Sensor-based ultra-precision spring performance monitoring method and system

Abstract

The application relates to the technical field of performance monitoring, and provides a sensor-based ultra-precision spring performance monitoring method and system, which comprises the following steps: performing first decision layer decision: identifying a specific characteristic signal according to time characteristics and energy characteristics; the specific characteristic signal refers to a signal with energy higher than a preset energy value and duration less than a preset time length; the specific characteristic signal is determined as an external interference signal; performing second decision layer decision: for signals not determined as external interference signals in the first decision layer, a signal with specific frequency distribution and cumulative change characteristics is identified according to corresponding frequency characteristics and corresponding appearance frequency change trend and cumulative energy change trend, and the signal with specific frequency distribution and cumulative change characteristics is determined as an internal damage signal. The application has the effect of improving the accuracy and reliability of ultra-precision spring performance monitoring.

Description

Technical Field

[0001] This invention relates to the technical field of performance monitoring, specifically to a sensor-based method and system for monitoring the performance of ultra-precision springs. Background Technology

[0002] In the field of precision mechanics, especially in applications with extremely high reliability requirements, such as deep space probes, the performance monitoring of ultra-precision springs is crucial to ensuring mission success. Under extreme operating conditions, the microstructure of these springs may undergo subtle changes, leading to performance degradation or even failure. Existing monitoring technologies often face challenges in complex environments, struggling to accurately distinguish between genuine damage signals and external interference. This can lead to system misjudgments and, consequently, serious consequences.

[0003] For example, in deep space exploration missions, ultra-precision springs in the probe's antenna system are used to provide precise restoring force and absorb minute vibrations. Engineers monitor the spring's strain distribution and internal damage by deploying distributed fiber Bragg grating sensors and high-frequency acoustic emission sensors. In the early stages of the mission, environmental interference was minimal, and the monitoring system functioned well. However, as the probe entered the outer asteroid belt, it frequently encountered impacts from tiny dust particles and high-energy cosmic rays. These impacts caused transient vibrations in the probe's structure, which were transmitted to the springs, resulting in the acoustic emission sensors frequently receiving strong vibration signals similar to those from internal microcracks. This led to frequent misjudgments by the monitoring system, issuing false damage alarms.

[0004] To avoid interference from invalid alarms, ground control had to significantly increase the trigger threshold for acoustic emission signals. While this adjustment reduced false alarms, it also caused weak but critical signals that truly represented early damage to be filtered out as background noise. For example, acoustic emission signals generated by fatigue cracks slowly propagating in vacuum and cryogenic environments had amplitudes and energies far below the adjusted high threshold, and therefore went undetected by the system. Ultimately, the undetected microcracks continued to propagate, leading to the brittle fracture of a spring during a routine operation, a permanent deviation in antenna pointing, and the failure of the mission's core objective.

[0005] Existing technologies struggle to accurately distinguish and effectively suppress misjudgments caused by external interference signals when deep space probes encounter high-intensity, transient external disturbances (such as particle impacts and high-energy cosmic rays) without reducing sensitivity to real micro-damage signals, thus avoiding the loss of critical fault information due to simply raising the detection threshold.

[0006] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0007] This application discloses a sensor-based method and system for monitoring the performance of ultra-precision springs, aiming to solve the technical problems of existing technologies in accurately distinguishing between real damage signals and external interference in complex environments, leading to system misjudgment, and the potential loss of key fault information when increasing the detection threshold to reduce false alarms.

[0008] The technical solution of this application is as follows: In a first aspect, this application discloses a sensor-based method for monitoring the performance of ultra-precision springs, comprising: The acoustic emission signal generated by the ultra-precision spring is monitored, and the acoustic emission signal is initially captured based on a preset low energy threshold to obtain a segment of the initially captured acoustic emission signal. The initially captured acoustic emission signal segments are processed to convert them into data signal segments; Extracting multidimensional features from signal segments in data form; multidimensional features include time features, energy features, and frequency features; A hierarchical judgment process is performed based on multidimensional features. The hierarchical judgment process includes: performing a first judgment layer judgment: identifying specific feature signals based on time and energy features; specific feature signals refer to signals with energy higher than a preset energy value and duration less than a preset duration; judging specific feature signals as external interference signals; performing a second judgment layer judgment: for signals that were not judged as external interference signals in the first judgment layer, identifying signals with specific frequency distribution and cumulative change characteristics based on the corresponding frequency characteristics and the corresponding frequency change trend and cumulative energy change trend, and judging signals with specific frequency distribution and cumulative change characteristics as internal damage signals.

[0009] Through this technical solution, this application can effectively distinguish between external interference signals and internal damage signals through a layered judgment process, avoiding the problem of losing the real damage signal due to simply raising the threshold in traditional methods, thereby improving the accuracy and reliability of ultra-precision spring performance monitoring.

[0010] Secondly, this application also discloses a sensor-based ultra-precision spring performance monitoring system for performing sensor-based ultra-precision spring performance monitoring, comprising: The signal capture execution module is used to monitor the acoustic emission signal generated by the ultra-precision spring, and to perform preliminary capture of the acoustic emission signal based on a preset low energy threshold to obtain a preliminary captured acoustic emission signal segment; The signal processing execution module is used to process the initially captured acoustic emission signal segments to convert them into data signal segments; The multidimensional feature extraction module is used to extract multidimensional features from signal segments in data form; the multidimensional features include time features, energy features, and frequency features. The hierarchical judgment execution module is used to perform a hierarchical judgment process based on multi-dimensional features. The hierarchical judgment process includes: performing a first judgment layer judgment: identifying specific feature signals based on time and energy characteristics; specific feature signals refer to signals with energy higher than a preset energy value and duration less than a preset duration; judging specific feature signals as external interference signals; performing a second judgment layer judgment: for signals not judged as external interference signals in the first judgment layer, identifying signals with specific frequency distribution and cumulative change characteristics based on corresponding frequency characteristics and corresponding frequency change trends and cumulative energy change trends, and judging signals with specific frequency distribution and cumulative change characteristics as internal damage signals.

[0011] This application provides a hardware and software integrated system that uses modular design to capture, process, extract features, and determine the layers of acoustic emission signals, thereby effectively distinguishing between external interference and internal damage. This provides a complete solution for the performance monitoring of ultra-precision springs and solves the problems of misjudgment and information loss in the prior art.

[0012] Beneficial Effects: This application discloses a sensor-based method for monitoring the performance of ultra-precision springs. By monitoring the acoustic emission signals generated by the ultra-precision springs and initially capturing them based on a preset low-energy threshold, sensitivity to weak damage signals is ensured. Subsequently, the captured signals are processed and multi-dimensional features, including time features, energy features, and frequency features, providing a comprehensive data foundation for subsequent refined judgment. The key lies in its layered judgment process: the first judgment layer efficiently identifies and eliminates external interference signals with high energy and short duration based on time and energy features, effectively avoiding false alarms caused by simply raising the threshold in traditional methods. On this basis, the second judgment layer further combines frequency features, frequency change trends, and cumulative energy change trends to identify signals with specific frequency distributions and cumulative change characteristics, thereby accurately determining them as internal damage signals.

