Clamp health state evaluation method and system based on internet of things

Abstract

The present application belongs to the field of intelligent manufacturing and equipment operation and maintenance technology, and specifically relates to a fixture health state evaluation method and system based on the Internet of Things. The method comprises: arranging at least three displacement sensors on the fixture positioning plane, arranging a vibration sensor on the fixture body, and collecting multi-point displacement and vibration response; performing online parameter identification based on a simplified digital twin model of the fixture finite element to obtain a digital twin parameter vector including contact stiffness, fixture frame stiffness, and support plane inclination angle; under the boundary conditions of a standard workpiece, a nominal clamping force, and a typical cutting load, simulating to obtain the maximum deformation of the workpiece, and constructing a deformation occupation ratio with the allowable geometric tolerance, and accordingly defining a monotonically decreasing fixture health index HI. The present application realizes the interpretable quantitative evaluation and life prediction of the fixture health state, and reduces the quality fluctuation and sudden shutdown risk caused by fixture degradation.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing and equipment operation and maintenance technology, specifically relating to a method and system for assessing the health status of fixtures based on the Internet of Things. Background Technology

[0002] With the development of CNC machining, intelligent manufacturing, and flexible production lines, tooling fixtures not only perform the functions of workpiece positioning and clamping, but their geometric accuracy, rigidity, and long-term stability also directly affect the dimensional accuracy, form and position accuracy of the workpiece, and the stability of the machining process. In actual production, key components such as machine tool spindles and cutting tools are usually equipped with relatively complete online monitoring and predictive maintenance methods. However, due to the diverse structural forms and complex operating conditions, the health status assessment and life cycle management methods for fixtures are relatively weak. It is difficult to detect and prevent product quality fluctuations and sudden downtime caused by fixture degradation in a timely manner.

[0003] For fixture condition monitoring, existing technologies have begun to explore the introduction of multi-source data fusion and model-based judgment. For example, Chinese patent document CN114227378A discloses a method for detecting fixture condition. This method acquires fixture condition data, environmental data, and machine tool equipment data of the fixture under test. It then performs outlier removal and feature fusion on the fixture condition data and environmental data to form target fused data. This target fused data is input into a condition monitoring model associated with the machine tool equipment data, outputting anomaly scores. When the anomaly score exceeds a threshold, the fixture is determined to be in an abnormal state. This approach can achieve online monitoring of fixture condition to a certain extent, avoiding product quality degradation caused by fixture anomalies. However, such methods are mostly based on statistical features and data-driven models. The output "anomaly scores" lack clear physical meaning and are difficult to reflect the evolution process of key physical parameters such as fixture contact stiffness, frame stiffness, and support plane inclination. They also do not provide quantitative indicators directly related to workpiece geometric tolerances, nor do they address the prediction of remaining life based on the fixture degradation process.

[0004] In the digitalization of clamping processes, digital twin technology has been introduced into the field of clamping force control and deformation analysis for sensitive structures such as thin-walled parts. Chinese patent document CN112926152A discloses a digital twin-driven method for precise control and optimization of clamping force for thin-walled parts. This method utilizes IoT technology to equip the thin-walled part, fixture, and machine tool with various sensors to acquire clamping force and clamping deformation data. Based on this, digital twin virtual models are established, including a geometric model of the thin-walled part, a finite element deformation simulation model, and a clamping deformation prediction model. These models are used to map and simulate the clamping process of the thin-walled part in real time. Furthermore, a clamping force optimization model is constructed to solve for the clamping force range that meets the clamping deformation error requirements, thereby improving the machining accuracy of thin-walled parts. This scheme demonstrates that combining IoT measurement, finite element simulation, and digital twin technology for closed-loop optimization of the "clamping process-workpiece deformation-clamping force control" is feasible. However, this paper mainly focuses on the clamping deformation and clamping force selection of thin-walled parts. The core of the digital twin model is the deformation prediction of the workpiece under different clamping conditions, rather than the geometric accuracy degradation and stiffness attenuation of the fixture body in long-term use. Its optimization results are also mainly aimed at meeting the deformation error in a single machining operation. It has not established a health index system for fixture life management, nor has it considered the periodic identification and degradation modeling of the fixture physical parameters under the same working conditions.

[0005] In recent years, the academic community has also proposed digital twin models of fixtures for precision manufacturing. For example, Ma Songhua et al., in their paper "Digital Twin Model of Fixtures Supporting Rapid Design and Performance Tracking," proposed constructing a high-fidelity model of fixtures based on digital twin theory to support rapid variation design and performance tracking of fixtures. By establishing a closed loop of "physical fixture—virtual model—data interaction," they verified the impact of fixtures on the clamping stability and positioning accuracy of precision parts machining over a long lifespan, achieving collaborative optimization between the fixture design and application stages. This type of work demonstrates the feasibility and value of digital twins for fixtures at the system level, providing a virtual platform for performance analysis of fixtures throughout their entire life cycle. However, their focus remains on model structure, parameter configuration, and simulation verification during the design phase. They lack specific implementation paths for how to use a limited number of sensors and repeatable working conditions on the production floor to identify fixture physical parameters online, construct quantifiable health indicators, and predict lifespan.

[0006] In specific sub-fields, technological solutions such as chip fixture debugging systems based on digital twins have emerged. For example, Chinese patent document CN120178694A proposes a chip fixture debugging system based on digital twins. Through a state perception fusion module, a positioning correction judgment module, a contact adjustment execution module, a closed-loop feedback correction module, and a data linkage optimization module, it performs time-axis synchronous matching of multiple types of sensor data, extracts the intersection features of position offset and pressure change, and achieves high-precision attitude recognition and debugging path optimization. This solution fully leverages the advantages of digital twins in fixture attitude correction and contact adjustment during the debugging phase. However, its focus remains on the response efficiency and attitude accuracy of the debugging process, without establishing a model framework for health status assessment and remaining life prediction to address issues such as stiffness degradation, support plane tilting, and changes in contact conditions during long-term use of the fixture.

[0007] Based on the above existing technologies, it can be seen that:

[0008] On the one hand, fixture status detection methods based on multi-source data fusion and status monitoring models (such as CN114227378A) can identify whether the fixture is in an abnormal state, but the output results are mostly abstract abnormal scores, which do not have physical meaning directly corresponding to the workpiece geometric tolerance. They are difficult to use for quantitative evaluation of the impact of the fixture on machining accuracy, and also lack the characterization of the fixture degradation process and remaining life.

[0009] On the other hand, research on clamping process optimization methods driven by digital twins (such as CN112926152A) and fixture digital twin models mainly focuses on clamping force selection, workpiece deformation control, and fixture design performance verification. It emphasizes the instantaneous performance of "workpiece / fixture under given design parameters" rather than "physical parameter degradation of fixture during long-term operation" and its cumulative impact on subsequent workpiece quality. Related solutions usually do not periodically excite the fixture under unified standard working conditions, nor do they identify parameters such as contact stiffness, frame stiffness, and support plane tilt angle online based on simplified finite element models, and they do not utilize workpiece allowable tolerances to construct a fixture health index.

