Distributed lubrication control method and system with single-point state feedback

Abstract

This invention relates to the field of industrial control systems and discloses a distributed lubrication control method and system with single-point state feedback. The method includes: a control unit running built-in digital process control instructions to retrieve a control parameter matrix and modulate pulse waveform sequences using shaft motion data; controlling multiplexed actuators to activate in a time-division manner and causing adjacent multiplexed actuators to generate a reference overlapping phase difference, thus generating a pressure interference signal within the pressure feedback manifold; acquiring the pressure interference signal at a sampling frequency of 1000Hz and generating a time-domain pressure data sequence; calculating and separating the flow resistance state parameters using autocorrelation transformation and time-domain differential integration; and adaptively correcting the drive pulse width of the corresponding branch by comparing the reference impedance range through a boundary comparison program. This invention eliminates the phase interleaving and aliasing of pressure waves caused by fluid mass transfer delay and pipeline elastic expansion, improving the stability and real-time response capability of the dynamic closed-loop feedback regulation system under underactuated constraints.

Description

Technical Field

[0001] This invention relates to the field of industrial control system technology, and more specifically, to a distributed lubrication control method and system with single-point state feedback. Background Technology

[0002] Currently, in multi-axis linkage precision manufacturing equipment, distributed lubrication systems ensure the operational stability of each moving pair. In order to reduce the interface resource occupation of the control system, the industry generally adopts a single-point feedback control system with a single pressure sensor on the main lubrication pipeline. The controller uses a time-division sequence arranged on the time axis to identify the local operating conditions of branches of multiple discrete pipelines in turn. When the reversing solenoid valve of a specific branch opens, the fluid in the main pipeline is replenished to that branch, thereby generating transient pressure swing and recovery waveforms at the single-point pressure sensor in the main pipeline. The control unit analyzes the characteristic structure of the time-domain response waveform to invert the flow resistance state of the corresponding branch.

[0003] Besides the inherent limitations imposed by hardware aspects such as single-point sensing and the physical arrangement of discrete piping, existing control methods also have significant shortcomings in dealing with high-frequency alternating cutting conditions. For example, Chinese invention patent application CN114263837A discloses a centralized lubrication intelligent control system and method, which uses a control cabinet to perform time-domain polling of the opening times of all lubrication points and opens the corresponding distribution valve when the opening time and pressure meet the requirements. However, this time-sharing control scheme based on discrete time-domain polling implicitly relies on the idealized premise that the actions of each lubrication branch are absolutely isolated and do not interfere with each other on the time axis. However, under the continuous high-frequency cutting conditions of machining centers, the lubrication action cycle is greatly compressed, leading to time-sharing isolation detection. Faced with fluid dynamic constraints, conventional improvement approaches to address signal confusion typically focus on extending dead-zone latency to eliminate interference or increasing overall oil supply to avoid under-lubrication risks. However, extending latency disrupts the machining center's tight cycle time, and increasing overall oil supply not only leads to lubricant overflow and fluid contamination but also fails to eliminate wear caused by coking and blockage in local pipelines. Furthermore, solutions such as adding isolation check valves to each branch or adding sensor hardware at the terminal not only increase the complexity of internal wiring and hardware costs but also make sensor cables more prone to fatigue breakage in dynamic reversing environments. This puts the control system in a difficult dilemma between increasing hardware channel redundancy and improving decoupling control robustness.

[0004] Therefore, the technical problem to be solved by this invention is how to achieve dynamic decoupling between the single-point pressure feedback signal and the multi-output execution end conditions under continuous excitation conditions with multivariable nonlinear coupling by utilizing control loop logic reconstruction, and avoiding state determination deadlock caused by the intertwining of elastic expansion of flexible pipeline and mass transfer time delay phase. Summary of the Invention

[0005] To address the problems in the background art, this invention provides a distributed lubrication control method and system with single-point state feedback, comprising the following steps:

[0006] Step S101: Retrieve the preset control parameter matrix from the internal control unit and obtain the axis excitation signal obtained by feedforward deduction of axis motion data from the CNC interpolation buffer. Control the axis excitation signal and the control parameter matrix to perform phase modulation to generate a pulse waveform sequence that controls the time-division multiplexing actuator to start.

[0007] Step S102: The pulse waveform sequence control timing and counting module is used to drive the multi-channel distribution actuator to start in a time-sharing manner, so that the working period of two adjacent multi-channel distribution actuators generates a preset reference overlapping phase difference, and a pressure interference signal is excited in the pressure feedback manifold of fluid transmission.

