A platform cap health monitoring method based on fiber grating sensing
Through fiber grating strain sensor parameter optimization and BP neural network identification model, automated platform cap damage detection is realized, solving the problem of low warning reliability caused by manual analysis in traditional methods, and improving monitoring efficiency and accuracy.
Patent Information
- Application Number
- CN202411162004.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-08-23
AI Technical Summary
The traditional platform cap monitoring method based on fiber grating sensing requires manual analysis of monitoring data, and cannot quickly identify the damage to the platform cap structure, resulting in low early warning reliability.
Establish a parameter optimization model for fiber grating strain sensors, use weighted least squares algorithm to obtain the optimal structural parameters, and build a damage recognition model based on BP neural network, automatically analyze the deformation timing data to identify damage, correct it in response to user feedback, and improve damage Accuracy of identification.
It improves the efficiency and accuracy of platform cap damage detection, reduces dependence on professionals, reduces labor costs, and adapts to environmental changes through model corrections, ensuring the accuracy and stability of long-term operation.
Smart Images

Figure CN118690140B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of platform cap structure monitoring, and specifically is a platform cap health monitoring method based on fiber grating sensing. Background Art
[0002] With the rapid development of railway transportation, platform safety has received increasing attention.
[0003] As an important part of the platform structure, the platform cap’s health status monitoring is particularly important, which is directly related to the safety and comfort of passengers.
[0004] At present, the platform cap health monitoring system needs to monitor key indicators such as the structural deformation and crack development of the platform cap in real time and continuously to ensure that potential safety hazards can be discovered in a timely manner. In the existing technology, the fiber grating sensor uses the grating structure inside the optical fiber to reflect light of a specific wavelength. When external physical quantities (such as temperature and stress) act on the optical fiber, the grating period will change, which will cause the reflected wavelength to shift. By monitoring the change of this wavelength, high-precision measurement of external physical quantities can be achieved.
[0005] The traditional platform cap monitoring method based on fiber grating sensing requires manual analysis of the acquired monitoring data to confirm the condition of the platform cap. It is unable to quickly identify the damage of the platform cap structure reflected in the data, cannot provide timely warning, and has low reliability. Summary of the invention
[0006] The purpose of an embodiment of the present invention is to provide a platform cap health monitoring method based on fiber grating sensing, aiming to solve the technical problem that the traditional platform cap monitoring method based on fiber grating sensing needs to manually analyze the acquired monitoring data to confirm the condition of the platform cap, cannot quickly identify the damage of the platform cap structure reflected in the data, cannot provide timely warning, and has low reliability.
[0007] To achieve the above object, the present invention provides the following technical solutions.
[0008] In one embodiment of the present invention, a method for monitoring the health of a platform cap based on fiber grating sensing is provided, comprising the following steps:
[0009] A parameter optimization model for fiber Bragg grating strain sensors is established, and the weighted least squares algorithm is used to weight the sampling points to obtain the optimized response value and the optimal structural parameters of the sensor.
[0010] The deformation time series data of the platform cap is obtained by using fiber Bragg grating sensing with optimized structural parameters;
[0011] A damage identification model based on BP neural network is constructed, deformation time series data is used as input of the damage identification model, and damage identification results of the platform cap are output to the user end;
[0012] In response to the feedback result of the user end, the status result of the platform cap is corrected to obtain the damage identification result of the platform cap;
[0013] According to the difference between the damage appraisal result and the damage identification result, a correction algorithm is introduced to correct the parameters of the damage identification model, and the corrected damage identification model is used to perform damage identification on the input deformation time series data of the platform cap.
[0014] As a further limitation of the solution of the present invention, in the step of establishing a parameter optimization model of a fiber Bragg grating strain sensor, the parameter optimization model is expressed as:
[0015] (1);
[0016] Indicates the uniformity of sensor distribution; Indicates the maximum and minimum capacitance ratio indicators; Indicates constraints; It represents the angle between the light beam and the optical axis plane; and They represent the corresponding weights of the objective function, and are the structural parameters of the sensor respectively.
