Micro-vibration risk prediction and scheduling management system for laser precision machining

Abstract

The application relates to the technical field of scheduling management, and discloses a micro-vibration risk prediction scheduling management system for laser precision machining, which comprises the following steps: firstly, a known scheduling table of heavy equipment and a target work order precision grade are acquired, a scheduling state and an exposure time length are generated; secondly, an allowable deviation limited by a deformation allowance and a vibration weight are obtained, a decay function is constructed and is associated with the scheduling state in a time domain, residual vibration is obtained to synthesize a source end speed; then, a machine displacement is synthesized by combining a propagation delay and a path attenuation; the machine displacement is locally exposed and weightedly integrated at a candidate time by using the vibration weight, a prediction deviation is calculated, and a risk coefficient reflecting allowance consumption is obtained; finally, feasible starting time is extracted in a resource available interval, a scheduling instruction is issued, and an audit record is generated. The application realizes automatic coupling of equipment scheduling and dynamic interference avoidance.

Description

Technical Field

[0001] This invention relates to the field of scheduling and management technology, and more specifically, to a micro-vibration risk prediction and scheduling management system for laser precision machining. Background Technology

[0002] In the daily operation of modern high-precision laser processing workshops, the cutting action of laser processing does not rely on a solid tool with mechanical rigidity, but rather on the highly concentrated energy distribution of the laser beam within space. When the workpiece is disturbed by environmental micro-vibrations caused by the operation of surrounding heavy-duty machinery (such as large punch presses, overhead cranes, etc.), the relative position of the focal spot and the processed surface will undergo uncontrollable instantaneous shift, directly causing the processing boundaries of micro-holes and narrow slits to exceed the process tolerances.

[0003] Current on-site production scheduling generally faces a dilemma: while the Manufacturing Execution System (MES) knows the start and end calendars of the equipment, it is unaware that after heavy-duty equipment stops, damping wakes can remain in the environmental floor and plant components for several minutes. To avoid scrap risks, on-site staff often blindly increase the experience-based buffer time (e.g., forcing the production line to wait silently for ten minutes after a punch press stops), and this crude downtime severely squeezes effective production time. Conversely, without buffers, the defect rate of high-precision work orders will remain high. Existing scheduling software generally lacks quantitative assessment methods for environmental dynamic transmission lags and energy decay patterns; the precision labels on process cards can only serve as priority references for queuing and cannot be truly transformed into avoidance indicators on the scheduling timeline. Summary of the Invention

[0004] This invention provides a micro-vibration risk prediction and scheduling management system for laser precision machining, which solves the technical problems mentioned in the background art.

[0005] This invention provides a micro-vibration risk prediction, scheduling, and management system for laser precision machining. It is applied to a factory scheduling system comprising a data acquisition unit, a data processing unit, and a manufacturing execution system, and is configured to execute:

[0006] The known schedule of heavy equipment is obtained and time-aligned with the accuracy level of the target work order to generate a schedule status that characterizes the dynamic interference cycle of the equipment, as well as the exposure duration that characterizes the sensitive time window of the processing disturbance.

[0007] Based on the aforementioned accuracy level, the allowable deviation for the limited machining deformation allowance is determined, along with the vibration weight used to filter out invalid interferences other than the structure's natural frequency.

[0008] A decay function characterizing the energy dissipation of the environmental medium is constructed, and the scheduling state is correlated with the decay function in the time domain to obtain the residual vibration containing the damped tail wave, thereby synthesizing the source end velocity reflecting the source excitation intensity.

[0009] By combining the propagation delay and path attenuation that characterize spatial transmission hindrance, the source velocity is transmitted and synthesized, and converted into the machine displacement of the physical machine tool facing the processing of the target work order.

[0010] Candidate start times for scheduling simulation are set, and the machine displacement is locally exposed and weighted by the vibration weight at the candidate start times to calculate the prediction deviation characterizing the predicted profile offset. The prediction deviation is then compared and converted with the allowable deviation to obtain the risk coefficient reflecting the margin consumption status.

[0011] Within a continuous time interval where the risk coefficient does not exceed the limit and the resource availability status of the manufacturing execution system is true, extract the extreme nodes that satisfy the scheduling boundary and align them with the system scheduling scale to obtain the start time to avoid micro-vibration interference.

[0012] Based on the stated start time, a dispatch entry instruction is issued, and the risk coefficient is decoupled and decomposed according to the vibration source to form an audit record for tracing the vibration source.

[0013] The beneficial effects of this invention are as follows: By performing cross-domain analysis of the scheduling status of heavy equipment with the local energy distribution and damping attenuation law of the laser beam, it is possible to quantify the predicted contour offset caused by environmental interference without relying on real-time sensing probes or adding fixed empirical buffer times. This invention can automatically output start-up instructions with precise avoidance times in the manufacturing execution system based on the actual accuracy requirements of the work order. This not only effectively reduces the residual vibration impact of complex excitation sources in the factory area on high-precision machining, but also avoids capacity waste caused by overly conservative scheduling while ensuring resource availability. Attached Figure Description

[0014] Figure 1 This is a flowchart of the micro-vibration risk prediction and scheduling management system for laser precision machining according to the present invention. Detailed Implementation

[0015] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, features described in some examples may be combined in other examples.

[0016] like Figure 1 As shown, the micro-vibration risk prediction and scheduling management system for laser precision machining is applied to a factory scheduling system that includes a data acquisition unit, a data processing unit, and a manufacturing execution system, and is configured to execute:

[0017] The known schedule of heavy equipment is obtained and time-aligned with the accuracy level of the target work order to generate a schedule status that characterizes the dynamic interference cycle of the equipment, as well as the exposure duration that characterizes the sensitive time window of the processing disturbance.

