Intelligent trial hitting rental service method and system for racket sports equipment

Abstract

This invention relates to the field of intelligent management technology for sports equipment, specifically a method and system for intelligent trial-play rental services for racket-type sports equipment. This solution receives trial-play request data sent by user terminals through a server, generates a rental session identifier, and establishes a rental record. Upon equipment return, it receives load-displacement curve data and elastic recovery trajectory data uploaded by a programmable loading module, generating dynamic stiffness characteristic operator data. Based on the initial characteristic fingerprint data throughout the entire lifecycle, it performs recursive residual analysis to form a damage time series database. When a failure threshold is reached or a fracture event occurs, it calculates the damage contribution ratio data and generates liability determination result data. This achieves quantitative recording of structural integrity and calculation of damage contribution ratios, improving the accuracy of liability determination.

Description

Technical Field

[0001] This invention relates to the field of intelligent management of sports equipment, specifically a method and system for providing intelligent trial rental services for racket-type sports equipment. Background Technology

[0002] With the development of intelligent sports equipment management systems, the application of racket-type sports equipment in trial rental, self-service sharing, and unattended management scenarios is gradually becoming more widespread.

[0003] Existing intelligent sports equipment management typically involves a complete rental and return loop consisting of servers, smart cabinets, and user terminals. It also incorporates technologies such as identity recognition, rental record storage, and anomaly detection during equipment circulation to achieve automated management under conditions of high-frequency, multi-user continuous use.

[0004] The inventors of this application have discovered the following technical problems with the aforementioned technology: In the intelligent trial rental scenario of racket-type sports equipment, the same racket will generate invisible micro-cracks, interlaminar fatigue, or cumulative micro-damage such as local stiffness reduction during high-frequency, multi-user continuous use; such micro-damage is insufficient to trigger obvious fracture or appearance abnormalities after a single use, but reaches a critical state after multiple accumulations, and may cause structural failure during a normal use, thus leading to the system determining responsibility solely based on the "last rental time node," resulting in erroneous accountability under delayed failure conditions. Existing technologies lack technical means to quantitatively record the structural performance degradation introduced by each use; existing detection methods focus on single-time anomaly identification and cannot establish a degradation data sequence of structural performance changes over time, making it impossible to calculate the damage contribution ratio based on the physical degradation process and failing to solve the problem of quantifying responsibility under delayed failure conditions. Summary of the Invention

[0005] This application provides a method and system for intelligent trial rental services of racket-type sports equipment. In traditional sports equipment rental scenarios, due to the highly concealed and delayed-cumulative nature of internal structural damage to rackets, conventional surface inspections cannot detect microscopic fractures or delaminations. To fundamentally solve the technical problem of consumer disputes caused by the difficulty in determining liability for hidden internal injuries, this application abandons the simple static logic of determining liability solely based on appearance or the final moment of fracture, as is commonly used in the field. Instead, it delves into the mechanical evolution mechanism of racket physical materials and constructs a novel intelligent liability determination system based on dynamic stiffness decay and time series tracking. To achieve the above objectives, the embodiments of this application disclose the following technical solutions:

[0006] On the one hand, this solution discloses a method for providing intelligent trial rental services for racket-type sports equipment, including the following steps:

[0007] Step S1: The server receives the trial call request data sent by the user terminal, performs identity verification and payment status verification on the trial call request data, generates a rental session identifier and stores it in the session management database. The verification and identifier generation of the trial call request data here is not only to realize the regular business flow. Its core technical motivation is to establish a unique digital mapping relationship between the user's real physical identity and the microscopic material damage that may occur in the future at the underlying logic, so as to provide a trust foundation for the final non-repudiation liability determination.

[0008] Step S2: The server sends an opening control command to the smart cabinet control unit according to the rental session identifier, and receives the equipment identification data returned by the smart cabinet control unit. The server binds the equipment identification data with the rental session identifier to generate a rental record. This binding action locks the spatiotemporal uniqueness of a specific loading test object at both the physical space and information system levels, and eliminates data crosstalk and feature pollution that are easily generated when multiple devices are transferred concurrently from a mechanism perspective.

[0009] Step S3: When the equipment is returned, the server receives the load-displacement curve data and elastic recovery trajectory data uploaded by the programmable loading module built into the smart cabinet, extracts features from the load-displacement curve data, and generates dynamic stiffness feature operator data. The programmable loading module, as the physical boundary setter for in-situ testing, avoids secondary environmental variation errors caused by the remote transfer and testing of equipment. The physical necessity of extracting dynamic stiffness feature operator data lies in overcoming the inherent defect that a single static mechanical parameter can never characterize the evolution of microcracks or interlayer delamination in polymer carbon fiber composite materials.

[0010] Step S4: The server calls the initial feature fingerprint data of the entire life cycle corresponding to the equipment, performs recursive residual analysis on the dynamic stiffness feature operator data and the initial feature fingerprint data of the entire life cycle, generates the damage increment data of this rental period and writes it into the damage time series database. The initial feature fingerprint data of the entire life cycle establishes the reference zero point of absolute no damage for a specific physical equipment. The fundamental reason for using recursive residual analysis instead of conventional numerical subtraction is that it can accurately extract the nonlinear damage increment at the micro level from the complex macroscopic mechanical hysteresis performance in the mathematical dimension, ensuring the purity of damage quantification.

[0011] Step S5: When the server detects that the equipment has reached the failure threshold or a breakage event has occurred, it reads the historical damage increment data in the damage time series database, calculates the damage contribution ratio data corresponding to each rental session, generates liability determination result data, and sends it to the user terminal. This liability determination mechanism completely subverts the simple and crude logic in the field that "breakage means the end user is fully responsible." It transforms the microscopic hidden physical damage quantification indicators extracted in the previous steps into dynamic financial liability determination basis with high legal persuasiveness across scales, generating a very strong system synergy effect.

[0012] Furthermore, to eliminate the technical pain point that conventional single instantaneous impact tests easily miss the blind zone of viscoelastic hysteresis deformation inside polymer composites, this application does not employ a simple single extreme value test. Instead, it uses multi-stage quasi-static loading control commands to gradually excite and comprehensively capture the deep internal mechanical response of the material. Simultaneously, this application profoundly reveals the inherent physical law that hidden internal damage exhibits more significant energy dissipation during the unloading phase. It creatively shifts the focus to a later stage, transforming the recovery path during the unloading phase into highly quantifiable elastic recovery trajectory data, thus laying a solid foundation of underlying physical verification logic for accurately assessing the integrity of carbon structures.

