A collaborative control system and method for a linear guide grinding machine

By constructing a thermo-coupled four-terminal mechanical impedance network and cascaded transfer functions, the coupling state of grinding force and heat is quantified in real time, and the excitation source of feed rate adjustment is predicted and suppressed. This solves the problem of cross-coupling between grinding force and grinding heat in the grinding machine control system, and improves machining accuracy and stability.

CN123103522APending Publication Date: 2026-11-17NANJING NANTE PRECISE MASCH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611181552.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-05
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing grinding machine control systems cannot effectively identify and respond to the cross-coupling characteristics between grinding force and grinding heat, resulting in increased machining vibration and a lack of rapid response and model update mechanisms, which affects machining accuracy and stability.

Method used

A thermo-coupled four-terminal mechanical impedance network is constructed. The mutual impedance parameters are quantified in real time through an identification module. The response of the feed rate adjustment is predicted by a cascaded transfer function, thereby achieving synergistic optimization of grinding force control and vibration suppression.

Benefits of technology

It improves the machining accuracy and stability of linear guide grinding machines. By quantifying the thermo-mechanical coupling state in real time, it predicts and suppresses the excitation source of the feed speed adjustment, and achieves coordinated optimization control of grinding force and vibration.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123103522A_ABST
    Figure CN123103522A_ABST
Patent Text Reader

Abstract

This application relates to the technical field of intelligent control of CNC grinding machines, and discloses a collaborative control system and method for linear guide grinding machines. It includes a four-terminal impedance module that establishes a thermo-coupled four-terminal mechanical impedance network using the grinding force port and thermal port as its two ends, outputting an impedance matrix. An identification module calculates the current values ​​of the mutual impedance parameters in real time and generates a coupling deviation trigger signal based on the rate of change of the deviation from the reference value. A cascaded collaborative module models the force control subsystem and vibration support subsystem as two two-port networks respectively, calculates the overall transfer function through cascade characteristics, and a collaborative controller controls the cascaded collaborative module to re-identify the overall transfer function when a coupling deviation trigger signal is received. When outputting the feed rate adjustment amount, it predicts the response amplitude at the vibration port based on the current overall transfer function. When the comprehensive vibration index exceeds a safety threshold, it imposes amplitude constraints on the adjustment amount, realizing the joint control of grinding force and vibration, and improving machining stability and accuracy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the technical field of intelligent control of CNC grinding machines, and discloses a collaborative control system and method for linear guide grinding machines. Background Technology

[0002] In the grinding process of an online guide grinding machine, there is an inherent coupling relationship between grinding force and grinding heat. The work done by the grinding force is converted into grinding heat, and the grinding heat, in turn, affects the generation and transmission of grinding force by changing the mechanical properties of the material and the contact state of the grinding zone. This force-heat coupling characteristic is not constant, but continuously evolves with changes in grinding depth, feed rate, cooling conditions, and workpiece material.

[0003] Currently, existing grinding machine control systems typically treat grinding force control and vibration suppression as two independent control objectives: the force control loop maintains a constant grinding force by adjusting the feed rate, while the vibration monitoring system independently collects vibration signals and alarms or shuts down the machine when limits are exceeded. The lack of a mathematical model describing the transmission relationship between force control output and vibration response between these two subsystems means that the force control loop cannot predict the response at the vibration port when outputting feed rate adjustments. This adjustment itself may become a source of excitation for the system, exacerbating machining vibration. Furthermore, existing technologies primarily rely on independent acquisition and threshold comparison of temperature and force signals for identifying force-thermal coupling states. This only determines whether preset safety boundaries are exceeded, failing to quantitatively characterize the cross-coupling characteristics between the grinding force and thermal ports. More critically, when the thermo-mechanical coupling state changes abruptly due to variations in grinding conditions, the force-thermal coupling transmission relationship changes accordingly, leading to a mismatch between the original system model and control parameters and the actual working conditions. However, existing control strategies lack rapid response and model update mechanisms for these coupling abrupt changes. Summary of the Invention

[0004] The purpose of this section is to outline some aspects of the embodiments of this application and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents, and such simplifications or omissions should not be construed as limiting the scope of this application.

[0005] The specific inventive objective of this application is as follows:

[0006] This application achieves online quantitative sensing of the force-thermal interaction state by constructing a thermo-coupled four-terminal mechanical impedance network. Under the deviation change rate triggered re-identification mechanism, the feed rate adjustment is constrained by the output amplitude of the cascaded transfer function of the two-port network, thereby achieving synergistic optimization of grinding force control and vibration suppression and improving the machining accuracy and stability of the linear guide grinding machine.

[0007] To address the aforementioned technical problems, this application provides a collaborative control system and method for a linear guide grinding machine.

[0008] On the one hand, this application provides a collaborative control system and method for a linear guide grinding machine, including:

[0009] The four-terminal impedance module uses the grinding force port and the heat port as the two ends to establish a four-terminal mechanical impedance network for the thermo-coupling of the linear guide grinding machine, and outputs an impedance matrix containing mutual impedance parameters.

[0010] The identification module calculates the current value of the mutual impedance parameter in real time and generates a coupling deviation trigger signal based on the rate of change of the deviation between the current value and the reference value.

[0011] The cascaded collaborative module models the force control subsystem as a first two-port network and the vibration support subsystem as a second two-port network, and calculates the overall transfer function through the cascaded characteristics of the two-port networks;

[0012] The collaborative controller is connected to the identification module and the cascaded collaborative module respectively. When the collaborative controller receives the coupling deviation trigger signal, it controls the cascaded collaborative module to re-identify the overall transfer function. When the force control loop needs to output the feed speed adjustment amount, the collaborative controller predicts the response amplitude of the current speed adjustment amount at the vibration port based on the current overall transfer function, and obtains the comprehensive vibration index based on the predicted response amplitude and the current measured vibration characteristics. When the comprehensive vibration index exceeds the safety threshold, the amplitude of the feed speed adjustment amount is actively reduced until the comprehensive vibration index corresponding to the reduced feed speed adjustment amount is not higher than the safety threshold.

[0013] As a preferred embodiment of the linear guide grinding machine collaborative control system and method of this application, wherein:

[0014] The grinding force port is the first port of the four-terminal mechanical impedance network, used to connect the mechanical signal variables of the grinding force during the grinding process. The mechanical signal variables include the grinding force signal and the displacement signal caused by the grinding force.

[0015] The hot port is the second port of the four-terminal mechanical impedance network, used to connect the thermal signal variables of grinding heat during the grinding process. The thermal signal variables include the temperature signal of the grinding zone and the heat flow signal.

[0016] The grinding force port and the heat port serve as the input and output ports of the four-terminal mechanical impedance network, respectively. The four-terminal mechanical impedance network uses the mechanical signal variable of the grinding force port as the input excitation and the thermal signal variable of the heat port as the output response to establish a coupling and transmission relationship between the grinding force field and the temperature field.

[0017] As a preferred embodiment of the linear guide grinding machine collaborative control system and method of this application, wherein:

[0018] The mechanical signal variables of the grinding force port and the thermal signal variables of the heat port are collected respectively. The mechanical signal variables include the grinding force signal F and the displacement X under the action of the grinding force. The thermal signal variables include the grinding zone temperature T and the heat flux Q.

[0019] Using the signal variables of the grinding force port and the heat port as boundary conditions, the network equation of the four-terminal mechanical impedance network is established.

[0020] The diagonal elements of the impedance matrix and The self-impedance parameters of the grinding force port and the heat port are respectively, characterizing the intrinsic transmission characteristics between the physical quantities of each port; the off-diagonal elements of the impedance matrix and These are mutual impedance parameters. Characterizes the coupling and transmission properties of the reaction between the temperature field and the grinding force field. Characterizes the coupling and transfer properties of the displacement field to the heat flow field. The change in the amount of grinding force indirectly reflects the coupling efficiency of grinding force work being converted into grinding heat.

[0021] A preset excitation signal is applied to the grinding force port and the heat port through the network equation, and the response signal is collected. The parameter values ​​of each element in the impedance matrix are calculated by the identification algorithm. The parameter values ​​include self-impedance parameters and mutual impedance parameters.

[0022] The calculated impedance matrix of the self-impedance parameter and the mutual impedance parameter is used as the output of the four-terminal mechanical impedance network.

[0023] As a preferred embodiment of the linear guide grinding machine collaborative control system and method of this application, wherein:

[0024] The identification module takes the impedance matrix output by the four-terminal impedance module as input and continuously captures the real-time response data of the grinding force port and the heat port during the dynamic cutting process when the grinding machine is in continuous processing state.

[0025] The identification module uses the force port excitation signal and the heat port response signal acquired in the current control cycle as information input, and uses the parameter estimation results of the historical cycle as a priori reference. By recursively comparing and correcting the information input with the priori reference cycle by cycle, the estimated value of the mutual impedance parameter is updated.

[0026] As a preferred embodiment of the linear guide grinding machine collaborative control system and method of this application, wherein:

[0027] The identification module also stores the reference value of the mutual impedance parameter of the linear guide grinding machine under the calibration conditions. The reference value is used to characterize the reference characteristics of the grinding machine under normal thermo-coupling conditions. After the identification module obtains the current value of the mutual impedance parameter in real time, it first calculates the instantaneous deviation between the current value and the reference value.

[0028] The identification module performs a time-series comparison between the instantaneous deviation at the current moment and the instantaneous deviation at historical moments, and calculates the rate of change of the instantaneous deviation over time.

[0029] When the rate of change exceeds the preset rate tolerance range, it is determined that the current thermo-coupling state is undergoing a sudden deviation that is not gradual.

[0030] Simultaneously with the determination outputting a positive result, the identification module immediately generates the coupling deviation trigger signal and sends the trigger signal to the cooperative controller to trigger the cooperative controller to initiate a re-identification of the overall transfer function.

[0031] As a preferred embodiment of the linear guide grinding machine collaborative control system and method of this application, wherein:

[0032] The overall transfer function is used to describe the response generated at the vibration port after the feed speed command passes through the force control link and the vibration transmission link;

[0033] The cascaded collaborative module defines the physical variables of the input port of the force control subsystem as the feed speed command signal output after being adjusted by the collaborative controller, and defines the physical variables of the output port as the dynamic grinding force signal actually output after the feed speed command is acted upon by the force control execution link. It also establishes a first transmission correlation between the input port and the output port, reflecting the transmission characteristics from the speed command to the grinding force, thereby abstracting the force control subsystem as a whole into a first two-port network with unidirectional signal transmission characteristics.

[0034] The cascaded collaborative module defines the physical variable of the input port of the vibration support subsystem as the dynamic grinding force signal output by the output port of the force control subsystem, and defines the physical variable of its output port as the vibration response signal generated at the actual vibration monitoring port after the dynamic grinding force is transmitted through the machine tool structure. It also establishes a second transmission correlation between the input port and the output port, reflecting the transmission characteristics from grinding force to vibration response, thereby abstracting the entire vibration support subsystem into a second two-port network with unidirectional signal transmission characteristics.

[0035] As a preferred embodiment of the linear guide grinding machine collaborative control system and method of this application, wherein:

[0036] After completing the modeling of the first two-port network and the second two-port network, the cascaded coordination module obtains the first transmission correlation parameter corresponding to the first two-port network and the second transmission correlation parameter corresponding to the second two-port network. The cascaded coordination module cascades the second two-port network after the first two-port network according to the direction of signal flow in the actual physical link, so that the output port of the first two-port network and the input port of the second two-port network logically coincide into the same cascaded node.

[0037] The cascaded collaborative module first maps the feed speed command signal into a grinding force signal at an intermediate node through the first transmission association relationship, and then maps the grinding force signal at the intermediate node into a vibration response signal at the final node through the second transmission association relationship.

[0038] As a preferred embodiment of the linear guide grinding machine collaborative control system and method of this application, wherein:

[0039] When the cooperative controller receives the coupling deviation trigger signal, it controls the cascaded cooperative module to re-identify the overall transfer function. The cooperative controller monitors the coupling deviation trigger signal output by the identification module in real time. The trigger signal is a pulse-type or level-type trigger identifier, which is used to indicate to the cooperative controller that the current thermal coupling state has suddenly deviated.

[0040] When the cooperative controller detects the rising edge of the coupling deviation trigger signal, it immediately sends a forced re-identification command to the cascaded cooperative module. Upon receiving the forced re-identification command, the cascaded cooperative module suspends the current control auxiliary function based on the old overall transfer function, re-acquires the real-time input-output response data of the force control subsystem and the vibration support subsystem under the current operating conditions, and re-executes the transmission correlation identification of the first two-port network and the second two-port network and the composite derivation of the overall transfer function based on the re-acquired data to obtain an updated overall transfer function that matches the current thermo-coupling state.

[0041] After the cascaded collaborative module completes the re-identification of the overall transfer function, the collaborative controller replaces the old overall transfer function with the updated overall transfer function.

[0042] As a preferred embodiment of the linear guide grinding machine collaborative control system and method of this application, wherein:

[0043] After the force control loop inside the cooperative controller completes the calculation of the feed rate adjustment amount for the current control cycle, but before outputting the adjustment amount to the force control subsystem, the cooperative controller first intercepts the feed rate adjustment amount and applies it as an input excitation to the overall transfer function currently held by the cascaded cooperative module. Through the mapping relationship of the overall transfer function, the pre-response amplitude that the feed rate adjustment amount will generate at the vibration monitoring port after passing through the force control link and the vibration transmission link is calculated.

[0044] The collaborative controller has a built-in comprehensive vibration index evaluation logic. The evaluation logic integrates the pre-response amplitude with other vibration characteristic parameters measured at the current vibration monitoring port to obtain the comprehensive vibration index value that is expected to be caused by the actual output of the feed speed adjustment. The collaborative controller compares the comprehensive vibration index value with a pre-stored safety threshold. If the comprehensive vibration index value is lower than or equal to the safety threshold, the collaborative controller determines that the feed speed adjustment is a safe adjustment and allows it to be output normally to the force control subsystem.

[0045] If the comprehensive vibration index value is higher than the safety threshold, the cooperative controller determines that the feed rate adjustment is a risk adjustment and actively reduces the amplitude of the feed rate adjustment before outputting it to the force control subsystem. The cooperative controller only outputs the reduced feed rate adjustment to the force control subsystem when the comprehensive vibration index value corresponding to the reduced feed rate adjustment after remapping by the overall transfer function is not higher than the safety threshold.

