Method for evaluating service life of TF-SAW filter and related equipment

Abstract

The invention relates to a TF-SAW filter life evaluation method and related equipment, and the method comprises the following steps: applying an excitation signal to a TF-SAW filter, collecting response data, and generating a frequency domain response signal; performing state feature extraction on the TF-SAW filter based on the frequency domain response signal to generate a state feature sequence; performing degradation analysis on the TF-SAW filter based on the state feature sequence to generate a degradation track sequence; performing failure probability calculation on the TF-SAW filter based on the degradation track sequence to obtain a failure probability; and evaluating the service life of the TF-SAW filter based on the failure probability, and generating a service life evaluation report, thereby solving the problem that the TF-SAW filter is easy to generate performance degradation or even failure in a complex electromagnetic environment and a long-time working condition.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of TF-SAW filter, and particularly relates to a life evaluation method of TF-SAW filter and related equipment. BACKGROUND

[0002] With the rapid development of wireless communication technology, the performance requirements of radio frequency front-end devices are increasingly improved. TF-SAW (Thin Film Bulk Acoustic Resonator-Surface Acoustic Wave) filter is widely concerned due to its excellent high-frequency characteristics, good integrability and small volume. Under complex electromagnetic environment and long-time working conditions, the TF-SAW filter is prone to performance degradation and even failure, which affects the stability and reliability of the system. Therefore, it is of great significance to develop a life evaluation method for TF-SAW filter to ensure the safe operation of the equipment and reduce the maintenance cost. SUMMARY

[0003] The technical problem solved by the present application is to provide a life evaluation method of TF-SAW filter and related equipment, which solves the problem that the TF-SAW filter is prone to performance degradation and even failure under complex electromagnetic environment and long-time working conditions.

[0004] To solve the above technical problem, one technical solution adopted by the present application is that an excitation signal is applied to the TF-SAW filter and response data is collected to generate a frequency domain response signal; state feature extraction is performed on the TF-SAW filter based on the frequency domain response signal to generate a state feature sequence; degradation analysis is performed on the TF-SAW filter based on the state feature sequence to generate a degradation trajectory sequence; failure probability calculation is performed on the TF-SAW filter based on the degradation trajectory sequence to obtain a failure probability; and life evaluation is performed on the TF-SAW filter based on the failure probability to generate a life evaluation report.

[0005] To solve the above technical problem, one technical solution adopted by the present application is that an acquisition unit is provided for applying an excitation signal to the TF-SAW filter and collecting response data to generate a frequency domain response signal; an extraction unit is provided for performing state feature extraction on the TF-SAW filter based on the frequency domain response signal to generate a state feature sequence; an analysis unit is provided for performing degradation analysis on the TF-SAW filter based on the state feature sequence to generate a degradation trajectory sequence; a calculation unit is provided for performing failure probability calculation on the TF-SAW filter based on the degradation trajectory sequence to obtain a failure probability; and an evaluation unit is provided for performing life evaluation on the TF-SAW filter based on the failure probability to generate a life evaluation report.

[0006] To solve the above technical problems, another technical solution adopted by the present application is to provide an electronic device comprising a memory and a processor, the memory storing program instructions, and the processor fetching the program instructions from the memory to execute the life evaluation method of the TF-SAW filter.

[0007] To solve the above technical problems, another technical solution adopted by the present application is to provide a computer-readable storage medium comprising stored program data, which, when executed by a processor, is used to implement the life evaluation method of the TF-SAW filter.

[0008] The above scheme applies an excitation signal to the TF-SAW filter and collects response data to generate a frequency domain response signal; performs state feature extraction on the TF-SAW filter based on the frequency domain response signal to generate a state feature sequence; performs degradation analysis on the TF-SAW filter based on the state feature sequence to generate a degradation trajectory sequence; calculates the failure probability of the TF-SAW filter based on the degradation trajectory sequence to obtain the failure probability; and evaluates the life of the TF-SAW filter based on the failure probability to generate a life evaluation report, solving the problem of performance degradation or even failure of the TF-SAW filter under complex electromagnetic environment and long working conditions, effectively capturing the subtle changes of the device at different degradation stages, and enhancing the sensitivity to early degradation features. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiment or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0010] Figure 1 is a step schematic diagram of the life evaluation method of the TF-SAW filter in an embodiment of the present application; Figure 2 is a structural block diagram of the life evaluation device of the TF-SAW filter in an embodiment of the present application; Figure 3 is a structural schematic block diagram of a computer device of an embodiment of the present application.

[0011] The object implementation, functional characteristics and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0012] With reference to the drawings and embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0013] Specifically, the life evaluation method of the TF-SAW filter in the embodiment includes the following steps: As shown in Figure 1 , Figure 1 is a life evaluation method of a TF-SAW filter in an embodiment of the present application, including the following steps: Step S1, an excitation signal is applied to the TF-SAW filter and response data is collected to generate a frequency domain response signal.

[0014] Specifically, in a wireless communication system, when evaluating the performance of a TF-SAW filter, an excitation signal needs to be applied to the TF-SAW filter and response data needs to be collected to generate a frequency domain response signal, which is a key step to ensure its stability and reliability. When implementing this process, first, a signal generator is used to generate an excitation signal with precise control. The frequency range and power level of the excitation signal need to be determined according to the design parameters of the TF-SAW filter to be tested to ensure the effectiveness of the test. After the excitation signal is applied to the TF-SAW filter, high-precision data acquisition equipment is used to capture the response data processed by the filter. These response data contain all the information about how the filter affects the input signal. Then, the response data in the time domain is converted into a frequency domain response signal, usually using mathematical methods such as Fast Fourier Transform (FFT) to complete the conversion, so as to analyze the frequency response characteristics of the filter. For example, in the application scenario of a mobile communication base station, considering the complex environment and strong electromagnetic interference of the base station, in order to ensure the quality and stability of signal transmission, the TF-SAW filter used must be strictly tested for performance. In this process, we apply an excitation signal in a specific frequency range according to the above method and carefully record the response data. Then, by converting the obtained frequency domain response signal, we can clearly observe the important indicators such as the gain and insertion loss of the filter at different frequencies, so as to judge whether it meets the design requirements and the reliability under long-term working conditions, which provides strong support for ensuring the safety of the base station operation and reducing maintenance costs. The whole process not only tests the basic functions of the filter, but also lays a foundation for further research on its life evaluation.

