System for monitoring state of fiber composite structure

Abstract

The invention discloses a state monitoring system for a fiber composite structure, and relates to the technical field of fiber composite structures, and the state monitoring system comprises a data processing module, a state analysis module, a model prediction module and a monitoring early warning module. Through the state analysis module, key factors influencing the damage characteristics of the fiber composite structure can be quickly identified, the diagnosis efficiency of structural damage reasons is improved, the key factors are analyzed through the model prediction module, the damage trend of the fiber composite structure at the future moment can be predicted, and an early warning signal is sent out in time; therefore, the failure risk of the structure in future use is effectively avoided, the safety is improved, through the monitoring and early warning module, feedback is conducted based on early warning signals, the external working condition is actively adjusted in combination with the working state of the fiber composite structure, the service life of the fiber composite structure is prolonged, and the maintenance cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of fiber composite structures, and in particular to a system for monitoring the state of fiber composite structures. Background Art

[0002] Fiber composite materials are widely used in aerospace, wind power, rail transit, automobile and other fields due to their light weight, high stiffness, high modulus, high reliability and gradually reduced manufacturing cost. However, in these key application fields, the safety of fiber composite structures still faces huge challenges. Therefore, the condition monitoring of fiber composite structures has become a research focus. Structural health monitoring technology (SHM) is an efficient online condition monitoring method. It obtains sensor feedback signals (such as stress, strain, temperature, etc.) embedded in fiber composite structures in real time, and analyzes the information related to the structural state by combining advanced signal processing technology and modeling methods. SHM technology can identify the damage characteristics and integrity of fiber composite structures, and realize real-time evaluation of the structural health state. This condition monitoring technology not only improves the reliability of fiber composite structures under complex loads and harsh environments, but also provides data support for structural design and safety management, which is of great significance to the application safety in key fields.

[0003] In the existing technology, it is not easy to accurately extract the damage characteristic parameters of the fiber composite structure, and it is not easy to conduct in-depth analysis of the characteristic parameters, and it is not easy to quickly identify the key factors affecting the damage characteristics of the fiber composite structure, which reduces the efficiency of diagnosing the cause of structural damage and cannot provide a core basis for structural health management. At the same time, it is not easy to predict the damage trend of the fiber composite structure at future times, and it is not easy to issue early warning signals in time. Therefore, the failure risk of the structure in future use cannot be effectively avoided, which reduces safety.

[0004] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention proposes a system for condition monitoring of fiber composite structures, which solves the problems mentioned in the above background technology that the existing system is not convenient for accurately extracting damage characteristic parameters of fiber composite structures, and is not convenient for in-depth analysis of the characteristic parameters, and cannot quickly identify key factors affecting the damage characteristics of fiber composite structures, thereby reducing the efficiency of diagnosing the causes of structural damage and failing to provide a core basis for structural health management. At the same time, it is not convenient to predict the damage trend of the fiber composite structure at future times, and cannot issue early warning signals in time, thereby failing to effectively avoid the risk of failure of the structure in future use, thereby reducing safety issues.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A system for condition monitoring of a fiber composite structure, the system comprising: A data processing module is used to obtain real-time monitoring data of the fiber composite structure, pre-process the real-time monitoring data of the fiber composite structure, and extract damage characteristic parameters of the fiber composite structure; The state analysis module is used to analyze the damage characteristic parameters based on the damage characteristic parameters of the fiber composite structure using the damage identification analysis algorithm to identify the key factors affecting the damage characteristics of the fiber composite structure; The model prediction module is used to establish a damage identification model based on the key factors that affect the damage characteristics of the fiber composite structure, and use the damage identification model to warn of the damage trend of the fiber composite structure at future moments and issue a warning signal; A monitoring and early warning module is used to adjust the external working conditions of the fiber composite structure based on the issued early warning signal and in combination with the working state of the fiber composite structure, so as to optimize the use and maintenance of the fiber composite structure; Based on the damage characteristic parameters of the fiber composite structure, the damage characteristic parameters are analyzed using the damage identification analysis algorithm, and the key factors affecting the damage characteristics of the fiber composite structure are identified, including: Set the maximum number of iterations of the damage identification analysis algorithm and randomly generate several combinations of damage feature parameters; The correlation evaluation algorithm is used to conduct a preliminary correlation evaluation on each combination of damage characteristic parameters to obtain its conformity with the damage characteristics of the fiber composite structure; Check whether the new combination of damage characteristic parameters is within the feasible range. If not, adjust it back to the feasible range through mapping conditions. Otherwise, keep it unchanged. Randomly select several optimized damage characteristic parameter combinations for perturbation operation. If the perturbed combination exceeds the feasible range, adjust it back to the feasible range through mapping conditions. According to the probability selection principle, several optimized damage characteristic parameter combinations are selected as the initial set of the next generation, and the optimization range of the damage characteristic parameter combination is dynamically adjusted using a nonlinear function; It is determined whether the damage identification analysis algorithm has reached the maximum number of iterations. If not, a preliminary correlation evaluation is performed again. Otherwise, based on the optimization results, the key factors affecting the damage characteristics of the fiber composite structure are finally identified.

[0007] Furthermore, the data processing module is connected to the model prediction module through the state analysis module, and the model prediction module is connected to the monitoring and early warning module.

[0008] Furthermore, the data processing module includes: A data acquisition module is used to acquire real-time status data of the fiber composite structure through sensors and pre-process the status data; The feature extraction module is used to perform feature recognition on the preprocessed state data and extract damage feature parameters of the fiber composite structure.