[0013] Through the above technical solution, this application overcomes the shortcomings of existing technologies in extremely complex environments such as deep space exploration, which struggle to accurately distinguish and effectively suppress misjudgments caused by external interference signals without reducing sensitivity to real micro-damage signals, thus avoiding the loss of critical fault information due to simply increasing the detection threshold. This method can significantly improve the accuracy and reliability of ultra-precision spring performance monitoring, effectively preventing system failures and mission failures caused by misjudgments, and providing strong support for the long-term stable operation of precision machinery under harsh conditions. Attached Figure Description

[0014] Figure 1This is a flowchart of a sensor-based ultra-precision spring performance monitoring method in one embodiment of the present invention; Figure 2 This is a flowchart of a sensor-based method for monitoring the performance of ultra-precision springs, as described in another embodiment of the present invention. Figure 3 This is a system block diagram of a sensor-based ultra-precision spring performance monitoring system according to another embodiment of the present invention; Explanation of reference numerals in the attached figures: 1. Sensor-based ultra-precision spring performance monitoring system; 11. Signal acquisition execution module; 12. Signal processing execution module; 13. Multi-dimensional feature extraction module; 14. Hierarchical judgment execution module. Detailed Implementation

[0015] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0016] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0017] This application proposes a sensor-based method for monitoring the performance of ultra-precision springs, combining... Figure 1 As shown, it includes: S1, monitor the acoustic emission signal generated by the ultra-precision spring, and perform preliminary capture of the acoustic emission signal based on a preset low energy threshold to obtain a preliminary captured acoustic emission signal segment; S2, performs signal processing on the initially captured acoustic emission signal segments to convert them into data signal segments; S3 extracts multidimensional features from signal segments in data form; the multidimensional features include time features, energy features, and frequency features. S4, perform a hierarchical judgment process based on multi-dimensional features; the hierarchical judgment process includes: performing a first judgment layer judgment: identifying specific feature signals based on time and energy features; specific feature signals refer to signals with energy higher than a preset energy value and duration less than a preset duration; judging specific feature signals as external interference signals; performing a second judgment layer judgment: for signals not judged as external interference signals in the first judgment layer, based on the corresponding frequency features and the corresponding frequency change trend and cumulative energy change trend, identifying signals with specific frequency distribution and cumulative change characteristics, and judging signals with specific frequency distribution and cumulative change characteristics as internal damage signals.

[0018] To facilitate understanding of the sensor-based ultra-precision spring performance monitoring method described in this application, the key terms used in this document will be explained in a unified manner.

[0019] Acoustic emission signals refer to transient elastic waves generated by the release of internal stress during the deformation, damage, or fracture of a material. These transient elastic waves propagate along the material and can be sensed by sensors and converted into processable electrical signals. In the performance monitoring of ultra-precision springs, acoustic emission signals can characterize changes in the spring's internal microstructure and the evolution of damage, serving as an important information carrier for identifying internal damage.

[0020] The "low energy threshold" refers to the lower limit of energy used for preliminary screening of acoustic emission signals. Signals below this threshold are generally not included in subsequent analysis to reduce the impact of background noise and weak interference on the monitoring results.

[0021] "Data-form signal segment" refers to a digital signal segment formed after sampling, quantization, filtering, or other preprocessing of the original acoustic emission signal. This data-form signal segment serves as the input for subsequent feature extraction and analysis.

[0022] "Multidimensional features" refer to multiple types of characterization information extracted from signal segments in data form, used to describe the characteristics of acoustic emission signals from different dimensions. In this application, multidimensional features include at least time features, energy features, and frequency features. Time features may include duration, rise time, and peak occurrence time; energy features may include cumulative energy, peak amplitude, and root mean square value; frequency features may include dominant frequency, frequency band distribution, and average frequency.

[0023] "Specific characteristic signals" refer to external interference signals identified in the first determination layer based on time and energy characteristics. These signals are typically characterized by high energy and short duration, corresponding in characteristics to external disturbances such as instantaneous impacts or electromagnetic interference.

[0024] "Frequency variation trend" refers to the pattern of how the number of times or frequency of a certain type of acoustic emission signal changes over time during a continuous monitoring period.

[0025] "Cumulative energy change trend" refers to the pattern of cumulative energy change over time in a certain type of acoustic emission signal during a continuous monitoring period.

[0026] "Specific frequency distribution and cumulative change characteristics" refers to the combined characteristics of internal damage signals in terms of frequency distribution, frequency changes, and cumulative energy changes, which are used to distinguish internal damage signals from external interference signals.

[0027] Based on the above terminology definitions, the sensor-based ultra-precision spring performance monitoring method described in this application revolves around acoustic emission signal capture, signal digitization processing, multi-dimensional feature extraction, and hierarchical judgment and identification. By performing hierarchical analysis on the feature differences of different types of acoustic emission signals, the method can identify internal damage to ultra-precision springs.

[0028] During the signal acquisition phase, the acoustic emission signals generated by the ultra-precision spring are monitored, and preliminary capture is performed based on a low-energy threshold. Specifically, a piezoelectric sensor array can be deployed on the surface of the ultra-precision spring or its supporting structure. This allows the acoustic emission signals generated by microscopic changes within the spring to propagate along the material to the sensors, which then convert the mechanical vibrations into electrical signals and output them to the data acquisition unit. The data acquisition unit captures only acoustic emission signals with energy exceeding a preset low-energy threshold, thereby filtering out a significant amount of background noise and obtaining a preliminarily captured acoustic emission signal segment. In other embodiments, fiber optic acoustic emission sensors can also be used to sense acoustic emission events and perform preliminary screening based on a preset low-energy threshold after photoelectric conversion. Regardless of the sensing method used, the preliminary capture process is used to separate acoustic emission events with further analytical value from the original monitoring signal.

[0029] After initial acquisition, the acoustic emission signal segments undergo signal processing to convert them into data-based signal segments. This processing can be performed using an analog-to-digital converter (ADC) to convert the analog acoustic emission signal into a digital signal. During the conversion process, sampling, quantization, and digital filtering are incorporated to improve the accuracy of subsequent analysis. For example, the sampling rate can be set to 1MHz, the quantization precision to 16 bits, and a bandpass filter can be applied to the converted digital signal to suppress noise in non-target frequency bands, concentrating the retained signal within the frequency range relevant to spring damage. Through these processes, the original acoustic emission signal is uniformly converted into data-based signal segments suitable for computer analysis.

[0030] After obtaining signal segments in data form, multidimensional features are extracted. Time features characterize the signal's change state along the time axis, including duration, rise time, and peak occurrence time; energy features characterize the signal's intensity and release degree, including accumulated energy, peak amplitude, and root mean square value; frequency features characterize the signal's energy distribution in the frequency domain, including dominant frequency, frequency band distribution, and average frequency. Time features can be calculated by identifying the signal's start point, peak point, and decay endpoint; energy features can be obtained by integrating the square of the signal amplitude or by directly reading the peak amplitude; frequency features are extracted by converting the time-domain signal to the frequency-domain signal using a Fast Fourier Transform. The joint extraction of time, energy, and frequency features forms a multidimensional feature set for subsequent hierarchical determination.

[0031] After extracting multidimensional features, a hierarchical determination process is performed based on these features. This hierarchical determination process includes a first determination layer and a second determination layer. The first determination layer is used to preferentially identify signals with typical external interference characteristics, while the second determination layer is used to further identify acoustic emission signals related to internal damage, after eliminating obvious external interference.

[0032] In the first determination layer, specific characteristic signals are identified based on time and energy characteristics, and these specific characteristic signals are determined to be external interference signals. The specific characteristic signal meets the condition that its energy is higher than a preset energy value and its duration is less than a preset duration. For example, the preset energy value can be set to 100 mV²·s, and the preset duration to 50 microseconds. When the cumulative energy of a certain acoustic emission signal segment exceeds 100 mV²·s and its duration is less than 50 microseconds, the signal can be identified as a specific characteristic signal. This type of signal typically corresponds to instantaneous impacts, electromagnetic disturbances, or other short-term strong interferences. Its energy release is concentrated and its duration is short, which is inconsistent with the continuous characteristics formed during the internal damage evolution of the ultra-precision spring. Therefore, it is rejected in the first determination layer.