[0010] Furthermore, existing research on intelligent operation and maintenance of equipment throughout its entire life cycle driven by digital twins is mostly focused on large equipment and rotating machinery. The construction of health indicators is often based on abstract quantities such as vibration amplitude, energy spectrum, and statistical characteristics. There is a lack of ideas to directly use the "proportion of structural deformation on product tolerance" as a health metric, and there is also a lack of utilization of changes in geometric parameters (such as the inclination angle of the support plane) and the attenuation of contact stiffness as physical constraints in the degradation model.

[0011] Therefore, in industrial settings, how to combine digital twins and finite element analysis under IoT conditions to periodically identify the physical parameters of fixtures through dedicated excitation conditions for empty clamping, construct a health index with clear physical meaning by combining the maximum deformation of the workpiece with the allowable geometric tolerance, and perform degradation modeling and remaining life prediction under physical constraints, thereby providing a quantitative basis for maintenance decisions and process compensation, still requires further research and improvement. Summary of the Invention

[0012] The technical objective of this invention is to provide a quantitative assessment method and system for the health status of fixtures throughout their entire lifecycle, based on the Internet of Things and digital twin technologies. This method involves collecting multi-point displacement and vibration responses of the fixture under dedicated no-clamping conditions, combining this data with a simplified finite element digital twin model of the fixture, and identifying key physical parameters such as contact stiffness, frame stiffness, and the tilt angle of the supporting plane online. Furthermore, a health index with clear physical meaning is constructed based on the ratio of the maximum workpiece deformation to the allowable geometric tolerance. This further enables the prediction of the remaining usable life of the fixture under physical constraints, thereby providing a reliable quantitative basis for fixture maintenance decisions and process compensation, and reducing the risk of product quality fluctuations and sudden downtime caused by fixture degradation.

[0013] On the one hand, in order to achieve the above objectives, the present invention adopts the following technical solution:

[0014] A method for assessing the health status of a fixture based on the Internet of Things (IoT) includes the following steps:

[0015] S1. Arrange at least three displacement sensors on the fixture positioning plane and a vibration sensor on the fixture body. When the fixture is in an empty clamping state, apply a preset clamping force for loading and unloading and / or spindle speed step excitation through the machine tool control system to collect multi-point displacement and vibration response.

[0016] S2. Based on the finite element simplified digital twin model of the clamp, online parameter identification is performed on the response to obtain a digital twin parameter vector consisting of contact stiffness, clamp frame stiffness and support plane tilt angle.

[0017] S3. Under standard workpiece, nominal clamping force, and typical cutting load boundary conditions, the maximum deformation Δ of the workpiece is obtained by simulating using a digital twin model. max and with the workpiece's allowable geometric tolerance Δ allow The ratio r = Δ max / Δ allow Construct the fixture health index HI = f(r), where f is a monotonically decreasing function;

[0018] S4. Based on the degradation sequence of HI as a function of clamping cycles obtained from multiple tests, a degradation model is established under the physical constraints that HI does not increase monotonically and is correlated with the decrease in contact stiffness and the increase in the inclination angle of the support plane. The remaining number of clamping cycles when HI drops to a preset threshold is predicted as the remaining usable life (RUL) of the fixture. HI and RUL are output through the Internet of Things platform for maintenance and process compensation decisions.

[0019] Preferably, in S1, at least three displacement sensors are arranged at three non-collinear points on the fixture positioning plane. By performing planar fitting on the displacement increments of the three points, the overall rigid translation of the fixture positioning plane and the change of tilt angle around two mutually perpendicular axes are obtained. The tilt angle of the support plane is calculated from the change of tilt angle.

[0020] Preferably, the finite element simplified digital twin model of the fixture includes: a finite element simplified model with the fixture frame as an elastic body element, the contact area between the fixture and the machine tool table as a nonlinear spring element, and the clamping element as the point of concentrated force application. The online parameter identification only solves for the contact stiffness, the equivalent Young's modulus of the frame, and the equivalent offset of the support point, so as to reduce the amount of calculation and maintain correspondence with the sensor arrangement.

[0021] Preferably, the construction of the health index HI in step S3 includes: first calculating the deformation occupancy ratio r = Δ max / Δ allow When r is below the first preset threshold, HI is set to a healthy range value close to 1. When r is above the second preset threshold, HI is set to a failure range value close to 0. An exponential or piecewise linear mapping function f(r) is used to generate a continuously changing health index HI between the first preset threshold and the second preset threshold.

[0022] Preferably, the degradation model established in step S4 includes: smoothing and outlier removal of the discrete sequence of HI with the number of clamping cycles; fitting the degradation curve using a constrained nonlinear regression or time series prediction algorithm under the constraint that HI(k+1)≤HI(k); and adding a penalty term to the objective function that is consistent with the direction of HI change for the contact stiffness attenuation rate and the change rate of the support plane tilt angle, so that the degradation model conforms to the actual mechanical damage evolution law of the fixture.

[0023] Preferably, the method further includes: when the health index HI is lower than the warning threshold but higher than the failure threshold, calling the digital twin model to simulate the maximum deformation of the workpiece under different clamping force levels, workpiece support point layouts, or the addition of auxiliary support blocks, and comparing the corresponding deformation occupancy ratio r, selecting the process compensation scheme that can keep r within the allowable range and has the least impact on the production cycle without replacing the fixture, and pushing the process compensation scheme and fixture maintenance suggestions to the manufacturing execution system through the Internet of Things platform.

[0024] Preferably, when constructing the initial health benchmark of the digital twin parameter vector in step S2, the empty clamping detection data of multiple processing batches corresponding to the key dimensions and geometric tolerances of the workpiece are selected, and the fluctuation range is less than the preset value. It is required that the main mode frequency offset of the vibration spectrum in the data does not exceed the predetermined threshold, so as to ensure that the initial digital twin parameter vector corresponds to the health state of the fixture.

[0025] On the other hand, the present invention also provides an Internet of Things-based fixture health status assessment system, comprising:

[0026] A fixture measuring device is used to arrange at least three displacement sensors on the fixture positioning plane and a vibration sensor on the fixture body, and to configure a data acquisition node electrically connected to the sensors.

[0027] The machine tool control interface module is used to send preset clamping force loading-unloading and / or spindle speed step excitation commands to the machine tool control system when the fixture is in an unclamped state.

[0028] The IoT edge gateway is used to receive the multi-point displacement response and vibration response uploaded by the acquisition node and forward them to the host server through the Industrial Internet of Things protocol.

