A vertical band saw machine saw blade anti-jamming system

Abstract

The present application belongs to vertical band sawing machine technical field, specifically speaking, it is a vertical band sawing machine saw blade anti-jamming system, including sawing machine body and the saw blade arranged on the sawing machine body, further including data acquisition module and control execution module in communication connection with the data acquisition module; the data acquisition module includes infrared thermal imager, dynamic pressure sensor, conductivity electrode, tension cylinder pressure transmitter, miniature tension and pressure sensor, flexible array pressure film sensor, sound level meter, low frequency speed type vibration sensor, accelerometer and machine tool numerical control system installed on the machine tool; the control execution module includes: cooling chip removal evaluation unit, based on the tooth tip temperature rise scalar, pressure fluctuation root mean square value and conductivity change rate to construct the cooling chip removal efficiency model, and obtain the cooling chip removal efficiency coefficient; effectively overcome the great hysteresis and rough processing defects caused by passive single threshold monitoring in the existing band sawing machine anti-jamming control.

Description

Technical Field

[0001] This invention belongs to the technical field of vertical band saws, specifically a system for preventing the saw blade from jamming in a vertical band saw. Background Technology

[0002] Existing band saw anti-jamming control mostly adopts passive single threshold monitoring technology. Typically, a current sensor is installed at the spindle or feed servo motor end. When the system detects a sudden increase in motor load current that exceeds a preset fixed threshold, the machine tool CNC system triggers an alarm and forces a shutdown. Although some equipment has introduced local temperature or vibration monitoring, it merely uses isolated sensor data as a single over-limit shutdown condition or performs simple mechanical backtracking with fixed parameters.

[0003] However, existing technologies have significant drawbacks: First, single threshold monitoring has a large lag, making it unable to identify complex coupled causes such as chip blockage, clamp loosening, or saw blade stiffness degradation, and by the time a current surge is detected, substantial jamming has often already occurred; second, existing systems lack nonlinear fusion modeling of multi-source heterogeneous data such as thermodynamics, fluid dynamics, dynamics, and acoustics, making it impossible to proactively and quantitatively assess global sawing robustness; finally, existing processing methods are too crude, lacking an active control mechanism that adaptively adjusts the amplitude and frequency of the feed axis micro-motion backing based on real-time chip removal efficiency and health status, resulting not only in low processing efficiency but also a high risk of saw tooth breakage due to improper backing.

[0004] Therefore, the present invention provides a vertical band saw blade anti-jamming system. Summary of the Invention

[0005] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0006] The technical solution adopted by the present invention to solve its technical problem is: a vertical band saw blade anti-jamming system of the present invention includes a band saw body and a saw blade disposed on the band saw body, and also includes a data acquisition module and a control execution module that is communicatively connected to the data acquisition module;

[0007] The data acquisition module includes an infrared thermal imager, a dynamic pressure sensor, a conductivity electrode, a tension cylinder pressure transmitter, a miniature tensile and compressive sensor, a flexible array pressure film sensor, a sound level meter, a low-frequency velocity vibration sensor, an accelerometer, and a machine tool CNC system, all installed on the machine tool.

[0008] The control execution module includes:

[0009] The cooling chip removal evaluation unit constructs a cooling chip removal efficiency model based on the tooth tip temperature rise scalar, the root mean square value of pressure pulsation, and the rate of change of electrical conductivity, and obtains the cooling chip removal efficiency coefficient.

[0010] The stiffness assessment unit constructs a saw blade stiffness health model based on the microwave peak-valley difference of tension force, natural frequency offset, and lateral clamping force, and obtains the saw blade stiffness health coefficient.

[0011] The disturbance assessment unit constructs a clamping and dynamic disturbance model based on the mean and standard deviation of the contact surface pressure, the noise frequency band energy, and the foundation vibration velocity, and obtains the clamping and dynamic disturbance coefficients.

[0012] The robustness assessment unit constructs a dynamic robustness model based on the stiffness health coefficient, disturbance coefficient, high-frequency impact energy of the guide arm, and slip rate, and obtains the robustness coefficient.

[0013] The adaptive control unit constructs an optimization model based on the robustness coefficient and chip removal efficiency coefficient, and outputs the target feed axis micro-motion yielding amplitude and frequency.

[0014] The beneficial effects of this invention are as follows:

[0015] 1. The present invention discloses a vertical band saw blade anti-jamming system. By setting up a data acquisition module and a control execution module that communicates with the data acquisition module, it effectively overcomes the defects of the passive single threshold monitoring in the existing band saw anti-jamming control, which has great lag and crude processing method. By constructing a nonlinear fusion model of multi-physics coupling, it can proactively quantify and evaluate the global sawing robustness, accurately identify complex coupling causes such as chip blockage or saw blade stiffness degradation, and the adaptive control unit actively adjusts the micro-motion yielding amplitude and frequency of the feed axis based on the real-time evaluation results. This completely replaces the traditional forced stop or fixed mechanical retraction, effectively avoids saw tooth breakage and ensures processing continuity.

[0016] 2. The vertical band saw blade anti-jamming system described in this invention significantly improves the synchronization of multi-source heterogeneous sensor data fusion and the accuracy of model perception by setting a uniformly defined observation time window as a macroscopic benchmark reference system for acquiring dynamic physical parameters. This observation time window is strictly based on the single cutting cycle of the saw blade rotation, and its length is dynamically calculated by the ratio of the total ring length of the band saw blade to the saw blade linear speed read by the system in real time. This completely eliminates data deviation and misjudgment caused by differences in sampling frequency or time misalignment, providing a highly consistent calculation premise for subsequent performance and robustness evaluation models, and solidifying the data reliability of the anti-jamming system.

[0017] 3. The vertical band saw blade anti-jamming system of the present invention achieves high-order nonlinear safety closed-loop adjustment of the feed axis micro-motion retraction action by setting a dynamic optimization model that integrates robustness coefficient and chip removal efficiency coefficient. When outputting the target retraction amplitude, the system constructs the evaluation coefficient as a nonlinear amplification multiplier based on mechanical requirements. When the working conditions deteriorate, the driving retraction amplitude exhibits an exponentially rapid amplification response. At the same time, a minimum safety micro-motion frequency floor and an anti-overflow compensation amount are introduced in the frequency adjustment. The negative exponential function is used to make the retraction frequency smoothly approach the safety floor when the system becomes unstable. This effectively reduces the dynamic resonance excitation under extremely cold working conditions and provides a valuable window for the penetration and chip removal of cutting fluid, ensuring that the control system will not collapse under extreme conditions and maintaining the bottom-line anti-jamming action. Attached Figure Description

[0018] The invention will now be further described with reference to the accompanying drawings.

[0019] Figure 1 This is a perspective view of the present invention;

[0020] Figure 2 This is an overview diagram of the overall system architecture of the present invention;

[0021] Figure 3 This is a logic flowchart of the cooling chip removal evaluation unit in this invention;

[0022] Figure 4 This is a flowchart of the stiffness evaluation unit in this invention;

[0023] Figure 5 This is a flowchart of the disturbance evaluation unit in this invention;

[0024] Figure 6 This is a logic flowchart of the robustness evaluation unit in this invention;

[0025] Figure 7 This is a logic flowchart of the adaptive control unit in this invention.

[0026] In the picture: 1. The saw body; 2. The saw blade. Detailed Implementation

[0027] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0028] The present invention provides a vertical band saw blade anti-jamming system, which includes a band saw body 1 and a saw blade 2 disposed on the band saw body 1, and also includes a data acquisition module and a control execution module that is communicatively connected to the data acquisition module.

[0029] The data acquisition module includes an infrared thermal imager, dynamic pressure sensor, conductivity electrode, tension cylinder pressure transmitter, miniature tensile and compressive sensor, flexible array pressure film sensor, sound level meter, low-frequency velocity vibration sensor, accelerometer, and machine tool CNC system, all installed on the machine tool.

[0030] The control execution module includes:

[0031] The cooling chip removal evaluation unit constructs a cooling chip removal efficiency model based on the tooth tip temperature rise scalar, the root mean square value of pressure pulsation, and the rate of change of electrical conductivity, and obtains the cooling chip removal efficiency coefficient.

[0032] The stiffness assessment unit constructs a saw blade 2 stiffness health model based on the tension force microwave peak-valley difference, natural frequency offset, and lateral clamping force, and obtains the saw blade 2 stiffness health coefficient.

[0033] The disturbance assessment unit constructs a clamping and dynamic disturbance model based on the mean and standard deviation of the contact surface pressure, the noise frequency band energy, and the foundation vibration velocity, and obtains the clamping and dynamic disturbance coefficients.

[0034] The robustness assessment unit constructs a dynamic robustness model based on the stiffness health coefficient, disturbance coefficient, high-frequency impact energy of the guide arm, and slip rate to obtain the robustness coefficient;

[0035] The adaptive control unit constructs an optimization model based on the robustness coefficient and chip removal efficiency coefficient, and outputs the target feed axis micro-motion backing amplitude and frequency.

[0036] The data acquisition module is responsible for acquiring various physical parameters during the sawing process in real time, while the control execution module analyzes and processes these data and adjusts the operation of the saw based on the analysis results. For example, the data acquisition module can be configured to continuously monitor the sawing status through various sensors and transmit the acquired raw data to the control execution module for further processing.

[0037] These sensors can work independently or in combination to provide real-time data across multiple dimensions and physical fields. For example, an infrared thermal imager can periodically scan the temperature of the saw blade tips, a dynamic pressure sensor can monitor the pressure in the cutting fluid lines, a conductivity electrode can detect the chemical state of the cutting fluid, and the machine tool CNC system provides basic operating parameters.