[0033] In the second judgment layer, signals not identified as external interference signals in the first judgment layer are analyzed in conjunction with their corresponding frequency characteristics, frequency variation trends, and cumulative energy variation trends. Signals with specific frequency distributions and cumulative variation characteristics are identified and classified as internal damage signals. For the internal damage process of ultra-precision springs, such as microcrack initiation and propagation, the acoustic emission signals typically have high energy concentration within a specific frequency range. Furthermore, as the damage evolves, the frequency of occurrence and cumulative energy of the relevant signals exhibit a continuous changing pattern. For example, if the acoustic emission signals within a certain frequency range continuously increase during the monitoring period, and the corresponding cumulative energy increases synchronously, it indicates that the signals within that frequency range are not random noise but are more likely related to the evolution of internal damage. When a certain type of signal simultaneously meets the preset frequency distribution conditions, frequency variation conditions, and cumulative energy variation conditions, it can be identified as a signal with specific frequency distribution and cumulative variation characteristics and further classified as an internal damage signal.

[0034] In some implementations, signals not eliminated by the first judgment layer can be further analyzed for their spectral characteristics to identify the frequency concentration range corresponding to internal damage. For example, acoustic emission signals generated by the propagation of internal microcracks can exhibit high energy concentration in the range of 200kHz to 400kHz; simultaneously, within a continuous monitoring period, the frequency of signals appearing in this frequency range continuously increases, and the accumulated energy also shows an increasing trend. Based on the above frequency distribution and cumulative change characteristics, such signals can be identified as internal damage signals. Thus, the layered judgment process achieves a step-by-step convergence from external interference elimination to internal damage identification, making the performance monitoring results of the ultra-precision spring closer to its true damage state.

[0035] Optional, combined Figure 2 As shown, the second determination layer involves identifying signals that were not identified as external interference signals in the first determination layer. This is done by recognizing signals with specific frequency distributions and cumulative change characteristics based on their corresponding frequency features, frequency variation trends, and cumulative energy variation trends. The steps for determining signals with specific frequency distributions and cumulative change characteristics as internal damage signals include: A1, deploy an external impact sensing array to record the timestamps, peak amplitudes, and durations of external particle impact events, and generate energy characterization parameters for external impact events; A2, for each acoustic emission signal that is not determined as an external interference signal by the first determination layer, extract the corresponding timestamp and match it with the timestamp of the external impact event recorded by the external impact sensing array in a time window. A3, when the time window matching is successful and the energy characterization parameters of the external impact event and the energy characteristics of the acoustic emission signal meet the preset matching range, the acoustic emission signal is determined to be an external impact interference signal and is removed from the internal damage determination process of the second determination layer; the external impact interference signal is a type of external interference signal, used to perform further external source removal on signals that were not determined to be external interference signals in the first determination layer. A4. For acoustic emission signals that are not identified as external impact interference signals, based on the corresponding frequency characteristics and the corresponding frequency change trend and cumulative energy change trend, identify signals with specific frequency distribution and cumulative change characteristics, and determine signals with specific frequency distribution and cumulative change characteristics as internal damage signals.

[0036] Specifically, an external impact sensing array is deployed to independently sense and record impact events in the external environment of the ultra-precision spring, outside of the acoustic emission signal monitoring system. This external impact sensing array can consist of multiple high-sensitivity impact sensors or piezoelectric sensors, used to collect the timestamps, peak amplitudes, and durations corresponding to external impact events. Based on the collected peak amplitudes and durations, energy characterization parameters of the external impact events can be further generated. These energy characterization parameters characterize the intensity level of the external impact events and serve as a reference for subsequently determining whether the acoustic emission signal originates from external impact interference. By setting up an external impact sensing array independent of the acoustic emission monitoring link, an independent source of evidence can be provided for external interference identification, thereby improving the accuracy of identifying external impact interference signals.

[0037] For each acoustic emission signal not identified as an external interference signal in the first determination layer, its corresponding occurrence timestamp is extracted, and this timestamp is matched with the timestamp of the external impact event recorded by the external impact sensing array using a time window. This time window matching is used to determine whether there is a temporal correspondence between the acoustic emission signal and the external impact event. Specifically, a preset time window can be set around the timestamp of the external impact event, and it can be determined whether the occurrence time of the acoustic emission signal falls within this preset time window. The preset time window is used to cover the time deviation introduced after the external impact occurs due to signal propagation delay, sensor response differences, and data acquisition link processing delays. For example, a ±50 microsecond time window centered on the timestamp of the external impact event can be set to complete the temporal correlation determination between the acoustic emission signal and the external impact event.

[0038] When the time window matching is successful, it is further determined whether the energy characterization parameters of the external impact event and the energy characteristics of the acoustic emission signal meet a preset matching range. The preset matching range is used to constrain the correspondence between the intensity of the external impact event and the energy level of the acoustic emission signal, to avoid incorrectly associating essentially unrelated events simply because they are close in time. In other words, time window matching only indicates that the two may be temporally related, while energy matching is used to further confirm whether the two are consistent in event intensity. When the energy characterization parameters of the external impact event and the energy characteristics of the acoustic emission signal are within the preset matching range, it can be considered that the acoustic emission signal and the corresponding external impact event are matched in both temporal and energy dimensions; if the two meet the time window matching but the energy difference exceeds the preset matching range, it indicates that the acoustic emission signal should not be directly attributed to the external impact event.

[0039] When both time window matching and energy matching are achieved, the corresponding acoustic emission signal is identified as an external impact interference signal and removed from the internal damage assessment process of the second assessment layer. By adding a secondary screening process based on an external impact sensing array after the first assessment layer, external source interference signals not identified by the first assessment layer can be further filtered out, reducing the interference of external impact events on the internal damage identification results. Therefore, the signal set retained by the second assessment layer can more effectively characterize the acoustic emission features during the damage evolution process of the ultra-precision spring, improving the accuracy of the internal damage identification results.

[0040] In some preferred embodiments, the ultra-precision spring operates in a service environment with a risk of impact from minute particles, such as space debris or dust particles. When a minute particle impacts the spring surface, the spring generates a corresponding acoustic emission signal. Simultaneously, an external impact sensing array positioned near the spring independently detects the external impact event and records the impact timestamp T1, peak amplitude A1, and duration D1. The system further calculates the energy characterization parameter E_impact of the external impact event based on A1 and D1.

[0041] Simultaneously, the acoustic emission sensor detects the corresponding acoustic emission signal and obtains its timestamp T_AE and energy characteristic E_AE. The acoustic emission signal first enters the first determination layer; if its energy and duration characteristics do not meet the determination conditions for external interference signals in the first determination layer, it enters the second determination layer for further analysis. In the second determination layer, the system extracts the timestamp T_AE of the acoustic emission signal and performs time window matching with the timestamp T1 of the external impact event. If T_AE is within the ±50 microsecond time window corresponding to T1, the time window matching is considered successful. Subsequently, the system compares E_AE with E_impact; if the difference between the two is within a preset matching range, for example, controlled within ±10%, the acoustic emission signal is determined to be consistent with the external impact event in the energy dimension. Based on the above time matching and energy matching results, the acoustic emission signal can be determined as an external impact interference signal and removed from the subsequent internal damage determination process.