[0009] Furthermore, the correlation evaluation algorithm is used to conduct a preliminary correlation evaluation on each combination of damage characteristic parameters, and the degree of conformity with the damage characteristics of the fiber composite structure is obtained, including: Initialize the parameters of the correlation evaluation algorithm, and generate an initial damage characteristic parameter group in the damage characteristic parameter space through a correlation screening mechanism, where each damage characteristic parameter item corresponds to a different combination of damage characteristic parameters; Calculate the correlation conformity value of each damage characteristic parameter item in the initial damage characteristic parameter group, and select the damage characteristic parameter item that best conforms to the damage characteristics of the fiber composite structure as the initial optimal damage characteristic parameter item according to the conformity value; According to the correlation evaluation optimization formula, the positions of all damage characteristic parameter items in the initial damage characteristic parameter group are updated to generate a new damage characteristic parameter combination group; Based on the exploration mechanism, an opposing damage characteristic parameter group is generated, and the damage characteristic parameter item with the highest conformity value is selected from the original damage characteristic parameter group and the opposing damage characteristic parameter group by using the screening algorithm to enter the next round of optimization; Check whether the association evaluation algorithm has reached the maximum number of iterations. If not, continue to optimize. Otherwise, select the damage characteristic parameter item with the highest compliance value from the final damage characteristic parameter combination group as the optimal damage characteristic parameter combination that matches the damage characteristics of the fiber composite structure.

[0010] Furthermore, the association evaluation optimization formula is: ; In the formula, H i，w Indicates i The damage characteristic parameter item is w The position of the damage feature parameter item after the dimension is updated; e and e All represent adjustment coefficients; A w express w The upper limit of the dimension; B w express w The lower limit of the dimension; h represents the distance function; f i Indicates i The current position of the damage characteristic parameter item; f j Indicates j The current position of the damage characteristic parameter item; Indicates the number of the current damage characteristic parameter item in the group w The global optimal damage characteristic parameter item of dimension; N Indicates the total number of current damage characteristic parameter items.

[0011] Furthermore, based on the exploration mechanism, an opposing damage characteristic parameter group is generated, and a screening algorithm is used to select the damage characteristic parameter item with the highest conformity value from the original damage characteristic parameter group and the opposing damage characteristic parameter group to enter the next round of optimization, including: A number of original damage characteristic parameter combination groups are randomly generated in the damage characteristic parameter space, and each damage characteristic parameter item corresponds to a potential damage characteristic solution; Calculate the correlation conformity value of each damage characteristic parameter item, and evaluate the quality of each damage characteristic parameter item; Based on the updating rule of the screening algorithm, the positions of other damage characteristic parameter items in the original damage characteristic parameter group are adjusted to gradually approach the current optimal damage characteristic parameter item; Based on the exploration mechanism, a group of opposing damage characteristic parameters is generated, and the correlation conformity value of each damage characteristic parameter item is calculated; if the conformity value of any damage characteristic parameter item in the group of opposing damage characteristic parameters is higher than the current optimal damage characteristic parameter item, it is updated to the new optimal damage characteristic parameter item; The damage characteristic parameter items with the highest conformity value are selected from the original damage characteristic parameter group and the opposing damage characteristic parameter group to enter the next round of optimization.

[0012] Furthermore, the update rule formula is: ; In the formula, Indicates e The damage characteristic parameter item is c +The position at the iteration; Indicates e The damage characteristic parameter item is c The position at the iteration; Indicated in c The position of the optimal damage characteristic parameter item at the iteration; α represents the learning factor; d Represents a random number.

[0013] Furthermore, the model prediction module includes: A model building module for building a damage identification model based on the identified key factors affecting the damage characteristics of fiber composite structures; The prediction optimization module is used to optimize the parameters of the damage identification model using a parameter optimization algorithm, and to warn of the damage trend of the fiber composite structure at a future moment based on the optimized damage identification model and issue a warning signal.

[0014] Furthermore, based on the identified key factors affecting the damage characteristics of fiber composite structures, a damage identification model is constructed including: Obtain the key factors that affect the damage characteristics of fiber composite structures, and divide the key factors into training sets and test sets; Based on the fusion learning model, any key factor is used as the main reference feature and input into the integrated learning model together with the remaining key factors to construct multiple damage identification model sub-models; The performance of all constructed damage identification model sub-models is evaluated on the training set, and the best performing damage identification model sub-model is selected for fusion to form a damage identification model.

[0015] Furthermore, the parameters of the damage identification model are optimized using a parameter optimization algorithm, and based on the optimized damage identification model, an early warning is given to the damage trend of the fiber composite structure at a future moment. The early warning signal is issued including the following steps: Analyze the correlation between key factors using parameter optimization algorithms and preliminarily generate key factor data subsets; Randomly select a number of data from the key factor data subset as a training set, and calculate the optimization gain index related to the damage identification model parameters; Filter out the optimal parameter combination and divide the key factor data subset into several subsets of equal size as the initial training set; According to the fiber composite structure damage identification model, the damage change probability of the core parameters under different structural states is calculated to generate the characteristic pattern of the damage state; The parameter optimization algorithm is used to further verify the consistency of the damage characteristic pattern and select the subset of key factors with the highest classification accuracy as the input data of the optimized damage identification model; Combined with the optimized damage identification model and the probability of damage change, the influence of core parameters on the damage trend of fiber composite structures is analyzed; if the pattern is inconsistent, the key factor data is included in the damage change matrix for correction; otherwise, new parameter optimization data is introduced to gradually improve the weight allocation rules of the damage identification model; The optimized damage identification model is used to predict the damage trend of fiber composite structures in the future and issue early warning signals.

[0016] The beneficial effects of the present invention are: 1. The present invention uses a state analysis module to conduct an in-depth analysis of characteristic parameters, which can quickly identify key factors that affect the damage characteristics of fiber composite structures, improve the diagnostic efficiency of the causes of structural damage, and provide a core basis for structural health management. By analyzing key factors through a model prediction module, the damage trend of the fiber composite structure at future moments can be predicted, and early warning signals can be issued in time, thereby effectively avoiding the risk of failure of the structure in future use and improving safety. Through the monitoring and early warning module, not only feedback is provided based on the early warning signal, but also the external working conditions are actively adjusted in combination with the working state of the fiber composite structure, thereby extending the service life of the fiber composite structure and reducing maintenance costs.

[0017] 2. The present invention performs correlation analysis and conformity calculation through an association evaluation algorithm, so that the system can accurately screen out the key factors that have the most significant impact on the damage of the fiber composite structure from the complex damage data, and provide efficient support for subsequent status monitoring and trend prediction. The mechanism of dynamically adjusting parameter combinations and mapping conditions enables the system to flexibly adapt to different fiber composite structure damage characteristics, while ensuring the high accuracy and stability of the analysis results.