[0008] Step S103: Using the pressure transformation unit on the pressure feedback manifold, pressure interference signals are acquired at a sampling frequency of 1000Hz to generate a time-domain pressure data sequence containing continuous pressure sampling points.

[0009] Step S104: Input the time-domain pressure data sequence into the demodulation program, perform sliding window autocorrelation transformation and time-domain differential integration calculation within the time period corresponding to the reference overlapping phase difference, and separate the dimensionless flow resistance state parameters that characterize the resistance characteristics of each lubrication branch.

[0010] Step S105: The flow resistance state parameter is compared with the preset reference impedance range through the boundary comparison program. When the flow resistance state parameter exceeds the reference impedance range, the driving pulse width of the corresponding branch in the control parameter matrix is ​​adaptively corrected.

[0011] Preferably, before comparing the flow resistance state parameters with the reference impedance range in step S105, when the operating condition meets the preset resting period condition, the standard verification branch in the multi-way actuator is controlled to open for a preset time to output a test pulse and acquire the corresponding response waveform. The fluid viscosity variation coefficient is calculated based on the falling edge slope of the response waveform, and the reference impedance range is shifted and corrected using the fluid viscosity variation coefficient to compensate for the flow resistance drift caused by changes in ambient temperature.

[0012] Preferably, in step S105, when the flow resistance state parameter exceeds the upper limit of the reference impedance range, the driving pulse width of the corresponding branch in the control parameter matrix is ​​gradually increased according to the resistance deviation of the flow resistance state parameter from the reference impedance range by a preset step increment.

[0013] Preferably, step S104 includes the following sub-steps: Step S1041, within the time period corresponding to the reference overlapping phase difference, the time-domain pressure data sequence is extracted by a moving window, the autocorrelation function of the pressure interference signal within the moving window is calculated, and the autocorrelation change curve is obtained; Step S1042, the differential change rate of the autocorrelation change curve is calculated, and the differential change rate is integrated over time to separate the flow resistance state parameters.

[0014] Preferably, step S103 includes the following sub-steps: step S1031, the pressure conversion unit continuously acquires the fluid pressure in the pressure feedback manifold at a sampling frequency of 1000Hz and filters out noise in the preset frequency band to obtain continuous pressure data points; step S1032, the continuous pressure data points are combined according to the acquisition time sequence to construct an interference waveform and generate a time-domain pressure data sequence.

[0015] Preferably, when the driving pulse width of the corresponding branch in the control parameter matrix is ​​gradually increased according to the preset step increment, the flow resistance state parameter is reduced by a smoothing filter module, and the median filter module is used to calculate the median of the resistance deviation in multiple consecutive control cycles to hedge against fluctuations.

[0016] Preferably, when determining the reference impedance range during the calibration stage, each independent lubrication branch is opened sequentially, and the initial flow resistance value of each independent lubrication branch under normal flow conditions is obtained using the pressure transformation unit. Based on the initial flow resistance value plus and minus the preset allowable deviation, the reference impedance range corresponding to each independent lubrication branch is determined.

[0017] Preferably, when using the resistance evolution trend to provide an early warning of insufficient oil supply, the flow resistance state parameters of each independent lubrication branch are continuously recorded in each control cycle to generate a resistance change sequence over time; the rate of change of the change sequence over time is calculated to obtain the flow resistance rate parameter; when the flow resistance rate parameter of an independent lubrication branch continuously exceeds the preset upper limit of the rate of change, the corresponding early warning signal of insufficient oil supply is output.

[0018] Preferably, in step S105, when the flow resistance state parameter is lower than the lower limit of the reference impedance range, the drive pulse width of the corresponding branch in the control parameter matrix is ​​corrected to zero, and a leakage alarm signal is output.

[0019] A distributed lubrication control system with single-point state feedback is provided to implement a distributed lubrication control method with single-point state feedback. The system includes:

[0020] The signal modulation module is used to retrieve the preset control parameter matrix inside the self-control unit and obtain the axis excitation signal derived from the axis motion data feedforward of the CNC interpolation buffer. It controls the axis excitation signal to be phase-modulated with the control parameter matrix to generate a pulse waveform sequence for controlling the time-division control module to start, and transmits the pulse waveform sequence to the timing and counting module.

[0021] The timing and counting module is used to receive pulse waveform sequences and use the pulse waveform sequences to drive the distribution control module to start in a time-division manner, so that the working periods of two adjacent distribution control modules generate a preset reference overlapping phase difference, so as to generate a pressure interference signal in the pressure feedback main pipe.