[0018] Based on the aforementioned accuracy level, the allowable deviation for the limited machining deformation allowance is determined, along with the vibration weight used to filter out invalid interferences other than the structure's natural frequency.

[0019] A decay function characterizing the energy dissipation of the environmental medium is constructed, and the scheduling state is correlated with the decay function in the time domain to obtain the residual vibration containing the damped tail wave, thereby synthesizing the source end velocity reflecting the source excitation intensity.

[0020] By combining the propagation delay and path attenuation that characterize spatial transmission hindrance, the source velocity is transmitted and synthesized, and converted into the machine displacement of the physical machine tool facing the processing of the target work order.

[0021] Candidate start times for scheduling simulation are set, and the machine displacement is locally exposed and weighted by the vibration weight at the candidate start times to calculate the prediction deviation characterizing the predicted profile offset. The prediction deviation is then compared and converted with the allowable deviation to obtain the risk coefficient reflecting the margin consumption status.

[0022] Within a continuous time interval where the risk coefficient does not exceed the limit and the resource availability status of the manufacturing execution system is true, extract the extreme nodes that satisfy the scheduling boundary and align them with the system scheduling scale to obtain the start time to avoid micro-vibration interference.

[0023] Based on the stated start time, a dispatch entry instruction is issued, and the risk coefficient is decoupled and decomposed according to the vibration source to form an audit record for tracing the vibration source.

[0024] The micro-vibration risk prediction and scheduling management system for laser precision machining provided in this embodiment is applied to a factory scheduling system that includes a data acquisition unit, a data processing unit, and a manufacturing execution system. The data acquisition unit collects heavy equipment scheduling data, work order process data, factory environmental parameters, and vibration test data; the data processing unit executes algorithms for vibration prediction, deviation calculation, scheduling simulation, and audit record generation; and the manufacturing execution system provides resource status information, issues scheduling instructions, and stores production process data.

[0025] S201 eliminates system time differences by basing the local time of the manufacturing execution system on midnight of the current day, thus deriving an absolute time that represents a unified clock reference across the entire plant. The calculation formula is as follows:

[0026]

[0027] in, For absolute time, To generate the local time of the execution system, The system time difference is set to midnight of the day. The NTP network time protocol is used to synchronize the clocks of all equipment and systems in the plant, with a synchronization error of no more than 1 millisecond, to avoid misalignment between the scheduling status and vibration prediction time caused by the clock asynchrony of different equipment.

[0028] S202 extracts step change characteristics from the start-stop boundaries of multiple shifts for each heavy equipment, and combines them to generate a shift status to define the active time interval of the excitation source. The calculation formula is as follows:

[0029]

[0030] in, For the first The scheduling status of heavy equipment. This is a unit step function that outputs 1 when the input is greater than or equal to 0, and 0 when the input is less than 0. For time variables, Index for heavy equipment This represents the total number of heavy equipment. For the scheduling segment index, For the first Total number of shifts for each heavy equipment unit and The first The scheduling system defines start and stop boundaries for shift work. The manufacturing execution system's schedule uses a standard data structure, including fields for equipment number, shift segment number, start time, stop time, and shift type. The time format follows the ISO 8601 standard. For cross-day scheduling, shift segments with start times on the current day and stop times on the next day are split into current-day and next-day segments, and recorded in the corresponding date's scheduling status. Temporary orders and equipment maintenance periods are treated as special shift segments; the scheduling status for maintenance periods is set to 0. The scheduling status is a function of time; a value of 1 indicates that the corresponding heavy equipment is in an operational state, and a value of 0 indicates that the equipment is in a stopped state.

[0031] The Manufacturing Execution System's (MES) scheduling system employs a real-time synchronization mechanism with a synchronization cycle of one minute. When a temporary scheduling change occurs, the system immediately triggers the synchronization operation. Scheduling changes include three categories: adjustments to equipment start / stop times, temporary order insertions, and changes to equipment maintenance plans. Change information is pushed to this system via the MES event notification interface. Upon receiving the change information, the system automatically recalculates the scheduling status of the corresponding heavy equipment and updates the risk coefficients and start times for all subsequent work orders.

[0032] S203 obtains the total processing length of the target work order, calculates the processing path runtime based on the scanning speed, and combines the auxiliary pause time required for clamping and curing to obtain the exposure time used to define the anti-interference zone. The calculation formula is:

[0033]

[0034] in, For the target work order index, For exposure duration, The total processing length of the target work order is the total length of the laser scanning path. The scanning speed in laser processing refers to the speed at which the laser head moves relative to the workpiece surface. The auxiliary stoppage period includes five stages that must be included in the interference prevention zone: workpiece clamping, positioning, curing, laser preheating, and optical path calibration. The workpiece cooling period is not included in the interference prevention zone. The auxiliary stoppage period is obtained from the process document corresponding to the work order. Each process document predefines the standard auxiliary time for the corresponding processing type. If it is not specified in the process document, the historical average auxiliary time of the same type of work order over the past three months is used. The exposure time is the total time interval during the entire processing that needs to avoid micro-vibration interference.

[0035] Typical auxiliary downtime values ​​for different machining types are as follows: 30 to 60 seconds for micro-hole machining, 60 to 120 seconds for narrow-slit machining, and 120 to 300 seconds for curved surface machining. For special processes, such as ultra-precision machining or large-size workpiece machining, the auxiliary downtime can be defined separately in the process document, and the defined value takes precedence over the standard auxiliary time. When the auxiliary time for a special process is not specified in the process document, the historical average auxiliary time of similar special work orders over the past month is used.