[0017] As a further limitation of the solution of the present invention, the step of using a weighted least squares algorithm to weight the sampling points to obtain an optimized response value and obtain the optimal structural parameters of the sensor includes:
[0018] Sampling of structural parameters and calculation of the response values of the sampled samples:
[0019] (2);
[0020] In formula (2), Represents the random distribution coefficient of sample data; represents the design distribution error; Represents the difference between sampled data and simulated data;
[0021] (3);
[0022] In formula (3), represents the function order; Represents the scale factor of the sampled data; Indicates the conversion coefficient between simulated data and known data; represents the sensor parameter optimization model;
[0023] The sampling data is weighted in combination with the sample response value;
[0024] And determine whether the least squares algorithm converges. If it converges, combine the linear function to output the optimal structural parameters of the sensor.
[0025] As a further limitation of the solution of the present invention, the criterion for judging whether the least squares algorithm has converged is: ;in, represents the response value of the kth generation optimal design, Indicates the preset convergence accuracy; Represents the sum of squares of the differences in data points; where: , and Represents the response value of the i-th and j-th sample data; , represents the sampled data fitting matrix.
[0026] As a further limitation of the solution of the present invention, in the step of outputting the optimal structural parameters of the sensor in combination with the linear function, the linear function is expressed as:
[0027] (4);
[0028] In formula (4), and are the sensor structural parameters to be optimized; Indicates the average value of the sampled values; represents a linear function; Represents the data point normalization coefficient.
[0029] As a further limitation of the solution of the present invention, the step of constructing a damage identification model based on a BP neural network includes:
[0030] Perform wavelet decomposition on the collected deformation time series data and map the data range to [1,2];
[0031] Extract the peak and valley values of the decomposed data and calculate the micro-strain sensed by the fiber Bragg grating sensor, and use the extreme value interpolation method to obtain the strain data envelope;
[0032] Smooth the strain data envelope, extract the low-frequency information of the first layer of wavelet decomposition, and perform exponential function curve fitting;
[0033] All strain data envelopes are used as the input set of the BP neural network. The Bayesian regularization method is selected to train the hidden layer of the BP neural network. During training, 80% of the input set is set as the training set, 10% as the validation set, and 10% as the test set. The number of training times is set to 1000 times, and the training is performed until the model converges to obtain a trained damage identification model.
[0034] In another embodiment of the present invention, a platform cap health monitoring system based on fiber grating sensing is provided, and the platform cap health monitoring system comprises:
[0035] The sensor parameter optimization module is used to establish a parameter optimization model for the fiber Bragg grating strain sensor, and adopts a weighted least squares algorithm to weight the sampling points to obtain the optimized response value and obtain the optimal structural parameters of the sensor;
[0036] A data acquisition module, used for acquiring deformation time series data of the platform cap by using fiber Bragg grating sensing with optimized structural parameters;
[0037] A damage identification module is used to construct a damage identification model based on a BP neural network, use the deformation time series data as the input of the damage identification model, and output the damage identification result of the platform cap to the user end;
[0038] A result correction module, used for correcting the status result of the platform cap in response to the feedback result of the user terminal, and obtaining the damage identification result of the platform cap;
[0039] The model correction module is used to introduce a correction algorithm to correct the parameters of the damage identification model according to the difference between the damage appraisal result and the damage identification result, and use the corrected damage identification model to perform damage identification on the input deformation time series data of the platform cap.
[0040] In yet another embodiment of the present invention, a computer device is provided, comprising:
[0041] At least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the platform cap health monitoring method based on fiber grating sensing.
[0042] In yet another embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the platform cap health monitoring method based on fiber grating sensing.