[0042] Through the above processing method, the acoustic emission response caused by external impact events is no longer identified solely based on the time and energy characteristics in the first judgment layer. Instead, it undergoes secondary confirmation by combining the temporal and energy evidence provided by an independent external impact sensing array. This reduces the probability of external impact signals being misjudged as internal damage signals of the ultra-precision spring, allowing the monitoring results to more accurately reflect the damage state of the ultra-precision spring itself.

[0043] Optionally, for acoustic emission signals not identified as external impact interference signals, the steps of identifying signals with specific frequency distribution and cumulative variation characteristics based on the corresponding frequency characteristics, the corresponding frequency change trend, and the cumulative energy change trend, and determining signals with specific frequency distribution and cumulative variation characteristics as internal damage signals include: Deploy temperature sensor arrays and micro-strain sensor arrays to monitor local temperature changes and minute deformations of the ultra-precision spring and its corresponding surrounding structure in real time; Acquire local temperature change data and microstrain data that occur at times adjacent to acoustic emission signals that are not identified as external impact interference signals; Based on local temperature change data and microstrain data, determine whether the acoustic emission signal is a non-damage internal event caused by thermal stress or material relaxation. For acoustic emission signals that are not determined to be non-damaging internal events, signals with specific frequency distribution and cumulative change characteristics are identified based on the corresponding frequency characteristics, the corresponding frequency change trend and cumulative energy change trend, and signals with specific frequency distribution and cumulative change characteristics are determined to be internal damage signals.

[0044] Specifically, the deployment of temperature sensor arrays and micro-strain sensor arrays aims to provide real-time environmental and material state information for the ultra-precision spring and its surrounding structure. The temperature sensor array is used to accurately measure local temperature, while the micro-strain sensor array is used to monitor minute deformations of the material, which may be caused by temperature changes, material creep, or stress relaxation. These sensors are strategically placed in critical areas of the ultra-precision spring to ensure the representativeness of the acquired data. Acquiring local temperature change data and micro-strain data adjacent to the occurrence time of acoustic emission signals not identified as external impact interference signals means that after the acoustic emission signal is captured and undergoes initial external interference removal, the system retrieves temperature and micro-strain data closely related to the time point of the acoustic emission signal's occurrence. This temporal proximity ensures a causal correlation between the acoustic emission signal and changes in the environmental or material state. In practical applications, based on the local temperature change data and micro-strain data, it is determined whether the acoustic emission signal is a non-damaging internal event caused by thermal stress or material relaxation. The purpose is to use this auxiliary information to more finely classify the source of the acoustic emission signal. For example, when an acoustic emission signal occurs simultaneously with a significant local temperature rise or fall, and microstrain data also show deformation consistent with thermal expansion or contraction, the acoustic emission signal is likely derived from thermal stress rather than structural damage. Similarly, if an acoustic emission signal is associated with slow deformation (creep) or stress release (relaxation) of the material under constant load, it may also be identified as a non-damaging internal event. This determination process can be performed based on preset thresholds, models, or machine learning algorithms to distinguish the characteristic patterns of damaged and non-damaging signals. Thus, only acoustic emission signals that are neither external interference signals nor non-damaging internal events caused by thermal stress or material relaxation will be further analyzed based on their frequency characteristics, frequency variation trends, and cumulative energy variation trends, ultimately identifying and determining them as true internal damage signals.

[0045] Optionally, for acoustic emission signals not determined to be non-damaging internal events, the step of identifying signals with specific frequency distribution and cumulative change characteristics based on the corresponding frequency characteristics, the corresponding frequency change trend, and the cumulative energy change trend, and determining signals with specific frequency distribution and cumulative change characteristics as internal damage signals includes: Read signal segments in data form from different locations and with different frequency components; Real-time monitoring of local environmental parameters of the ultra-precision spring and its corresponding surrounding structure; local environmental parameters include temperature, radiation dose rate and vacuum level; Based on local environmental parameters, adjust the specific frequency range and energy concentration threshold associated with early microcracks; Based on the energy concentration threshold, the energy of data signal segments with different positions and different frequency components is checked to see if the energy is concentrated in the adjusted specific frequency range, and the check results are obtained. The frequency change trend and cumulative energy change trend of the corresponding data form signal segments are tracked and checked, and the frequency change threshold and cumulative energy threshold are adjusted according to local environmental parameters. Based on the adjusted occurrence frequency threshold and cumulative energy threshold, the threshold determination criteria are determined. When the occurrence frequency trend index corresponding to the occurrence frequency change trend and the cumulative energy trend index corresponding to the cumulative energy change trend meet the threshold judgment conditions, the data form signal segment with the inspection result is judged as an internal damage signal; where the occurrence frequency trend index is the occurrence frequency growth rate index within the sliding time window; and the cumulative energy trend index is the cumulative energy growth rate index within the sliding time window.

[0046] Specifically, reading signal segments in the form of data from different locations and with different frequency components refers to acquiring signal data with spatial and frequency dimensions from acoustic emission data that has undergone initial acquisition, signal processing, and multi-layer screening. These signal segments in the form of data contain the occurrence of acoustic emission events in different regions of the ultra-precision spring and their energy distribution in different frequency ranges. Real-time monitoring of the local environmental parameters of the ultra-precision spring and its corresponding surrounding structures can be understood as continuously acquiring key physical quantities in the working environment of the ultra-precision spring by deploying appropriate sensors. Local environmental parameters include temperature, radiation dose rate, and vacuum level. These parameters have direct or indirect effects on the acoustic emission characteristics of the material and the signal propagation path. For example, temperature changes affect the elastic modulus and sound velocity of the material, radiation dose rate may accelerate material aging, and vacuum level affects the attenuation of sound waves in the medium.

[0047] In practical applications, adjusting the specific frequency range and energy concentration threshold associated with early microcracks based on local environmental parameters aims to adapt the identification criteria to environmental changes. Acoustic emission signals generated by early microcracks typically exhibit specific frequency distributions and energy concentration characteristics, but these characteristics are not static and change dynamically with environmental parameters. By adjusting these thresholds in real time, the characteristic signals of microcracks can be captured more accurately. Furthermore, based on the energy concentration threshold, it is examined whether the energy of signal fragments with different locations and frequency components is concentrated within the adjusted specific frequency range, yielding the examination results. This step aims to initially screen signals that conform to the characteristics of early microcracks in terms of frequency and energy.

[0048] Furthermore, the frequency and cumulative energy trends of the data-form signal segments with "yes" results are tracked, and the frequency and cumulative energy thresholds are adjusted based on local environmental parameters. This indicates that this application focuses not only on the instantaneous characteristics of individual signals but also on their dynamic evolution trends. The frequency trend index is the frequency growth rate index within a sliding time window, and the cumulative energy trend index is the cumulative energy growth rate index within a sliding time window. These indicators reflect the initiation and development of damage. By adjusting the judgment thresholds for these trends based on local environmental parameters, false trends caused by environmental changes can be avoided. Therefore, based on the adjusted frequency and cumulative energy thresholds, threshold judgment conditions are determined. When the frequency trend index corresponding to the frequency change trend and the cumulative energy trend index corresponding to the cumulative energy change trend meet the threshold judgment conditions, the data-form signal segment with "yes" results is judged as an internal damage signal. This constitutes a comprehensive judgment logic that combines the frequency and energy characteristics of the signal with its time-varying trend characteristics, and all judgment criteria can be adaptively adjusted according to the real-time environment, thereby improving the accuracy and robustness of internal damage identification.