[0046] A collaborative control method for a linear guide grinding machine includes:

[0047] A four-terminal mechanical impedance network with thermal coupling is established using the grinding force port and the heat port as the two ends. An impedance matrix containing mutual impedance parameters is identified and output by applying excitation and acquiring the response.

[0048] Using the impedance matrix as input, the current value of the mutual impedance parameter is updated recursively cycle by cycle. The instantaneous deviation between the current value and the reference value and its rate of change are calculated. When the rate of change exceeds a preset tolerance range, a coupling deviation trigger signal is generated.

[0049] The force control subsystem is modeled as a first two-port network, the vibration support subsystem is modeled as a second two-port network, and the first two-port network and the second two-port network are cascaded and combined along the signal flow direction to derive the overall transfer function from the feed speed command to the vibration response.

[0050] When the coupling deviation trigger signal is received, the current control auxiliary function is paused, data is re-acquired, and the cascade modeling and transfer function calculation steps are re-executed to replace the old overall transfer function with the updated overall transfer function.

[0051] Before outputting the feed rate adjustment amount, it is input into the current global transfer function to calculate the pre-response amplitude at the vibration port, and the comprehensive vibration index is evaluated accordingly. When the comprehensive vibration index exceeds the safety threshold, the amplitude of the feed rate adjustment amount is reduced before output.

[0052] The beneficial effects of this application are as follows:

[0053] This application constructs a four-terminal mechanical impedance network using the grinding force port and the heat port as its two ends. By outputting an impedance matrix containing mutual impedance parameters, the force-thermal coupling relationship is transformed into impedance parameters, which can quantitatively describe the strength and changing trend of the force-thermal coupling. By calculating the current value of the mutual impedance parameters in real time, and generating a coupling deviation trigger signal based on the rate of change of the deviation between the current value and the reference value rather than relying solely on the absolute value of the deviation, the gradual drift and sudden deviation of the thermo-mechanical coupling state are effectively distinguished, avoiding frequent false triggers caused by fluctuations in normal operating conditions and improving the accuracy of abnormal state identification.

[0054] This application models the force control subsystem and the vibration support subsystem as two two-port networks respectively, and derives the overall transfer function through the cascaded characteristics of the two-port networks. The overall transfer function can describe the response characteristics generated at the vibration port after the feed speed command passes through the force control link and the vibration transmission link, so that the influence of the force control output on the vibration can be quantified and predicted before the output.

[0055] Before the actual output of the feed rate adjustment amount, this application predicts the response amplitude at the vibration port based on the current overall transfer function. When the comprehensive vibration index exceeds the safety threshold, it actively imposes amplitude constraints on the adjustment amount, which can suppress it before the adjustment amount becomes an excitation source. This achieves synergistic optimization of grinding force control and vibration suppression, thereby effectively improving the machining accuracy and stability of the linear guide grinding machine. Attached Figure Description

[0056] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained through these drawings without creative effort. Wherein:

[0057] Figure 1 The overall architecture diagram of the linear guide grinding machine collaborative control system provided in this application;

[0058] Figure 2 A flowchart illustrating the construction and identification of a four-terminal mechanical impedance network for a linear guide grinding machine collaborative control system provided in this application;

[0059] Figure 3 A schematic diagram illustrating the two-port network cascading and overall transfer function derivation of a linear guide grinding machine collaborative control system provided in this application;

[0060] Figure 4 This application provides a flowchart of the feed speed adjustment prediction and amplitude constraint control for a linear guide grinding machine collaborative control system.

[0061] Figure 5 Example of PID control parameters for a control panel of a linear guide grinding machine collaborative control system provided in this application;

[0062] Figure 6 The control panel of the linear guide grinding machine collaborative control system provided in this application displays the results of detecting self-impedance and mutual impedance. Detailed Implementation

[0063] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the specific embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0064] Many specific details are set forth in the following description in order to provide a full understanding of this application. However, this application may also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the spirit of this application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0065] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of this application. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0066] Example 1

[0067] like Figure 1 As shown, this embodiment provides a linear guide grinding machine collaborative control system, including:

[0068] The four-terminal impedance module uses the grinding force port and the heat port as the two ends to establish a four-terminal mechanical impedance network for the thermo-coupling of the linear guide grinding machine, and outputs an impedance matrix containing mutual impedance parameters.

[0069] The grinding force port is the first port of the four-terminal mechanical impedance network, used to connect the mechanical signal variables of the grinding force during the grinding process. The mechanical signal variables include the grinding force signal and the displacement signal caused by the grinding force.

[0070] The grinding force signal F is the cutting force generated by the interaction between the grinding wheel and the workpiece during the grinding process. It is used to represent the mechanical excitation of the grinding force port. The grinding force includes the normal component Fn perpendicular to the workpiece surface and the tangential component Ft along the grinding speed direction.

[0071] The grinding force signal F is acquired in real time by a force sensor installed on the grinding wheel spindle support of the linear guide grinding machine's worktable. Specifically, a piezoelectric triaxial force gauge or a strain gauge force sensor can be used, fixed between the workpiece fixture and the worktable. The force generated by the interaction between the grinding wheel and the workpiece is transmitted to the sensor through the fixture. The sensor converts the mechanical signal into a charge or voltage signal, which is then sent to the control system after passing through a charge amplifier and a data acquisition card.

[0072] The grinding force signal F, as one of the input excitation variables of the grinding force port in the four-terminal mechanical impedance network, together with the displacement X, constitutes the boundary condition of the port. The real-time change of the grinding force signal F reflects the change of the grinding state, and the mapping relationship between the heat flow Q generated by its work and the displacement X is the mutual impedance. The coupling and transfer characteristics of grinding force work being converted into grinding heat are described.

[0073] The grinding force port is the first port of the four-terminal mechanical impedance network, used to connect the mechanical signal variables of the grinding force during the grinding process. The mechanical signal variables include the grinding force signal F and the displacement signal X caused by the grinding force.

[0074] The grinding force signal F serves as the input excitation variable of the grinding force port, and the displacement X serves as the output response variable of the port under the same excitation, thus constituting the complete mechanical boundary conditions of the grinding force port.

[0075] Displacement X is the dynamic displacement response generated along the force transmission path of a linear guide grinding machine, such as the grinding head, worktable, or grinding wheel spindle, under the action of grinding force. It is a core physical quantity describing the mechanical response at the grinding force end. This displacement reflects the elastic deformation of the machine tool structure under the excitation of grinding force, and its amplitude and phase are directly related to the magnitude and frequency components of the grinding force.

[0076] Engineering source and detection method: Displacement X is acquired in real time by a displacement sensor installed near the point of application of the grinding force. Specifically, an eddy current displacement sensor or a laser displacement sensor can be used, fixed on a stationary component such as the machine tool bed, with the probe facing the surface to be measured on the grinding head or worktable. This non-contact measurement measures the relative displacement change under the action of the grinding force, and the sampling frequency is synchronized with the force sensor.

[0077] Displacement X and grinding force signal F constitute the mechanical signal variables of the grinding force port. In the four-terminal mechanical impedance network, the ratio between the grinding force signal F and displacement X is the self-impedance parameter. It characterizes the intrinsic transmission characteristics of the grinding force port itself.

[0078] The hot port is the second port of the four-terminal mechanical impedance network, used to connect the thermal signal variables of grinding heat during the grinding process. The thermal signal variables include the temperature signal of the grinding zone and the heat flow signal.

[0079] Temperature T is the transient temperature of the grinding arc zone during the grinding process, used to describe the thermal excitation at the hot end. The grinding arc zone is the area where the grinding wheel contacts the workpiece, and the temperature of the zone reflects the intensity and distribution of grinding heat.

[0080] Temperature T is obtained in real time by a temperature sensor installed near the grinding arc zone.

[0081] Specifically, embedded thermocouples can be used, with the thermocouple hot junction embedded in the workpiece surface, or an infrared thermometer can be used to non-contactly measure the surface temperature of the grinding arc zone. The thermocouple signal is sent to the data acquisition system after cold junction compensation and amplification, while the infrared thermometer directly outputs the temperature signal to capture the transient changes in the temperature of the grinding arc zone.

[0082] Temperature T, as a thermal signal variable at the hot port of the four-terminal mechanical impedance network, together with heat flow Q, constitutes the boundary condition of the thermal domain at this port. The real-time change of temperature T reflects the generation and accumulation of grinding heat, and the mapping relationship between temperature T and the grinding force signal F is the mutual impedance parameter. This describes the reaction coupling and transmission characteristics of the temperature field on the grinding force field.

[0083] Heat flux Q is the rate of heat transfer per unit area through the grinding arc region per unit time. It is used to describe the thermal response of the hot port. The magnitude of heat flux reflects the process of heat transfer from grinding to the workpiece, grinding wheel, grinding fluid and surrounding medium.

[0084] The heat flux Q is measured directly by a heat flux sensor. A preferred example of direct measurement is to use a thin-film heat flux sensor, which is attached to the workpiece surface near the grinding arc area. The sensor outputs a voltage signal by converting the heat flux signal.

[0085] Heat flux Q and temperature T together constitute the complete thermal signal variable at the hot port. In a four-terminal mechanical impedance network, the ratio between temperature T and heat flux Q is the self-impedance parameter. It is used to characterize the heat transfer characteristics of the hot port itself, reflecting the heat diffusion and heat accumulation characteristics of the grinding arc zone.

[0086] The grinding force port and the thermal port serve as the input and output ports of the four-terminal mechanical impedance network, respectively. The four-terminal mechanical impedance network uses the mechanical signal variable of the grinding force port as input excitation and the thermal signal variable of the thermal port as output response to establish a coupling and transmission relationship between the grinding force field and the temperature field. Alternatively, it uses the thermal signal variable of the thermal port as input excitation and the mechanical signal variable of the grinding force port as output response to establish a reaction coupling and transmission relationship between the temperature field and the grinding force field.

[0087] The grinding force port and the heat port serve as the input and output ports of the four-terminal mechanical impedance network, respectively. The four-terminal mechanical impedance network establishes a coupling and transmission relationship between the grinding force field and the temperature field by taking the mechanical signal variables of the grinding force port, including the grinding force signal F and the displacement X, as the input boundary and the thermal signal variables of the heat port, including the temperature T and the heat flux Q, as the output boundary.

[0088] Specifically, the grinding force signal F performs work and is converted into grinding heat, causing the temperature T in the grinding zone to rise. This is determined by the off-diagonal elements of the impedance matrix. Quantitatively characterizing the positive coupling transmission characteristics of the grinding force field to the heat flow field; simultaneously, the increase in temperature T leads to softening of the workpiece material and changes in the coefficient of friction, which in turn alters the magnitude and distribution of the grinding force signal F. The reverse coupling transmission characteristics of heat to force are determined by the off-diagonal elements of the impedance matrix. Quantitative characterization; the cross-coupling transfer characteristics between two ports, i.e., mutual impedance parameters. and The intrinsic transfer characteristics of each port, i.e., its self-impedance parameter. and Together they form a four-terminal impedance matrix.

[0089] like Figure 2 As shown, mechanical signal variables of the grinding force port and thermal signal variables of the heat port are collected respectively. The mechanical signal variables include the grinding force signal F and the displacement X under the action of the grinding force. The thermal signal variables include the grinding zone temperature T and the heat flux Q.

[0090] Using the signal variables of the grinding force port and the heat port as boundary conditions, the network equation of the four-terminal mechanical impedance network is established.

[0091] The network equation is expressed in the form of an impedance matrix as follows: the grinding force signal F and temperature T form a generalized force vector, and the displacement X and heat flow Q form a vector. X and Q can be linearly mapped through a 2×2 order impedance matrix.

[0092] Specifically, the network equation is described as follows: the force vector equals the impedance matrix multiplied by the flow vector. Here, the first element of the force vector is the grinding force signal F, and the second element is the temperature T; the first element of the flow vector is the displacement X, and the second element is the heat flux Q.

[0093] The grinding force signal F and displacement X describe the mechanical energy transfer relationship at the grinding force port, that is, the grinding force is the excitation applied to the machine tool structure, and the displacement is the response generated under the current excitation; the temperature T and heat flow Q describe the thermal energy transfer relationship at the heat port, that is, the temperature is the thermal potential excitation of the grinding arc zone, and the heat flow is the thermal response generated under the current excitation.

[0094] The diagonal elements of the impedance matrix and The self-impedance parameters of the grinding force port and the heat port are respectively, characterizing the intrinsic transmission characteristics between the physical quantities of each port; the off-diagonal elements of the impedance matrix and These are mutual impedance parameters. Characterizes the coupling and transmission properties of the reaction between the temperature field and the grinding force field. Characterizes the coupling and transfer properties of the displacement field to the heat flow field. The change in the amount of grinding force indirectly reflects the coupling efficiency of grinding force work being converted into grinding heat.

[0095] Specifically, Let be the first diagonal element of the impedance matrix, and be the self-impedance parameter of the grinding force port. The intrinsic transmission characteristics of the grinding force port itself are used to characterize the following: the ratio between the grinding force signal F and the displacement X under the condition of zero heat flux boundary (Q=0) at the hot port. =F / X reflects the structural dynamic stiffness characteristics of a linear guide grinding machine under grinding force.

[0096] The second diagonal element of the impedance matrix represents the self-impedance parameter of the hot port. The heat transfer characteristics of the hot port itself are characterized by the ratio between heat flux Q and temperature T under the zero-force boundary condition (F=0) at the grinding force port. =Q / T, this parameter reflects the heat conduction and heat storage characteristics of the grinding arc zone.

[0097] The first off-diagonal element of the impedance matrix is ​​one of the mutual impedance parameters. This is used to characterize the reaction coupling transmission characteristics of the temperature field on the grinding force field, specifically: the response relationship of the grinding force signal F with temperature T under the boundary condition that the displacement X remains constant, i.e. =F / T describes the strength of the coupling effect on the grinding force after the temperature rise in the grinding zone causes the workpiece material to soften and the friction coefficient to change.