[0029] The digital twin modeling and parameter identification module, deployed on the host server, is used to perform online parameter identification on the response based on the simplified digital twin model of the clamp with finite element, and obtain a digital twin parameter vector composed of contact stiffness, clamp frame stiffness and support plane tilt angle.

[0030] The health index calculation module is used to call the digital twin model to simulate and obtain the maximum deformation Δmax of the workpiece under the boundary conditions of standard workpiece, nominal clamping force and typical cutting load, and construct the fixture health index HI=f(r) based on the ratio r of Δmax to the workpiece's allowable geometric tolerance Δallow.

[0031] The degradation modeling and life prediction module is used to establish a degradation model based on the degradation sequence of the health index HI obtained from multiple empty clamping tests as a function of the number of clamping cycles. Under the physical constraints that HI does not increase monotonically and is correlated with the decrease in contact stiffness and the increase in the inclination angle of the support plane, the module predicts the remaining usable life (RUL) of the fixture.

[0032] The maintenance and process compensation decision module is used to output the Health Index (HI) and Remaining Usable Life (RUL) through the IoT platform, and generate maintenance work orders and process compensation suggestions.

[0033] Preferably, the digital twin modeling and parameter identification module includes: a model simplification unit, used to simplify the three-dimensional structure of the fixture into a finite element model containing elastic frame elements, nonlinear contact spring elements, and clamping concentrated loads; and a parameter solving unit, used to use the least squares method, Bayesian estimation, or other optimization algorithms to fit the three-point displacement and vibration response output by the model simulation to the actual measurement data, thereby solving for the three sets of physical parameters: contact stiffness, equivalent frame stiffness, and equivalent offset of the support point.

[0034] Preferably, the maintenance and process compensation decision module is further configured to: when the health index HI is lower than the warning threshold and the remaining usable life RUL is less than the preset upper limit, call the health index calculation module and the digital twin modeling and parameter identification module to simulate the maximum deformation of the workpiece under different clamping forces, support point arrangements and auxiliary support block combinations, screen out the process compensation scheme that meets the allowable geometric tolerance of the workpiece and has the least impact on the production cycle, and push the scheme together with the maintenance plan for fixture correction, realignment or replacement to the manufacturing execution system and the maintenance management system.

[0035] On the other hand, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method.

[0036] On the other hand, the present invention also provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the method.

[0037] This invention acquires highly sensitive multi-point response data under a dedicated excitation condition of no-clamping by arranging at least three displacement sensors on the fixture positioning plane and combining them with vibration sensors. Using a simplified digital twin model based on the finite element method, physical parameters such as contact stiffness, fixture frame stiffness, and support plane tilt angle are identified online, achieving real-time and interpretable tracking of fixture geometric accuracy and stiffness degradation. Furthermore, a health index HI is constructed using the ratio of the maximum workpiece deformation to the allowable geometric tolerance under standard working conditions. This directly quantifies the fixture's health status as the degree of occupancy of workpiece tolerances, enabling the health assessment results to be correlated with machining dimensions and shape. A clear one-to-one correspondence is established between the quality of the fixture and the remaining usable life (RUL). Furthermore, by introducing physical constraints such as monotonic degradation of the health index, stiffness decay, and increased inclination angle of the support plane, a degradation model is established, which significantly improves the stability and reliability of the RUL prediction. In conjunction with the Internet of Things (IoT) platform, the HI (Health Index) and RUL are linked with maintenance work orders and process compensation schemes, enabling enterprises to proactively arrange correction, realignment, or replacement, as well as adjust process parameters such as clamping force and support point layout before fixture failure. This effectively reduces batch deviations and sudden downtime caused by fixture degradation, and improves the processing consistency, equipment utilization rate, and maintenance refinement level of the production line. Attached Figure Description

[0038] Figure 1 This is a schematic block diagram of the overall framework of the IoT-based fixture health status assessment system of the present invention.

[0039] Figure 2 This is a flowchart of the IoT-based fixture health status assessment method of the present invention.

[0040] Figure 3 This is a schematic diagram of the frame structure of the simplified digital twin model with finite element method of the present invention. Detailed Implementation

[0041] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of the present invention.

[0042] I. Terminology Explanation

[0043] To facilitate understanding of the present invention by those skilled in the art, the relevant terms are explained first:

[0044] 1. Empty clamping state

[0045] "Empty clamping state" refers to a situation where the fixture is in its normal clamping posture, the clamping mechanism applies a set clamping force, but no workpiece is clamped at the clamping station, and there is only contact constraint between the fixture and the machine tool table. In this state, the displacement and vibration response of the fixture mainly come from the structural stiffness of the fixture itself, the contact stiffness, and the connection stiffness with the machine tool table. This avoids interference from the workpiece stiffness and mass on the response, and is beneficial for repeatable health checks.

[0046] 2. Fixture positioning plane

[0047] The "fixture positioning plane" is typically the supporting plane where the fixture contacts the workpiece, or the mounting reference plane between the fixture and the machine tool table. It can be a single machining plane or an equivalent plane formed by a combination of multiple support blocks and support pins. This invention arranges at least three displacement sensors on this positioning plane and obtains the rigid translation and tilt changes of the plane through three-point displacement fitting, which is used to characterize the degradation of the fixture positioning accuracy.

[0048] 3. Simplified digital twin model using finite element method

[0049] A simplified digital twin model of finite element (FEM) refers to a finite element simulation model established for a specific fixture. While retaining the key structural features that affect the overall stiffness and positioning accuracy of the fixture, it simplifies local details, models the fixture frame as an elastic body element, models the contact area between the fixture and the machine tool table as a nonlinear spring element, and models the clamping elements as concentrated load application points. In this way, the mechanical characteristics of the fixture-machine tool system can be described with a small number of equivalent physical parameters, enabling online rapid simulation and parameter identification.

[0050] 4. Digital Twin Parameter Vector

[0051] The digital twin parameter vector is the set of physical parameters that drive the simplified finite element model described above, for example:

[0052] ;

[0053] in, This refers to the equivalent contact stiffness between the fixture and the machine tool table. The equivalent Young's modulus of the fixture frame; , These respectively represent the positioning plane around two orthogonal axes (e.g. shaft and The tilt angle of the axis; It can represent the equivalent offset of the support point, etc. This invention includes at least three types of parameters: contact stiffness, frame stiffness, and the tilt angle of the support plane.

[0054] 5. Number of clamping cycles

[0055] The clamping cycle count refers to the cumulative number of complete "clamping-releasing" cycles that the clamp undergoes during use. express. It is highly correlated with fatigue damage accumulation and wear degree, and is an important measure of fixture life.

[0056] Maximum deformation of workpiece With allowable geometric tolerances

[0057] The maximum deformation value of a key position or machined surface of a workpiece obtained through simulation using a digital twin model under given working conditions (standard workpiece, nominal clamping force, and typical cutting load); the dimensional or geometric tolerance value allowed for this process in the drawings or process documents.