[0036] S301 establishes a correlation model including peak energy and spatial attenuation distribution factor, constructing a local energy characterizing the spatial distribution density of beam energy. The calculation formula is:

[0037]

[0038] in, Local energy, i.e., radial distance from the center of the beam. Energy density at that location This refers to the spatial relative coordinate distance, specifically the radial distance from a point on the machining surface to the center of the beam. Peak energy, i.e., the maximum energy density at the center of the beam. The radius represents the spatial distribution factor and corresponds to the beam waist radius of the Gaussian beam. This model is built upon the energy distribution characteristics of the fundamental mode Gaussian beam and is applicable to most industrial laser processing scenarios.

[0039] S302 derives the boundary radius characterizing the actual area of ​​material removal based on the geometric span at which the local energy extraction energy density decays to the material ablation threshold. The calculation formula is as follows:

[0040]

[0041] in, Let the boundary radius be , The ablation threshold is the minimum energy density required to remove material from a workpiece. It can be obtained in two ways: first, by referencing standard ablation threshold data from the NIST Chemistry WebBook; and second, through testing with standard samples. The testing method involves using the same laser parameters as the target work order to perform single-point ablation on a standard sample, measuring the diameter of the ablation pit, and then using the Gaussian beam energy distribution formula to deduce the ablation threshold. Typical reference ranges for common processed materials are: 1.2 to 2.5 joules per square centimeter for stainless steel, 0.3 to 0.8 joules per square centimeter for monocrystalline silicon, and 3.0 to 5.0 joules per square centimeter for alumina ceramics. When the local energy equals the ablation threshold, the corresponding radial distance is the actual boundary radius of the laser processing. This method avoids the processing boundary definition errors caused by non-physical tools and accurately determines the effective range of laser processing.

[0042] Material ablation threshold testing must be conducted under standard environmental conditions, with the ambient temperature controlled between 20 and 25 degrees Celsius and the relative humidity between 40% and 60%. Before testing, the surface of the standard sample must be cleaned to remove oil and oxide layers. When the actual processing environment deviates from the standard environment by more than the above range, the material ablation threshold needs to be corrected. The correction factor is obtained through linear interpolation of the ambient temperature and humidity.

[0043] S303 establishes a mapping mechanism between accuracy levels and process standards to obtain the allowable deviation of the corresponding process standards. The calculation formula is:

[0044]

[0045] in, The allowable deviation is the maximum permissible form and position deviation of the machined contour. For mapping functions, This refers to the accuracy level of the target work order. The accuracy level adopts the standard tolerance levels specified in GB / T1800.1-2009, coded as IT01 to IT18. The mapping function is constructed by establishing a mapping table from accuracy level to permissible deviation for different machining types. The permissible deviation for micro-hole machining corresponds to the diameter tolerance of the standard tolerance level, the permissible deviation for narrow slit machining corresponds to the width tolerance of the standard tolerance level, and the permissible deviation for curved surface machining corresponds to the surface profile tolerance of the standard tolerance level. The mapping table is pre-stored in the system database and can be updated and maintained according to process standards.

[0046] For non-standard accuracy levels, linear interpolation is used for mapping. First, the upper and lower standard tolerance levels corresponding to the non-standard accuracy level are determined. Then, the corresponding allowable deviation is calculated based on the ratio of the difference between the non-standard accuracy level and the two standard levels. When the non-standard accuracy level is higher than the highest standard level IT01, the allowable deviation is 0.5 times the allowable deviation of IT01 level; when the non-standard accuracy level is lower than the lowest standard level IT18, the allowable deviation is twice the allowable deviation of IT18 level.

[0047] S304 uses the boundary radius and scanning speed to define the local duration of the focal spot's action over the processing surface. The calculation formula is:

[0048]

[0049] in, This refers to the localized duration of the laser beam, specifically the time it takes for the laser spot to pass over a fixed point on the workpiece surface. The effective diameter of the focal spot is approximately... By combining the scanning speed, the dwell time of the focal spot at a single point can be calculated, and this duration determines the degree of influence of vibrations at different frequencies on the processing accuracy.

[0050] S305 extracts frequency domain attenuation characteristics based on the local action duration and the preset center frequencies of each evaluation frequency band, filters out environmental noise in non-resonant frequency bands, and derives the vibration weight for each evaluation frequency band. The calculation formula is as follows:

[0051]

[0052] in, For the first Vibration weights for each evaluation frequency band. For frequency band indexing, To assess the total number of frequency bands, For the first The evaluation frequency bands are determined based on the inherent frequencies of the plant structure and the response characteristics of laser processing. They are typically divided into 5 to 10 bands, covering a frequency range from 1 Hz to 1000 Hz. The low-frequency band (1 to 10 Hz) corresponds to the overall structural vibration of the plant; the mid-frequency band (10 to 100 Hz) corresponds to the vibration of the equipment foundation; and the high-frequency band (100 to 1000 Hz) corresponds to the vibration of the laser machine itself. A recommended frequency band configuration for a typical plant area is: 1 to 5 Hz, 5 to 10 Hz, 10 to 20 Hz, 20 to 50 Hz, 50 to 100 Hz, 100 to 200 Hz, 200 to 500 Hz, and 500 to 1000 Hz, for a total of 8 bands. The number and range of evaluation frequency bands can be adjusted based on the actual structural test results of the plant area. This formula, in the form of a Singer function, characterizes the cumulative influence of vibrations at different frequencies over a given local duration. When the product of the vibration frequency and the local duration is an integer, the vibration effects cancel each other out, and the weight approaches zero. When the product is a half-integer, the weight reaches its maximum value. Vibration weights can filter out non-resonant frequency noise that has a minor impact on machining accuracy, retaining only the frequency bands with significant effects.