[0013] Furthermore, addressing the issue of extremely low signal-to-noise ratio in extracting microstructural damage features from racket-type equipment, this solution breaks away from the traditional characterization method that relies solely on ultimate tensile or bending strength. Its core defense logic lies in performing cross-dimensional feature fusion processing: combining hysteresis area characteristic data, representing microscopic friction and crack propagation loss, with stiffness change characteristic data, representing macroscopic resistance to deformation. This fusion mechanism, based on the unique mechanical loop characteristics of composite materials, can amplify the signal weight of early-stage micro-injuries, exhibiting physical sensitivity and detection robustness unmatched by conventional mathematical statistical methods in this field.

[0014] Furthermore, due to the unavoidable deviations in resin content or manufacturing tolerances in carbon cloth lamination during industrial mass production, the initial mechanical properties of different individual devices naturally exhibit dispersion. To avoid fatal misjudgments in subsequent damage residual assessments caused by these inherent differences, this application establishes a mandatory individualized mechanical gene archiving mechanism. By extracting and solidifying the initial characteristic fingerprint data throughout the entire lifecycle in a non-destructive state, this solution fundamentally filters out the background noise interference caused by fluctuations in factory yield, ensuring that the mathematical comparison for each subsequent lease liability determination has an absolutely pure and exclusive judgment origin.

[0015] Furthermore, during the long-term service life of sports equipment rentals, mechanical degradation is a complex composite of normal fatigue aging and abnormal external force damage. If the direct interpolation method commonly used in conventional business systems is employed, it is easy to unreasonably shift the natural depreciation costs that the service provider should bear to the user. Therefore, this application creatively introduces time-recursive filtering to decouple and separate the residuals in the time-series frequency domain. This logic is not a mere imitation of conventional algorithms, but rather precisely delineates the technical boundary between normal wear and tear and human-caused damage, accurately identifying nonlinear abrupt changes in data, and endowing this system with extremely strong legal resistance capabilities.

[0016] Furthermore, this solution's determination of equipment retirement milestones does not rely on manual experience or a rough calculation of usage counts, but is deeply anchored in big data mining of historical macroscopic fracture event physical precursors. By dynamically comparing real-time extracted cumulative damage data with failure thresholds possessing predictive warning attributes, this system achieves a significant paradigm shift from passive post-event liability determination to proactive pre-event mechanical lifespan interception. This correlation mechanism, which generates failure trigger identification data, can forcibly prevent the continued circulation of high-risk equipment at the system level before physical fracture triggers potential sports-related personal injury.

[0017] Furthermore, in the face of cumulative fracture events ultimately caused by the covert damage inflicted by multiple historical users at different nodes, this application completely abandons the industry's unfair terms of end-point full liability in its algorithm mechanism. By deeply mining the incremental damage data at each stage hidden in the time series database and using high-dimensional normalization processing mapping, this solution scientifically reconstructs the extremely vague and complex physical evolution history into damage contribution ratio data with clear responsibility and precise proportions, achieving a perfect cross-border reconstruction of complex physical and mechanical events into a basis for fair determination of responsibility in digital finance.

[0018] Furthermore, to ensure the operational efficiency and rigidity of the intelligent leasing business loop, this application, after establishing the liability allocation data, directly establishes a strong correlation channel for automated processing with the underlying payment and settlement network. The significance of this step goes far beyond simple financial arithmetic; it transforms the highly abstract microscopic mechanical loss physical quantity at the front end into directly realized compensation amount data on the user terminal through deterministic transformation logic. This cross-domain linkage effect, seamlessly connecting physical damage detection to financial payment and settlement, significantly reduces the cost of manual intervention in after-sales service due to disputes.

[0019] Furthermore, considering that ordinary users are naturally resistant to technical challenges and have concerns about black boxes when assigning liability for invisible microscopic damage, this application makes the underlying argumentation logic of the system explicit. By forcibly packaging and storing core technical features such as dynamic stiffness characteristic operator data and residual sequence data in an independent cloud environment, the system provides a traceable, timestamped, and tamper-proof end-to-end visualized chain of evidence for each liability assignment. This data packaging action completes the user trust loop in the business logic, greatly improving the commercial feasibility of this core underlying patent technology in the civilian rental market.

[0020] On the other hand, this solution discloses an intelligent trial rental service system for racket-type sports equipment, including: a server, an intelligent cabinet control unit, a programmable loading module, and a user terminal; the intelligent cabinet control unit is used to receive the door opening control command sent by the server and return equipment identification data; the programmable loading module is used to collect load-displacement curve data and elastic recovery trajectory data and upload them to the server; the user terminal is used to send trial request data and receive liability determination result data.

[0021] The system hardware architecture constructed in this application is by no means a simple physical stacking of conventional IoT terminals and back-end servers in this field. Its core non-obviousness lies in the construction of a highly specialized cyber-physical system, in which the programmable loading module acts as an extremely sensitive physical probe, directly penetrating the microscopic mechanical anomalies at the material's underlying layer; while the server relies on a complex computing architecture to perform residual decoupling and liability determination analysis. All independent physical components collaborate deeply under a unified digital leasing timeline, jointly transforming the previously hidden damage evolution, which was in a mechanical black box state, into a highly transparent, fully quantified, fully automated digital closed-loop control system.

[0022] This invention provides a smart trial rental service method and system for racket-type sports equipment. By acquiring load-displacement curve data and elastic recovery trajectory data of the equipment at the end of each rental cycle, dynamic stiffness characteristic operator data is constructed. This data is then combined with recursive residual analysis using the initial characteristic fingerprint data generated when the equipment is first stored, enabling the quantitative recording and continuous accumulation of structural performance degradation introduced in each rental cycle. Compared to methods that determine responsibility solely based on the time of fracture, this invention constructs a cumulative damage evolution path based on a damage time-series database. When a failure threshold is reached, the damage contribution ratio corresponding to each rental session can be decomposed in reverse, thereby achieving responsibility allocation based on the physical degradation process. This solution is applicable to racket-type equipment with different structural types. Even when the damage formation time and failure time are inconsistent, it can still achieve dynamic tracking and objective judgment of the structural state, improving the accuracy of failure identification and the technical rationality of responsibility allocation, while enhancing the system's data closed-loop management capabilities and the traceability of test results. Attached Figure Description

[0023] Figure 1 This is a system structure diagram of an embodiment of the present invention;

[0024] Figure 2 This is a general flowchart of an embodiment of the present invention;

[0025] Figure 3 This is a flowchart illustrating the detection and generation process of an embodiment of the present invention;

[0026] Figure 4 This is a diagram showing the generation of the dynamic stiffness feature operator in an embodiment of the present invention.