[0058] 6. Deformation Occupancy Ratio

[0059] The deformation occupancy ratio is defined as follows:

[0060] in, This represents the maximum deformation of the workpiece under standard operating conditions, and represents the allowable tolerance for the corresponding process. This reflects the extent to which workpiece deformation occupies the tolerance under the current fixture condition.

[0061] 7. Health Index

[0062] Health Index This invention constructs a dimensionless health index, with a value range of [value range missing]. Through a monotonically decreasing function Deformation occupancy ratio Mapped to :when When the deformation is small (much smaller than the tolerance), A value close to 1 indicates that the clamp is healthy; when When it is close to or exceeds 1, A value close to 0 indicates that the fixture is nearing or has reached a failure state.

[0063] 8. Remaining usable lifespan

[0064] Remaining usable lifespan This refers to the number of clamping cycles from the current position. From now on, under the condition of continued use in the future, the health index Reduced to a preset failure threshold The corresponding remaining number of clamping cycles:

[0065] ;

[0066] in, This represents the number of failure cycles predicted by the model.

[0067] 9. Standard Operating Conditions

[0068] Standard operating conditions refer to a set of unified boundary and load conditions used for health index calculation and fixture condition comparison, including: standard workpiece model, nominal clamping force, typical cutting load and matching workpiece material parameters, so that simulation results of different detection cycles are comparable.

[0069] II. System Structure

[0070] Reference Figure 1 The present invention provides an IoT-based fixture health status assessment system, comprising: a fixture measuring device; a machine tool control interface module; an IoT edge gateway; a host server; a manufacturing execution system and maintenance management system; and a process and tolerance database.

[0071] 1. Fixture measuring device

[0072] The fixture measuring device is installed on the fixture to be monitored and is used to collect multi-point displacement and vibration response of the fixture under no-clamping conditions. It mainly includes:

[0073] At least three displacement sensors are fixed at three non-collinear positions on the positioning plane of the fixture, forming an approximate triangular arrangement, to measure displacement changes at different positions on the plane under no-clamp conditions;

[0074] At least one vibration acceleration sensor: installed in a relatively rigid area on the fixture body, such as the side wall of the fixture frame or near the reinforcing ribs, to collect the vibration response of the fixture under excitation conditions;

[0075] Data acquisition node: Electrically connected to the sensor, used to provide power, complete data sampling and A / D conversion, and after timestamping the acquired displacement and vibration data, it is sent to the IoT edge gateway via wired or wireless means.

[0076] Displacement sensors can be inductive, LVDT, capacitive, or laser displacement sensors; vibration sensors can be piezoelectric accelerometers or MEMS accelerometers. Data acquisition nodes can be implemented using industrial-grade microcontrollers, ARM processors, or FPGAs.

[0077] 2. Machine tool control interface module

[0078] The machine tool control interface module can be an industrial PC or embedded controller installed in the machine tool control cabinet, connected to the machine tool numerical control system (CNC) or PLC via a bus, for:

[0079] When the fixture is in an unclamped state, a preset clamping force loading-unloading command is sent to the machine tool control system;

[0080] Control the spindle to perform a step change in speed to achieve a step excitation of the spindle speed;

[0081] Information such as excitation type, amplitude, timestamp, and clamping cycle count is uploaded to the host server and edge gateway for time alignment with sensor data.

[0082] 3. IoT Edge Gateway

[0083] The IoT edge gateway connects to the data acquisition node via fieldbus or industrial Ethernet and is used for:

[0084] It receives multi-point displacement and vibration responses;

[0085] Receive excitation status data transmitted by the machine tool control interface module;

[0086] Perform preliminary caching, synchronization, and preprocessing (such as simple filtering and compression) on the data;

[0087] Data is uploaded to the host server via industrial Ethernet, 5G, or Wi-Fi.

[0088] 4. Host server

[0089] The host server is the computational core of the method of this invention, internally deploying multiple functional modules: a digital twin modeling and parameter identification module; a health index calculation module; a degradation modeling and life prediction module; a maintenance and process compensation decision-making module; and a data storage and management module. The digital twin modeling and parameter identification module includes a model simplification unit and a parameter solving unit, used to maintain the simplified finite element model and solve for the digital twin parameter vector; the health index calculation module is used for simulation calculations under standard operating conditions. And build The degradation modeling and lifetime prediction module is used to generate... Degenerate sequences and prediction The maintenance and process compensation decision module is used for comprehensive... and Generate maintenance and process compensation solutions.

[0090] 5. MES / Maintenance System and Process and Tolerance Database

[0091] The MES / maintenance system connects to a host server to receive health assessment results and maintenance recommendations, and formulates specific execution plans based on production scheduling, tool change plans, and other information. The process and tolerance database stores standard workpiece parameters, nominal clamping forces, typical cutting loads, and corresponding geometric tolerances for each product and process, providing the foundational data for health index calculations.

[0092] III. Method Implementation Process and Specific Technical Route

[0093] like Figure 2 As shown, the IoT-based fixture health status assessment method of this embodiment includes the following steps:

[0094] S1: Multi-point displacement and vibration response acquisition under no-clamping excitation condition;

[0095] S2: Online parameter identification based on simplified digital twin model of finite element method;

[0096] S3: Standard Operating Condition Simulation and Health Index Build;

[0097] S4: Degenerative Sequences and Physically Constrained Degenerative Model Establishment predict;

[0098] Auxiliary steps: Generation of process compensation scheme and construction of initial health baseline.

[0099] S1: Multi-point displacement and vibration response acquisition

[0100] 1. Sensor Setup and Calibration

[0101] Select three spatially non-collinear points on the fixture positioning plane. , , Install displacement sensors, preferably forming an approximately equilateral triangle with the lines connecting the three points to improve fitting accuracy and sensitivity to plane tilt. During sensor installation, ensure the measuring rod or probe is perpendicular to the positioning plane for direct measurement of normal displacement. Install vibration sensors in the rigid area of ​​the fixture frame, securing them with screws or high-strength adhesive; the installation direction can be vertical. (Direction), or a suitable axial direction can be selected according to the main vibration direction of the fixture. After installation, the displacement sensor is zero-point calibrated so that all displacement outputs are zero or the same reference value when the fixture is not under force or excited; the vibration sensor is sensitivity calibrated to ensure that the coefficient between the output voltage and acceleration meets the calibration value.

[0102] 2. Establishing the empty clamping state

[0103] After the fixture completes a batch of machining, the workpiece is unloaded, the clamping mechanism is opened, and then a no-clamping operation is performed to ensure that the contact interface and clamping state between the fixture and the machine tool table are consistent with the actual machining, but no workpiece is placed on the workstation. At this time, the current clamping force or hydraulic pressure can be read and recorded through the machine tool control interface module.