[0049] Optionally, adjusting the specific frequency range and energy concentration threshold associated with early microcracks based on local environmental parameters includes: Based on local environmental parameters, identify the corresponding action mode; the action mode includes superposition or cancellation of the transmission characteristics of data signal segments; Based on the mode of action, a set of correction factors is generated; the set of correction factors contains correction amounts for different frequency ranges and energy concentration thresholds. The set of correction factors is applied to a preset specific frequency range and energy concentration threshold to obtain the adjusted specific frequency range and energy concentration threshold.

[0050] Specifically, local environmental parameters refer to factors such as temperature, radiation dose rate, and vacuum level of the ultra-precision spring and its surrounding environment. Changes in these parameters directly affect the propagation characteristics of acoustic emission signals within the spring material and at the sensor-spring interface. For example, increased temperature may lead to changes in the sound velocity or increased attenuation in the material, radiation may cause changes in the microstructure of the material, and changes in vacuum level may affect the coupling efficiency of acoustic emission signals in the medium.

[0051] Identifying the corresponding mode of action refers to analyzing how local environmental parameters collectively influence the transmission characteristics of signal segments in data form. This influence may manifest as a superposition effect, where multiple environmental parameters jointly enhance or weaken a certain characteristic of the signal; or it may manifest as a cancellation effect, where the influence of one environmental parameter is partially or completely canceled out by another. For example, in a high-temperature, high-radiation environment, temperature may cause signal attenuation, while radiation may change the elastic modulus of materials; the effects of both on the signal frequency distribution may be superimposed or partially canceled out.

[0052] Based on the identified interaction modes, a set of correction factors is generated. This set of correction factors is a group of quantified correction values ​​used to precisely adjust specific frequency ranges and energy concentration thresholds. These corrections are pre-established based on different combinations of environmental parameters and their interaction modes, using experimental data, physical models, or machine learning methods. For example, when a certain superposition interaction mode is identified, the set of correction factors may include corrections to broaden or narrow a specific frequency range, as well as corrections to increase or decrease the energy concentration threshold.

[0053] Finally, the generated set of correction factors is applied to a preset specific frequency range and energy concentration threshold. The preset specific frequency range and energy concentration threshold are empirical or theoretical settings based on the characteristics of early microcrack signals under standard environmental conditions or ideal states. By applying the set of correction factors, these preset values ​​can be dynamically and precisely adjusted according to real-time local environmental parameters, thereby obtaining adjusted specific frequency ranges and energy concentration thresholds that better suit the current environmental conditions.

[0054] Optionally, the steps of adjusting the occurrence frequency threshold and the cumulative energy threshold based on local environmental parameters include: Based on local environmental parameters, identify the influence patterns of environmental parameter combinations on the transmission characteristics of data signal segments; the influence patterns include nonlinear or non-monotonic effects on the cumulative energy threshold and the occurrence frequency threshold; among them, the nonlinear effect is the effect of non-proportional changes between changes in local environmental parameters and the threshold adjustment amount; the non-monotonic effect is the effect of changes in local environmental parameters causing the threshold adjustment direction to reverse in different value ranges and having an inflection point. Based on the influence pattern, a dynamic threshold function is generated; the dynamic threshold function is used to calculate the corresponding occurrence frequency threshold and cumulative energy threshold according to real-time local environmental parameters. By applying a dynamic threshold function, the corresponding occurrence frequency threshold and cumulative energy threshold are calculated based on real-time local environmental parameters, and these are used to replace the preset fixed thresholds to obtain the adjusted occurrence frequency threshold and cumulative energy threshold.

[0055] Specifically, identifying the influence patterns of environmental parameter combinations on the transmission characteristics of signal segments in data form refers to determining, through analysis of historical data, experimental results, or physical models, how different local environmental parameters (such as temperature, radiation dose rate, and vacuum level) and their combinations collectively affect the propagation, attenuation, and characteristic manifestations of acoustic emission signals. This influence pattern may not be a simple linear superposition but may include nonlinear or non-monotonic effects. Nonlinear effects refer to a non-proportional relationship between changes in local environmental parameters and the threshold adjustment amount. For example, a doubling of temperature may not simply double the threshold adjustment amount but rather change in a squared or exponential manner. Non-monotonic effects indicate that when local environmental parameters change, the direction of threshold adjustment may reverse at a certain inflection point. For example, in low-temperature regions, an increase in temperature may lead to a decrease in the threshold, but in high-temperature regions, an increase in temperature may lead to an increase in the threshold.

[0056] Furthermore, based on the identified influence patterns, a dynamic threshold function is generated. This dynamic threshold function is a mathematical model or algorithm that takes real-time local environmental parameters as input and outputs precisely calculated occurrence frequency thresholds and cumulative energy thresholds. This function can embed nonlinear or non-monotonic mappings to accurately reflect the influence of environmental parameters on the thresholds. For example, the function can be a polynomial function, an exponential function, a piecewise function, or a complex mapping constructed based on a machine learning model (such as a neural network).

[0057] Therefore, by applying this dynamic threshold function, the most suitable occurrence frequency threshold and cumulative energy threshold under the current operating conditions can be dynamically calculated based on real-time acquired local environmental parameters. This dynamic calculation result will replace the preset fixed threshold or simple linear correction threshold in traditional methods, thus making threshold adjustment more precise and adaptive.

[0058] Optionally, the steps for identifying the influence patterns of combinations of environmental parameters on the transmission characteristics of signal segments in data form, based on local environmental parameters, include: Obtain fatigue parameters for the spring material; these parameters include service time, cumulative stress cycle count, and historical maximum stress value. Based on local environmental parameters and fatigue level parameters, a dynamic correction process for the influence mode is performed. The dynamic correction process includes: identifying the influence mode based on the local environmental parameters; correcting the degree and inflection point of nonlinear or non-monotonic changes in the influence mode based on the fatigue level parameters; temporarily adjusting the sensitivity of the influence mode when the local environmental parameters reach preset extreme values; and outputting the dynamically corrected influence mode. Among these, the sensitivity is a parameter that determines the adjustment intensity of the influence mode to the occurrence frequency threshold and the cumulative energy threshold under changes in environmental parameters. It is used to temporarily amplify or suppress the threshold adjustment amplitude when the local environmental parameters reach extreme values.

[0059] Specifically, fatigue parameters are indicators characterizing the cumulative damage experienced by a spring material during its service life. These parameters may include service time, cumulative stress cycles, and the highest historical stress value. Service time reflects the total duration of the material's exposure to environmental and load conditions; cumulative stress cycles quantify the frequency and total amount of cyclic loads the material experiences; and the highest historical stress value records the highest stress level the material has ever endured. These parameters collectively provide a comprehensive assessment of the current fatigue state of the spring material, aiming to provide crucial material condition information for subsequent influence mode correction.

[0060] The dynamic correction process for influence modes aims to refine the initially identified influence modes by considering the fatigue level of the spring material. This process first identifies an initial influence mode based on local environmental parameters, describing the fundamental impact of these parameters on signal transmission characteristics. Based on this, the degree of nonlinear or non-monotonic change and the location of inflection points in the influence mode are corrected according to the acquired fatigue level parameters. For example, as material fatigue increases, the influence of certain environmental parameters on signal transmission may change from linear to nonlinear, or its direction of influence may reverse at a specific threshold. These changes are quantified and adjusted using fatigue level parameters. Furthermore, when local environmental parameters reach preset extreme values, such as a sudden temperature increase or an abnormally high radiation dose rate, a temporary sensitivity adjustment is made to the influence mode. Sensitivity can be understood as the adjustment intensity parameter of the influence mode to the occurrence frequency threshold and cumulative energy threshold under changes in environmental parameters. Its purpose is to temporarily amplify or suppress the threshold adjustment amplitude when local environmental parameters reach extreme values, ensuring the accuracy and robustness of monitoring under abnormal operating conditions.

