Multi-step sensor layout optimization method for large structure deformation monitoring
By using a multi-step deduplication sensor layout optimization method, leveraging strain field characteristics and inverse finite element algorithm, the sensor layout is optimized, solving the problems of large number and high cost of sensors in deformation monitoring of large aerospace structures, and achieving high-precision and lightweight monitoring.
Patent Information
- Application Number
- CN202310643010.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-06-01
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2043-06-01
AI Technical Summary
Large aerospace structures require a huge number of strain sensors for deformation monitoring due to their large size and complex strain distribution, resulting in high monitoring costs and failure to meet lightweight requirements.
A multi-step deduplication sensor layout optimization method is adopted. Through finite element simulation, measured strain optimization and inverse finite element algorithm, the sensor layout is optimized. By utilizing the gradient, amplitude, symmetry and linear relationship of the strain field, the number of sensors is reduced and the monitoring accuracy is improved.
It enables high-precision deformation monitoring of large structures, meeting the lightweight requirements of aerospace, while reducing monitoring costs.
Smart Images

Figure CN116757021B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of engineering structure deformation monitoring, specifically involving a multi-step redundancy reduction sensor layout optimization method for large structure deformation monitoring. Background Technology
[0002] Large aerospace structures, such as large deployable antennas in orbit, are subjected to rapid temperature changes and non-uniform temperature loads during operation. These complex thermal loads can cause unpredictable thermal deformation of the antenna structure, severely affecting the normal operation of the spacecraft. Deformation monitoring is a crucial part of structural health monitoring, enabling real-time perception of the spacecraft's operational status and providing key technologies for subsequent structural condition assessment and analysis. Real-time monitoring and effective control of structural deformation and vibration of large antenna structures is a critical issue in the field of aerospace technology.
[0003] In strain-based structural deformation monitoring methods, fiber Bragg grating sensors possess advantages such as light weight, small size, strong anti-interference capability, and good electromagnetic insulation performance, making them suitable for deformation monitoring of on-orbit antenna structures with densely packed electromagnetic sensitive elements. By acquiring real-time strain data and combining it with advanced signal processing and deformation reconstruction algorithms, online real-time sensing of aerospace structures can be achieved. Among the developed strain-based deformation reconstruction algorithms, the inverse finite element method (IFM) has gained widespread attention due to its advantages of not requiring structural material or load characteristics and offering high reconstruction accuracy.
[0004] However, measured strain input is one of the key factors for the inverse finite element algorithm to achieve high-precision real-time reconstruction. Large-scale structural deformation monitoring requires a large number of strain sensors to be deployed on the structure. Due to limitations in engineering conditions such as monitoring cost and sensor weight, it is difficult to meet the lightweight requirements of aerospace structural deformation monitoring. Summary of the Invention
[0005] To address the shortcomings of the existing technologies, the present invention aims to provide a multi-step deduplication sensor layout optimization method for large-scale structural deformation monitoring, thereby solving the problems of the large number of strain sensors, high monitoring costs, and failure to meet the lightweight requirements of aerospace deformation monitoring caused by the large size of the deformation monitoring object and the complex strain distribution.
[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:
[0007] The present invention provides a multi-step redundancy reduction sensor layout optimization method for large structure deformation monitoring, comprising the following steps:
[0008] 1) Establish a finite element model of a large structure using finite element simulation software. Extract the finite element simulation strain fields in three directions (0°, 45°, and 90°) on the surface of the structure to be monitored from the finite element model. Based on the distribution characteristics of the finite element strain fields in the three directions, arrange measurement points at extreme value locations and gradient change directions to obtain the initial layout of strain sensors in the three directions (0°, 45°, and 90°).
[0009] 2) Based on the initial layout of strain sensors in the three directions of 0°, 45° and 90° in step 1), the sensors are arranged on the large structure. The strain is measured in real time during the thermal deformation experiment of the large structure, and the measured strain fields in the three directions of 0°, 45° and 90° of the structure are fitted by fitting functions respectively.
