A method and system for milling control of a bearing ring

Abstract

The application belongs to the technical field of milling, and particularly relates to a milling control method and system for bearing rings, which method comprises the following steps: acquiring and analyzing vibration signals and spindle driving power of a processing system in real time, calculating a regenerative effect dominant index for quantifying the severity of chatter based on the ratio of the amplitude of chatter to the amplitude of the passing frequency of a tool tooth, multiplying a nonlinear function of the index and the time integral of the spindle power to construct a dynamic cumulative disturbance factor that can comprehensively reflect the current stable state and the historical cumulative disturbance of the system, and dynamically adjusting the adaptive step of an FXLMS control algorithm according to the disturbance factor, so that more efficient and robust active suppression of processing chatter is realized under the premise of ensuring the stability of the system.

Description

Technical Field

[0001] This invention relates to the field of milling technology. More specifically, this invention relates to a milling control method and system for bearing rings. Background Technology

[0002] Bearing rings are key load-bearing components in precision machinery, especially large or thin-walled bearing rings, whose machining process has extremely stringent requirements for final dimensional accuracy and surface quality. However, due to their relatively low structural rigidity, these workpieces are easily subjected to resonance from the periodic cutting force of the tool during milling, resulting in severe self-excited vibration, i.e., machining chatter.

[0003] Once this chatter occurs, it will cause obvious chatter marks on the workpiece surface, severely deteriorating the surface roughness. More importantly, the dynamically changing cutting force will also cause elastic deformation of the workpiece, directly destroying its roundness and wall thickness uniformity, resulting in substandard product precision.

[0004] To address this challenge, active vibration control technology has been introduced into the milling process. Among them, active vibration control based on the FXLMS (Filtered-X Least Mean Squares) algorithm is considered an effective technical approach. This method monitors machining vibration in real time by arranging sensors and actuators on the system. The controller generates a cancellation signal that is equal in magnitude but opposite in phase to the vibration signal based on the monitoring signal, and drives the actuator to generate a reverse force, thereby suppressing machining chatter.

[0005] However, existing standard FXLMS-based algorithms typically employ a fixed adaptive step size. The weights of the control filter are updated, but in the actual milling process of bearing rings, the machining system is a complex time-varying system. For example, as the cutting progresses, the workpiece mass decreases, the cutting point position changes, and the wear of the tool and the accumulation of machining heat effects will all cause changes in the system transfer function, i.e., the dynamic characteristics of the system.

[0006] Fixed step size Unable to adapt to this time-varying characteristic: if If the value is set too high, the algorithm's updates may be too aggressive when the system undergoes drastic dynamic changes, potentially leading to model mismatch and failure to converge, or even system instability; conversely, if... If the value is set too small, the algorithm converges too slowly and cannot track and effectively suppress rapidly changing chatter in time. Therefore, how to make the step size of the FXLMS algorithm adapt to the dynamic changes in the milling process is a technical problem that needs to be solved urgently. Summary of the Invention

[0007] To address the technical problem that the existing FXLMS algorithm, due to its fixed step size, cannot adapt to the time-varying characteristics of milling systems, the present invention provides solutions in the following aspects.

[0008] In a first aspect, the present invention provides a milling control method for bearing rings, comprising: acquiring vibration signals of a machining system and spindle drive power of a machine tool during the milling process of the bearing ring; performing spectral analysis on the vibration signals to obtain a vibration spectrum; identifying the tooth passage frequency amplitude and chatter amplitude in the vibration spectrum, wherein the chatter amplitude is the maximum amplitude other than the tooth passage frequency; calculating a regeneration effect dominance index based on the ratio of the chatter amplitude to the tooth passage frequency amplitude; calculating a system dynamic cumulative disturbance factor based on the product of a nonlinear function of the regeneration effect dominance index and the time integral of the spindle drive power from the start of machining; determining an adaptive step size of an FXLMS control algorithm according to the system dynamic cumulative disturbance factor, wherein the adaptive step size is inversely proportional to the system dynamic cumulative disturbance factor; and applying the adaptive step size to update the control parameters of the FXLMS control algorithm.