[0027] Figure 5 This is a flowchart of the recursive residual analysis process according to an embodiment of the present invention;

[0028] Figure 6 This is a schematic diagram of the load-displacement curve and hysteresis area according to an embodiment of the present invention;

[0029] Figure 7 This is a schematic diagram of the structure of the smart cabinet according to an embodiment of the present invention;

[0030] Among them, 1. racket socket; 2. speaker; 3. display. Detailed Implementation

[0031] 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 a part of the embodiments of the present invention, and not all of them. 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. In the following description, numerous specific details are set forth to provide a comprehensive understanding of the present invention. The present invention may be practiced without some or all of these specific details. In other instances, well-known processes have not been described in detail to avoid unnecessarily obscuring the present invention.

[0032] When used in conjunction with the terms "comprising," "method comprising," or similar language in this specification and appended claims, the singular forms "a," "some," and "the" include plural references unless the context clearly indicates otherwise. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0033] Example

[0034] In this embodiment, the "programmable loading module" is defined as an electromechanical integrated component installed inside the intelligent cabinet for applying controllable mechanical loading to equipment and collecting response data. It includes a loading actuator, a torque sensor, a displacement sensor, and an embedded control circuit. The loading actuator is preferably an electric lead screw or stepper motor drive mechanism for outputting controllable loading displacement; the torque sensor is used to collect real-time torque data during the loading process; the displacement sensor is used to collect displacement data during the loading process; and the embedded control circuit is used to execute multi-stage quasi-static loading control commands and upload the collected data to the server.

[0035] "Load-displacement curve data" is defined as a set of corresponding data formed by real-time torque data and displacement data after time synchronization processing, which is used to reflect the mechanical response behavior of the equipment during loading and unloading.

[0036] "Elastic recovery trajectory data" is defined as a set of data formed by displacement recovery paths during the unloading phase, used to characterize the elastic recovery and hysteretic energy dissipation characteristics of the structure.

[0037] "Dynamic stiffness feature operator data" is defined as a multi-dimensional feature vector formed by feature extraction and feature fusion processing of load-displacement curve data and elastic recovery trajectory data, used to characterize the structural integrity state.

[0038] "Initial Feature Fingerprint Data of the Entire Life Cycle" is defined as the benchmark dynamic stiffness feature operator data obtained based on the standard loading process when the equipment is first put into storage, which is used as a reference benchmark for the subsequent performance evolution of the equipment.

[0039] "Recursive residual analysis processing" is defined as a process of performing cycle-by-cycle difference calculations on dynamic stiffness feature operator data and initial feature fingerprint data throughout the entire life cycle, and then combining this with a time recursive filtering algorithm to perform trend separation on the residual sequence data.

[0040] The core concept of this solution lies in breaking the long-standing static and subjective judgment logic in the sports equipment rental industry, which relies on "relying on appearance for damage assessment" or "end users bearing full responsibility for breakage." Traditional technologies, limited by physical space, cannot track microscopic delamination of the resin matrix or latent fractures of carbon fiber inside the equipment. This application creatively introduces in-situ mechanical testing hardware at the smart locker node, transforming complex macroscopic mechanical hysteresis phenomena into digital, nonlinear microscopic damage increments. By precisely decoupling and separating natural fatigue aging and abnormal violent damage at the algorithmic level, a fully automated trust chain is constructed, leading directly from microscopic physical detection to digital financial compensation, thereby providing an irrefutable basis for non-obviousness.

[0041] Example 1

[0042] This embodiment discloses a method for providing an intelligent trial rental service for racket-type sports equipment, including the following steps:

[0043] The server receives trial call request data sent by the user terminal, performs identity verification and payment status verification on the trial call request data, generates a rental session identifier, and stores it in the session management database.

[0044] In this embodiment, when a user wishes to rent a tennis racket, they initiate a request by scanning the device's QR code using a smartphone application. Upon receiving the trial request data, which includes the user's account, equipment type, and timestamp, the server calls the credit authorization interface in the background to verify the user's identity and freezes the corresponding deposit pre-authorization to complete the payment status verification. After successful verification, the server uses a hash algorithm to generate a globally unique UUID as the rental session identifier and writes it as the primary key into the session management database deployed in the cloud.

[0045] The technical motivation for implementing this step is to establish a strong correlation between the physical operating entity and subsequent detection data at the bottom layer of the information system. By generating and storing a globally unique rental session identifier, it is possible to ensure that no data crosstalk occurs at any human-computer interaction node in complex concurrent rental scenarios, laying a preliminary data foundation for ultimately achieving irrefutable accountability.

[0046] The server sends an opening control command to the smart cabinet control unit based on the rental session identifier, and receives the equipment identification data returned by the smart cabinet control unit. The server then binds the equipment identification data with the rental session identifier to generate a rental record.

[0047] After obtaining legal identification, the server sends an authentication-encrypted door-opening control command to the designated rental cabinet via IoT protocols such as MQTT. Upon receiving the command, the smart cabinet control unit (such as a built-in PLC controller) activates the electromagnetic lock mechanism to unlock the door. When the user removes the racket, the RFID reader inside the cabinet instantly captures the passive tag embedded in the racket handle and reports the unique equipment identification data. The server then establishes a mapping between this equipment identification data and the aforementioned rental session identifier in its database, thereby officially generating and activating a rental record.

[0048] This step, through the interplay of hardware sensing and software logic, locks in the spatiotemporal uniqueness of a specific test object at both the physical and informational levels. This instant binding mechanism fundamentally eliminates the business logic loophole of "user A lending out racket B but being recorded as racket C," ensuring that all subsequent physical test data has an absolutely pure data source.

[0049] Upon equipment return, the server receives load-displacement curve data and elastic recovery trajectory data uploaded by the programmable loading module built into the smart cabinet. It then extracts features from the load-displacement curve data to generate dynamic stiffness feature operator data. The acquisition of the load-displacement curve data includes: the programmable loading module applying multi-order quasi-static loading control commands to the equipment; the programmable loading module's built-in torque and displacement sensors collecting real-time torque and displacement data during the loading process, respectively; and the server performing time synchronization processing on the real-time torque and displacement data to generate standardized load-displacement curve data, and then processing the unloading... The recovery path of each stage is segmented and fitted to obtain elastic recovery trajectory data, which serves as input data for generating the dynamic stiffness feature operator data. The generation of the dynamic stiffness feature operator data includes: the server performing slope segmentation calculation on the load-displacement curve data to extract stiffness change feature data for the loading and unloading stages; performing energy dissipation analysis on the elastic recovery trajectory data to extract hysteresis area feature data; and performing feature fusion processing on the stiffness change feature data and the hysteresis area feature data to generate the dynamic stiffness feature operator data used to characterize structural integrity.

[0050] The moment the user inserts the racket into the smart cabinet's slot after a trial run, the programmable loading module is triggered.