[0098] The second off-diagonal element of the impedance matrix is ​​one of the mutual impedance parameters. This is used to characterize the positive coupling transfer characteristics of the grinding force field to the heat flow field, specifically: the response relationship of heat flow Q as displacement X changes under the boundary condition of constant temperature T, i.e. =Q / X, parameter By describing the efficiency of mechanical energy conversion into heat flow, the coupling characteristics of grinding force work being converted into grinding heat are indirectly reflected. The relationship between the grinding force signal F and the temperature T can be derived from... Obtained by inverting the matrix.

[0099] A preset excitation signal is applied to the grinding force port and the heat port through the network equation, and the response signal is collected. The parameter values ​​of each element in the impedance matrix are calculated by the identification algorithm. The parameter values ​​include self-impedance parameters and mutual impedance parameters.

[0100] The calculated impedance matrix of the self-impedance parameter and the mutual impedance parameter is used as the output of the four-terminal mechanical impedance network.

[0101] When the grinding machine is in a static or idling state, a preset sweep frequency force excitation signal is applied to the grinding force port through a signal generator, and a preset heat flow excitation signal is applied to the heat port simultaneously. The excitation signal adopts a sweep frequency sine signal or a pseudo-random binary sequence signal to fully excite the response characteristics of each frequency component.

[0102] The preset heat flow excitation signal can be obtained by applying pulsed heat flow through heating elements or laser heat sources.

[0103] Furthermore, while applying the excitation signal, the force signal F and displacement signal X at the grinding force port, as well as the temperature signal T and heat flow signal Q at the hot port, are simultaneously acquired.

[0104] In this application, a preferred method for implementing the identification algorithm includes: performing a Fast Fourier Transform on the acquired time-domain excitation signal and response signal, converting them to the frequency domain, and then calculating the transfer function value at each frequency point. Specifically, in the network equation F= ·X+ ·T and Q= ·X+ In step T, by applying independent excitation signals to the two ports respectively, two sets of linearly independent input-output data are obtained. These data are then substituted into the above equations and solved simultaneously to calculate the values ​​of the four impedance parameters at each frequency point. The identification algorithm can employ the least squares estimation method.

[0105] It should be noted that those skilled in the art can perform the above identification process either offline during the calibration stage before grinding, or online during the processing by superimposing a small-amplitude sweep frequency signal. Regardless of whether an offline or online method is used, the core of the identification algorithm lies in using two independent excitation signals from the two ports to decouple the four impedance parameters from the coupled equations, thereby obtaining their respective independent estimates.

[0106] Furthermore, the impedance matrix of the calculated self-impedance parameters and mutual impedance parameters is used as the output of the four-terminal mechanical impedance network.

[0107] Specifically, the four impedance parameters calculated by the identification algorithm will be... , , , Assembled into a 2×2 order impedance matrix according to a matrix format, and output to the identification module in a preset data format. In the output impedance matrix, the diagonal elements... and These are the self-impedance parameters, and the off-diagonal elements are... and These are the mutual impedance parameters. This impedance matrix serves as the final output of the four-terminal mechanical impedance network.

[0108] By pairing the grinding force signal F with the displacement X as the boundary condition of the mechanical domain and the temperature T with the heat flow Q as the boundary condition of the thermal domain, and by achieving synchronous acquisition of four signals under the same system clock reference, the phase relationship and timing consistency between the mechanical signal variables and the thermal signal variables are ensured, avoiding timing misalignment between signals in different physical domains caused by inconsistent sampling times of each sensor, thereby improving the accuracy of mutual impedance parameter identification.

[0109] By establishing network equations using a pairing method with grinding force signal F and temperature T as generalized forces and displacement X and heat flow Q as generalized flows, mechanical and thermal variables that originally belonged to different physical domains are incorporated into the same mathematical framework for unified description. This quantifies the coupling efficiency of force-generated heat and the feedback intensity of thermal reaction on force, overcoming the shortcomings of existing technologies that can only generally determine the mutual influence between force and heat but cannot quantify the degree of coupling in each direction.

[0110] By applying independent excitation signals to the two ports respectively, two sets of linearly independent input and output data are obtained, which enables the four impedance parameters to be completely decoupled from the coupled network equations and independently identified, ensuring the uniqueness and accuracy of the estimated values ​​of each impedance parameter.

[0111] The identification module calculates the current value of the mutual impedance parameter in real time and generates a coupling deviation trigger signal based on the rate of change of the deviation between the current value and the reference value.

[0112] Specifically, after the grinding machine enters the continuous grinding process, the identification module receives dynamic response data from the grinding force port and the heat port in real time via the data bus. The real-time response data of the grinding force port includes the real-time measured value of the grinding force signal F and the real-time measured value of the displacement X, which are synchronously collected by a piezoelectric triaxial force gauge installed between the worktable and the workpiece fixture and an eddy current displacement sensor installed between the machine tool bed and the grinding head at a preset sampling frequency and then sent to the identification module. The real-time response data of the heat port includes the real-time measured value of the grinding zone temperature T and the real-time measured value of the heat flow Q, which are synchronously collected by an embedded thermocouple and a heat flow sensor attached to the workpiece surface at a preset sampling frequency and then sent to the identification module.

[0113] The identification module also includes a circular data buffer, which stores the sampled values ​​of four physical quantities (F, X, T, Q) in the most recent N control cycles in chronological order. After the identification module completes the data reception for each cycle, it triggers a parameter update operation.

[0114] The identification module takes the impedance matrix output by the four-terminal impedance module as input and continuously captures the real-time response data of the grinding force port and the heat port during the dynamic cutting process when the grinding machine is in continuous processing state.

[0115] The identification module uses the force port excitation signal and the heat port response signal acquired in the current control cycle as information input, and uses the parameter estimation results of the historical cycle as a priori reference. By recursively comparing and correcting the information input with the priori reference cycle by cycle, the estimated value of the mutual impedance parameter is updated.

[0116] The identification module uses the force port excitation signal and the heat port response signal acquired in the current control cycle as the information input, and uses the parameter estimation results of the historical cycle as the prior reference. By recursively comparing and correcting the information input with the prior reference cycle by cycle, the estimated value of the mutual impedance parameter is updated.

[0117] The information input is a set of real-time data newly acquired within the current control cycle that has not yet been used for parameter updates. Specifically, within the current control cycle t, the information input includes the following four synchronously sampled physical quantities: the current cycle grinding force measurement value F(t) and displacement measurement value X(t) at the grinding force port, and the current cycle temperature measurement value T(t) and heat flow measurement value Q(t) at the heat port. These reflect the actual force-thermal dynamic response state of the grinding machine at the current moment and are the basic data for the identification algorithm to update parameters.

[0118] Among them, the force port variables include the grinding force signal F(t) and displacement X(t), where X(t) is the input excitation of the force port and F(t) is the output response of the force port; the heat port variables include temperature T(t) and heat flux Q(t), where T(t) is the input excitation of the heat port and Q(t) is the output response of the heat port.

[0119] The prior reference is the estimated value of the mutual impedance parameters obtained after the parameter update was completed in the previous control cycle, i.e., time t-1, including the estimated value of the mutual impedance parameters in the previous cycle. (t-1) and (t-1), and the estimated value of the self-impedance parameter of the previous cycle. (t-1) and (t-1), the parameter estimation results of the historical period of t-1 contain the information accumulation of all historical data up to the previous period, and are the starting reference point for the parameter update of the current period.

[0120] The prior reference is stored in the parameter register inside the identification module. After the parameter update is completed in each control cycle, the contents of the register are updated to the estimated value of the current cycle, which serves as the prior reference for the next cycle.

[0121] In this application, a preferred method for recursive comparison and correction includes:

[0122] The recursive comparison and correction process is performed by the identification module in each control cycle as follows:

[0123] The parameter estimates from the previous cycle, i.e., the prior reference, are used as the predicted values ​​for the current cycle. They are substituted into the pre-established impedance network equations and combined with the excitation signals (X(t) and T(t)) in the current cycle's information input to calculate the predicted response values ​​that should be generated under the current predicted parameters, i.e., the predicted values ​​of displacement and heat flow.

[0124] Specifically, the impedance network equation expresses the mapping relationship between the generalized force vector and the generalized flow vector in matrix form, that is:

[0125]

[0126] In the prediction step, the identification module uses the grinding force signal F(t) and temperature T(t) actually measured in the current control cycle as the known generalized force input vector, and the estimated value of the impedance matrix updated in the previous control cycle. Substituting the known prediction parameters into the above matrix equation and performing matrix inversion on the equation, we can solve for the generalized flow vectors corresponding to the current prediction parameters, namely the predicted displacement value Xpred(t) and the predicted heat flux value Qpred(t).

[0127] Assuming that the current force-thermal coupling characteristics are the same as those of the previous cycle, i.e., the impedance parameters have not changed abruptly, under the excitation of the grinding force signal F(t) and temperature T(t) measured in the current cycle, what displacement response and thermal flow response should the grinding force port and the thermal port produce respectively?

[0128] The calculated predicted response value is compared with the actual X(t) and Q(t) collected in the current period's information input, and the deviation between the two is calculated. The deviation is used to reflect the degree of inconsistency between the output predicted based on historical parameters and the actual measured output. Its magnitude directly characterizes the degree of adaptation of the previous period's parameter estimate to the current operating condition.

[0129] The calculated displacement prediction value Xpred(t) and heat flow prediction value Qpred(t) are compared with the actual displacement measurement value X(t) and heat flow measurement value Q(t) collected in the current control cycle, respectively, and the deviation between the two sets of corresponding variables is calculated. Specifically:

[0130] Displacement deviation = X(t) - Xpred(t), the deviation reflects the degree to which the measured value of the grinding force port displacement deviates from the predicted value under the excitation of displacement X(t);

[0131] Heat flux deviation = Q(t) - Qpred(t), the deviation reflects the degree to which the measured value of the heat flux at the hot port deviates from the predicted value under temperature T(t) excitation.

[0132] If the deviation is zero, it indicates that the current actual force-thermal coupling characteristics are completely consistent with the previous cycle, and the impedance parameters do not need to be updated; if the deviation is not zero, it indicates that the force-thermal coupling characteristics have changed, and the impedance parameters need to be corrected accordingly based on the magnitude and direction of the deviation.

[0133] Based on the calculated deviation, the prior benchmark is weighted and corrected according to a predetermined recursive correction rule. The correction logic is as follows: the estimated value of the new period parameter is equal to the estimated value of the previous period parameter plus a correction term. The magnitude of the correction term is proportional to the calculated deviation. The proportionality coefficient, i.e. the correction gain, is automatically adjusted according to the reliability of historical data and the uncertainty of the current information, so that the parameter with the greater deviation receives a larger correction.

[0134] The weighted correction of the prior reference according to the predetermined recursive correction rule specifically refers to multiplying the calculated displacement deviation and heat flow deviation by a preset correction gain coefficient, and then using these as adjustment amounts for each element of the impedance matrix, which are then superimposed on the parameter estimates of the previous period.

[0135] A larger gain coefficient results in a faster response of the parameter to the measurement data of the current cycle, but also a greater sensitivity to measurement noise; a smaller gain coefficient results in a more stable parameter update, but a slower tracking speed of changes in operating conditions. The range of the correction gain coefficient G is 0.01 ≤ G ≤ 0.5. The correction gain coefficient is determined according to the following rules: when the deviation increases in the same direction for three or more consecutive control cycles, the correction gain coefficient is increased to accelerate the parameter tracking speed; when the deviation oscillates alternately between positive and negative values ​​between adjacent control cycles, the correction gain coefficient is decreased to suppress parameter estimation fluctuations. The initial value of the correction gain coefficient is set to 0.1, and it is adaptively adjusted cycle by cycle according to the above rules during the recursive identification process.

[0136] A preferred adaptive adjustment method is as follows: ,in This represents the number of periods in which the deviation changes continuously in the same direction. A positive value indicates continuous increase, and a negative value indicates continuous decrease. When the deviation oscillates alternately between positive and negative values ​​between adjacent periods... The range of the corrected gain coefficient is 0.01 ≤ G ≤ 0.5.

[0137] The correction process is calculated as follows: using the mutual impedance parameter Taking the update as an example, its current period's estimated value (t) equals the estimated value of the previous period. (t-1) Add a correction term, which is equal to the preset correction gain coefficient multiplied by the heat flow deviation of the current cycle, and the self-impedance parameter. The correction term is calculated using the displacement deviation, which corresponds to the deviation of the response variables at different ports in the network equation.

[0138] The correction methods for other impedance parameters are the same as those described above, using their respective deviations as the basis for correction. The correction depends on the reaction relationship between temperature T and grinding force signal F, and the deviation is obtained by calculating the corresponding error term in the network equation. The correction depends on the deviation relationship between temperature T and heat flux Q.

[0139] The obtained corrected parameter values ​​are used as the estimated values ​​of the mutual impedance parameters for the current cycle, written into the parameter register, replacing the original prior reference, completing the parameter update for the current cycle, and used as the prior reference for the next cycle.

[0140] The identification module performs a data acquisition, recursion, and update operation in each control cycle, so that the current value of the mutual impedance parameter is continuously tracked and updated as the grinding conditions change. After completing the parameter update, the identification module outputs the estimated value of the mutual impedance parameter at the current moment to the cooperative controller as data for determining the current thermo-coupling state.

[0141] The identification module also stores the reference value of the mutual impedance parameter of the linear guide grinding machine under the calibration conditions. The reference value is used to characterize the reference characteristics of the grinding machine under normal thermo-coupling conditions. After the identification module obtains the current value of the mutual impedance parameter in real time, it first calculates the instantaneous deviation between the current value and the reference value.

[0142] After the grinding machine is installed and debugged, and before formal machining, a complete four-terminal impedance identification process is first performed under standard calibration conditions. A preferred example of the calibration conditions is that the grinding machine spindle speed is 80%-100% of the rated speed, the feed rate is 50%-70% of the rated feed rate, the coolant flow rate and pressure are the rated values, the workpiece material is the standard test piece material, and the grinding machine reaches thermal equilibrium after running under no-load for 30 minutes, and then continuously grinds at least 5 strokes with constant cutting parameters in a stable machining state. Under the current state, the mechanical and thermal coupling characteristics of the grinding machine are in a stable and repeatable normal range, which can represent the healthy working state of the grinding machine.