[0010] 3) Analyze the measured strain fields in the 0° and 90° directions. Delete measurement points with small gradient changes according to the gradient redundancy principle, delete measurement points with measured values less than 50 microstrains according to the amplitude redundancy principle, and delete measurement points with measured value deviations of less than 50 microstrains at symmetrical positions according to the symmetry redundancy principle. This will result in the optimized sensor layout in the 0° and 90° directions.
[0011] 4) Based on the principle of linear relationship redundancy elimination, construct the mapping relationship between the measured strain in the 0° and 90° directions and the strain in the 45° direction, thereby deleting some strain measurement points in the 45° direction and obtaining the optimized layout of the measured strain sensor.
[0012] 5) Adjust the strain measurement point to the center area of the reverse unit and analyze the deformation reconstruction error distribution. Further optimize the sensor layout optimized in steps 3) and 4).
[0013] Furthermore, the initial layout of the strain sensor in step 1) specifically includes:
[0014] 11) Input the material and load information of the large structure into the finite element simulation software, establish the finite element calculation model of the large structure, extract the finite element calculation strain of the surface to be monitored, draw the gradient distribution diagram of the strain field in the three directions of 0°, 45° and 90° respectively, and mark the location of the strain maximum value.
[0015] 12) For the strain fields in the three directions in step 11), arrange strain measurement points at the extreme values and large values of strain amplitude, and arrange strain measurement points along the direction of strain gradient change.
[0016] 13) Arrange strain measurement points near the edge of the surface to be tested on the large structure to improve the accuracy of fitting the strain field and obtain the initial layout of the three-direction linear strain sensors on the surface to be monitored of the large structure.
[0017] Furthermore, the gradient redundancy elimination principle in step 3) specifically includes:
[0018] Based on the analysis of the measured strain gradients on the surface in the 0° and 90° directions, the strain measurement points with maximum values are retained, a strain gradient interval threshold is set between adjacent strain measurement points, redundant strain measurement points with strain changes less than the interval threshold are deleted, and the zeroing threshold is adjusted until the maximum deviation of the fitted strain field before and after optimization is less than the error limit.
[0019] Furthermore, the amplitude redundancy elimination principle in step 3) specifically includes:
[0020] Based on the analysis of the measured strain amplitudes of the surface in the 0° and 90° directions, a zeroing threshold is set, strain measurement points with strain amplitudes less than the zeroing threshold are deleted, and the strain value of the strain measurement point is set to zero. The zeroing threshold is then adjusted until the maximum deviation of the fitted strain field before and after optimization is less than the error limit.
[0021] Furthermore, the symmetry redundancy elimination principle in step 3) specifically includes:
[0022] Based on the analysis of the symmetry of the measured strain on the surface in the 0° and 90° directions, a symmetry error threshold is set, and strain measurement points with strain errors less than the symmetry error threshold at symmetric positions are deleted. That is, strain measurement points on one side are retained, and the strain values of the strain measurement points on the other side are obtained indirectly based on the symmetry. The symmetry error threshold is then adjusted until the maximum deviation of the fitted strain field before and after optimization is less than the error limit.
[0023] Furthermore, the linear relationship redundancy elimination principle in step 4) specifically includes:
[0024] 41) Fit the strain fields at 0° and 90° at M time points respectively, and extract the strain values at 45° at M time points. Construct the mapping relationship between the linear strain at 45° at M time points and the strain at 0° and 90°. Construct the relationship between the measured strain at 0°, 90° and 45° using a bivariate linear function. The strain relationship at the i-th strain measurement point is expressed as:
[0025] ε 45,i =p 1,i ε 0,i +p 2,i ε 90,i +p 3,i
[0026] In the formula, ε 45,i ε 0,i and ε 90,i Let p be the strain values in three directions at the i-th measurement point. 1,i p 2,i and p 3,i The strain coefficient at the i-th measurement point;
[0027] 42) Fit the strain fields in the 45° direction before and after optimization respectively and calculate the maximum deviation between them. Adjust the number of strain measurement points in the 45° direction to be retained until the maximum deviation of the fitted strain fields before and after optimization is less than the error limit.