[0018] 3. The present invention constructs multiple damage identification sub-models by integrating learning models, and integrates the sub-models with the best performance to form a final damage identification model, which significantly improves the recognition accuracy of damage characteristics of fiber composite structures, ensures that the system can accurately reflect the damage state of the structure, uses parameter optimization algorithms to analyze the correlation of key factors, screens the optimal parameter combination, and gradually adjusts the model parameters by optimizing the gain index, so that the model can run with optimal performance, thereby improving the analysis and identification efficiency of damage characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 4 is a principle block diagram of a system for condition monitoring of a fiber composite structure according to an embodiment of the present invention.

[0021] In the figure: 1. Data processing module; 2. State analysis module; 3. Model prediction module; 4. Monitoring and early warning module. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0023] In the description of the present invention, unless otherwise specified, the meaning of "plurality" is two or more. In addition, the terms "first", "second", "third", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0024] According to an embodiment of the present invention, a system for condition monitoring of a fiber composite structure is provided.

[0025] The present invention will now be further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to an embodiment of the present invention, a system for monitoring the condition of a fiber composite structure comprises: A data processing module is used to obtain real-time monitoring data of the fiber composite structure, pre-process the real-time monitoring data of the fiber composite structure, and extract damage characteristic parameters of the fiber composite structure; It should be explained that a variety of sensors are embedded in the fiber composite structure, such as optical fiber sensors, strain sensors, accelerometers, temperature sensors, etc. These sensors can collect physical signals (stress, strain, vibration, temperature, etc.) of the structure under different working conditions in real time.

[0026] Specifically, real-time monitoring data includes: 1) Mechanical parameters: Stress / strain data: reflects the degree of deformation inside the material.

[0027] Vibration data: Dynamic response of the structure (acceleration, displacement, velocity).

[0028] 2) Environmental parameters: Temperature data: The performance changes of materials at different ambient temperatures.

[0029] Humidity Data: Effect of external humidity on material properties.

[0030] 3) Structural status data: Crack / delamination signal: Detected by acoustic emission sensor or ultrasonic sensor.

[0031] Fiber break signal: The local damage location is obtained by the optical fiber sensor.

[0032] 4) Dynamic response data: changes in natural frequency, damping characteristics, etc., obtained through dynamic tests.

[0033] 5) Electrical property data (if it is a smart material): resistance, capacitance or conductivity change signal, which is used to detect changes inside the material.

[0034] Specifically, the damage characteristic parameters of the fiber composite structure include: 1) Strain characteristic parameters: Strain increment: The change in local strain value relative to normal working conditions.

[0035] Strain gradient: The difference in strain values ​​at different locations, characterizing local anomalies.

[0036] 2) Vibration characteristic parameters: Modal parameters: including changes in modal frequency, modal vibration shape and damping ratio.

[0037] Acceleration characteristics: peak value, root mean square value (RMS), etc. of the acceleration signal.

[0038] 3) Acoustic emission characteristic parameters: Acoustic emission count: The number of acoustic wave signals when cracks or delamination occur.

[0039] Acoustic emission energy: The intensity of the sound waves characterizes the degree of damage.

[0040] 4) Crack characteristic parameters: Crack length / depth: crack size parameter.

[0041] Crack Growth Rate: The rate at which a crack grows over time.

[0042] 5) Temperature characteristic parameters: local temperature anomalies or temperature gradients (related to thermal fatigue or delamination damage).

[0043] 6) Frequency domain characteristic parameters: Power Spectral Density (PSD): The energy distribution of a vibration signal in the frequency domain.

[0044] Frequency offset: The offset of the structural modal frequency relative to the reference value.

[0045] 7) Image characteristic parameters (non-contact monitoring): thermal gradient characteristics in thermal imaging data. Displacement field or strain field distribution of surface deformation.

[0046] 8) Health Index (HI): A health status score calculated by a health monitoring algorithm based on a variety of characteristic data, used to quantify the integrity of the structure.

[0047] The state analysis module is used to analyze the damage characteristic parameters based on the damage characteristic parameters of the fiber composite structure using the damage identification analysis algorithm to identify the key factors affecting the damage characteristics of the fiber composite structure; Specifically, the key factors affecting the damage characteristics of fiber composite structures include: 1) Key factors related to material properties: Fiber strength and stiffness: The strength and stiffness of fiber materials directly determine the load-bearing capacity of the composite structure. Fiber damage (such as breakage and wear) will seriously affect the overall performance.

[0048] Matrix properties: The toughness, elastic modulus and fatigue resistance of the matrix material have an important influence on the damage characteristics of the composite structure, such as matrix cracks or debonding.

[0049] Fiber-matrix interface bonding strength: Insufficient interface bonding performance may lead to interface debonding, which is one of the common damage modes of composite materials.

[0050] Lamination direction and thickness: Fibers in different lamination directions have significantly different bearing capacities for loads in different directions. Incorrect lamination design may cause local damage.

[0051] 2) Key factors related to mechanical loading: Stress concentration: Holes, edges or inclusions in composite materials can cause local stress concentrations, which can easily lead to cracks or fiber breakage.

[0052] Fatigue loading: Long-term alternating loading can lead to degradation of material properties, such as fatigue crack or delamination growth.

[0053] Impact load: Instantaneous high-intensity impact can cause fiber breakage, matrix destruction or delamination damage.

[0054] Multiaxial loading effect: The damage behavior of composite materials under multiaxial loading is complex, and loads in different directions will affect the damage mode.

[0055] 3) Environmental factors: Temperature effect: High temperature may cause matrix softening or decomposition, low temperature may cause brittle damage; temperature cycling can lead to thermal fatigue cracks.

[0056] Humidity effect: Moisture absorption causes matrix expansion and reduced interfacial bonding performance, further inducing damage.

[0057] Corrosion and aging: Long-term exposure to harsh environments (such as chemical corrosion and ultraviolet radiation) will cause material performance degradation.

[0058] Thermomechanical coupling effect: Under the combined action of temperature and mechanical load, the thermal expansion and contraction of the material may induce cracks or delamination.

[0059] 4) Structural characteristics: Geometry: Stress concentration is prone to occur at the edges and corners of complex structures, leading to damage.

[0060] Manufacturing defects: Defects during the manufacturing process such as pores, uneven fiber distribution or debonding between layers will accelerate damage.