[0043] Compared with the prior art, the platform cap health monitoring method and system based on fiber grating sensing of the present invention have the following beneficial effects:
[0044] First, by establishing a parameter optimization model for the fiber Bragg grating strain sensor and using the weighted least squares algorithm to weight the sampling points to obtain the optimized response value, the optimal structural parameters of the sensor are obtained to improve the measurement accuracy of the sensor; and by obtaining the optimal structural parameters of the sensor, the fiber Bragg grating strain sensor can better adapt to the monitoring needs of the platform cap, improve the measurement sensitivity, and be able to continuously and accurately monitor the deformation of the platform cap in a complex environment;
[0045] Secondly, the present invention can efficiently identify the damage of the platform cap from the deformation time series data by constructing a damage identification model based on the BP neural network, thereby improving the efficiency of damage detection, and the automated damage identification process reduces the dependence on professionals and reduces the labor cost;
[0046] Thirdly, the present invention corrects the status result of the platform cap in response to the feedback result of the user end, and obtains the damage identification result of the platform cap, which is helpful to discover and correct the possible problems of the model, thereby improving the accuracy of damage identification;
[0047] Fourthly, the present invention can quickly adapt to new changes, such as changes in data characteristics caused by environmental factors, material aging, etc., by correcting the parameters of the damage identification model; by integrating the feedback results of managers, it reduces the misjudgment that may occur when relying solely on machine learning; each iteration is an optimization of the model, ensuring the accuracy and stability of long-term operation. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention.
[0049] Figure 1 This is a system structure diagram of a platform cap health monitoring method based on fiber grating sensing according to the present invention;
[0050] Figure 2 This is a flow chart of a method for monitoring the health of a platform cap based on fiber grating sensing according to the present invention;
[0051] Figure 3 It is a sub-flow chart of the platform cap health monitoring method based on fiber grating sensing of the present invention;
[0052] Figure 4 This is a structural block diagram of a platform cap health monitoring system based on fiber grating sensing according to the present invention. DETAILED DESCRIPTION
[0053] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0054] At present, the platform cap health monitoring system needs to monitor key indicators such as the structural deformation and crack development of the platform cap in real time and continuously to ensure that potential safety hazards can be discovered in time.
[0055] To this end, the system architecture design of the platform cap health monitoring method provided in the embodiment of the present disclosure includes a fiber grating sensor array, a signal demodulator, a data transmission unit and a data processing and analysis module; wherein the sensor array is responsible for collecting the strain data of the platform cap; the signal demodulator converts the optical signal into an electrical signal; the data transmission unit transmits the electrical signal to the central processing system; the data processing and analysis module is responsible for data storage, analysis and early warning; and provides an intuitive monitoring platform for operators.
[0056] The principle of the fiber grating sensor to measure the deformation or cracking of the platform cap is:
[0057] The broadband light source emits light with a certain width, which is incident on the grating of the fiber Bragg grating strain sensor through the fiber coupler. The incident light is affected by the wavelength selectivity of the grating, and the light that meets the conditions is reflected and enters the fiber demodulator again through the fiber coupler. The change of strain distribution on the surface of the object under test will cause the central wavelength of the reflected light to change. By analyzing the offset of the central wavelength of the reflected light, the change data of the deformation or cracking of the platform cap can be obtained.
[0058] Furthermore, in the system architecture, with respect to the fiber grating sensor layout scheme, FBG sensors are reasonably arranged according to the structural characteristics and usage environment of the platform cap to cover all key monitoring points; the sensors should be arranged along the length direction of the platform cap and densely arranged in areas where cracks may occur to ensure that tiny structural changes can be accurately captured.
[0059] The data transmission adopts wired or wireless mode to ensure the stability and security of data transmission.
[0060] The following is a detailed description of the specific implementation of the platform cap health monitoring method and system based on fiber grating sensing of the present invention in conjunction with specific embodiments.
[0061] like Figure 1As shown, it is a system structure diagram provided by the present disclosure. In the platform cap health monitoring method based on fiber Bragg grating sensing provided by the present disclosure, the structural deformation time series data of the platform cap is obtained by fiber Bragg grating sensing. After the deformation time series data is processed, on the one hand, it is used to add to the training set; on the other hand, the processed deformation time series data is input into the damage identification model as the deformation data to be measured; the training set is used to train the BP neural network model, and then obtain the damage identification model for identifying the health status of the platform cap;
[0062] Furthermore, based on the damage identification result of the platform cap, corresponding maintenance measures are taken for the platform cap.