[0061] Optionally, the steps for identifying the influencing patterns based on local environmental parameters include: Preliminary identification is performed based on local environmental parameters to obtain the initial influencing pattern; Multiple local environmental parameter sensors are deployed and distributed in a grid to obtain temperature, radiation dose rate and vacuum level in different local areas; Based on the temperature, radiation dose rate, and vacuum level of different local regions, the influence patterns of each local region are identified separately. Spatial interpolation is performed on the influence patterns of local areas to generate an overall influence pattern distribution map of the spring. Based on the influence pattern distribution map, the initial influence pattern is corrected to obtain the influence pattern.

[0062] Specifically, during the dynamic correction process of the impact mode, a preliminary identification is first performed based on local environmental parameters to obtain the initial impact mode. This preliminary identification can be based on a preset model or empirical data to make a preliminary judgment on the relationship between local environmental parameters and the threshold impact mode.

[0063] Furthermore, to obtain more refined environmental parameter information, multiple local environmental parameter sensors are deployed. These sensors are distributed in a grid pattern on the ultra-precision spring and its surrounding structure to acquire environmental parameters such as temperature, radiation dose rate, and vacuum level in different local areas in real time. For example, a thermocouple array can be used to monitor temperature, a Geiger counter array to monitor radiation dose rate, and a miniature vacuum gauge array to monitor vacuum level.

[0064] Based on the temperature, radiation dose rate, and vacuum level data obtained from these different local regions, the influence patterns of each local region were identified. This means that the influence patterns of environmental parameters on the acoustic emission signal transmission characteristics (and consequently on the cumulative energy threshold and occurrence frequency threshold) may differ for different parts of the spring, requiring localized analysis.

[0065] Subsequently, spatial interpolation is performed on the influence patterns of these local areas. Spatial interpolation is a method of estimating data for unknown points using data from known points; algorithms such as Kriging interpolation, inverse distance weighted interpolation, or spline interpolation can be used. Through spatial interpolation, an overall influence pattern distribution map of the spring can be generated, which can intuitively show the continuous changes in the influence patterns of environmental parameters at various parts of the spring.

[0066] Finally, based on the impact pattern distribution map, the initial impact patterns obtained earlier were revised to obtain more accurate and comprehensive impact patterns. This revision took into account the spatial heterogeneity of environmental parameters, making the identification of impact patterns closer to actual working conditions.

[0067] Optionally, the steps of correcting the degree and inflection point location of nonlinear or non-monotonic changes in the influence mode based on fatigue level parameters include: The fatigue degree parameters are analyzed to identify the type of fatigue mechanism of the spring material; the fatigue mechanism types include one or more of high-cycle fatigue, low-cycle fatigue, or creep fatigue. Based on the type of fatigue mechanism, select the corresponding set of correction rules from the preset correction rule library; the set of correction rules defines the adjustment method for the degree of nonlinear or non-monotonic change and the position of inflection point in the influence mode; The set of correction rules is applied to the influence model to correct the degree and inflection point location of nonlinear or nonmonotonic changes in the influence model.

[0068] Specifically, analyzing fatigue parameters involves comprehensively evaluating parameters such as the spring's service life, cumulative stress cycles, and historical maximum stress value to determine the primary fatigue mechanism affecting the current spring material. For example, high-cycle fatigue is typically associated with a high number of cycles and relatively low stress levels; low-cycle fatigue involves a low number of cycles and high plastic strain; and creep fatigue mainly occurs under high temperatures and prolonged continuous loading. Identifying these fatigue mechanism types is fundamental to ensuring that subsequent modifications are targeted.

[0069] The pre-defined correction rule base can be understood as a database containing various correction rules. These rules are established based on in-depth research and experimental verification of material behavior and acoustic emission signal characteristics under different fatigue mechanisms. Each set of correction rules is tailored to a specific fatigue mechanism type, defining in detail how to adjust the degree of nonlinear or non-monotonic changes in the influencing mode and the inflection point position. For example, correction rules for high-cycle fatigue may focus on the impact of microcrack initiation and propagation on high-frequency acoustic emission signals, and adjust the nonlinearity of the influence of environmental parameters on the threshold accordingly; while correction rules for creep fatigue may focus more on the impact of material creep damage on low-frequency acoustic emission signals, and adjust the inflection point position of the threshold as a function of temperature.

[0070] In practical applications, applying the set of correction rules to the influence model specifically refers to selecting the appropriate set of correction rules from the correction rule base based on the identified fatigue mechanism type, and then refining the influence model initially identified based on local environmental parameters according to the adjustment methods defined in the set. This refinement may include adjusting the weights of specific parameters in the influence model, changing the coefficients of nonlinear functions, or redefining the inflection point positions of non-monotonic changes. The aim is to make the influence model more accurately reflect the actual impact of local environmental parameters on the acoustic emission signal transmission characteristics and threshold determination under a specific fatigue state.

[0071] In some preferred embodiments, assuming a high-precision spring, after long-term service, exhibits fatigue parameters showing an extremely high cumulative stress cycle count but relatively low stress values ​​per cycle, this indicates that the spring is primarily in a high-cycle fatigue state. The system identifies "high-cycle fatigue" as the current fatigue mechanism type based on these fatigue parameters. Subsequently, the system selects a set of correction rules specifically for high-cycle fatigue from a pre-defined correction rule library. This rule set may define how the nonlinear influence of local environmental parameters such as temperature and radiation dose rate on the frequency distribution and energy concentration threshold of microcrack acoustic emission signals should be adjusted under high-cycle fatigue conditions, and within what range of environmental parameters might the inflection point of the influence direction occur. For example, the rule might indicate that in the later stages of high-cycle fatigue, the material's sensitivity to temperature changes increases significantly, leading to a greater degree of nonlinearity in threshold adjustment. The system applies these correction rules to the current influence mode, thereby precisely adjusting the degree and inflection point position of nonlinear or non-monotonic changes in the influence mode, enabling the subsequently dynamically calculated frequency threshold and cumulative energy threshold to more accurately meet the spring performance monitoring needs under high-cycle fatigue conditions.

[0072] This application also discloses a sensor-based ultra-precision spring performance monitoring system for performing sensor-based ultra-precision spring performance monitoring, combined with... Figure 3 As shown, the sensor-based ultra-precision spring performance monitoring system 1 includes: The signal acquisition execution module 11 is used to monitor the acoustic emission signal generated by the ultra-precision spring, and to perform preliminary acquisition of the acoustic emission signal based on a preset low energy threshold to obtain a preliminary captured acoustic emission signal segment. The signal processing execution module 12 is used to process the initially captured acoustic emission signal segments to convert them into data signal segments; The multidimensional feature extraction module 13 is used to extract multidimensional features from signal segments in data form; the multidimensional features include time features, energy features and frequency features. The hierarchical judgment execution module 14 is used to perform a hierarchical judgment process based on multi-dimensional features. The hierarchical judgment process includes: performing a first judgment layer judgment: identifying specific feature signals based on time and energy features; specific feature signals refer to signals with energy higher than a preset energy value and duration less than a preset duration; judging specific feature signals as external interference signals; performing a second judgment layer judgment: for signals that were not judged as external interference signals in the first judgment layer, identifying signals with specific frequency distribution and cumulative change characteristics based on the corresponding frequency characteristics and the corresponding frequency change trend and cumulative energy change trend, and judging signals with specific frequency distribution and cumulative change characteristics as internal damage signals.

