A click life test system for a mouse microswitch

Abstract

The application belongs to the technical field of electronic component testing, and discloses a click life test system for a mouse micro switch; acoustic pulse sequences, electrical guide communication signals and key post displacement trajectory data are acquired through a data acquisition module, an acoustic-electric synchronous deviation is extracted by using a synchronous analysis module, effective working distance and dynamic contact stiffness are determined through a working calculation module, micro impedance noise entropy is calculated by using an impedance analysis module, and adaptive driving force parameters are output in combination with a parameter generation module. A deformation analysis module is applied to track the position change of the maximum deflection point, a stiffness-deformation coupling degradation atlas is constructed, a multi-dimensional index is calculated by fusing a degradation calculation module, a touch degradation index is calculated, dynamic optimization of a parameter adjustment module is realized, and accurate prediction of a life prediction module is realized. The application realizes multi-dimensional collaborative analysis, adaptive testing and nonlinear prediction, and improves the accuracy and reliability of micro switch life testing.

Description

Technical Field

[0001] This invention relates to the field of electronic component testing technology, and more specifically, to a click life testing system for a mouse micro switch. Background Technology

[0002] Current microswitch lifespan testing technologies have numerous shortcomings, failing to meet the dual requirements of high reliability and superior tactile feedback in modern electronic devices. Traditional testing methods employ a single-dimensional evaluation model, often focusing only on electrical conduction or simple mechanical travel, neglecting the importance of acoustic characteristics and tactile experience. This leads to many superficially qualified products quickly exhibiting "double-click failure" or "poor tactile feedback" issues during actual use. In high-frequency usage scenarios such as gaming mice, the primary user complaint is not complete switch failure, but rather abnormal tactile feedback or inconsistent response. These subtle changes often cannot be effectively captured and quantified within the existing testing framework. Furthermore, traditional testing uses constant parameters and fixed conditions, failing to simulate varying user habits and environmental factors, such as the alternation of rapid clicks and long presses, or the dynamic changes in material properties under different temperature and humidity conditions. This results in significant discrepancies between laboratory data and actual lifespan. Current technologies have failed to establish the intrinsic relationship between reed deformation characteristics, contact state, and acoustic response, resulting in a severe lack of understanding of complex failure mechanisms. For example, the synergistic effect between contact oxide layer formation and reed fatigue is completely ignored, failing to explain the "self-recovery" and "intermittent failure" phenomena observed in many switches during use. In terms of prediction methods, traditional linear models are overly simplistic and fail to reflect the rapid degradation characteristics commonly seen at the end of a switch's lifespan, causing a sudden deterioration in user experience without warning. Existing systems cannot identify early signs of switch degradation, forcing manufacturers to passively respond to existing batch failures and miss the golden opportunity for preventative improvements.

[0003] In view of this, the present invention proposes a click life testing system for mouse microswitches to solve the above problems. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a click life testing system for a mouse microswitch, comprising: The data acquisition module is used to acquire the acoustic pulse sequence, electrical conduction signal and key post displacement trajectory data of the micro switch during the click test. The acoustic pulse sequence includes the acoustic characteristic waveform generated by the deformation of the internal spring and the collision of the contacts. The synchronization analysis module is used to perform time-domain alignment between the acoustic pulse sequence and the electrical conduction signal, and extract the time difference between the peak time of the acoustic pulse and the rising edge time of the electrical conduction, which is denoted as the acoustic-electric synchronization deviation. The work calculation module is used to calculate the effective work distance from the deformation of the reed to the closure of the contact point based on the displacement trajectory data of the key post and the acoustic-electric synchronization deviation, and to derive the dynamic contact stiffness based on the effective work distance. The impedance analysis module is used to inject a high-frequency microampere-level constant current probe signal across the contacts of a microswitch, collect the micro-voltage fluctuation signal during the clicking process, extract the high-frequency noise energy distribution in the micro-voltage fluctuation signal, and calculate the micro-impedance noise entropy. The parameter generation module is used to generate adaptive driving force parameters based on dynamic contact stiffness and micro impedance noise entropy. The striking execution module is used to drive the electromagnetic striking mechanism to perform a clicking action on the micro switch according to the adaptive driving force parameters, and to collect the dynamic deformation curve of the reed in real time during the striking process. The deformation analysis module is used to extract the position of the maximum deflection point in the dynamic deformation curve of the reed, track the offset trajectory of the position of the maximum deflection point with the number of clicks, and construct a stiffness-deformation coupling degradation map by combining the effective work distance. The degradation calculation module is used to integrate acoustic-electric synchronization deviation, microscopic impedance noise entropy and stiffness-deformation coupling degradation spectrum to calculate multi-dimensional tactile degradation index. The parameter adjustment module is used to dynamically adjust the frequency scanning range of the high-frequency microampere constant current probe signal and update the micro impedance noise entropy based on the multi-dimensional tactile degradation index. The lifespan prediction module is used to output the click lifespan prediction result of the micro switch when the multi-dimensional tactile degradation index meets the lifespan termination judgment condition.

[0005] The technical effects and advantages of the mouse microswitch click life testing system of the present invention are as follows: This invention achieves a comprehensive technological breakthrough in microswitch life testing. Multi-dimensional parameter collaborative analysis improves the accuracy of degradation feature extraction, solving the problem that single parameters are insufficient to reflect complex failure mechanisms. The adaptive testing strategy overcomes the disconnect between fixed parameters and actual use, making life assessment more practical. Acoustic-electrical synchronous analysis and deformation feature tracking enable precise location of degradation sources, providing a microscopic perspective for failure mechanism research. Microscopic impedance noise entropy overcomes the sensitivity bottleneck of early degradation detection, making preventative intervention possible. The self-learning mechanism of this invention enhances adaptability to different switch types, and the multiple verification system strengthens the reliability of test results. Attached Figure Description

[0006] Figure 1 This is a schematic diagram of a mouse microswitch click life test system according to the present invention. Detailed Implementation

[0007] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0008] This application provides a click life testing system for mouse microswitches. The system's execution entities include, but are not limited to, microswitch testing platforms, electronic component life assessment centers, tactile performance analysis systems, and quality inspection platforms, which can be considered general computing nodes in this application. The testing system includes, but is not limited to, at least one cloud-based life analysis engine, a distributed life testing system, and an intelligent degradation detector.

[0009] Please see Figure 1 In this embodiment of the invention, a click life testing system for a mouse microswitch includes: The data acquisition module is used to acquire the acoustic pulse sequence, electrical conduction signal, and key post displacement trajectory data of the microswitch during the click test. The acoustic pulse sequence includes the acoustic characteristic waveform generated by the deformation of the internal reed and the collision of the contacts, which is acquired in real time by a high-sensitivity piezoelectric sensor. The electrical conduction signal is acquired by a micro-current detection circuit to collect the voltage change of the contact conduction state, and the key post displacement trajectory data is recorded by a high-precision laser displacement sensor to record the movement process of the key post. These three types of data constitute a complete physical characteristic description of the microswitch click process, reflecting the acoustic, electrical, and mechanical characteristics, respectively, providing comprehensive raw data for subsequent analysis and ensuring the integrity and accuracy of the test.

[0010] The synchronization analysis module is used to align the acoustic pulse sequence with the electrical conduction signal in the time domain, extracting the time difference between the peak moment of the acoustic pulse and the rising edge of the electrical conduction signal, denoted as the acoustic-electric synchronization deviation. This module first performs bandpass filtering on the acoustic pulse sequence to extract characteristic signals within the resonant frequency band of the reed. Then, it precisely locates the rising edge of the electrical conduction signal and calculates the time difference between the two signals by comparing their key time points. The acoustic-electric synchronization deviation is a key indicator for evaluating the electromechanical conversion efficiency within the switch. This deviation gradually changes with the increase of the switch's service life, reflecting the degradation process of the contact state.