[0038] The control execution module may include a cooling and chip removal evaluation unit, a stiffness evaluation unit, a disturbance evaluation unit, a robustness evaluation unit, and an adaptive control unit. These units work together to form a closed-loop control system to achieve the anti-jamming function of the saw blade 2.

[0039] The cooling chip removal evaluation unit constructs a cooling chip removal efficiency model based on the scalar value of tooth tip temperature rise, the root mean square value of pressure pulsation, and the rate of change of conductivity, and obtains the cooling chip removal efficiency coefficient. Specifically, this unit can receive temperature data from an infrared thermal imager, pressure data from a dynamic pressure sensor, and conductivity data from a conductivity electrode. By comprehensively analyzing these parameters, such as by using a weighted average or judgment based on empirical rules, a preliminary cooling chip removal efficiency coefficient can be calculated. For example, when the tooth tip temperature increases, the pressure pulsation increases, or the conductivity changes drastically, the cooling chip removal efficiency can be considered to have decreased.

[0040] The stiffness assessment unit constructs a saw blade 2 stiffness health model based on the peak-to-valley difference of the tension force microwave, the natural frequency offset, and the lateral clamping force, and obtains the saw blade 2 stiffness health coefficient. This unit can receive tension force data from the tension cylinder pressure transmitter, vibration data from the accelerometer, and lateral clamping force data from the miniature tensile and compressive sensor. By comprehensively analyzing these parameters, such as by using threshold comparison or judgment based on empirical rules, a saw blade 2 stiffness health coefficient is generated. For example, excessive tension force fluctuation, natural frequency deviation, or insufficient lateral clamping force may all indicate that there is a problem with the stiffness of the saw blade 2.

[0041] The disturbance assessment unit constructs a clamping and dynamic disturbance model based on the mean and standard deviation of the contact surface pressure, the noise frequency band energy, and the foundation vibration velocity, and obtains the clamping and dynamic disturbance coefficients. This unit can receive contact surface pressure data from the flexible array pressure film sensor, noise data from the sound level meter, and foundation vibration data from the low-frequency velocity type vibration sensor. By comprehensively analyzing these parameters, such as by using statistical analysis or judgment based on empirical rules, the stability of the clamping and the level of dynamic disturbance are evaluated, thereby obtaining the clamping and dynamic disturbance coefficients. For example, uneven contact surface pressure distribution, abnormal noise energy, or excessive foundation vibration may indicate the presence of clamping or dynamic disturbance problems.

[0042] The robustness assessment unit constructs a dynamic robustness model based on the aforementioned stiffness health coefficient, disturbance coefficient, guide arm high-frequency impact energy, and slip rate, and obtains robustness coefficients. This unit integrates the coefficients obtained by the previous assessment unit with the guide arm high-frequency impact energy data from the accelerometer and the slip rate data provided by the machine tool CNC system. By comprehensively analyzing these parameters, such as by using weighted average or judgment based on empirical rules, a coefficient representing the overall robustness of the sawing system is generated. For example, when the stiffness health coefficient is high, the disturbance coefficient is low, the impact energy is small, and the slip rate is stable, the system robustness is relatively high.

[0043] The adaptive control unit constructs an optimization model based on the aforementioned robustness coefficient and chip removal efficiency coefficient, and outputs the target feed axis micro-motion retraction amplitude and frequency. This unit receives the robustness coefficient and chip removal efficiency coefficient as input, and calculates the feed axis micro-motion retraction amplitude and frequency that need to be adjusted by comprehensively analyzing these coefficients, such as by using a lookup table or judgment based on empirical rules. For example, when the robustness coefficient and chip removal efficiency coefficient are low, the system will output a larger retraction amplitude and a higher retraction frequency to avoid jamming.

[0044] First, the infrared thermal imager in the data acquisition module monitors the continuous rise in temperature of the saw blade tip 2 in real time. The dynamic pressure sensor detects that the pressure pulsation at the front end of the cutting fluid nozzle begins to increase. At the same time, the conductivity electrode also detects that the conductivity change rate of the cutting fluid exceeds the normal range. The machine tool CNC system provides basic operating parameters such as the current saw blade linear speed.

[0045] Next, the cooling and chip removal evaluation unit in the control execution module receives this data. Based on this data, the unit calculates using its internal cooling and chip removal efficiency model and finds that the cooling and chip removal efficiency coefficient is continuously decreasing, indicating that the cooling and chip removal capabilities of the cutting fluid are weakening and there is a risk of chip blockage.

[0046] Meanwhile, the stiffness assessment unit continued to work. The tension cylinder pressure transmitter detected a slight microwave peak-valley difference in the tension force. After processing, the vibration signal collected by the accelerometer showed a slight shift in the natural frequency of the saw blade 2. The miniature tension and compression sensor showed that the lateral clamping force remained stable. Based on this information, the stiffness assessment unit calculated that the stiffness health coefficient of the saw blade 2 had decreased slightly, but was still within an acceptable range, indicating that the saw blade 2 itself had not yet experienced a serious stiffness problem.

[0047] In addition, the disturbance assessment unit also monitors the environment and clamping status. The flexible array pressure film sensor shows that the pressure distribution on the clamping contact surface is uniform, the noise frequency band energy detected by the sound level meter is at a normal level, and the low-frequency velocity type vibration sensor does not detect abnormal base vibration velocity. Thus, the disturbance assessment unit calculates that the clamping and dynamic disturbance coefficient remain at a high level, indicating that the clamping is stable and the external disturbance is small.

[0048] Subsequently, the robustness assessment unit integrates the coefficients output by the cooling chip removal assessment unit, stiffness assessment unit, and disturbance assessment unit, as well as data such as the high-frequency impact energy and slip rate of the guide arm. Due to the decrease in cooling chip removal efficiency coefficient, even if the stiffness health coefficient and disturbance coefficient are relatively stable, the robustness assessment unit will still calculate a decreasing robustness coefficient, reflecting that the overall anti-jamming capability of the system is weakening.

[0049] Finally, the adaptive control unit receives the decreased robustness coefficient and cooling chip removal efficiency coefficient. Based on its internal optimization model, the adaptive control unit immediately calculates and outputs new target feed axis micro-motion yielding amplitude and frequency. For example, the system may instruct the feed axis to operate with a slightly increased micro-motion yielding amplitude and a slightly increased micro-motion yielding frequency. These adjustment commands are sent to the machine tool CNC system, which then adjusts the motion trajectory of the feed axis.

[0050] Therefore, before the saw blade 2 actually jams, the system effectively alleviates the problems of chip blockage and poor heat dissipation by actively adjusting the micro-motion yielding strategy of the feed axis, thus preventing the saw blade 2 from jamming and maintaining the continuity of processing.

[0051] This application integrates multiple sensors, such as an infrared thermal imager, a dynamic pressure sensor, and a conductivity electrode, through a data acquisition module. This enables comprehensive real-time data acquisition of multi-source heterogeneous data, including thermodynamics, fluid dynamics, dynamics, and acoustics, during the sawing process. This allows the system to proactively identify potential jamming risks rather than reactively responding to jamming events that have already occurred. For example, in the above example, the system begins to assess when abnormalities occur in tooth tip temperature rise, pressure pulsation, and cutting fluid conductivity, much earlier than the alarm point of traditional current surge.

[0052] Furthermore, the control execution module of this application constructs a nonlinear fusion model coupled with multiple physics fields through a cooling and chip removal evaluation unit, a stiffness evaluation unit, a disturbance evaluation unit, and a robustness evaluation unit. This overcomes the limitations of existing technologies that treat isolated sensor data as a single out-of-limit shutdown condition. In the above example, the system does not simply shut down because a certain parameter exceeds the limit, but comprehensively evaluates multiple factors such as cooling and chip removal efficiency, saw blade 2 stiffness health, clamping and dynamic disturbances, and finally quantifies them into a robustness coefficient. This comprehensive evaluation mechanism can identify complex coupled causes such as chip blockage, clamp loosening, or saw blade 2 stiffness degradation, providing a more accurate understanding of the sawing state.

[0053] Ultimately, the adaptive control unit constructs an optimization model based on robustness coefficients and chip removal efficiency coefficients, outputting the target feed axis micro-motion yielding amplitude and frequency, thus achieving active adaptive control. This contrasts sharply with the crude forced shutdown or simple fixed-parameter mechanical retraction in existing technologies. In the example above, the system dynamically adjusts the feed axis micro-motion yielding strategy based on real-time evaluation results, rather than simply stopping the machine. This refined active intervention mechanism not only effectively avoids saw blade jamming and reduces the risk of saw tooth breakage, but also improves equipment reliability and processing quality while ensuring processing efficiency. Therefore, this application provides a band saw anti-jamming solution that moves from passive response to active prevention, from single judgment to multi-dimensional fusion, and from coarse intervention to refined adaptation, demonstrating significant technological advancement.

[0054] When acquiring dynamic physical characteristics, the data acquisition module uniformly defines an observation time window as a macroscopic benchmark reference system. The observation time window is based on the single cutting cycle of the saw blade 2 rotation, and its time length is uniquely determined by the ratio of the total annular length of the band saw blade 2 preset by the system to the real-time linear speed of the saw band read by the machine tool CNC system.

[0055] When acquiring dynamic physical parameters, the data acquisition module uniformly defines the observation time window. As a macroscopic benchmark reference system, the observation time window Defined as the single cutting cycle of two revolutions of the saw blade, the formula is as follows:

[0056]

[0057] In the formula, The unit is ; The system presets the total length of the band saw blade in two loops, in units of... ; The saw blade linear speed is read in real time from the instruction register of the machine tool's CNC system, and the unit is _____. .