[0104] 3. Incentive Sequence Design and Execution

[0105] In the no-clamp state, the machine tool is subjected to a preset excitation sequence via the machine tool control interface module. The excitation includes, but is not limited to:

[0106] 1. Clamping force loading-unloading excitation: Controlling the clamping force at... and The loading and unloading cycles vary in a stepped manner, for example, 3 to 5 loading-unloading cycles are performed within a single testing cycle, each time loading to the nominal clamping force. of Uninstall to Preload;

[0107] 2. Spindle speed step excitation: Control the spindle to step from a lower speed to a medium speed, and then step back to the initial speed, such as... wait.

[0108] The time curve of the excitation sequence is recorded by the machine tool control interface module and timestamped.

[0109] 4. Data Acquisition and Plane Fitting

[0110] The acquisition node acquires displacement signals from three points at a preset sampling frequency. , , and vibration signals Plane fitting is performed on the displacement increment within each excitation cycle to decompose the overall rigid translation and the tilt angle of the support plane. Let the spatial coordinates of the three points be... The displacement increment is Assume the plane displacement field can be expressed as:

[0111] ;

[0112] in, This represents the overall rigid translation amount. , With and around , The tilt angle of the axis is relevant. The solution is obtained by fitting using the least squares method. , , Then, by combining the geometric dimensions of the fixture, the support plane can be calculated around... shaft and Changes in the tilt angle of the axis .

[0113] This plane fitting process enables the present invention to extract overall translation and tilt information from three-point displacement data, providing a more stable data foundation for subsequent parameter identification.

[0114] S2: Online parameter identification based on simplified digital twin model of finite element method

[0115] 1. Simplified Finite Element Model Structure

[0116] like Figure 3 As shown, a simplified finite element model is established for a specific fixture, including:

[0117] Fixture frame model: Modeled using solid elements or beam elements, with material parameters derived from equivalent Young's modulus. Definitions of Poisson's ratio, density, etc. As one of the parameters to be identified;

[0118] Contact interface model: Several nonlinear spring elements are set in the contact area between the bottom surface of the fixture and the machine tool table, which are equivalent to contact stiffness. Based on the concentration of the contact area, it can be equivalent to a finite number of springs;

[0119] Clamping force application point: Set a concentrated force node at the application position of the clamping mechanism to apply the clamping load;

[0120] Sensing points: Measurement points are arranged in the model at positions corresponding to the three displacement sensors and one vibration sensor, which are used to output the simulated displacement and vibration response.

[0121] The model simplification unit is responsible for importing and simplifying the three-dimensional structure of the fixture into the above form, ensuring that the simulation scale is suitable for online calculation.

[0122] 2. Matching relationship between parameter vector and response

[0123] Let the digital twin parameter vector be: ;in, For contact stiffness; The equivalent Young's modulus of the frame; The static tilt angle of the supporting plane; This represents the equivalent offset of the support point or other sensitive parameters. For a single empty clamping test, the displacement and vibration response at several typical time points are extracted from the collected data to construct a measurement vector:

[0124] ;

[0125] Using the finite element model, under a given Under the same excitation conditions, the simulated response vector is obtained:

[0126] ;

[0127] 3. Objective function construction and optimization solution

[0128] To find the parameter vector that best fits the actual measurement data, the following objective function is constructed:

[0129] ;

[0130] in, This is an index of selected typical time points. For the number of time points, It is a 2-norm.

[0131] Using least squares, Bayesian estimation, or other optimization algorithms (such as the Levenberg-Marquardt algorithm) to... By minimizing the solution, we obtain:

[0132] ;

[0133] These are estimated values ​​for parameters such as contact stiffness, frame stiffness, and support plane tilt angle corresponding to this test. Those skilled in the art can adjust the parameter dimensions according to the complexity of the fixture structure, and it can still be implemented within the required range.

[0134] To avoid unreasonable solution results, physical constraints can be added to each parameter during the optimization process, for example... Not lower than a certain lower limit Not exceeding the theoretical upper limit of materials, etc.

[0135] S3: Health Index Build

[0136] 1. Standard operating condition loading and simulation

[0137] Retrieve the standard workpiece model and nominal clamping force for the target process from the process and tolerance database. and typical cutting load parameters. The parameter vector obtained in the current detection cycle. By applying a simplified finite element model and standard working condition boundary conditions, the deformation field at key locations of the workpiece under these conditions is calculated, and the maximum deformation value is extracted.

[0138] ;

[0139] in, This is a set of key measurement areas for the workpiece. For the center of the region The amount of displacement.

[0140] 2. Deformation Occupancy Ratio With health index

[0141] Based on the allowable geometric tolerances in the process drawings, calculate the deformation occupancy ratio: r = Δ max / Δ allow ,in, .

[0142] Based on this, a health index is constructed. In a preferred embodiment, two thresholds are selected. , ,satisfy And construct piecewise functions :

[0143] when hour, ;

[0144] This indicates that the deformation is much smaller than the tolerance, and the fixture is in good condition.

[0145] when hour, ;

[0146] in, To adjust the coefficient for the descent speed. At this time... exist Smooth decrease within the interval;

[0147] when hour, ;

[0148] in, To rapidly decrease the slope, A value close to 0 indicates that the fixture is nearing or has reached a failure state.

[0149] This invention utilizes deformation occupancy ratio By directly linking the health status of the fixture to the geometric tolerances of the workpiece, Each numerical range has an intuitive physical meaning, which is a significant technical advantage compared to simply identifying "abnormal scores".

[0150] S4: Degradation Model and RUL Prediction

[0151] 1. Degenerate sequence construction and preprocessing

[0152] This invention periodically executes S1 to S3 according to a preset strategy throughout the entire lifespan of the fixture (e.g., every time it experiences...). (either through a clamping cycle or once per shift), resulting in a series of health indices. and the corresponding number of clamping cycles This forms a degenerate sequence: Before degradation modeling, the above sequence is smoothed and outliers are removed. For example, moving average, Savitzky-Golay filtering and other methods are used to remove high-frequency noise, and outliers that deviate significantly from the trend are removed or replaced to avoid interference from measurement noise on the degradation model.

[0153] 2. Degenerate model with monotonicity constraints

[0154] From a physical perspective, damage and stiffness degradation in fixtures are generally irreversible; therefore, the health index should be adjusted accordingly. It is monotonically non-increasing. Therefore, constraints are added when fitting the degradation curve:

[0155] ;

[0156] In one implementation, an exponential degradation model can be used:

[0157] ;

[0158] in, , , The parameters to be fitted are... , , Simultaneous constraints during least-squares fitting. about The discrete values ​​satisfy the condition of being monotonically non-increasing.