[0011] The work calculation module calculates the effective work distance from reed deformation to contact closure based on the key post displacement trajectory data and the acoustic-electric synchronization deviation, and derives the dynamic contact stiffness based on this effective work distance. This module analyzes the displacement process of the key post from activation to maximum travel, and combines this with the acoustic-electric synchronization deviation to accurately calculate the effective displacement distance from the start of reed deformation to complete contact closure. This distance is a crucial parameter for measuring switch triggering efficiency and directly affects the user's tactile feedback experience. By calculating the ratio of the derivative of the real-time driving force to the effective work distance, the dynamic contact stiffness can be obtained. This parameter reflects the elastic characteristics of the switch reed and the contact state, and is a fundamental indicator for predicting switch lifespan.

[0012] The impedance analysis module injects a high-frequency microampere-level constant current probe signal across the contacts of a microswitch, acquiring the micro-voltage fluctuation signal during the switching process. It then extracts the high-frequency noise energy distribution from this signal and calculates the micro-impedance noise entropy. This module precisely analyzes the micro-resistance characteristics of the contacts under contact conditions by injecting a frequency-adjustable probe signal. Wavelet packet decomposition technology is used to perform multi-band analysis on the micro-voltage fluctuation signal, extracting the energy distribution of different frequency bands and calculating the information entropy index to quantify the contact quality. The micro-impedance noise entropy is a highly sensitive indicator for evaluating the micro-state of the contact surface, enabling early detection of degradation phenomena such as oxidation, wear, and material migration.

[0013] The parameter generation module generates adaptive driving force parameters based on dynamic contact stiffness and micro-impedance noise entropy. This module compares the current dynamic contact stiffness with a preset benchmark value, and combines this with the changing trend of the micro-impedance noise entropy to comprehensively evaluate the degradation state of the switch and dynamically adjust the driving force parameters for subsequent tests. This adaptive adjustment optimizes test conditions according to the real-time state of the switch, ensuring test reliability while simulating various stress conditions faced by the switch in actual use, thus improving the practical value and predictive accuracy of the test results.

[0014] The striking mechanism module drives an electromagnetic striking mechanism to perform a clicking action on a micro switch based on adaptive driving force parameters, and collects the dynamic deformation curve of the reed in real time during the striking process. This module controls the striking rod to perform a precise pressing action through a high-precision electromagnetic actuator, while a side-mounted laser displacement sensor array scans and records the displacement changes of multiple marked points on the reed surface. Using spatial spline interpolation technology, the discrete displacement data is reconstructed into a continuous deformation curve, comprehensively capturing the dynamic deformation process of the reed during the striking process, providing detailed mechanical characteristic data for subsequent degradation analysis.

[0015] The deformation analysis module extracts the location of the maximum deflection point in the dynamic deformation curve of the reed, tracks the offset of the maximum deflection point with the number of clicks, and constructs a stiffness-deformation coupling degradation map by combining it with the effective working distance. This module analyzes the maximum deformation location of the reed during each click, monitors the migration pattern of this location with the number of uses, and constructs a degradation map by combining it with the changing trend of the effective working distance. This map visually demonstrates the evolution of the coupling relationship between the reed's elastic and deformation characteristics over its service life, and is an important basis for predicting switch life.

[0016] The degradation calculation module integrates acoustic-electric synchronization deviation, microscopic impedance noise entropy, and stiffness-deformation coupling degradation spectrum to calculate a multi-dimensional tactile degradation index. This module uses multi-feature fusion technology to integrate degradation indices from different physical dimensions into a unified evaluation system, comprehensively assessing the overall degradation degree of the switch. The multi-dimensional tactile degradation index comprehensively reflects the coordinated degradation process of the switch's acoustic, electrical, and mechanical characteristics, providing a more comprehensive and reliable criterion for lifespan prediction.

[0017] The parameter adjustment module dynamically adjusts the frequency scanning range of the high-frequency microampere-level constant current probe signal based on a multi-dimensional tactile degradation index, updating the microscopic impedance noise entropy. This module adaptively adjusts the frequency characteristics of the probe signal by comparing the tactile degradation index with a preset threshold, optimizing the monitoring effect on the contact state. When the degradation level is low, the initial frequency scanning range is maintained; when degradation worsens, the detection range is expanded towards higher frequencies; when degradation is severe, the focus is on the most sensitive frequency band, improving detection sensitivity and efficiency. This dynamic adjustment mechanism significantly enhances the system's ability to monitor changes in the microscopic state of the contact.

[0018] The lifespan prediction module outputs the predicted click lifespan of a microswitch when the multi-dimensional tactile degradation index meets the lifespan termination criteria. This module continuously monitors the rate of change of the degradation index; when the degradation rate continuously exceeds a preset threshold and the degradation index exceeds a limit threshold, the switch is deemed to have reached the end of its lifespan. By analyzing the historical trend of the degradation rate and combining it with nonlinear extrapolation technology, the theoretical lifespan of the switch is predicted and integrated with actual test results to generate a final lifespan prediction report, providing a scientific basis for product design and quality control.

[0019] The modules are connected via wired and / or wireless means to enable data transmission between them.

[0020] In this embodiment of the invention, the detailed implementation steps of time-domain alignment of the acoustic pulse sequence and the electrical conduction signal, and extraction of the time difference between the peak time of the acoustic pulse and the rising edge time of the electrical conduction, denoted as the acoustic-electric synchronization deviation, include: Bandpass filtering is applied to the acoustic pulse sequence to extract the filtered acoustic signal within the reed's resonant frequency band, and the envelope of the filtered acoustic signal is obtained. Bandpass filtering is a fundamental step in extracting the characteristic acoustic signal of the reed, improving the signal-to-noise ratio by isolating the acoustic components of the reed's resonant frequency band. A Butterworth bandpass filter is used, with the filtering frequency range typically set to 500Hz-5kHz, corresponding to the main resonant frequency range of a typical microswitch reed. The envelope of the filtered signal is extracted using Hilbert transform, making the transient impact characteristics of the reed trigger more apparent. The envelope extraction formula is: in, For the envelope, This is the filtered acoustic signal. for Hilbert transform; The envelope highlights the energy change trend of the acoustic signal, making it easier to accurately locate the critical moment when the reed deformation triggers.

[0021] The zero-crossing point of the rising edge of the electrical conduction signal is extracted and recorded as the conduction reference time. Rising edge extraction is a crucial step in locating the instant the contact closes, determining the conduction moment by detecting rapid changes in the electrical signal. The extraction process first performs low-pass filtering on the conduction signal to eliminate high-frequency noise interference; then, the first derivative of the signal is calculated to identify the time period with the largest slope; finally, linear interpolation is used to accurately locate the moment when the signal crosses a preset threshold, typically selecting the 50% point during the signal's rise from 10% to 90% as the zero-crossing point, and recording its corresponding timestamp as the conduction reference time. This method can resist the influence of signal amplitude fluctuations and provide a stable time reference.

[0022] Within a preset time window before the conduction reference time, the first peak point of the envelope is found and recorded as the acoustic trigger time. Peak detection is the core step in locating the acoustic trigger point. The initial moment of reed deformation is determined by searching for the local maximum value of the envelope. The detection process, within a 10ms window before the conduction reference time, uses a combination of sliding window method and threshold judgment to identify points that meet the peak condition. The peak condition is defined as follows: the amplitude of a local point is greater than all points before and after it (usually = 5-10), and the amplitude exceeds three times the average value of the background noise. When multiple points meet the condition, the earliest time is selected as the acoustic trigger time. This reflects the initial moment when the reed begins to deform and is an important time marker for analyzing the reed's dynamic characteristics.