[0053] The triggering conditions for adjusting the evaluation frequency band include the completion of factory structure renovation, the addition of heavy equipment, and the replacement of the vibration isolation system on the laser processing machine. The adjustment method is to conduct a comprehensive vibration test on the factory area, collect 24-hour environmental vibration data, perform spectral analysis on the vibration data to determine the frequency range of the main vibration energy distribution, and then re-divide the evaluation frequency band according to the energy distribution to ensure that each frequency band contains at least one main vibration peak.

[0054] S401 obtains the attenuation constants characterizing the system damping dissipation rate for each heavy piece of equipment in each evaluation frequency band. Based on these attenuation constants, a time-dependent exponential attenuation model is constructed, forming an attenuation function characterizing the structural damping properties. The calculation formula is as follows:

[0055]

[0056] in, For the first The heavy equipment in the first The attenuation constant of the evaluation frequency band, The attenuation function is used. The attenuation constant is obtained through on-site vibration testing. The testing instrument uses a piezoelectric accelerometer with a sensitivity of 100 mV / g and a sampling frequency no less than 10 times the center frequency of the highest evaluation frequency band. Test points are set up on the ground surface at 1m, 2m, 5m, and 10m around the foundation of the heavy equipment. Sensors in three directions are placed at each test point to measure vibrations in the horizontal X, horizontal Y, and vertical Z directions, respectively. The data processing flow is as follows: the vibration signal after the equipment is shut down is bandpass filtered to obtain the vibration signal for each evaluation frequency band. Then, the signal amplitude of each frequency band is subjected to exponential fitting, and the time constant obtained from the fitting is the attenuation constant for that frequency band. The typical range of the attenuation constant is 0.1s to 10s, with the attenuation constant in the high-frequency band being smaller than that in the low-frequency band. The exponential attenuation model can accurately describe the attenuation process of vibration energy caused by structural damping over time.

[0057] During the attenuation constant test, heavy equipment must be operated under rated load for at least 30 minutes continuously before being suddenly stopped. Vibration signals are collected within 60 seconds after shutdown. Outlier data is removed using the 3σ criterion, marking data points exceeding three times the standard deviation of the mean as outliers. The remaining data is then subjected to exponential fitting. If more than 10% of the data points are removed, the test must be repeated.

[0058] S402 fuses the scheduling status and attenuation function through time-domain convolution to obtain the residual vibration characterizing the structural inertial excitation that continues after the equipment stops. The calculation formula is as follows:

[0059]

[0060] in, For residual vibration, For the integration time variable, time-domain convolution combines the operating state of heavy equipment with the damping characteristics of the structure, enabling accurate calculation of the damping tailwaves that persist after equipment shutdown. This vibration is a source of interference that is easily overlooked in traditional scheduling methods. For discrete-time systems, convolution can be implemented using the Fast Fourier Transform to improve computational efficiency.

[0061] S403 extracts the reference velocity representing the inherent excitation intensity of heavy equipment at a reference location, and combines it with residual vibration for time-domain modulation to obtain the source-end velocity after removing static interference and dynamically superimposing the scheduling state. The calculation formula is:

[0062]

[0063] in, For reference speed, i.e., the first The effective value of vibration velocity measured at a fixed reference position of a heavy piece of equipment under normal operating conditions. The reference velocity is the velocity at the source of vibration generated by heavy equipment at different times. The reference location is chosen on the ground surface 0.5 meters from the edge of the heavy equipment foundation, avoiding the equipment's anchor bolts and areas of local resonance. Test conditions include three scenarios: no-load operation, rated load operation, and maximum load operation. The effective value of the vibration velocity is measured under each of the three scenarios, and the effective value of the vibration velocity under rated load operation is taken as the reference velocity. The typical range of the reference velocity is 0.1 mm / s to 10 mm / s; the reference velocity for large punch presses is greater than that for overhead cranes.

[0064] When heavy equipment operates in multiple modes, such as single-stroke and continuous-stroke modes for a punch press, or no-load and heavy-load modes for an overhead crane, the reference speed for each mode must be measured separately. The system automatically selects the corresponding reference speed for source-end speed calculation based on the equipment's operating mode in the schedule. When the operating mode is not specified in the schedule, the reference speed of the most commonly used operating mode is used.

[0065] S501 extracts the time hysteresis characteristics, representing the spatial transmission lag of vibration, based on the physical distance between the heavy equipment and the physical platform and the wave velocity of the environmental medium. The calculation formula is as follows:

[0066]

[0067] in, To delay the spread, For the first The straight-line physical distance from a heavy piece of equipment to a laser processing physical machine. The wave velocity refers to the Rayleigh wave velocity, which is the velocity of vibration propagating within the factory floor structure. Typical Rayleigh wave velocities for different flooring materials are: 2000-3000 m / s for reinforced concrete floors, 1800-2500 m / s for epoxy floors, and 3500-4500 m / s for granite floors. The field testing method uses the falling weight method, where a weight is dropped at distances of 1, 2, and 3 meters from the sensor, and the time difference between the vibration reaching the sensor is measured to calculate the wave velocity. The propagation delay is the time required for the vibration to travel from the vibration source to the processing machine.