[0159] 3. Penalty items for contact stiffness and tilt angle constraints

[0160] This invention further considers that the decrease in contact stiffness and the increase in the tilt angle of the support plane are important physical characteristics of fixture degradation. Therefore, a penalty term consistent with the direction of change in contact stiffness and tilt angle is introduced into the degradation model. Let the contact stiffness of the actual detection cycle be... The angle of inclination is Define the stiffness attenuation rate and the tilt angle change rate as follows:

[0161] ;

[0162] Construct the comprehensive loss function:

[0163] ;

[0164] in, These are predicted values ​​from the degradation model; Penalties for violating the monotonicity constraint; and , , Related to situations where the directions of change are inconsistent. These are the weighting coefficients.

[0165] By minimizing Under the condition of satisfying physical constraints, the parameters of the degenerate model are obtained, so that the fitting results are consistent with the historical data. The data is consistent with the actual degradation trends of stiffness and tilt angle.

[0166] 4. RUL Prediction and Output

[0167] Let the failure threshold of the health index be... ,Notice Solve the equation:

[0168] ;

[0169] get The current clamping cycle count is Then the remaining usable lifetime is:

[0170] ;

[0171] The degradation modeling and lifetime prediction module will determine the current detection cycle. and The data is sent to the maintenance and process compensation decision-making module and pushed to the MES / maintenance system via the IoT platform, supporting production personnel in making maintenance plans and process adjustment decisions.

[0172] Furthermore, when the health index Below the warning threshold However, it is still above the failure threshold. In this invention, without immediately replacing the fixture, a digital twin model is used to explore process compensation strategies. The specific steps are as follows:

[0173] 1) In the current parameter vector Based on this, multiple candidate process variable combinations are set, such as different clamping forces. Different workpiece support point layouts and solutions for adding or adjusting the position of auxiliary support blocks;

[0174] 2) Simulate each combination under standard or near-actual machining conditions and calculate the corresponding maximum workpiece deformation. Then calculate the deformation occupancy ratio: r （j） =Δ max （j） / Δ allow ;

[0175] 3) From satisfying (For example Among the combinations, the one with the least impact on production cycle time, energy consumption, and operational complexity is selected as the recommended process compensation scheme;

[0176] 4) The maintenance and process compensation decision module will combine this solution with... , The maintenance recommendations are pushed to the MES / maintenance system to guide temporary process adjustments on-site.

[0177] Furthermore, to improve the accuracy of health assessments, this invention preferably establishes an initial health baseline after the fixture is newly put into use or after a major overhaul:

[0178] 1) Select several processing batches, and the workpiece dimensions and geometric tolerances of these batches are all qualified and have small fluctuations.

[0179] 2) Multiple empty clamping tests shall be performed during the aforementioned batches, and the frequency deviation of the main mode of vibration spectrum shall not exceed a preset threshold, for example, not exceeding 2%;

[0180] 3) Input these detection data into the digital twin modeling and parameter identification module to obtain a set of parameter vectors representing health status. and corresponding health index (Usually close to 1);

[0181] 4) In subsequent testing, all parameter changes can be relative to... By making comparisons, it is easier to observe the relative changes in contact stiffness and tilt angle, thereby further improving the stability of degradation modeling.

[0182] Specific application example: Health assessment and life prediction of automotive transmission housing machining fixtures

[0183] 1. Application Scenarios and Technological Background

[0184] An automotive parts manufacturer uses a specialized multi-station combination fixture to rough and finish machine aluminum alloy transmission housings on a horizontal machining center production line. The fixture includes:

[0185] It is connected to the machine tool worktable via three adjustable support pads;

[0186] The housing is positioned and clamped using a hydraulic clamping cylinder and a positioning pin;

[0187] Each shift processes approximately For each workpiece, the critical flatness tolerance is Δ. \allow =0.03mm.

[0188] Under traditional usage, after several months of continuous use, this fixture occasionally exhibited out-of-tolerance flatness issues in certain batches of transmission housings, with these out-of-tolerance issues appearing intermittently and in batches. Previously, the company relied solely on manual experience to determine whether the fixture needed calibration, which presented the following problems:

[0189] 1) There are no quantitative indicators for evaluating the health of fixtures;

[0190] 2) It is difficult to detect the degradation of fixture stiffness and the tilting of the support plane in a timely manner;

[0191] 3) Temporary shutdowns for corrections led to passive adjustments to production plans, resulting in high scrap and rework rates.

[0192] To this end, the company deployed the IoT-based fixture health status assessment method and system of the present invention on the fixture to perform online health monitoring and lifespan management of the fixture.

[0193] 2. Sensor Placement and System Deployment

[0194] On the dedicated fixture for the transmission housing, the sensors are arranged and the system is deployed according to step S1 of the present invention:

[0195] 1) Select three spatially non-collinear points on the shell positioning plane. , , Inductive displacement sensors are installed at the three points, which roughly form an equilateral triangle, with the measurement direction perpendicular to the positioning plane.

[0196] 2) Install a piezoelectric accelerometer on the side plate of the fixture frame near the hydraulic cylinder to collect vibration response;

[0197] 3) Install the data acquisition node and connect it to the IoT edge gateway via industrial Ethernet;

[0198] 4) Establish communication between the machine tool control system and the machine tool control interface module of this invention, enabling it to:

[0199] Perform the "dry clamping" action when there is no workpiece.

[0200] The clamping force is controlled according to a preset program for loading and unloading.

[0201] Control the spindle in and Step excitation is performed between intervals.

[0202] After system deployment is complete, and after fixture replacement or overhaul, the initial health baseline construction is performed first, selecting continuous... One shift, total The empty clamping test data of the workpieces were used to confirm that the workpiece dimensions were all qualified and the fluctuations were small, and the digital twin parameter vectors in the healthy state were identified from this data. and initial health index .

[0203] 3. Periodic detection and parameter identification process

[0204] In actual production, enterprises set a target of approximately [amount missing] per unit of production. One empty clamping health check is performed every clamping cycle (approximately 5 shifts). The following describes three representative check cycles.

[0205] 3.1 First test (approximately one week after commissioning)

[0206] Total number of clamping cycles: ;

[0207] According to the present invention, S1 is used for empty clamping excitation, and the three-point displacement response and vibration response are collected;

[0208] The change in the tilt angle of the supporting plane was obtained by plane fitting calculation. All less than ;

[0209] The collected data was input into a simplified finite element digital twin model, and S2 online parameter identification was performed to obtain:

[0210] Contact stiffness Change relative to initial value ;

[0211] Equivalent Young's Modulus of the Frame Basically unchanged;

[0212] Support plane tilt angle The difference from the initial value is negligible.

[0213] In step S3, the maximum deformation of the workpiece is simulated under standard workpiece, nominal clamping force, and typical cutting load boundary conditions:

[0214] ;

[0215] Due to the tolerance of this process, the deformation percentage is:

[0216] r1=Δ max,1 / Δ allow =0.012 / 0.03=0.4;

[0217] Corresponding health index Approximately (Within the healthy range, only slightly below the initial state).