[0023] The time difference between the acoustic trigger moment and the conduction reference moment is calculated as the initial acoustic-electric time difference for a single click. Time difference calculation is a direct method for quantifying acoustic-electric synchronization, assessing the internal electromechanical conversion efficiency by measuring the interval between two key moments. The calculation formula is: in, For the first The initial acoustic-electric time difference of each click For the corresponding conduction reference time, This corresponds to the acoustic triggering time. Normally, the acoustic triggering time is earlier than the conduction reference time. A positive value reflects the time required from the onset of reed deformation to the completion of contact closure, typically ranging from 1 to 5 ms. This time difference usually changes as the switch's lifespan increases, making it a sensitive indicator for monitoring switch degradation.

[0024] The initial acoustic-electric time difference (AETD) is extracted continuously from multiple click cycles. The variance of the initial AETD is calculated, and the variance is added to the mean of the most recent initial AETD to obtain the AETD synchronization deviation. Statistical processing is a key step in improving measurement stability, eliminating the influence of random fluctuations through statistical analysis of multiple measurements. The process first records the initial AETD for N consecutive clicks (usually N=10-20), then calculates the mean and variance of these data, and finally adds the variance to the mean of the most recent measurement to obtain the final AETD synchronization deviation. The formula for calculating the AETD synchronization deviation is: in, For acoustic-electric synchronization deviation, The mean of the most recent N measurements. This represents the variance of these N measurements. This calculation method not only considers the average level of the time difference but also incorporates the degree of fluctuation in the time difference, which can more comprehensively reflect the acoustic-electric synchronization performance and stability of the switch, providing a reliable time reference for subsequent calculation of the working distance.

[0025] In this embodiment of the invention, the detailed implementation steps for calculating the effective working distance from reed deformation to contact closure based on the key post displacement trajectory data and the acoustic-electric synchronization deviation, and deriving the dynamic contact stiffness based on the effective working distance, include: This process extracts the displacement-time curve of the key column from its initial press to its maximum travel from the key column's displacement trajectory data. Displacement trajectory extraction is a fundamental step in analyzing the key movement process, obtaining a complete motion curve by processing the raw displacement data. The extraction process first denoises the raw displacement data using the Savitzky-Golay filtering algorithm to eliminate high-frequency noise while preserving key features of the curve. Then, it identifies the displacement start point and the maximum travel point to determine the complete pressing process. Finally, it reconstructs the continuous displacement-time curve from the discrete sampling points using cubic spline interpolation, improving the accuracy and reliability of subsequent analysis. The displacement-time curve comprehensively records the key column's movement process, including the acceleration, deceleration, and rebound phases, providing a precise spatial reference for calculating the effective working distance.

[0026] The acoustic-electric synchronization deviation is mapped onto the time axis of the displacement-time curve, and the displacement change between the acoustic trigger moment and the conduction reference moment is extracted and recorded as the effective work distance. Displacement mapping is a key step in determining the work interval, accurately locating the effective deformation range through time-space correspondence. The mapping process first maps the acoustic trigger moment and the conduction reference moment onto the displacement-time curve, and then extracts the displacement change between these two moments as the effective work distance. The calculation formula is:

[0027] in, For the effective working distance, This is a displacement-time curve function. and These are the conduction reference time and the acoustic trigger time, respectively. The effective working distance reflects the distance the button post moves from the start of reed deformation to the complete closure of the contact. It directly relates to the user's perceived button trigger force and response characteristics, and is an important parameter for evaluating the tactile quality of the switch.

[0028] The real-time driving force during the striking process is acquired, and the ratio of the real-time driving force to the derivative of the effective working distance is calculated, denoted as the dynamic contact stiffness. Stiffness calculation is a core step in analyzing the elastic characteristics of the reed, evaluating the contact state through the force-displacement relationship. The calculation process first obtains real-time driving force data during the pressing process from the electromagnetic striking mechanism, either directly measured by a force sensor or indirectly derived through the driving current; then, the force and displacement data are aligned on the time axis to establish a force-displacement relationship; finally, the derivative of the force with respect to the displacement within the effective working range is calculated as the dynamic contact stiffness. The formula for calculating the dynamic contact stiffness is: in, Let F be the dynamic contact stiffness, F be the real-time driving force, and x be the displacement. In practical calculations, the derivative is usually approximated using the finite difference method: Dynamic contact stiffness is a key indicator for measuring the mechanical characteristics of microswitches. It directly reflects the elastic performance of the reed and the contact state of the contacts. As the service life of the switch increases, this parameter usually changes, making it an effective means of monitoring the mechanical degradation of the switch.

[0029] In this embodiment of the invention, a high-frequency microampere-level constant current probe signal is injected across the two ends of the microswitch contact to collect the micro-voltage fluctuation signal during the clicking process. The detailed implementation steps for extracting the high-frequency noise energy distribution in the micro-voltage fluctuation signal and calculating the micro-impedance noise entropy include: With the microswitch in the open state, a high-frequency microampere-level constant current probe signal with a linearly swept frequency is injected into both ends of the contacts. Probe signal injection is a fundamental step in analyzing the micro-impedance of the contacts. By controlling a current source to apply a specific signal to the contacts, the electrical characteristics of the contacts are probed. The injection process uses a high-precision programmable constant current source to generate a micro-current signal with an amplitude of 10-100 μA, with the frequency varying from 1 MHz to 50 MHz in a linearly swept manner, and a scan period of 10 ms. The use of a constant current source instead of a constant voltage source ensures a stable probe current when the contact impedance changes, improving measurement sensitivity and repeatability. The linearly swept frequency method comprehensively covers the frequency response characteristics of the contacts, providing rich frequency domain information for subsequent spectrum analysis. The microampere-level current avoids electrothermal interference to the contacts while maintaining a sufficient signal-to-noise ratio, making it suitable for long-term reliable measurements.

[0030] During the time intervals between the instant the contacts close and open, micro-voltage fluctuation signals across the contacts are simultaneously acquired. Voltage acquisition is a crucial step in recording the contact response, using a high-precision sampling system to capture minute voltage changes. The acquisition process employs a high-speed data acquisition card with a sampling rate of 100MHz or higher to ensure the capture of high-frequency components; the sampling precision is 16 bits or higher to guarantee the ability to resolve weak signals; the acquisition time window is precisely synchronized with the instantaneous changes in the contact state, typically from 1ms before closing to 5ms after closing, and from 5ms before opening to 1ms after opening. These two time windows represent the most active phases of contact state change, and the micro-voltage fluctuation signals contain rich information about the dynamic characteristics of the contacts, making them an important data source for analyzing the microscopic state of the contacts.

[0031] Wavelet packet decomposition was performed on the micro-voltage fluctuation signal to obtain multiple sub-band signals of different frequency ranges. Wavelet packet decomposition is an advanced method for multi-resolution signal analysis, capable of finely dividing the signal in the time and frequency domains and extracting local features. The decomposition process uses the Daubechies wavelet basis (db4), with a typical decomposition level of 5-6 layers, generating 32-64 sub-bands. Compared with traditional Fourier analysis, wavelet packet decomposition has better time-frequency localization capabilities, simultaneously capturing the transient characteristics and frequency domain distribution of the signal, making it particularly suitable for analyzing non-stationary signals during contact closure / opening processes. Each sub-band signal reflects the energy distribution and time-varying characteristics of the original signal within a specific frequency range, providing fine-grained frequency domain decomposition for noise entropy calculation.