[0143] Under calibrated operating conditions, by applying a preset frequency sweep excitation signal to the grinding force port and heat port, and acquiring the response signal, the impedance matrix under the current operating condition is identified. The identification module extracts the first mutual impedance parameter from the impedance matrix. ref refers to the calibration value of the reaction coupling strength between the temperature field and the grinding force field, and the second mutual impedance parameter. ref is the calibration value for the positive coupling efficiency of the grinding force field and the temperature field. ref and The ref value is written as a reference value into the non-volatile register inside the identification module, serving as a reference standard for subsequent judgment of whether the coupling state has deviated.

[0144] It should be noted that, ref characterizes the degree of change in grinding force signal F when the temperature T in the grinding zone changes by one unit under calibrated conditions. It reflects the coupling sensitivity of the temperature field to the force field, and the unit is N / ℃.

[0145] ref characterizes the degree of change in heat flux Q for each unit change in displacement X under calibrated conditions, with units of (W / m²) / m, and is used to represent the benchmark value of the coupling efficiency of force-generated heat.

[0146] The identification module compares the instantaneous deviation at the current moment with the instantaneous deviation at a historical moment in time sequence, and calculates the rate of change of the instantaneous deviation over time. The rate of change is used to characterize the degree of swiftness of the grinding machine's thermo-coupling state deviating from the normal reference characteristics.

[0147] After the identification module completes the recursive update of the mutual impedance parameters in each control cycle, it reads the estimated value of the mutual impedance parameters for the current control cycle from the parameter register, which is the first estimated value of the mutual impedance parameters for the current cycle. (t) and estimated value of the second mutual impedance parameter (t) reads the pre-stored reference value of the mutual impedance parameter from the reference value register. ref and ref.

[0148] A preferred method for calculating instantaneous deviation:

[0149] Calculate the instantaneous deviation between the two mutual impedance parameters:

[0150] Instantaneous deviation of the first mutual impedance This deviation is a signed numerical value, with units of N / ℃. When Δ When (t) is positive, the reaction coupling strength between the current temperature field and the grinding force field is greater than that under calibration conditions; Δ When (t) is negative, it indicates that it is less than the calibration state; when the absolute value is close to zero, it indicates that it is basically consistent with the calibration state.

[0151] Instantaneous deviation of the second mutual impedance This deviation is a signed numerical value, with units of ℃ / N. When Δ When Δ is positive, the positive coupling efficiency of the current grinding force field to the temperature field is greater than that of the calibration state; when Δ When (t) is negative, it indicates that it is less than the calibration state; when the absolute value is close to zero, it indicates that it is basically consistent with the calibration state.

[0152] The instantaneous deviations of the first and second mutual impedances constitute the deviation vector [Δ] for the current control cycle. (t), Δ [t], the deviation vector is used to represent the direction and magnitude of the deviation of the current thermo-mechanical coupling state from the calibration reference. When each component of the deviation vector is close to zero, it indicates that the current thermo-mechanical coupling state is consistent with the calibration state; when any component of the deviation vector deviates significantly from zero, it indicates that the corresponding coupling path has changed.

[0153] When the rate of change exceeds the preset rate tolerance range, it is determined that the current thermo-coupling state is undergoing a sudden deviation that is not gradual.

[0154] The identification module maintains a buffer of historical deviation data stored in chronological order. This buffer uses a first-in, first-out (FIFO) queue structure to store the most recent N control cycles, where N ≥ 2. The specific value is set according to the control cycle duration to ensure that short-term noise is smoothed while retaining sufficient instantaneous deviation data with changing trends. Each storage unit records the complete deviation vector [Δ] for one control cycle. (k), Δ [(k)] and its corresponding timestamp, after calculating the new instantaneous deviation in each control cycle, push the new data into the queue and remove the oldest data in the queue.

[0155] A preferred method for calculating the rate of change:

[0156] The instantaneous deviation Δ is calculated in the current control cycle. (t) and Δ After (t), the identification module reads the instantaneous deviation Δ from the deviation history record buffer of the previous control cycle. (t-1) and Δ (t-1), calculate the rate of change of each of the two mutual impedance parameters respectively.

[0157] When there are at least two adjacent cycles of data in the buffer, the identification module calculates the rate of change as follows:

[0158] The rate of change of the instantaneous deviation of the first mutual impedance , where Ts is the duration of the control cycle, in seconds.

[0159] The rate of change of the instantaneous deviation of the second mutual impedance , where Ts is the duration of the control cycle, in seconds.

[0160] The aforementioned rate of change characterizes the magnitude of the instantaneous deviation of each mutual impedance parameter within a unit of time. The larger the rate of change, the more drastic the change of the instantaneous deviation within a unit of time, indicating that the mechanical-thermal coupling state is changing rapidly. The smaller the rate of change or the closer it is to zero, the more gradual the change of the instantaneous deviation within a unit of time, indicating that the mechanical-thermal coupling state is relatively stable or only undergoes slow drift.

[0161] When the buffer stores data from multiple consecutive periods, the identification module can use a polynomial fitting method to calculate the rate of change. The instantaneous deviation data from the most recent M periods in the buffer are used as sample points, where 3≤M≤N. A first-order linear function is fitted using the least squares method, and the slope of the linear function is taken as the rate of change for the current period. This can effectively suppress the interference of single-period measurement noise on the rate of change calculation results, making the estimation of the rate of change smoother and more stable.

[0162] The identification module has a preset rate tolerance range for the rate of change of the mutual impedance parameter. The method for setting this tolerance range is as follows:

[0163] After the benchmark value identification is completed during the grinding machine calibration phase, the rate of change of the mutual impedance parameter is continuously monitored. Using the standard deviation σ of the monitored rate of change data within this time period as the benchmark, the rate tolerance range is set to [-k·σ, +k·σ], where k is a preset confidence coefficient, the value of which is determined based on the actual working stability of the grinding machine (k=3.0 in this embodiment). The preferred value range is 2.5≤k≤4.0, ensuring that the rate tolerance range reflects the statistical distribution characteristics of the rate of change of the grinding machine under normal working conditions, i.e., including the observed rate of change under normal working conditions within the tolerance range.

[0164] When there are significant changes in the grinding machine configuration, workpiece material, or machining process, the above calibration process must be repeated to update the rate tolerance range. The tolerance range can be pre-calibrated and stored as multiple parameter sets according to different working conditions (workpiece material, grinding parameters, cooling conditions, etc.). Before machining, select the calibration value closest to the current working conditions as the benchmark; during machining, the upper limit of the tolerance range can also be dynamically fine-tuned based on historical data. If the workpiece material is changed or the grinding parameters are significantly altered, recalibration or retrieval of the corresponding parameters from the pre-stored database is required.

[0165] The identification module has built-in deviation rate threshold determination logic, which determines the rate of change during the current control cycle. (t) and After calculating (t), the decision logic compares the two rates of change with the preset rate tolerance range:

[0166] like (t) exceeds [-k· ,+k· The range of ], or Exceeding [-k· ,+k· The range of [ ] is determined by the logic that the current thermo-coupling state is undergoing a sudden, non-gradual deviation. and These are the standard deviations of the rates of change of the first and second mutual impedances obtained during the calibration phase, respectively.

[0167] The above determination is based on the condition that the rate of change of any mutual impedance parameter exceeds the limit, which is triggered to ensure that sudden changes on any coupling path can be detected in a timely manner.

[0168] When the absolute value of the instantaneous deviation is large but its rate of change is within the tolerance range, the judgment logic does not output a positive result. For example, after a long period of tool processing, the mutual impedance parameter gradually changes with the increase of tool wear. The instantaneous deviation accumulates to a large value, but the rate of change per unit time remains within a small range. This is a predictable gradual evolution and does not require triggering emergency intervention.

[0169] When the rate of change exceeds the tolerance range, the judgment logic outputs a positive result regardless of the magnitude of the absolute value of the instantaneous deviation. For example, if the coolant is suddenly interrupted during grinding, the grinding depth suddenly increases due to a feed system malfunction, or local hard spots appear in the workpiece material, the coupling state changes rapidly in a short period of time with a large rate of change. This is a sudden anomaly that requires immediate triggering of a response. Using the rate of change rather than the absolute value of the instantaneous deviation as the triggering criterion effectively avoids misjudging long-standing large deviations as sudden anomalies and also avoids missing sudden medium deviations.

[0170] Simultaneously with the determination outputting a positive result, the identification module immediately generates the coupling deviation trigger signal and sends the trigger signal to the cooperative controller to trigger the cooperative controller to initiate a re-identification of the overall transfer function.

[0171] A preferred method for generating a trigger signal is as follows: when the determination logic determines that the current thermal coupling state is undergoing a sudden deviation, the identification module immediately generates a coupling deviation trigger signal at the signal output port of the physical layer. The deviation trigger signal is a pulse-type trigger identifier, i.e., a level-changing pulse with a duration of not less than two control cycles of the identification module, or a level-type trigger identifier, i.e., a transition from a low level to a high level and holding until a reset command is received.

[0172] The identification module connects to the co-controller via an internal data bus or dedicated hardware signal line. After generating a trigger signal, the identification module sends the trigger signal to the interrupt input or event capture port of the co-controller in real time. The pulse-type trigger signal triggers the co-controller's interrupt response on the rising edge, enabling the co-controller to execute the interrupt service routine at the start of the next control cycle after the trigger signal arrives. The level-type trigger signal continuously indicates the trigger status by being high. When the co-controller completes the re-identification process, it sends a reset command to the identification module. After receiving the command, the identification module restores the trigger signal to a low level, completing a full trigger-response-reset cycle.

[0173] Upon receiving the trigger signal, the collaborative controller initiates a re-identification process for the overall transfer function in the cascaded collaborative module. It suspends the current control auxiliary function based on the old overall transfer function, re-collects the real-time input-output response data of the force control subsystem and vibration support subsystem under the current operating conditions, re-executes the identification of the transmission correlation between the first two-port network and the second two-port network, and the composite derivation of the overall transfer function. It then obtains an updated overall transfer function that matches the current thermo-coupling state and replaces the old overall transfer function with the updated overall transfer function. The trigger signal is the start command for the adaptive update process.

[0174] The cascaded collaborative module models the force control subsystem as a first two-port network and the vibration support subsystem as a second two-port network, and calculates the overall transfer function through the cascaded characteristics of the two-port networks.

[0175] The overall transfer function is used to describe the response generated at the vibration port after the feed speed command passes through the force control link and the vibration transmission link;

[0176] Specifically, during the grinding process, the feed rate command output by the co-controller serves as the initial input signal for the entire control chain. This speed command first enters the force control subsystem, and after being processed by the force control execution components, including the servo driver, servo motor, ball screw, and feed mechanism, it is converted into an actual dynamic grinding force acting on the grinding contact area. This dynamic grinding force then travels through the mechanical structures such as the machine bed, guide rails, spindle, and fixtures, ultimately generating a corresponding vibration response signal at the vibration monitoring port. The overall transfer function represents the complete mapping relationship between the feed rate command and the vibration response, describing the combined response effect produced after the same speed command flows sequentially through two different subsystems.

[0177] The cascaded collaborative module uses the causal transmission relationship of port physical variables as the modeling basis;

[0178] Specifically, the causal transmission relationship means that during the transmission of a signal or physical quantity along the system link from the previous stage to the next stage, the output of each stage is the result of the input of that stage acting on the internal physical processes of that stage. That is, the input is the cause and the output is the effect, and there is a physical causal temporal relationship between the input and the output. That is, the input signal exists before the output signal, and the output signal is generated by the input signal acting on the internal dynamic characteristics of the system. When modeling, the cascaded collaborative module first identifies the input physical quantities and output physical quantities of the force control subsystem and the vibration support subsystem, as well as the physical transmission path between the input and the output. Based on this, it determines the definition of the port variables of each subsystem, thereby establishing a causal transmission model that conforms to the actual physical signal flow.

[0179] like Figure 3As shown, the cascaded collaborative module defines the physical variable of the input port of the force control subsystem as the feed speed command signal output after being adjusted by the collaborative controller, and defines the physical variable of the output port as the dynamic grinding force signal actually output after the feed speed command is acted upon by the force control execution link. It also establishes a first transmission correlation relationship between the input port and the output port, reflecting the transmission characteristics from the speed command to the grinding force, thereby abstracting the force control subsystem as a whole into a first two-port network with unidirectional signal transmission characteristics.

[0180] It should be noted that the physical boundaries of the force control subsystem are: the input port is located at the command signal input terminal of the servo driver, and the output port is located at the grinding force application point in the grinding contact area; the physical components included in the force control subsystem can be a servo driver, a servo motor, a ball screw, a feed mechanism, and a grinding head spindle system. When identifying the first transmission correlation, the grinding depth, spindle speed, coolant flow rate, and pressure are kept constant. The identification signal is applied within a continuous cutting segment where the workpiece material removal is uniform, to eliminate the interference of workpiece material inhomogeneity and cutting parameter fluctuations on the accuracy of the transfer function identification.

[0181] Specifically, the force control subsystem is the physical component in a linear guide grinding machine responsible for executing feed speed control commands and generating grinding force. Its physical components include a servo driver, a servo motor, a ball screw, a feed mechanism, and a grinding head spindle system. During the grinding process, the feed speed command signal output by the co-controller serves as the control command input. This signal is converted into a motor speed control signal by the servo driver. The servo motor drives the ball screw to rotate, and the screw converts the rotational motion into linear motion of the feed mechanism. The feed mechanism pushes the workpiece or grinding wheel to achieve grinding feed, ultimately generating grinding force at the contact interface between the grinding wheel and the workpiece.

[0182] When the cascaded collaborative module abstracts the force control subsystem into a first-two-port network, its input port variable is the speed adjustment amount output by the collaborative controller, in mm / min, which is the control command input of the force control subsystem; the output port variable is the actual grinding force output by the force control execution link, in N, which is the execution result of the force control subsystem. The physical relationship between the input and output variables is as follows: changes in the feed speed command will lead to changes in the movement speed of the feed mechanism, which in turn changes the contact state and cutting depth between the grinding wheel and the workpiece, ultimately causing a corresponding change in the grinding force.

[0183] The first transmission correlation is the mapping relationship between the feed speed command signal and the dynamic grinding force output signal. It reflects the dynamic transmission behavior of all force control links, such as the response characteristics of the servo drive, the transmission characteristics of the motor and the lead screw, the motion characteristics of the feed mechanism, and the force generation characteristics of the grinding process. By abstracting the subsystem into a two-port network, its complex electromechanical coupling details are encapsulated into a unified port transmission correlation. The correlation is obtained through the system identification method.