[0068] The correction factor for the effect of different floor thicknesses on Rayleigh wave velocity is as follows: 0.8 when the floor thickness is less than 0.2 meters; 1.0 when the floor thickness is between 0.2 and 0.5 meters; and 1.2 when the floor thickness is greater than 0.5 meters. The corrected environmental medium wave velocity is the typical value multiplied by the correction factor.

[0069] S502 constructs a spatial energy dissipation model based on the distance attenuation coefficient and physical distance to obtain the path attenuation characterizing the medium absorption effect. The calculation formula is:

[0070]

[0071] in, For path decay, For the first The range attenuation coefficient for each evaluation frequency band. The empirical formula for the range attenuation coefficient is: ,in Let be the center frequency of the b-th evaluation frequency band, expressed in Hertz. The field calibration method involves placing sensors at different distances from the heavy equipment to measure the vibration amplitude of each frequency band, and obtaining the distance attenuation coefficient through exponential fitting. This model describes the energy attenuation caused by medium absorption during vibration propagation in space.

[0072] The correction factors for distance attenuation coefficients for different media are as follows: 1.0 for reinforced concrete flooring, 1.2 for epoxy flooring, and 0.8 for granite flooring. The corrected distance attenuation coefficient is calculated by multiplying the empirical formula value by the correction factor.

[0073] S503 extracts the isolation ratio, which characterizes the frequency transmission properties of the vibration isolation facility. Combining this with path attenuation, reference velocity, and residual vibration corrected for propagation delay, the machine speed, reflecting the intensity of the disturbance, is derived using the root-mean-square energy synthesis rule. The calculation formula is:

[0074]

[0075] in, For the first The machine speed in each evaluation frequency band, The vibration isolation ratio is the ratio of the vibration isolation system of the machine tool to the vibration isolation ratio of the first... The transmissibility of vibration in the evaluation frequency band is considered. The isolation ratio can be obtained from the frequency response curve provided by the vibration isolator manufacturer or through on-site vibration isolation effect testing. The on-site testing method involves placing sensors on the foundation and machine platform of the vibration isolation system, respectively, and measuring the vibration amplitude at the same frequency. The ratio of the amplitude on the machine platform to the amplitude on the foundation is the isolation ratio. For multi-order vibration isolation systems, the total isolation ratio is the product of the isolation ratios of each order. The typical range of the isolation ratio is 0.01 to 10. At the natural frequency of the vibration isolation system, an isolation ratio greater than 1 will amplify the vibration. When there are two or more heavy machines of the same model and operating speed, the vibrations will coherently superimpose, requiring correction to the root mean square (RMS) synthesis method. The correction method involves calculating the total vibration velocity of the coherent sources for a coherent source group, and then performing a RMS synthesis with the vibration velocities of other incoherent sources. The total vibration velocity of the coherent vibration sources is the algebraic sum of the vibration velocities of each coherent vibration source. The coherence coefficient of equipment of the same model and rotation speed is taken as 0.8, and the coherence coefficient of equipment of different models is taken as 0.

[0076] The method for identifying coherent vibration sources is as follows: Operating status signals and vibration signals of all heavy equipment are collected, and the cross-correlation coefficients between the vibration signals of different equipment are calculated. When the cross-correlation coefficient is greater than 0.7, it is determined to be a coherent vibration source group. The system automatically maintains a list of coherent vibration source groups, and when the operating status of the equipment changes, the cross-correlation coefficients are recalculated and the list is updated.

[0077] The aging correction factor for the vibration isolation system is determined based on its service life: 1.0 for less than 1 year; 1.1 for 1 to 3 years; 1.3 for 3 to 5 years; and 1.5 for more than 5 years. The corrected vibration isolation ratio is the original vibration isolation ratio multiplied by the correction factor. The system undergoes on-site testing every six months to update the vibration isolation ratio and aging correction factor.

[0078] S504 combines machine speed and vibration weights, and removes the frequency dimension through angular frequency conversion to obtain the machine displacement including exposure time gating filtering. The calculation formula is:

[0079]

[0080] in, For the first The machine tool displacement is evaluated within a specific frequency band. The displacement and velocity of simple harmonic motion satisfy a basic physical relationship. By combining the vibration weights, the machine tool displacement after frequency filtering can be obtained. This displacement is the direct cause of the laser processing contour offset.

[0081] S601 extracts the instantaneous intensity of the target work order at different processing stages and performs maximum value normalization to obtain the exposure weight characterizing the difference in sensitivity at each processing stage. The calculation formula is as follows:

[0082]

[0083] in, For exposure weight, The integral relative time within the effective time window is the time offset relative to the candidate start time. For the target work order in relative time The instantaneous intensity at that point, i.e., the laser output power. The normalized baseline maximum intensity refers to the maximum instantaneous intensity during the processing of the target work order. Instantaneous intensity data is obtained from the real-time output of the laser controller, acquired via the OPCUA protocol, with a time resolution of at least 1 millisecond. For variable power processing periods, the curve showing the instantaneous intensity changing over time is stored in the process file, and the power curve in the process file is directly called during processing. Processing periods with higher instantaneous intensity are more sensitive to vibration interference, and therefore have a greater exposure weight.

[0084] The method for compensating for laser power fluctuations involves installing a power sensor at the output end of the laser processing machine to collect the laser output power in real time at a frequency of no less than 100 Hz. The real-time power value is compared with the set power value in the process document to calculate the power fluctuation coefficient. This coefficient is then used to correct the instantaneous intensity. The corrected instantaneous intensity is the set instantaneous intensity multiplied by the power fluctuation coefficient.