[0009] This invention constructs a regenerative effect dominant index by analyzing vibration signals in real time to quantify the severity of chatter. Combined with the time integral of the spindle drive power, it calculates a system dynamic cumulative disturbance factor that comprehensively reflects the current stable state of the system and historical cumulative disturbances. Based on this factor, the adaptive step size of the FXLMS algorithm is dynamically adjusted. When the system is stable, a larger step size is used to achieve rapid convergence. When the system tends to be unstable, the step size is automatically reduced to ensure the stability of the algorithm. Thus, in the complex time-varying milling process, more efficient and robust suppression of machining chatter is achieved.

[0010] Preferably, the method further includes: after the milling process begins, acquiring vibration acceleration signals on the fixture in real time using an accelerometer; and simultaneously acquiring the spindle speed in real time through the CNC system interface. and the number of teeth of the cutting tool .

[0011] Preferably, the method for obtaining the vibration spectrum includes: performing segmented windowing processing on the acquired vibration acceleration signal, and applying a fast Fourier transform to obtain the vibration spectrum within each time window. .

[0012] Preferably, the frequency amplitude of the cutting teeth is the vibration spectrum. medium frequency The amplitude of the spectral peak at that point, the frequency at which the blade passes through. , Main spindle speed This represents the number of teeth on the cutting tool.

[0013] This invention clarifies a method for identifying the amplitude of the cutting tooth passing frequency. By multiplying the spindle speed by the number of cutting teeth, the passing frequency of the cutting teeth is accurately located. This allows for the accurate extraction of the forced vibration component caused by the periodic impact of the cutting tool from a complex vibration spectrum. This provides key benchmark parameters for comparing chatter (i.e., self-excited vibration) with forced vibration and quantifying system stability.

[0014] Preferably, the method for obtaining the flutter amplitude is: searching the vibration spectrum. Find the entire frequency band except for the frequency The maximum value among the spectral peak amplitudes of frequencies other than those mentioned above is denoted as the flutter amplitude. .

[0015] This invention clarifies the method for obtaining the chatter amplitude, namely, finding the maximum spectral peak amplitude in the entire spectrum except for the tooth passing frequency. This can effectively and accurately capture the self-excited chatter component that plays a dominant role in the current processing state, and achieve effective separation of chatter and forced vibration, providing another core input for accurately calculating the dominant index of the regeneration effect.

[0016] Preferably, the nonlinear function of the dominant index of the regeneration effect is a logarithmic function that monotonically increases with respect to the dominant index of the regeneration effect. , for The dominant index is the regeneration effect at any given time.

[0017] This invention employs a monotonically increasing logarithmic function to nonlinearly amplify the dominant exponent of the regeneration effect. This results in a relatively gradual change in the disturbance factor during the initial stage of flutter, when the exponent is small, while the growth of the disturbance factor is more significant when flutter intensifies, when the exponent is large. This improves the sensitivity of the control system to instability trends.

[0018] Preferably, the time integral of the spindle drive power from the start of machining is: , for Spindle drive power at any given moment; Indicates the time from the start of processing to The cumulative integral at each moment; This is the preset processing cycle characteristic time constant.

[0019] This invention accumulates the historical energy input of the machining process, which can indirectly reflect the long-term impact of slow time-varying factors such as tool wear and thermal effects on the dynamic characteristics of the system. This makes the final system dynamic cumulative disturbance factor not only reflect the current state, but also include historical information, making the decision-making basis more comprehensive.

[0020] Preferably, the formula for calculating the adaptive step size is: ;in: for Adaptive step size at each time step; for The system's dynamic cumulative disturbance factor at any given time; The preset maximum allowable step size, and , To control the length of the filter, The filtered reference signal The power, the filtered reference signal Through the secondary channel model vibration spectrum The result of filtering; This is the slope control parameter; This serves as a reference perturbation threshold.

[0021] This invention uses the Sigmoid function to smoothly map the system's dynamically accumulated disturbance factor to the step size value. When the disturbance factor is below the threshold, the step size is close to the preset maximum value to ensure rapid convergence. When the disturbance factor exceeds the threshold, the step size will decrease smoothly and rapidly, thereby adopting a conservative strategy to avoid algorithm divergence, resulting in smoother transitions and more stable control near the state transition point.

[0022] Preferably, applying the adaptive step size to update the control parameters of the FXLMS control algorithm includes: updating the weight vector of the control filter of the FXLMS control algorithm using the adaptive step size, wherein the update method is: ;in, and These are the weight vectors of the control filter before and after the update, respectively; for Adaptive step size at each time step; for The vibration acceleration signal at any given moment; for The filtered reference signal at time 10:00.