[0184] Specifically, a preset sweep frequency speed command signal is applied to the input port of the force control subsystem. The excitation signal adopts a short-time sweep frequency signal with limited amplitude, and the signal application time is ensured to be no more than two grinding strokes to avoid a noticeable impact on the surface quality of the workpiece. The grinding force response signal of the output port is collected synchronously, and the transfer function from the speed command to the grinding force is calculated by the transfer function identification algorithm. The transfer function is the mathematical expression of the first transmission correlation relationship.

[0185] Two-port networks have unidirectional signal transmission characteristics, meaning that signals can only be transmitted from the input port to the output port in the forward direction, and the signal at the output port will not affect the signal at the input port in the reverse direction.

[0186] The cascaded collaborative module defines the physical variable of the input port of the vibration support subsystem as the dynamic grinding force signal output by the output port of the force control subsystem, and defines the physical variable of its output port as the vibration response signal generated at the actual vibration monitoring port after the dynamic grinding force is transmitted through the machine tool structure. It also establishes a second transmission correlation between the input port and the output port, reflecting the transmission characteristics from grinding force to vibration response, thereby abstracting the entire vibration support subsystem into a second two-port network with unidirectional signal transmission characteristics.

[0187] The physical boundary of the vibration support subsystem can be: the input port is located at the grinding force application point in the grinding contact area, and the output port is located at the installation position of the vibration monitoring sensor, such as the spindle bearing seat or the bed guide rail; the physical components included in the vibration support subsystem can be the grinding contact area, grinding wheel spindle, spindle bearing, bed structure, guide rail slider and worktable.

[0188] Specifically, the vibration support subsystem refers to the structural component in a linear guide grinding machine responsible for bearing and transmitting grinding forces and generating a vibration response. Its physical components include the grinding contact area, grinding wheel spindle, spindle bearings, bed structure, guide rails, and worktable. During the grinding process, the dynamic grinding force output by the force control subsystem acts on the contact interface between the grinding wheel and the workpiece. This force, as the excitation source, propagates through the transmission paths of the grinding wheel, spindle, spindle bearings, bed, guide rails, and worktable to various structural components of the machine tool, thus triggering a vibration response in the machine tool structure.

[0189] When the cascaded collaborative module abstracts the vibration support subsystem into a second two-port network, its input port variable is the dynamic grinding force at the output port of the first two-port network, measured in N, which is the excitation source input of the vibration support subsystem. The output port variable is the vibration response or vibration acceleration signal generated at the actual vibration monitoring port after the grinding force is transmitted through the machine tool structure, measured in m / s², which is the response output of the vibration support subsystem under excitation. The physical relationship between the input and output variables is as follows: the grinding force acts as an external excitation on the machine tool structure, exciting the structure's inherent modes and generating a vibration response with corresponding amplitude and frequency. Its transmission characteristics depend on the stiffness, damping, and mass distribution of the machine tool structure.

[0190] The second transmission correlation is the mapping relationship between the dynamic grinding force signal and the vibration response signal, reflecting the transmission behavior of all vibration transmission links, including the force transmission characteristics of the grinding contact area, the support characteristics of the spindle bearing, the modal characteristics of the bed structure, and the damping characteristics of the guide rail slider. By abstracting the subsystem into a two-port network, its complex internal structural dynamics details are encapsulated into a unified port transmission correlation, which is obtained through a system identification method.

[0191] Specifically, a preset sweep frequency excitation signal is applied to the input port of the vibration support subsystem, and the vibration response signal of the output port is collected synchronously. The transfer function from grinding force to vibration response is calculated by the transfer function identification algorithm. The transfer function is the mathematical expression of the second transmission correlation relationship.

[0192] Two-port networks also have unidirectional signal transmission characteristics, meaning that the signal can only be transmitted positively from the grinding force at the input port to the vibration response at the output port, which conforms to the physical causal law of the grinding force as the excitation source and the vibration response as the passive response result in the vibration support subsystem.

[0193] The first two-port network and the second two-port network have a causal matching relationship in terms of port physical variables. That is, the output port physical variables of the first two-port network and the input port physical variables of the second two-port network are completely consistent in terms of physical type and dimension. They are both dynamic grinding force signals, which enables the two two-port networks to have physical compatibility for cascading connection.

[0194] Specifically, in the signal link, the causal matching relationship means that the output physical quantity of the previous subsystem is exactly the input physical quantity of the next subsystem. In essence, they describe the transmission process of the same physical quantity. The dynamic grinding force signal output by the output port of the first two-port network is the actual grinding force acting on the grinding contact area in physical terms. The actual grinding force also serves as the input excitation source of the vibration support subsystem. The port variables of the two networks are both force signals in physical type, both Newtons (N) in dimension, and both correspond to the grinding contact area in physical location.

[0195] This causal matching relationship ensures that the two two-port networks can be cascaded according to the actual direction of signal flow. The output port of the first two-port network and the input port of the second two-port network coincide to form the same cascade node. The physical variable at the node is both the output response of the preceding network and the input excitation of the following network. The two are continuous in time and identical in physical form, and there is no requirement for any physical quantity conversion or dimension transformation.

[0196] When abstracting two actual physical subsystems into a two-port network model, only by ensuring the consistency and continuity of the physical variables at the two network interfaces can the two independently identified transmission relationships be serially combined along the signal flow direction to form a holistic transmission function from the feed speed command to the vibration response. If the physical variables at the two network interfaces are inconsistent, such as one outputting force and the other requiring displacement input, or if the dimensions are mismatched, the two networks cannot be directly cascaded, requiring an additional conversion step, increasing modeling complexity and sources of error. In this application, the direct transmission of the dynamic grinding force signal between the two networks requires no physical quantity conversion, ensuring the accuracy and simplicity of the cascaded combination.

[0197] After completing the modeling of the first two-port network and the second two-port network, the cascaded coordination module obtains the first transmission correlation parameter corresponding to the first two-port network and the second transmission correlation parameter corresponding to the second two-port network. The cascaded coordination module cascades the second two-port network after the first two-port network according to the direction of signal flow in the actual physical link, so that the output port of the first two-port network and the input port of the second two-port network logically coincide into the same cascaded node.

[0198] The first transmission correlation parameter is obtained as follows: During the aforementioned first two-port network modeling process, a preset sweep frequency speed command excitation signal is applied to the input port of the force control subsystem. The dynamic grinding force response signal of the output port is then synchronously acquired. The frequency domain transfer function from the feed speed command to the dynamic grinding force is calculated using a transfer function identification algorithm. This transfer function is expressed in the form of a frequency domain response—that is, the ratio of the dynamic grinding force amplitude at the output port to the speed command amplitude at the input port, and their phase difference, at different frequency points (typically ranging from 0Hz to 200Hz, covering the bandwidth of the force control loop). This set of frequency domain response data constitutes the first transmission correlation parameter and is stored in the parameter register within the cascaded collaborative module.

[0199] The second transmission correlation parameter is obtained as follows: In the aforementioned second two-port network modeling process, after applying a preset sweep frequency force excitation signal to the input port of the vibration support subsystem and synchronously acquiring the vibration response signal of the output port, the frequency domain transfer function from the dynamic grinding force to the vibration response is calculated by the transfer function identification algorithm. The frequency domain transfer function is expressed in the form of frequency domain response, i.e., at different frequency points. A preferred example is that the range can be 0Hz to 1000Hz.

[0200] The ratio of the vibration response amplitude at the output port to the grinding force amplitude at the input port, and their phase difference, within the main modal frequency range covering the machine tool structure, constitute the second transmission correlation parameter, which is stored in the parameter register inside the cascaded collaborative module.

[0201] The cascaded collaborative module first maps the feed speed command signal to a grinding force signal at an intermediate node via the first transmission correlation, and then maps the grinding force signal at the intermediate node to a vibration response signal at the final node via the second transmission correlation. By sequentially cascading and combining the above two mapping relationships along the signal flow direction, a complete mapping relationship directly from the feed speed command signal to the vibration response signal is derived, and this complete mapping relationship is the overall transmission function.

[0202] Specifically, the feed rate command signal output by the co-controller first enters the force control subsystem, and after being acted upon by the force control execution link, it outputs dynamic grinding force; the grinding force then enters the vibration support subsystem, and after being transmitted through the machine tool structure, it generates a vibration response. The complete flow path of the signal from the starting end to the ending end is: feed rate command, force control subsystem, dynamic grinding force, vibration support subsystem, vibration response.

[0203] The cascaded collaborative module connects the second two-port network representing the vibration support subsystem to the first two-port network representing the force control subsystem, according to the aforementioned physical signal flow direction. Specifically, the output port variable of the first two-port network is the grinding force signal, and the input port variable of the second two-port network is also the grinding force signal. Both the output port variable of the first two-port network and the input port variable of the second two-port network are force signals with the dimension of Newtons (N).

[0204] The output port of the first two-port network and the input port of the second two-port network are overlapped to form the same cascade node. The cascade node corresponds to the grinding contact area, that is, the dynamic grinding force output by the first network is used as the excitation input of the second network here.

[0205] The mapping refers to the process of converting an input signal into an output signal through a transfer function. That is, given the amplitude and frequency of the input excitation signal, the amplitude and phase of the output response signal at that frequency are calculated through the transfer function.

[0206] Specifically, when performing a step-by-step mapping operation in the frequency domain, the cascaded cooperative module includes a first-level mapping and a second-level mapping;

[0207] The first-level mapping includes: a cascaded collaborative module acquiring the feed speed command signal. If the speed command signal is a time-domain signal, it is first converted into a frequency-domain representation using a fast Fourier transform (FFT), by decomposing it into the amplitude and phase information of each frequency component. For each frequency component of the speed command signal, the amplitude is multiplied by the gain value of the first transmission correlation parameter at the same frequency point, and the phase is added to the phase difference of the first transmission correlation parameter at the same frequency point to obtain the frequency-domain representation of the grinding force signal at the intermediate node at the same frequency point. After performing the calculation on all frequency components one by one, the complete frequency-domain representation of the dynamic grinding force signal at the intermediate node is obtained, which is the grinding force signal mapped to the intermediate node by the first transmission correlation. If a time-domain form of the grinding force signal is required, an inverse fast Fourier transform is performed on the obtained frequency-domain result.

[0208] The second-level mapping includes: using the frequency domain representation of the grinding force signal at the intermediate node obtained from the first-level mapping as the input of the second-level mapping; for each frequency component of the grinding force signal, multiplying the amplitude of the frequency component by the gain value of the second transmission correlation parameter at the same frequency point, and adding the phase difference of the second transmission correlation parameter at the same frequency point to the phase to obtain the frequency domain representation of the vibration response signal at the final node at the same frequency point; after performing the calculation on all frequency components one by one, the complete frequency domain representation of the vibration response signal at the final node is obtained, which is the vibration response signal at the final node mapped by the second transmission correlation.

[0209] It is important to note that the first-level mapping is performed to obtain the intermediate grinding force signal, and then the second-level mapping is performed using the intermediate grinding force signal as input to obtain the final vibration response signal. Each level of mapping is a direct application of the corresponding transmission correlation, and the input-output relationship of the mapping process corresponds exactly to the signal transmission path in the actual physical link. When the input signal contains multiple frequency components, the above frequency-point mapping operation is applied to each frequency component. Utilizing the superposition principle of linear systems, for any excitation signal containing multiple frequency components, it can be decomposed into each frequency component, mapped separately, and then synthesized.

[0210] By sequentially cascading and combining the two mapping relationships along the signal flow direction, a complete mapping relationship directly between the feed speed command signal and the vibration response signal is derived. This complete mapping relationship is the overall transfer function.

[0211] Specifically, the cascaded composite operation method is as follows: since the first-level mapping and the second-level mapping each have a definite gain value and phase difference on the same set of frequency points, during cascaded composite, for each frequency point, the gain value of the first transmission correlation at that frequency point is multiplied by the gain value of the second transmission correlation at the same frequency point to obtain the total gain value of the overall transfer function at that frequency point; the phase difference of the first transmission correlation at that frequency point is added to the phase difference of the second transmission correlation at the same frequency point to obtain the total phase difference of the overall transfer function at that frequency point.

[0212] For any frequency point f, the gain of the overall transfer function Htotal(f) is |Htotal(f)| = | (f)|×| (f)|, which is the product of the first-stage transfer function gain and the second-stage transfer function gain; the phase of the overall transfer function Htotal(f) is ∠Htotal(f) = ∠ (f)+∠ (f) represents the algebraic sum of the phases of the first-stage transfer function and the second-stage transfer function. (f) represents the transfer function value of the first transmission correlation at frequency f. (f) is the transfer function value of the second transmission correlation at frequency f.

[0213] After performing the composite operation of gain multiplication and phase addition on each frequency point, the complete frequency domain response of the overall transfer function on the frequency point sequence is obtained. The frequency domain response directly describes the mapping relationship between the feed speed command signal and the vibration response signal, without the need to separately acquire or process the grinding force signal data.

[0214] The overall transfer function is a frequency domain transfer function. When the feed speed command signal is a sine wave of a certain frequency, the amplitude of the sine wave of the vibration response signal at the same frequency is equal to the input amplitude multiplied by the total gain, and the phase is either lagging behind or leading the input signal by the total phase difference. Using the overall transfer function, the response amplitude and phase that will be generated at the vibration port can be calculated in advance before the actual output of the feed speed adjustment.

[0215] In the frequency domain, the cascaded composition of two transfer functions equals their product. Since both two-port networks in this application are unidirectional transmission networks, with signals transmitted only from the input port to the output port in the forward direction, and their port physical variables are perfectly matched (first network output = second network input), the two transfer functions can be directly cascaded and composed. The overall transfer function obtained after cascade composition is equivalent to the result obtained by identifying the two subsystems as a single system. This equivalence ensures that the overall transfer function derived from the cascade composition reflects the actual transmission characteristics from the feed rate command to the vibration response.

[0216] The overall transfer function is used to characterize the response amplitude and phase transfer characteristics of each feed rate adjustment output by the cooperative controller after flowing sequentially through the force control subsystem and the vibration support subsystem at the vibration monitoring port.

