A method for optimizing the arrangement of structural monitoring optical fiber sensors

Abstract

The application discloses a kind of structural monitoring optical fiber sensor multi-objective optimization arrangement methods, step S1, the best sensor quantity selection is determined using modal metric function;Step S2, sensor multi-objective optimization arrangement is carried out in combination with sensor coverage model, weight factor and perception overlap cost function model.The technical scheme of the present application solves the problem that the monitoring coverage is redundant and the overall deformation of the structure cannot be accurately reflected due to uneven distribution of sensor quantity and layout, and simultaneously combines a visual measurement system to reduce the monitoring error introduced by the traditional line positioning method, thereby improving the sensor arrangement efficiency and positioning accuracy.

Description

Technical Field

[0001] This invention belongs to the field of structural deformation monitoring technology, and particularly relates to a method for optimizing the arrangement of fiber optic sensors for structural monitoring. Background Technology

[0002] With the development of large-scale composite material assembly fields such as aerospace, structural health monitoring throughout the product lifecycle plays a crucial role in ensuring the overall performance of structures. During the assembly process, large composite products experience structural deformation due to stress, impact, and other loads. Structural deformation monitoring, as a vital component of health monitoring, involves installing sensors on the surface or internal components and using online data collection to identify structural health and assess structural safety performance. The placement of sensors, as the foundation for acquiring structural health monitoring information, has a profound impact on the effective acquisition of structural information. Ideally, increasing the number of sensors would improve the ability to perceive detailed structural information. However, in the assembly of large-scale composite structures in aerospace and other fields, the number of sensors is typically limited due to the large number of components, complex assembly environments, and data maintenance constraints. Therefore, optimizing the placement of a limited number of sensors to maximize the efficiency of structural condition data acquisition is of great significance to the development of structural health monitoring.

[0003] Currently, with the development of intelligent optimization algorithms, numerous studies are based on these algorithms to optimize sensor placement. However, optimization using a single objective function often neglects the number and angular distribution of sensors, leading to excessive overlap and coverage issues during the optimization process. Due to the complexity and multi-objective nature of real-world problems, the applicability of single-objective optimization methods for sensor optimization research has decreased, and multi-objective optimization is gradually taking a dominant role in sensor placement. While current multi-objective optimization algorithms have improved the issue of sensor quantity, there is still a significant gap between the optimized placement and the constraints of actual applications, easily leading to problems such as insufficient modal coverage and decreased sensitivity in key areas, resulting in missing sensor monitoring data. Therefore, it is necessary to ensure the overall identifiability of the structure under the premise of lightweight sensors, while considering the sensor's sensing range, the overall structural changes, and the sensor angles. Summary of the Invention

[0004] The technical problem to be solved by this invention is to provide an optimized arrangement method for fiber optic sensors for structural monitoring, which solves the problem of redundant monitoring coverage caused by uneven distribution of the number and layout of sensors, which makes it impossible to accurately reflect the overall deformation of the structure. Under the premise of meeting the requirement of lightweight sensors, the method combines the overall trend of structural changes, allocates weights according to different regions, and adjusts the sensor arrangement angle according to the sensing range to improve the overall utilization rate of sensors, ensuring the monitoring of key areas of the structure while achieving effective global monitoring of the measured structure.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A method for optimizing the arrangement of fiber optic sensors for structural monitoring includes:

[0007] Step S1: Use the modal attitude function to determine the optimal number of sensors;

[0008] Step S2: Combine the sensor coverage model, weighting factor and perception overlap cost function model to optimize the sensor multi-objective layout.

[0009] As a preferred option, in step S1, seven modal attitude parameters are selected, namely effective independence, Fisher information matrix, information entropy, kinetic energy, average driving point residual, eigenvalue vector product, and modal confidence. One of the seven modal attitude parameters and the number of sensors are used as optimization objectives. The Pareto solution obtained between the modal function and the number of sensors is analyzed to determine the optimal number of sensors.

[0010] As a preferred option, the optimal mode function is selected based on the Hypervolume index.

[0011] Preferably, in step S2, based on the sensor coverage model, weighting factor, and sensing overlap cost function model, the NSGA-II multi-objective optimization algorithm is adopted. By comprehensively considering multiple conflicting objective functions, the global optimal solution is found based on the Pareto front. According to the number of selected sensors, the optimal model attitude function and sensor coverage rate are used as objective functions, and the sensor position and angle are optimized based on the structural weighting factor and sensor overlap cost function. The optimal position, angle, overlap cost, and coverage rate of the sensors under the optimal number are output.

