Interlayer composite vibration isolation structure performance evaluation method based on vibration feedback
By generating parameters through resonance analysis and classification fault analysis, optimizing sensor deployment, and using machine learning to generate reinforcement strategies, the problem of insufficient adaptiveness of sensor deployment strategies is solved, and efficient vibration isolation performance evaluation and real-time monitoring are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ANHUI HIGHWAY BRIDGE ENG CO LTD
- Filing Date
- 2026-04-28
- Publication Date
- 2026-06-09
Smart Images
Figure CN122174678A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of building structure analysis technology, and more specifically, to a method for evaluating the performance of inter-story composite vibration isolation structures based on vibration feedback. Background Technology
[0002] In modern engineering structures, floor-to-floor vibration isolation structures are widely used to isolate the effects of environmental vibrations on the superstructure. The core performance of composite vibration isolation structures relies on the coordinated work of two distinct types of supports: steel spring supports provide continuous flexible support and vibration isolation, while limiting supports act as safety boundaries, preventing excessive displacement under extreme loads. To ensure the long-term safe and reliable operation of such critical systems, it is essential to construct a sensor monitoring network capable of real-time sensing of their operational status and performance evaluation.
[0003] Currently, in this technical field, traditional sensor network design and performance evaluation methods mainly suffer from the following technical defects and shortcomings: 1. Sensor deployment is disconnected from system functional objectives; 2. Sensor hardening strategies are static, lack adaptability, and are not cost-conscious.
[0004] To address the above problems, this invention proposes a solution. Summary of the Invention
[0005] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a method for evaluating the performance of interlayer composite vibration isolation structures based on vibration feedback. The present invention generates a sensor layout set by optimizing the sensor layout based on a first parameter, and generates a reinforcement strategy by machine learning based on the sensor layout set and a second parameter, thereby solving the problem of insufficient adaptive sensor layout strategy in the performance evaluation of interlayer composite vibration isolation structures.
[0006] To achieve the above objectives, the present invention provides the following technical solution: The vibration feedback-based performance evaluation method for interlayer composite vibration isolation structures includes the following steps: acquiring the mechanical data of the composite vibration isolation structure, generating a first parameter through resonance analysis, and generating a second parameter through classification fault analysis; based on the first parameter, generating a sensor deployment set by optimizing the sensor layout, and generating a reinforcement strategy for each sensor deployment node through machine learning based on the sensor deployment set and the second parameter, forming a failure-resistant sensor network; acquiring real-time monitoring data of the failure-resistant sensor network, and obtaining a quantitative evaluation result of the vibration isolation performance through transmissibility and boundary state analysis.
[0007] In a preferred embodiment, the composite vibration isolation structure includes a steel spring support and a limiting support; the first parameter is the comprehensive modal contribution of the support; and the second parameter is the support-enhanced failure risk parameter.
[0008] In a preferred embodiment, acquiring the mechanical data of the composite vibration isolation structure and generating a first parameter through resonance analysis includes: performing modal analysis on the pre-constructed dynamic model of the composite vibration isolation structure to acquire the natural frequencies and mode shape data of multiple modes, and extracting the modal displacement data corresponding to each support; calculating the principal mode participation based on the position of each support by weighting the modal displacement data with the natural frequencies; calculating the coherence function based on the support response frequency and the response frequency of the superstructure, and averaging the coherence function within the preset vibration isolation key frequency band to obtain the vibration isolation coherence; and weightedly fusing the principal mode participation and vibration isolation coherence to obtain the first parameter.
[0009] In a preferred embodiment, the step of generating the second parameter through classification fault analysis includes: acquiring historical operation and maintenance data of the composite vibration isolation structure; constructing the support type parameters of the steel spring support based on the static compression and load parameters; constructing the support type parameters of the limiting support based on the historical impact count and limiting gap; and integrating the support type parameters, environmental data, and acquired historical operation and maintenance data to calculate the second parameter.
[0010] In a preferred embodiment, the step of generating a sensor deployment set by optimizing the sensor deployment based on the first parameter includes: constructing a sensor deployment model with the objective function of maximizing the weighted sum of the first parameter and the decision variable, using whether or not to deploy sensors on candidate supports as decision variables; setting total budget constraints, steel spring support coverage constraints, and limit support coverage constraints; and solving the sensor deployment model using a hierarchical sequence method under the constraints to obtain the sensor deployment set.