[0023] Secondly, the present invention provides a milling control system for bearing rings, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned milling control method for bearing rings is implemented.

[0024] By adopting the above technical solution, a milling control method for bearing rings is generated into a computer program and stored in a memory for loading and execution by a processor. This allows for the creation of a terminal device based on the memory and processor, facilitating its use.

[0025] The beneficial effects of this invention are as follows:

[0026] This invention constructs a regenerative effect dominant index by analyzing vibration signals in real time to quantify the severity of chatter. Combined with the time integral of the spindle drive power, it calculates a system dynamic cumulative disturbance factor that comprehensively reflects the current stable state of the system and historical cumulative disturbances. Based on this factor, the adaptive step size of the FXLMS algorithm is dynamically adjusted. When the system is stable, a larger step size is used to achieve rapid convergence. When the system tends to be unstable, the step size is automatically reduced to ensure the stability of the algorithm. Thus, in the complex time-varying milling process, more efficient and robust suppression of machining chatter is achieved. Attached Figure Description

[0027] Figure 1 This is a flowchart schematically illustrating a milling control method for bearing rings according to the present invention;

[0028] Figure 2 A schematic diagram illustrating a vibration acceleration signal;

[0029] Figure 3 A schematic diagram illustrating the spindle drive power;

[0030] Figure 4 A schematic diagram illustrating the results of the spectral analysis is provided.

[0031] Figure 5 A schematic diagram illustrating the dominant index of the regeneration effect and the system dynamic cumulative disturbance factor;

[0032] Figure 6 A schematic diagram illustrating the adaptive step size is shown. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0035] This invention discloses a milling control method for bearing rings, referring to... Figure 1 This includes steps S1-S4:

[0036] S1: During the milling process of the bearing ring, the vibration signal of the machining system and the spindle drive power of the machine tool are acquired. The vibration signal is subjected to spectrum analysis to obtain the vibration spectrum. The amplitude of the cutting tooth passing frequency and the chatter amplitude are identified in the vibration spectrum.

[0037] To suppress machining chatter, the system is equipped with active vibration control components, including:

[0038] (1) Accelerometer: mounted on the fixture, used to collect vibration acceleration signals during the machining process in real time. .

[0039] (2) Actuator: mounted on the clamp, used to apply a counteracting force.

[0040] (3) CNC system interface: used to read machining status data from the machine tool CNC system in real time, including: spindle speed Number of teeth on the cutting tool and spindle drive power .

[0041] (4) Active vibration controller: As the core of the system, it receives data from the acceleration sensor and the CNC system interface, executes steps S1 to S4, and generates control signals to drive the actuator.

[0042] Specifically, after the milling process begins, the vibration acceleration signal on the fixture is collected in real time by an accelerometer. Simultaneously, the spindle speed is acquired in real time through the CNC system interface. Number of teeth on the cutting tool and spindle drive power .

[0043] Furthermore, the collected data The signal is segmented and windowed, and then Fast Fourier Transform (FFT) is applied to obtain the vibration spectrum within each time window. .

[0044] Furthermore, based on the spindle speed and the number of teeth of the cutting tool Calculate the current pass frequency of the cutter teeth. , In the vibration spectrum In the middle, the frequency was identified. The amplitude of the spectral peak at that point is denoted as the amplitude of the passing frequency of the blade. This amplitude represents the magnitude of the forced vibration caused by the periodic impact of the cutting tool.

[0045] Simultaneously, search the vibration spectrum. Find the entire frequency band except for the frequency The maximum value among the spectral peak amplitudes of frequencies other than those mentioned above is denoted as the flutter amplitude. Maximum value corresponding frequency This is the current primary flutter frequency.

[0046] Thus, by acquiring the processing signals in real time and performing spectral analysis, it is possible to accurately separate the signals representing forced vibrations. and representing self-excited flutter This provides basic data for subsequent evaluation of system stability.

[0047] S2: The dominant index of regeneration effect is calculated based on the ratio of the chatter amplitude to the pass frequency amplitude of the cutter teeth.

[0048] To quantify the severity of self-excited flutter relative to forced vibration, a regeneration effect dominance index was constructed. When the regeneration effect dominates the index A high value indicates that flutter is dominant and the system is unstable; and the regeneration effect dominates the index. Calculated using the following formula:

[0049]

[0050] in, This represents the flutter amplitude. This represents the amplitude of the cutting tooth's passing frequency.