[0012] For monitoring large composite structures, based on the optimized number of sensors, combined with the optimal modal function and sensor perception model, and considering the weighting factors and overlap cost function for different regions, the sensor positions and angles were optimized. With the optimal number of sensors, not only was sensor coverage overlap and blind spots reduced, ensuring monitoring of critical areas while also considering overall structural coverage, but a visual tracking projection positioning system was also used to determine the positions and angles of the fiber optic sensors, reducing monitoring errors introduced by traditional line-marking positioning methods, thereby improving sensor deployment efficiency and positioning accuracy. The feasibility of this optimized deployment method was verified by combining random point placement with binocular visual scanning, improving the accuracy and reliability of structural deformation monitoring in practical applications. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0014] Figure 1 This is a flowchart of the fiber optic sensor optimization arrangement method for structural monitoring according to an embodiment of the present invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] Example 1:

[0018] Currently, in the health monitoring of large composite structures, there are issues such as uneven sensor distribution leading to excessive overlap or insufficient sensor modal coverage resulting in monitoring blind spots, and decreased sensor sensitivity leading to missing sensor monitoring data. Figure 1 As shown, this embodiment of the invention provides a method for optimizing the arrangement of fiber optic sensors for structural monitoring, comprising:

[0019] Step S1: Sensor number selection based on modal attitude function

[0020] To achieve lightweight sensor design and optimal sensor quantity selection, seven modal attitude functions were selected: effective independence, Fisher information matrix, information entropy, kinetic energy, average driving point residual, eigenvalue vector product, and modal confidence. One of these seven modal attitude functions and the number of sensors were used as optimization objectives. The Pareto solutions obtained between the modal functions and the number of sensors were analyzed to determine the optimal number of sensors.

[0021] Step S2: Optimal Mode Function Selection Based on Hypervolume Index

[0022] Based on the Hypervolume (HV) index, the Pareto solutions obtained from the above seven mode attitude quantitative functions and the number of sensors are evaluated using hypervolume assessment to measure the diversity and balance of the obtained solution set. According to the trend of the obtained solution set, it is divided into linear solutions and convex solutions. The optimal mode attitude quantitative function is selected from the two solution set trends using the HV index. The expression for the HV index is:

[0023]

[0024] In the formula, v(x,P) r ) represents the solution in the non-dominated solution set D and the reference point P. r The resulting spatial hypervolume is used to determine the optimal modulus parameters based on the hypervolume value.

[0025] Step S3: Establishment of FBG fiber optic sensing coverage model

[0026] To achieve the optimal number of sensors while ensuring maximum sensor coverage within the measured structure area, an FBG sensing coverage model needs to be established. When solving for optimal sensor coverage, a probabilistic sensing model is often introduced to analyze the sensor coverage problem. Due to the complexity of the sensor monitoring environment in practical applications, the sensing probability is converted into a decay function form based on different conditions, letting... Let N represent the joint sensing coverage of N sensors. Then, the average coverage of N sensors over M nodes is:

[0027]

[0028] Step S4, Weighting Factors and Angle Overlap Cost Function Model

[0029] To improve monitoring coverage and optimize sensor resource utilization when the number of sensors is limited, a weighting factor is introduced to optimize sensor location distribution. Before determining the weighting factor, to avoid increased data partitioning complexity due to uneven distribution of deformation datasets, partitioning is performed based on the overall trend of structural deformation and the definitions of nominal and critical values ​​in structural analysis. The minimum and maximum nominal values, and then the minimum and maximum critical values, are selected sequentially to partition the deformation data. Based on prior knowledge, deformation data between the minimum and maximum nominal values ​​should account for approximately 50% of the total deformation data, as data in this range tends to be closer to the sample mean. Deformation data between the minimum and maximum critical values ​​should account for approximately 95% of the total deformation data; as the deformation data approaches the critical value, it tends to deviate from the mean. After partitioning, weights are defined based on the mean values ​​of data in low, medium, and high deformation regions. This method is less susceptible to extreme values ​​and has a certain degree of stability. Furthermore, the mean values ​​of different deformation regions can intuitively reflect the degree of deformation in that region, conforming to the overall trend of data characteristics. The mean values ​​of the low, medium, and high deformation regions are set to μ.l μ m μ h Its expression is:

[0030]