[0028] Furthermore, step 5) specifically includes:
[0029] 51) Based on the element division results of the large structure inverse finite element algorithm, modify the optimized triaxial strain sensor layout in steps 3) and 4), adjust the position of some strain measurement points to be located within the inverse element, and avoid them near the boundary of the inverse element; obtain the strain of the inverse element that lacks strain measurement points by strain interpolation fitting.
[0030] 52) Fit the linear strain fields in the three directions of 0°, 90° and 45° under the initial layout and optimized layout of the sensor respectively, and calculate the plane strain components at the centroid positions of the upper and lower surfaces of each inverse unit using the following formula;
[0031] ε xx =ε0
[0032] ε yy =ε 90
[0033] γ xy =ε0+ε 90 -2ε 45
[0034] Where ε0 is the strain along the 0-degree direction, ε 90 For the strain along the 90-degree direction, ε 45 For the strain along the 45-degree direction, ε xx ε represents the normal strain along the x-axis of the structural surface. yy For the normal strain along the y-axis of the surface of a large structure, γ xy The shear strain is given by the xy plane; the membrane strain e is calculated based on the plane strain components of the upper and lower surfaces of the i-th inverse element and the structural thickness H. i and curvature k i :
[0035]
[0036]
[0037] Where z = H / 2 represents the upper surface of the structure, and z = -H / 2 represents the lower surface of the structure; the membrane strain e of the i-th reverse unit i and curvature k i Substituting into the local sub-reconstruction equations of the inverse finite element method:
[0038] k e ·ae =f e
[0039] Element equivalent external force matrix f e The specific expression is as follows:
[0040] f e =∫ A (B m,i T e i +H 2 B κ,i T k i )dA
[0041] Among them, a e Let k be the nodal displacement of the i-th inverse element to be solved. e B m,i B κ,i All are constant matrices related to the size of the i-th inverse element, and A is the area of the i-th inverse element. The local sub-reconstruction equations of all inverse elements are calculated, and through coordinate transformation and element assembly, the global reconstruction equations of the structure are obtained:
[0042] KU = F
[0043] In the formula, K is the equivalent stiffness matrix and F is the equivalent external force matrix. After adding boundary constraints, the reconstructed displacement vector U of all inverse element nodes is obtained. The strain fields fitted by the initial and optimized layouts of the sensor are calculated and used as inputs for the inverse finite element method to calculate the difference between the two reconstructed displacements. Strain measurement points are added in the region with the largest difference in reconstructed displacements.
[0044] 53) Adjust the positions of the strain in the three directions of 0°, 45° and 90°, refit the strain fields in the three directions of 0°, 90° and 45° as input for the inverse finite element method, and calculate the difference between the reconstructed displacement and the strain field under the initial sensor layout as input for the algorithm, until the maximum deviation of the reconstructed displacement before and after optimization is less than the error limit, and complete the multi-step redundancy elimination sensor layout optimization.
[0045] The beneficial effects of this invention are:
[0046] This invention establishes a large-scale structural finite element calculation model and adopts a multi-step redundancy elimination sensor layout optimization method, which optimizes the strain sensor layout based on simulation strain optimization step, measured strain optimization step, and reconstruction algorithm optimization step.
[0047] This invention fully utilizes the gradient, amplitude, symmetry characteristics, and linear strain relationship of the strain field on the surface to be monitored, and integrates advanced deformation reconstruction algorithms to achieve high-precision deformation monitoring of large aerospace structures with a smaller number of strain sensors. It solves the problem of a large number of strain monitoring sensors for large structures under complex deformation monitoring requirements, and significantly reduces the cost of strain monitoring while meeting the lightweight requirements of aerospace structure deformation monitoring. Attached Figure Description
[0048] Figure 1 This is a flowchart of the method of the present invention.