[0063] The present invention can automatically analyze the monitoring data obtained by fiber optic Bragg grating sensing to confirm the condition of the platform cap, can quickly identify the damage of the platform cap structure reflected in the data, can provide timely warning, and has high reliability.
[0064] like Figure 2 As shown, in one embodiment of the present disclosure, a platform cap health monitoring method based on fiber grating sensing is provided, and the monitoring method comprises the following steps:
[0065] Step S1: Establish a parameter optimization model for the fiber Bragg grating strain sensor, use a weighted least squares algorithm to weight the sampling points to obtain an optimized response value, and obtain the optimal structural parameters of the sensor;
[0066] Specifically, in the embodiment of the present disclosure, in the step of establishing a parameter optimization model for a fiber Bragg grating strain sensor, the parameter optimization model is expressed as:
[0067] (1);
[0068] In formula (1), Indicates the uniformity of sensor distribution; Indicates the maximum and minimum capacitance ratio indicators; Indicates constraints; It represents the angle between the light beam and the optical axis plane; and They represent the corresponding weights of the objective function, and are the structural parameters of the sensor respectively.
[0069] Furthermore, in the embodiment of the present disclosure, the sensitivity of the conventional fiber Bragg grating is low and the measurement accuracy cannot be guaranteed. Therefore, the parameters of the sensor need to be optimized. Specifically, the weighted least squares algorithm is used to weight the sampling points to obtain the optimized response value, and the step of obtaining the optimal structural parameters of the sensor includes:
[0070] Sampling of structural parameters and calculation of the response values of the sampled samples:
[0071] (2);
[0072] In formula (2), Represents the random distribution coefficient of sample data; represents the design distribution error; Represents the difference between sampled data and simulated data;
[0073] (3);
[0074] In formula (3), represents the function order; Represents the scale factor of the sampled data; Indicates the conversion coefficient between simulated data and known data; represents the sensor parameter optimization model;
[0075] Furthermore, the present disclosure also performs weighted processing on the sampled data in combination with the sample response value;
[0076] And determine whether the least squares algorithm converges. If it converges, combine the linear function to output the optimal structural parameters of the sensor.
[0077] Among them, in the embodiment of the present disclosure, the criterion for judging whether the least squares algorithm converges is: ;in, represents the response value of the kth generation optimal design, Indicates the preset convergence accuracy; Represents the sum of squares of the data point differences;
[0078] in: , and represents the response value of the i-th and j-th sample data; further, , represents the sampled data fitting matrix.
[0079] Further, in the embodiment of the present disclosure, in the step of outputting the optimal structural parameters of the sensor in combination with the linear function, the linear function is expressed as:
[0080] (4);
[0081] In formula (4), and are the sensor structural parameters to be optimized; Indicates the average value of the sampled values; represents a linear function; Represents the data point normalization coefficient.
[0082] Optionally, in the process summary of weighted processing of the sampled data in the present invention, a larger weight is given to the data points in the middle section, and a smaller weight is given to the sampled data points at the beginning and end of the sampling.
[0083] For further information, please refer to Figure 2 The platform cap health monitoring method based on fiber grating sensing in the embodiment of the present disclosure further includes:
[0084] Step S2: using the fiber Bragg grating sensor with optimized structural parameters to obtain the deformation time series data of the platform cap;
[0085] Specifically, in the disclosed embodiment, the fiber Bragg grating strain sensor with optimized parameters is installed at the key parts of the platform cap, such as the supporting structure, beam or the area with concentrated force. During the installation process, it is necessary to ensure that the sensor is well coupled with the platform cap structure to ensure that the sensor can accurately capture the tiny deformation. After installation, the wavelength change of the reflected light from the sensor is continuously monitored by a dedicated demodulator and converted into strain data;
[0086] Furthermore, the strain data collected by the present invention is analyzed through time series to form deformation time series data of the platform cap. These data reflect the deformation of the platform cap at different time points and can be used to analyze the structural health status of the platform cap and warn of potential damage.