[0217] Specifically, when the cooperative controller needs to output a certain feed rate adjustment amount in the control process, it first inputs the frequency domain characteristics of the adjustment amount, including the amplitude and phase information of each frequency component, into the overall transfer function. The response amplitude generated at the vibration port after the adjustment amount passes through the entire cascaded link is calculated by gain product and phase summation. If the response amplitude would cause the overall vibration index to exceed the safety threshold, the cooperative controller applies amplitude constraints to the adjustment amount before outputting it.

[0218] The overall transfer function is used to characterize the response amplitude and phase transfer characteristics of each feed rate adjustment output by the cooperative controller after flowing sequentially through the force control subsystem and the vibration support subsystem at the vibration monitoring port.

[0219] When the identification module determines that a sudden deviation in the thermal coupling state has occurred and generates a coupling deviation trigger signal, the rising edge or level transition of the trigger signal is input to the co-controller via the hardware interrupt pin, and the co-controller captures the trigger signal in the next instruction cycle.

[0220] Upon detecting a coupling deviation trigger signal, the cooperative controller immediately performs the following operations: suspends the currently executing force control loop routine control task and saves the current control state and intermediate variables; sends a forced re-identification command to the cascaded cooperative module via the data bus, the forced re-identification command containing a re-identification request identifier and the current timestamp information; and switches its own operating state to re-identification waiting mode, suspending the prediction and constraint operations of the feed rate adjustment amount in re-identification waiting mode, and waiting for the cascaded cooperative module to complete the re-identification.

[0221] After receiving the forced re-identification command, the cascaded collaborative module executes the re-identification process, which includes: pausing the control auxiliary function of the current old overall transfer function; re-acquiring the input and output response data of the force control subsystem and the vibration support subsystem under the current actual processing state; re-executing the transmission association identification of the first two-port network and the second two-port network based on the re-acquired data; and re-deriving the updated overall transfer function according to the aforementioned two-port network cascade composite method.

[0222] After the cascaded collaborative module completes re-identification, it sends the updated overall transfer function to the collaborative controller via the data bus. Upon receiving the updated overall transfer function, the collaborative controller replaces the old overall transfer function with the updated one and switches its operating state from re-identification waiting back to normal operation mode, resuming the execution of the suspended force control loop's routine control tasks.

[0223] Specifically, in each control cycle, the force control loop within the co-controller first calculates the required feed rate adjustment Δv(t) for the current control cycle based on the deviation between the measured grinding force and the target value, according to a preset force control algorithm such as a PID control algorithm. It should be noted that those skilled in the art can easily understand how to calculate the required feed rate adjustment for the current control cycle using a PID algorithm or other force control algorithms, based on the technical content disclosed in this application and common knowledge of PID control algorithms. Figure 5 As shown, the PID control parameters used in the force control loop in this embodiment are proportional coefficient Kp=1.0, integral coefficient Ki=0.1, and derivative coefficient Kd=0.05. The vibration safety threshold is set to 4.5 m / s² according to the ISO grinding force signal ISO 10816 standard and the grinding machine machining accuracy requirements. The figure also shows the reference values ​​of the mutual impedance parameters under calibration conditions. ref and The value is indicated by a normalized value and is used as a reference benchmark in the identification module to determine coupling deviation.

[0224] After the force control loop within the co-controller completes the calculation of the feed rate adjustment Δv(t) for the current control cycle, before outputting the adjustment to the force control subsystem, the co-controller first retrieves the feed rate adjustment from the output buffer of the force control loop. The feed rate adjustment Δv(t) exists in the co-controller's memory as a time-domain digital signal. Predictive calculations require converting it to a frequency-domain representation to obtain the amplitude and phase information of the adjustment signal at each frequency component.

[0225] The frequency domain representation of the feed rate adjustment is input into the overall transfer function currently held by the cascaded cooperative module. The overall transfer function is a frequency domain transfer function, expressed in terms of amplitude-frequency characteristics and phase-frequency characteristics at each frequency point. During the prediction calculation, for each frequency component of the feed rate adjustment signal, the amplitude of the feed rate adjustment signal is multiplied by the total gain value of the overall transfer function at the same frequency point to obtain the amplitude of the response component that the frequency component will generate at the vibration monitoring port; its phase is added to the total phase difference value of the overall transfer function at the same frequency point to obtain the phase of the response component that the frequency component will generate at the vibration monitoring port, thus obtaining the frequency domain representation of the vibration response signal. The root mean square value or peak value of the amplitude of each frequency component in the frequency domain representation is the pre-response amplitude Apred that the feed rate adjustment will generate at the vibration monitoring port. The obtained frequency domain result can also be subjected to an inverse fast Fourier transform to obtain the time domain vibration acceleration waveform, from which characteristic values ​​such as peak acceleration and root mean square acceleration can be extracted.

[0226] If the prediction result indicates that the adjustment amount is safe, then the output is normal; if the prediction result indicates that the adjustment amount poses a risk, then amplitude constraints are applied before output.

[0227] Specifically, the collaborative controller has built-in comprehensive vibration index evaluation logic, which is implemented as follows:

[0228] The collaborative controller acquires the current measured vibration signal from the vibration monitoring port in real time through an accelerometer installed on the spindle bearing housing or the machine bed, and extracts multiple vibration characteristic parameters from the signal, including: the peak value of the current measured vibration acceleration, the root mean square value of the current measured vibration acceleration, and the spectral energy value of the current measured vibration signal in a specific frequency band.

[0229] The evaluation logic integrates the pre-response amplitude Aspred with the measured vibration characteristic parameters using a weighted fusion method to obtain the expected comprehensive vibration index value Vtotal after the actual output of the feed rate adjustment Δv(t). The calculation method is as follows: the current measured vibration characteristic parameters constitute the baseline level of the current vibration state, and the pre-response amplitude Aspred represents the additional vibration contribution that the adjustment Δv(t) will add. The sum of the two is the expected comprehensive vibration index value, which is used to establish a complete mapping path from the feed rate adjustment Δv(t) to the comprehensive vibration index Vtotal: Δv(t) is first mapped to the pre-response amplitude Aspred through the global transfer function, and then fused with the current measured vibration level to obtain Vtotal.

[0230] The safety threshold Vthreshold is preset based on the vibration allowable standard and machining accuracy requirements of the linear guide grinding machine. Combined with the specific grinding machine model and machining accuracy level, it is set as the upper limit of the root mean square value of vibration acceleration or the upper limit of other vibration characteristic quantities set according to the specific grinding machine process requirements. The co-controller compares the calculated comprehensive vibration index value Vtotal with the pre-stored safety threshold Vthreshold.

[0231] If Vtotal≤Vthreshold, the co-controller determines that the feed rate adjustment is a safe adjustment and allows it to be output normally to the force control subsystem.

[0232] If Vtotal > Vthreshold, the collaborative controller determines that the feed rate adjustment is a risk adjustment and initiates the amplitude constraint process.

[0233] The amplitude constraint process includes:

[0234] When the current feed rate adjustment is determined to be a risk adjustment, the co-controller first calculates the preliminary constraint coefficient α = Vthreshold / Vtotal based on the ratio between the current comprehensive vibration index value Vtotal and the safety threshold Vthreshold. Since Vtotal > Vthreshold, the constraint coefficient α satisfies 0 < α < 1. After multiplying the adjustment amplitude by α, the corresponding theoretical comprehensive vibration index value will be equal to the safety threshold.

[0235] Since the overall transfer function used in this application is obtained by linearization near the operating point, the system can be considered linear for small-amplitude adjustments. Therefore, the constraint coefficient is calculated as α = Vthreshold / Vtotal (0 < α < 1). Multiplying the initial constraint coefficient α by the original feed rate adjustment Δv(t) yields the initially reduced adjustment Δv'(t) = α × Δv(t). Because the overall transfer function used in this application is obtained by linearization near the operating point, the system can be considered linear for small-amplitude adjustments. Therefore, a single amplitude reduction is sufficient to reduce the comprehensive vibration index value to the safe threshold level. Considering the slight nonlinearity in the actual system, the constraint coefficient can be further multiplied by the safety factor β (0.9 ≤ β ≤ 0.95), i.e., the final constraint coefficient αfinal = β·Vthreshold / Vtotal. The cooperative controller sends the reduced adjustment value as the final output value to the force control subsystem without requiring multiple iterations for verification.

[0236] The feed rate adjustment amount, after being adjusted and verified by the above constraint coefficients, is output as the final adjustment amount to the force control subsystem. The force control subsystem receives the adjustment amount after amplitude reduction and executes the corresponding feed motion.

[0237] When the cooperative controller receives the coupling deviation trigger signal, it controls the cascaded cooperative module to re-identify the overall transfer function. The cooperative controller monitors the coupling deviation trigger signal output by the identification module in real time. The trigger signal is a pulse-type or level-type trigger identifier, which is used to indicate to the cooperative controller that the current thermal coupling state has suddenly deviated.

[0238] Specifically, the collaborative controller and the identification module are connected via a dedicated hardware signal line. The signal line is connected to the external interrupt input terminal of the collaborative controller, such as the GPIO interrupt pin. The collaborative controller has an interrupt service routine set up inside. The routine is configured to work in rising edge trigger or level transition trigger mode, and runs continuously during the grinding process to monitor the level state changes on the signal line in real time.

[0239] The trigger signal has two optional implementation forms:

[0240] When the identification module determines that a sudden deviation in the thermal coupling state has occurred, the identification module generates a level transition pulse at its signal output port with a duration of not less than two control cycles of the identification module. If the pulse transitions from low level to high level and is maintained for a predetermined time, it will automatically return to low level. When the interrupt service routine of the co-controller detects the rising edge of the pulse, it is identified as a valid trigger event.

[0241] When the identification module determines that a sudden deviation in the thermal coupling state has occurred, the identification module switches the level of its signal output port from low to high and maintains this state until it receives a reset command sent by the co-controller after completing re-identification. The co-controller's interrupt service routine recognizes a valid trigger event when it detects the transition edge from low to high.

[0242] All collaborative controllers use the detection of a rising edge of a signal as a capture flag for trigger events. The operation of the trigger signal generation and identification module in the aforementioned steps to determine that a sudden deviation in the thermal coupling state has occurred and output a positive determination result occurs synchronously. Once the determination logic outputs a positive result, the identification module generates the trigger flag at its physical level signal output port.

[0243] When the cooperative controller detects the rising edge of the coupling deviation trigger signal, the cooperative controller immediately sends a forced re-identification command to the cascaded cooperative module;

[0244] Specifically, when the interrupt service routine of the co-controller detects the rising edge of the trigger signal, the co-controller in the interrupt response function:

[0245] The coordinating controller immediately suspends the currently executing force control loop routine control task, and saves the current control state and intermediate variables, including the measured grinding force value of the current control cycle, the intermediate value of the speed adjustment calculation, and the control status flag, to prevent data conflicts between the routine control task and the re-identification operation during the re-identification period.

[0246] The collaborative controller sends a forced re-identification command to the cascaded collaborative module via the data bus. The command includes the following information fields: command type identifier and sending timestamp. The collaborative controller sends the command synchronously and waits for a confirmation signal from the cascaded collaborative module to ensure successful command reception.

[0247] The collaborative controller switches its operating state to the waiting-for-re-identification-completion mode. To prevent force control instability due to sudden changes in operating conditions during re-identification, the collaborative controller simultaneously starts a safety transition timer. If the re-identification duration exceeds the predetermined time, such as 10 control cycles, it automatically forces the feed rate adjustment to 50% of the current value and issues an alarm signal to reduce the risk of vibration and ensure machining safety. In the waiting-for-re-identification-completion mode, the prediction and constraint operations of the feed rate adjustment are suspended, and the output of new control commands to the force control subsystem is stopped, but the existing control output remains unchanged to prevent the actuator from going out of control during re-identification. It waits for the cascaded collaborative module to complete the re-identification and return the updated overall transfer function.

[0248] To prevent the same trigger event from being repeatedly responded to during re-identification, the co-controller disables the trigger signal response on the interrupt pin after sending the forced re-identification command, and then unmasks it after the re-identification is completed.

[0249] After receiving the forced re-identification command, the cascaded collaborative module pauses the current control auxiliary function based on the old overall transfer function, and re-acquires the real-time input and output response data of the force control subsystem and the vibration support subsystem under the current working conditions. During the acquisition process, the inherent force and heat fluctuations and vibration signals in the grinding process are preferentially used as natural excitations. If the spectral energy of the natural excitation is insufficient, a small amplitude value can be superimposed on the control command. The small amplitude value can be a short-time sweep frequency signal of no more than 3%, and the application time is ensured to be no more than 2 grinding strokes to minimize the impact on the processing quality. Based on the re-acquired data, the transmission correlation identification of the first two-port network and the second two-port network and the composite derivation of the overall transfer function are re-executed to obtain an updated overall transfer function that matches the current thermo-coupling state.

[0250] Specifically, upon receiving a forced re-identification command, the cascaded collaborative module first confirms the receipt of a valid forced re-identification command, suspends any control auxiliary functions currently being executed based on the old overall transfer function, such as ongoing response amplitude prediction calculations, marks the currently held old overall transfer function as pending replacement, clears the internal temporary register used to store the current overall transfer function, and initializes the data buffer required for the re-identification process. Under the current actual machining state, the cascaded collaborative module applies a preset sweep frequency speed command excitation signal to the input port of the force control subsystem, i.e., the command input terminal of the servo driver, via the data bus. The frequency range covers the force control loop bandwidth, preferably 0Hz-200Hz, and simultaneously acquires the grinding force response at the output port of the force control subsystem, i.e., the grinding force sensor. Upon completion of the signal acquisition, the acquired time-domain input and output data are stored in the internal buffer of the cascaded collaborative module. The acquired time-domain excitation signal and response signal are subjected to Fast Fourier Transform (FFT) respectively. After conversion to the frequency domain, the amplitude ratio and phase difference of the response signal and excitation signal are calculated at each frequency point to obtain the transfer function value at each frequency point, forming the frequency domain expression of the first transmission correlation relationship, which serves as the updated first transmission correlation relationship parameter. Under the current actual machining state, the cascaded collaborative module applies a preset sweep frequency force excitation signal to the input port of the vibration support subsystem, i.e., near the grinding contact area, through a signal generator. The frequency range covers the main modal frequencies of the machine tool structure, typically 0Hz to 1000Hz, and simultaneously acquires the vibration response signal of the accelerometer at the output port of the vibration support subsystem, i.e., the vibration monitoring port.