[0061] Structural joints: Riveted, glued or bolted joints are weak areas of the composite structure and are prone to damage.

[0062] 5) Factors related to dynamic response: Natural frequency shift: The natural frequency of a fiber composite structure changes due to damage and can be used as an important damage indicator.

[0063] Changes in damping characteristics: An increase in the damping ratio may indicate crack or delamination damage in the structure.

[0064] Asymmetry in vibration response: Asymmetry can be a manifestation of localized damage or changes in mass distribution.

[0065] 6) Damage mode characteristics: Delamination damage: Delamination or delamination between fiber layers is a common damage mode, which is related to the ply orientation and interface strength.

[0066] Crack propagation path: The starting point and propagation direction of the crack are closely related to the stress distribution and material properties.

[0067] Fiber breakage and pullout: Fiber breakage or pullout from the matrix reflects changes in interfacial properties.

[0068] Matrix cracking: Matrix cracking under local high stress may induce overall damage extension.

[0069] 7) Factors related to signal characteristics: Stress / strain distribution: Abnormal changes in local stress and strain are often early signs of damage.

[0070] Acoustic emission signal: The characteristics of acoustic emission events such as frequency, amplitude and cumulative count reflect micro-damage processes such as fiber breakage and interface debonding.

[0071] Vibration modal characteristics: Changes in vibration modal frequency and shape are related to the location and severity of damage.

[0072] Thermal response characteristics: Local temperature anomalies detected by thermal imaging may reflect delamination or crack damage.

[0073] The model prediction module is used to establish a damage identification model based on the key factors that affect the damage characteristics of the fiber composite structure, and use the damage identification model to warn of the damage trend of the fiber composite structure at future moments and issue a warning signal; A monitoring and early warning module is used to adjust the external working conditions of the fiber composite structure based on the issued early warning signal and in combination with the working state of the fiber composite structure, so as to optimize the use and maintenance of the fiber composite structure; It needs to be explained that, based on the damage trend prediction data in the early warning signal (such as crack growth rate, delamination range, strain concentration area, etc.), the severity and development risk of the current damage of the fiber composite structure are evaluated; combined with the working status data of the fiber composite structure (such as current load, vibration intensity, temperature environment, etc.), the correlation between the damage risk and the working conditions is analyzed; if the damage risk is high, reduce the load intensity borne by the structure to avoid damage expansion; optimize the load distribution to reduce local stress concentration in the damaged area; if the early warning signal detects fatigue damage caused by vibration, adjust the operating frequency to avoid resonance; reduce the operating rate or dynamic response amplitude to alleviate vibration damage; if the temperature, humidity or corrosive environment aggravates the damage risk, improve the environmental conditions (such as reducing humidity, temperature or adding protective coatings); for structures in long-term high temperature environments, set up thermal insulation protection measures to reduce the risk of thermal fatigue damage; adjust the sensor layout density or sampling frequency according to the early warning signal, and conduct more intensive monitoring of high-risk areas; Provide real-time feedback to the monitoring system to verify whether the adjustment effect has significantly reduced the risk of damage expansion; determine maintenance priorities based on the severity of damage and the importance of the structure, and prioritize high-risk areas for maintenance; determine maintenance priorities based on the severity of damage and the importance of the structure, and prioritize high-risk areas for maintenance; optimize the long-term use and maintenance strategies of the structure in combination with historical damage data and early warning signals, such as adjusting the inspection cycle and replacing vulnerable parts in advance; set operating warning values ​​to avoid the structure from operating under overload or high-frequency conditions to prevent accelerated damage caused by overload; optimize the use strategy (such as reducing the workload and evenly distributing the pressure) based on the damage prediction results to delay the failure time of the structure; observe the damage expansion trend after the external conditions are adjusted, and evaluate the optimization effect through the changes in the damage characteristic parameters; update the health status score of the fiber composite structure in combination with the adjusted operating status and real-time monitoring data as a quantitative basis for the optimization results; and continuously iterate and adjust the external working conditions and maintenance strategies based on the evaluation results of the optimization effect.

[0074] Based on the damage characteristic parameters of the fiber composite structure, the damage characteristic parameters are analyzed using the damage identification analysis algorithm, and the key factors affecting the damage characteristics of the fiber composite structure are identified, including: Set the maximum number of iterations of the damage identification analysis algorithm and randomly generate several combinations of damage feature parameters; The correlation evaluation algorithm is used to conduct a preliminary correlation evaluation on each combination of damage characteristic parameters (i.e., the fireworks individual in the improved fireworks algorithm) to obtain its conformity with the damage characteristics of the fiber composite structure; Check whether the new damage characteristic parameter combination is within the feasible range (first perform mutation operation on the damage characteristic parameter combination). If not, adjust it back to the feasible range through mapping conditions (i.e., mapping rules). Otherwise, keep it unchanged. Randomly select several optimized damage characteristic parameter combinations for perturbation operation. If the perturbed combination exceeds the feasible range, adjust it back to the feasible range through mapping conditions. According to the probability selection principle (i.e., roulette principle), several optimized damage characteristic parameter combinations are selected as the initial set of the next generation (i.e., the initial group of the next generation of fireworks in the improved fireworks algorithm), and the optimization range of the damage characteristic parameter combination is dynamically adjusted using a nonlinear function (i.e., a nonlinear decreasing method) (i.e., the maximum explosion radius is nonlinearly decreased according to the number of iterations in the improved fireworks algorithm); It is determined whether the damage identification analysis algorithm has reached the maximum number of iterations. If not, a preliminary correlation evaluation is performed again. Otherwise, based on the optimization results, the key factors affecting the damage characteristics of the fiber composite structure are finally identified.

[0075] Specifically, the damage identification and analysis algorithm is an improved fireworks algorithm. Based on the analysis of the traditional fireworks algorithm, the explosion radius of the fireworks is adjusted by nonlinear decreasing to improve the optimization ability of the fireworks algorithm. The traditional fireworks algorithm is a swarm intelligence algorithm for global optimization solution with an explosion search mechanism. It has excellent performance and high efficiency in solving complex problems. The fireworks algorithm mainly includes four parts: explosion operator, mutation, mapping, and probability selection principle. The algorithm randomly selects several fireworks as the initial fireworks group, explodes and mutates each firework in the fireworks group, and maps the fireworks individuals that are outside the feasible domain after mutation so that they are still in the feasible domain. Finally, the individuals with the best fitness are retained, and several minus one fireworks individuals are selected through the probability selection principle (i.e., roulette principle) to form the next generation of fireworks initial group. Through iteration, the fireworks individuals are getting closer and closer to the global optimal solution.