[0011] In a preferred embodiment, the step of generating a hardening strategy for each sensor deployment node based on the sensor deployment set and a second parameter, thereby forming a failure-resistant sensor network, includes: constructing a multi-dimensional feature vector for each sensor node in the sensor deployment set, combined with historical operation and maintenance data; the multi-dimensional feature vector includes a first parameter, a second parameter, a support type parameter, and a node degree; inputting the multi-dimensional feature vector into a pre-trained machine learning model, which outputs a hardening strategy label; performing a corresponding hardening operation on each sensor node according to the hardening strategy label; and forming the final failure-resistant sensor network after completing the hardening operations on all nodes.
[0012] In a preferred embodiment, the step of inputting the multidimensional feature vector into a pre-trained machine learning model and outputting reinforcement strategy labels includes: labeling the strategy based on the multidimensional feature vector and historical operation and maintenance data; using the multidimensional feature vector as input, predicting the reinforcement strategy through the machine learning model; providing bias feedback based on the strategy labels, and training the machine learning model to minimize the prediction bias.
[0013] In a preferred embodiment, the step of performing corresponding hardening operations on each sensor node according to the hardening strategy label includes: the hardening strategy label is a discrete classification label; when the hardening strategy label is algorithm fault tolerance hardening, configuring a data verification and reconstruction algorithm for the sensor node; when the hardening strategy label is homogeneous backup hardening, adding a sensor of the same type near the sensor node; when the hardening strategy label is heterogeneous fusion hardening, adding a sensor based on different physical principles for the sensor node.
[0014] In a preferred embodiment, the step of acquiring real-time monitoring data from the anti-failure sensor network and obtaining a quantitative evaluation result of vibration isolation performance through transmissibility and boundary state analysis includes: acquiring the displacement spectrum amplitudes of predetermined key points of the superstructure and steel spring supports based on the anti-failure sensor network, and calculating the single-point displacement transmissibility of each steel spring support; weighting and fusing the single-point displacement transmissibility with the corresponding support comprehensive modal contribution to obtain the system-level transmissibility; averaging the system-level transmissibility based on the key vibration isolation frequency band to obtain the system-level vibration isolation transmissibility; calculating the limit support activation state index based on the activation duration of the limit support within the observation period and the total observation duration; comparing the system-level vibration isolation transmissibility with a preset vibration isolation performance threshold, and comparing the limit support activation state index with a preset safety state threshold to generate a quantitative evaluation result of vibration isolation performance; the quantitative evaluation result of vibration isolation performance includes performance level determination and abnormal warning information.
[0015] The technical effects and advantages of the vibration feedback-based method for evaluating the performance of interlayer composite vibration isolation structures in this invention are as follows: This invention generates a first parameter through resonance analysis and a second parameter through fault classification analysis, directly anchoring the optimization goal of sensor deployment to efficiently acquiring vibration isolation performance evaluation data, thus achieving the adaptability of sensor deployment strategy data. Based on the first parameter, a sensor deployment set is generated by optimizing the sensor deployment, achieving the response of sensor deployment strategy adaptive data. Based on the sensor deployment set and the second parameter, a reinforcement strategy for each sensor deployment node is generated through machine learning, which can intelligently predict the failure risk of each sensor node and adjust the reinforcement strategy accordingly, improving the system's tolerance to faults or failures and solving the problem of insufficient adaptive sensor deployment strategy in the performance evaluation of interlayer composite vibration isolation structures. Attached Figure Description
[0016] Figure 1 This is a schematic diagram of the performance evaluation method for interlayer composite vibration isolation structures based on vibration feedback provided in an embodiment of the present invention. Detailed Implementation
[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0018] Example 1, Figure 1 The present invention provides a method for evaluating the performance of interlayer composite vibration isolation structures based on vibration feedback, comprising the following steps: S1. Obtain the mechanical data of the composite vibration isolation structure, generate the first parameter through resonance analysis, and generate the second parameter through classification fault analysis; S2. Based on the first parameter, a sensor deployment set is generated by optimizing the sensor deployment. Based on the sensor deployment set and the second parameter, a hardening strategy for each sensor deployment node is generated through machine learning to form a failure-resistant sensor network. S3. Obtain real-time monitoring data from the anti-failure sensor network, and obtain quantitative evaluation results of vibration isolation performance through transmissibility and boundary state analysis.