[0022] The distribution control module, connected to the discrete lubrication pipeline, is used to activate in a time-sharing manner under the drive of the timing and counting module;

[0023] The pressure acquisition and transformation module is used to acquire the pressure interference signal in the pressure feedback manifold at a sampling frequency of 1000Hz, generate a time-domain pressure data sequence containing continuous pressure sampling points, and output the time-domain pressure data sequence to the data demodulation module.

[0024] The data demodulation module is used to receive time-domain pressure data sequences and, under the control of the demodulation program, perform sliding window autocorrelation transformation and time-domain differential integration calculation within the time period corresponding to the reference overlapping phase difference. It separates the dimensionless flow resistance state parameters that characterize the resistance characteristics of each lubrication branch and inputs the flow resistance state parameters to the boundary comparison correction module.

[0025] The boundary comparison correction module is used to receive the flow resistance state parameters and, under the control of the boundary comparison process, compare the flow resistance state parameters with the preset reference impedance range. When the flow resistance state parameters exceed the reference impedance range, the module adaptively corrects the driving pulse width of the corresponding branch in the control parameter matrix and feeds back the correction result to the signal modulation module.

[0026] The present invention has the following beneficial effects:

[0027] 1. In distributed lubrication control with single-point state feedback, the motion axis trajectory sequence in the CNC system interpolation buffer is read through the bus, and the acceleration and velocity characteristics of the corresponding axis are extracted. The parameter matrix is ​​adaptively reconstructed using the predicted lubrication consumption gradient. A microsecond-level hardware timer is scheduled to cause the opening time of adjacent solenoid valves to overlap in phase shift. A nonlinear interference waveform containing branch impedance state is actively constructed at the single-point sensor of the main pipeline. This operating mechanism relies on digital process control instructions to reconstruct the control loop logic, getting rid of the requirement that the state evolution of each discrete branch is absolutely isolated on the time axis in traditional isolated time-division detection. It eliminates the phase interleaving between the residual unloading wave and the subsequent loading wave caused by the mass transfer delay of the flexible pipeline, and transforms the overlapping interference signal into a known data stream carrying flow resistance characteristics.

[0028] 2. The demodulation module receives the time-domain discrete feature sequence generated by high-frequency sampling conversion. Within the bound phase shift overlap time span, it calls the internal sliding window autocorrelation operator to calculate the time-domain integral of the differential partial derivative of the interference envelope, and extracts the independent transient flow resistance contribution parameters of each of the two adjacent lubrication branches. This multi-channel decoupling and reconstruction mechanism transforms the spatially distributed node state differences into a nonlinear evolution process of the time envelope, constructing a multi-channel closed-loop feedback regulation logic suitable for industrial automation control. This solves the low-pass filtering distortion problem caused by the radial elastic expansion of the flexible pipe wall in traditional methods, avoids signal confusion of discrete terminals under continuous overlapping excitation, and provides stable data support for the subsequent state machine to determine whether coking, blockage or damage leakage has occurred in each branch oil circuit.

[0029] 3. The identification module integrates a multi-level timing phase differential impedance boundary follow-up compensation mechanism. By monitoring the CNC working condition state machine in real time, the control unit generates a verification pulse using a standard verification branch with constant impedance during a determined pre-rest period. Based on the drop slope of the feedback standard response waveform, the current viscosity degradation factor is dynamically calculated. The discrimination boundary of the comparison operator is translated online. This collaborative self-calibration path forcibly eliminates the global flow resistance parameter offset caused by the systematic increase in temperature, enabling the control loop to adaptively follow the physical and chemical changes of the fluid medium. This avoids impedance overlap false alarms under continuous heavy-load cutting thermal interference and ensures the accuracy of multi-branch state diagnosis under complex working conditions. Attached Figure Description

[0030] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings, in which several embodiments of the invention are illustrated by way of example and not limitation, wherein:

[0031] Figure 1 This is a flowchart of the distributed lubrication control method with single-point state feedback according to the present invention.

[0032] Figure 2 This is a structural diagram of the distributed lubrication control system with single-point state feedback according to the present invention. Detailed Implementation

[0033] The principles and spirit of the present invention will now be described with reference to several exemplary embodiments in conjunction with the accompanying drawings. It should be understood that these embodiments are provided merely to enable those skilled in the art to better understand and implement the present invention, and are not intended to limit the scope of the present invention in any way. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to fully convey the scope of the present invention to those skilled in the art.