[0076] It should be noted that, in order to facilitate understanding of the above technical solution of the present invention, the present invention will be described in detail below based on the damage characteristic parameters of the fiber composite structure in the actual process, using a damage identification analysis algorithm to analyze the damage characteristic parameters and identify the key factors affecting the damage characteristics of the fiber composite structure.

[0077] Step 1: Problem description: The real-time monitoring data of a fiber composite structure collected the following damage characteristic parameters (assuming a total of 8 characteristics): P 1: Strain concentration factor (between 0 and 1).

[0078] P 2: Peak value of vibration acceleration (0-10g).

[0079] P 3: Layering area (0-50mm²).

[0080] P 4: Crack length (0-20mm).

[0081] P 5: Temperature gradient (0-50℃ / m).

[0082] P 6: Humidity content (0-10%).

[0083] P 7: Modal frequency offset (0-5Hz).

[0084] P 8: Accumulated count of acoustic emission events (0-1000).

[0085] The goal is to identify the most important key factors affecting the damage characteristics of fiber composite structures from these parameters through algorithms.

[0086] Step 2: Damage identification and analysis: 1) Set the maximum number of iterations of the damage identification algorithm: The maximum number of iterations is set to 50, and 20 damage feature parameter combinations are initially generated.

[0087] Each combination is composed of random values ​​of feature parameters, and the initial combinations are shown in Table 1 (taking some examples).

[0088] Table 1: Initial combination diagram 2) Use the association evaluation algorithm to perform correlation analysis: The conformity of each characteristic parameter combination is calculated by the following formula: ; In the formula, oh n represents parameter weights; initial weights are assumed to be random values, for example: oh 1=0.2, oh 2=0.15, oh 3=0.1, oh 4=0.1, oh 5=0.15, oh 6=0.05, oh 7=0.1, oh 8=0.15.

[0089] Take the first combination as an example: F 1=0.2×0.25+0.15×5.4+0.1×30.2+0.1×12.1+0.15×20.3+0.05×4.5+0.1×2.3+0.15×500.

[0090] F 1=0.05+0.81+3.02+1.21+3.045+0.225+0.23+75=83.58.

[0091] Calculate the compliance of all combinations and select the combination with higher compliance.

[0092] 3) Check the feasibility of the characteristic parameter combination: Check whether each parameter value is within a reasonable range. For example, if P 2More than 10g or P 5 exceeds 50°C / m, the mapping condition is used to adjust it back to the feasible range.

[0093] Mapping formula: like P n > P max ,but P n = P max .

[0094] P n < P max ,but P n = P min .

[0095] Check the first combination: P 1=0.25 (reasonable).

[0096] P 2=5.4 (reasonable).

[0097] P 3=30.2 (reasonable).

[0098] P 4=12.1 (reasonable).

[0099] P 5=20.3 (reasonable).

[0100] P 6=4.5 (reasonable).

[0101] P7=2.3 (reasonable).

[0102] P 8=500 (reasonable).

[0103] If there are outliers, adjust them.

[0104] 4) Random perturbation and optimization: Perform perturbation operations on parameter combinations that meet the range. P 1Perturbation: ; In the formula, i represents the perturbation value (e.g. i =0.05).

[0105] like =0.25+0.05=0.3, if it is still within the range, keep it; if it is out of the range, adjust it back to a reasonable range.

[0106] 5) Select the next generation combination based on the probability selection principle: Select several combinations according to the probability distribution of compliance to enter the next generation: ; Compliance F 1=83.58, F 2=92.5, F Take 3=60.3 as an example: ; Combinations are selected according to probability to enter the next generation.

[0107] 6) Dynamically adjust the optimization range: Use nonlinear functions to narrow the optimization range and focus on searching areas with high compliance. For example: R new = R old ×0.9.

[0108] 7) Determine the end condition of iteration: If the maximum number of iterations (50 times) is reached, the iteration is stopped and the parameter combination with the highest degree of compliance is output as the result.

[0109] Step 3: Output key factors: After 50 iterations, the parameter combination with the highest compliance is output and its impact on damage is analyzed: Key factors for final identification: P 3 (layer area), P 4 (crack length), P 8 (Accumulated AE event count).

[0110] These factors have the most significant impact on the damage characteristics of fiber composite structures and are the focus of subsequent damage identification and trend prediction.

[0111] In this optional embodiment, the data processing module is connected to the model prediction module via the state analysis module, and the model prediction module is connected to the monitoring and early warning module.

[0112] In this optional embodiment, the data processing module includes: A data acquisition module is used to acquire real-time status data of the fiber composite structure through sensors and pre-process the status data; The feature extraction module is used to perform feature recognition on the preprocessed state data and extract damage feature parameters of the fiber composite structure.

[0113] Specifically, the data acquisition module is connected to the feature extraction module.

[0114] Specifically, the method for extracting damage characteristic parameters of the fiber composite structure includes: Time domain feature extraction: extract statistical features directly from the original time series signal.

[0115] Frequency domain feature extraction: The signal is converted from the time domain to the frequency domain through Fourier transform, and the following frequency domain features are extracted.

[0116] Spatial distribution feature extraction: For distributed sensors (such as fiber optic sensors or array sensors), spatial distribution features can be extracted.

[0117] Machine learning assisted feature extraction: principal component analysis (PCA), cluster analysis, autoencoder, etc.