[0085] Within the effective time window defining the exposure duration, S602 extracts the machine displacement of each evaluation frequency band shifted to the candidate start time. It then uses the exposure weights to perform weighted integration and averaging of the root mean square energy, yielding the prediction bias characterizing the cumulative deformation effect within a specific candidate time window. The calculation formula is as follows:

[0086]

[0087] in, For candidate start time The corresponding prediction bias, This is the candidate start time. The time axis of the machine tool displacement is shifted. This yields the machine displacement change over the entire exposure time when processing begins at the candidate time. The sum of the squares of the displacements in each frequency band is multiplied by the exposure weight, integrated, and then divided by the integral with the exposure weight to obtain the weighted root mean square displacement, which is the predicted total offset of the processing contour. For discrete-time data, the integration operation is performed using the trapezoidal integral method.

[0088] S603 quantifies the predicted deviation by proportionally matching the allowable deviation, resulting in a risk coefficient that reflects the degree of machining tolerance consumption. The calculation formula is:

[0089]

[0090] in, For candidate start time The corresponding risk coefficient. The risk coefficient represents the proportion of the predicted deviation to the allowable deviation. When the risk coefficient is less than or equal to 1, the predicted deviation is within the allowable range of the process, and the processing accuracy can meet the requirements; when the risk coefficient is greater than 1, the predicted deviation exceeds the allowable range, and the processing will produce defective products.

[0091] S701 extracts the queuing zero point set by the preconditions of the work order, traverses and filters the lower bound of the manufacturing execution system where the resource availability status is true and the risk coefficient does not exceed the limit on the time axis, and extracts the earliest feasible time that meets both environmental immunity and production scheduling requirements. The calculation formula is:

[0092]

[0093] in, The earliest feasible time, Queue zero point refers to the earliest time when the target work order meets all prerequisites, including completion of the previous process, arrival of materials, and completion of equipment initialization. This is a logical indicator function for the availability status of resources. This is a logic AND operator. Resources include laser processing machine, chiller, air compressor, operators, and tooling fixtures. The logic for determining resource availability is as follows: when all resources are available within a given time period... When all resources are idle, the resource availability status is true; otherwise, it is false. The priority rule for multiple resource conflicts is as follows: laser processing machine has the highest priority, followed by chiller and air compressor, and finally operators and tooling fixtures. When resource conflicts exist, the availability of higher-priority resources is prioritized. The traversal step size for candidate times is one-tenth of the fixed division unit of the scheduling granularity. To improve computational efficiency, a binary search optimization algorithm is used. First, the time interval with a risk coefficient less than or equal to 1 is determined, and then the earliest available resource time is found within that interval. The infimum operation is used to find the smallest candidate start time that satisfies all constraints.

[0094] The handling method for resource reservations and temporary occupancy is as follows: the system supports resource reservations up to 72 hours in advance, and reservation information is synchronized to this system through the Manufacturing Execution System (MES). When temporary resource occupancy occurs, such as due to equipment failure repair or emergency order insertion, the MES immediately pushes the temporary occupancy information to this system. When determining resource availability, the system prioritizes reserved and temporarily occupied resources, setting the resource availability status for the corresponding time period to false.

[0095] The method for adjusting the traversal step size for candidate times under different accuracy requirements is as follows: when the accuracy level of the target work order is higher than IT5, the traversal step size is one-twentieth of the fixed division unit of the scheduling granularity; when the accuracy level is between IT5 and IT10, the traversal step size is one-tenth of the fixed division unit of the scheduling granularity; when the accuracy level is lower than IT10, the traversal step size is one-fifth of the fixed division unit of the scheduling granularity.

[0096] S702, based on the fixed-time indexing rules of the manufacturing execution system, performs forced upward discretization and rounding on the earliest feasible time to obtain the start-up time after removing interference from continuous time fragments. The calculation formula is as follows:

[0097]

[0098] in, For the start of construction, This is a fixed unit of measurement for scheduling granularity, i.e., the minimum time interval for scheduling by the manufacturing execution system, typically 1 minute, 5 minutes, or 10 minutes. This is the floor operator. Discretization rounding aligns the earliest feasible moment in continuous time to the system scheduling scale, avoiding the generation of unexecutable time fragments.

[0099] S703 merges the start time with the baseline of midnight of the same day, converting it into a scheduling entry command that includes queuing guidance information and conforms to the system interface communication protocol. The calculation formula is:

[0100]

[0101] in, To dispatch entry commands, This function performs protocol conversion and string concatenation mapping. Scheduling instructions are transmitted using the OPCUA protocol. Standard fields for each instruction include work order number, equipment number, start time, processing parameters, operator number, and priority. The start time is in ISO 8601 standard format, accurate to the second. The exception handling mechanism is as follows: upon receiving a scheduling instruction, the Manufacturing Execution System (MES) returns a confirmation message within 10 seconds. If no confirmation message is received, the system will resend the instruction three times. After three failed resends, an audible and visual alarm is triggered, and the dispatcher is notified. This function converts the absolute start time into a standard instruction format recognizable by the MES and sends it to the shop floor execution layer via the system interface.

[0102] The feedback mechanism for the execution status of scheduling instructions is as follows: after receiving a scheduling instruction, the manufacturing execution system sequentially returns five status feedbacks: instruction received, instruction confirmed, equipment started, processing started, and processing completed. The system monitors the instruction execution status in real time, and triggers an exception handling process if no corresponding status feedback is received within a preset time. The preset time is 10 seconds for instruction received feedback, 30 seconds for instruction confirmed feedback, 60 seconds for equipment started feedback, and 120 seconds for processing started feedback.