[0051] To capture any transient abnormal impacts that may occur during a user's test play, this embodiment incorporates a three-axis MEMS accelerometer module inside the racket socket. The accelerometer is electrically connected to the intelligent cabinet control unit and is used to collect residual vibration attenuation curve data of the racket frame at the moment the equipment is returned.

[0052] The server performs spectral analysis on the vibration attenuation curve data, extracting the peak impact amplitude, dominant frequency distribution, and damping attenuation coefficient, and compares it with a standard safe impact spectral model. When the peak impact amplitude exceeds a preset safe impact threshold, abnormal impact event identification data is generated and written into the damage time series database.

[0053] The technical purpose of this impact detection module is to overcome the insensitivity of quasi-static loading detection to high-frequency transient damage. By capturing the high-frequency vibration characteristics generated by violent ground impacts or impacts from hard objects, abnormal events can be recorded in advance before micro-cracks have fully evolved into macro-stiffness decay, providing time-stamped evidence for subsequent responsibility analysis.

[0054] The stepper motor within this module applies multi-stage quasi-static loading control commands to the racket frame via push rods. During this process, a high-frequency sampling torque sensor and a photoelectric displacement sensor simultaneously acquire real-time torque and displacement data. The server receives the underlying hardware data and performs spatiotemporal alignment calibration.

[0055] Considering the difference in hardware interrupt frequencies and physical link transmission delays between the torque sensor and displacement sensor in the programmable loading module, the original sequences obtained by the two sensors exhibit asymmetric misalignment in timestamps, making it impossible to directly construct a closed power loop. Therefore, this embodiment introduces a spatiotemporal mapping matrix based on sampling interval compensation. By reconstructing the asynchronous sampling points using third-order splines, the time synchronization alignment equation is obtained:

[0056] ;

[0057] In this embodiment, the formula aims to eliminate time-delay distortion during industrial sensor acquisition. Due to physical differences between torque and displacement sensors in terms of underlying hardware response frequency, signal conditioning circuit group delay, and data bus transmission path length, the original data sequences acquired synchronously by the system exhibit asynchronous phase differences on the time axis. If the load-displacement function relationship is directly constructed without phase compensation, the synthesized curve will suffer from non-physical distortion, thereby compromising the reliability of the subsequent damage operator extraction.

[0058] Among them, the main formula A temporal reconstruction model for displacement data based on third-order spline interpolation was constructed. Among them, This represents the displacement eigenvalue mapped at the reference timestamp t after time alignment calibration. Polynomial coefficients. The system server obtains the data by performing spline constraint solving or least squares fitting on local neighborhood sampling points based on the original sampling sequence of the displacement sensor. In this mechanical testing scenario, the physical significance of selecting the third-order polynomial structure is to ensure that the zeroth (displacement), first (velocity), and second (acceleration) derivatives of the displacement function in the time dimension all have mathematical continuity, thereby satisfying the smoothness constraint of kinematic characteristics during dynamic mechanical loading.

[0059] The variable t is defined as the global synchronization reference benchmark of the server system, used to establish a unified spatiotemporal coordinate system for heterogeneous sensor data. Core parameter Defined as a system time delay compensation factor, its physical value integrates the cumulative contributions of hardware bus delay, signal amplifier response delay, and analog-to-digital conversion (ADC) cycle time. In practical engineering environments, The system exhibits nonlinear fluctuation characteristics due to ambient temperature drift and system computational load. The server calculates the current time in real-time by sending synchronization trigger pulses or system detection packets during the initialization phase of the return test cycle. Instantaneous values ​​are used to achieve microsecond-level signal phase locking.

[0060] Constraints in the formula An adaptive driving relationship between the master formula and the compensation factor was established. This integral expression characterizes the torque signal. The process of calculating the cross-correlation function with the displacement signal L is used to evaluate the time offset of the two signal sequences. Waveform consistency under [condition]. By executing [action / method]... The optimization process involves the server searching for and extracting the offset that maximizes the value of the cross-correlation function. and assign it to As input variables to the main formula, this coupling mechanism enables adaptive calibration in heterogeneous hardware environments, ensuring that the system can eliminate interference from environmental and hardware differences and forcibly align the sampling points of heterogeneous sensors at the physical level.

[0061] Subsequently, the server extracts stiffness change feature data of the loading and unloading stages through slope segmentation calculation, and performs integral calculation on the segmented fitted elastic recovery trajectory data of the unloading stage.

[0062] To quantify the nonlinear energy dissipation caused by delamination of the resin matrix inside the racket material, this application does not simply compare peak loads, but instead constructs a hysteresis energy flow mapping function to characterize microscopic losses by calculating the envelope area of ​​the loading path and the elastic recovery path within the displacement interval:

[0063] ;

[0064] In this embodiment, the hysteresis area integral calculation formula is used to quantify the microscopic mechanical losses within composite material devices. From a materials mechanics perspective, in a quasi-static loading cycle, if the matrix structure of a polymer composite material is intact, its loading and unloading paths theoretically remain highly consistent; however, if interlaminar delamination or microfiber fracture exists internally, the unloading process will generate irreversible internal friction energy dissipation. This physical quantity manifests as a closed hysteresis loop on the load-displacement curve, as shown on the left side of the formula. This refers to the hysteresis energy dissipation measure, which characterizes the degree of energy dissipation and serves as a core feature parameter for assessing the structural integrity of equipment and measuring damage increments.

[0065] The core integral term of the formula The numerical solution process for the power consumption envelope area within a single test cycle is described. The endpoints of the integration interval are also described. and These are defined as the initial zero point of displacement and the extreme point of maximum deformation in the trajectory of the programmable loading module, respectively, and are locked in real time by the photoelectric displacement sensor after clock synchronization. Within this range, The load-displacement mapping function during the loading phase is generated by the server performing polynomial smooth fitting on the collected discrete torque data lattice. This is the elastic recovery trajectory function during the unloading phase. (The function is defined by the displacement axis.) By continuously integrating the instantaneous difference between the loading and unloading functions, the system obtains the absolute value of the mechanical energy consumed by the equipment in one loading cycle. This area value is positively correlated with the degree of latent damage inside the material and is a direct criterion for quantifying microscopic physical damage.

[0066] Considering the significant viscoelastic characteristics of carbon fiber composites, whose mechanical response is greatly affected by ambient temperature, this formula introduces a correction parameter outside the integration domain. This refers to the ambient temperature compensation coefficient. Because the storage modulus and loss modulus of the resin matrix exhibit nonlinear changes with temperature fluctuations, a single mechanical integral value lacks comparability under different operating conditions. To eliminate measurement errors caused by environmental thermal effects, the temperature sensor built into the intelligent cabinet acquires real-time ambient temperature data at the instant of testing. The server dynamically matches the corresponding temperature characteristic according to a preset material modulus-temperature feature mapping table. The original integral area is then normalized. This step ensures that the calculation results reflect the actual physical damage to the equipment, rather than data drift caused by fluctuations in ambient temperature.