[0015] Step S2, state feature extraction is performed on the TF-SAW filter based on the frequency domain response signal to generate a state feature sequence.

[0016] Specifically, the state feature sequence is generated by extracting the state features of the TF-SAW filter based on the frequency domain response signal, which is a key step to realize the performance evaluation and life prediction. After obtaining the frequency domain response signal, the working state information contained in the signal needs to be further analyzed. Specifically, a plurality of characteristic parameters reflecting the current health status of the TF-SAW filter are extracted from the frequency domain response signal by a signal processing algorithm, such as center frequency offset, insertion loss change, bandwidth broadening degree, return loss, and passband ripple. The changes of these parameters are often closely related to the failure mechanisms such as material aging, electrode degradation, or interface defects. These characteristic parameters are orderly organized into a time series form of data set, i.e., the state feature sequence, which is used to represent the performance evolution trend of the filter at different working stages. For example, in the actual application scenario of mobile communication base station, the piezoelectric film and electrode structure of the TF-SAW filter may gradually be damaged due to long-term exposure to high-power radio frequency signals and complex electromagnetic interference, resulting in subtle distortion of the frequency domain response signal. By successively extracting the features from the continuously monitored frequency domain response signal and generating the corresponding state feature sequence, the performance degradation of the filter can be identified, and whether it needs to be maintained or replaced in advance can be determined, so as to ensure the stable operation of the communication system and reduce the maintenance cost in the later stage. Therefore, this step not only realizes the non-invasive perception of the internal state of the filter, but also provides a reliable data basis for the subsequent life modeling and fault warning.

[0017] Step S3, performing degradation analysis on the TF-SAW filter based on the state feature sequence to generate a degradation trajectory sequence.

[0018] Specifically, the degradation analysis of the TF-SAW filter based on the state feature sequence to generate a degradation trajectory sequence is a key link to realize the life assessment. After obtaining the state feature sequence, the change trend of the filter performance with time or working cycle needs to be identified and quantified through mathematical modeling and statistical analysis method. Specifically, each parameter in the state feature sequence such as the center frequency drift and the insertion loss increase value is fitted in time sequence, and a degradation model such as linear regression, exponential decay model or Weibull distribution is used to describe the evolution law, so as to reveal the gradual process of the TF-SAW filter from the normal working state to the performance degradation and even to the failure state. Since in the application scenario of the mobile communication base station, the TF-SAW filter is subjected to high frequency signal excitation and temperature cycle stress for a long time, micro cracks or delamination may occur at the interface between the internal piezoelectric material and the metal electrode, resulting in the state feature sequence obtained each time showing a trend of continuous and slow deterioration. Through continuous degradation analysis of these sequence data, a degradation trajectory sequence reflecting the performance degradation process can be generated. The sequence can not only intuitively show the evolution path of the filter health state, but also be used to predict the remaining useful life, provide a scientific basis for the base station maintenance, and further ensure the stability and reliability of the system operation, and effectively reduce the operation and maintenance cost caused by sudden failure.

[0019] Step S4, failure probability calculation of the TF-SAW filter based on the degradation trajectory sequence to obtain the failure probability.

[0020] Specifically, the failure probability calculation of the TF-SAW filter based on the degradation trajectory sequence is the core step to realize the reliability evaluation and life prediction. After obtaining the degradation trajectory sequence, the performance degradation trend reflected by the sequence is used to construct a statistical life model. Usually, the degradation trajectory sequence is input into a random process model such as Wiener process, Gamma process or nonlinear regression model by combining the accelerated degradation test data with the monitoring data under actual working conditions. Through parameter estimation and distribution fitting, the possibility of the TF-SAW filter reaching a certain failure threshold (such as insertion loss exceeding 3 dB or center frequency offset greater than 5%) at different time points is quantified. Since in the application scenario of mobile communication base station, the TF-SAW filter is long-term operated in high temperature, high humidity and strong electromagnetic interference environment, its degradation trajectory sequence often presents nonlinear and individual difference degradation characteristics, therefore, the initial model needs to be dynamically modified by combining the Bayesian updating method to improve the prediction accuracy. For example, when the degradation trajectory sequences of multiple groups of the same type of TF-SAW filter are continuously collected and included in the analysis model, the system can calculate the failure probability of the device within the next 1000 hours under the current working condition. If the probability exceeds the preset safety threshold (such as 10%), the warning mechanism is triggered, prompting the operation and maintenance personnel to replace the device in advance, so as to avoid communication interruption, ensure the continuous and stable operation of the base station system, and effectively reduce the maintenance cost and safety risk caused by sudden failure.

[0021] Step S5, life evaluation of the TF-SAW filter based on the failure probability, generating a life evaluation report.

[0022] Specifically, based on the failure probability, the life assessment report of the TF-SAW filter is generated, which is the final output link of the whole performance monitoring and reliability analysis process. On the basis of obtaining the failure probability, combined with the preset reliability index and system safety threshold, the residual service life of the TF-SAW filter is quantitatively inferred. Specifically, by extrapolating the trend of the failure probability with time, the time point corresponding to the critical failure probability (such as 50% or 80%) is determined, which is the predicted end of life of the filter. At the same time, considering the influence of working environment stress, signal load intensity and historical degradation trajectory sequence, the evaluation results are corrected and weighted to improve the engineering applicability of the prediction. Since in the application scenario of mobile communication base station, multiple TF-SAW filters work in parallel and bear the key signal filtering function, their life status is directly related to the communication quality and service continuity of the base station, therefore the generated life assessment report not only contains the residual life estimation value of a single device, but also integrates the failure risk level, health state classification, trend early warning suggestion and other contents for the operation and maintenance system to make reference decisions. For example, when the failure probability of a TF-SAW filter in a certain base station rises from 5% to 12% in continuous monitoring, the system assesses that its residual life has entered the warning interval, and accordingly generates a life assessment report and pushes it to the maintenance platform, prompting that the device should be replaced as a priority in the next maintenance period, so as to avoid link interruption caused by sudden failure, ensure the long-term stable operation of the communication system, and effectively reduce the maintenance cost and operation risk in the whole life cycle.