[0118] In this optional embodiment, a preliminary correlation evaluation is performed on each combination of damage characteristic parameters using a correlation evaluation algorithm to obtain the degree of conformity with the damage characteristics of the fiber composite structure, including: Initialize the parameters of the correlation evaluation algorithm, and generate an initial damage characteristic parameter group in the damage characteristic parameter space through a correlation screening mechanism, where each damage characteristic parameter item corresponds to a different combination of damage characteristic parameters; Calculate the correlation conformity value of each damage characteristic parameter item in the initial damage characteristic parameter group, and select the damage characteristic parameter item that best conforms to the damage characteristics of the fiber composite structure as the initial optimal damage characteristic parameter item according to the conformity value; According to the correlation evaluation optimization formula, the positions of all damage characteristic parameter items in the initial damage characteristic parameter group are updated to generate a new damage characteristic parameter combination group; Based on the exploration mechanism, an opposing damage characteristic parameter group is generated, and the damage characteristic parameter item with the highest conformity value is selected from the original damage characteristic parameter group and the opposing damage characteristic parameter group by using the screening algorithm to enter the next round of optimization; Check whether the association evaluation algorithm has reached the maximum number of iterations. If not, continue to optimize. Otherwise, select the damage characteristic parameter item with the highest compliance value from the final damage characteristic parameter combination group as the optimal damage characteristic parameter combination that matches the damage characteristics of the fiber composite structure.

[0119] Specifically, the association evaluation algorithm is an improved grasshopper optimization algorithm. The traditional grasshopper optimization algorithm GOA is a newer natural heuristic algorithm that simulates the social behavior of insects such as grasshoppers. Pests such as grasshoppers will affect crop yields, and their entire life cycle consists of three stages: eggs, larvae and adults. In the larval stage, the movement characteristics of grasshoppers include jumping and columnar rotation flight, small step length, slow movement, and predation on vegetation on the path; while in the adult stage, grasshoppers migrate over a large range in groups, with large steps and suddenness. The present invention uses completely random chaotic tent mapping to initialize the population to improve the traversal and diversity of the population; uses a learning automaton to update the adjustment coefficient that determines the search direction to balance global search and local development; introduces a refraction opposition learning position update mechanism for improvement to avoid the algorithm from falling into a local optimum.

[0120] In this optional embodiment, the association evaluation optimization formula is: ; In the formula, H i，w Indicates i The damage characteristic parameter item is w The position of the damage feature parameter item after the dimension is updated; e 1 and e 2 indicates adjustment coefficient; A w express w The upper limit of the dimension; B w express w The lower limit of the dimension; h represents the distance function; f i Indicates i The current position of the damage characteristic parameter item; f j Indicates j The current position of the damage characteristic parameter item; Indicates the number of the current damage characteristic parameter item in the group w The global optimal damage characteristic parameter item of dimension; N Indicates the total number of current damage characteristic parameter items.

[0121] In this optional embodiment, generating an opposing damage characteristic parameter group based on an exploration mechanism, and using a screening algorithm to select the damage characteristic parameter item with the highest conformity value from the original damage characteristic parameter group and the opposing damage characteristic parameter group to enter the next round of optimization includes: A number of original damage characteristic parameter combination groups are randomly generated in the damage characteristic parameter space, and each damage characteristic parameter item corresponds to a potential damage characteristic solution; Calculate the correlation conformity value of each damage characteristic parameter item, and evaluate the quality of each damage characteristic parameter item; Based on the updating rule of the screening algorithm, the positions of other damage characteristic parameter items in the original damage characteristic parameter group are adjusted to gradually approach the current optimal damage characteristic parameter item; Based on the exploration mechanism (i.e., kernel extreme learning machine), a group of opposing damage characteristic parameters is generated, and the correlation conformity value of each damage characteristic parameter item is calculated; if the conformity value of any damage characteristic parameter item in the group of opposing damage characteristic parameters is higher than the current optimal damage characteristic parameter item, it is updated to the new optimal damage characteristic parameter item; The damage characteristic parameter items with the highest conformity value are selected from the original damage characteristic parameter group and the opposing damage characteristic parameter group to enter the next round of optimization.

[0122] Specifically, the screening algorithm is a black hole optimization algorithm, and the idea of ​​the algorithm relatively completely describes the general characteristics of the black hole phenomenon in nature; based on these characteristics of the black hole optimization algorithm, the optimization search principle of the black hole optimization algorithm is relatively simple and easy to implement. The black hole optimization algorithm mainly simulates the actual black hole phenomenon, randomly arranges a certain number of stars in a certain search space, and determines and evaluates the fitness function of each star in the search space through statistical means, and selects a star with the best fitness value as the black hole. The boundary of this black hole is regarded as the area where the current global optimal solution is located, and the black hole itself is regarded as the current global optimal solution.

[0123] In this optional embodiment, the update rule formula is: ; In the formula, Indicates e The damage characteristic parameter item is c +1 position at iteration; Indicates e The damage characteristic parameter item is c The position at the iteration; Indicated in c The position of the optimal damage characteristic parameter item at the iteration; α represents the learning factor; dRepresents a random number.

[0124] In this optional embodiment, the model prediction module includes: A model building module for building a damage identification model based on the identified key factors affecting the damage characteristics of fiber composite structures; The prediction optimization module is used to optimize the parameters of the damage identification model using a parameter optimization algorithm, and to warn of the damage trend of the fiber composite structure at a future moment based on the optimized damage identification model and issue a warning signal.

[0125] Specifically, the model building module and the prediction optimization module are connected.

[0126] In this optional embodiment, based on the identified key factors affecting the damage characteristics of the fiber composite structure, constructing a damage identification model includes: Obtain the key factors that affect the damage characteristics of fiber composite structures, and divide the key factors into training sets and test sets; Based on the fusion learning model, any key factor is used as the main reference feature and input into the integrated learning model together with the remaining key factors to construct multiple damage identification model sub-models; The performance of all constructed damage identification model sub-models is evaluated on the training set, and the best performing damage identification model sub-model is selected for fusion to form a damage identification model.