[0034] A distributed lubrication control method with single-point state feedback includes the following steps:

[0035] Step S101: Retrieve the preset control parameter matrix from the internal control unit and obtain the axis excitation signal obtained by feedforward deduction of axis motion data from the CNC interpolation buffer. Control the axis excitation signal and the control parameter matrix to perform phase modulation to generate a pulse waveform sequence that controls the time-division multiplexing actuator to start.

[0036] Step S102: The pulse waveform sequence control timing and counting module is used to drive the multi-channel distribution actuator to start in a time-sharing manner, so that the working period of two adjacent multi-channel distribution actuators generates a preset reference overlapping phase difference, and a pressure interference signal is excited in the pressure feedback manifold of fluid transmission.

[0037] Step S103: Using the pressure transformation unit on the pressure feedback manifold, pressure interference signals are acquired at a sampling frequency of 1000Hz to generate a time-domain pressure data sequence containing continuous pressure sampling points.

[0038] Step S104: Input the time-domain pressure data sequence into the demodulation program, perform sliding window autocorrelation transformation and time-domain differential integration calculation within the time period corresponding to the reference overlapping phase difference, and separate the dimensionless flow resistance state parameters that characterize the resistance characteristics of each lubrication branch.

[0039] Step S105: The flow resistance state parameter is compared with the preset reference impedance range through the boundary comparison program. When the flow resistance state parameter exceeds the reference impedance range, the driving pulse width of the corresponding branch in the control parameter matrix is ​​adaptively corrected.

[0040] Preferably, before comparing the flow resistance state parameters with the reference impedance range in step S105, when the operating condition meets the preset resting period condition, the standard verification branch in the multi-way actuator is controlled to open for a preset time to output a test pulse and acquire the corresponding response waveform. The fluid viscosity variation coefficient is calculated based on the falling edge slope of the response waveform, and the reference impedance range is shifted and corrected using the fluid viscosity variation coefficient to compensate for the flow resistance drift caused by changes in ambient temperature.

[0041] Preferably, in step S105, when the flow resistance state parameter exceeds the upper limit of the reference impedance range, the driving pulse width of the corresponding branch in the control parameter matrix is ​​gradually increased according to the resistance deviation of the flow resistance state parameter from the reference impedance range by a preset step increment.

[0042] Preferably, step S104 includes the following sub-steps: Step S1041, within the time period corresponding to the reference overlapping phase difference, the time-domain pressure data sequence is extracted by a moving window, the autocorrelation function of the pressure interference signal within the moving window is calculated, and the autocorrelation change curve is obtained; Step S1042, the differential change rate of the autocorrelation change curve is calculated, and the differential change rate is integrated over time to separate the flow resistance state parameters.

[0043] Preferably, step S103 includes the following sub-steps: step S1031, the pressure conversion unit continuously acquires the fluid pressure in the pressure feedback manifold at a sampling frequency of 1000Hz and filters out noise in the preset frequency band to obtain continuous pressure data points; step S1032, the continuous pressure data points are combined according to the acquisition time sequence to construct an interference waveform and generate a time-domain pressure data sequence.

[0044] Preferably, when the driving pulse width of the corresponding branch in the control parameter matrix is ​​gradually increased according to the preset step increment, the flow resistance state parameter is reduced by a smoothing filter module, and the median filter module is used to calculate the median of the resistance deviation in multiple consecutive control cycles to hedge against fluctuations.

[0045] Preferably, when determining the reference impedance range during the calibration stage, each independent lubrication branch is opened sequentially, and the initial flow resistance value of each independent lubrication branch under normal flow conditions is obtained using the pressure transformation unit. Based on the initial flow resistance value plus and minus the preset allowable deviation, the reference impedance range corresponding to each independent lubrication branch is determined.

[0046] Preferably, when using the resistance evolution trend to provide an early warning of insufficient oil supply, the flow resistance state parameters of each independent lubrication branch are continuously recorded in each control cycle to generate a resistance change sequence over time; the rate of change of the change sequence over time is calculated to obtain the flow resistance rate parameter; when the flow resistance rate parameter of an independent lubrication branch continuously exceeds the preset upper limit of the rate of change, the corresponding early warning signal of insufficient oil supply is output.

[0047] Preferably, in step S105, when the flow resistance state parameter is lower than the lower limit of the reference impedance range, the drive pulse width of the corresponding branch in the control parameter matrix is ​​corrected to zero, and a leakage alarm signal is output.