[0051] For example:

[0052] (1) For working condition A, which represents stable cutting: in The frequency amplitude of the blade teeth detected at all times This indicates that forced vibration is significant, and the flutter amplitude is high. This indicates that the flutter is mild; at this point, the regeneration effect dominates the index. A lower value indicates that the system is stable.

[0053] (2) For condition B, which represents flutter: In At a certain moment, due to tool wear, the system tends to become unstable, and the amplitude of the passing frequency of the tool teeth is detected. And flutter amplitude This indicates severe flutter; at this point, the regeneration effect dominates the index. The value increased significantly, indicating that self-excited flutter has become the main component of vibration.

[0054] Thus, by constructing a regenerative effect dominant index, the complex vibration spectrum can be quantified into a single stability index, enabling real-time monitoring of the system's instability trend.

[0055] S3: The dynamic cumulative disturbance factor of the system is calculated by multiplying the nonlinear function of the regeneration effect dominant index with the time integral of the spindle drive power from the start of processing.

[0056] It should be noted that further calculation of the system's dynamic cumulative disturbance factor is required. This factor is used to characterize the degree of change in the dynamic characteristics of a system, which is mainly caused by the accumulated thermal and mechanical effects during processing. The logic behind its construction is that the degree of accumulated disturbance in the system is not only related to the total input energy but also closely related to the current stable state of the system; this stable state is dominated by the regenerative effect. Characterization, when the system is already on the edge of instability, i.e. At higher levels, even a tiny energy input can cause drastic changes in the system's dynamics.

[0057] Specifically, the system's dynamic cumulative disturbance factor Calculated using the following formula:

[0058]

[0059] in: for The system's dynamic cumulative disturbance factor at any given time; for The dominant index of regeneration effect at any given time; for Spindle drive power at any given moment; Indicates the time from the start of processing to The cumulative integral at each moment; The preset processing cycle characteristic time constant is used to normalize the integration result, making it dimensionless. In this embodiment, the processing cycle characteristic time constant is... It equals the total processing time for a typical workpiece.

[0060] For example:

[0061] (1) For working condition A, which represents stable cutting: in At 30 seconds, the dominant index of the regeneration effect , arrive average power After normalization, it equals 1, therefore the power integral term... for ;at this time, System dynamic cumulative disturbance factor at time 1 =(1+0.182)×0.3=0.355.

[0062] (2) For condition B, which represents flutter: In At the second, the regeneration effect dominates the index. , arrive average power After normalization, it equals 1, therefore the power integral term... for ;at this time, System dynamic cumulative disturbance factor at time 1 =(1+0.182)×1.8=5.025.

[0063] (3) Comparison and It can be seen that, in At that time, due to the dominant index of the regeneration effect The nonlinear amplification effect and longer energy accumulation result in a higher dynamic accumulation perturbation factor in the system. The value is much higher than At any given moment, it accurately reflects that the system dynamics have undergone significant changes and are more sensitive to disturbances.

[0064] Thus, by calculating this cumulative perturbation factor, the accumulated energy of the processing history can be combined with the stability of the current state, providing a more comprehensive basis for decision-making regarding the adaptive step size.

[0065] S4: Determine the adaptive step size of the FXLMS control algorithm based on the system's dynamic cumulative disturbance factor, and apply the adaptive step size to the control parameter update of the FXLMS control algorithm.

[0066] It should be noted that the calculated dominant index of the regeneration effect is used. Dynamically adjust the adaptive step size of the FXLMS algorithm The adjustment logic is as follows: when the system undergoes drastic dynamic changes... When the value is high, a smaller step size should be used. To ensure the stability of the algorithm and avoid divergence due to model mismatch; once the system is stable... When the value is low, a larger step size can be used. To accelerate convergence and achieve rapid tracking and suppression of vibrations; the sigmoid function provides a smooth mapping relationship, when Below the reference threshold At that time, step length Approaching the preset maximum value To achieve rapid tracking; when Exceed At that time, step length This will reduce the uncertainty of the system model smoothly, thus automatically adopting a conservative strategy to avoid algorithm divergence when the uncertainty of the system model increases.