[0023] In a specific scenario, the excitation signal is applied to the TF-SAW filter and the response data is collected to generate a frequency domain response signal, which includes: The excitation signal is applied to the TF-SAW filter, and the bandwidth parameter analysis of the excitation signal is performed to obtain a bandwidth feature vector. Based on the bandwidth feature vector, the frequency domain response of the TF-SAW filter is measured to obtain a frequency domain response feature, and the dynamic characteristic analysis of the frequency domain response feature is performed to obtain a frequency domain response signal.

[0024] Specifically, applying an excitation signal to the TF-SAW filter and collecting response data to generate a frequency domain response signal is a basic step for realizing performance monitoring and life evaluation. This process not only requires precise control of the input excitation, but also requires in-depth analysis of its dynamic response characteristics in the frequency domain. In specific implementation, first, an excitation signal is applied to the TF-SAW filter, which is usually generated by a vector network analyzer or a radio frequency signal generator. The frequency range of the excitation signal needs to cover the design working frequency band of the filter. For example, in the application scenario of a mobile communication base station, if the TF-SAW filter is used for 5G NR n78 frequency band (3.3-3.8 GHz), the excitation signal should be swept or stepped in this range to fully stimulate the response behavior of the filter. At the same time, the bandwidth parameter of the excitation signal needs to be analyzed, and the actual output bandwidth characteristics, including passband width, cutoff frequency, roll-off slope and other key indicators, are extracted, and these parameters are constructed into a numerical bandwidth feature vector. This vector not only reflects the frequency spectrum characteristics of the excitation signal itself, but also provides a reference benchmark for subsequent measurements, ensuring the consistency and repeatability of test conditions. Then, based on the bandwidth feature vector, the frequency domain response of the TF-SAW filter is measured, that is, in the same test environment, the output signal processed by the filter is collected by using a high-precision receiving end device to obtain its gain, phase, return loss and insertion loss and other frequency domain response characteristics. These frequency domain response characteristics reflect the selectivity, matching performance and energy loss of the filter at different frequency points, and are the core basis for evaluating the functional integrity of the filter. Subsequently, dynamic characteristic analysis is performed on the frequency domain response characteristics, focusing on the stability, non-linear distortion degree and resonance peak shift trend under continuous excitation. Through sliding window method or time-frequency joint analysis technology, transient fluctuations and long-term drift phenomena in the response signal are identified, and finally a frequency domain response signal containing complete amplitude and phase information is generated. For example, during the long-term operation of the base station, if the piezoelectric layer performance of the TF-SAW filter is degraded due to thermal stress accumulation, the frequency domain response characteristics of the filter will show passband widening and insertion loss increasing. Dynamic characteristic analysis can capture these subtle changes and encode them as frequency domain response signals for subsequent state feature extraction and degradation modeling, thereby providing data support for system-level reliability management. The whole process ensures the quantifiability and traceability of the whole link from excitation input to response output, and lays a solid foundation for subsequent life evaluation.

[0025] In specific scenarios, the state feature extraction of the TF-SAW filter based on the frequency domain response signal generates a state feature sequence, including: Based on the frequency domain response signal, the TF-SAW filter is subjected to surface acoustic wave frequency domain analysis to obtain surface acoustic wave frequency domain characteristics. The acoustic surface wave frequency domain characteristics are used to evaluate an acoustoelectric coupling loss characteristic of the TF-SAW filter, to obtain an acoustoelectric coupling loss parameter, and to decompose a loss characteristic in the acoustoelectric coupling loss parameter to obtain a loss component sequence; An acoustic surface wave propagation attenuation coefficient in the loss component sequence is calculated, and a dispersion effect analysis is performed on the acoustic surface wave propagation attenuation coefficient to obtain a dispersion parameter; A state of the TF-SAW filter is evaluated based on the dispersion parameter to obtain a state characteristic vector, and a multi-dimensional feature combination is performed on the state characteristic vector to obtain a state characteristic sequence.