[0019] This embodiment generates a first parameter through resonance analysis and a second parameter through fault classification analysis, enabling the optimization goal of sensor deployment to be directly anchored to the efficient acquisition of vibration isolation performance evaluation data, thus achieving the adaptability of sensor deployment strategy data. Based on the first parameter, a sensor deployment set is generated by optimizing the sensor deployment, realizing the response of sensor deployment strategy adaptive data. Based on the sensor deployment set and the second parameter, a reinforcement strategy for each sensor deployment node is generated through machine learning, which can intelligently predict the failure risk of each sensor node and adjust the reinforcement strategy accordingly, improving the system's tolerance to faults or failures and solving the problem of insufficient adaptive sensor deployment strategy in the performance evaluation of interlayer composite vibration isolation structures.
[0020] S1. Obtain the mechanical data of the composite vibration isolation structure, generate the first parameter through resonance analysis, and generate the second parameter through classification fault analysis.
[0021] In this embodiment, the composite vibration isolation structure includes a steel spring support and a limiting support; the first parameter is the comprehensive modal contribution of the support; the second parameter is the support-enhanced fault risk parameter.
[0022] In this embodiment, obtaining the mechanical data of the composite vibration isolation structure and generating the first parameter through resonance analysis includes: Modal analysis was performed on the dynamic model of the pre-constructed composite vibration isolation structure to obtain the natural frequencies and mode shapes of multiple modes, and the modal displacement data corresponding to each support were extracted. Based on each support location, the modal displacement data and natural frequency are weighted to calculate the participation of the principal vibration mode; Based on the support response frequency and the superstructure response frequency, the coherence function is calculated, and the average value of the coherence function within the preset key frequency band of vibration isolation is obtained to obtain the vibration isolation coherence. The first parameter is obtained by weighted fusion of the main mode participation and the vibration isolation coherence.
[0023] This embodiment uses resonance analysis to closely link the vibration isolation effect with the dynamic response of the superstructure.
[0024] It should be noted that the natural frequency refers to the inherent dynamic characteristics of the composite vibration isolation structure, which is used to indicate the inherent vibration characteristics of the structure under conditions without external excitation. The modal energy order is determined by the natural frequency.
[0025] It should be noted that the participation of the main vibration mode is used to characterize the response intensity of a certain support location in several dominant vibration modes of the composite vibration isolation structure. It reflects the degree of participation of the support in the overall dynamic behavior of the structure. The higher the participation, the more significant the influence of the support on the overall vibration response of the composite vibration isolation structure.
[0026] In this embodiment, the formula for calculating the participation of the principal vibration mode is as follows:
[0027] In the formula, For support Participation of the principal mode, This is the upper limit of the modal order. For modal weights, and , For the first First-order modal energy, For the first First mode in support Modal displacement components at that location, This indicates the position of the support.
[0028] In this embodiment, the specific formula for calculating vibration isolation coherence is as follows:
[0029]
[0030] In the formula, For support The vibration isolation coherence, For the critical frequency band of vibration isolation, For the critical frequency band of vibration isolation, For support With respect to the superstructure reference point at frequency coherence function, For signal and cross-spectral density, For signal and cross-spectral density, For signal and The cross-spectral density.
[0031] In this embodiment, the formula for calculating the comprehensive modal contribution of the support is as follows:
[0032] In the formula, For the contribution of the support to the overall modal characteristics, To preset weights, For support Participation of the principal mode, For support The vibration isolation coherence.
[0033] It should be noted that the vibration isolation coherence is used to characterize the degree of influence of the dynamic response of the support position on the response of the superstructure within the critical frequency band of vibration isolation.
[0034] It should be noted that the critical frequency band for vibration isolation refers to the frequency range that has a decisive impact on the vibration isolation performance evaluation results during the operation of the vibration isolation system and needs to be monitored and analyzed in detail.
[0035] In this embodiment, generating the second parameter through fault classification analysis includes: Obtain historical operation and maintenance data for composite vibration isolation structures; The support type parameters of the steel spring support are constructed based on the static compression and load parameters. The bearing type parameters of the limiting bearing are constructed based on the historical number of impacts and the limiting gap; The second parameter is obtained by integrating the support type parameter, environmental data, and acquired historical operation and maintenance data.
[0036] This embodiment constructs support type parameters through classification fault analysis, and assesses the failure risk of the support by combining environmental data and historical operation and maintenance data, thereby enhancing the system's ability to predict future failures.