[0058] Observation time window It is a preset time interval used to limit the data acquisition module to sample and process data within a specific time period. Its function is to provide a unified time reference benchmark for multi-source heterogeneous data, ensuring that all data correspond to the same physical state or process when performing data fusion and model calculation.

[0059] Observation time window This can be implemented at the data processing layer through software algorithms, for example, by defining the data buffer size or sampling period, or by precise control through hardware timers. A macroscopic reference system refers to a unified, global time or spatial coordinate system against which all relevant data and events are synchronized and aligned. In this application, it specifically refers to providing a unified time dimension reference point for data acquisition to solve the problem of asynchronous data acquisition from different sensors. Besides a time reference, a macroscopic reference system can also be a spatial reference, for example, by using a unified coordinate system to align data from different sensors in spatial locations, and by defining the observation time window. Defined as the single-cycle cutting period of the saw blade 2, it refers to the time required for the band saw blade 2 to complete one full cycle of motion. It aims to ensure that the system can capture the physical state information of all key points of the saw blade 2 on the entire circular path within each observation cycle, thereby ensuring the integrity and representativeness of the data. This cycle can be determined by monitoring specific marker points on the saw blade 2 or by calculating the linear velocity and length of the saw blade 2.

[0060] The system presets the total length of the band saw blade in two loops. This refers to the fixed perimeter of the band saw blade 2 used in the vertical band saw. This length is a known and relatively stable physical parameter, which is usually measured during the installation of the saw blade 2 or obtained from the specifications of the saw blade 2. This length can be stored in advance in the parameter table of the machine tool CNC system.

[0061] The instruction register of a CNC machine tool system is a hardware or software unit within the system used to store and manage machine tool operating instructions and status parameters. It can reflect the current operating status of the machine tool in real time, such as spindle speed, feed rate, and tool position. In addition to the instruction register, real-time operating parameters can also be obtained through the PLC (Programmable Logic Controller) or HMI (Human Machine Interface) interface of the CNC system.

[0062] Real-time reading of saw blade linear speed This refers to the actual operating speed of the band saw blade 2 during the sawing process. This speed is dynamic and is affected by factors such as load and motor speed. This parameter is obtained in real time through the instruction register of the machine tool's CNC system, which ensures the observation time window. The calculation is based on the current actual working conditions, thereby realizing the dynamic adaptive adjustment of the time window. In addition to reading from the instruction register, the linear speed of the saw belt can also be directly measured by the encoder or speed sensor installed on the saw belt.

[0063] The data acquisition module can adjust the data acquisition time window in real time according to the actual operating speed of the saw blade 2. When the saw blade linear speed... When changes occur, the observation time window The system will also be adjusted accordingly to always maintain the same time required for saw blade 2 to complete one full cycle. This means that regardless of the speed of saw blade 2, within each observation time window... All physical parameters collected internally (such as temperature data from the infrared thermal imager, pressure data from the dynamic pressure sensor, and conductivity data from the conductivity electrode) will correspond to the complete physical process of the saw blade 2 running one revolution on the entire circular path. This unified time reference eliminates calculation errors and model inaccuracies caused by data asynchrony, greatly improving the real-time performance and accuracy of the entire anti-jamming system's perception of the sawing status, and providing a reliable basis for the subsequent adaptive control unit to output the target feed axis micro-motion yielding amplitude and frequency.

[0064] By uniformly defining the observation time window The cutting cycle of a single rotation of the band saw blade 2 is based on the total annular length of the band saw blade 2. and real-time saw band linear speed Dynamic calculations ensure that all collected physical parameters correspond to the physical process of a complete cycle of saw blade 2, eliminating data deviations caused by fluctuations in saw blade 2 rotation speed or differences in sampling frequency. This provides a unified and synchronized time reference for the cooling and chip removal evaluation unit, stiffness evaluation unit, disturbance evaluation unit, and robustness evaluation unit. This data synchronization mechanism enables subsequent evaluation models to perform accurate calculations based on consistent physical states, significantly improving the real-time performance and accuracy of anti-jamming control. For example, when evaluating cooling and chip removal efficiency, key parameters such as tooth tip temperature rise, pressure pulsation, and conductivity change rate can accurately reflect the conditions within the same sawing cycle, avoiding misjudgments caused by data misalignment. Similarly, when evaluating the stiffness health and dynamic disturbances of saw blade 2, the synchronization of relevant parameters also ensures the reliability of the evaluation results.

[0065] Ultimately, this precise and real-time state perception capability provides a solid data foundation for the adaptive control unit to output the target feed axis micro-motion retraction amplitude and frequency, enabling the anti-jamming system to respond more accurately to complex changes in the sawing process, effectively preventing saw blade 2 from jamming, and improving processing efficiency and saw blade 2 service life.

[0066] The cooling chip removal efficiency model in the cooling chip removal evaluation unit is constructed by the following logic: the tooth tip temperature rise scalar, the root mean square value of pressure pulsation, and the rate of change of conductivity are normalized dimensionlessly based on the system's preset benchmark critical parameters; the normalized characteristic quantities are weighted and aggregated, and the cooling chip removal efficiency coefficient is output through a preset nonlinear decay function; when the comprehensive deviation of each characteristic quantity increases, the cooling chip removal efficiency coefficient approaches zero nonlinearly and rapidly to characterize the sudden deterioration of the cutting lubricant film.

[0067] The formula for the cooling chip removal efficiency model in the cooling chip removal evaluation unit is:

[0068]

[0069] In the formula, The cooling chip removal efficiency coefficient is dimensionless. This is a scalar value for the temperature rise at the tooth tip, in units of... ; This is the root mean square value of the pressure pulsation, in units of... ; This represents the rate of change of cutting fluid conductivity, in units of... ; , , These are preset constant benchmarks for temperature rise, pressure pulsation, and conductivity change rate, respectively; , , These are dimensionless weighting coefficients, and .

[0070] Among them, the cooling chip removal efficiency model is a mathematical model designed to quantify the cooling capacity of the coolant and the chip removal efficiency during the sawing process. This model can be implemented as a software module in the control execution module, and the efficiency coefficient is calculated by processing sensor data in real time.

[0071] Cooling chip removal efficiency coefficient It is a dimensionless index used to represent the overall performance level of a cooling and chip removal system. Its value is typically between 0 and 1, where 1 represents optimal performance. It is a scalar value representing the tooth tip temperature rise. This is the temperature rise at the tips of the two saw teeth, directly reflecting the heat load in the sawing area and the cooling effect of the coolant. This temperature rise scalar can be obtained by measuring the tip temperature and subtracting the ambient or reference temperature, for example, through non-contact measurement using an infrared thermal imager, or by measuring the root mean square value of pressure pulsation. This indicates the degree of pressure fluctuation within the cutting fluid delivery system. High pressure pulsation may indicate blockage in the flow path, abnormal pump operation, or uneven distribution of cutting fluid, all of which can affect cooling and chip removal. This value can be obtained by statistically calculating the instantaneous pressure signal.

[0072] Rate of change of cutting fluid conductivity It reflects the rate of change of the cutting fluid conductivity over time and can be used as an indicator of the degree of cutting fluid contamination or degradation, because chips, debris, or other contaminants can change the conductivity of the cutting fluid, thereby affecting its chip removal ability and overall effectiveness. This rate of change can be obtained by continuously monitoring the conductivity of the cutting fluid and calculating its time derivative.

[0073] Preset constants for temperature rise, pressure pulsation, and rate of change of conductivity. , , It is a reference value used to compare current operating conditions, typically determined through experimental calibration under ideal or benchmark cutting conditions or according to manufacturer specifications, and stored in the configuration parameters of the control execution module. It is a dimensionless weighting coefficient. , , The weights are used to assign the relative importance of each input parameter in the cooling chip removal efficiency model. The sum of these weights is 1 to ensure the normalization of the weighted sum. These weights can be adjusted according to the specific cutting material, machining conditions or operating requirements.

[0074] This application's solution integrates three key physical parameters—tooth tip temperature rise scalar, pressure pulsation root mean square value, and cutting fluid conductivity change rate—into a cooling and chip removal efficiency model, achieving a comprehensive quantitative evaluation of the cooling and chip removal status during sawing. These parameters are first normalized according to a preset constant benchmark and then weighted using dimensionless weighting coefficients to reflect their relative importance in evaluating cooling and chip removal efficiency. Subsequently, these weighted and normalized parameters are input into an exponential decay function to generate the final cooling and chip removal efficiency coefficient. This nonlinear fusion method ensures that when any one or more parameters deviate from their ideal benchmark value, the efficiency coefficient exhibits a sensitive and rapid decreasing trend, thus enabling timely detection of performance degradation in the cooling and chip removal system. This multi-parameter evaluation method overcomes the limitations of traditional single-parameter monitoring, providing a comprehensive and real-time insight into the thermal state of the sawing area and chip removal efficiency. Furthermore, this cooling and chip removal evaluation unit works closely with the data acquisition module, which operates within a uniformly defined observation time window. The system obtains the required dynamic physical parameters, ensuring the synchronization and representativeness of the model input data. This makes the calculation results of the cooling chip removal efficiency coefficient more accurate and reliable, providing a solid data foundation for subsequent adaptive control.

[0075] The specific methods for acquiring characteristic quantities in the cooling chip removal evaluation unit include: the tooth tip temperature rise scalar is represented as the difference between the highest area temperature extracted from the region of interest of the sawing exit scanned by the infrared thermal imager and the ambient reference temperature within the observation time window; the pressure pulsation root mean square value is represented as the root mean square characteristic value of the transient pressure fluctuation collected by the dynamic pressure sensor within the observation time window; the conductivity change rate is represented as the absolute value of the change obtained by discretely differentiating the signal collected by the conductivity electrode with the observation time window as the step size, in order to filter out high-frequency transient noise interference.