[0067] From the perspective of the logical connection of the algorithm architecture, the integral operation in the first half of the formula and the compensation coefficient in the second half constitute a cascaded processing flow of feature extraction and environmental denoising. The definite integral structure is responsible for solving the electrical signal output by the underlying sensor into a mechanical energy dissipation component with clear physical meaning; the compensation coefficient, as a calibration operator, maps this physical quantity to the feature space under the standard reference environment. The two are coupled through product logic, ensuring that the output dynamic stiffness feature operator data has high working condition robustness, thus providing deterministic data input for subsequent high-precision recursive residual analysis and effectively avoiding the risk of false alarms in damage identification caused by environmental interference.

[0068] Finally, the server performs feature fusion processing on the stiffness slope and hysteresis area, and encapsulates it to generate high-dimensional dynamic stiffness feature operator data.

[0069] Conventional, single-instantaneous impact tests are prone to missing the viscoelastic hysteresis deformation blind zone within polymer composites such as carbon fibers. This embodiment abandons conventional testing methods, introducing multi-stage loading and focusing on energy dissipation during the unloading phase (extracting hysteresis area characteristic data) in the algorithm. This is because once a microcrack is formed, its stiffness change is not significant during loading, but the internal friction during unloading significantly increases the hysteresis area. This in-depth application of the physical mechanism greatly amplifies the signal-to-noise ratio of early-stage micro-injuries, breaking through the industry bottleneck that non-destructive testing must rely on large X-ray equipment.

[0070] The server pre-acquires the benchmark dynamic stiffness characteristic data generated when the equipment is first received into the warehouse, and stores the benchmark dynamic stiffness characteristic data as initial characteristic fingerprint data for the entire life cycle. The server calls the initial characteristic fingerprint data for the entire life cycle corresponding to the equipment, performs recursive residual analysis processing on the dynamic stiffness characteristic operator data and the initial characteristic fingerprint data for the entire life cycle, generates damage increment data for this rental period, and writes it into the damage time series database. The initial characteristic fingerprint data for the entire life cycle includes benchmark load-displacement curve data and benchmark elastic recovery trajectory data obtained by executing the standard loading process through the programmable loading module when the equipment is first received into the warehouse. The server performs feature extraction processing on the benchmark load-displacement curve data and benchmark elastic recovery trajectory data to generate benchmark dynamic stiffness characteristic operator data, and stores the benchmark dynamic stiffness characteristic operator data as initial characteristic fingerprint data for the entire life cycle. The initial characteristic fingerprint data of the life cycle; the recursive residual analysis processing includes: the server performing cycle-by-cycle difference calculation on the dynamic stiffness characteristic operator data and the initial characteristic fingerprint data of the entire life cycle to generate residual sequence data; performing time recursive filtering on the residual sequence data to separate natural fatigue trend data and nonlinear mutation data; using the nonlinear mutation data as the damage increment data of the current lease cycle and writing it into the damage time series database; the failure threshold is obtained by the server through statistical modeling based on the damage time series data of historical sample equipment before fracture; after each time the server writes the damage increment data of the current lease cycle, it updates the cumulative damage field in the damage time series database to obtain the current cumulative damage data and compares it with the failure threshold. When the current cumulative damage data reaches or exceeds the failure threshold, failure trigger identification data is generated.

[0071] After receiving the current operator data, the server first retrieves the baseline dynamic stiffness characteristic operator data generated during the standard loading process when the equipment was first put into storage, i.e., the initial characteristic fingerprint data throughout its entire life cycle. Subsequently, it performs cycle-by-cycle difference calculations to generate residual sequence data. Crucially, the server calls a time-frequency domain analysis model to perform time-recursive filtering on this residual sequence data.

[0072] During long-term use, equipment naturally ages by exhibiting a slow, low-frequency linear decrease in stiffness, while rough handling results in a high-frequency, sudden shift in nonlinear operator behavior. This embodiment constructs a state transition space and recursively approximates the operator observations of the current period with the predicted values ​​of the previous period to obtain the residual state equation for separating nonlinear abrupt changes in data.

[0073] ;

[0074] In the liability determination calculation logic constructed in this embodiment, the core technical purpose of the time recursive filtering algorithm formula set is to achieve decoupling processing of the mechanical attenuation signal of composite materials in the frequency domain, so as to quantify and separate the natural fatigue evolution component and the single nonlinear physical failure component of the equipment. The pre-equation of this formula set... In essence, it is a first-order exponential smoothing state transition model used to calculate and update the cumulative natural fatigue trend estimate of the equipment. In this equation, the system uses the k-th rental period as a baseline and retrieves the historical fatigue baseline state of the previous period stored in the database. This is then summed with the current period's physical deviation using a weighted average. The physical deviation term on the right-hand side of the equation... This is derived from the dynamic stiffness feature operator data extracted in real time. Subtract the initial lifecycle feature fingerprint data established when the equipment first enters the warehouse. The absolute value of the global macroscopic stiffness decay since the initial state was calculated. To suppress the numerical disturbance of the global baseline by a single anomalous loading, a fatigue evolution decay factor was introduced into the equation. This parameter is a low-sensitivity weight pre-fixed in the system's underlying database based on the standard SN fatigue life curve of carbon fiber composite materials. By applying this weight, the system can construct a low-frequency tracking mechanism with mathematical inertia, ensuring that the fatigue trend estimate strictly follows the slow and smooth aging law in materials science, thereby shielding the interference of high-frequency abrupt noise at the algorithm level.

[0075] Corresponding to the aforementioned state transition equations for long-period, slowly varying trends, the post-residual equations of the formula set... This constitutes a mechanism for extracting and quantifying nonlinear high-frequency damage data. In this equation, the server directly uses the feature operators of two adjacent rental periods. and Differential calculations are performed to determine the extreme values ​​of relative stiffness drops within a single test interval. Given that carbon fiber composites lack physical self-healing properties after matrix cracking or fiber breakage, any negative stiffness increment observed in the measurement can be attributed to background drift of the hardware sensor or systematic errors caused by environmental thermal expansion. Therefore, this scheme introduces a nonlinear activation function. By setting strict forward conduction mapping rules, all non-positive residual values ​​are forcibly truncated and zeroed. Furthermore, to compensate for the reasonable minor mechanical losses inevitably caused by legitimate and compliant test firing operations, the equation further applies a deduction operation to the difference term. The deducted natural drift tolerance is adjusted by a preset tolerance coefficient. Compared with the historical baseline noise mean of the underlying sensor The product is then used to obtain the data. After the above differential, error reduction, and nonlinear activation processing, the system finally outputs the incremental damage data for this lease period. This represents a quantitative indicator of pure and absolute abnormal physical damage.