[0067] Specifically, this is achieved through a Sigmoid function, based on the dominant index of the regeneration effect. Generate adaptive step size The specific calculation formula is as follows:

[0068]

[0069] in: for Adaptive step size at each time step; for The system's dynamic cumulative disturbance factor at any given time; The preset maximum allowable step size is the maximum update rate that the system can allow when it is most stable. , To control the length of the filter, The filtered reference signal The power; The slope control parameter determines the sensitivity of the step size change; The reference perturbation threshold represents the critical point at which the system dynamics begin to change significantly; when Exceed At this time, the step size should be significantly reduced.

[0070] in, Filtered reference signal at time 1 Through the secondary channel model vibration spectrum The result of filtering; the secondary channel refers to the physical response path from the actuator to the accelerometer, including the actuator itself, the fixture structure, and the dynamic characteristics of the sensor; secondary channel model. It is a digital filter used to simulate this real physical path in the algorithm. The method for obtaining it is as follows: Before active control begins, the system performs a system identification. During this stage, the controller drives the actuator to emit a known broadband signal, such as white noise, while the accelerometer measures the response. By analyzing the relationship between the input white noise and the output response, the filter coefficients simulating the path are calculated using the standard LMS algorithm, thus obtaining the secondary channel model. .

[0071] For slope control parameters :like If the value is too large, the Sigmoid function curve becomes very steep, once... Just over Adaptive step size It will immediately and drastically change from Dropping to near 0 provides the fastest protective response, but it may also... around Fluctuations cause step size jitter; if If the step size is too small, the sigmoid function curve is very flat, resulting in an adaptive step size. Will follow The slope control parameter decreases slowly and gradually as it increases, providing a smoother transition. However, this may also lead to the step size decreasing too slowly when the system dynamically deteriorates, reducing algorithm stability. Therefore, the slope control parameter... The value range is [0.5, 10]. In this embodiment, the slope control parameter is... Setting it to 2 provides a relatively responsive yet still smooth transition.

[0072] For reference perturbation threshold : It is the center point of the Sigmoid function, when When, adaptive step size Exactly equal to ,if If the setting is too low, the system will be overly conservative; even slight perturbations will cause it to drastically reduce the step size, resulting in slow convergence and poor suppression. If the setting is too high, the system will become overly aggressive and will wait... If the step size is only reduced when the system is already close to flutter, it's too late; the algorithm may have already diverged due to model mismatch. Therefore, a reference perturbation threshold should be used. The value range is [1, 4.5]. In this embodiment, the reference perturbation threshold is used. Setting it to 3 provides a reasonable buffer between the stable value and the flutter value, enabling it to respond quickly to a stable state while also taking a conservative approach when flutter begins to accumulate.

[0073] For example, hour:

[0074] (1) In At any given time, the system dynamically accumulates the disturbance factor. ,but Adaptive step size at time ,at this time much smaller Adaptive step size Close to the maximum value The algorithm converges quickly.

[0075] (2) In At any given moment, the system dynamically accumulates the disturbance factor. ,but Adaptive step size at time ,at this time Exceeded Adaptive step size The size is significantly reduced, thereby decreasing the algorithm's update rate and ensuring its stability when the uncertainty of the system model increases.

[0076] Finally, this adaptive step size This is applied to the weight update formula of the FXLMS algorithm, replacing the original fixed step size. The new weight update formula is:

[0077]

[0078] in, and These are the weight vectors of the control filter before and after the update, respectively; for Adaptive step size at each time step; for The vibration acceleration signal at any given moment; for The filtered reference signal at time 10:00.

[0079] Thus, by generating an adaptive step size negatively correlated with the system disturbance state and applying it to the FXLMS algorithm, the control system achieves the fastest convergence speed while ensuring stability, thereby effectively suppressing machining chatter in the time-varying process and ultimately ensuring the machining quality of the bearing ring.

[0080] For example: Figure 2 This is a schematic diagram of vibration acceleration signals. Figure 3 This is a schematic diagram of the spindle drive power. At 1.0s, the amplitude of the vibration acceleration signal increases significantly, and the spindle drive power also increases accordingly. Figure 4 This is a schematic diagram of the spectrum analysis results, where the frequency amplitude of the blade remains basically constant, while the flutter amplitude is... =Rise sharply after 1.0s; Figure 5 This diagram illustrates the dominant index of the regeneration effect and the system's dynamic cumulative disturbance factor. The trend of the dominant index of the regeneration effect is consistent with the amplitude of the tooth-passing frequency. =After 1.0s, it rises sharply. The system's dynamic cumulative disturbance factor simultaneously considers the regeneration effect dominant index and the power integral. Before 1.0s, the regeneration effect dominated the slow growth of the exponent. =1.0s later, due to the simultaneous increase in the regeneration effect dominance index and the principal shaft driving power, the growth slope of the system's dynamic cumulative disturbance factor becomes significantly steeper, and in Successfully crossed near s =3.0 threshold; Figure 6 This is a schematic diagram of the adaptive step size. Before s, , When the term approaches 0, the adaptive step size remains at its maximum value. Nearby, in After s, , The rapid increase in the term causes the adaptive step size to be quickly reduced.