[0049] Figure 2 This is a schematic diagram of the composite honeycomb sandwich panel structure and fixture positions used in the example.
[0050] Figure 3a This is a schematic diagram of the initial layout of the finite element strain at 0° on the surface to be monitored and the sensor at 0°.
[0051] Figure 3b This is a schematic diagram of the initial layout of the finite element strain in the 90° direction of the surface to be monitored and the 90° direction sensor.
[0052] Figure 3c This is a schematic diagram of the finite element strain in the 45° direction of the surface to be monitored and the initial layout of the sensor in the 45° direction.
[0053] Figure 4a This is a schematic diagram of the actual stress field measured at 0° on the surface to be monitored.
[0054] Figure 4b This is a schematic diagram of the strain field measured at 90° on the surface to be monitored.
[0055] Figure 4c This is a schematic diagram of the strain field measured at 45° on the surface to be monitored.
[0056] Figure 5 This is a schematic diagram showing the fitting relationship between the measured strain at 0° and 90° and the strain at 45° at typical locations.
[0057] Figure 6 This is a schematic diagram of the reverse unit partitioning.
[0058] Figure 7 This is a schematic diagram of the deformation reconstruction result of the optimized sensor layout under typical temperature difference conditions.
[0059] Figure 8 This is a schematic diagram showing the variation of the maximum error in antenna structure displacement reconstruction over time.
[0060] Figure 9 A schematic diagram of the optimized layout of the triaxial linear strain sensor on the surface to be monitored. Detailed Implementation
[0061] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to embodiments and accompanying drawings. The content mentioned in the embodiments is not intended to limit the present invention.
[0062] Reference Figure 1 As shown, the present invention provides a multi-step redundancy reduction sensor layout optimization method for large structure deformation monitoring, comprising the following steps:
[0063] 1) Establish a finite element model of a large structure using finite element simulation software. Extract the finite element simulation strain fields in three directions (0°, 45°, and 90°) on the surface of the structure to be monitored from the finite element model. Based on the distribution characteristics of the finite element strain fields in the three directions, arrange measurement points at extreme value locations and gradient change directions to obtain the initial layout of strain sensors in the three directions (0°, 45°, and 90°).
[0064] 2) Based on the initial layout of strain sensors in the three directions of 0°, 45° and 90° in step 1), the sensors are arranged on the large structure. The strain is measured in real time during the thermal deformation experiment of the large structure, and the measured strain fields in the three directions of 0°, 45° and 90° of the structure are fitted by fitting functions respectively.
[0065] 3) Analyze the measured strain fields in the 0° and 90° directions. Delete measurement points with small gradient changes according to the gradient redundancy principle, delete measurement points with measured values less than 50 microstrains according to the amplitude redundancy principle, and delete measurement points with measured value deviations of less than 50 microstrains at symmetrical positions according to the symmetry redundancy principle. This will result in the optimized sensor layout in the 0° and 90° directions.
[0066] 4) Based on the principle of linear relationship redundancy elimination, construct the mapping relationship between the measured strain in the 0° and 90° directions and the strain in the 45° direction, thereby deleting some strain measurement points in the 45° direction and obtaining the optimized layout of the measured strain sensor.
[0067] 5) Adjust the strain measurement point to the center area of the reverse unit and analyze the deformation reconstruction error distribution. Further optimize the sensor layout optimized in steps 3) and 4).
[0068] This embodiment takes a large composite sandwich panel structure as the object, and describes the implementation process of the present invention by applying a thermal load to one side of the structure to obtain the structural strain field, and carrying out sensor layout optimization and deformation reconstruction.