[0103] At a predetermined start-up time, S801 extracts vibration weights, isolation ratios, path attenuation, reference velocity, and residual vibration after translation to construct a multi-dimensional energy transfer chain. Normalized energy integration is then performed over the exposure duration to derive individual risks characterizing the absolute contribution of specific equipment scheduling and specific frequency bands to the final deformation error. These risks are then arranged in order to generate a list of individual risks. The calculation formula is as follows:

[0104]

[0105] in, For single-item risks, i.e., the first The heavy equipment in the first The contribution of vibrations generated in each evaluation frequency band to the final processing deviation is determined. By integrating each link in the vibration energy transfer chain, the absolute contribution of each vibration source and each frequency band is quantified. All individual risks are then sorted from largest to smallest to generate a list of individual risks.

[0106] S802 merges and serializes the work order number, accuracy level, allowable deviation, start time, corresponding risk coefficient, individual risk list for tracing interference sources, and schedule identifier for version tamper-proofing to generate audit records that solidify the responsibility traceability chain. The calculation formula is:

[0107]

[0108] in, For audit records, For work order number, This is a list of individual risks extracted by dimensionality reduction based on all individual risks sorted by numerical value. This serves as the schedule identifier. The schedule identifier is generated using the SHA-256 hash algorithm. The raw data range for hash calculation includes the numbers of all heavy equipment, schedule segment numbers, start times, stop times, and schedule types. The hash value is 256 bits long and stored as a hexadecimal string. The verification method is as follows: when it is necessary to verify whether the schedule has been tampered with, the hash value of the current schedule is recalculated and compared with the schedule identifier in the audit record. If they match, it has not been tampered with; otherwise, it has been tampered with. The serialization operation combines all information into structured data according to a fixed format and stores it in the system database. When processing quality problems occur, the source of interference and the responsible link can be quickly located through the audit record.

[0109] Audit logs are stored encrypted using the AES-256 encryption algorithm. Access permissions for audit logs are divided into three levels: Level 1 is for system administrators, who can view all audit logs; Level 2 is for process engineers, who can view audit logs for their specific workshop; and Level 3 is for operators, who can only view audit logs for the work orders they are responsible for. Authentication is required to access audit logs, and all access operations are recorded in the system log.

[0110] The embodiments of this example have been described above. However, this example is not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms based on the guidance of this example, and all of them are within the protection scope of this example.

Claims

1. A micro-vibration risk prediction and scheduling management system for laser precision machining, applied to a factory scheduling system comprising a data acquisition unit, a data processing unit, and a manufacturing execution system, characterized in that... Configured for execution: The known schedule of heavy equipment is obtained and time-aligned with the accuracy level of the target work order to generate a schedule status that characterizes the dynamic interference cycle of the equipment, as well as the exposure duration that characterizes the sensitive time window of the processing disturbance. Based on the aforementioned accuracy level, the allowable deviation for the limited machining deformation allowance is determined, along with the vibration weight used to filter out invalid interferences other than the structure's natural frequency. A decay function characterizing the energy dissipation of the environmental medium is constructed, and the scheduling state is correlated with the decay function in the time domain to obtain the residual vibration containing the damped tail wave, thereby synthesizing the source end velocity reflecting the source excitation intensity. By combining the propagation delay and path attenuation that characterize spatial transmission hindrance, the source velocity is transmitted and synthesized, and converted into the machine displacement of the physical machine tool facing the processing of the target work order. Candidate start times for scheduling simulation are set, and the machine displacement is locally exposed and weighted by the vibration weight at the candidate start times to calculate the prediction deviation characterizing the predicted profile offset. The prediction deviation is then compared and converted with the allowable deviation to obtain the risk coefficient reflecting the margin consumption status. Within a continuous time interval where the risk coefficient does not exceed the limit and the resource availability status of the manufacturing execution system is true, extract the extreme nodes that satisfy the scheduling boundary and align them with the system scheduling scale to obtain the start time to avoid micro-vibration interference. Based on the stated start time, a dispatch entry instruction is issued, and the risk coefficient is decoupled and decomposed according to the vibration source to form an audit record for tracing the vibration source.

2. The micro-vibration risk prediction and scheduling management system for laser precision machining according to claim 1, characterized in that, The system obtains the known work schedule of heavy equipment and aligns it with the accuracy level of the target work order to generate a work schedule status that characterizes the dynamic interference cycle of the equipment, as well as the exposure duration that characterizes the sensitive time window for processing disturbances, including: The local time of the manufacturing execution system is based on the zero point of the day to eliminate system time difference, so as to obtain the absolute time that represents the unified clock reference of the entire plant. For each of the heavy equipment, step change features are extracted based on the start-stop boundaries of its multi-segment schedule, and combined to generate the schedule state used to define the active time interval of the excitation source. The total processing length of the target work order is obtained, the running time of the processing path is calculated based on the scanning speed, and the auxiliary pause time required for clamping and curing is combined to obtain the exposure time used to define the anti-interference zone.