[0087] Furthermore, the health monitoring method of the embodiment of the present disclosure also includes:
[0088] Step S3: constructing a damage identification model based on BP neural network, taking the deformation time series data as the input of the damage identification model, and outputting the damage identification result of the platform cap to the user end;
[0089] For details, please refer to Figure 3 In the embodiment of the present disclosure, the steps of step S3 specifically include:
[0090] Step S31: performing wavelet decomposition on the collected deformation time series data, and mapping the data range to [1,2];
[0091] Step S32: extracting the peak and valley values of the decomposed data and calculating the microstrain sensed by the fiber grating sensor, and obtaining the strain data envelope using the extreme value interpolation method;
[0092] Step S33: smoothing the strain data envelope, extracting the low-frequency information of the first layer of wavelet decomposition and performing exponential function curve fitting;
[0093] Step S34: All strain data envelopes are used as the input set of the BP neural network, and the Bayesian regularization method is selected to train the hidden layer of the BP neural network. During training, 80% of the input set is set as the training set, 10% as the validation set, and 10% as the test set; the number of training times is set to 1000, and the training is performed until the model converges to obtain a trained damage identification model.
[0094] After the model training is completed, it needs to be tested on the test set to evaluate the damage identification accuracy of the model. At this stage, cross-validation can be used to optimize the model parameters and avoid overfitting.
[0095] Furthermore, in the disclosed embodiment, the deformation time series data that needs to be damaged is used as the input of the trained damage identification model, and the damage identification result of the platform cap is output to the user end.
[0096] Furthermore, before the deformation time series data is input into the damage identification model, data preprocessing is first required, and the preprocessing includes cleaning the data to remove noise and outliers, standardizing or normalizing the data, and filling in missing values to ensure data integrity.
[0097] Preferably, the BP neural network provided by the present disclosure is composed of an input layer, a plurality of hidden layers and an output layer;
[0098] Among them, the input layer receives the extracted feature data, the hidden layer processes this data, and finally the output layer generates the damage identification results.
[0099] Among them, the damage results of the platform cap include deformation, cracks, etc.; the present disclosure is mainly used to detect the degree of deformation and cracks, and further determine whether maintenance is needed; the degree includes: normal, minor damage, general damage and serious damage; for normal results, no maintenance is required; for minor damage and general damage, the management personnel will evaluate; for serious damage, the strategy of notifying maintenance is directly adopted.
[0100] The deformation is manifested as sinking, expansion or other forms of deformation of the plastering layer or structural layer of the platform cap; the analysis involves the use of steel formwork concrete with smooth surface and no scratch treatment for the base layer during the original construction, high mud content in the plastering cement, and failure to maintain the plastering surface in time or as required;
[0101] The cracks are caused by improper setting of expansion joints, tensile cracks or bursting cracks; or environmental changes and large temperature differences may cause expansion and contraction of the material, thereby causing cracks.
[0102] For further information, please refer to Figure 2 , the health monitoring method of the embodiment of the present disclosure further includes:
[0103] Step S4: in response to the feedback result of the user end, the status result of the platform cap is corrected to obtain the damage identification result of the platform cap;
[0104] Step S5: According to the difference between the damage appraisal result and the damage identification result, a correction algorithm is introduced to correct the parameters of the damage identification model, and the corrected damage identification model is used to perform damage identification on the input deformation time series data of the platform cap.
[0105] Among them, in step S4, the present disclosure provides the results of the damage identification model to the management personnel, so as to facilitate the management personnel to evaluate the damage identification structure of the platform cap to determine whether it is necessary to take measures for the deformation, cracks, etc. produced in the platform cap, wherein the status of the platform cap given by the management personnel based on the damage identification results is returned to the damage identification model as a damage identification result.