[0076] The specific method for obtaining the basic parameters in the cooling chip removal efficiency model is as follows:

[0077]

[0078] in, To preset the ambient temperature, For the infrared thermal imager installed near the sawing exit, within the time window Temperature of the highest pixel point at the tooth tip extracted by internal scanning;

[0079]

[0080] In the formula, The transient pressure is collected by a dynamic pressure sensor installed on the cutting fluid line at the nozzle tip. The average pressure within the time window;

[0081]

[0082] in, The instantaneous conductivity measurement value is collected by a conductivity electrode installed on the cutting fluid return line.

[0083] Infrared thermal imagers are used to measure the temperature distribution on the surface of an object in a non-contact manner. During the sawing process, they can be installed near the sawing exit and acquire the temperature information of the tooth tips in real time by scanning the tooth tip area of ​​the saw blade. The infrared thermal imager can use an uncooled microbolometer array, such as a vanadium oxide (VOx) or amorphous silicon (a-Si) detector, to achieve cost-effectiveness and compactness; or it can use a cooled detector, such as a mercury cadmium telluride (MCT) or indium antimonide (InSb) detector, to provide higher sensitivity and faster response speed, thereby capturing the instantaneous temperature changes of the tooth tips more accurately.

[0084] Secondly, the dynamic pressure sensor is used to monitor pressure fluctuations in the cutting fluid line. Installing this sensor on the line upstream of the cutting fluid nozzle allows for real-time acquisition of the transient pressure of the cutting fluid. This dynamic pressure sensor can be a piezoelectric sensor, which has high-frequency response characteristics and can accurately capture minute pressure pulsations during the cutting fluid injection process. Furthermore, the conductivity electrode is used to reflect changes in ion concentration or impurity content in the cutting fluid. Installing this electrode on the cutting fluid return line allows for real-time monitoring of the instantaneous conductivity of the cutting fluid. The conductivity electrode can be a two-electrode or four-electrode sensor, which calculates conductivity by measuring the resistance of the solution, and is suitable for monitoring conventional cutting fluids; or, an inductive conductivity sensor can be used, which avoids electrode contamination by measuring the induced current, and is suitable for cutting fluid environments containing a lot of particles or oil.

[0085] Based on this, the scalar value of the tooth tip temperature rise The acquisition is based on the infrared thermal imager within the observation time window. Temperature of the highest pixel point at the tooth tip extracted by internal scanning With preset ambient temperature The difference between them serves to quantify the heat accumulation in the tooth tip area during sawing, directly reflecting the balance between the generation and dissipation of cutting heat, and is a key indicator for evaluating the cooling effect.

[0086] Root mean square value of pressure pulsation The acquisition is based on transient pressure collected by a dynamic pressure sensor. During the observation time window The root mean square (RMS) calculation within the fluid is used to characterize the stability and uniformity of the cutting fluid injection. A larger value indicates a lower uniformity. An abnormal value may indicate an unstable or uneven supply of cutting fluid, thus affecting cooling and chip removal.

[0087] Rate of change of cutting fluid conductivity The acquisition is based on the instantaneous conductivity collected by the conductivity electrode. During the observation time window The rate of change within the fluid indirectly reflects the change in the concentration of chips or abrasive particles in the cutting fluid. When chip removal is not smooth, chips accumulate in the cutting fluid, which may lead to a significant change in conductivity, thus indicating a decrease in chip removal efficiency.

[0088] During the sawing process, an infrared thermal imager was strategically installed near the saw exit to continuously monitor the temperature distribution at the tips of the two saw blade teeth, and within the aforementioned observation time window... Inside, the highest pixel temperature at the tooth tip is precisely captured. Meanwhile, a dynamic pressure sensor is installed on the cutting fluid line at the nozzle tip to collect the transient pressure of the cutting fluid in real time. In addition, conductivity electrodes are deployed on the cutting fluid return line to continuously monitor the instantaneous conductivity of the cutting fluid. These sensors are in the same observation window Internal synchronous data acquisition ensures the alignment of different physical quantities in the time dimension. Subsequently, using this synchronous data, calculations are performed through a specific mathematical model: the scalar value of tooth tip temperature rise. pass With preset ambient temperature The difference obtained directly quantifies the accumulation of cutting heat; the root mean square value of pressure pulsation. By analyzing transient pressure In the time window The integral operation within the timeframe is used to characterize the stability of the cutting fluid injection; the rate of change of cutting fluid conductivity is also measured. Then through instantaneous conductivity In the time window The rate of change within the fluid is calculated, indirectly reflecting the concentration of chips in the cutting fluid. Through the above-mentioned collaborative work, the limitations of traditional single-sensor monitoring can be overcome, and multi-dimensional information such as thermal, fluid dynamics and chemical composition can be organically combined, so that the cooling chip removal efficiency model can more realistically reflect the complex environment of the sawing area, thus providing a solid data foundation for subsequent adaptive control decisions. This effectively solves the problem of inaccurate model evaluation caused by the lack of precise quantification methods in traditional solutions.

[0089] The stiffness health model of saw blade 2 in the stiffness assessment unit is constructed by extracting the microwave peak-valley difference of tension force, the natural frequency offset, and the relative deviation of lateral clamping force from the factory standard reference value. Based on the nonlinear hysteresis decay law of material fatigue damage accumulation, a mathematical model containing high-power inverse mapping terms is constructed. As the deviation of each characteristic term accumulates, the stiffness health coefficient of saw blade 2 shows a nonlinear accelerated decay trend.

[0090] The formula for the saw blade 2 stiffness health model in the stiffness evaluation unit is:

[0091]

[0092] In the formula, The stiffness health coefficient of the saw blade is dimensionless. The peak-to-valley difference in tension force is expressed in units of... ; This is the natural frequency offset, in units of ; Lateral clamping force, unit: ; This is the standard tension reference, in units of ; The initial natural frequency reference, in units of ; This serves as the standard clamping force reference, with units of... ; , , It is a dimensionless constant; It is a positive even number.

[0093] Among them, the stiffness health coefficient of saw blade 2 It is a dimensionless index used to quantitatively evaluate the stiffness health of saw blade 2 under working conditions. Its value is usually between 0 and 1, where 1 indicates that the stiffness of saw blade 2 is in an ideal healthy state, and the smaller the value, the more serious the stiffness degradation. This coefficient serves as the quantitative basis for comprehensively evaluating the stiffness state of saw blade 2 and provides decision input for subsequent adaptive control. It is obtained by nonlinearly fusing multiple physical parameters, including the peak-valley difference of tension force. The tension of the saw blade 2 during operation is the difference between the instantaneous maximum and minimum values ​​of its tension force within an observation period. This parameter reflects the uniformity and stability of the tension state of the saw blade 2. A large peak-to-valley difference may indicate that the saw blade 2 has problems such as local slack, wear of the tensioning mechanism, or deformation of the saw blade 2 itself. These will affect the overall stiffness of the saw blade 2. This parameter can be obtained by real-time monitoring of the tension force by a force sensor or pressure sensor installed on the tensioning mechanism and statistical analysis within a specific time window.

[0094] natural frequency offset The natural frequency refers to the deviation between the actual natural frequency of the saw blade 2 under working conditions and its initial or standard natural frequency. Natural frequency is an important indicator of structural dynamics. Factors such as wear, fatigue, cracks, or thermal deformation of the saw blade 2 can cause changes in its mass and stiffness distribution, leading to a shift in its natural frequency. This shift is a key dynamic characteristic for assessing the stiffness degradation of the saw blade 2. This parameter can be obtained by installing vibration sensors (such as accelerometers) on or near the saw blade 2 to collect vibration signals, then extracting the current natural frequency through modal analysis, spectral analysis, or natural response identification, and comparing it with a preset initial natural frequency.

[0095] Lateral clamping force The guide mechanism applies a lateral constraint force to the saw blade 2. The guide mechanism maintains the straightness and stability of the saw blade 2 during the cutting process through lateral clamping force, preventing it from lateral swaying or twisting. Insufficient or abnormal fluctuations in clamping force will directly affect the geometric stiffness of the saw blade 2, making it more prone to instability under cutting load. This parameter can be directly measured by installing a miniature force sensor on the guide block or its adjustment mechanism.

[0096] Standard tension reference Initial natural frequency reference Standard clamping force benchmark These are the physical quantity reference values ​​corresponding to the ideal or healthy working conditions of the saw blade 2. They serve as the benchmark for evaluating the current condition of the saw blade 2. All deviations are calculated relative to these benchmark values, which are usually calibrated through experimental measurement, empirical setting, or data provided by the manufacturer during the installation and commissioning of the saw blade 2 or before the new saw blade 2 is put into use.

[0097] Dimensionless weighting coefficient , , This is used to adjust the relative importance of different physical quantities in the stiffness health model. Since different physical quantities may have varying degrees of influence on the stiffness health of saw blade 2, setting different weighting coefficients can make the model more accurately reflect the actual situation. These coefficients can be determined through expert experience, historical data analysis, machine learning algorithm optimization, or experimental calibration. Positive even numbers It is an exponential parameter in the model, used to amplify abnormal fluctuations deviating from the baseline state. Because it is an even number, it ensures the non-negativity of the deviation term; because it is a positive number, the larger the deviation, the greater the negative impact on the health coefficient. This value makes the model more sensitive to small deviations, thus providing a more obvious early warning in the early stages of saw blade stiffness degradation. A suitable value for this parameter is usually determined through experimental verification and system optimization. The value is adjusted to balance the model's sensitivity and robustness.