[0032] The proportion of the energy of each sub-band signal to the total energy of all sub-band signals is calculated and denoted as the energy probability of that sub-band. Energy probability calculation is a core step in spectral feature extraction, quantitatively describing the frequency domain characteristics of a signal through energy distribution. The calculation process first calculates the energy of each sub-band signal:

[0033] in, For the first Energy of individual frequency bands Let be the i-th wavelet coefficient in this frequency band, and be the number of coefficients in this frequency band. Then calculate the sum of the energies of all sub-bands: Where M is the total number of sub-bands. Finally, the energy probability of each sub-band is calculated: Energy probability distribution intuitively reflects the distribution pattern of signal energy in the frequency domain. When the contacts are in normal contact, the energy is usually concentrated in a specific frequency band, while when the contacts degrade or have poor contact, the energy distribution is more dispersed. It is a sensitive indicator for detecting changes in the contact state.

[0034] The energy probability of all sub-bands is calculated using the information entropy formula to obtain the microscopic impedance noise entropy. Entropy calculation is a classic method for quantifying distribution uncertainty, evaluating the complexity of the contact state at the touch point through information theory indicators. The calculation uses the Shannon entropy formula:

[0035] Where H is the microscopic impedance noise entropy, For the first The energy probability of each sub-band, M, is the total number of sub-bands. The theoretical range of entropy is 0 to... In actual measurements, the entropy value is typically between 2 and 6. A lower entropy value indicates a more concentrated energy distribution, corresponding to a more stable contact state; a higher entropy value indicates a more dispersed energy distribution, corresponding to a less stable contact state. Microscopic impedance noise entropy can sensitively reflect microscopic changes on the contact surface, such as degradation phenomena like oxidation, wear, and material migration. It is a high-level indicator for assessing the health status of contacts and provides an important basis for generating adaptive driving force parameters.

[0036] In this embodiment of the invention, the detailed implementation steps for generating adaptive driving force parameters based on dynamic contact stiffness and microscopic impedance noise entropy include: The dynamic contact stiffness is compared with a preset stiffness benchmark value to calculate the stiffness attenuation ratio. Stiffness comparison is a fundamental step in assessing the degree of mechanical degradation, quantifying the change in spring elasticity by comparing it to a benchmark value. The comparison process first determines the stiffness benchmark value, typically using average test data from new switches or standard values ​​from product specifications; then, the ratio of the currently tested dynamic contact stiffness to the benchmark value is calculated as the stiffness attenuation index. The calculation formula is:

[0037] in, This represents the stiffness attenuation ratio. The dynamic contact stiffness is currently being measured. This is a preset stiffness reference value. Under normal circumstances, as the switch's service life increases, the dynamic contact stiffness typically decreases. The value gradually decreases from its initial value close to 1, when When the stiffness drops below a certain threshold (typically 0.6-0.7), it indicates that the mechanical characteristics of the switch have significantly degraded. The stiffness attenuation ratio is an important basis for adjusting the driving force and directly affects the force control strategy for subsequent tests.

[0038] The contact degradation coefficient is calculated by comparing the microscopic impedance noise entropy with a preset noise entropy baseline. Entropy comparison is a core step in assessing the degree of electrical degradation, quantifying changes in contact quality through changes in entropy value. The comparison process first determines the noise entropy baseline, typically using the average test value or theoretical expected value of a new switch; then, it calculates the ratio of the currently measured microscopic impedance noise entropy to the baseline, serving as a quantitative indicator of contact degradation. The calculation formula is: in, Here, H represents the contact degradation coefficient, and H represents the currently measured microscopic impedance noise entropy. To preset the noise entropy baseline, The theoretical maximum entropy value (usually taken as...) (M represents the number of sub-bands). The contact degradation coefficient ranges from 0 to 1; the closer the value is to 1, the worse the contact condition. As the number of times the switch is used increases, the entropy value usually rises, reflecting the gradual deterioration of the contact surface condition, and is an important factor to consider when adjusting the drive parameters.

[0039] The force gain adjustment factor is obtained by multiplying the stiffness attenuation ratio by the contact deterioration coefficient. Parameter fusion is a key step in integrating the mechanical and electrical degradation states, combining information from different dimensions through mathematical operations. The fusion process uses a product form, considering the influence of both factors simultaneously to ensure that significant changes in either factor are reflected in the final result. The calculation formula is as follows: in, For force gain adjustment factor, This represents the stiffness attenuation ratio. This is the contact deterioration coefficient. The characteristic of the product form is that when both factors are in a good state (…),… Approaching 1, When the force gain adjustment factor is close to 0, the driving force adjustment is relatively small; when any factor degrades significantly, the force gain adjustment factor increases, and the driving force requires a larger adjustment. This nonlinear fusion method can sensitively respond to changes in the switching state, providing precise control coefficients for adaptive driving force generation.

[0040] The peak force and pressing speed of the base click force curve are scaled using a force gain adjustment factor to generate adaptive driving force parameters. Curve scaling is the final step in generating adaptive test parameters, adjusting the force curve characteristics to suit the current switch state. The scaling process first determines the base click force curve, typically the force-time relationship under standard test conditions; then, based on the force gain adjustment factor, the key parameters of the curve are adjusted, mainly including peak force and pressing speed (i.e., the rate of force rise). The adjustment formula is: in, and These are the adjusted peak force and pressing speed, respectively. and These are the corresponding parameters of the base curve. and To adjust the weighting coefficients, This is the force gain adjustment factor. The weighting coefficients are typically determined through experimental optimization and reflect the sensitivity of different parameters to the test results. The adjusted adaptive driving force parameters can adjust the test intensity according to the real-time state of the switch, maintaining mild test conditions in the early stages of switch degradation and appropriately increasing the test intensity as degradation intensifies. This more realistically simulates the stress distribution under actual usage conditions, improving the practicality of the test and the accuracy of predictions.

[0041] In this embodiment of the invention, the detailed implementation steps of driving the electromagnetic striking mechanism to perform a clicking action on the micro switch according to the adaptive driving force parameters, and collecting the dynamic deformation curve of the reed in real time during the striking process include: The adaptive driving force parameters are converted into the driving current waveform of the electromagnetic striking mechanism. Parameter conversion is a fundamental step in executing precise clicks, mapping the force parameters into current control signals through a control algorithm. The conversion process first establishes the correspondence between electromagnetic force and driving current, typically based on the characteristic curve or theoretical model of the electromagnetic actuator; then, the required current-time curve is calculated inversely from the required force-time curve; finally, the actual driving waveform is output through a digital-to-analog conversion circuit. To improve control accuracy, the system employs a combination of pre-calibrated lookup tables and real-time feedback correction to compensate for the effects of nonlinear factors such as temperature changes and hysteresis. The driving current waveform typically includes a rapid rise segment, a plateau holding segment, and a slow fall segment, corresponding to the key striking, holding, and release processes, respectively. The timing characteristics of the waveform directly affect the tactile feel of the click and the effectiveness of the test.

[0042] The pressing action is driven by a current waveform, which activates the striking lever. A laser displacement sensor array positioned on the side of the microswitch scans multiple marked points on the reed surface. The click action is the core operation of the test, simulating real-world usage through precisely controlled mechanical movements. During execution, the electromagnetic striking mechanism generates a precise thrust based on the driving current, which is transmitted to the switch button via the striking lever, completing one full press. Simultaneously, the side-mounted laser displacement sensor array monitors the reed's deformation in real time, with a sampling frequency typically above 10kHz to ensure the capture of rapid deformation processes. The sensor array consists of 5-10 independent laser displacement sensors, evenly distributed at key locations on the reed surface. These locations are pre-marked to improve measurement repeatability. Each sensor typically has a measurement accuracy better than 1μm, accurately capturing minute deformation changes and providing high-quality raw data for subsequent deformation analysis.