3. The micro-vibration risk prediction and scheduling management system for laser precision machining according to claim 2, characterized in that, Based on the aforementioned accuracy level, the permissible deviation for limiting the machining deformation allowance is derived, along with vibration weights used to filter out invalid interferences other than the structure's natural frequency, including: Establish a correlation model that includes peak energy and spatial attenuation distribution factor, and construct a local energy that characterizes the spatial distribution density of beam energy; Based on the geometric span when the local energy extraction energy density decays to the material ablation threshold, the boundary radius characterizing the actual area of ​​material removal is obtained to avoid the machining boundary definition error caused by non-physical tools. A mapping mechanism between the accuracy level and the process standard is established to obtain the allowable deviation corresponding to the process standard; The local duration of the focal spot's action over the processing surface is defined by the boundary radius and the scanning speed. Based on the local action duration and the preset center frequency of each evaluation frequency band, frequency domain attenuation features are extracted, environmental noise in non-resonant frequency bands is filtered out, and the vibration weights corresponding to each evaluation frequency band are obtained.

4. The micro-vibration risk prediction and scheduling management system for laser precision machining according to claim 3, characterized in that, A decay function characterizing the energy dissipation of the environmental medium is constructed, and the scheduling state is correlated with the decay function in the time domain to obtain the residual vibration including the damped tail wave. This is used to synthesize the source-end velocity reflecting the intensity of the source excitation, including: Obtain the attenuation constant of each of the heavy equipment in each evaluation frequency band to characterize the system damping dissipation rate, and construct a time-dependent exponential attenuation model based on the attenuation constant to form the attenuation function characterizing the structural damping characteristics; The scheduling status and the attenuation function are fused using a time-domain convolution operation to obtain the residual vibration that characterizes the structural inertial excitation that continues to exist after the equipment stops. The reference velocity representing the inherent excitation intensity of the heavy equipment at the reference position is extracted, and time-domain modulation is performed in combination with the residual vibration to obtain the source velocity after removing static interference and dynamically superimposing the scheduling state.

5. The micro-vibration risk prediction and scheduling management system for laser precision machining according to claim 4, characterized in that, Combining propagation delay and path attenuation, which characterize spatial transmission hindrance, the source velocity is propagated and synthesized into a machine displacement of the physical machine tool facing the target work order, including: Based on the physical distance between the heavy equipment and the physical platform and the time hysteresis characteristics of the environmental medium wave velocity extraction, the propagation delay characterizing the spatial transmission lag of vibration is obtained. A spatial energy dissipation model is constructed based on the distance attenuation coefficient and the physical distance to obtain the path attenuation characterizing the medium absorption effect; The vibration isolation ratio, which characterizes the frequency transmission characteristics of the vibration isolation facility, is extracted. Combined with the path attenuation, the reference speed, and the residual vibration corrected according to the propagation delay, the machine speed reflecting the intensity of the disturbance is obtained through the root mean square energy synthesis law. By combining the machine speed and the vibration weight, and removing the frequency dimension through angular frequency conversion, the machine displacement including exposure time gating filtering is obtained.

6. The micro-vibration risk prediction and scheduling management system for laser precision machining according to claim 5, characterized in that, Candidate start times for scheduling simulation are set. The vibration weight is used to perform local exposure-weighted integration of the machine displacement at the candidate start times to calculate the prediction deviation characterizing the predicted profile offset. The prediction deviation is then compared and converted with the allowable deviation to obtain a risk coefficient reflecting the margin consumption status, including: The instantaneous intensity of the target work order at different processing time periods is extracted and the maximum value is normalized to obtain the exposure weight that characterizes the difference in sensitivity at the processing stage. Within the effective time window that defines the exposure duration, the machine displacement of each evaluation frequency band is extracted and shifted to the candidate start time. The root mean square energy is weighted and averaged using the exposure weight to obtain the prediction bias that characterizes the cumulative deformation effect within a specific candidate time window. The predicted deviation is proportionally quantified to the allowable deviation to obtain the risk coefficient, which reflects the degree of processing tolerance consumption.

7. The micro-vibration risk prediction and scheduling management system for laser precision machining according to claim 6, characterized in that, Within a continuous time interval where the risk coefficient does not exceed the limit and the resource availability status of the manufacturing execution system is true, extract the extreme nodes that satisfy the scheduling boundary and align them with the system scheduling scale to derive the start-up time to avoid micro-vibration interference, including: Extract the queuing zero point set by the preconditions of the work order, traverse and filter the lower bound of the resource availability status of the manufacturing execution system on the time axis where the risk coefficient does not exceed the limit, and extract the earliest feasible time that meets the dual requirements of environmental disturbance resistance and production scheduling. Based on the fixed time indexing rules of the manufacturing execution system, the earliest feasible time is forcibly discretized and rounded upward to obtain the start-up time after removing interference from continuous time fragments; The start time is merged with the reference point of midnight of the day and converted into the scheduling entry instruction that includes queuing guidance information and conforms to the system interface communication protocol.

8. The micro-vibration risk prediction and scheduling management system for laser precision machining according to claim 7, characterized in that, Based on the stated start time, a dispatch entry instruction is issued, and the risk coefficient is decoupled and decomposed according to the vibration source to form an audit record tracing the vibration source, including: At the determined start time, the vibration weight, the vibration isolation ratio, the path attenuation, the reference speed, and the residual vibration after translation are extracted to construct a multi-dimensional energy transfer chain. Normalized energy integration is performed within the exposure time to obtain individual risks that characterize the absolute contribution of specific equipment scheduling and specific frequency bands to the final deformation error. These individual risks are then arranged in order to generate a list of individual risks. The work order number, the accuracy level, the allowable deviation, the start time, the risk coefficient corresponding to that time, the single risk list used to trace the source of interference, and the schedule identifier used for version anti-tampering are merged and serialized to generate the audit record that solidifies the responsibility traceability link.

Images