[0076] Engineering analysis of the underlying coupling logic of the two sets of equations in the system algorithm architecture reveals that they constitute a closed-loop, mutually exclusive, and decoupled relationship of baseline dynamic reconstruction and high-frequency residual truncation. The preceding state transition equation operates smoothly throughout the entire equipment service life, its core function being to define and isolate the normal material depreciation and decay baseline borne by the operator. The subsequent residual equation, based on this baseline evolution law, specifically performs discretized abnormal increment extraction for single leasing events. Although both share the feature operator data stream output by the same physical hardware acquisition system, they implement strict reverse isolation in frequency domain processing and data attribution mapping. The former solidifies long-term evolution laws through low-pass filtering characteristics, while the latter uses nonlinear rectification characteristics to filter high-frequency fracture features. This deterministic isolation operation of composite material failure signals based on mathematical modeling effectively establishes the empirical boundary of a single abnormal load impact under complex service conditions, providing underlying data support with engineering rigor and logical uniqueness for subsequent multi-user responsibility allocation procedures.

[0077] The extracted nonlinear mutation data is ultimately stored as incremental damage data for this lease period. Simultaneously, the server dynamically assesses the current cumulative damage data.

[0078] Considering the randomness and cumulative effect of composite material failure, this application abandons the fixed threshold judgment mode. Instead, it combines the damage gradient change rate of historical fracture samples, constructs a dynamic safety boundary equation based on the Weibull distribution mapping, and constructs a failure judgment inequality based on the mapping relationship between the current cumulative damage data and the failure probability:

[0079] ;

[0080] In the equipment safety service assessment system constructed in this embodiment, the cumulative damage and dynamic failure threshold determination formula is used to establish a nonlinear mapping relationship between the microscopic damage increment of composite materials and the probability of macroscopic structural failure. Its core technical purpose is to predict the fracture risk of equipment through a probabilistic statistical model, thereby achieving preventive maintenance and safety risk prevention. This formula uses a two-parameter Weibull distribution as the underlying reliability modeling framework, which is consistent with the random failure characteristics of composite materials under fatigue loads. Defined as the predicted instantaneous failure probability of equipment under the current cumulative damage state, this value is a continuous variable in the range [0,1], which directly represents the statistical risk of catastrophic fracture of the equipment structure.

[0081] In the engineering acquisition of variables and the analysis of data structure dimensions, the numerator... This represents the sum of all nonlinear damage increments accumulated from the initial receipt of the equipment to the end of the current k-th rental period. The server obtains this value by retrieving all historical residual components associated with the unique identifier of the equipment from the damage time-series database and performing a time-domain summation operation, thus reflecting the entire lifecycle of the equipment's internal microstructure evolution. The corresponding denominator term... The initial characteristic fingerprint data of the equipment is used to normalize the cumulative damage, converting it into a dimensionless relative loss ratio. To further accurately characterize the fatigue failure mechanism unique to composite materials, shape parameters are introduced into the formula. With scale parameters These two parameters are not fixed preset values, but are obtained by the system through maximum likelihood estimation or least squares regression fitting of the damage trajectories of fractured samples in the historical database; among them, This determines the rate at which the failure probability increases with damage, reflecting the concentration of brittle fracture in the carbon fiber resin matrix. It is then used as a scale factor to correct the characteristic lifetime boundary of the material.

[0082] The right-hand side of the determination equation The logical criteria boundaries for the system to execute hardware interlocking control were established. Parameters Defined as a preset failure trigger warning threshold, this threshold is a static parameter embedded in the server configuration module based on industry safety standards and ultimate tensile / shear strength test data provided by equipment manufacturers, combined with a certain safety margin coefficient. The calculated instantaneous failure probability... When this boundary is exceeded, the inequality holds, and the system immediately generates failure trigger flag data. This logical judgment process directly drives the underlying intelligent cabinet control unit to perform a physical door locking operation, thereby achieving closed-loop control from physical measurement and probabilistic modeling to safety prevention at the algorithm level. This dynamic threshold judgment mechanism, based on reliability engineering theory, can more accurately capture the failure characteristics of composite materials with a sudden increase in probability at the end of their service life compared to the traditional fixed stiffness threshold discrimination method, providing rigorous algorithmic support for ensuring personal safety during the leasing process.

[0083] If the preset failure threshold is exceeded, failure trigger identification data is generated, and the control unit locks the hatch, preventing the next user from renting it.

[0084] To enhance the objective confirmation of macroscopic structural damage to the equipment, in this embodiment, the intelligent cabinet is further equipped with an industrial-grade image acquisition unit and an image analysis and processing unit. The image acquisition unit includes at least one high-resolution CMOS camera, located above the racket socket, with the lens optical axis facing the racket frame and handle connection area; a ring-shaped supplementary light source is provided around the image acquisition unit to provide stable and uniform lighting conditions in the enclosed environment of the cabinet.

[0085] While the equipment is returned and the programmable loading module detection process is triggered, the server sends an image acquisition trigger command to the image acquisition unit. The camera acquires multi-angle static image data of the current equipment, and the raw image stream is preprocessed by the embedded edge computing module in the cabinet, including distortion correction, grayscale equalization and noise filtering, to generate standardized structural detection image data.

[0086] The server processes the structural inspection image data using a crack recognition algorithm. The crack recognition algorithm is based on a defect classification model constructed using a convolutional neural network, or a traditional image recognition algorithm based on edge enhancement and connected component analysis. It performs pixel-level scanning of the racket frame area, extracts crack length features, fracture section continuity features, and fiber exposure features, and generates structural integrity judgment label data.

[0087] When the structural integrity assessment tag data indicates the presence of a through-crack or complete fracture, the server generates fracture confirmation tag data and performs a logical fusion assessment with the failure trigger tag data calculated based on the damage time series. If either of the two is true, the system executes a door locking operation and enters the responsibility tracing process.

[0088] By introducing a visual confirmation mechanism independent of the mechanical operator detection path, this embodiment constructs a dual-channel redundant detection system of "microscopic performance degradation prediction - macroscopic fracture confirmation", thereby avoiding the risk of misjudgment caused by relying solely on probability model judgment and ensuring that fracture events are supported by a chain of physical evidence.

[0089] In real-world equipment service, material degradation is a complex signal resulting from the superposition of "natural aging" and "malicious human damage." The technical necessity of this step lies in smoothing out factory tolerances by introducing initial characteristic fingerprint data throughout the entire lifecycle; then, time-recursive filtering is used to eliminate normal wear (natural fatigue trend data) that should be borne by the service provider. This logic precisely locks in the cost of malicious and unauthorized operations by users, providing a strong defense for fair liability determination within the system, while the proactive failure warning mechanism completely eliminates the safety hazard of injury from broken frames.