[0031] In the formula, n l n m n h x represents the number of data points in the low, medium, and high deformation regions, respectively. l x m x h These represent the deformation data points in each region. Given the mean of the dataset for each region, after normalization, the weights ω for low, medium, and high deformation regions are determined. l ω m and ω h They are represented as follows:

[0032]

[0033] To reduce data redundancy caused by overlap between sensors, the sensor angles need to be optimized. Let the sensor angle adjustment range be 0° to 360°. The Euclidean distance between the two sensor centers can be determined based on the sensor coordinates. Let (x... a ,y a ), (x b ,y b ) represent the corresponding position coordinates of sensor a and sensor b, respectively, (θ) a ,θ b If ) represents the sensor rotation angle, then its angular overlap cost function can be expressed as:

[0034]

[0035] Its optimization objective must satisfy the minimization of the overlapping cost function:

[0036]

[0037] According to the theorem for calculating the area of ​​irregular polygons, the overlapping area of ​​the sensing regions between sensor a and sensor b is:

[0038]

[0039] In the formula, (x i ,y i ),(x i+1 ,y i+1 ),(x n ,y n () represents the vertex coordinates of the overlapping polygon. Similarly, the area of ​​the sensor placement constraint region is:

[0040]

[0041] In the formula, (x j ,y j ),(x j+1 ,y j+1 ),(x m ,y m () represents the vertex coordinates of the constrained polygon. At this point, the ratio of the overlapping area perceived by the N sensors at different angles to the area of ​​the constrained region is:

[0042]

[0043] Step S5: Multi-objective optimization method based on NSGA-II

[0044] Based on the above model, the NSGA-II (Non-dominated Sorting Genetic Algorithm II) multi-objective optimization algorithm is adopted. By comprehensively considering multiple conflicting objective functions, the global optimum is found based on the Pareto front. Depending on the number of selected sensors, the optimal model attitude function and sensor coverage are used as objective functions. The sensor positions and angles are optimized based on the structural weight factor and the sensor overlap cost function, outputting the optimal sensor positions, angles, overlap costs, and coverage for the optimal number of sensors.

[0045] This invention first uses a modal attitude function to determine the optimal number of sensors, achieving lightweight sensor design. Second, it combines a sensor coverage model, weighting factors, and a perception overlap cost function model for multi-objective sensor optimization. Finally, based on the optimization results, it is applied to structural deformation monitoring to verify the feasibility of the optimization method. A visual tracking projection positioning system is used to determine the position and angle of the fiber optic sensors, reducing monitoring errors introduced by traditional line-marking positioning methods, thereby improving sensor placement efficiency and positioning accuracy. Results show that the absolute error of deformation monitoring is within 0.1 mm, the relative error is within 6.47%, and the root mean square error is 0.066 mm. This method allocates weights according to different regions based on the overall deformation trend of the structure and optimizes the sensor placement angle in conjunction with the sensing range, reducing coverage redundancy between sensors. With lightweight design as a prerequisite, it improves the overall utilization efficiency of the sensors and ensures effective monitoring of the deformation of the measured structure.

[0046] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for optimizing the arrangement of fiber optic sensors for structural monitoring, characterized in that, include: Step S1: Use the modal attitude function to determine the optimal number of sensors; Step S2: Combine the sensor coverage model, weighting factors, and sensing overlap cost function model to optimize the sensor multi-objective layout; In step S1, seven modal attitude parameters are selected, namely effective independence, Fisher information matrix, information entropy, kinetic energy, average driving point residual, eigenvalue vector product and modal confidence. One of the seven modal attitude parameters and the number of sensors are used as optimization objectives. The Pareto solution obtained between the modal function and the number of sensors is analyzed to determine the optimal number of sensors. Selecting the optimal mode function based on the Hypervolume metric; In step S2, based on the sensor coverage model, weighting factor, and sensing overlap cost function model, the NSGA-II multi-objective optimization algorithm is adopted. By comprehensively considering multiple conflicting objective functions, the global optimal solution is found based on the Pareto front. According to the number of selected sensors, the optimal model attitude function and sensor coverage rate are used as objective functions. The sensor position and angle are optimized based on the structural weighting factor and sensor overlap cost function. The optimal sensor position, angle, overlap cost, and coverage rate under the optimal number of sensors are output. In step S2, a weighting factor is introduced to optimize the sensor location distribution; before the weighting factor is determined, the structure is partitioned according to the overall trend of structural deformation and the definitions of nominal and critical values ​​in the structural analysis. After partitioning, weights are defined based on the average values ​​of the data from the low, medium, and high deformation regions.

Images