[0069] In this embodiment, the object of deformation and reconstruction is the antenna plate, which is a composite honeycomb sandwich panel structure, such as... Figure 2As shown, the dimensions are 3125mm × 1500mm × 20mm (length × width × thickness). A heating element is installed on the lower surface. During heating, a temperature gradient exists along the thickness direction, causing the honeycomb sandwich panel to warp. In the simulation strain optimization step, the material parameters of the honeycomb sandwich panel are input into ABAQUS software to establish a finite element calculation model. Based on measured temperature data, a thermal load is applied, and the simulated strain field of the surface without heating elements in three directions (0°, 90°, and 45°) is extracted. Strain measurement points are arranged at the extreme values and locations of large strain amplitudes, along the direction of strain gradient change, and near the edge of the surface to be tested. This results in the initial layout of the triaxial strain sensor network on the large structure's surface to be monitored. A total of 99 measurement points are arranged in the strain sensor network, with 33 points each in the 0°, 90°, and 45° directions. The initial layout of the triaxial strain sensor network on the surface to be monitored is shown below. Figures 3a-3c As shown.
[0070] In the measured strain optimization step, according to Figures 3a-3c The initial layout of the three strain sensors shown is as follows: sensors are arranged on the structure, and the measured strain of the structure is measured in the thermal deformation experiment. Fitting functions are then used to fit the measured strain fields in the 0°, 45°, and 90° directions of the structure, respectively. Figures 4a-4c As shown, the measured strain fields in the 0° and 90° directions were analyzed. Based on the gradient redundancy elimination principle, a strain gradient interval threshold of 30 με was set, and measurement points with small gradient changes were deleted. Based on the amplitude redundancy elimination principle, a strain gradient interval threshold of 30 με was set, and measurement points with small measured values were deleted. Based on the symmetry redundancy elimination principle, a strain zeroing threshold of 30 με was set, and measurement points with similar measured values at symmetrical positions were deleted. Based on the linear relationship redundancy elimination principle, a mapping relationship between the measured strain in the 0° and 90° directions and the strain in the 45° direction was constructed, thus obtaining the optimized measured strain sensor layout. Figure 5 The mapping relationship between strain at 0° and 90° and strain at 45° is shown at 10 moments during typical position heating process.
[0071] In the optimization step of the reconstruction algorithm, based on the element partitioning results of the inverse finite element algorithm for large structures, such as... Figure 6 As shown, the positions of some strain measurement points are adjusted to be located inside the reverse element, avoiding them near the boundary of the reverse element. Figure 7 Deformation reconstruction results based on the optimized sensor layout are presented under typical heating conditions. Within 10 minutes of heating, the maximum reconstruction deviation of 25 displacement measurement points in the main antenna structure region after the optimized strain sensor layout does not exceed 0.46 mm, and is less than 10% of the maximum displacement. Figure 7As shown, this demonstrates that the existing strain sensor layout and deformation reconstruction algorithm can successfully monitor the deformation of the satellite antenna structure. The current sensor layout meets the requirements for strain field fitting error and displacement field reconstruction error.
[0072] The revised and optimized triaxial strain sensor layout includes 43 strain sensors arranged at 31 locations: 21 at 0°, 20 at 90°, and 2 at 45°. Figure 7 As shown. Therefore, the multi-step deduplication sensor layout optimization method for deformation monitoring of large structures can effectively solve the problems of large number of strain monitoring sensors and complex layout process of large satellite antenna structures under complex deformation monitoring requirements. It can achieve lightweight strain monitoring while ensuring a certain strain monitoring accuracy, reduce strain monitoring cost, and greatly improve sensor layout efficiency.
[0073] This invention has many specific applications. The above description is only a preferred embodiment of this invention. It should be noted that for those skilled in the art, several improvements can be made without departing from the principle of this invention, and these improvements should also be considered within the scope of protection of this invention.