[0098] By introducing the peak-valley difference of the tension force microwave, the change in the uniformity of the force on the saw blade 2 during the rotation cycle can be reflected, revealing whether there is local slack or abnormal tension in the saw blade 2. By introducing the natural frequency offset, the changes in the dynamic characteristics of the saw blade 2 caused by wear, fatigue, or thermal deformation can be captured, which is an important dynamic indicator for judging the stiffness degradation of the saw blade 2. By introducing the lateral clamping force deviation, the constraint capability of the guide mechanism on the saw blade 2 can be monitored in real time, ensuring that the saw blade 2 maintains the expected geometric stiffness during the cutting process. These parameters are uniformly defined in the observation time window when the data acquisition module acquires dynamic physical parameters. As a macroscopic benchmark reference system, the observation time window Defined as the single cutting cycle of the saw blade's two rotations, thus ensuring the synchronization and consistency of data acquisition, the model adopts exponential normalization processing to convert the deviations of the above three dimensions into dimensionless health coefficients.

[0099] Among them, parameters As a positive even number, it can effectively amplify abnormal fluctuations that deviate from the baseline state, making the health coefficient more sensitive to stiffness degradation. This fusion modeling method based on multi-source physical parameters can more accurately characterize the real-time stiffness health status of the saw blade 2, providing a reliable decision basis for subsequent adaptive control, thereby effectively avoiding the risk of saw blade 2 jamming or breaking due to insufficient stiffness.

[0100] The specific methods for obtaining characteristic quantities in the stiffness evaluation unit include: the peak-valley difference of the tension force microwave is the difference between the peak value and the valley value of the tension force signal within the observation time window; the natural frequency offset is the absolute value of the difference between the current natural frequency and the initial frequency obtained by self-vibration response identification based on accelerometer signals; and when calculating the deviation of the lateral clamping force, the absolute value of the difference between the current measured value and the reference value is used to normalize the difference relative to the reference value in order to avoid the abnormal division-by-zero singularity of the calculation when the clamping force is lost.

[0101] The specific method for obtaining the basic parameters in the saw blade 2 stiffness health model is as follows:

[0102] The tension force signal acquired by the tension cylinder pressure transmitter installed at the pressure test port of the tension cylinder within the time window The difference between the maximum and minimum values ​​within;

[0103]

[0104] in, The current natural frequency is obtained by collecting vibration signals from an accelerometer mounted on the guide arm and identifying them through natural vibration response.

[0105] The real-time lateral clamping force is directly collected by a miniature tension / compression sensor installed after the adjusting screw of the movable guide block.

[0106] Among them, the tension cylinder pressure transmitter is a sensor that can convert the hydraulic pressure inside the tension cylinder into a standard electrical signal output. Its core function is to monitor the pressure change inside the cylinder in real time. This pressure directly reflects the tension state of the saw blade 2. For example, a piezoresistive pressure transmitter can be used. Its internal pressure-sensitive resistor deforms under pressure, causing a change in resistance value, and then outputs a voltage or current signal proportional to the pressure.

[0107] The difference between the maximum and minimum values ​​of the tension signal within the time window quantifies the performance of saw blade 2 within a specific observation time window. The dynamic fluctuation range of the internal tension reflects the stability of the saw blade 2 tensioning system and the dynamic response of the saw blade 2 under force. This difference can be obtained by sampling the continuous pressure signal output by the tensioning cylinder pressure transmitter through a digital signal processor (DSP) or microcontroller, within each observation time window. Within the time window, record the maximum and minimum values ​​of all sampling points, and then calculate their difference; or use an analog peak hold circuit to track and hold the peak and valley values ​​of the signal in real time, and read and calculate the difference at the end of the time window.

[0108] An accelerometer is a sensor that measures the acceleration of an object. It is used to monitor the vibration of the saw blade 2 or the guide arm. These vibrations contain dynamic characteristic information of the saw blade 2, such as its natural frequency. For example, a piezoelectric accelerometer can be used, which utilizes the principle that piezoelectric ceramics generate charges when subjected to vibration to convert mechanical vibration into an electrical signal. Natural response identification is a technique that determines the system's modal parameters such as natural frequency and damping ratio by analyzing the vibration response signal of the system under no external excitation or only random excitation. For example, a Fast Fourier Transform (FFT) can be used to perform spectral analysis on the vibration time-domain signal collected by the accelerometer. The natural frequency is determined by identifying the peak frequency of energy concentration in the spectrum. The current natural frequency refers to the natural vibration frequency exhibited by the saw blade 2 under the current working state. It is closely related to the material properties, geometry, tension, and stiffness of the saw blade 2. When the saw blade 2 experiences wear, fatigue, changes in tension, or stiffness degradation, its natural frequency will change.

[0109] A miniature tensile or compressive sensor is a sensor capable of measuring minute tensile or compressive forces. It is typically characterized by its small size and high accuracy, and is used to directly measure the lateral clamping force exerted by the movable guide block on the saw blade 2. For example, a resistance strain gauge sensor can be used, which measures force based on the principle of the change in resistance value of the strain gauge when it deforms under stress. The real-time lateral clamping force refers to the instantaneous lateral constraint force exerted by the movable guide block on the saw blade 2 during its operation. The magnitude of this force directly affects the lateral stiffness, vibration characteristics, and sawing stability of the saw blade 2.

[0110] By clarifying the acquisition paths of three core physical parameters—the peak-to-valley difference of tension force microwave, the natural frequency offset, and the lateral clamping force—high-precision real-time data support was provided for the aforementioned saw blade 2 stiffness health model. A tension cylinder pressure transmitter was installed at the tension cylinder pressure measuring port to monitor the pressure changes within the tension cylinder in real time. This pressure signal was observed within the observation time window. The difference between the maximum and minimum values ​​within the microwave tension, i.e., the peak-to-valley difference of the tension force, can effectively capture the dynamic fluctuations of the saw blade 2 tension during operation, thus reflecting the tension stability of the saw blade 2 within the cutting cycle. This dynamic monitoring method overcomes the limitations of traditional static tension force measurement and more realistically reflects the stress state of the saw blade 2 under actual working conditions. At the same time, the accelerometer installed on the guide arm continuously collects the vibration signals of the saw blade 2. These vibration signals, after self-vibration response identification processing, can accurately extract the current natural frequency of the saw blade 2. .

[0111] By using the current inherent frequency Compared with the initial natural frequency reference By comparison, the natural frequency offset is calculated. This offset can sensitively detect changes in the stiffness of the saw blade 2 caused by wear, fatigue, or changes in stress state. This frequency offset-based monitoring method can effectively avoid the limitations of a single static indicator and provide early warning of stiffness degradation of the saw blade 2. In addition, the miniature tensile and compressive sensor is cleverly installed after the adjusting screw of the movable guide block to directly collect the real-time lateral clamping force applied by the movable guide block to the saw blade 2. This parameter directly reflects the constraint strength of the guide block on saw blade 2, ensuring that the stiffness assessment model can perceive changes in external clamping constraints in real time. The peak-to-valley difference of the tension force microwave, the natural frequency offset, and the real-time lateral clamping force are used as key inputs to the saw blade 2 stiffness health model, jointly constructing a comprehensive, dynamic, and accurate saw blade 2 stiffness health assessment system. Through these real-time, multi-dimensional physical parameters, the system can more accurately quantify the stiffness health coefficient of saw blade 2. This multi-source data fusion strategy significantly improves the accuracy and foresight of the health status assessment of the saw blade 2, effectively solving the problem that traditional single or static monitoring methods cannot accurately predict the stiffness degradation and jamming risk of the saw blade 2.

[0112] The clamping and dynamic disturbance model in the disturbance evaluation unit includes a clamping loosening penalty term and a dynamic disturbance penalty term. The clamping loosening penalty term is based on the mean and standard deviation of the contact surface pressure extracted by the flexible array pressure film sensor to construct a proportional function characterizing the uniformity of force. The dynamic disturbance penalty term superimposes the dimensionless noise frequency band energy and the foundation vibration velocity in the characteristic space using orthogonal norms and uses a negative exponential function to map the attenuation, outputting the clamping and dynamic disturbance coefficients.

[0113] The clamping and dynamic disturbance model formulas and parameter acquisition methods in the disturbance evaluation unit are as follows:

[0114]

[0115] In the formula, The clamping and dynamic disturbance coefficients are dimensionless. and These are the mean and standard deviation of the contact surface pressure, respectively, in units of... All of these are obtained by extracting the pressure matrix output from a flexible array pressure film sensor embedded in the jaws of the clamp;

[0116] Energy in the noise frequency band, unit: ,in The signal is a frequency domain signal acquired by a sound level meter installed on the operating side of the machine tool;

[0117] The root mean square value of the basic vibration velocity is extracted by a low-frequency velocity-type vibration sensor installed on the base of the saw, in units of... ; and As a preset benchmark; and The weights are orthogonal penalty weights.

[0118] The clamping and dynamic disturbance model in this disturbance assessment unit aims to quantify the impact of external environmental disturbances and workpiece clamping status on the vertical band saw during the sawing process. It generates a dimensionless clamping and dynamic disturbance coefficient by comprehensively analyzing multiple physical parameters. This coefficient can intuitively reflect the external robustness of the sawing process. This model provides a systematic method for identifying and evaluating non-saw blade 2 performance factors that may cause saw blade 2 to jam or cause processing abnormalities.

[0119] Clamping and dynamic disturbance coefficients It is a comprehensive index used to characterize the stability of workpiece clamping during the sawing process and the degree of influence of the external environment (such as noise and machine tool foundation vibration) on the sawing process. The value of this coefficient is usually between 0 and 1. A higher value indicates that the clamping state is more stable, the external disturbance is smaller, and the sawing process is more robust; a lower value indicates that the clamping is unstable or the external disturbance is larger.