[0081] In summary, this approach achieves rapid convergence using a larger step size when the system is stable, and automatically reduces the step size to ensure the stability of the control algorithm when the system becomes unstable.

[0082] The present invention also discloses a milling control system for bearing rings, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a milling control method for bearing rings according to the present invention.

[0083] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

Claims

1. A milling control method for bearing rings, characterized in that, include: During the milling process of the bearing ring, the vibration signal of the machining system and the spindle drive power of the machine tool are acquired; The vibration signal is subjected to spectral analysis to obtain the vibration spectrum; the amplitude of the tooth passing frequency and the chatter amplitude are identified in the vibration spectrum, where the chatter amplitude is the maximum amplitude other than the tooth passing frequency; The dominant index of regeneration effect is calculated based on the ratio of the flutter amplitude to the pass frequency amplitude of the cutter teeth. The dynamic cumulative disturbance factor of the system is calculated by multiplying the nonlinear function of the dominant index of the regeneration effect with the time integral of the spindle drive power from the start of machining. The adaptive step size of the FXLMS control algorithm is determined based on the system's dynamic cumulative disturbance factor, wherein the adaptive step size is inversely proportional to the system's dynamic cumulative disturbance factor; and the adaptive step size is applied to the update of the control parameters of the FXLMS control algorithm. The formula for calculating the adaptive step size is: ； for Adaptive step size at each time step; for The system's dynamic cumulative disturbance factor at any given time; The preset maximum allowable step size, and , To control the length of the filter, The filtered reference signal The power, the filtered reference signal Through the secondary channel model vibration spectrum The result of filtering; This is the slope control parameter; This serves as a reference perturbation threshold.

2. The milling control method for bearing rings according to claim 1, characterized in that, The method further includes: After the milling process begins, the vibration acceleration signal on the fixture is collected in real time by an accelerometer; at the same time, the spindle speed is acquired in real time through the CNC system interface. and the number of teeth of the cutting tool .

3. The milling control method for bearing rings according to claim 2, characterized in that, The method for obtaining the vibration spectrum includes: The acquired vibration acceleration signals are segmented and windowed, and then a fast Fourier transform is applied to obtain the vibration spectrum within each time window. .

4. The milling control method for bearing rings according to claim 1, characterized in that, The cutting teeth pass through a vibration spectrum with a frequency amplitude. medium frequency The amplitude of the spectral peak at that point, the frequency at which the blade passes through. , Main spindle speed This represents the number of teeth on the cutting tool.

5. The milling control method for bearing rings according to claim 1, characterized in that, The method for obtaining the flutter amplitude is as follows: Search vibration spectrum Find the entire frequency band except for the frequency The maximum value among the spectral peak amplitudes of frequencies other than those mentioned above is denoted as the flutter amplitude. .

6. The milling control method for bearing rings according to claim 1, characterized in that, The nonlinear function of the dominant index of the regeneration effect is a monotonically increasing logarithmic function of the dominant index of the regeneration effect. , for The dominant index is the regeneration effect at any given time.

7. The milling control method for bearing rings according to claim 1, characterized in that, The spindle drive power, integral over time from the start of machining, is: , for Spindle drive power at any given moment; Indicates the time from the start of processing to The cumulative integral at each moment; This is the preset processing cycle characteristic time constant.

8. The milling control method for bearing rings according to claim 1, characterized in that, Applying the adaptive step size to the control parameter update of the FXLMS control algorithm includes: The weight vector of the control filter in the FXLMS control algorithm is updated using the adaptive step size, and the update method is as follows: ； in, and These are the weight vectors of the control filter before and after the update, respectively; for Adaptive step size at each time step; for The vibration acceleration signal at any given moment; for The filtered reference signal at time 10:

00.

9. A milling control system for bearing rings, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement a milling control method for bearing rings according to any one of claims 1-8.

Images