[0037] In this embodiment, the formula for the support type parameter is as follows:
[0038]
[0039] For the first Parameters for each type of steel spring support For the first Each limit support type parameter , , , To preset weights, This is the current static compression amount. To design static compression, This represents the actual vertical load. To design vertical loads, For the number of historical impacts, For reference, the number of impacts This is the current limit gap. To design the limiting clearance.
[0040] It should be noted that static compression refers to the amount of compressive deformation that occurs when the support is subjected to a static load. It should be noted that the limiting gap refers to the functional non-contact distance between the limiting support and the limited structure.
[0041] In this embodiment, the specific formula for calculating the support-enhanced failure risk parameter is as follows:
[0042] In the formula, For the support enhancement failure risk parameters, , , , To preset weights, These are the normalized environmental parameters. The normalized historical fault index, Mean time between failures (MTBF) For support type parameters, It can be 1 or 2.
[0043] S2. Based on the first parameter, a sensor deployment set is generated by optimizing the sensor deployment. Based on the sensor deployment set and the second parameter, a hardening strategy for each sensor deployment node is generated through machine learning to form a failure-resistant sensor network.
[0044] In this embodiment, generating a sensor deployment set by optimizing the sensor deployment based on the first parameter includes: Using whether or not to deploy sensors on candidate supports as decision variables, a sensor deployment model is constructed with the objective function of maximizing the weighted sum of the first parameter and the decision variables. And set total budget constraints, steel spring support coverage constraints, and limit support coverage constraints; Under constraints, the sensor deployment model is solved using the hierarchical sequence method to obtain the sensor deployment set.
[0045] This embodiment generates a sensor deployment set by optimizing the sensor deployment, taking into account not only performance indicators but also budget and constraints, thereby improving the economy and operability of implementation.
[0046] In this embodiment, the objective function formula is as follows:
[0047] In the formula, Let be the objective function. To define the decision variables, 1 represents the position of the support. Install sensors, Contribution to the overall modal characteristics of the support.
[0048] In this embodiment, the total budget constraint is as follows:
[0049] In the formula, For support The deployment cost This is the upper limit of the total budget.
[0050] In this embodiment, the steel spring support covering constraint is specifically as follows:
[0051] In the formula, This is a collection of steel spring supports.
[0052] In this embodiment, the limiting support covering constraint is specifically as follows:
[0053] In the formula, This is a set of limit supports.
[0054] It should be noted that the hierarchical sequence method is an existing technical means, and will not be described in detail in this embodiment.
[0055] In this embodiment, the step of generating a hardening strategy for each sensor deployment node through machine learning based on the sensor deployment set and the second parameter to form a failure-resistant sensor network includes: For each sensor node in the sensor deployment set, a multi-dimensional feature vector is constructed by combining historical operation and maintenance data. The multi-dimensional feature vector includes a first parameter, a second parameter, a support type parameter, and a node degree. Input multidimensional feature vectors into a pre-trained machine learning model and output reinforcement policy labels; Based on the hardening strategy label, perform the corresponding hardening operation for each sensor node; After all nodes have been hardened, the final failure-resistant sensor network is formed.
[0056] In this embodiment, the step of inputting multi-dimensional feature vectors into a pre-trained machine learning model and outputting reinforcement strategy labels includes: Based on multidimensional feature vectors, policy labels are assigned according to historical operation and maintenance data; Using multidimensional feature vectors as input, a machine learning model is used to predict reinforcement strategies. Bias feedback is based on policy labels, and the machine learning model is trained to minimize prediction bias.
[0057] This embodiment improves the sensor network's resilience by developing differentiated hardening strategies for each sensor node using a machine learning model, enabling the system to operate stably for a long time.
[0058] In this embodiment, performing the corresponding hardening operation on each sensor node according to the hardening strategy label includes: The reinforcement strategy label is a discrete classification label; When the reinforcement strategy label is algorithm fault tolerance reinforcement, configure the data verification and reconstruction algorithm for the sensor node; When the hardening strategy is labeled as homogeneous backup hardening, add a sensor of the same type near the sensor node. When the hardening strategy is labeled as heterogeneous fusion hardening, sensors based on different physical principles are added to the sensor node.
[0059] This embodiment improves the fault tolerance and reliability of the sensor network by flexibly adjusting different types of hardening strategies according to the actual needs of the sensor nodes.
[0060] It should be noted that the node degree represents the number of direct connections a sensor node has with other sensor nodes in the network.