[0120] in, and The mean and standard deviation of the contact surface pressure are obtained from the pressure matrix output by a flexible array pressure film sensor mounted in the jaws of the clamp. A flexible array pressure film sensor is a thin-film device that measures the pressure distribution on a contact surface. It is typically composed of multiple layers of materials, including pressure-sensitive materials and an electrode array. When subjected to pressure, its resistance or capacitance changes, thus measuring the pressure. This sensor is flexible, thin, and bendable, allowing it to fit closely to irregular surfaces and provide high spatial resolution pressure distribution data. Besides resistive or capacitive film sensors, piezoelectric film sensors can also be used. These sensors generate an electric charge when subjected to pressure, and the amount of charge is measured to reflect the pressure magnitude. A flexible array pressure film sensor typically consists of multiple independent sensing units, each corresponding to a measurement point, forming a two-dimensional pressure distribution matrix. By statistically analyzing the pressure readings of all sensing units in this matrix, the mean pressure on the contact surface can be calculated. and pressure standard deviation The pressure mean reflects the overall clamping force, while the pressure standard deviation characterizes the uniformity of the clamping force. For example, the mean can be obtained by summing all effective pressure values ​​in the matrix and then dividing by the number of effective units, and the standard deviation can be obtained by calculating the square root of the average of the sum of squares of the deviations of each pressure value from the mean.

[0121] The energy in the noise frequency band, of which The frequency domain signal acquired by a sound level meter installed on the machine tool's operating side is a sound level meter, an instrument used to measure sound pressure level or sound intensity level. It typically includes a microphone, signal processing circuitry, and a display unit. It converts sound waves into electrical signals and amplifies, filters, and analyzes these signals. Besides common digital sound level meters, high-sensitivity electret or piezoelectric microphones can also be used, along with a dedicated signal acquisition and processing module, to acquire acoustic signals in the frequency domain. It is obtained by performing a Fourier transform (e.g., Fast Fourier Transform, FFT) on the time-domain sound pressure signal collected by the sound level meter. It represents the intensity distribution of different frequency components in the sound signal. By analyzing the frequency domain signal, noise characteristics within a specific frequency range can be identified, and the noise frequency band energy can be determined. This refers to the total energy of sound signals within a specific frequency range. This frequency range is usually related to abnormal noises that may be generated during the sawing process (such as chip clogging, saw tooth wear, resonance, etc.). It is determined by analyzing the frequency domain signal. Integrating within a specified frequency band allows for the quantification of noise energy within that band.

[0122] The root mean square (RMS) value of the fundamental vibration velocity is extracted by a low-frequency velocity-type vibration sensor mounted on the base of the sawing machine. This low-frequency velocity-type vibration sensor is specifically designed to measure vibration velocity in the low-frequency range. It typically employs the principle of a seismic detector or a scheme integrating an accelerometer and an integrating circuit. The seismic detector measures velocity by generating an induced electromotive force through the movement of a coil in a magnetic field, while the accelerometer measures acceleration. The velocity is then obtained through integration. Mounted on the sawing machine base, it aims to capture the inherent vibrations or environmental vibrations generated by the entire machine tool structure during operation. These vibrations may be transmitted through the workpiece or saw blade, affecting sawing stability. The RMS value of the vibration velocity... It is a commonly used indicator for measuring vibration intensity, reflecting the average level of vibration energy. For time-domain vibration velocity signals acquired by low-frequency velocity-type vibration sensors, the root mean square value can be calculated within a certain time window to obtain the vibration intensity. This value can effectively characterize the overall vibration level of the machine tool base, providing a benchmark for assessing the interference of the external environment on the sawing process.

[0123] and As a preset benchmark; and As the orthogonal penalty weight, the preset benchmark is a pre-defined reference value used to measure the energy of the measured noise frequency band. and the root mean square value of foundation vibration velocity Normalization is performed or the data is used as a benchmark for comparison. These benchmark values ​​are usually determined experimentally under ideal or standard operating conditions or based on experience, representing acceptable noise and vibration levels. Orthogonal penalty weights are used to adjust the relative importance of noise frequency band energy and foundation vibration velocity in the clamping and dynamic disturbance model. They are dimensionless constants. By adjusting their values, the model can be optimized according to the actual application scenario and the sensitivity to different disturbance sources, so that the model can more accurately reflect the actual disturbance situation. For example, if the noise is considered to have a greater impact on the sawing process, the weights can be appropriately increased. The value of .

[0124] By embedding a flexible array pressure film sensor into the jaws of a clamp, the contact pressure distribution between the workpiece and the clamp is monitored in real time, and the average pressure of the contact surface is extracted from the data. and standard deviation The mean pressure reflects the overall strength of the clamping force, while the standard deviation characterizes the uniformity and stability of the clamping. By incorporating the ratio of the mean pressure to the standard deviation into the model, the reliability of workpiece clamping can be effectively assessed, avoiding machining abnormalities caused by unstable clamping. Simultaneously, a sound level meter is installed on the machine tool's operating side to continuously collect ambient acoustic signals, and through frequency domain analysis, calculates the noise frequency band energy within a specific frequency band. This energy value can capture abnormal acoustic features that may occur during the sawing process, such as noise caused by chip clogging, saw tooth wear, or machine tool resonance, thereby identifying potential external interference. In addition, a low-frequency velocity vibration sensor is installed on the saw base to extract the root mean square value of the basic vibration velocity. This parameter quantifies the environmental vibration generated by the overall operation of the machine tool, which helps to distinguish the inherent vibration of the machine tool itself from the dynamic response during the sawing process, ensuring a more accurate assessment of sawing disturbances.

[0125] The average contact surface pressure obtained above Standard deviation Noise frequency band energy and the root mean square value of foundation vibration velocity These parameters are input into the clamping and dynamic disturbance model, which uses a nonlinear function form, and are compared with a preset benchmark. , and orthogonal penalty weights , By combining these factors, the final clamping and dynamic disturbance coefficients can be calculated. This fusion approach allows the model to comprehensively consider the impact of different disturbance sources and weight them according to their relative importance. In this way, the clamping and dynamic disturbance model can provide a quantitative and comprehensive assessment of external disturbances, including the clamping and dynamic disturbance coefficients. It is then passed to the robustness assessment unit, along with the saw blade 2 stiffness health coefficient. Together with other parameters, a dynamic robustness model is constructed, enabling the entire system to distinguish between processing abnormalities caused by a decline in the performance of the saw blade 2 itself (such as stiffness degradation) and those caused by external environmental disturbances (such as unstable clamping, noise, and foundation vibration). This precise disturbance assessment provides the adaptive control unit with a more accurate decision-making basis, allowing it to output an optimized target feed axis micro-motion yield amplitude based on the actual sawing state and external disturbance conditions. With frequency This enables more precise and proactive anti-jamming control, significantly improving the robustness and processing efficiency of the sawing process.

[0126] The dynamic robustness model in the robustness assessment unit includes a basic robustness term and a dynamic anomaly penalty term. The basic robustness term is constructed based on the geometric mean characteristics of the saw blade stiffness health coefficient and the clamping and dynamic disturbance coefficients. The dynamic anomaly penalty term is based on the high-frequency impact energy of the guide arm extracted after high-pass filtering and the slip rate obtained based on the speed difference between the driving and driven wheels. It uses a soft saturation limiting function to perform nonlinear suppression mapping to avoid assessment oscillations caused by high-frequency extreme anomalies.

[0127] The formula and parameter acquisition method for the dynamic robustness model in the robustness assessment unit are as follows:

[0128]

[0129] In the formula, The robustness coefficient is dimensionless. The maximum impact energy benchmark is given in units of... ; and This is the magnification factor;

[0130] For the high-frequency impact energy of the guide arm, among which, The high-frequency acceleration is the vibration signal obtained from the accelerometer mounted on the guide arm after high-pass filtering, and the unit is 1. ;

[0131] Let be the slip ratio, where and The angular velocities of the driving and driven wheels are read by the servo encoder of the machine tool's CNC system, in units of... , and The radius of the wheel is given in units of 1. .

[0132] Among them, the high-frequency impact energy of the guide arm It is an indicator that measures the instantaneous high-frequency impact load borne by saw blade 2 at the guide arm. It extracts the high-frequency acceleration by high-pass filtering the vibration signal collected by the accelerometer. Then, the square of the high-frequency acceleration within the observation time window The energy value, obtained by integration, can capture the instantaneous abnormal vibrations generated by the saw blade 2 at the guide position. These vibrations are often precursors to saw blade 2 instability, tooth breakage, or jamming. This can be achieved by acquiring the raw vibration signal with an accelerometer and then using a digital signal processor (DSP) or a microcontroller's built-in digital filter module to implement a high-pass filtering function. A suitable cutoff frequency is set to filter out low-frequency background noise, thereby obtaining high-frequency acceleration. Alternatively, the output signal of the accelerometer can be preprocessed using an analog high-pass filter circuit, and then the filtered analog signal can be converted into a digital signal by an analog-to-digital converter (ADC) for subsequent calculations.

