parameters; Step S13: selecting the initial layup parameters of the finite element blade as optimization design variables, taking the consistency of the structural characteristic parameters of the finite element blade with the structural characteristic parameters of the actual wind turbine composite blade as the optimization goal, and combining an optimization algorithm to solve and obtain the optimized layup parameters of the finite element blade; Step S14: According to the optimized ply parameters of the finite element blade, the initial blade finite element model is adjusted and modified to obtain an optimized blade finite element model, and the optimized blade finite element model is verified to obtain a blade finite element model that is consistent with the structural characteristic parameters of the actual wind turbine composite blade.

3. The global strain reconstruction method for wind turbine composite blades according to claim 1, characterized in that: The steps of establishing the blade load-strain field reduced-order prediction model include: Step S51: compiling the data of the blade global strain dataset E into a column matrix and combining them in chronological order to obtain a strain snapshot matrix A; Step S52: performing modal decomposition on the strain snapshot matrix A using an intrinsic orthogonal decomposition method to obtain a strain snapshot matrix B of a dominant mode; Step S53: obtaining the equivalent load time series spectrum of the n operating parameter combinations according to the blade load inversion model; Step S54: constructing a training database Z using the equivalent load time series spectra of the n working condition parameter combinations as input and the strain snapshot matrix B as output; Step S55: using a radial basis function neural network and based on the training database Z, constructing a blade load-strain field reduced-order prediction model.

4. A global strain reconstruction system for composite blades of wind turbines, characterized in that: Includes the following modules: Module M1: used to establish a blade finite element model that is consistent with the structural characteristic parameters of the actual wind turbine composite blade; Module M2: used to determine u blade root strain test positions, where u is an integer greater than 2; Module M3: used to establish blade load inversion model; The module M3 includes the following modules: Module M31: used to set n combinations of operating parameters, where n is an integer greater than 2. The operating parameters include wind speed, wind direction, and turbulence intensity. Based on the blade element momentum theory, the blade is divided into m blade elements, where m is an integer greater than 2. The lift L of the jth blade element under the i-th operating parameter combination is obtained. ij , resistance D ij , torque M ij , i=1,2,…n, j=1,2,…m; Module M32: used to apply the lift force L in the blade finite element model ij , resistance D ij , torque M ij , obtain the strain ε of the qth blade root strain test position under the i-th working condition parameter combination iq , q=1,2,…u, obtain the blade global strain data set E under the n operating parameter combinations; Module M33: used to lift the ij , resistance D ij , torque M ij and the strain ε iq Establish load inversion training sample data set T; Module M34: used for inverting the training sample data set T based on the load, with the strain ε iq As input, take the lift force L ij , resistance D ij , torque M ij As output, the blade load inversion model is obtained based on the radial basis function neural network with output feedback; Module M4: for performing a strain dynamic test on the actual wind turbine composite blade to obtain a strain time series spectrum of the u blade root strain test positions, and based on the blade load inversion model, obtain an equivalent load time series spectrum of the actual wind turbine composite blade; Module M5: used to establish a blade load-strain field reduced-order prediction model; Module M6: used to obtain a blade global strain cloud diagram based on the blade load-strain field reduced-order prediction model using the equivalent load time series spectrum of the actual wind turbine composite blade as input.

5. The global strain reconstruction system for wind turbine composite blades according to claim 4, characterized in that: The module M1 includes the following modules: Module M11: used to determine the initial layup parameters of the finite element blade and establish the initial blade finite element model based on the size, load, material, process level, and installation method of the actual wind turbine composite blade; Module M12: used to obtain the structural characteristic parameters of the actual wind turbine composite blade using the initial blade finite element model; the structural characteristic parameters of the actual wind turbine composite blade include stiffness distribution characteristics, mass characteristics and modal parameters; Module M13: for selecting the initial layup parameters of the finite element blade as optimization design variables, taking the consistency of the structural characteristic parameters of the finite element blade with the structural characteristic parameters of the actual wind turbine composite blade as the optimization goal, and combining the optimization algorithm to solve and obtain the optimized layup parameters of the finite element blade; Module M14: used to adjust and modify the initial blade finite element model according to the optimized ply parameters of the finite element blade to obtain an optimized blade finite element model, verify the optimized blade finite element model, and obtain a blade finite element model that is consistent with the structural characteristic parameters of the actual wind turbine composite blade.

6. The global strain reconstruction system for wind turbine composite blades according to claim 4, characterized in that: The module M5 includes the following modules: Module M51: used to compile the data of the blade global strain data set E into a column matrix and combine them in chronological order to obtain a strain snapshot matrix A; Module M52: for performing modal decomposition on the strain snapshot matrix A using an intrinsic orthogonal decomposition method to obtain a strain snapshot matrix B of a dominant mode; Module M53: obtaining the equivalent load time series spectrum of the n operating parameter combinations according to the blade load inversion model; Module M54: used to construct a training database Z using the equivalent load time series spectrum of the n working condition parameter combinations as input and the strain snapshot matrix B as output; Module M55: used to construct a blade load-strain field reduced-order prediction model based on the training database Z using a radial basis function neural network.

Images