[0043] Spatial spline interpolation is performed on the displacement values ​​of multiple marker points collected at the same time to obtain the spatial deflection distribution of the reed surface at that moment. Spatial interpolation is a key step in constructing a continuous deformation curve, reconstructing the complete surface shape from discrete measurement points using mathematical methods. The interpolation process employs a cubic spline interpolation algorithm, processing the multi-point displacement data at each time point to obtain a continuous function covering the entire reed surface. Cubic spline interpolation is chosen because it ensures the smoothness and continuity of the curve while accurately reflecting local details. For N measurement points, cubic spline interpolation generates N-1 segments of cubic polynomials, each segment having the following form:

[0044] in, Let be the segment of the cubic spline interpolation function, representing the interval . The above is a mathematical description of the surface deformation of the reed. For the first The location coordinates of each sampling point The independent variable represents the spatial position coordinates on the surface of the reed. The coefficient of the constant term indicates the position. Displacement value at that location, The coefficient of the linear term represents the position. The first derivative value at that point, The coefficient of the quadratic term is denoted by and is located at . The second derivative at that point is related. The coefficient of the cubic term controls the degree of curve distortion within the interval; The coefficients are determined through continuity and smoothness conditions to ensure that the function values, first derivatives, and second derivatives of adjacent segments are continuous at the connection points. This interpolation method can accurately reproduce the deflection distribution of the reed and maintain reasonable shape estimation even in regions with sparse measurement points.

[0045] The spatial deflection distribution at all times is arranged chronologically to generate the reed's dynamic deformation curve. Temporal integration is the final step in forming a complete dynamic view, constructing the deformation evolution process by organizing the time series. The integration process first sorts all spatial deflection distributions according to timestamps, forming a three-dimensional dataset of time-space-displacement. Then, through data visualization technology, this dataset is converted into an intuitive dynamic deformation curve, displaying the complete deformation history of the reed during the click process. The deformation curve typically includes key stages such as the initial static state, rapid deformation process, maximum deflection state, elastic recovery process, and final static state, comprehensively recording the reed's dynamic characteristics. This dynamic curve not only includes the spatial deformation distribution but also the rate of change in the temporal dimension, providing a rich source of information for subsequent deformation analysis and degradation assessment.

[0046] In this embodiment of the invention, the detailed implementation steps for extracting the position of the maximum deflection point in the dynamic deformation curve of the reed, tracking the offset trajectory of the maximum deflection point position with the number of clicks, and combining it with the effective work distance to construct a stiffness-deformation coupling degradation map include: The spatial coordinates of the point with the maximum deflection value are extracted from the dynamic deformation curve of the reed corresponding to each click, and recorded as the current maximum deflection point location. Maximum deflection point extraction is a fundamental step in deformation analysis, determining key deformation characteristics by finding extreme points. The extraction process first identifies the moment of maximum deformation in the complete dynamic deformation curve, typically corresponding to the instant the button is pressed to its maximum travel; then, it searches for the maximum value point in the spatial deflection distribution at that moment and records its coordinates. The search employs a mesh subdivision method, first finding an approximate location on a coarse mesh, and then precisely locating it on a fine mesh near that location, improving computational efficiency and accuracy. The maximum deflection point usually corresponds to the location of the reed with the greatest force; its spatial location and deflection value directly reflect the reed's deformation characteristics and load-bearing capacity, and are important indicators for assessing the reed's health status. As the switch's lifespan increases, the location and deflection value of the maximum deflection point may change; tracking this change helps in the early detection of signs of reed fatigue and degradation.

[0047] The spatial Euclidean distance between the current maximum deflection point and the reference maximum deflection point obtained during the initial click cycle is calculated and denoted as the single-click deflection offset. Offset calculation is a crucial step in quantitatively assessing deformation changes, reflecting the evolution of the deformation pattern through distance measurement. The calculation uses the three-dimensional Euclidean distance formula:

[0048] in, This represents the deflection offset in a single instance. The spatial coordinates of the point with the current maximum deflection. The spatial coordinates of the maximum deflection point are used as a reference. The reference point is usually chosen as the average position during the initial stage of switch testing (the first 10-20 clicks), representing the initial health state of the switch. The Euclidean distance intuitively reflects the degree of movement of the maximum deflection point in three-dimensional space. The larger the distance, the more significant the change in deformation mode, which may indicate changes in the properties of the reed material or damage to the local structure. The trend of the single deflection offset with the number of clicks is an important basis for evaluating the fatigue evolution of the reed, which usually shows a nonlinear process of slow initial growth followed by accelerated growth.

[0049] The stiffness-deformation coupling characteristic value is obtained by calculating the ratio of the single deflection offset to the effective work distance of the corresponding number of clicks. Characteristic value calculation is a core step in constructing the degradation index, reflecting the coupling relationship of multiple physical quantities through the ratio of key parameters. The calculation formula is: in, These are stiffness-deformation coupled eigenvalues. This represents the deflection offset in a single instance. This represents the effective working distance corresponding to the number of clicks. This ratio cleverly combines deformation space characteristics and mechanical behavior characteristics, providing a more comprehensive reflection of the reed's health status. When the reed is in good condition, the deformation mode is stable, the deflection offset is small, and the effective working distance is stable. The value is small and stable; as the reed begins to degrade, the deformation mode changes, the deflection offset increases, and the effective working distance may decrease, leading to... The value increases significantly, sensitively indicating changes in the reed's state. This coupling eigenvalue eliminates the limitations of a single parameter, providing a more comprehensive and sensitive degradation indicator.

[0050] A stiffness-deformation coupling degradation graph is constructed with the number of clicks on the horizontal axis and the stiffness-deformation coupling characteristic value on the vertical axis. Graph construction is the final step in visualizing the degradation process, intuitively displaying the dynamic evolution of degradation through a two-dimensional chart. The construction process first collects sufficient click test data, typically covering more than 50% of the switch's expected lifespan; then, the stiffness-deformation coupling characteristic value is calculated for each test point; finally, these data points are sorted by the number of clicks to form a continuous degradation curve. To improve the readability and smoothness of the graph, moving averages or multinomial fitting are typically used to process the original data, eliminating the influence of random fluctuations. The stiffness-deformation coupling degradation graph intuitively shows the trend of switch performance changes over its lifespan. A typical graph exhibits three stages: an initial stable period (characteristic values ​​remain at low levels), a transitional degradation period (characteristic values ​​rise slowly), and an accelerated degradation period (characteristic values ​​increase rapidly). By analyzing the inflection points and slope changes of the graph, key turning points in switch lifespan can be identified, providing important basis for degradation index calculation and lifespan prediction.

[0051] In this embodiment of the invention, the detailed implementation steps for calculating the multi-dimensional tactile degradation index by integrating acoustic-electric synchronization deviation, microscopic impedance noise entropy, and stiffness-deformation coupling degradation spectrum include: The slope of the curve corresponding to the current click count is extracted from the stiffness-deformation coupled degradation map and denoted as the fatigue evolution rate. Slope extraction is a fundamental step in quantifying the degradation rate, reflecting the dynamic characteristics of degradation through derivative analysis. The extraction process uses a local linear regression method, fitting a straight line within a data window near the current click count (typically the top and bottom 10% of data points), and calculating the slope of the line as the local degradation rate. The calculation formula is:

[0052] in, For fatigue evolution rate, These are stiffness-deformation coupled eigenvalues. For the number of clicks, This represents the current number of clicks. In practical calculations, the derivative value is usually approximated using finite difference or linear regression methods. The fatigue evolution rate directly reflects the instantaneous speed of the degradation process and is a key indicator for predicting lifespan. A lower evolution rate indicates that the switch is in a stable state, while a higher evolution rate foreshadows the arrival of an accelerated degradation phase, serving as an important signal for triggering early warnings and predictions.