[0251] After the data acquisition is completed, the acquired time-domain input and output data are stored in the internal buffer of the cascaded collaborative module.

[0252] The time-domain excitation signal and response signal acquired in step four are subjected to fast Fourier transforms respectively. After being transformed to the frequency domain, the amplitude ratio and phase difference of the response signal and the excitation signal are calculated at each frequency point to obtain the transfer function value at each frequency point. The frequency domain expression of the second transmission correlation relationship includes the gain value and phase difference value at each frequency point, which are used as the updated second transmission correlation relationship parameters.

[0253] The updated first transmission correlation parameters and the updated second transmission correlation parameters obtained in step five are cascaded and composited according to the aforementioned two-port network cascade composite method. For each frequency point, the gain value of the first transmission correlation at that frequency point is multiplied by the gain value of the second transmission correlation at the same frequency point to obtain the total gain value of the overall transmission function at that frequency point. The phase difference of the first transmission correlation at that frequency point is added to the phase difference of the second transmission correlation at the same frequency point to obtain the total phase difference value of the overall transmission function at that frequency point. After performing the above operations for all frequency points one by one, the complete frequency domain expression of the updated overall transmission function is obtained.

[0254] The cascaded collaborative module sends the updated overall transfer function derived in step six to the collaborative controller via the data bus, and sends a re-identification completion response signal to notify the collaborative controller that the re-identification process has ended.

[0255] After the cascaded collaborative module completes the re-identification of the overall transfer function, the collaborative controller replaces the old overall transfer function with the updated overall transfer function.

[0256] Specifically, while in a continuous waiting state, the cooperative controller receives the updated overall transfer function data and re-identification completion response signal from the cascaded cooperative module via the data bus. The cooperative controller performs integrity verification on the received updated overall transfer function data, such as checksum or CRC check, to confirm that no errors occurred during data transmission. If the verification passes, the replacement process begins; if the verification fails, a retransmission request is sent to the cascaded cooperative module, requesting the retransmission of the updated overall transfer function data.

[0257] The cooperative controller writes the updated overall transfer function that has passed the verification into its internal parameter register used to store the overall transfer function, replacing the old overall transfer function data stored in the register, so that the updated overall transfer function becomes the latest model currently held.

[0258] The co-controller unmasks the interrupt pin and resumes listening to the trigger signal; it switches the working state from waiting for re-identification to complete back to normal operation mode; it resumes the execution of the suspended force control loop routine control tasks, and performs response amplitude prediction and amplitude constraint of the feed speed adjustment based on the updated overall transfer function.

[0259] After the replacement is completed, the co-controller sends a reset command to the identification module, notifying the identification module that the current thermal coupling state has been responded to. After receiving the reset command, the identification module restores the trigger signal to a low level, completing a complete trigger response reset cycle.

[0260] like Figure 4As shown, after the force control loop inside the cooperative controller completes the calculation of the feed rate adjustment amount for the current control cycle, but before outputting the adjustment amount to the force control subsystem, the cooperative controller first intercepts the feed rate adjustment amount and applies it as an input excitation to the overall transfer function currently held by the cascaded cooperative module. Through the mapping relationship of the overall transfer function, the pre-response amplitude that the feed rate adjustment amount will generate at the vibration monitoring port after passing through the force control link and the vibration transmission link is calculated.

[0261] Specifically, in each control cycle, the force control loop calculates the feed rate adjustment Δv(t) according to the deviation between the current measured grinding force and the target value using a PID algorithm. Before Δv(t) is output to the force control subsystem, the co-controller intercepts the adjustment signal from the output buffer of the force control loop and converts the time-domain signal into a frequency-domain representation through a fast Fourier transform—decomposing it into the amplitude and phase information of each frequency component. For each frequency component, its amplitude is multiplied by the total gain of the overall transfer function at the same frequency point, and its phase is added to the total phase difference of the overall transfer function at the same frequency point. After performing the above operations on all frequency components, the frequency-domain representation of the vibration response signal is synthesized. The amplitudes of each frequency component in the frequency-domain representation are combined into a single amplitude, which is the pre-response amplitude Apred. This amplitude represents the expected vibration response magnitude that will be generated at the vibration port if the current feed rate adjustment is output at the current amplitude.

[0262] The collaborative controller has a built-in comprehensive vibration index evaluation logic. The evaluation logic combines the pre-response amplitude with other vibration characteristic parameters measured at the current vibration monitoring port to obtain the comprehensive vibration index value that is expected to be caused by the actual output of the feed speed adjustment.

[0263] A preferred example: The calculation method for the comprehensive vibration index value Vtotal is as follows:

[0264] Extract the root mean square value of the current measured vibration signal within the frequency range covered by the overall transfer function (e.g., 0–200 Hz). , will pre-response amplitude Power superposition with measured values ​​in the frequency band: Where VRMS is the current measured root mean square value of vibration acceleration, Apred is the pre-response amplitude, and α is the weighted fusion coefficient. The value of α ranges from 0.5 to 0.8. The value of α is determined by the least squares fitting method during the grinding machine calibration stage to minimize the mean square error between the predicted response amplitude and the measured response amplitude. The calibrated α is stored in the collaborative controller.

[0265] The collaborative controller compares the comprehensive vibration index value with a pre-stored safety threshold. If the comprehensive vibration index value is lower than or equal to the safety threshold, the collaborative controller determines that the feed rate adjustment is a safe adjustment and allows it to be output normally to the force control subsystem.

[0266] If the comprehensive vibration index value is higher than the safety threshold, the cooperative controller determines that the feed rate adjustment is a risk adjustment and actively reduces the amplitude of the feed rate adjustment before outputting it to the force control subsystem. The cooperative controller only outputs the reduced feed rate adjustment to the force control subsystem when the comprehensive vibration index value corresponding to the reduced feed rate adjustment after remapping by the overall transfer function is not higher than the safety threshold.

[0267] It should be noted that when re-acquiring real-time input / output response data, the cascaded collaborative module uses a short-time sweep frequency signal with limited amplitude superimposed on the current control command as the identification excitation source. The excitation signal is applied within the steady-state cutting segment of continuous grinding. When the current processing state of the grinding machine does not allow the superposition of external excitation signals, the cascaded collaborative module uses the inherent grinding force signal fluctuations during the grinding process as a natural excitation source to collect the input / output response data under natural conditions for identification.

[0268] Example 2

[0269] A collaborative control method for a linear guide grinding machine includes:

[0270] According to the description of the four-terminal impedance module in Example 1, a four-terminal mechanical impedance network with thermal coupling is established with the grinding force port and the heat port as the two ends. The impedance matrix containing mutual impedance parameters is identified and output by applying excitation and collecting response.

[0271] According to the description of the identification module in Embodiment 1, the current value of the mutual impedance parameter is updated recursively on a cycle-by-cycle basis using the impedance matrix as input, the instantaneous deviation between the current value and the reference value and its rate of change are calculated, and a coupling deviation trigger signal is generated when the rate of change exceeds a preset tolerance range.

[0272] According to the description of the cascaded cooperative module in Example 1, the force control subsystem is modeled as a first two-port network, the vibration support subsystem is modeled as a second two-port network, and the first two-port network and the second two-port network are cascaded and combined along the signal flow direction to derive the overall transfer function from the feed speed command to the vibration response.

[0273] According to the description of the collaborative controller in Embodiment 1, when the coupling deviation trigger signal is received, the current control auxiliary function is paused, data is re-acquired, and the cascade modeling and transfer function calculation steps are re-executed to replace the old overall transfer function with the updated overall transfer function.

[0274] According to the description of the cooperative controller in Embodiment 1, before the feed speed adjustment amount is output, it is input into the current global transfer function to calculate the pre-response amplitude at the vibration port, and the comprehensive vibration index is evaluated accordingly. When the comprehensive vibration index exceeds the safety threshold, the amplitude of the feed speed adjustment amount is reduced before it is output.

[0275] Example 3

[0276] The linear guide grinding machine used in this embodiment is a certain model of CNC linear guide dedicated forming grinding machine, with a spindle rated power of 15kW, a maximum spindle speed of 3600r / min, and a grinding wheel linear speed of 25m / s~30m / s. The workpiece being processed is a 45# steel linear guide with a hardness of HRC58~62. The grinding method is plunge forming grinding, and the grinding fluid is a water-based emulsion with a flow rate of 20L / min.

[0277] The data sources shown in Table 1 include:

[0278]

[0279] The grinding force signal F is acquired by a Kistler 9257B piezoelectric triaxial force gauge, which is installed between the worktable and the workpiece fixture. The range is -5kN to +10kN, the sensitivity is approximately -7.5pC / N, and the sampling frequency is set to 10kHz.

[0280] Displacement X is acquired by an eddy current displacement sensor, which is installed between the machine tool bed and the grinding head. The model is Micro-EpsiloneddyNCDT3005, with a range of 2mm, a resolution of 0.25μm, and a sampling frequency of 10kHz synchronized with the force sensor.

[0281] Temperature T is acquired by a K-type embedded thermocouple, with the thermocouple hot junction embedded 0.8 mm below the surface of the workpiece. The temperature measurement range is 0℃~800℃, the response time is no more than 1ms, and the sampling frequency is set to 5kHz.

[0282] The heat flux Q is acquired by a gSKINXR type thin-film heat flux sensor, which is attached to the workpiece surface near the grinding arc area. The range is -200kW / m² to +200kW / m², the response time is no more than 5ms, and the sampling frequency is set to 5kHz synchronously with the temperature sensor.

[0283] After the grinding machine reached thermal equilibrium through 30 minutes of no-load operation, it was continuously ground for 5 strokes at constant cutting parameters: spindle speed 2800 r / min, feed rate 1.2 m / min, and depth of grinding 0.03 mm / pass. Coolant flow rate and pressure were at rated values, and the workpiece material was standard 45# steel. Under these conditions, the force-thermal coupling characteristics of the grinding machine were within a stable and repeatable normal range. The measured reference value of the normal grinding force was approximately 420 N, and the reference value of the grinding arc temperature was approximately 185 °C.

[0284] Under calibration conditions, by applying a 0Hz–200Hz sweep frequency force excitation signal to the grinding force port and a 0Hz–100Hz sweep frequency heat flow excitation signal to the heat port simultaneously, with an amplitude of 20kW / m², the impedance matrix under calibration conditions is obtained by acquiring the response signals and calculating them using a frequency domain identification algorithm.

[0285] The self-impedance of the grinding force port reflects the dynamic stiffness of the machine tool structure ≈ 1.8 × N / m (100Hz reference value) The self-impedance of the hot port reflects the heat transfer characteristics of the grinding arc region ≈ 3.2 × W / (℃·m²); For thermal and mechanical reverse mutual impedance, =F / T, reflecting the reaction coupling of the temperature field on the grinding force signal F ≈2.1N / ℃; The mutual impedance of displacement to heat flow, and its change indirectly reflects the change in the efficiency of force-generated heat coupling.

[0286] in, ref=2.1N / ℃ and The reference value of ref=0.38(W / m²) / m is stored in the non-volatile register of the identification module as the mutual impedance parameter.

[0287] After the grinding machine enters continuous machining mode, the identification module runs recursive identification with a control cycle of Ts=1ms. The identification algorithm converges within the first 10 cycles, and the estimated value of the mutual impedance parameter stabilizes at... ≈2.15N / ℃ The instantaneous deviation Δ stabilizes near the calibration reference value. ≈0.05N / ℃, Δ ≈0.02 (W / m²) / m, rate of change The data obtained during the calibration phase were kept within a small range and within the preset tolerance range. =1.2(N / ℃) / s =0.5(℃ / N) / s, k=3.0, tolerance range is -3.6, +3.6(N / ℃) / s and -(0.5), +(0.5)((W / () / m) / s within.

[0288] like Figure 6 As shown, within the current control cycle, the elements in the real-time impedance matrix output online by the four-terminal impedance module are as follows: =1.2340、 =0.0116、 =0.0085、 =0.5670, Figure 6 The self-resistance and mutual resistance are displayed as dimensionless normalized schematic values ​​to illustrate the matrix structure and calculation process; the instantaneous deviation of the mutual impedance parameters calculated synchronously by the identification module is Δ. =-0.00039、Δ =0.00050, the corresponding rate of change is =0.0391、 =0.0500, at this point, since both the deviation and the rate of change are within the preset tolerance range, the statistical results obtained during the calibration phase are... =1.2(N / ℃) / s =0.5(℃ / N) / s, k=3.0, the identification module did not trigger the coupling deviation trigger signal, and the system is in normal monitoring state.

[0289] When the grinding process reached the 45th second, the coolant flow rate suddenly dropped to 30% of the rated value due to blockage in the coolant lines. The temperature in the grinding arc zone rose sharply from 195°C to approximately 320°C within 3 seconds. The identification module detected this. The estimated value changed significantly over five consecutive control periods, with the instantaneous deviation Δ The rate of change jumps to more than 10 times the initial value. The average value over 5 cycles far exceeded the upper limit of the tolerance range. The identification module immediately determined that a sudden deviation had occurred in the thermal coupling state and generated a coupling deviation trigger signal, which was sent to the cooperative controller.

[0290] After the co-controller captures the trigger signal via a hardware interrupt, it pauses the regular force control task and sends a forced re-identification command to the cascaded co-control module. The cascaded co-control module then re-acquires data under the current operating conditions after the coolant flow is restored—applying a 0Hz–200Hz sweep frequency speed command excitation signal to the force control subsystem input port while simultaneously acquiring the grinding force response; and applying a 0Hz–1000Hz sweep frequency force excitation signal to the vibration support subsystem input port while simultaneously acquiring the vibration acceleration response. Through re-identification and cascaded composite derivation, the total gain of the updated overall transfer function at 100Hz changes from 0.38 (W / m²) / m in the old function to 0.62 (m / s²) / (mm / min)—meaning that the same feed rate adjustment will generate approximately 63% additional vibration response at the vibration port after coolant failure. The co-controller replaces the old function with the updated overall transfer function and resumes regular control.