[0073] Specifically, the signal acquisition execution module can be configured to consist of an acoustic emission sensor, a preamplifier, and an analog-to-digital converter (ADC). The acoustic emission sensor is responsible for converting the mechanical vibrations generated by the spring into electrical signals; the preamplifier amplifies the weak electrical signals to improve the signal-to-noise ratio; and the ADC converts the analog signals into digital signals for subsequent digital processing. As a preferred implementation, this module can be integrated into a compact hardware unit, directly mounted near the ultra-precision spring, to minimize signal transmission loss and external interference.

[0074] The signal processing execution module can be implemented as a software program running on a digital signal processor (DSP) or embedded microcontroller. This module receives digital signal segments from the signal acquisition execution module and performs operations such as filtering, denoising, and sampling rate conversion to convert them into standardized data signal segments. For example, a bandpass filter can be used to remove noise within a specific frequency range, ensuring the accuracy of subsequent feature extraction. In practical applications, this module can also be implemented using a field-programmable gate array (FPGA) to provide higher processing speed and parallel processing capabilities.

[0075] The multidimensional feature extraction module can be implemented as a software algorithm running on the main processor or as part of an application-specific integrated circuit (ASIC). This module receives signal segments in data form and calculates time features (such as duration and rise time), energy features (such as cumulative energy and peak amplitude), and frequency features (such as dominant frequency and bandwidth distribution). For example, the frequency components of the signal can be analyzed using the Fast Fourier Transform (FFT) algorithm, and the signal's energy and duration can be calculated through statistical analysis. To improve efficiency, this module can employ a parallel computing architecture to process feature extraction tasks from multiple signal segments simultaneously.

[0076] The hierarchical decision-making execution module can be implemented as decision logic software running on a central processing unit (CPU) or as hardware logic circuitry. This module receives multi-dimensional features and makes judgments according to preset hierarchical decision rules. In the first decision layer, external interference signals are identified and eliminated based on time and energy characteristics. In the second decision layer, the remaining signals are further analyzed for their frequency characteristics, frequency variation trends, and cumulative energy variation trends to identify internal damage signals. This module is designed to ensure the accuracy and real-time performance of decisions; for example, complex decision logic can be implemented using state machines or expert systems.

[0077] The sensor-based ultra-precision spring performance monitoring system proposed in this application represents a significant advancement over existing technologies in addressing the challenges of ultra-precision spring performance monitoring in complex environments.

[0078] Traditional monitoring systems often struggle to accurately distinguish and effectively suppress false alarms caused by high-intensity, transient external interference without reducing their sensitivity to real micro-damage signals. For example, in deep space exploration missions, when a probe encounters particle impacts, existing systems may frequently issue false alarms because they cannot effectively distinguish between external impacts and internal damage, or they may raise the detection threshold to avoid false alarms, thus missing crucial early damage signals.

[0079] The system in this application effectively overcomes the aforementioned problems through its modular design and refined hierarchical judgment mechanism. The signal acquisition execution module and the signal processing execution module work together to ensure comprehensive acquisition and high-quality preprocessing of all potential event signals. The multi-dimensional feature extraction module provides a rich and comprehensive data foundation for subsequent intelligent judgment. Most importantly, the hierarchical judgment execution module achieves accurate differentiation between external interference signals and internal damage signals through two progressive judgment logics. The first judgment layer efficiently identifies and eliminates transient, high-energy external impacts, such as instantaneous vibrations caused by particle impacts, thus avoiding false alarms. The second judgment layer focuses on identifying internal damage signals that may have lower energy but possess clear damage characteristics, such as the unique frequency distribution and accumulation trend generated by slowly expanding fatigue cracks. This system design enables the monitoring system to effectively suppress external interference and maintain high sensitivity to weak internal damage signals in complex and variable environments, thereby significantly improving the accuracy and reliability of ultra-precision spring performance monitoring, avoiding the loss of critical fault information, and ensuring the service safety of ultra-precision springs under extreme conditions.

[0080] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A sensor-based ultra-precision spring performance monitoring method, characterized in that, include: The acoustic emission signal generated by the ultra-precision spring is monitored, and the acoustic emission signal is initially captured based on a preset low energy threshold to obtain a preliminarily captured acoustic emission signal segment; The initially captured acoustic emission signal segments are processed to convert them into data signal segments; Extracting multidimensional features from signal segments in data form; The multidimensional features include time features, energy features, and frequency features; A hierarchical determination process is performed based on the aforementioned multidimensional features; The layered determination process includes: performing a first determination layer determination: identifying a specific characteristic signal based on the time characteristics and the energy characteristics; the specific characteristic signal refers to a signal whose energy is higher than a preset energy value and whose duration is less than a preset duration; determining the specific characteristic signal as an external interference signal; performing a second determination layer determination: for signals that were not determined as external interference signals in the first determination layer, identifying signals with specific frequency distribution and cumulative change characteristics based on the corresponding frequency characteristics and the corresponding frequency change trend and cumulative energy change trend, and determining signals with specific frequency distribution and cumulative change characteristics as internal damage signals.

2. The sensor-based method for monitoring the performance of ultra-precision springs according to claim 1, characterized in that, The second determination layer determination, for signals not determined as external interference signals in the first determination layer, involves identifying signals with specific frequency distribution and cumulative change characteristics based on the corresponding frequency characteristics and the corresponding frequency change trend and cumulative energy change trend, and determining signals with specific frequency distribution and cumulative change characteristics as internal damage signals. Deploy an external impact sensing array to record the timestamps, peak amplitudes, and durations of external particle impact events, and generate energy characterization parameters for external impact events; For each acoustic emission signal that is not identified as an external interference signal by the first determination layer, the corresponding timestamp is extracted and matched with the timestamp of the external impact event recorded by the external impact sensing array for time window matching. When the time window matching is successful and the energy characterization parameter of the external impact event and the energy characteristics of the acoustic emission signal meet the preset matching range, the acoustic emission signal is determined to be an external impact interference signal and is removed from the internal damage determination process of the second determination layer; the external impact interference signal is a type of external interference signal, used to perform further external source removal on signals that were not determined to be external interference signals in the first determination layer. For acoustic emission signals that are not identified as external impact interference signals, signals with specific frequency distribution and cumulative change characteristics are identified based on the corresponding frequency characteristics, the corresponding frequency change trend and cumulative energy change trend, and signals with specific frequency distribution and cumulative change characteristics are identified as internal damage signals.

3. The sensor-based method for monitoring the performance of ultra-precision springs according to claim 2, characterized in that, The step of identifying signals with specific frequency distribution and cumulative variation characteristics based on the corresponding frequency characteristics, frequency variation trends, and cumulative energy variation trends of acoustic emission signals not identified as external impact interference signals, and determining signals with specific frequency distribution and cumulative variation characteristics as internal damage signals, includes: Deploy temperature sensor arrays and micro-strain sensor arrays to monitor local temperature changes and minute deformations of the ultra-precision spring and its corresponding surrounding structure in real time; Acquire local temperature change data and microstrain data that occur at times adjacent to acoustic emission signals that are not identified as external impact interference signals; Based on the local temperature change data and the micro-strain data, it is determined whether the acoustic emission signal is a non-damaging internal event caused by thermal stress or material relaxation. For acoustic emission signals that are not determined to be non-damaging internal events, signals with specific frequency distribution and cumulative change characteristics are identified based on the corresponding frequency characteristics, the corresponding frequency change trend and cumulative energy change trend, and signals with specific frequency distribution and cumulative change characteristics are determined to be internal damage signals.