[0053] The acoustic-electric synchronization deviation, microscopic impedance noise entropy, and fatigue evolution rate were normalized to obtain normalized acoustic-electric deviation, normalized noise entropy, and normalized fatigue rate. Normalization is a crucial step in achieving multi-index fusion, making different physical quantities comparable by unifying their dimensional ranges. The maximum-minimum normalization method was used to map each index to the [0,1] interval. After normalization, all indices fall within the same numerical range, eliminating the influence of dimensional and scale differences, thus creating conditions for subsequent weighted fusion and ensuring that different physical quantities have reasonable relative importance during the fusion process.

[0054] Based on the current click count and its corresponding lifecycle stage, weight coefficients for normalized acoustic-electrical deviation, normalized noise entropy, and normalized fatigue rate are dynamically assigned. Weight assignment is the core step in achieving adaptive fusion, dynamically adjusting the importance of different indicators to adapt to the degradation characteristics of different stages. The assignment process first divides the switch lifecycle into initial, stable, and degradation stages based on the morphological characteristics of the stiffness-deformation coupled degradation spectrum; then, it determines the current stage based on the current click count; finally, it assigns different weight coefficients to different stages according to a preset weight assignment strategy. A typical weight assignment strategy is as follows: Initial stage (0-10% lifespan):

[0055] Stable phase (10-70% of lifespan):

[0056] Degradation stage (70-100% lifespan):

[0057] in, , and These are the weighting coefficients for acoustic-electric deviation, noise entropy, and fatigue rate, respectively, and they satisfy... This dynamic weighting strategy reflects the relative importance of different degradation indices at different lifecycle stages: the initial stage is most sensitive to acoustic-electric deviations, the mid-stage shows significant changes in micro-impedance, and the late stage is most indicative of fatigue evolution rate. Dynamic weighting enables the fused degradation index to more accurately reflect the actual state of the switch, improving the accuracy of lifespan prediction.

[0058] The multi-dimensional tactile degradation index is obtained by multiplying each normalized index by its corresponding weight coefficient and then summing the results. Index fusion is the final step in generating the comprehensive evaluation, integrating multi-dimensional information through weighted averaging. The fusion uses a linear weighted summation method.

[0059] in, The multidimensional tactile degradation index ranges from [0,1], with values ​​closer to 1 indicating more severe degradation. This linear fusion method is simple, intuitive, and computationally efficient, while maintaining sufficient flexibility and adaptability through dynamic weight adjustment. The multidimensional tactile degradation index comprehensively considers changes in acoustic, electrical, and mechanical characteristics, fully reflecting the switch's health status and serving as a core basis for system decision-making. The index increases with the switch's lifespan, typically exhibiting a non-linear growth trend, especially accelerating towards the end of its lifespan. This closely matches the actual degradation process of the switch, providing a reliable basis for accurate lifespan prediction.

[0060] In this embodiment of the invention, the detailed implementation steps for dynamically adjusting the frequency scanning range of the high-frequency microampere-level constant current probe signal and updating the microscopic impedance noise entropy based on a multi-dimensional tactile degradation index include: The multi-dimensional tactile degradation index is compared with two thresholds: a first degradation threshold and a second degradation threshold, where the first degradation threshold is less than the second. Threshold comparison is a fundamental step in decision-making and classification, transforming continuous indicators into discrete decisions through preset standards. The comparison process first determines two key thresholds: the first degradation threshold is typically set at 0.3-0.4, corresponding to the critical point when the switch enters a moderate degradation state; the second degradation threshold is typically set at 0.7-0.8, corresponding to the critical point when the switch enters a severe degradation state. Then, the current multi-dimensional tactile degradation index is compared with these two thresholds to determine the degradation level. This three-level classification method (mild, moderate, severe) simplifies the decision-making logic while retaining sufficient accuracy, adapting to the classification needs of most switch degradation situations and providing a clear decision basis for subsequent parameter adjustments.

[0061] When the multidimensional tactile degradation index is less than the first degradation threshold, the initial frequency scan range of the high-frequency microampere-level constant current probe signal is kept unchanged. Maintaining the initial range is the default strategy in the early stages, ensuring baseline data consistency by preserving standard parameters. During this maintenance, the probe signal frequency scan range is typically set to 1MHz-50MHz, using a linear sweep mode with a scan period of 10ms. This frequency range covers the characteristic frequency bands of most microswitch contact state changes, comprehensively capturing the impedance characteristics in the initial stage. When the degradation index is low, the contact state is relatively stable, and the standardized frequency scan provides a reliable benchmark measurement, facilitating comparison with data from subsequent stages and identifying minute degradation changes. Data acquisition in the initial stage also helps establish an individual characteristic model of the switch, providing a reference basis for subsequent adaptive adjustments.

[0062] When the multi-dimensional tactile degradation index falls between the first and second degradation thresholds, the frequency scan range is shifted towards higher frequencies by a preset bandwidth, increasing the detection weight of the high-frequency band. Frequency shifting is an optimization strategy in the mid-term stage, adjusting the detection focus to capture specific degradation features. The shifting process maintains the total bandwidth unchanged, but simultaneously shifts the scan start and end frequencies towards higher frequencies by a preset bandwidth (typically 10-15MHz), making the scan range 10MHz-60MHz or 15MHz-65MHz. This shifting strategy is based on the observation that as the micro-state of the contact surface degrades, surface roughness and oxidation increase, and impedance characteristics typically change more significantly and contain richer information at high frequencies. By increasing the weight of high-frequency detection, these micro-changes can be captured more sensitively, improving the sensitivity and specificity of noise entropy calculation. This adjustment strategy is the core of adaptive measurement, optimizing test parameters according to the actual state of the switch and improving detection efficiency.

[0063] When the multidimensional tactile degradation index exceeds the second degradation threshold, the bandwidth of the frequency scan range is reduced, focusing on the frequency band with the highest current microscopic impedance noise entropy. Bandwidth focusing is a refined strategy in the later stages, improving the analysis accuracy of key frequency bands by concentrating detection resources. The focusing process first analyzes the previous scan data to identify the frequency band with the largest noise entropy contribution; then, the scan bandwidth is reduced to 1 / 3 to 1 / 2 of the original, focusing around this key frequency band; finally, the sampling density in this frequency band is increased to enhance signal quality. For example, if a significant entropy contribution is found in the 30MHz-45MHz range, the scan range is adjusted to 25MHz-50MHz, and the sampling rate is doubled. This bandwidth focusing strategy significantly improves the monitoring accuracy of specific degradation patterns, enabling the capture of more subtle changes and providing high-quality evidence for accurate assessment of the end of the device's lifespan. This refined monitoring is particularly important when switches are nearing the end of their lifespan, helping to accurately capture precursory signals of final failure.

[0064] The probe signal is re-injected using the adjusted frequency scan range, and the micro impedance noise entropy is recalculated and updated. Parameter update is the final step in completing the feedback loop, updating degradation metrics through optimized measurements. The update process follows the same procedure as the initial impedance analysis, but uses adjusted frequency parameters: first, a probe signal with updated frequency characteristics is injected into the contact; then, micro-voltage fluctuation signals across the contact are acquired; next, sub-band signals are extracted using wavelet packet decomposition; finally, the energy probability distribution and micro impedance noise entropy are calculated. The updated entropy value reflects a more accurate assessment of the contact state under optimized detection conditions, revealing subtle degradation features that standard measurements might overlook. This feedback optimization mechanism enables the system to continuously track dynamic changes in the switching state, adapt to degradation characteristics at different stages, maintain the effectiveness and sensitivity of measurements, and provide high-quality real-time data support for the lifetime prediction module.