[0106] Furthermore, in step S5, based on the damage identification result, the embodiment of the present disclosure further modifies the parameters of the damage identification model. Specifically, an iterative feedback network is adopted to integrate the feedback of the management personnel into the training process of the model through multiple rounds of iterations to modify the parameters of the damage identification model.
[0107] Among them, according to the damage identification results fed back by the management personnel, the weight and bias parameters of the BP neural network are adjusted; the BP neural network is retrained using the adjusted parameters;
[0108] Furthermore, each time there is a difference between the damage appraisal result and the damage identification result, the damage appraisal result and the damage identification result are synchronously saved; the difference between the damage appraisal result and the damage identification result is quantified, and the weight and bias parameters of the BP neural network are adjusted.
[0109] The present invention can quickly adapt to new changes, such as changes in data characteristics caused by environmental factors, material aging, etc., by correcting the parameters of the damage identification model; by integrating the feedback results of managers, it reduces the misjudgment that may occur when relying solely on machine learning; each iteration is an optimization of the model, ensuring the accuracy and stability of long-term operation.
[0110] Please refer to Figure 4 In another embodiment of the present invention, a platform cap health monitoring system based on fiber grating sensing is provided, and the platform cap health monitoring system comprises:
[0111] The sensor parameter optimization module 601 is used to establish a parameter optimization model of the fiber Bragg grating strain sensor, and adopt a weighted least squares algorithm to weight the sampling points to obtain an optimized response value, so as to obtain the optimal structural parameters of the sensor;
[0112] The data acquisition module 602 is used to acquire the deformation time series data of the platform cap by using the fiber Bragg grating sensor with optimized structural parameters;
[0113] The damage identification module 603 is used to construct a damage identification model based on a BP neural network, use the deformation time series data as the input of the damage identification model, and output the damage identification result of the platform cap to the user end;
[0114] A result correction module 604 is used to correct the status result of the platform cap in response to the feedback result of the user end, and obtain the damage identification result of the platform cap;
[0115] The model correction module 605 is used to introduce a correction algorithm to correct the parameters of the damage identification model according to the difference between the damage appraisal result and the damage identification result, and use the corrected damage identification model to perform damage identification on the input deformation time series data of the platform cap.
[0116] The disclosed platform cap health monitoring method and system based on fiber Bragg grating sensing combines fiber Bragg grating strain sensors to obtain deformation, cracks and other damage data of the platform cap, combines the advantages of data analysis and machine learning models, and has high accuracy and real-time performance. In specific implementation, it is also necessary to adjust the parameters and algorithm selection of the damage identification model according to different equipment and application scenarios to improve the adaptability and effect of the system.
[0117] In this embodiment, the platform cap health monitoring system based on fiber grating sensing adopts the steps of the aforementioned platform cap health monitoring method based on fiber grating sensing during execution. Therefore, the operation process of the platform cap health monitoring system based on fiber grating sensing will not be introduced in detail in this embodiment.
[0118] The fiber grating sensing-based platform cap health monitoring method and system of the present invention establish a parameter optimization model based on the fiber grating strain sensor, adopt a weighted least squares algorithm to weight the sampling points to obtain the optimized response value, and obtain the optimal structural parameters of the sensor; use the fiber grating sensor with structural parameter tuning to obtain the deformation time series data of the platform cap; construct a damage identification model based on the BP neural network, use the deformation time series data as the input of the damage identification model, and output the damage identification result of the platform cap to the user end; in response to the feedback result of the user end, correct the status result of the platform cap to obtain the damage identification result of the platform cap; according to the difference between the damage identification result and the damage identification result, introduce a correction algorithm to correct the parameters of the damage identification model, and use the corrected damage identification model to identify the damage of the input deformation time series data of the platform cap, which greatly improves the structural safety and maintenance efficiency of the platform cap and has significant economic and social benefits.
[0119] In one embodiment, a computer device is provided in an embodiment of the present invention, comprising at least one processor, and a memory communicatively connected to the at least one processor;
[0120] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the steps of the platform cap health monitoring method based on fiber grating sensing.