[0026] Specifically, state feature extraction is performed on the TF-SAW filter based on the frequency domain response signal to generate a state feature sequence, which is a key link for in-depth understanding of the internal working state and potential degradation mechanism of the TF-SAW filter. This process starts from the frequency domain response signal and gradually strips out multi-dimensional features reflecting the health status of the device through a series of physical mechanism-driven analysis methods. In specific implementation, first, surface acoustic wave frequency domain analysis is performed on the TF-SAW filter based on the frequency domain response signal, that is, the resonant peak, anti-resonance point and phase jump region of the passband edge related to the surface acoustic wave propagation are identified in the obtained frequency domain response signal, so as to extract the key information such as the main modal frequency, group delay fluctuation and energy concentration of the surface acoustic wave, and form the surface acoustic wave frequency domain features. These features are directly related to the sound wave excitation efficiency and propagation integrity between the piezoelectric film and the electrode structure in the filter. On this basis, the acoustoelectric coupling loss characteristics of the TF-SAW filter are further evaluated using the surface acoustic wave frequency domain features, because the decrease of the acoustoelectric conversion efficiency is often an early sign of device aging. By establishing a mapping relationship between the acoustoelectric coupling coefficient and the insertion loss, and combining the measured frequency domain response data, the acoustoelectric coupling loss parameter is back calculated, which comprehensively reflects the energy loss in the process of converting electrical energy into mechanical energy. Subsequently, the acoustoelectric coupling loss parameter is decomposed into loss components of multiple physical sources such as dielectric loss, electrode resistance loss, interface scattering loss and radiation loss, thereby obtaining a loss component sequence, which can finely depict the influence weight of different failure mechanisms on the overall performance. Next, the surface acoustic wave propagation attenuation coefficient in the loss component sequence is calculated, that is, the energy attenuation degree per unit propagation distance, which is extremely sensitive to material aging, micro-crack propagation and other defects. Then, the surface acoustic wave propagation attenuation coefficient is analyzed for frequency dispersion effect, the nonlinear trend of which with frequency change is investigated, and features such as wave velocity shift and group velocity fluctuation are extracted to obtain frequency dispersion parameters. These parameters can reveal the propagation anomalies caused by interface degradation or stress accumulation when the sound wave propagates in the multi-layer structure. Finally, the state of the TF-SAW filter is evaluated based on the frequency dispersion parameters, for example, whether there is electrode delamination or piezoelectric layer polarization recession, and these evaluation results are quantified into a set of comparable numerical values to form a state feature vector. Then, through multi-dimensional feature combination in the time dimension of the state feature vectors in multiple monitoring periods, indicators such as frequency shift, loss growth and dispersion intensification are fused to form a state feature sequence with time sequence characteristics, which is used for subsequent degradation modeling. For example, during the long-term operation of a mobile communication base station, if the electrode metal of a certain TF-SAW filter diffuses due to continuous high-power work, the acoustoelectric coupling loss will gradually increase, and the resistance loss proportion in the loss component sequence will rise, while the frequency dispersion parameters show unstable wave propagation at high frequencies. The finally generated state feature sequence clearly presents this degradation trend, providing data support for accurate life prediction and preventive maintenance.

[0027] In a specific scenario, the state feature sequence is used to perform degradation analysis on the TF-SAW filter to generate a degradation trajectory sequence, including: Based on the state feature sequence, a performance degradation parameter of the TF-SAW filter is calculated, and a time domain correlation calculation is performed on the performance degradation parameter to obtain degradation correlation data; Based on the degradation correlation data, a degradation process analysis is performed on the TF-SAW filter to obtain a degradation process feature group, and a degradation trajectory fitting is performed on the degradation process feature group to generate a degradation trajectory sequence.

[0028] Specifically, the TF-SAW filter is analyzed based on the state feature sequence to generate a degradation trajectory sequence, which is a key bridge connecting state monitoring and life prediction. This process aims to extract a quantitative path reflecting the device performance degradation rule from the multi-dimensional and time-series state features. In specific implementation, first, the performance degradation parameters of the TF-SAW filter are calculated based on the state feature sequence. These parameters are usually selected from indicators sensitive to device aging, such as center frequency drift, insertion loss increment, quality factor drop rate, and acoustic-electric coupling coefficient attenuation rate. They are all derived from the state feature vectors extracted in the previous step and form sequence data with monotonic or trend change characteristics on the time axis. Subsequently, time-domain correlation calculation is performed on the performance degradation parameters. That is, the Pearson correlation coefficient, mutual information, or dynamic time warping (DTW) method is used to analyze the synergistic change relationship between the degradation parameters in the time evolution process, identify which parameters show strong coupling or consistent change trend in long-term operation, and obtain degradation correlation data. This data not only reveals the internal relationship between different failure mechanisms, but also provides a basis for variable selection for subsequent modeling. On this basis, the degradation process of the TF-SAW filter is analyzed based on the degradation correlation data, focusing on identifying whether the degradation behavior presents linear, exponential, power-law, or segmented evolution characteristics. For example, in the application scenario of mobile communication base stations, the TF-SAW filter works in a high-temperature and high-humidity environment with high-frequency excitation for a long time. Its degradation process often goes through three stages: initial slight fluctuation, medium-term accelerated degradation, and late-stage rapid failure. Through clustering and trend segmentation of the degradation correlation data, degradation process feature groups characteristic of each stage can be extracted, including initial degradation rate, inflection point occurrence time, acceleration change rate, and other key features. These feature groups not only reflect the actual degradation path of the device in a complex electromagnetic environment, but also provide support for building a degradation model with clear physical meaning. Then, the degradation trajectory fitting is performed on the degradation process feature groups. The least squares method, Kalman filtering, or nonlinear regression models (such as Weibull, Logistic, or Arrhenius models) are used to fit the discrete degradation data points into a continuous, smooth, and predictive curve, i.e., to generate a degradation trajectory sequence. This sequence can intuitively show the whole process of the TF-SAW filter from the healthy state to the functional degradation, and can be used to extrapolate its remaining useful life. For example, when the insertion loss of a TF-SAW filter in a base station rises from 0.8 dB to 1.5 dB with working time, and the degradation correlation data shows that it is highly positively correlated with the center frequency offset, the fitted degradation trajectory sequence will show a slow-to-fast growth trend, indicating that the device has entered the accelerated aging stage. Therefore, the system can give an early warning and arrange replacement, thereby avoiding communication link interruption and ensuring the stability and reliability of the base station operation.The whole degradation analysis process realizes the transformation from multi-dimensional state perception to single degradation path, and lays a solid foundation for subsequent failure probability calculation and life evaluation report generation.

[0029] In a specific scenario, the degradation process analysis of the TF-SAW filter based on the degradation correlation data obtains a degradation process feature group, including: The parameter coupling analysis is performed on the degradation correlation data to obtain a parameter coupling matrix, and the primary and secondary factors of the parameter coupling matrix are identified to obtain a degradation factor weight sequence; Based on the degradation factor weight sequence, the TF-SAW filter is subjected to multi-parameter synergistic analysis to obtain a parameter synergistic interaction map, and the parameter synergistic interaction map is subjected to feature extraction to obtain a degradation feature vector; The degradation process feature group is obtained by performing dynamic evolution analysis on the degradation feature vector.