[0065] In this embodiment of the invention, the detailed implementation steps for outputting the click life prediction result of the micro switch when the multi-dimensional tactile degradation index meets the lifespan termination determination condition include: The first derivative of the multi-dimensional haptic degradation index as a function of clicks is calculated in real time and denoted as the degradation rate. Derivative calculation is a fundamental step in monitoring degradation dynamics, assessing the acceleration of degradation through the rate of change. The calculation process employs a numerical differentiation method, calculating the rate of change based on continuous degradation index data. The calculation formula is:

[0066] in, For degradation rate, For the number of clicks The multidimensional tactile degradation index of time, The sampling interval is typically 100-500 clicks. To reduce the impact of noise, the original derivative values ​​are usually processed using a moving average to smooth short-term fluctuations. The degradation rate is a sensitive indicator of the switch's lifetime status. It typically remains at a low level in the early stages of its lifetime, but increases significantly near the end of its lifetime, exhibiting typical accelerated degradation characteristics. Continuous monitoring of the degradation rate can promptly capture the inflection point of accelerated degradation, providing a dynamic basis for determining the end of the lifetime.

[0067] When the degradation rate exceeds a preset rate threshold multiple times consecutively, and the multi-dimensional tactile degradation index exceeds the limit degradation threshold, the lifespan termination criteria are met. Termination determination is a critical decision in lifespan testing, and multiple conditions ensure its reliability. The determination process first sets a rate threshold (typically 3-5 times the initial degradation rate) and a limit degradation threshold (typically 0.85-0.95); then, it continuously monitors the degradation rate, recording the number of times it exceeds the threshold consecutively; simultaneously, it compares the current degradation index with the limit threshold; finally, when both conditions are met simultaneously (the degradation rate exceeds the threshold 3-5 times consecutively, and the degradation index exceeds the limit threshold), the lifespan termination switch is activated. This dual-condition determination mechanism avoids misjudgments that may arise from a single indicator, particularly combating the influence of temporary fluctuations and outliers in the degradation curve, thus improving the accuracy and robustness of the termination determination. Lifespan termination determination directly affects the reliability of the test results and is a critical output decision of the entire testing system.

[0068] The actual number of clicks when the end-of-life criteria are met is recorded, and the degradation rate change sequence from multiple previous click cycles is extracted. Data recording is a preparatory step for predictive analysis, supporting extrapolation calculations by preserving key historical data. The recording process first marks the current number of clicks as the observed actual lifespan point; then, it backtracks to extract degradation rate data from the previous 20%-30% of the lifespan cycles, forming a complete rate change sequence; finally, the sequence data is preprocessed, including outlier detection, missing value imputation, and data standardization, to ensure the data quality for subsequent analysis. The degradation rate sequence contains the degradation dynamics as the switch approaches the end of its lifespan, reflecting the cumulative effects of physical processes such as material fatigue and contact wear. It is an important source of information for predicting theoretical lifespan and can reveal the inherent laws and limiting states of the degradation process through mathematical models.

[0069] This algorithm uses a degradation rate variation sequence for nonlinear extrapolation to predict the theoretical number of clicks required to reach a degradation threshold. Nonlinear extrapolation is a core algorithm in lifetime prediction, extending observed trends through mathematical models to estimate theoretical limits. The extrapolation process first uses a degradation rate sequence and then fits it with an appropriate nonlinear model; commonly used models include exponential models, power-law models, and Weibull models. Then, using the fitted model parameters, it predicts the theoretical number of clicks required when the degradation exponent reaches its limit threshold. For example, if an exponential model is chosen to describe the change in degradation rate with the number of clicks:

[0070] in, For the number of clicks The rate of degradation at that time For reference, the number of clicks The rate of degradation at that time Lambda This is the exponential growth coefficient. The parameter is determined by fitting historical data. and Then, by integrating the degradation rate, the degradation exponent is obtained as a function of the number of clicks. Solving for the number of clicks required for this function to reach its limit threshold yields the theoretical lifetime prediction. Nonlinear extrapolation considers the accelerating characteristics of the degradation process, enabling a reasonable estimation of the complete lifetime curve based on limited observational data, thus providing a theoretical basis for lifetime prediction.

[0071] The actual click count and the theoretical click count are weighted and fused to generate a click lifetime prediction result, which is then output. Result fusion is the final step in balancing experience and theory, improving prediction accuracy by integrating multi-source information. The fusion process uses a weighted average method, combining actual observations and model predictions:

[0072] in, For the final click lifetime prediction result, This represents the actual number of clicks achieved in the test. This is the number of clicks predicted by theoretical extrapolation. The weighting coefficients (typically 0.6-0.8) reflect the level of confidence in actual observations. This fusion strategy considers both direct evidence from actual testing and the predictive power of theoretical models, generating a more balanced and reliable lifetime estimate. The final lifetime prediction results are output in a standard report format, including predicted click counts, 95% confidence intervals, degradation curve charts, and key degradation index analysis, providing comprehensive technical support for product design, quality control, and reliability assessment, helping manufacturers optimize product performance and lifespan.

[0073] This invention achieves accurate testing and prediction of the click life of microswitches through multi-dimensional acquisition and analysis of acoustic pulse sequences, electrical conduction signals, and key post displacement trajectory data. The multi-modal sensor fusion and adaptive testing method of this invention can comprehensively evaluate the health status of microswitches, accurately locate degradation sources and evolution paths, and effectively predict click life.

[0074] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0075] It should be noted that all formulas in this manual are calculated by removing dimensions and taking their numerical values. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0076] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A click life testing system for mouse microswitches, characterized in that, include: The data acquisition module is used to acquire the acoustic pulse sequence, electrical conduction signal, and key post displacement trajectory data of the micro switch during the click test process; The synchronization analysis module is used to perform time-domain alignment between the acoustic pulse sequence and the electrical conduction signal, and extract the time difference between the peak time of the acoustic pulse and the rising edge time of the electrical conduction, which is denoted as the acoustic-electric synchronization deviation. The work calculation module is used to calculate the effective work distance from the deformation of the reed to the closure of the contact point based on the displacement trajectory data of the key post and the acoustic-electric synchronization deviation, and to derive the dynamic contact stiffness based on the effective work distance. The impedance analysis module is used to inject a high-frequency microampere-level constant current probe signal across the contacts of the microswitch, collect the micro-voltage fluctuation signal during the clicking process, extract the high-frequency noise energy distribution in the micro-voltage fluctuation signal, and calculate the micro-impedance noise entropy. The parameter generation module is used to generate adaptive driving force parameters based on the dynamic contact stiffness and the micro impedance noise entropy. The striking execution module is used to drive the electromagnetic striking mechanism to perform a clicking action on the micro switch according to the adaptive driving force parameters, and to collect the dynamic deformation curve of the reed in real time during the striking process. The deformation analysis module is used to extract the position of the maximum deflection point in the dynamic deformation curve of the reed, track the offset trajectory of the position of the maximum deflection point with the number of clicks, and construct a stiffness-deformation coupling degradation map by combining the effective work distance. The degradation calculation module is used to fuse the acoustic-electric synchronization deviation, the microscopic impedance noise entropy, and the stiffness-deformation coupling degradation spectrum to calculate a multi-dimensional tactile degradation index. The parameter adjustment module is used to dynamically adjust the frequency scanning range of the high-frequency microampere constant current probe signal and update the micro impedance noise entropy based on the multi-dimensional tactile degradation index. The lifespan prediction module is used to output the click lifespan prediction result of the micro switch when the multi-dimensional tactile degradation index meets the lifespan termination judgment condition.