Claims
1. A multi-step redundancy-eliminating sensor layout optimization method for deformation monitoring of large structures, characterized in that, The steps are as follows: 1) Establish a finite element model of a large structure using finite element simulation software. Extract the finite element simulation strain fields in three directions (0°, 45°, and 90°) on the surface of the structure to be monitored from the finite element model. Based on the distribution characteristics of the finite element strain fields in the three directions, arrange measurement points at extreme value locations and gradient change directions to obtain the initial layout of strain sensors in the three directions (0°, 45°, and 90°). 2) Based on the initial layout of strain sensors in the three directions of 0°, 45° and 90° in step 1), the sensors are arranged on the large structure. The strain is measured in real time during the thermal deformation experiment of the large structure, and the measured strain fields in the three directions of 0°, 45° and 90° of the structure are fitted by fitting functions respectively. 3) Analyze the measured strain fields in the 0° and 90° directions. Delete measurement points with small gradient changes according to the gradient redundancy principle, delete measurement points with measured values less than 50 microstrains according to the amplitude redundancy principle, and delete measurement points with measured value deviations of less than 50 microstrains at symmetrical positions according to the symmetry redundancy principle. This will result in the optimized sensor layout in the 0° and 90° directions. 4) Based on the principle of linear relationship redundancy elimination, construct the mapping relationship between the measured strain in the 0° and 90° directions and the strain in the 45° direction, thereby deleting some strain measurement points in the 45° direction and obtaining the optimized layout of the measured strain sensor. 5) Adjust the strain measurement point to the center area of the reverse unit and analyze the deformation reconstruction error distribution. Further optimize the sensor layout optimized in steps 3) and 4).
2. The multi-step redundancy reduction sensor layout optimization method for large structure deformation monitoring according to claim 1, characterized in that, The initial layout of the strain sensor in step 1) specifically includes: 11) Input the material and load information of the large structure into the finite element simulation software, establish the finite element calculation model of the large structure, extract the finite element calculation strain of the surface to be monitored, draw the gradient distribution diagram of the strain field in the three directions of 0°, 45° and 90° respectively, and mark the location of the strain maximum value. 12) For the strain fields in the three directions in step 11), arrange strain measurement points at the extreme values and large values of strain amplitude, and arrange strain measurement points along the direction of strain gradient change. 13) Arrange strain measurement points near the edge of the surface to be tested on the large structure to improve the accuracy of fitting the strain field and obtain the initial layout of the three-direction linear strain sensors on the surface to be monitored of the large structure.
3. The multi-step redundancy reduction sensor layout optimization method for large structure deformation monitoring according to claim 1, characterized in that, The gradient redundancy elimination principle in step 3) specifically includes: Based on the analysis of the measured strain gradients on the surface in the 0° and 90° directions, the strain measurement points with maximum values are retained, a strain gradient interval threshold is set between adjacent strain measurement points, redundant strain measurement points with strain changes less than the interval threshold are deleted, and the zeroing threshold is adjusted until the maximum deviation of the fitted strain field before and after optimization is less than the error limit.
4. The multi-step redundancy reduction sensor layout optimization method for large structure deformation monitoring according to claim 1, characterized in that, The amplitude redundancy elimination principle in step 3) specifically includes: Based on the analysis of the measured strain amplitudes of the surface in the 0° and 90° directions, a zeroing threshold is set, strain measurement points with strain amplitudes less than the zeroing threshold are deleted, and the strain value of the strain measurement point is set to zero. The zeroing threshold is then adjusted until the maximum deviation of the fitted strain field before and after optimization is less than the error limit.
5. The multi-step redundancy reduction sensor layout optimization method for large structure deformation monitoring according to claim 1, characterized in that, The symmetry redundancy elimination principle in step 3) specifically includes: Based on the analysis of the symmetry of the measured strain on the surface in the 0° and 90° directions, a symmetry error threshold is set, and strain measurement points with strain errors less than the symmetry error threshold at symmetric positions are deleted. That is, strain measurement points on one side are retained, and the strain values of the strain measurement points on the other side are obtained indirectly based on the symmetry. The symmetry error threshold is then adjusted until the maximum deviation of the fitted strain field before and after optimization is less than the error limit.