[0127] It should be explained that through preliminary analysis, the key factors affecting the damage characteristics of fiber composite structures are determined, such as delamination area, crack length, strain gradient, vibration frequency offset, cumulative count of acoustic emission events, etc. These key factors are parameters closely related to damage extracted from a large amount of monitoring data; the historical monitoring data are divided into training sets and test sets according to a certain ratio, where the training set is used for model training and the test set is used to verify the performance of the model. When dividing, it is necessary to ensure that the distribution of data is as consistent as possible to avoid overfitting or underfitting of the model; use ensemble learning methods, such as random forests or gradient boosting decision trees, as the basic algorithm for modeling. Ensemble learning methods can combine the advantages of multiple models to improve recognition accuracy; for each key factor, it is used as the main reference feature and input into the model together with other factors. The design idea of ​​each sub-model is to focus on a key factor while considering the influence of other factors, so as to establish multiple different sub-models; use the training set data to train each sub-model so that it can learn the relationship between the key factor and the damage characteristics of the fiber composite structure. During the training process, the parameters of the model need to be continuously adjusted to ensure that the sub-model can accurately capture the damage characteristics; the performance of each sub-model on the training set is evaluated, and a series of indicators (such as accuracy, error, etc.) are used to measure the recognition effect of the model. The purpose of this step is to find the best performing sub-model; based on the evaluation results, the sub-model with the best performance is selected as the core of the subsequent model fusion. This selection requires a comprehensive consideration of the recognition accuracy and stability of the model; the output results of multiple sub-models are fused to form a more accurate and stable damage recognition model. Common fusion methods include weighted averaging, voting decisions, or stacking with more advanced meta-learning models; during the model fusion process, the weights of each sub-model can be dynamically adjusted according to its performance. The better the sub-model, the greater its influence in the final model, thereby improving the accuracy of recognition; the fused damage recognition model is applied to the test set data to evaluate its performance on unseen data. The performance of the model on the test set can reflect its generalization ability and actual application effect; if the performance of the model on the test set is poor, it is necessary to readjust the model parameters, data division or fusion strategy until the model achieves the expected effect; integrate the trained damage identification model into the monitoring system so that it can receive the monitoring data of key factors in real time and output the damage identification results or prediction information.

[0128] In this optional embodiment, the parameters of the damage identification model are optimized by using a parameter optimization algorithm, and based on the optimized damage identification model, an early warning is given for the damage trend of the fiber composite structure at a future moment, and issuing an early warning signal includes the following steps: Analyze the correlation between key factors using parameter optimization algorithms and preliminarily generate key factor data subsets; Randomly select a number of data from the key factor data subset as a training set, and calculate the optimization gain index related to the damage identification model parameters; Filter out the optimal parameter combination and divide the key factor data subset into several subsets of equal size as the initial training set; According to the fiber composite structure damage identification model, the damage change probability of the core parameters under different structural states is calculated to generate the characteristic pattern of the damage state; The parameter optimization algorithm is used to further verify the consistency of the damage characteristic pattern and select the subset of key factors with the highest classification accuracy as the input data of the optimized damage identification model; Combined with the optimized damage identification model and the probability of damage change, the influence of core parameters on the damage trend of fiber composite structures is analyzed; if the pattern is inconsistent, the key factor data is included in the damage change matrix for correction; otherwise, new parameter optimization data is introduced to gradually improve the weight allocation rules of the damage identification model; The optimized damage identification model is used to predict the damage trend of fiber composite structures in the future and issue early warning signals.

[0129] Specifically, the parameter optimization algorithm is the Apriori algorithm, which is a classic association rule mining algorithm used to discover frequent itemsets and association rules from a data set. The algorithm is based on the important property that all subsets of frequent itemsets are also frequent, which effectively reduces the search space.

[0130] In summary, with the help of the above technical solutions of the present invention, the present invention performs correlation analysis and conformity calculation through the association evaluation algorithm, so that the system can accurately screen out the key factors that have the most significant impact on the damage of the fiber composite structure from the complex damage data, and provide efficient support for subsequent state monitoring and trend prediction. The mechanism of dynamically adjusting parameter combinations and mapping conditions enables the system to flexibly adapt to different fiber composite structure damage characteristics, while ensuring the high accuracy and stability of the analysis results. The present invention constructs multiple damage identification sub-models through a fusion learning model, and fuses the best performing sub-models to form a final damage identification model, which significantly improves the recognition accuracy of the damage characteristics of the fiber composite structure, ensures that the system can accurately reflect the damage state of the structure, uses a parameter optimization algorithm to analyze the correlation of key factors, screens the optimal parameter combination, and gradually adjusts the model parameters by optimizing the gain index, so that the model can run with optimal performance, and improves the analysis and recognition efficiency of damage characteristics.

[0131] 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, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A system for condition monitoring of a fiber composite structure, characterized in that: include: Data processing module; The state analysis module is used to analyze the damage characteristic parameters based on the damage characteristic parameters of the fiber composite structure using the damage identification analysis algorithm to identify the key factors affecting the damage characteristics of the fiber composite structure; The model prediction module is used to establish a damage identification model based on the key factors that affect the damage characteristics of the fiber composite structure, and use the damage identification model to warn of the damage trend of the fiber composite structure at future moments and issue a warning signal; Monitoring and early warning module; The damage characteristic parameters based on the fiber composite structure and analyzing the damage characteristic parameters using a damage identification analysis algorithm include: Set the maximum number of iterations of the damage identification analysis algorithm and randomly generate several combinations of damage feature parameters; The correlation evaluation algorithm is used to conduct a preliminary correlation evaluation on each damage characteristic parameter combination to obtain its conformity with the damage characteristics of the fiber composite structure.

2. A system for condition monitoring of a fiber composite structure according to claim 1, characterized in that: The key factors that affect the damage characteristics of fiber composite structures include: Check whether the new combination of damage characteristic parameters is within the feasible range. If not, adjust it back to the feasible range through mapping conditions. Otherwise, keep it unchanged. Randomly select several optimized damage characteristic parameter combinations for perturbation operation. If the perturbed combination exceeds the feasible range, adjust it back to the feasible range through mapping conditions. According to the probability selection principle, several optimized damage characteristic parameter combinations are selected as the initial set of the next generation, and the optimization range of the damage characteristic parameter combination is dynamically adjusted using a nonlinear function; It is determined whether the damage identification analysis algorithm has reached the maximum number of iterations. If not, a preliminary correlation evaluation is performed again. Otherwise, based on the optimization results, the key factors affecting the damage characteristics of the fiber composite structure are finally identified.

3. A system for condition monitoring of a fiber composite structure according to claim 1, characterized in that: The data processing module comprises: A data acquisition module is used to acquire real-time status data of the fiber composite structure through sensors and pre-process the status data; The feature extraction module is used to perform feature recognition on the preprocessed state data and extract damage feature parameters of the fiber composite structure.