[0090] When the server detects that equipment has reached a failure threshold or a breakage event has occurred, it reads historical damage increment data from the damage time series database, calculates the damage contribution ratio data corresponding to each rental session, generates liability determination result data, and sends it to the user terminal. The generation of the damage contribution ratio data includes: the server reading the damage increment data for the current rental period stored in the damage time series database according to rental session identifiers; normalizing the damage increment data for the current rental period corresponding to each rental session identifier to generate damage contribution ratio data corresponding to each rental session; associating the damage contribution ratio data with the rental record to form liability allocation data; and in the... After the liability determination result data is generated, the server sends the liability allocation data to the payment and settlement module inside the server; the payment and settlement module calculates the corresponding compensation amount data based on the liability allocation data, and pushes the compensation amount data to the corresponding user terminal to complete the online deduction or settlement process; the system also includes a cloud storage unit; while generating the liability determination result data, the server encapsulates the dynamic stiffness characteristic operator data, the residual sequence data, and the damage contribution ratio data into detection report data and stores it in the cloud storage unit; the server responds to the user terminal's query request and returns the detection report data to the user terminal.

[0091] Suppose a racket breaks completely after its Nth rental. The server traces the damage time-series database and extracts the nonlinear damage increment data from all rentals of the racket from its manufacturing date to its breakage. This data is then normalized using a weighted allocation algorithm.

[0092] To address the challenge of assigning responsibility for progressive equipment failures caused by multi-user involvement, this embodiment employs an energy-weighted backtracking method. The total damage throughout the equipment's lifecycle, from warehousing to failure, is considered the denominator, while the damage increment under a specific session identifier is considered the numerator. Based on a time-weighted correction for nonlinear damage, a responsibility allocation equation is constructed:

[0093] ;

[0094] In this embodiment, a normalized calculation formula for damage contribution based on time-series nonlinear increments is used to transform the discrete physical damage components generated by the equipment through multiple users throughout its entire life cycle into a logical basis with uniquely determined financial compensation weights. The formula constructs a weighted normalization model with a time gain coefficient, aiming to solve the problem of liability determination caused by the nonlinearity of damage accumulation during composite material failure, that is, to correct the physical characteristic that the closer to the equipment failure critical point, the higher the induction weight of unit damage increment to the final fracture.

[0095] At the level of variable definition and data acquisition, output variables This represents the damage contribution ratio data corresponding to the j-th rental session. The numerator contains... The damage increment for the j-th rental period, extracted by a time-recursive filtering algorithm, reflects the net nonlinear damage generated during the usage period of a specific user. To compensate for the "damage acceleration effect" caused by material fatigue, a time-weighted factor is introduced into the formula. Where N represents the total number of rentals of the equipment from its initial use until the macroscopic failure, obtained by counting rental record entries on the server; j is the time-series index number of the current rental session. Parameters Defined as the damage evolution acceleration sensitivity coefficient, this value is an empirical constant preset based on the material accelerated aging test data, used to adjust the sensitivity of the time weight to the final contribution.

[0096] The denominator of the formula Weighted damage summation was performed for all rental periods throughout the equipment's entire lifecycle, thus constructing a normalized baseline with a total of 100%. The server retrieved all historical damage increments under this equipment identifier from the damage time-series database through source tracing. The system iterates through the data and calculates the physical damage from a single rental based on its corresponding temporal position i. Through a proportional mapping between the numerator and denominator, the system effectively transforms the absolute value of the physical damage from a single rental into a relative value of the total contribution over the entire rental period. This algorithm ensures that even when the absolute amounts of damage caused by different users are equal, the differences in their temporal distribution (i.e., ...) will lead to a more balanced and efficient system. (Different), the final compensation ratio will also be allocated differently according to the evolution law of physical damage.

[0097] From the perspective of logical consistency in engineering liability determination, this formula achieves a deterministic mathematical mapping from microscopic mechanical attenuation to macroscopic economic loss assessment. It ultimately converges the deeply black-boxed mechanical behavior of composite materials (load-displacement hysteresis, recursive residual analysis, Weibull failure probability, etc.) into transparent, traceable, and statistically significant percentage data. This normalization process based on time series weights not only eliminates the logical flaws of the traditional approach where end-users bear full responsibility, but also provides the financial settlement module with underlying algorithmic support that conforms to the causal laws of mechanics, ensuring the rigor of the liability determination report in both engineering technology and legal logic dimensions.

[0098] The damage contribution ratio data for each rental session is correlated with the corresponding rental records to form precise liability allocation data, which is then pushed to the payment and settlement module. The settlement module multiplies this ratio by the original value of the equipment to determine the specific compensation amount and initiates online deductions from multiple user terminals. Simultaneously, the system packages the relevant dynamic stiffness characteristic operator data and residual sequence data into tamper-proof PDF inspection report data in the background, stores it in the cloud, and makes it available for review and query by the penalized user terminals.

[0099] This step completely resolves the technical pain point of mutual blame-shifting caused by hidden internal damage. Through rigorous mathematical normalization, the mechanical attenuation indicators, previously hidden in a black box, are transformed into transparent and legally convincing liability-sharing data, achieving a seamless connection between complex physical and mechanical events and fair digital financial liability determination. Simultaneously, the mechanism of encapsulating the evidence chain and providing access to test report data fundamentally eliminates user doubts about automatic deductions, constructing a complete closed loop of business trust.

[0100] Example 2

[0101] This embodiment discloses an intelligent trial-play rental service system for racket-type sports equipment, including: a server, an intelligent cabinet, an intelligent cabinet control unit, a programmable loading module, and a user terminal; the intelligent cabinet is equipped with an intelligent cabinet control unit and a programmable loading module, the intelligent cabinet control unit is used to receive door opening control commands sent by the server and return equipment identification data; the programmable loading module is used to collect load-displacement curve data and elastic recovery trajectory data and upload them to the server; the user terminal is used to send trial-play request data and receive liability determination result data.

[0102] This embodiment discloses an intelligent cabinet, which is equipped with multiple racket sockets, a locking mechanism, a speaker, and a display, the display of which can display a QR code.

[0103] The racket socket is used to insert and fix racket-type sports equipment;

[0104] The locking mechanism is electrically connected to the intelligent cabinet control unit and is used to control the opening and closing of the racket socket;

[0105] The speaker is used to play voice prompts;

[0106] The display is used to show rental status information and liability determination results.

[0107] In one embodiment, the smart cabinet further includes an image acquisition unit and an edge computing module; the image acquisition unit is used to acquire image data of the equipment's external structure; the edge computing module is used to preprocess the image data and upload it to the server. The smart cabinet also includes an accelerometer module for acquiring impact and vibration data of the equipment.