4. The sensor-based ultra-precision spring performance monitoring method according to claim 3, characterized in that, The step of identifying signals with specific frequency distribution and cumulative variation characteristics based on the corresponding frequency characteristics, frequency variation trends, and cumulative energy variation trends of acoustic emission signals that are not determined to be non-damaging internal events, and determining signals with specific frequency distribution and cumulative variation characteristics as internal damage signals, includes: Read signal segments in data form from different locations and with different frequency components; Real-time monitoring of local environmental parameters of the ultra-precision spring and its corresponding surrounding structure; the local environmental parameters include temperature, radiation dose rate, and vacuum level. Based on the local environmental parameters, adjust the specific frequency range and energy concentration threshold associated with early microcracks; Based on the energy concentration threshold, check whether the energy of data signal segments with different positions and different frequency components is concentrated in the adjusted specific frequency range, and obtain the check result; Track the frequency change trend and cumulative energy change trend of the data form signal segment corresponding to the inspection result being yes, and adjust the frequency threshold and cumulative energy threshold according to the local environmental parameters; Based on the adjusted occurrence frequency threshold and cumulative energy threshold, the threshold determination criteria are determined. When the occurrence frequency trend index corresponding to the occurrence frequency change trend and the cumulative energy trend index corresponding to the cumulative energy change trend meet the threshold judgment condition, the data form signal segment with the inspection result being yes is judged as an internal damage signal; wherein, the occurrence frequency trend index is the occurrence frequency growth rate index within the sliding time window; and the cumulative energy trend index is the cumulative energy growth rate index within the sliding time window.

5. The sensor-based method for monitoring the performance of ultra-precision springs according to claim 4, characterized in that, The step of adjusting the specific frequency range and energy concentration threshold related to early microcracks based on the local environmental parameters includes: Based on the local environmental parameters, the corresponding action mode is identified; the action mode includes superposition or cancellation of the transmission characteristics of data signal segments; Based on the aforementioned mode of operation, a set of correction factors is generated; the set of correction factors includes correction amounts for different frequency ranges and energy concentration thresholds. The set of correction factors is applied to a preset specific frequency range and energy concentration threshold to obtain the adjusted specific frequency range and energy concentration threshold.

6. The sensor-based method for monitoring the performance of ultra-precision springs according to claim 4, characterized in that, The step of adjusting the occurrence frequency threshold and the cumulative energy threshold based on the local environmental parameters includes: Based on the local environmental parameters, identify the influence patterns of environmental parameter combinations on the transmission characteristics of data-form signal segments; the influence patterns include nonlinear or non-monotonic influences on the cumulative energy threshold and the occurrence frequency threshold; wherein, the nonlinear influence is the influence of a non-proportional change between the change of the local environmental parameters and the threshold adjustment amount; the non-monotonic influence is the influence of the change of the local environmental parameters causing the threshold adjustment direction to reverse in different value ranges and having an inflection point; Based on the aforementioned influence pattern, a dynamic threshold function is generated; the dynamic threshold function is used to calculate the corresponding occurrence frequency threshold and cumulative energy threshold according to real-time local environmental parameters. By applying the dynamic threshold function, the corresponding occurrence frequency threshold and cumulative energy threshold are calculated based on real-time local environmental parameters, and the preset fixed thresholds are replaced to obtain the adjusted occurrence frequency threshold and cumulative energy threshold.

7. The sensor-based method for monitoring the performance of ultra-precision springs according to claim 6, characterized in that, The step of identifying the influence pattern of the combination of environmental parameters on the transmission characteristics of the data-form signal segment based on the local environmental parameters includes: Obtain fatigue parameters of the spring material; the fatigue parameters include service time, cumulative stress cycle count, and historical maximum stress value; Based on the local environmental parameters and the fatigue level parameters, a dynamic correction process for the influence mode is performed. This dynamic correction process includes: identifying the influence mode based on the local environmental parameters; correcting the degree and inflection point of nonlinear or non-monotonic changes in the influence mode based on the fatigue level parameters; temporarily adjusting the sensitivity of the influence mode when the local environmental parameters reach preset extreme values; and outputting the dynamically corrected influence mode. The sensitivity is a parameter representing the adjustment intensity of the influence mode to the frequency threshold and cumulative energy threshold under environmental parameter changes, used to temporarily amplify or suppress the threshold adjustment amplitude when local environmental parameters reach extreme values.

8. The sensor-based method for monitoring the performance of ultra-precision springs according to claim 7, characterized in that, The step of identifying the influence pattern based on the local environmental parameters includes: Based on the local environmental parameters, a preliminary identification is performed to obtain the initial influencing pattern; Multiple local environmental parameter sensors are deployed and distributed in a grid to obtain temperature, radiation dose rate and vacuum level in different local areas; Based on the temperature, radiation dose rate, and vacuum level of different local regions, the influence patterns of each local region are identified separately. Spatial interpolation is performed on the influence patterns of local areas to generate an overall influence pattern distribution map of the spring. Based on the aforementioned influence pattern distribution map, the initial influence pattern is corrected to obtain the influence pattern.

9. The sensor-based method for monitoring the performance of ultra-precision springs according to claim 7, characterized in that, The step of correcting the degree and inflection point position of nonlinear or non-monotonic changes in the influence mode based on the fatigue degree parameter includes: The fatigue degree parameters are analyzed to identify the type of fatigue mechanism of the spring material; the fatigue mechanism type includes one or more of high-cycle fatigue, low-cycle fatigue, or creep fatigue. Based on the fatigue mechanism type, a corresponding set of correction rules is selected from a preset set of correction rules; the set of correction rules defines the adjustment method for the degree of nonlinear or non-monotonic change and the position of inflection point in the influence mode. The set of correction rules is applied to the influence pattern to correct the degree and inflection point position of nonlinear or non-monotonic changes in the influence pattern.

10. A sensor-based ultra-precision spring performance monitoring system, used to perform sensor-based ultra-precision spring performance monitoring, characterized in that, include: The signal capture execution module is used to monitor the acoustic emission signal generated by the ultra-precision spring, and to perform preliminary capture of the acoustic emission signal based on a preset low energy threshold to obtain a preliminary captured acoustic emission signal segment; The signal processing execution module is used to process the initially captured acoustic emission signal segments to convert them into data signal segments; A multidimensional feature extraction module is used to extract multidimensional features from signal segments in data form; the multidimensional features include time features, energy features, and frequency features. The hierarchical determination execution module is used to perform a hierarchical determination process based on the multidimensional features; The hierarchical determination process includes: performing a first-level determination: identifying specific feature signals based on the time characteristics and the energy characteristics; A specific characteristic signal refers to a signal whose energy is higher than a preset energy value and whose duration is less than a preset duration; the specific characteristic signal is determined to be an external interference signal; a second determination layer is performed: for signals that are not determined to be external interference signals in the first determination layer, based on the corresponding frequency characteristics and the corresponding frequency change trend and cumulative energy change trend, signals with specific frequency distribution and cumulative change characteristics are identified, and signals with specific frequency distribution and cumulative change characteristics are determined to be internal damage signals.

Images