4. A system for condition monitoring of a fiber composite structure according to claim 1, characterized in that: The use of the correlation evaluation algorithm to perform a preliminary correlation evaluation on each combination of damage characteristic parameters to obtain the degree of conformity with the damage characteristics of the fiber composite structure includes: Initialize the parameters of the correlation evaluation algorithm, and generate an initial damage characteristic parameter group in the damage characteristic parameter space through a correlation screening mechanism, where each damage characteristic parameter item corresponds to a different combination of damage characteristic parameters; Calculate the correlation conformity value of each damage characteristic parameter item in the initial damage characteristic parameter group, and select the damage characteristic parameter item that best conforms to the damage characteristics of the fiber composite structure as the initial optimal damage characteristic parameter item according to the conformity value; According to the correlation evaluation optimization formula, the positions of all damage characteristic parameter items in the initial damage characteristic parameter group are updated to generate a new damage characteristic parameter combination group; Based on the exploration mechanism, an opposing damage characteristic parameter group is generated, and the damage characteristic parameter item with the highest conformity value is selected from the original damage characteristic parameter group and the opposing damage characteristic parameter group by using the screening algorithm to enter the next round of optimization; Check whether the association evaluation algorithm has reached the maximum number of iterations. If not, continue to optimize. Otherwise, select the damage characteristic parameter item with the highest compliance value from the final damage characteristic parameter combination group as the optimal damage characteristic parameter combination that matches the damage characteristics of the fiber composite structure.

5. A system for condition monitoring of a fiber composite structure according to claim 4, characterized in that: The association evaluation optimization formula is: ； In the formula, H i，w Indicates i The damage characteristic parameter item is w The position of the damage feature parameter item after the dimension is updated; ε 1 and ε 2 indicates adjustment coefficient; A w express w The upper limit of the dimension; B w express w The lower limit of the dimension; h represents the distance function; f i Indicates i The current position of the damage characteristic parameter item; f j Indicates j The current position of the damage characteristic parameter item; Indicates the number of the current damage characteristic parameter item in the group w The global optimal damage characteristic parameter item of dimension; N Indicates the total number of current damage characteristic parameter items.

6. A system for condition monitoring of a fiber composite structure according to claim 5, characterized in that: The generation of opposing damage characteristic parameter groups based on the exploration mechanism and the use of a screening algorithm to select the damage characteristic parameter items with the highest conformity value from the original damage characteristic parameter group and the opposing damage characteristic parameter group to enter the next round of optimization include: A number of original damage characteristic parameter combination groups are randomly generated in the damage characteristic parameter space, and each damage characteristic parameter item corresponds to a potential damage characteristic solution; Calculate the correlation conformity value of each damage characteristic parameter item, and evaluate the quality of each damage characteristic parameter item; Based on the updating rule of the screening algorithm, the positions of other damage characteristic parameter items in the original damage characteristic parameter group are adjusted to gradually approach the current optimal damage characteristic parameter item; Based on the exploration mechanism, a group of opposing damage characteristic parameters is generated, and the correlation conformity value of each damage characteristic parameter item is calculated; if the conformity value of any damage characteristic parameter item in the group of opposing damage characteristic parameters is higher than the current optimal damage characteristic parameter item, it is updated to the new optimal damage characteristic parameter item; The damage characteristic parameter items with the highest conformity value are selected from the original damage characteristic parameter group and the opposing damage characteristic parameter group to enter the next round of optimization.

7. A system for condition monitoring of a fiber composite structure according to claim 6, characterized in that: The update rule formula is: ； In the formula, Indicates e The damage characteristic parameter item is c +1 position at iteration; Indicates e The damage characteristic parameter item is c The position at the iteration; Indicated in c The position of the optimal damage characteristic parameter item at the iteration; α represents the learning factor; δ Represents a random number.

8. A system for condition monitoring of a fiber composite structure according to claim 1, characterized in that: The model prediction module includes: A model building module for building a damage identification model based on the identified key factors affecting the damage characteristics of fiber composite structures; The prediction optimization module is used to optimize the parameters of the damage identification model using a parameter optimization algorithm, and to warn of the damage trend of the fiber composite structure at a future moment based on the optimized damage identification model and issue a warning signal.

9. A system for condition monitoring of a fiber composite structure according to claim 8, characterized in that: The damage identification model is constructed based on the identified key factors affecting the damage characteristics of the fiber composite structure, including: Obtain the key factors that affect the damage characteristics of fiber composite structures, and divide the key factors into training sets and test sets; Based on the fusion learning model, any key factor is used as the main reference feature and input into the integrated learning model together with the remaining key factors to construct multiple damage identification model sub-models; The performance of all constructed damage identification model sub-models is evaluated on the training set, and the best performing damage identification model sub-model is selected for fusion to form a damage identification model.

10. A system for condition monitoring of a fiber composite structure according to claim 9, characterized in that: The method of optimizing the parameters of the damage identification model by using a parameter optimization algorithm, and issuing an early warning of the damage trend of the fiber composite structure at a future moment based on the optimized damage identification model, and issuing an early warning signal comprises the following steps: Analyze the correlation between key factors using parameter optimization algorithms and preliminarily generate key factor data subsets; Randomly select a number of data from the key factor data subset as a training set, and calculate the optimization gain index related to the damage identification model parameters; Filter out the optimal parameter combination and divide the key factor data subset into several subsets of equal size as the initial training set; According to the fiber composite structure damage identification model, the damage change probability of the core parameters under different structural states is calculated to generate the characteristic pattern of the damage state; The parameter optimization algorithm is used to further verify the consistency of the damage characteristic pattern and select the subset of key factors with the highest classification accuracy as the input data of the optimized damage identification model; Combined with the optimized damage identification model and the probability of damage change, the influence of core parameters on the damage trend of fiber composite structures is analyzed; if the pattern is inconsistent, the key factor data is included in the damage change matrix for correction; otherwise, new parameter optimization data is introduced to gradually improve the weight allocation rules of the damage identification model; The optimized damage identification model is used to predict the damage trend of fiber composite structures in the future and issue early warning signals.

Images