[0030] Specifically, based on the degradation correlation data, a degradation process analysis is performed on the TF-SAW filter to obtain a degradation process feature group, which is a key link for in-depth revealing the evolution law of the internal failure mechanism of the device. This process not only focuses on the change trend of a single performance parameter, but also emphasizes the interaction and collaborative evolution among multiple degradation indicators. In specific implementation, first, a parameter coupling analysis is performed on the degradation correlation data. The degradation correlation data is obtained by performing time domain correlation calculation on the performance degradation parameters in the early stage, and reflects the linkage relationship of key parameters such as center frequency drift, insertion loss increase, and quality factor decrease in the time dimension. By constructing a correlation coefficient matrix or mutual information matrix between parameters, and further using gray correlation analysis, partial correlation analysis or dynamic time warping (DTW) distance measurement method, the mutual influence strength between parameters is quantified, so as to obtain a parameter coupling matrix. The matrix represents the coupling degree between different degradation variables in the form of numerical values. For example, electrode aging may cause the synchronous rise of insertion loss and return loss, and this strong coupling relationship will be reflected as a high value element in the matrix. On this basis, a primary and secondary factor identification is performed on the parameter coupling matrix, that is, by using mathematical methods such as eigenvalue decomposition, principal component analysis (PCA) or entropy weight method, the key parameters playing a leading role in the overall degradation process are identified. For example, if it is found that the insertion loss and multiple other parameters all present high coupling values, it can be determined that the insertion loss is the main factor of degradation, and the other parameters are secondary response variables. Thus, a degradation factor weight sequence is generated, which sorts the importance of each degradation parameter in the form of weight coefficients, providing priority basis for subsequent analysis. Then, based on the degradation factor weight sequence, a multi-parameter collaborative analysis is performed on the TF-SAW filter. The high-weight parameters are used as driving variables to construct the evolution trajectory in the multi-dimensional space, and the dynamic interaction relationship between parameters is drawn by using visualization technology or graph neural network model to form a parameter collaborative action atlas. The atlas can intuitively show how each parameter influences each other in different working stages, for example, the frequency offset is dominant in the early stage, the loss and bandwidth widening collaborative deterioration appear in the middle stage, and the overall performance collapse appears in the later stage. Subsequently, a feature extraction is performed on the parameter collaborative action atlas, and an image recognition algorithm or a topological feature extraction method is used to identify key structural patterns such as cluster aggregation, chain conduction or mutation transition from the atlas, and convert them into a numerical degradation feature vector containing degradation starting point, acceleration inflection point, coupling strength peak value and other information. Finally, a dynamic evolution analysis is performed on the degradation feature vector, and a time sliding window technology or a state transition model is combined to track the change trajectory of the degradation feature vector in the long-term operation, identify the stage characteristics of the degradation process such as the early stable period, the middle acceleration period and the late failure precursor period, and finally integrate them into the degradation process feature group.For example, in the application scenario of mobile communication base station, the performance of the piezoelectric layer of a certain TF-SAW filter is degraded due to long-term high temperature and high humidity operation. The degradation correlation data shows that the insertion loss is highly negatively correlated with the acoustic-electric coupling coefficient. After parameter coupling analysis, it is identified that the degradation of the electrode-dielectric interface is the main cause. The generated degradation process feature set clearly reflects the transition process from slow degradation to sudden degradation of the device, providing a solid mechanism support for predicting the remaining life and formulating maintenance strategies.

[0031] In specific scenarios, the failure probability calculation of the TF-SAW filter based on the degradation trajectory sequence is performed to obtain the failure probability, including: An initially set failure threshold range is obtained, and the intersection of the failure threshold range and the degradation trajectory sequence is extracted to obtain an initial failure time point; Based on the initial failure time point, a statistical distribution fitting is performed to obtain an initial failure time distribution function, and a parameter sensitivity analysis is performed on the initial failure time distribution function to obtain a parameter sensitive weight; The initial failure time distribution function is optimized and adjusted using the parameter sensitive weight to obtain an optimized failure time distribution function, and quantile number estimation is performed on the optimized failure time distribution function to generate quantile number failure time; Based on the quantile number failure time, the failure probability under different confidence levels is calculated.