6. The multi-step redundancy-eliminating sensor layout optimization method for large structure deformation monitoring according to claim 1, characterized in that, The linear relationship redundancy elimination principle in step 4) specifically includes: 41) Fit the strain fields at 0° and 90° at M time points respectively, and extract the strain values at 45° at M time points. Construct the mapping relationship between the linear strain at 45° at M time points and the strain at 0° and 90°. Construct the relationship between the measured strain at 0°, 90° and 45° using a bivariate linear function. The strain relationship at the i-th strain measurement point is expressed as: e 45,i =p 1,i e 0,i +p 2,i e 90,i +p 3,i In the formula, ε 45,i ε 0,i and ε 90,i Let p be the strain values in three directions at the i-th measurement point. 1,i p 2,i and p 3,i The strain coefficient at the i-th measurement point; 42) Fit the strain fields in the 45° direction before and after optimization respectively and calculate the maximum deviation between them. Adjust the number of strain measurement points in the 45° direction to be retained until the maximum deviation of the fitted strain fields before and after optimization is less than the error limit.
7. The multi-step redundancy reduction sensor layout optimization method for large structure deformation monitoring according to claim 1, characterized in that, Step 5) specifically includes: 51) Based on the element division results of the large structure inverse finite element algorithm, modify the optimized triaxial strain sensor layout in steps 3) and 4), adjust the position of some strain measurement points to be located within the inverse element, and avoid them near the boundary of the inverse element; obtain the strain of the inverse element that lacks strain measurement points by strain interpolation fitting. 52) Fit the linear strain fields in the three directions of 0°, 90° and 45° under the initial layout and optimized layout of the sensor respectively, and calculate the plane strain components at the centroid positions of the upper and lower surfaces of each inverse unit using the following formula; e xx =ε0 e yy =e 90 c xy =ε0+ε 90 -2e 45 Where ε0 is the strain along the 0-degree direction, ε 90 For the strain along the 90-degree direction, ε 45 For the strain along the 45-degree direction, ε xx ε represents the normal strain along the x-axis of the structural surface. yy For the normal strain along the y-axis of the surface of a large structure, γ xy The shear strain is given by the xy plane; the membrane strain e is calculated based on the plane strain components of the upper and lower surfaces of the i-th inverse element and the structural thickness H. i and curvature k i : Where z = H / 2 represents the upper surface of the structure, and z = -H / 2 represents the lower surface of the structure; the membrane strain e of the i-th reverse unit i and curvature k i Substituting into the local sub-reconstruction equations of the inverse finite element method: k e ·a e =f e Element equivalent external force matrix f e The specific expression is as follows: f e =∫ A (B m,i T yes i +H 2 B κ,i T to i )dA Among them, a e Let k be the nodal displacement of the i-th inverse element to be solved. e B m,i B κ,i All are constant matrices related to the size of the i-th inverse element, and A is the area of the i-th inverse element. The local sub-reconstruction equations of all inverse elements are calculated, and through coordinate transformation and element assembly, the global reconstruction equations of the structure are obtained: KU = F Where K is the equivalent stiffness matrix and F is the equivalent external force matrix. After adding boundary constraints, the reconstructed displacement vector U of all inverse element nodes is obtained. The strain fields fitted by the initial and optimized layouts of the sensor are calculated and used as inputs for the inverse finite element method to calculate the difference between their reconstructed displacements. Strain measurement points are added in the region with the largest difference in reconstructed displacements. 53) Adjust the positions of the strain in the three directions of 0°, 45° and 90°, refit the strain fields in the three directions of 0°, 90° and 45° as input for the inverse finite element method, and calculate the difference between the reconstructed displacement and the strain field under the initial sensor layout as input for the algorithm, until the maximum deviation of the reconstructed displacement before and after optimization is less than the error limit, and complete the multi-step redundancy elimination sensor layout optimization.
Citation Information
Patent Citations
Time compensation method for multi-channel online gas mass spectrometry measurement
CN111007138A
Mark point-based global absolute phase alignment method, storage medium and system
CN113074667A