[0218] The test results indicate that the fixture is still in good condition and requires no additional maintenance or process adjustments.

[0219] 3.2 Second test (approximately 2 months after use)

[0220] Total number of clamping cycles: ;

[0221] Execute the same empty clamping excitation sequence, collect data and perform plane fitting to obtain the support plane around the [missing information]. The change in axis tilt angle is approximately , around The change in axis tilt angle is approximately ;

[0222] Digital twin parameter identification results:

[0223] 1) Contact stiffness The value decreased by approximately relative to the initial value. ;

[0224] 2) Frame stiffness Almost unchanged;

[0225] 3) Inclination angle of the supporting plane , There is a noticeable deviation.

[0226] Under standard operating conditions, the maximum deformation of the workpiece was obtained through simulation:

[0227] ;

[0228] Deformation occupancy ratio:

[0229] ;

[0230] According to the preset The mapping function, at which point the health index is approximately It is already significantly lower than its initial state. The degradation modeling and lifetime prediction module will... and The degradation sequence was incorporated, and an exponential degradation model was initially fitted.

[0231] During this phase, the quality department's statistics revealed an increase in the proportion of critical planar dimensions of the transmission housing approaching the upper tolerance limit, but actual deviations remained relatively few. Through the health index indicators of this invention, maintenance engineers can now predict the continued degradation of the fixture's condition.

[0232] 3.3 Third test (approximately 4 months of use)

[0233] Total number of clamping cycles: ;

[0234] The results of the empty clamping test show that the support plane is around axis, The changes in axis tilt angle respectively reached , ;

[0235] The digital twin parameter vector further degenerates:

[0236] The contact stiffness decreases by approximately [percentage missing] relative to the initial value. ;

[0237] The significantly increased tilt angle of the support plane indicates a marked change in the contact state between the fixture and the machine tool table.

[0238] The maximum deformation of the workpiece was obtained through simulation under standard working conditions:

[0239] ;

[0240] Deformation occupancy ratio:

[0241] ;

[0242] pass The mapping yields a health index:

[0243] ;

[0244] At this time, the workpiece inspection data also showed that the flatness of a small number of transmission housings had approached or even slightly exceeded the upper limit of tolerance, resulting in small-batch rework and scrap.

[0245] 4. The effectiveness of degradation modeling and RUL prediction

[0246] by , , The degradation sequence is constructed from subsequent detection points, and the degradation modeling and lifetime prediction module meets the following requirements: Fitting an exponentially degenerate model under the constraint of monotonically non-increasing:

[0247] ;

[0248] Furthermore, a penalty term related to the decrease in contact stiffness and the increase in the tilt angle of the supporting plane is introduced into the objective function, so that the fitted degradation curve is consistent with the physical degradation trend (decreased stiffness and increased tilt angle).

[0249] Set failure threshold (Experience shows that this value corresponds to the proportion of deformation occupied) (Approximately close to the tolerance limit), please solve:

[0250] ;

[0251] At the third testing moment The model predicts approximately the number of failure cycles. ,therefore:

[0252] ;

[0253] This means that under current usage conditions, the clamp can still be used safely for approximately [time period missing]. After one clamping cycle (approximately 15-20 shifts), exceeding this range significantly increases the risk of deviations in the transmission housing.

[0254] Through this invention, enterprises can, before a large number of batches of fixtures actually exhibit out-of-tolerance behavior, [according to...]. Develop maintenance plans in advance, such as scheduling a planned downtime window in the coming week to calibrate or replace fixtures, to avoid the passive situation of "temporary downtime + mass scrapping".

[0255] 5. Process compensation effect based on health index

[0256] After the third test results came out, the health index Within the warning range, the enterprise, based on the process compensation scheme function of this invention, used a digital twin model to simulate various combinations of process parameters, including:

[0257] 1) Adjust the clamping force from the nominal value Reduce to ;

[0258] 2) Slightly adjust the position of the workpiece support points to make the support area more evenly distributed in the deformation-sensitive area;

[0259] 3) Add an auxiliary support block on one side to offset some of the additional deformation caused by tilting.

[0260] Simulation results show that, without increasing machining time or changing cutting parameters, the maximum workpiece deformation is reduced to: (10% reduction in clamping force + adjustment of support point position + addition of a single-sided auxiliary support block) after adopting the combined scheme of "reducing clamping force by 10% + adjusting support point position + adding a single-sided auxiliary support block".

[0261] ;

[0262] The deformation occupancy ratio becomes:

[0263] ;

[0264] The corresponding health index is from Temporarily upgraded to:

[0265] ;

[0266] Subsequently, this process compensation scheme was applied in actual production, and the results of critical flatness tests on the transmission housing over five consecutive shifts showed that:

[0267] 1) The proportion of out-of-tolerance parts was reduced from approximately [previous amount] before compensation. Reduce to ;

[0268] 2) The average flatness of the workpiece is close to the median tolerance, and the distribution is more concentrated;

[0269] 3) No temporary shutdowns occurred due to fixture problems, and the production plan remained stable.

[0270] This result verifies that the "process compensation strategy based on health index and digital twin simulation" proposed in this invention can effectively delay quality risks before the fixture is replaced, and has a significant effect on reducing scrap rate and avoiding sudden downtime.

[0271] 6. Comparative Analysis with Traditional Solutions

[0272] To compare the technical effects of this invention, the company statistically analyzed key indicators before and after implementing this invention (taking a 4-month operating cycle as an example):

[0273] 1) Scrap rate and repair rate

[0274] Before implementing this invention: the scrap rate related to the critical flatness of the transmission housing was approximately %, with a return rate of approximately %. ;

[0275] After implementing this invention: the scrap rate was reduced to Around [time], the return rate dropped to [percentage]. %about.

[0276] 2) Scheduled Fixture Maintenance vs. Unexpected Downtime

[0277] Traditional maintenance practices show that, on average, there is an unplanned downtime of 4 to 6 hours due to fixture problems every 3 to 4 months.

[0278] After the introduction of this invention: fixture maintenance is completely transformed into planned downtime, through... Advance planning and forecasting prevented any sudden downtime due to fixture failure within four months.

[0279] 3) Quality stability

[0280] Before implementing this invention, the flatness test results of the workpiece showed a clear tendency to converge near the upper tolerance.

[0281] After implementing this invention, the product's geometric accuracy distribution is more concentrated in the middle range of the tolerance through the health index and process compensation scheme.

[0282] The foregoing description of embodiments of the present invention, through which those skilled in the art are able to implement or use the present invention, will be readily apparent to those skilled in the art. Various modifications to these embodiments will be readily apparent to those skilled in the art. The general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novelty disclosed herein.