[0291] In a certain control cycle after the resumption of normal control, the force control loop calculates the required feed rate adjustment Δv(t) = +0.15 mm / min based on the deviation between the current measured grinding force of 470 N and the target value of 420 N using the PID control algorithm.

[0292] Before Δv(t) is output to the force control subsystem, the co-controller intercepts the adjustment amount and inputs it into the updated overall transfer function for prediction. After converting Δv(t) to the frequency domain, each frequency component is multiplied by the total gain of the overall transfer function (0.62(m / s²) / (mm / min) at 100Hz), and the resulting composite is the pre-response amplitude Apred = 2.8m / s².

[0293] The accelerometer mounted on the spindle bearing housing measures the current root mean square value of vibration acceleration to be 3.5 m / s². The comprehensive vibration index evaluation logic weights and merges the pre-response amplitude with the measured value—taking a weighting coefficient k=0.6, Vtotal=3.5+0.6×2.8=5.18 m / s².

[0294] The safety threshold Vthreshold is set to 4.5 m / s² according to the standard and the grinding machine's machining accuracy requirements. Vtotal = 5.18 m / s² > 4.5 m / s², and the co-controller determines that this adjustment amount is a risk adjustment amount.

[0295] The cooperative controller calculates the constraint coefficient α = Vthreshold / Vtotal = 4.5 / 5.18 ≈ 0.869. Considering the slight nonlinearity of the actual system, a safety factor β = 0.95 is taken, and the final constraint coefficient αfinal = 0.95 × 0.869 ≈ 0.826. Multiplying the original adjustment Δv(t) = +0.15 mm / min by αfinal yields the final adjustment Δvfinal ≈ +0.124 mm / min, which is directly output to the force control subsystem for execution without the need for multiple iterations for verification.

[0296] The collaborative controller outputs the final adjustment amount, Δvfinal = +0.124 mm / min, after amplitude reduction, to the force control subsystem for execution, rather than the original risk adjustment amount. The entire process of prediction and constraint described above is completed before the actual output of the feed rate adjustment amount, which belongs to feedforward prediction and constraint.

[0297] It is important to note that the constructions and arrangements of this application shown in several different exemplary embodiments are merely illustrative. Although only two embodiments are described in detail in this disclosure, those who consult this disclosure will readily understand that many modifications are possible without substantially departing from the novel teachings and advantages of the subject matter described in this application. These modifications may include, for example, changes in the size, dimensions, structure, shape, and proportions of various elements, as well as parameter values ​​(e.g., temperature, pressure, etc.), installation arrangements, the use of materials, colors, orientations, etc. For example, an element shown as integrally formed may be composed of multiple parts or elements, the position of elements may be inverted or otherwise altered, and the nature or number or position of discrete elements may be changed or altered. Therefore, all such modifications are intended to be included within the scope of this application. The order or sequence of any process or method steps may be changed or rearranged by alternative embodiments. Any "apparatus plus function" clause is intended to cover, and not only structurally equivalent but also equivalent structures, the structures performing the functions described herein. Other substitutions, modifications, alterations, and omissions may be made in the design, operation, and arrangement of the exemplary embodiments without departing from the scope of this application. Therefore, this application is not limited to a particular embodiment, but extends to various modifications that still fall within the scope of the appended claims.

[0298] Furthermore, in order to provide a concise description of exemplary embodiments, not all features of actual embodiments (i.e., those features that are not relevant to the best mode of performing this application as currently considered, or those features that are not relevant to implementing this application) may be omitted.

[0299] It should be understood that numerous specific implementation decisions can be made during the development of any practical implementation, such as in any engineering or design project. Such development efforts may be complex and time-consuming, but for those of ordinary skill in the art who benefit from this disclosure, the development effort will be a routine task in design, manufacturing, and production without requiring extensive experimentation.

[0300] It should be noted that the above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit it. Although this application has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of this application without departing from the spirit and scope of the technical solutions of this application, and all such modifications and substitutions should be covered within the scope of the claims of this application.

Claims

1. A collaborative control system for a linear guide grinding machine, characterized in that, include: The four-terminal impedance module uses the grinding force port and the heat port as the two ends to establish a four-terminal mechanical impedance network for the thermo-coupling of the linear guide grinding machine, and outputs an impedance matrix containing mutual impedance parameters. The identification module calculates the current value of the mutual impedance parameter in real time and generates a coupling deviation trigger signal based on the rate of change of the deviation between the current value and the reference value. The cascaded collaborative module models the force control subsystem as a first two-port network and the vibration support subsystem as a second two-port network, and calculates the overall transfer function through the cascaded characteristics of the two-port networks; The collaborative controller is connected to the identification module and the cascaded collaborative module respectively. When the collaborative controller receives the coupling deviation trigger signal, it controls the cascaded collaborative module to re-identify the overall transfer function. When the force control loop needs to output the feed speed adjustment amount, the collaborative controller predicts the response amplitude of the current speed adjustment amount at the vibration port based on the current overall transfer function. When the response amplitude causes the comprehensive vibration index to exceed the safety threshold, it implements amplitude constraint on the feed speed adjustment amount.

2. The linear guide grinding machine collaborative control system according to claim 1, characterized in that: The grinding force port is the first port of the four-terminal mechanical impedance network, used to connect the mechanical signal variables of the grinding force during the grinding process. The mechanical signal variables include the grinding force signal and the displacement signal caused by the grinding force. The hot port is the second port of the four-terminal mechanical impedance network, used to connect the thermal signal variables of grinding heat during the grinding process. The thermal signal variables include the temperature signal of the grinding zone and the heat flow signal. The grinding force port and the heat port serve as the input and output ports of the four-terminal mechanical impedance network, respectively. The four-terminal mechanical impedance network uses the mechanical signal variable of the grinding force port as the input excitation and the thermal signal variable of the heat port as the output response to establish the transmission relationship between the grinding force field and the temperature field.

3. The linear guide grinding machine collaborative control system according to claim 2, characterized in that: The mechanical signal variables of the grinding force port and the thermal signal variables of the heat port are collected respectively. The mechanical signal variables include the grinding force signal F and the displacement X under the action of the grinding force. The thermal signal variables include the grinding zone temperature T and the heat flux Q. Using the signal variables of the grinding force port and the heat port as boundary conditions, the network equation of the four-terminal mechanical impedance network is established. The diagonal elements of the impedance matrix and The self-impedance parameters of the grinding force port and the heat port are respectively, characterizing the intrinsic transmission characteristics between the physical quantities of each port; the off-diagonal elements of the impedance matrix and These are mutual impedance parameters. Characterizes the coupling and transmission properties of the reaction between the temperature field and the grinding force field. Characterizes the coupling and transfer properties of the displacement field to the heat flow field. The change in the amount of grinding force indirectly reflects the coupling efficiency of grinding force work being converted into grinding heat. A preset excitation signal is applied to the grinding force port and the heat port through the network equation, and the response signal is collected. The parameter values ​​of each element in the impedance matrix are calculated by the identification algorithm. The parameter values ​​include self-impedance parameters and mutual impedance parameters. The calculated impedance matrix of the self-impedance parameter and the mutual impedance parameter is used as the output of the four-terminal mechanical impedance network.

4. The linear guide grinding machine collaborative control system according to claim 1, characterized in that: The identification module takes the impedance matrix output by the four-terminal impedance module as input and continuously captures the real-time response data of the grinding force port and the heat port during the dynamic cutting process when the grinding machine is in continuous processing state. The identification module uses the force port excitation signal and the heat port response signal acquired in the current control cycle as information input, and uses the parameter estimation results of the historical cycle as a priori reference. By recursively comparing and correcting the information input with the priori reference cycle by cycle, the estimated value of the mutual impedance parameter is updated.

5. The linear guide grinding machine collaborative control system according to claim 1, characterized in that: The identification module also stores the reference value of the mutual impedance parameter of the linear guide grinding machine under the calibration conditions. The reference value is used to characterize the reference features of the grinding machine under normal thermo-coupling conditions. After acquiring the current value of the mutual impedance parameter in real time, the identification module first calculates the instantaneous deviation between the current value and the reference value. The identification module performs a time-series comparison between the instantaneous deviation at the current moment and the instantaneous deviation at historical moments, and calculates the rate of change of the instantaneous deviation over time. When the rate of change exceeds the preset rate tolerance range, it is determined that the current thermo-coupling state is undergoing a sudden deviation that is not gradual. Simultaneously with the determination outputting a positive result, the identification module immediately generates the coupling deviation trigger signal and sends the trigger signal to the cooperative controller to trigger the cooperative controller to initiate a re-identification of the overall transfer function.

6. The linear guide grinding machine collaborative control system according to claim 5, characterized in that: The overall transfer function is used to describe the response generated at the vibration port after the feed speed command passes through the force control link and the vibration transmission link; The cascaded collaborative module defines the physical variables of the input port of the force control subsystem as the feed speed command signal output after being adjusted by the collaborative controller, and defines the physical variables of the output port as the dynamic grinding force signal actually output after the feed speed command is acted upon by the force control execution link. It also establishes a first transmission correlation between the input port and the output port, reflecting the transmission characteristics from the speed command to the grinding force, thereby abstracting the force control subsystem as a whole into a first two-port network with unidirectional signal transmission characteristics. The cascaded collaborative module defines the physical variable of the input port of the vibration support subsystem as the dynamic grinding force signal output by the output port of the force control subsystem, and defines the physical variable of its output port as the vibration response signal generated at the actual vibration monitoring port after the dynamic grinding force is transmitted through the machine tool structure. It also establishes a second transmission correlation between the input port and the output port, reflecting the transmission characteristics from grinding force to vibration response, thereby abstracting the entire vibration support subsystem into a second two-port network with unidirectional signal transmission characteristics.

7. The linear guide grinding machine collaborative control system according to claim 1, characterized in that: After completing the modeling of the first two-port network and the second two-port network, the cascaded coordination module obtains the first transmission correlation parameter corresponding to the first two-port network and the second transmission correlation parameter corresponding to the second two-port network. The cascaded coordination module cascades the second two-port network after the first two-port network according to the direction of signal flow in the actual physical link, so that the output port of the first two-port network and the input port of the second two-port network logically coincide into the same cascaded node. The cascaded collaborative module first maps the feed speed command signal into a grinding force signal at an intermediate node through the first transmission association relationship, and then maps the grinding force signal at the intermediate node into a vibration response signal at the final node through the second transmission association relationship.

8. The linear guide grinding machine collaborative control system according to claim 1, characterized in that: When the cooperative controller receives the coupling deviation trigger signal, it controls the cascaded cooperative module to re-identify the overall transfer function. The cooperative controller monitors the coupling deviation trigger signal output by the identification module in real time. The trigger signal is a pulse-type or level-type trigger identifier, which is used to indicate to the cooperative controller that the current thermal coupling state has suddenly deviated. When the cooperative controller detects the rising edge of the coupling deviation trigger signal, it immediately sends a forced re-identification command to the cascaded cooperative module. Upon receiving the forced re-identification command, the cascaded cooperative module suspends the current control auxiliary function based on the old overall transfer function, re-acquires the real-time input-output response data of the force control subsystem and the vibration support subsystem under the current operating conditions, and re-executes the transmission correlation identification of the first two-port network and the second two-port network and the composite derivation of the overall transfer function based on the re-acquired data to obtain an updated overall transfer function that matches the current thermo-coupling state. After the cascaded collaborative module completes the re-identification of the overall transfer function, the collaborative controller replaces the old overall transfer function with the updated overall transfer function.

9. The linear guide grinding machine collaborative control system according to claim 1, characterized in that: After the force control loop inside the cooperative controller completes the calculation of the feed rate adjustment amount for the current control cycle, but before outputting the adjustment amount to the force control subsystem, the cooperative controller first intercepts the feed rate adjustment amount and applies it as an input excitation to the overall transfer function currently held by the cascaded cooperative module. Through the mapping relationship of the overall transfer function, the pre-response amplitude that the feed rate adjustment amount will generate at the vibration monitoring port after passing through the force control link and the vibration transmission link is calculated. The collaborative controller has a built-in comprehensive vibration index evaluation logic. The evaluation logic integrates the pre-response amplitude with other vibration characteristic parameters measured at the current vibration monitoring port to obtain the comprehensive vibration index value that is expected to be caused by the actual output of the feed speed adjustment. The collaborative controller compares the comprehensive vibration index value with a pre-stored safety threshold. If the comprehensive vibration index value is lower than or equal to the safety threshold, the collaborative controller determines that the feed speed adjustment is a safe adjustment and allows it to be output normally to the force control subsystem. If the comprehensive vibration index value is higher than the safety threshold, the cooperative controller determines that the feed rate adjustment is a risk adjustment and actively reduces the amplitude of the feed rate adjustment before outputting it to the force control subsystem. The cooperative controller only outputs the reduced feed rate adjustment to the force control subsystem when the comprehensive vibration index value corresponding to the reduced feed rate adjustment after remapping by the overall transfer function is not higher than the safety threshold.

10. A method for coordinated control of a linear guide grinding machine, used to implement the coordinated control system for a linear guide grinding machine as described in any one of claims 1-9, characterized in that, include: A four-terminal mechanical impedance network with thermal coupling is established using the grinding force port and the heat port as the two ends. An impedance matrix containing mutual impedance parameters is identified and output by applying excitation and acquiring the response. Using the impedance matrix as input, the current value of the mutual impedance parameter is updated recursively cycle by cycle. The instantaneous deviation between the current value and the reference value and its rate of change are calculated. When the rate of change exceeds a preset tolerance range, a coupling deviation trigger signal is generated. The force control subsystem is modeled as a first two-port network, the vibration support subsystem is modeled as a second two-port network, and the first two-port network and the second two-port network are cascaded and combined along the signal flow direction to derive the overall transfer function from the feed speed command to the vibration response. When the coupling deviation trigger signal is received, the current control auxiliary function is paused, data is re-acquired, and the cascade modeling and transfer function calculation steps are re-executed to replace the old overall transfer function with the updated overall transfer function. Before outputting the feed rate adjustment amount, it is input into the current global transfer function to calculate the pre-response amplitude at the vibration port, and the comprehensive vibration index is evaluated accordingly. When the comprehensive vibration index exceeds the safety threshold, the amplitude of the feed rate adjustment amount is reduced before output.