[0061] It should be noted that the machine learning model and training process are existing technologies, and will not be described in detail in this embodiment.
[0062] S3. Obtain real-time monitoring data from the anti-failure sensor network, and obtain quantitative evaluation results of vibration isolation performance through transmissibility and boundary state analysis.
[0063] In this embodiment, the acquisition of real-time monitoring data from the failure-resistant sensor network, and the obtaining of a quantitative evaluation result of vibration isolation performance through transmissibility and boundary state analysis, includes: Based on the failure-resistant sensor network, the displacement spectrum amplitudes of the predetermined superstructure key points and steel spring supports are obtained, and the single-point displacement transfer rate of each steel spring support is calculated. The system-level transfer rate is obtained by weighting and fusing the displacement transfer rate of each single point with the corresponding comprehensive modal contribution of the support. Based on the key frequency band of vibration isolation, the system-level transmissibility is averaged by frequency band to obtain the system-level vibration isolation transmissibility. The activation state index of the limit support is calculated based on the activation duration of the limit support within the observation period and the total observation duration. The system-level vibration isolation transmissivity is compared with the preset vibration isolation performance threshold, and the limit support activation state index is compared with the preset safety state threshold to generate a quantitative evaluation result of vibration isolation performance. The quantitative evaluation results of vibration isolation performance include performance level determination and abnormal warning information.
[0064] In this embodiment, when the activation state index of the limit support is greater than the safety state threshold, the system performance level is immediately determined to be the activation state of the limit device, and the highest level safety alarm is triggered. When the system-level vibration isolation transmissibility is greater than the vibration isolation performance threshold, the system performance level is judged to be a decrease in vibration isolation effectiveness. When the system-level vibration isolation transmissibility is less than or equal to the vibration isolation performance threshold, the system performance level is judged as excellent, and a performance warning is triggered.
[0065] It should be noted that the key points of the superstructure refer to one or more specific locations in the building structure above the vibration isolation layer that are the most sensitive or important for vibration control and performance evaluation.
[0066] In this embodiment, the formula for calculating the single-point displacement transmissibility is as follows:
[0067] In the formula, For single-point displacement transmissivity, The key point of the superstructure is frequency displacement spectrum amplitude, For steel spring supports at frequency The displacement spectrum amplitude.
[0068] In this embodiment, the specific formula for calculating the system-level transfer rate is as follows:
[0069] In the formula, For system-level transfer rate, Contribution to the overall modal characteristics of the support.
[0070] In this embodiment, the specific formula for calculating the system-level vibration isolation transmissibility is as follows:
[0071] In the formula, For system-level vibration isolation transmissibility, For the critical frequency band of vibration isolation, This refers to the critical frequency band width for vibration isolation.
[0072] In this embodiment, the formula for calculating the activation state index of the limiting support is as follows:
[0073] In the formula, The limit support activation state index, The cumulative duration of limit activation during the observation period. This represents the total observation time.
[0074] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.
[0075] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.
[0076] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0077] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.
[0078] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
[0079] In conclusion, 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 principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for evaluating the performance of interlayer composite vibration isolation structures based on vibration feedback, characterized in that, Includes the following steps: The mechanical data of the composite vibration isolation structure are obtained, the first parameter is generated through resonance analysis, and the second parameter is generated through classification fault analysis. Based on the first parameter, a sensor deployment set is generated by optimizing the sensor deployment. Based on the sensor deployment set and the second parameter, a hardening strategy for each sensor deployment node is generated through machine learning, forming a failure-resistant sensor network. Real-time monitoring data from the failure-resistant sensor network is acquired, and the vibration isolation performance is quantitatively evaluated through transmissibility and boundary state analysis.
2. The method for performance evaluation of interlayer composite vibration isolation structures based on vibration feedback according to claim 1, characterized in that, The composite vibration isolation structure includes steel spring supports and limiting supports; the first parameter is the comprehensive modal contribution of the supports; the second parameter is the support-enhanced failure risk parameter.
3. The method for performance evaluation of interlayer composite vibration isolation structures based on vibration feedback according to claim 2, characterized in that, The acquisition of mechanical data of the composite vibration isolation structure, and the generation of the first parameter through resonance analysis, includes: Modal analysis was performed on the dynamic model of the pre-constructed composite vibration isolation structure to obtain the natural frequencies and mode shapes of multiple modes, and the modal displacement data corresponding to each support were extracted. Based on each support location, the modal displacement data and natural frequency are weighted to calculate the participation of the principal vibration mode; Based on the support response frequency and the superstructure response frequency, the coherence function is calculated, and the average value of the coherence function within the preset key frequency band of vibration isolation is obtained to obtain the vibration isolation coherence. The first parameter is obtained by weighted fusion of the main mode participation and the vibration isolation coherence.