Claims

1. A method for assessing the health status of a fixture based on the Internet of Things, characterized in that, Includes the following steps: S1. Arrange at least three displacement sensors on the fixture positioning plane and a vibration sensor on the fixture body. When the fixture is in an empty clamping state, apply a preset clamping force for loading and unloading and / or spindle speed step excitation through the machine tool control system to collect multi-point displacement and vibration response. S2. Based on the finite element simplified digital twin model of the clamp, online parameter identification is performed on the response to obtain a digital twin parameter vector consisting of contact stiffness, clamp frame stiffness and support plane tilt angle. S3. Under standard workpiece, nominal clamping force, and typical cutting load boundary conditions, the maximum deformation Δ of the workpiece is obtained by simulating using a digital twin model. max and with the workpiece's allowable geometric tolerance Δ allow The ratio r = Δ max / Δ allow Construct the fixture health index HI = f(r), where f is a monotonically decreasing function; S4. Based on the degradation sequence of HI as a function of clamping cycle number obtained from multiple detections, a degradation model is established under the physical constraints that HI does not increase monotonically and is correlated with the decrease in contact stiffness and the increase in the inclination angle of the support plane. The remaining number of clamping cycles when HI drops to a preset threshold is predicted as the remaining usable life (RUL) of the fixture. HI and RUL are output through the Internet of Things platform for maintenance and process compensation decisions. The construction of the health index HI in step S3 includes: first calculating the deformation occupancy ratio r = Δ max / Δ allow When r is below the first preset threshold, HI is set to a healthy range value close to 1. When r is above the second preset threshold, HI is set to a failure range value close to 0. An exponential or piecewise linear mapping function f(r) is used to generate a continuously changing health index HI between the first preset threshold and the second preset threshold.

2. The method according to claim 1, characterized in that, In S1, at least three displacement sensors are arranged at three non-collinear points on the fixture positioning plane. By performing planar fitting on the displacement increments of the three points, the overall rigid translation of the fixture positioning plane and the change of tilt angle around two mutually perpendicular axes are obtained. The tilt angle of the support plane is calculated from the change of tilt angle.

3. The method according to claim 1, characterized in that, The finite element simplified digital twin model of the fixture includes: a finite element simplified model with the fixture frame as an elastic body element, the contact area between the fixture and the machine tool table as a nonlinear spring element, and the clamping element as the point of concentrated force application. The online parameter identification only solves for the contact stiffness, the equivalent Young's modulus of the frame, and the equivalent offset of the support point, so as to reduce the amount of calculation and maintain correspondence with the sensor arrangement.

4. The method according to claim 1, characterized in that, The degradation model established in step S4 includes: smoothing and outlier removal of the discrete sequence of HI with the number of clamping cycles; fitting the degradation curve with constrained nonlinear regression or time series prediction algorithm under the constraint that HI(k+1)≤HI(k); and adding a penalty term to the objective function that is consistent with the direction of HI change for the contact stiffness attenuation rate and the change rate of the support plane tilt angle, so that the degradation model conforms to the actual mechanical damage evolution law of the fixture.

5. The method according to claim 1, characterized in that, Also includes: When the health index HI is lower than the warning threshold but higher than the failure threshold, the digital twin model is invoked to simulate the maximum deformation of the workpiece under different clamping force levels, workpiece support point layouts, or the addition of auxiliary support blocks. The corresponding deformation occupancy ratio r is compared, and the process compensation scheme that can keep r within the allowable range and has the least impact on the production cycle without replacing the fixture is selected. The process compensation scheme and the fixture maintenance suggestions are pushed to the manufacturing execution system through the Internet of Things platform. And / or, when constructing the initial health benchmark of the digital twin parameter vector in step S2, select the empty clamping detection data of multiple processing batches with all the key dimensions and geometric tolerances of the corresponding workpiece being qualified and the fluctuation range being less than the preset value, and require that the main mode frequency offset of the vibration spectrum in the data does not exceed the predetermined threshold, so as to ensure that the initial digital twin parameter vector corresponds to the fixture health state.

6. A fixture health status assessment system based on the Internet of Things, characterized in that, This method is used to implement the method of any one of claims 1-5, comprising: A fixture measuring device is used to arrange at least three displacement sensors on the fixture positioning plane and a vibration sensor on the fixture body, and to configure a data acquisition node electrically connected to the sensors. The machine tool control interface module is used to send preset clamping force loading-unloading and / or spindle speed step excitation commands to the machine tool control system when the fixture is in an unclamped state. The IoT edge gateway is used to receive the multi-point displacement response and vibration response uploaded by the acquisition node and forward them to the host server through the Industrial Internet of Things protocol. The digital twin modeling and parameter identification module, deployed on the host server, is used to perform online parameter identification on the response based on the simplified digital twin model of the clamp with finite element, and obtain a digital twin parameter vector composed of contact stiffness, clamp frame stiffness and support plane tilt angle. The health index calculation module is used to call the digital twin model to simulate and obtain the maximum deformation Δ of the workpiece under standard workpiece, nominal clamping force, and typical cutting load boundary conditions. max And according to Δ max With respect to the workpiece's allowable geometric tolerance Δ allow The ratio r is used to construct the fixture health index HI = f(r); The degradation modeling and life prediction module is used to establish a degradation model based on the degradation sequence of the health index HI obtained from multiple empty clamping tests as a function of the number of clamping cycles. Under the physical constraints that HI does not increase monotonically and is correlated with the decrease in contact stiffness and the increase in the inclination angle of the support plane, the module predicts the remaining usable life (RUL) of the fixture. The maintenance and process compensation decision module is used to output the Health Index (HI) and Remaining Usable Life (RUL) through the IoT platform, and generate maintenance work orders and process compensation suggestions.

7. The system according to claim 6, characterized in that, The digital twin modeling and parameter identification module includes: a model simplification unit, used to simplify the three-dimensional structure of the fixture into a finite element model containing elastic frame elements, nonlinear contact spring elements, and clamping concentrated loads; and a parameter solving unit, used to use the least squares method, Bayesian estimation, or other optimization algorithms to fit the three-point displacement and vibration response output by the model simulation to the actual measurement data, thereby solving for the three physical parameter sets of contact stiffness, equivalent frame stiffness, and equivalent offset of support points. And / or, the maintenance and process compensation decision module is further configured to: when the health index HI is lower than the warning threshold and the remaining usable life RUL is less than the preset upper limit, call the health index calculation module and the digital twin modeling and parameter identification module to simulate the maximum deformation of the workpiece under different clamping forces, support point arrangements and auxiliary support block combinations, screen out the process compensation scheme that meets the allowable geometric tolerance of the workpiece and has the least impact on the production cycle, and push the scheme together with the maintenance plan for fixture correction, realignment or replacement to the manufacturing execution system and the maintenance management system.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, it implements the method of any one of claims 1–5.

9. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the method of any one of claims 1–5.

Images