[0121] In one embodiment, the present invention further provides a computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable the computer to execute the steps of the platform cap health monitoring method based on fiber grating sensing.
[0122] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be realized by instructing the relevant hardware through a computer program represented by computer instructions, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory.
[0123] Non-volatile memory may include read-only memory, magnetic tape, floppy disk, flash memory or optical storage, etc. Volatile memory may include random access memory or external cache memory. As an illustration and not limitation, RAM may be in various forms, such as static random access memory or dynamic random access memory, etc.
[0124] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A platform cap health monitoring method based on fiber grating sensing, characterized in that: The platform cap health monitoring method based on fiber grating sensing includes the following steps: A parameter optimization model for fiber Bragg grating strain sensors is established, and the weighted least squares algorithm is used to weight the sampling points to obtain the optimized response value and the optimal structural parameters of the sensor. The deformation time series data of the platform cap is obtained by using fiber Bragg grating sensing with optimized structural parameters; A damage identification model based on BP neural network is constructed, deformation time series data is used as input of the damage identification model, and damage identification results of the platform cap are output to the user end; In response to the feedback result of the user end, the status result of the platform cap is corrected to obtain the damage identification result of the platform cap; According to the difference between the damage appraisal result and the damage identification result, a correction algorithm is introduced to correct the parameters of the damage identification model, and the corrected damage identification model is used to perform damage identification on the input deformation time series data of the platform cap; In the step of establishing the parameter optimization model of the fiber Bragg grating strain sensor, the parameter optimization model is expressed as: (1); In formula (1), Indicates the uniformity of sensor distribution; Indicates the maximum and minimum capacitance ratio indicators; Indicates constraints; It represents the angle between the light beam and the optical axis plane; and They represent the corresponding weights of the objective function, and are the sensor structure parameters respectively; The steps of weighting the sampling points by using the weighted least squares algorithm to obtain the optimized response value and obtain the optimal structural parameters of the sensor include: Sampling of structural parameters and calculation of the response values of the sampled samples: (2); In formula (2), Represents the random distribution coefficient of sample data; represents the design distribution error; Represents the difference between sampled data and simulated data; (3); In formula (3), represents the function order; Represents the scale factor of the sampled data; Indicates the conversion coefficient between simulated data and known data; represents the sensor parameter optimization model; The sampling data is weighted in combination with the sample response value; And determine whether the least squares algorithm converges. If it converges, combine the linear function to output the optimal structural parameters of the sensor.
2. The method for monitoring the health of a platform cap based on fiber grating sensing according to claim 1 is characterized in that: The criterion for judging whether the least squares algorithm has converged is: ;in, represents the response value of the kth generation optimal design, Indicates the preset convergence accuracy; Represents the sum of squares of the data point differences; where: , and Represents the response value of the i-th and j-th sample data; , represents the sampled data fitting matrix.
3. The method for monitoring the health of a platform cap based on fiber grating sensing according to claim 2 is characterized in that: In the step of combining the linear function to output the optimal structural parameters of the sensor, the linear function is expressed as: (4); In formula (4), and are the sensor structural parameters to be optimized; Indicates the average value of the sampled values; represents a linear function; Represents the data point normalization coefficient.
4. The method for monitoring the health of a platform cap based on fiber grating sensing according to claim 3 is characterized in that: The steps of constructing a damage identification model based on BP neural network include: Perform wavelet decomposition on the collected deformation time series data and map the data range to [1,2]; Extract the peak and valley values of the decomposed data and calculate the micro-strain sensed by the fiber Bragg grating sensor, and use the extreme value interpolation method to obtain the strain data envelope; Smooth the strain data envelope, extract the low-frequency information of the first layer of wavelet decomposition, and perform exponential function curve fitting; All strain data envelopes are used as the input set of the BP neural network, and the Bayesian regularization method is selected to train the hidden layer of the BP neural network until the model converges to obtain a trained damage identification model.