2. The system according to claim 1, characterized in that, The step of aligning the acoustic pulse sequence with the electrical conduction signal in the time domain, and extracting the time difference between the peak time of the acoustic pulse and the rising edge time of the electrical conduction, denoted as the acoustic-electric synchronization deviation, includes: The acoustic pulse sequence is bandpass filtered to extract the filtered acoustic signal within the reed resonance frequency band, and the envelope of the filtered acoustic signal is obtained. Extract the zero-crossing time of the rising edge of the electrical conduction signal and record it as the conduction reference time; Within a preset time window before the conduction reference time, find the first peak point of the envelope and record it as the acoustic trigger time; Calculate the time difference between the acoustic triggering time and the conduction reference time, and use it as the initial acoustic-electric time difference for a single click; The initial acoustic-electric time difference is extracted continuously for multiple click cycles, the variance of the initial acoustic-electric time difference is calculated, and the variance is added to the mean of the most recent initial acoustic-electric time difference, which is recorded as the acoustic-electric synchronization deviation.

3. The system according to claim 1, characterized in that, The calculation of the effective working distance from reed deformation to contact closure based on the displacement trajectory data of the key post and the acoustic-electric synchronization deviation, and the derivation of dynamic contact stiffness based on the effective working distance, includes: Extract the displacement-time curve of the key column from the start to the maximum travel from the key column displacement trajectory data; The acoustic-electric synchronization deviation is mapped onto the time axis of the displacement-time curve, and the displacement change between the acoustic triggering moment and the conduction reference moment is extracted and recorded as the effective work distance. The real-time driving force during the impact process is obtained, and the ratio of the derivative of the real-time driving force to the effective working distance is calculated and denoted as the dynamic contact stiffness.

4. The system according to claim 1, characterized in that, The process of injecting a high-frequency microampere-level constant current probe signal across the microswitch contacts, acquiring micro-voltage fluctuation signals during the clicking process, extracting the high-frequency noise energy distribution in the micro-voltage fluctuation signals, and calculating the micro-impedance noise entropy includes: When the microswitch is in the open state, a high-frequency microampere-level constant current probe signal with a frequency that varies in a linear frequency sweep manner is injected into both ends of the contact. During the time interval between the instant the contact closes and the instant it opens, the micro-voltage fluctuation signal at both ends of the contact is collected synchronously. The micro-voltage fluctuation signal is decomposed by wavelet packet to obtain multiple sub-band signals of different frequency bands; Calculate the proportion of the energy of each sub-band signal to the sum of the total energy of all sub-band signals, and denot it as the energy probability of that sub-band; The energy probability of all sub-bands is calculated using the information entropy formula to obtain the microscopic impedance noise entropy.

5. The system according to claim 1, characterized in that, The process of generating adaptive driving force parameters based on the dynamic contact stiffness and the microscopic impedance noise entropy includes: The dynamic contact stiffness is compared with a preset stiffness reference value to calculate the stiffness attenuation ratio; The contact degradation coefficient is calculated by comparing the microscopic impedance noise entropy with a preset noise entropy baseline. Multiply the stiffness attenuation ratio by the contact deterioration coefficient to obtain the force gain adjustment factor; The adaptive driving force parameters are generated by scaling the peak force and pressing speed of the basic click force curve using the force gain adjustment factor.

6. The system according to claim 1, characterized in that, The step of driving the electromagnetic striking mechanism to perform a clicking action on the micro switch according to the adaptive driving force parameters, and collecting the dynamic deformation curve of the spring during the striking process in real time, includes: The adaptive driving force parameters are converted into the driving current waveform of the electromagnetic striking mechanism; The driving current waveform drives the striking rod to perform a pressing action, and a laser displacement sensor array set on the side of the micro switch scans multiple marked points on the surface of the reed. Spatial spline interpolation is performed on the displacement values ​​of multiple marker points collected at the same time to obtain the spatial deflection distribution on the surface of the reed at that time. Arrange the spatial deflection distribution at all times in chronological order to generate the dynamic deformation curve of the reed.

7. The system according to claim 1, characterized in that, The process of extracting the position of the maximum deflection point in the dynamic deformation curve of the reed, tracking the offset trajectory of the maximum deflection point position with the number of clicks, and combining it with the effective work distance to construct a stiffness-deformation coupling degradation map includes: Extract the spatial coordinates of the point with the largest deflection value from the dynamic deformation curve of the reed corresponding to each click, and record it as the current position of the maximum deflection point; Calculate the spatial Euclidean distance between the current maximum deflection point position and the reference maximum deflection point position obtained in the initial click cycle, and record it as the single deflection offset; The stiffness-deformation coupling characteristic value is obtained by calculating the ratio between the single deflection offset and the effective work distance corresponding to the number of clicks. The stiffness-deformation coupling degradation map is constructed with the number of clicks on the horizontal axis and the stiffness-deformation coupling characteristic value on the vertical axis.

8. The system according to claim 1, characterized in that, The multi-dimensional tactile degradation index is calculated by integrating the acoustic-electric synchronization deviation, the microscopic impedance noise entropy, and the stiffness-deformation coupling degradation spectrum, including: Extract the slope of the spectrum corresponding to the current number of clicks from the stiffness-deformation coupled degradation spectrum, and record it as the fatigue evolution rate; The acoustic-electric synchronization deviation, the microscopic impedance noise entropy, and the fatigue evolution rate are normalized respectively to obtain normalized acoustic-electric deviation, normalized noise entropy, and normalized fatigue rate. Based on the current lifecycle stage of the click count, dynamically allocate weight coefficients for the normalized acoustic-electric deviation, the normalized noise entropy, and the normalized fatigue rate; The multidimensional tactile degradation index is obtained by multiplying each normalized index by its corresponding weight coefficient and summing the results.

9. The system according to claim 1, characterized in that, The step of dynamically adjusting the frequency scanning range of the high-frequency microampere-level constant current probe signal and updating the microscopic impedance noise entropy based on the multi-dimensional tactile degradation index includes: The multidimensional tactile degradation index is compared with a first degradation threshold and a second degradation threshold within a range, wherein the first degradation threshold is less than the second degradation threshold; When the multidimensional tactile degradation index is less than the first degradation threshold, the initial frequency scanning range of the high-frequency microampere constant current probe signal remains unchanged; When the multidimensional tactile degradation index is between the first degradation threshold and the second degradation threshold, the frequency scanning range is shifted to the high-frequency direction by a preset bandwidth to increase the detection weight of the high-frequency band. When the multidimensional tactile degradation index is greater than the second degradation threshold, the bandwidth of the frequency scanning range is reduced, focusing on the frequency band with the largest current microscopic impedance noise entropy; The probe signal is re-injected using the adjusted frequency scan range, and the micro impedance noise entropy is recalculated and updated.

10. The system according to claim 1, characterized in that, When the multi-dimensional tactile degradation index meets the lifespan termination determination condition, the click lifespan prediction result of the micro switch is output, including: The first derivative of the multi-dimensional tactile degradation index as a function of the number of clicks is calculated in real time and denoted as the degradation rate. When the degradation rate exceeds the preset rate threshold multiple times consecutively, and the multi-dimensional tactile degradation index exceeds the limit degradation threshold, the lifespan termination determination condition is determined to be met. Record the actual number of clicks when the lifespan termination determination condition is met, and extract the degradation rate change sequence of multiple previous click cycles; The theoretical number of clicks to reach the extreme degradation threshold is predicted by nonlinear extrapolation using the degradation rate change sequence. The actual number of clicks and the theoretical number of clicks are weighted and fused to generate the click lifetime prediction result and output it.

Images