4. The method for performance evaluation of interlayer composite vibration isolation structures based on vibration feedback according to claim 3, characterized in that, The generation of the second parameter through fault classification analysis includes: Obtain historical operation and maintenance data for composite vibration isolation structures; The support type parameters of the steel spring support are constructed based on the static compression and load parameters. The bearing type parameters of the limiting bearing are constructed based on the historical number of impacts and the limiting gap; The second parameter is obtained by integrating the support type parameter, environmental data, and acquired historical operation and maintenance data.
5. The method for performance evaluation of interlayer composite vibration isolation structures based on vibration feedback according to claim 4, characterized in that, The process of generating a sensor deployment set based on the first parameter and optimizing the sensor deployment includes: Using whether or not to deploy sensors on candidate supports as decision variables, a sensor deployment model is constructed with the objective function of maximizing the weighted sum of the first parameter and the decision variables. And set total budget constraints, steel spring support coverage constraints, and limit support coverage constraints; Under constraints, the sensor deployment model is solved using the hierarchical sequence method to obtain the sensor deployment set.
6. The method for performance evaluation of interlayer composite vibration isolation structures based on vibration feedback according to claim 5, characterized in that, The method of generating a hardening strategy for each sensor deployment node based on the sensor deployment set and the second parameter through machine learning to form a failure-resistant sensor network includes: For each sensor node in the sensor deployment set, a multi-dimensional feature vector is constructed by combining historical operation and maintenance data. The multi-dimensional feature vector includes a first parameter, a second parameter, a support type parameter, and a node degree. Input multidimensional feature vectors into a pre-trained machine learning model and output reinforcement policy labels; Based on the hardening strategy label, perform the corresponding hardening operation for each sensor node; After all nodes have been hardened, the final failure-resistant sensor network is formed.
7. The method for performance evaluation of interlayer composite vibration isolation structures based on vibration feedback according to claim 6, characterized in that, The process of inputting multi-dimensional feature vectors into a pre-trained machine learning model and outputting reinforcement strategy labels includes: Based on multidimensional feature vectors, policy labels are assigned according to historical operation and maintenance data; Using multidimensional feature vectors as input, a machine learning model is used to predict reinforcement strategies. Bias feedback is based on policy labels, and the machine learning model is trained to minimize prediction bias.
8. The method for performance evaluation of interlayer composite vibration isolation structures based on vibration feedback according to claim 7, characterized in that, The step of performing corresponding hardening operations on each sensor node based on the hardening strategy label includes: The reinforcement strategy label is a discrete classification label; When the reinforcement strategy label is algorithm fault tolerance reinforcement, configure the data verification and reconstruction algorithm for the sensor node; When the hardening strategy is labeled as homogeneous backup hardening, add a sensor of the same type near the sensor node. When the hardening strategy is labeled as heterogeneous fusion hardening, sensors based on different physical principles are added to the sensor node.
9. The method for performance evaluation of interlayer composite vibration isolation structures based on vibration feedback according to claim 8, characterized in that, The acquisition of real-time monitoring data from the anti-failure sensor network, through transmissibility and boundary state analysis, yields a quantitative evaluation result of vibration isolation performance, including: Based on the failure-resistant sensor network, the displacement spectrum amplitudes of the predetermined superstructure key points and steel spring supports are obtained, and the single-point displacement transfer rate of each steel spring support is calculated. The system-level transfer rate is obtained by weighting and fusing the displacement transfer rate of each single point with the corresponding comprehensive modal contribution of the support. Based on the key frequency band of vibration isolation, the system-level transmissibility is averaged by frequency band to obtain the system-level vibration isolation transmissibility. The activation state index of the limit support is calculated based on the activation duration of the limit support within the observation period and the total observation duration. The system-level vibration isolation transmissivity is compared with the preset vibration isolation performance threshold, and the limit support activation state index is compared with the preset safety state threshold to generate a quantitative evaluation result of vibration isolation performance. The quantitative evaluation results of vibration isolation performance include performance level determination and abnormal warning information.