[0032] Specifically, the failure probability calculation of the TF-SAW filter based on the degradation trajectory sequence is the core quantitative step to realize its reliability evaluation and life prediction. This process compares the degradation trajectory sequence with the preset performance tolerance boundary, and then establishes a statistical model to predict the possibility of device functional failure at future time points. In specific implementation, first, the pre-set failure threshold range needs to be obtained, which is determined according to the design specification and system application requirements of the TF-SAW filter. For example, in the application scenario of mobile communication base station, the insertion loss exceeding 2.5 dB, the center frequency offset greater than ±50 MHz or the return loss degradation to less than 10 dB are defined as functional failure state. These indicators constitute a multi-dimensional or single-dimensional failure threshold range, which is used as a standard to judge whether the device exits the normal working interval. Subsequently, the intersection points of the failure threshold range and the degradation trajectory sequence are extracted, that is, the time point at which the degradation trajectory curve first crosses the failure threshold is found. This time point is the initial failure time point, which reflects the critical moment when the device enters functional failure from performance degradation under the current working condition. A group of initial failure time points can be obtained through multiple samples or multiple monitoring paths, providing a data basis for subsequent statistical modeling. On this basis, statistical distribution fitting is carried out based on the initial failure time points. Common life distribution models such as Weibull distribution, normal distribution, lognormal distribution or gamma distribution are used to fit the probability density function of these time points, so as to obtain the initial failure time distribution function. This function describes the time distribution law of the TF-SAW filter group under the same stress condition, and is a mathematical expression form for evaluating its reliability. However, due to factors such as environmental fluctuations, manufacturing deviations and measurement noises in actual applications, the distribution function obtained by initial fitting may have parameter uncertainty. Therefore, parameter sensitivity analysis needs to be further carried out on the initial failure time distribution function to evaluate the influence degree of shape parameter, scale parameter or location parameter change on the failure probability prediction result, identify the key parameter that has the greatest influence on the model output, and obtain the parameter sensitivity weight accordingly. This weight reflects the relative importance of each distribution parameter in the prediction process. Then, the initial failure time distribution function is optimized and adjusted using the parameter sensitivity weight, that is, the parameters with higher sensitivity are given higher confidence or a Bayesian updating mechanism is introduced for dynamic correction. Combined with the acceleration degradation test data or field monitoring feedback information, the adaptability and robustness of the model are improved, so as to obtain the optimized failure time distribution function. Compared with the initial model, this function can more accurately reflect the randomness and nonlinearity of the real degradation process.Subsequently, quantile estimation is performed on the optimized failure time distribution function, i.e., the failure time values at different cumulative probability levels (such as 10%, 50%, and 90%) are calculated to generate quantile failure times, for example, P10 represents that 10% of the devices are expected to fail before this time, and P50 is the median life. These quantiles provide multi-level life prediction results to meet the needs of different risk preference operation and maintenance decision-making. Finally, the failure probability at different confidence levels is calculated based on the quantile failure time, for example, if the optimized model predicts that the cumulative failure probability at the 5000th hour of base station operation is 8% and rises to 23% at the 8000th hour, it indicates that the failure risk of the TF-SAW filter significantly increases as the working time increases. For example, in a certain 5G base station dense deployment area, the operation and maintenance system continuously monitors the degradation trajectory sequences of multiple TF-SAW filters and finds that they accelerate degradation in a high-temperature environment, with initial failure time points concentrated in the 6000-7500 hour interval. After confirming that the shape parameter is the most sensitive through parameter sensitivity analysis, the Weibull model is optimized using field data, and the optimized failure time distribution function generated shows that under a 90% confidence level, the life of this batch of devices does not exceed 8200 hours. Based on this, a preventive replacement plan can be developed to avoid sudden communication interruptions. The entire failure probability calculation process realizes a closed loop from degradation trajectories to risk quantification, providing a scientific basis for ensuring the long-term stable operation of base station systems and reducing maintenance costs.

[0033] In specific scenarios, the intersection of the failure threshold range and the degradation trajectory sequence is extracted to obtain the initial failure time point, including: The failure threshold range is subjected to segment discretization processing to obtain threshold discrete points. Based on the threshold discrete points, intersection detection is performed on the degradation trajectory sequence to obtain a trajectory intersection sequence, and the time intersection coordinates of the trajectory intersection sequence are extracted to obtain the initial failure time point.

[0034] Specifically, extracting the intersection of the failure threshold range and the degradation trajectory sequence to obtain the initial failure time point is a crucial geometric and numerical analysis step in the TF-SAW filter failure prediction process. This step aims to provide a reliable time reference for subsequent statistical modeling by accurately identifying the specific moment when the degradation behavior crosses the safety boundary. In practice, the failure threshold range is first discretized by dividing a continuous failure judgment interval (e.g., insertion loss greater than 2.5 dB is considered a failure) into multiple equally spaced or adaptively distributed threshold discrete points. For example, the 2.5 dB to 3.0 dB interval is divided into 6 discrete points with a step size of 0.1 dB, forming an ordered set of numerical values, which are the threshold discrete points. This discretization not only improves the resolution of intersection detection but also facilitates efficient comparison and searching in digital signal processing systems. It is particularly suitable for situations where there are fluctuations or measurement noise in the degradation trajectory sequence, thereby avoiding misjudgments or missed judgments caused by a single threshold judgment. Based on this, intersection detection is performed on the degradation trajectory sequence based on the threshold discrete points. The degradation trajectory sequence is typically time-series data composed of performance degradation parameters (such as center frequency offset, insertion loss value, etc.) at multiple time points, representing a curve that evolves over time. By performing a crossover detection algorithm on this curve for each threshold discrete point, for example, using linear interpolation to determine whether two adjacent sampling points cross a certain threshold level, if a crossover occurs, the intersection position is recorded. All such intersections constitute the trajectory crossover sequence. This trajectory crossover sequence not only includes the point that first crosses the failure boundary but may also include multiple crossovers caused by short-term fluctuations. Therefore, it needs to be filtered in conjunction with physical meaning, and usually the earliest and continuously rising crossover point is retained as a valid intersection. Subsequently, the time crossover coordinates of the trajectory crossover sequence are extracted, that is, the timestamp information corresponding to each intersection point is obtained. These time coordinates reflect the specific time when the performance parameters of the TF-SAW filter first touch or exceed the failure threshold in actual operation. After data cleaning and consistency verification, the initial failure time point is finally determined. For example, in a mobile communication base station application scenario, a certain TF-SAW filter operates in a high-power RF environment for a long time. Its insertion loss gradually increases over time. The degradation trajectory sequence shows that this parameter slowly increases from an initial 0.9 dB to 2.6 dB. When the system sets 2.5 dB as the failure threshold and discretizes it into segments of 2.50, 2.52, 2.54, etc., by scanning the degradation trajectory sequence point by point, it was detected that the filter first crossed the 2.50 dB threshold discrete point around the 7200th hour. The precise intersection time was obtained through linear interpolation at 7215 hours, which is the extracted initial failure time point. This time point becomes the basic data for subsequently constructing the failure time distribution function, providing empirical evidence for evaluating the reliability of devices in the same batch, and thus supporting the optimization of base station maintenance strategies and the precise scheduling of spare parts.

[0035] In a specific scenario, the life assessment report is generated based on the failure probability of the TF-SAW filter. The life prediction sequence is obtained by performing life prediction analysis on the TF-SAW filter based on the failure probability. The risk level assessment result is obtained by performing risk level assessment on the life prediction sequence. The evaluation report content of the TF-SAW filter is generated based on the risk level assessment result, and the evaluation report content is formatted and arranged to obtain the life assessment report.