[0133] slip This indicator reflects the relative slippage between the band saw blade 2 and the drive and driven wheels. It quantifies this by comparing the difference in linear velocity between the drive and driven wheels, which is calculated from their respective angular velocities and wheel radii. This indicator can monitor the slippage between the saw blade and the drive wheels in real time. When the slippage rate increases, it usually indicates insufficient saw blade tension, abnormal load, or a problem with the transmission system. These issues can lead to unstable operation of the saw blade 2 and increase the risk of jamming. This can be achieved by using a servo encoder built into the machine tool's CNC system to read the real-time angular velocities of the drive and driven wheels. and And combined with the preset wheel radius and The speed is calculated in real time by the processor in the control execution module. Alternatively, independent speed sensors, such as photoelectric encoders or Hall sensors, can be installed on the driving wheel and the driven wheel respectively to transmit the collected speed signals to the control execution module, which then calculates the angular velocity and the slip rate.

[0134] The dynamic robustness model is a mathematical model that comprehensively evaluates the stability of saw blade 2 under dynamic operating conditions. It incorporates the stiffness health coefficient of saw blade 2. Clamping and dynamic disturbance coefficients High-frequency impact energy of the guide arm and slip Multiple parameters are used to output a dimensionless robustness coefficient through a nonlinear functional relationship. This model can proactively and quantitatively evaluate global sawing robustness, overcoming the lag of traditional single-threshold monitoring and providing a reliable decision-making basis for adaptive control. It can be implemented in the central processing unit (CPU) of the control execution module by programming the mathematical model, receiving parameters in real time from the stiffness evaluation unit, disturbance evaluation unit, and data acquisition module, performing calculations, and outputting robustness coefficients. In addition, the model can also be deployed in a dedicated FPGA (Field Programmable Gate Array) or ASIC (Application-Specific Integrated Circuit) to achieve high-speed parallel computing and meet real-time requirements.

[0135] The stiffness assessment unit has provided the saw blade 2 stiffness health factor. This reflects the structural integrity and tension of the saw blade 2 itself; the disturbance evaluation unit provides clamping and dynamic disturbance coefficients. The study characterized the effects of vibrations on the contact surface between the saw blade 2 and the workpiece, as well as environmental vibrations. Vibration signals were acquired using an accelerometer mounted on the guide arm, and high-frequency acceleration was extracted after high-pass filtering. The high-frequency impact energy of the guide arm can then be calculated. This energy value can capture instantaneous abnormal vibrations generated by the saw blade 2 at the guide position. These high-frequency impacts are often precursors to saw blade 2 instability, tooth breakage, or jamming. Simultaneously, the angular velocities of the drive and driven wheels are read in real time by the servo encoder of the machine tool's CNC system. and Combined with wheel radius and The slip ratio was calculated. slip This reflects the degree of slippage between the saw blade and the drive wheel. An increase in this level usually indicates insufficient saw blade tension or abnormal load, suggesting a potential problem with the transmission system and increasing the risk of jamming.

[0136] The dynamic robustness model organically combines these parameters through the hyperbolic tangent function. Impact energy With slip Perform nonlinear mapping and use the amplification factor and Adjustments are made to enable the model to smoothly handle fluctuations in input parameters, while also ensuring the robustness coefficient remains stable when the system tends to become unstable. The design of the model not only considers the static and quasi-static characteristics of the saw blade 2, but also incorporates the high-frequency dynamic response and transmission system status. This allows for a more comprehensive and forward-looking assessment of the overall operational robustness of the saw blade 2. Through this nonlinear fusion modeling of multi-source heterogeneous data, the scheme overcomes the limitations of relying solely on single or local parameter evaluation, providing a more reliable and accurate decision-making basis for subsequent adaptive control and effectively avoiding the risk of jamming caused by the instantaneous instability of the saw blade 2.

[0137] By introducing the high-frequency impact energy of the guide arm and the slip ratio, and integrating them into the dynamic robustness model, the system accurately captures the risk of instantaneous instability of the saw blade 2 under high-speed motion. The calculation of the high-frequency impact energy of the guide arm enables the system to detect impact loads that may cause saw teeth to break or jam in a timely manner, while the monitoring of the slip ratio can effectively warn of the risk of jamming caused by abnormal transmission system. This multi-dimensional and dynamic evaluation mechanism enables the robustness coefficient to reflect the real operating state of the saw blade 2 more comprehensively and proactively, thereby providing a more reliable basis for subsequent adaptive control, significantly reducing the risk of saw blade 2 jamming and damage, and improving the stability and safety of the sawing process.

[0138] When the optimization model in the adaptive control unit outputs the micro-motion retraction amplitude of the target feed axis, it constructs the robustness coefficient and chip removal efficiency coefficient as nonlinear amplification multipliers based on the physical chip breaking and retraction mechanics requirements, which are applied to the current retraction amplitude setting value of the machine tool. When the robustness coefficient or chip removal efficiency coefficient decreases to zero and the working condition deteriorates, the micro-motion retraction amplitude of the target feed axis exhibits an exponentially rapid amplification response relative to the current setting value.

[0139] The formula for the target feed axis micro-motion retraction amplitude model output by the adaptive control unit is:

[0140]

[0141] In the formula, The target feed axis micro-adjustment relief amplitude, in units of ; The setpoint for the yield amplitude currently read by the machine tool's CNC system, in units of ; The adaptive gain adjustment coefficient is dimensionless. It is the amplitude expansion index, and .

[0142] Among them, the target feed axis micro-motion retraction amplitude This refers to the maximum displacement of the feed axis during the sawing process, caused by the periodic small retraction of the saw blade 2 to avoid jamming or reduce cutting load. By dynamically adjusting this amplitude, the contact state between the saw blade 2 and the workpiece can be effectively controlled, thus preventing the saw blade 2 from jamming. This amplitude is typically achieved by controlling the servo motor of the feed axis through the machine tool's CNC system, causing it to reciprocate according to the calculated target amplitude. Current retraction amplitude setting value. This refers to the reference value of the feed axis micro-motion retraction amplitude preset by the machine tool CNC system or set by the operator under the current working conditions. This value serves as the starting point for adaptive adjustment and can be input from the human-machine interface of the machine tool CNC system.

[0143] Adaptive gain adjustment coefficient It is a dimensionless coefficient used to adjust the sensitivity and response strength of adaptive control; its role is to improve the robustness of the control system. and chip removal efficiency coefficient The degree of response to changes, in order to avoid over- or under-adjustment and ensure the stability of the system response, can be determined through extensive experimental calibration.

[0144] Amplitude Expansion Index It is a positive even exponent greater than 1, and its function is to introduce nonlinear adjustment characteristics, so that when the system state deteriorates (i.e., robustness or chip removal efficiency decreases), it can produce a more significant yield amplitude adjustment, thereby enhancing the anti-jamming effect. This exponent can also be determined through experimental calibration, empirical setting, or optimization algorithms. (Robustness coefficient) This is a dimensionless coefficient used to quantify the dynamic stability of the saw blade 2 of a vertical band saw during the sawing process. It comprehensively reflects the stiffness health of the saw blade 2, the tightness of the clamping, and the impact of external dynamic disturbances on the sawing process. It is a key indicator for assessing the risk of saw blade 2 jamming. This coefficient is obtained by the robustness assessment unit through modeling and calculation based on data from multiple dimensions, including the stiffness health coefficient of the saw blade 2, the clamping and dynamic disturbance coefficients, the high-frequency impact energy of the guide arm, and the slip rate. Chip removal efficiency coefficient. It is a dimensionless coefficient used to quantify the actual working effect of the cooling and chip removal system of a vertical band saw. This coefficient is obtained by the cooling and chip removal evaluation unit through modeling and calculation based on data such as the scalar value of tooth tip temperature rise, the root mean square value of pressure pulsation, and the rate of change of cutting fluid conductivity.

[0145] The adaptive control unit first obtains the current yield amplitude setpoint from the machine tool CNC system. Simultaneously, the robustness assessment unit and the cooling and chip removal assessment unit respectively calculated the robustness coefficient reflecting the dynamic health status of the saw blade 2. And the chip removal efficiency coefficient, which reflects the cooling and chip removal effect. The product of these two coefficients The overall operating state of the sawing system is comprehensively quantified. When this product term decreases (indicating a decrease in system robustness or chip removal efficiency), it is expressed through an exponential function. The effect will cause the target feed axis to fret back amplitude. Compared to Significantly increased adaptive gain adjustment coefficient and amplitude expansion index These factors together determine the sensitivity and nonlinear amplification effect of this adjustment.

[0146] Robustness coefficients obtained by fusing and modeling multi-source heterogeneous data and chip removal efficiency coefficient By incorporating this into the adjustment model of the yield amplitude, this scheme overcomes the lag and limitations of traditional single threshold monitoring, and achieves refined and forward-looking control of the sawing process, thereby effectively avoiding the problem of saw tooth breakage or reduced processing efficiency caused by improper yield amplitude of saw blade 2 under complex working conditions.

[0147] The formula for the target feed axis micro-motion yielding frequency model output by the adaptive control unit is:

[0148]

[0149] In the formula, The target feed axis micro-adjustment frequency, in units of ; The setpoint for the retraction frequency is currently read by the CNC system of the machine tool, in units of... ; The minimum safe micro-motion frequency set for the system to prevent servo synchronization loss, in units of ; This is the frequency gain adjustment coefficient, which is dimensionless. This is a very small positive compensation amount used to prevent system overflow when divided by zero.