[0108] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them; although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications can still be made to the specific implementation methods of the present invention or equivalent substitutions can be made to some technical features without departing from the spirit of the technical solutions of the present invention, and all such modifications should be covered within the scope of the technical solutions claimed in the present invention.

Claims

1. A method for providing intelligent trial rental services for racket-type sports equipment, characterized in that, A smart trial rental service system for racket-type sports equipment is provided. The system includes a server, a smart cabinet control unit, a programmable loading module, a user terminal, a session management database, and an injury time series database. The method includes the following steps: Step S1: The server receives the trial call request data sent by the user terminal, performs identity verification and payment status verification on the trial call request data, generates a rental session identifier and stores it in the session management database. Step S2: The server sends an opening control command to the smart cabinet control unit according to the rental session identifier, and receives the equipment identifier data returned by the smart cabinet control unit. The server binds the equipment identifier data with the rental session identifier to generate a rental record. Step S3: When the equipment is returned, the server receives the load-displacement curve data and elastic recovery trajectory data uploaded by the programmable loading module, performs feature extraction on the load-displacement curve data, and generates dynamic stiffness feature operator data. Step S4: The server pre-acquires the benchmark dynamic stiffness characteristic data generated when the equipment is first put into storage, and stores the benchmark dynamic stiffness characteristic data as the initial characteristic fingerprint data of the entire life cycle. The server calls the initial characteristic fingerprint data of the entire life cycle corresponding to the equipment and the dynamic stiffness characteristic operator data to perform recursive residual analysis processing, generate the damage increment data of this rental period and write it into the damage time series database. Step S5: When the server detects that the equipment has reached the failure threshold or a breakage event has occurred, it reads the historical damage increment data in the damage time series database, calculates the damage contribution ratio data corresponding to each rental session, generates the responsibility determination result data, and sends it to the user terminal.

2. The method according to claim 1, characterized in that, The acquisition of the load-displacement curve data includes: the programmable loading module applying multi-order quasi-static loading control commands to the equipment; the programmable loading module internally collecting real-time torque data and displacement data during the loading process; the server performing time synchronization processing on the real-time torque data and displacement data to generate standardized load-displacement curve data, and performing piecewise fitting processing on the recovery path during the unloading stage to obtain elastic recovery trajectory data; the elastic recovery trajectory data is used as input data for generating dynamic stiffness feature operator data.

3. The method according to claim 1, characterized in that, The generation of the dynamic stiffness feature operator data includes: the server performing slope segmentation calculation on the load-displacement curve data to extract stiffness change feature data during the loading and unloading stages; performing energy dissipation analysis on the elastic recovery trajectory data to extract hysteresis area feature data; and performing feature fusion processing on the stiffness change feature data and the hysteresis area feature data to generate dynamic stiffness feature operator data for characterizing structural integrity.

4. The method according to claim 1, characterized in that, The initial feature fingerprint data of the entire life cycle includes the reference load-displacement curve data and reference elastic recovery trajectory data obtained by executing the standard loading process through the programmable loading module when the equipment is first put into storage; the server performs feature extraction processing on the reference load-displacement curve data and reference elastic recovery trajectory data to generate reference dynamic stiffness feature operator data, and stores the reference dynamic stiffness feature operator data as the initial feature fingerprint data of the entire life cycle.

5. The method according to claim 1, characterized in that, The recursive residual analysis process includes: the server performing cycle-by-cycle difference calculation on the dynamic stiffness feature operator data and the initial feature fingerprint data of the entire life cycle to generate residual sequence data; performing time recursive filtering on the residual sequence data to separate natural fatigue trend data and nonlinear mutation data; and using the nonlinear mutation data as the damage increment data for this lease cycle and writing it into the damage time series database.

6. The method according to claim 1, characterized in that, The failure threshold is obtained by the server through statistical modeling based on the damage time series data corresponding to the historical sample equipment before fracture. After each write of the incremental damage data for the current rental period, the server updates the cumulative damage field in the damage time series database to obtain the current cumulative damage data and compares it with the failure threshold. When the current cumulative damage data reaches or exceeds the failure threshold, failure trigger identification data is generated.

7. The method according to claim 1, characterized in that, The generation of the damage contribution ratio data includes: the server reading the damage increment data of the current rental period stored in the damage time series database according to the rental session identifier; normalizing the damage increment data of the current rental period corresponding to each rental session identifier to generate damage contribution ratio data corresponding to each rental session; and associating the damage contribution ratio data with the rental records to form responsibility allocation data.

8. The method according to claim 1, characterized in that, After the liability determination result data is generated, the server sends the liability allocation data to the payment and settlement module inside the server; the payment and settlement module calculates the corresponding compensation amount data based on the liability allocation data, and pushes the compensation amount data to the corresponding user terminal to complete the online deduction or settlement process.

9. The method according to claim 1, characterized in that: The system also includes a cloud storage unit; while generating liability determination result data, the server encapsulates dynamic stiffness characteristic operator data, residual sequence data, and damage contribution ratio data into detection report data and stores it in the cloud storage unit; the server responds to the user terminal's query request and returns the detection report data to the user terminal.

10. A smart trial rental service system for racket-type sports equipment, characterized in that, include: Server, intelligent cabinet control unit, programmable loading module, user terminal, session management database, damage time series database and cloud storage unit; The server is used for: Receive trial call request data sent by user terminal, perform identity verification and payment status verification, generate rental session identifier and write it to session management database; The system sends an opening control command to the smart cabinet control unit based on the rental session identifier, and receives the equipment identifier data returned by the smart cabinet control unit to generate a rental record. When the equipment is returned, the load-displacement curve data and elastic recovery trajectory data uploaded by the programmable loading module are received, and dynamic stiffness characteristic operator data are generated. The initial feature fingerprint data of the entire life cycle stored in the session management database or the damage time series database is called and recursively residual analysis is performed with the dynamic stiffness feature operator data to generate the damage increment data of the current lease cycle and write it into the damage time series database. Based on the comparison between the current cumulative damage data in the damage time series database and the failure threshold, failure trigger identification data is generated when the failure threshold is reached or exceeded. When generating liability determination result data, the incremental damage data of the current rental period stored in the damage time series database according to the rental session identifier is read, the damage contribution ratio data is calculated, liability allocation data is formed and sent to the user terminal. The server also includes a payment and settlement module, which is used to generate compensation amount data based on liability sharing data and complete deduction or settlement processing. The server is also used to encapsulate dynamic stiffness feature operator data, residual sequence data, and damage contribution ratio data into detection report data and store them in the cloud storage unit.

Images