[0036] Specifically, based on the failure probability, the life assessment report of the TF-SAW filter is generated, which is the final output link of the whole health monitoring and reliability management process. This process converts the abstract probability calculation result into an operable engineering decision basis. In specific implementation, first, based on the failure probability, the life prediction analysis of the TF-SAW filter is performed, that is, by extrapolating the time point at which the failure probability reaches a preset risk threshold (such as 10% or 50%), the remaining service life of the filter is determined according to the change curve of the failure probability with time. This process usually adopts Monte Carlo simulation, Bayesian update or degradation model inversion method, and comprehensively considers the historical degradation trajectory sequence, the stress level of the working environment and the individual differences of the device, so as to generate a set of prediction results with time sequence characteristics, that is, a life prediction sequence. This sequence not only contains the expected life median, but also covers the life interval under different confidence levels, such as the range of P10 (10% probability of failure) to P90 (90% probability of failure), providing multi-dimensional life judgment basis for operation and maintenance personnel. On this basis, the risk level assessment of the life prediction sequence is performed, that is, the prediction result is mapped into a pre-defined risk classification system. For example, the remaining life greater than 5000 hours and the failure probability less than 5% is classified as "low risk", the remaining life between 2000-5000 hours and the failure probability between 5%-15% is classified as "medium risk", and the remaining life less than 2000 hours or the failure probability more than 15% is classified as "high risk". Through setting rules or introducing fuzzy logic algorithm, the risk level assessment result is obtained by comprehensively considering the factors such as life length, degradation rate and environmental severity. The result is presented in the form of grade label (such as R1, R2, R3) or color identification (green, yellow, red), which is convenient for quickly identifying the device state. Subsequently, the evaluation report content of the TF-SAW filter is generated based on the risk level assessment result, including device number, installation position, current health status, remaining life prediction value, risk level, trend analysis graph (such as degradation trajectory and failure probability curve) and maintenance suggestion (such as "it is suggested to replace in the next quarter" or "it can continue to monitor and run"). Finally, the evaluation report content is formatted and arranged according to the communication industry standard or enterprise internal specification, and is organized into a structured document format (such as PDF or XML), embedded with charts, time stamps and digital signatures, to form the final life assessment report. For example, in the application scenario of mobile communication base station, a TF-SAW filter is analyzed, and the failure probability reaches 12% at 8000 hours. The life prediction sequence shows that the P50 life is 8200 hours, and the system determines it as "medium risk" accordingly. The generated life assessment report is automatically pushed to the base station operation and maintenance platform, prompting the maintenance personnel to check the device in the subsequent inspection, and preparing spare parts, so as to realize the closed-loop management from data perception to decision support, effectively guarantee the stable operation of the communication system and optimize the maintenance resource allocation.

[0037] Please refer toFigure 2 , Figure 2 is a framework diagram of an embodiment of the life evaluation device of the TF-SAW filter. As shown in Figure 2 , the life evaluation device of the TF-SAW filter includes a collection unit for applying an excitation signal to the TF-SAW filter and collecting response data to generate a frequency domain response signal; an extraction unit for performing state feature extraction on the TF-SAW filter based on the frequency domain response signal to generate a state feature sequence; an analysis unit for performing degradation analysis on the TF-SAW filter based on the state feature sequence to generate a degradation trajectory sequence; a calculation unit for calculating the failure probability of the TF-SAW filter based on the degradation trajectory sequence to obtain the failure probability; and an evaluation unit for evaluating the life of the TF-SAW filter based on the failure probability to generate a life evaluation report.

[0038] With reference to Figure 3 , the computer device in the embodiment of the present application can have an internal structure as shown in Figure 3 . The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The database of the computer device is configured to store corresponding data in the embodiment. The network interface of the computer device is configured to communicate with an external terminal through a network connection. The computer program is executed by the processor to implement the above method.

[0039] Those skilled in the art can understand that Figure 3 the structure shown in the embodiment is only a block diagram of part of the structure related to the present application, and does not constitute a limitation on the computer device to which the present application is applied.

[0040] The embodiment of the present application also provides a computer readable storage medium having a computer program stored thereon, and the computer program is executed by the processor to implement the above method. It can be understood that the computer readable storage medium in the embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0041] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiment methods can be included. Any reference to memory, storage, database or other medium provided by the present application and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0042] In some embodiments, the apparatus provided by the embodiments of the present disclosure has functions or includes modules that can be used to perform the methods described in the above method embodiment, and the specific implementation can refer to the description of the above method embodiment. For brevity, it will not be repeated here.

[0043] The above description of each embodiment tends to emphasize the differences between the embodiments, and the same or similar parts can be mutually referred to, and for brevity, will not be repeated here.

[0044] In several embodiments provided in the present application, it should be understood that the disclosed methods and apparatuses can be implemented in other ways. For example, the above-described apparatus implementation is only schematic, for example, the division of modules or units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the shown or discussed mutual ones can be indirect coupling or communication connection through some interfaces, devices or units, and can be electrical, mechanical or other forms.

[0045] The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the present embodiment.

[0046] In addition, the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0047] If the integrated unit is realized in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the methods of the various embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0048] If the technical solutions of the present application involve personal information, the product applying the technical solutions of the present application has been explicitly informed of the personal information processing rules before processing the personal information, and has obtained the personal independent consent. If the technical solutions of the present application involve sensitive personal information, the product applying the technical solutions of the present application has obtained the personal independent consent before processing the sensitive personal information, and at the same time meets the requirement of "explicit consent". For example, at the personal information collection device such as camera, a clear and prominent mark is set to inform that the personal information collection range has been entered, and the personal information will be collected. If the person voluntarily enters the collection range, it is regarded as agreeing to collect the personal information; or on the device for processing personal information, the personal information processing rules are informed by using obvious marks / information, and the personal authorization is obtained by means of pop-up information or asking the person to upload his / her personal information, etc. The personal information processing rules can include personal information processor, personal information processing purpose, processing method, and personal information type, etc.