[0150] in, This represents the desired frequency of the feed axis's micro-movement retraction operation. This frequency is an output value dynamically adjusted by the system based on real-time operating conditions. It determines the speed at which the saw blade 2 performs periodic micro-retractions during the cutting process. A higher retraction frequency means more frequent micro-retractions, which may be used for fine control of the cutting process or to address slight jamming tendencies; while a lower retraction frequency may mean a longer continuous cutting time. This refers to the feed axis micro-motion and retraction frequency currently set by the machine tool's CNC system. This value is usually manually input by the operator based on experience or preset process parameters. It serves as the basic reference point for adaptive control, and the system will make dynamic adjustments based on this. It introduces an extremely low safety micro-motion frequency baseline, whose core function is to provide a physical protection boundary under extreme operating conditions, ensuring that no matter how deteriorate the operating conditions, the system can still maintain a minimum level of anti-jamming and yielding action, completely avoiding control system shock. It is a dimensionless frequency gain adjustment coefficient, whose function is to adjust the response sensitivity of the adaptive control system to changes in operating conditions. This coefficient determines the robustness coefficient. and cooling chip removal efficiency coefficient When a change occurs, the rate at which the frequency decays. It is a very small positive compensation amount, the main purpose of which is to prevent extremely poor cooling and chip removal efficiency under extreme operating conditions. The computational overflow occurs when the value approaches zero, ensuring the numerical stability of the control algorithm under all operating conditions.

[0151] The solution in this application introduces a dynamic robustness coefficient. With cooling chip removal efficiency coefficient A closed-loop adaptive adjustment mechanism for the feed axis micro-motion retraction frequency was constructed. This mechanism first reads the current retraction frequency setpoint from the machine tool CNC system. As a baseline, the system then utilizes robustness coefficients. and cooling chip removal efficiency coefficient The nonlinear functional relationship is used to dynamically correct the frequency, where the robustness coefficient is... This comprehensively reflects the current stiffness health status of saw blade 2 and the level of external disturbance. When robustness decreases, it causes... As the value increases, the system, through the operation of the negative exponential function term, causes the exponential decay multiplier to decrease sharply, thereby actively increasing the micro-motion yield frequency of the target feed axis. Towards the minimum safe micro-motion frequency baseline By employing a smooth approximation attenuation strategy, this frequency attenuation approach, which introduces a safety baseline, not only helps reduce the dynamic excitation of saw blade 2 in a low robustness state, avoids resonance, and provides the system with a longer response time to cope with potential jamming risks, but also ensures that the system maintains a minimum anti-jamming retreat action under extremely harsh working conditions.

[0152] Meanwhile, the cooling chip removal efficiency coefficient As a denominator term, it reflects the limiting effect of chip removal status on frequency adjustment. When chip removal efficiency is low, The value will also increase, leading to fast towards Approaching the bottom line, this bottom-line protection adjustment helps to reduce the cutting amount per unit time or change the cutting rhythm when chip removal is not smooth, thus giving more time for the cutting fluid to penetrate and for the chips to be removed, thereby improving the chip removal environment in the cutting area and preventing jamming caused by chip accumulation. The solution of this application can intelligently and safely adjust the feed axis micro-movement retraction frequency according to the real-time health status of the saw blade 2 and the dynamic changes in the cutting environment, thereby realizing active optimization control of the sawing process.

[0153] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A vertical band saw blade anti-jamming system, comprising a band saw body (1) and a saw blade (2) disposed on the band saw body (1), characterized in that: It also includes a data acquisition module and a control execution module that is communicatively connected to the data acquisition module; The data acquisition module includes an infrared thermal imager, a dynamic pressure sensor, a conductivity electrode, a tension cylinder pressure transmitter, a miniature tensile and compressive sensor, a flexible array pressure film sensor, a sound level meter, a low-frequency velocity vibration sensor, an accelerometer, and a machine tool CNC system, all installed on the machine tool. The control execution module includes: The cooling chip removal evaluation unit constructs a cooling chip removal efficiency model based on the tooth tip temperature rise scalar, the root mean square value of pressure pulsation, and the rate of change of electrical conductivity, and obtains the cooling chip removal efficiency coefficient. The stiffness assessment unit constructs a stiffness health model of the saw blade (2) based on the peak-valley difference of the tension force microwave, the natural frequency offset and the lateral clamping force, and obtains the stiffness health coefficient of the saw blade (2). The disturbance assessment unit constructs a clamping and dynamic disturbance model based on the mean and standard deviation of the contact surface pressure, the noise frequency band energy, and the foundation vibration velocity, and obtains the clamping and dynamic disturbance coefficients. The robustness assessment unit constructs a dynamic robustness model based on the stiffness health coefficient, disturbance coefficient, high-frequency impact energy of the guide arm, and slip rate, and obtains the robustness coefficient. The adaptive control unit constructs an optimization model based on the robustness coefficient and chip removal efficiency coefficient, and outputs the target feed axis micro-motion yielding amplitude and frequency.

2. The anti-jamming system for vertical band saw blades according to claim 1, characterized in that: When acquiring dynamic physical characteristics, the data acquisition module uniformly defines an observation time window as a macroscopic benchmark reference system. The observation time window is based on the single cutting cycle of the saw blade (2) rotation, and its time length value is uniquely determined by the ratio of the total ring length of the band saw blade (2) preset by the system to the real-time linear velocity of the saw blade read by the CNC system of the machine tool.

3. The anti-jamming system for vertical band saw blades according to claim 2, characterized in that: The cooling chip removal efficiency model in the cooling chip removal evaluation unit is constructed by the following logic: the tooth tip temperature rise scalar, the root mean square value of pressure pulsation, and the rate of change of conductivity are normalized dimensionlessly based on the preset benchmark critical parameters of the system; the normalized characteristic quantities are weighted and aggregated, and the cooling chip removal efficiency coefficient is output through a preset nonlinear decay function; when the comprehensive deviation of the characteristic quantities increases, the cooling chip removal efficiency coefficient approaches zero nonlinearly and rapidly to characterize the sudden deterioration of the cutting lubricant film.

4. The anti-jamming system for vertical band saw blades according to claim 3, characterized in that: The specific methods for acquiring the characteristic quantities in the cooling chip removal evaluation unit include: the tooth tip temperature rise scalar is represented as the difference between the highest region temperature extracted from the region of interest of the sawing exit scanned by the infrared thermal imager and the ambient reference temperature within the observation time window; the pressure pulsation root mean square value is represented as the root mean square characteristic value of the transient pressure fluctuation collected by the dynamic pressure sensor within the observation time window; the conductivity change rate is represented as the absolute value of the change obtained by discretely differentiating the signal collected by the conductivity electrode with the observation time window as the step size, in order to filter out high-frequency transient noise interference.

5. The anti-jamming system for vertical band saw blades according to claim 4, characterized in that: The stiffness health model of the saw blade (2) in the stiffness assessment unit is constructed by extracting the microwave peak-valley difference of tension force, the natural frequency offset, and the relative deviation of lateral clamping force from the factory standard reference value; based on the nonlinear hysteresis decay law of material fatigue damage accumulation, a mathematical model containing high-power inverse mapping terms is constructed. As the deviation of each characteristic term accumulates, the stiffness health coefficient of the saw blade (2) shows a nonlinear accelerated decay trend.

6. The anti-jamming system for vertical band saw blades according to claim 5, characterized in that: The specific methods for obtaining the characteristic quantities in the stiffness evaluation unit include: the peak-valley difference of the tension force microwave is the difference between the peak value and the valley value of the tension force signal within the observation time window; the natural frequency offset is the absolute value of the difference between the current natural frequency and the initial frequency obtained by self-vibration response identification based on the accelerometer signal; and when calculating the deviation of the lateral clamping force, the absolute value of the difference between the current measured value and the reference value is used to normalize the difference relative to the reference value in order to avoid the abnormal division-by-zero singularity of the operation when the clamping force is lost.

7. The anti-jamming system for vertical band saw blades according to claim 6, characterized in that: The clamping and dynamic disturbance model in the disturbance evaluation unit includes a clamping loosening penalty term and a dynamic disturbance penalty term. The clamping loosening penalty term is based on the mean and standard deviation of the contact surface pressure extracted by the flexible array pressure film sensor to construct a proportional function characterizing the uniformity of force. The dynamic disturbance penalty term superimposes the dimensionless noise frequency band energy and the foundation vibration velocity in the characteristic space using orthogonal norms and attenuates them using a negative exponential function to output the clamping and dynamic disturbance coefficients.

8. The anti-jamming system for vertical band saw blades according to claim 7, characterized in that: The dynamic robustness model in the robustness assessment unit includes a basic robustness term and a dynamic anomaly penalty term. The basic robustness term is constructed based on the geometric mean characteristics of the stiffness health coefficient of the saw blade (2) and the clamping and dynamic disturbance coefficients. The dynamic anomaly penalty term is based on the high-frequency impact energy of the guide arm extracted after high-pass filtering and the slip rate obtained based on the speed difference between the main and driven wheels. It uses a soft saturation limiting function to perform nonlinear suppression mapping to avoid the assessment oscillation caused by high-frequency extreme anomalies.

9. The anti-jamming system for vertical band saw blades according to claim 8, characterized in that: When the optimization model in the adaptive control unit outputs the target feed axis micro-motion backing amplitude, based on the physical chip breaking and avoidance mechanics requirements, it constructs the robustness coefficient and chip removal efficiency coefficient as nonlinear amplifying multipliers acting on the machine tool's current backing amplitude setting value; when the robustness coefficient or chip removal efficiency coefficient decreases to zero and the working condition deteriorates, the target feed axis micro-motion backing amplitude exhibits an exponentially rapid amplification response relative to the current setting value.

10. The anti-jamming system for vertical band saw blades according to claim 9, characterized in that: The formula for the target feed axis micro-motion yielding frequency model output by the adaptive control unit is: In the formula, The target feed axis micro-adjustment frequency, in units of ; The set value of the yield frequency currently read by the CNC system of the machine tool, in units of ; The minimum safe micro-motion frequency set for the system to prevent servo synchronization loss, in units of ; This is the frequency gain adjustment coefficient, which is dimensionless. This is a very small positive compensation amount used to prevent system overflow when divided by zero.

Images