Claims

1. A method for lifetime assessment of a TF-SAW filter, characterized in that, Includes the following steps: An excitation signal is applied to the TF-SAW filter and response data is acquired to generate a frequency domain response signal; Based on the frequency domain response signal, the state features of the TF-SAW filter are extracted to generate a state feature sequence; Degradation analysis is performed on the TF-SAW filter based on the state feature sequence to generate a degradation trajectory sequence; The failure probability of the TF-SAW filter is calculated based on the degradation trajectory sequence to obtain the failure probability. Based on the failure probability, the lifetime of the TF-SAW filter is assessed, and a lifetime assessment report is generated.

2. The lifetime assessment method for a TF-SAW filter according to claim 1, characterized in that, The process of applying an excitation signal to the TF-SAW filter and acquiring response data to generate a frequency domain response signal includes: An excitation signal is applied to the TF-SAW filter, and bandwidth parameter analysis is performed on the excitation signal to obtain a bandwidth feature vector; The frequency domain response of the TF-SAW filter is measured based on the bandwidth feature vector to obtain the frequency domain response characteristics. The dynamic characteristics of the frequency domain response characteristics are then analyzed to obtain the frequency domain response signal.

3. The lifetime assessment method for a TF-SAW filter according to claim 1, characterized in that, The step of extracting state features from the TF-SAW filter based on the frequency domain response signal to generate a state feature sequence includes: Based on the frequency domain response signal, surface acoustic wave frequency domain analysis is performed on the TF-SAW filter to obtain surface acoustic wave frequency domain characteristics. The acoustic-electric coupling loss characteristics of the TF-SAW filter are evaluated using the frequency domain characteristics of the surface acoustic wave, the acoustic-electric coupling loss parameters are obtained, and the loss characteristics in the acoustic-electric coupling loss parameters are decomposed to obtain the loss component sequence. Calculate the surface acoustic wave propagation attenuation coefficient in the loss component sequence, and perform dispersion effect analysis on the surface acoustic wave propagation attenuation coefficient to obtain dispersion parameters; The state of the TF-SAW filter is evaluated based on the dispersion parameters to obtain a state feature vector, and the state feature vector is combined with multi-dimensional features to obtain a state feature sequence.

4. The lifetime assessment method for a TF-SAW filter according to claim 1, characterized in that, The degradation analysis of the TF-SAW filter based on the state feature sequence to generate a degradation trajectory sequence includes: The performance degradation parameters of the TF-SAW filter are calculated based on the state feature sequence, and the time-domain correlation of the performance degradation parameters is calculated to obtain degradation correlation data. The degradation process of the TF-SAW filter is analyzed based on the degradation correlation data to obtain a degradation process feature group, and a degradation trajectory sequence is fitted to the degradation process feature group.

5. The lifetime assessment method for a TF-SAW filter according to claim 1, characterized in that, The calculation of the failure probability of the TF-SAW filter based on the degradation trajectory sequence to obtain the failure probability includes: Obtain a pre-set failure threshold range, and extract the intersection of the failure threshold range and the degradation trajectory sequence to obtain the initial failure time point; Based on the initial failure time point, a statistical distribution is fitted to obtain the initial failure time distribution function, and a parameter sensitivity analysis is performed on the initial failure time distribution function to obtain the parameter sensitivity weights. The initial failure time distribution function is optimized and adjusted using the parameter-sensitive weights to obtain an optimized failure time distribution function. Then, quantile estimation is performed on the optimized failure time distribution function to generate quantile failure times. The failure probability at different confidence levels is calculated based on the quantile failure time.

6. The lifetime assessment method for a TF-SAW filter according to claim 5, characterized in that, The step of extracting the intersection of the failure threshold range and the degradation trajectory sequence to obtain the initial failure time point includes: The failure threshold range is discretized into segments to obtain threshold discrete points; Intersection detection is performed on the degraded trajectory sequence based on the threshold discrete points to obtain the trajectory intersection sequence, and the time intersection coordinates of the trajectory intersection sequence are extracted to obtain the initial failure time point.

7. The lifetime assessment method for a TF-SAW filter according to claim 1, characterized in that, The lifetime assessment of the TF-SAW filter based on the failure probability, generating a lifetime assessment report, includes: Based on the failure probability, a lifetime prediction analysis is performed on the TF-SAW filter to obtain a lifetime prediction sequence. The risk level assessment is performed on the lifetime prediction sequence to obtain the risk level assessment result; Based on the risk level assessment results, an assessment report for the TF-SAW filter is generated, and the assessment report is formatted and organized to obtain a lifespan assessment report.

8. A lifetime assessment device for a TF-SAW filter, characterized in that, include: The acquisition unit is used to apply an excitation signal to the TF-SAW filter and acquire response data to generate a frequency domain response signal. The extraction unit is used to extract state features from the TF-SAW filter based on the frequency domain response signal and generate a state feature sequence. Analysis unit, used to perform degradation analysis on the TF-SAW filter based on the state feature sequence, and generate degradation trajectory sequence; The calculation unit is used to calculate the failure probability of the TF-SAW filter based on the degradation trajectory sequence to obtain the failure probability; An evaluation unit is used to evaluate the lifetime of the TF-SAW filter based on the failure probability and generate a lifetime evaluation report.

9. A computer device, characterized in that, The device includes a memory and a processor that are coupled to each other. The memory stores program instructions, and the processor executes the program instructions to implement the lifetime assessment method for the TF-SAW filter according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The device stores program instructions that can be executed by a processor, the program instructions being used to implement the lifetime assessment method for the TF-SAW filter according to any one of claims 1 to 7.

Images