[0048] A distributed lubrication control system with single-point state feedback is provided to implement a distributed lubrication control method with single-point state feedback. The system includes:

[0049] The signal modulation module is used to retrieve the preset control parameter matrix inside the self-control unit and obtain the axis excitation signal derived from the axis motion data feedforward of the CNC interpolation buffer. It controls the axis excitation signal to be phase-modulated with the control parameter matrix to generate a pulse waveform sequence for controlling the time-division control module to start, and transmits the pulse waveform sequence to the timing and counting module.

[0050] The timing and counting module is used to receive pulse waveform sequences and use the pulse waveform sequences to drive the distribution control module to start in a time-division manner, so that the working periods of two adjacent distribution control modules generate a preset reference overlapping phase difference, so as to generate a pressure interference signal in the pressure feedback main pipe.

[0051] The distribution control module, connected to the discrete lubrication pipeline, is used to activate in a time-sharing manner under the drive of the timing and counting module;

[0052] The pressure acquisition and transformation module is used to acquire the pressure interference signal in the pressure feedback manifold at a sampling frequency of 1000Hz, generate a time-domain pressure data sequence containing continuous pressure sampling points, and output the time-domain pressure data sequence to the data demodulation module.

[0053] The data demodulation module is used to receive time-domain pressure data sequences and, under the control of the demodulation program, perform sliding window autocorrelation transformation and time-domain differential integration calculation within the time period corresponding to the reference overlapping phase difference. It separates the dimensionless flow resistance state parameters that characterize the resistance characteristics of each lubrication branch and inputs the flow resistance state parameters to the boundary comparison correction module.

[0054] The boundary comparison correction module is used to receive the flow resistance state parameters and, under the control of the boundary comparison process, compare the flow resistance state parameters with the preset reference impedance range. When the flow resistance state parameters exceed the reference impedance range, the module adaptively corrects the driving pulse width of the corresponding branch in the control parameter matrix and feeds back the correction result to the signal modulation module.

[0055] Example 1: In a high-speed, heavy-duty, multi-axis linkage machining center with a distributed lubrication network, continuous high-frequency cutting occurs. The spindle bearing, guide rail slider, and lead screw pair generate dynamic and uneven lubrication consumption. The distributed lubrication network is connected by a main pipeline to multiple discrete electro-hydraulic valves. A single pressure transformation unit for single-point status feedback is installed on the main pipeline. Mass transfer delay occurs during long-distance fluid transmission, and the flexible pipe wall undergoes radial elastic deformation and volume dissipation. Simultaneously, vibration and noise interference caused by high-frequency cutting cause the residual unloading wave generated by the closure of the previous lubrication branch and the loading step wave generated by the opening of the next lubrication branch to produce nonlinear phase interleaving and aliasing interference in the pressure feedback main pipe. This results in distortion of the total pressure waveform corresponding to the time-division isolated detection on the time axis, making it impossible to directly identify the transient flow resistance state of each friction pair node. There is a risk of under-lubrication or coking blockage in local lubrication branches, leading to component damage.

[0056] The control unit retrieves a preset control parameter matrix and obtains the axis excitation signal generated by feedforward extrapolation of axis motion data in the CNC interpolation buffer. The axis excitation signal is then phase-modulated with the control parameter matrix to generate a pulse waveform sequence that controls the time-division multiplexing actuator. The control execution entity is a hardware timer / counter integrated within the control unit. This counter performs a discrete-time algebraic summation on the pre-stored reference start-up pulse steps of each independent channel in the control parameter matrix and the timing shift generated by the linear mapping of the axis excitation signal. The summation result is converted into the leading edge moment of the start-up level of each directional solenoid valve control register. This ensures that when two adjacent directional solenoid valves alternately operate, the start-up period of the preceding directional solenoid valve overlaps with the start-up period of the following directional solenoid valve by a fixed discrete working period of 5ms to 15ms. This, in turn, excites a nonlinear interleaved envelope signal containing the waveform characteristics of the overlapping unloading response of the two channels within the main fluid transmission pipeline. This signal reshapes the reference overlap phase difference, and the pulse waveform sequence is then used to drive the time-division multiplexing actuator to start, enabling the first... Road distribution actuator and the first The working period of the route distribution actuator generates a preset reference overlap phase difference. A pressure interference signal is generated within the pressure feedback manifold. The pressure transformation unit on the manifold continuously acquires this signal at a sampling frequency of 1000Hz to generate a time-domain pressure data sequence containing continuous pressure sampling points. This time-domain pressure data sequence is input into a demodulation program, and the phase difference is compared with the reference. During the corresponding phase-shift overlap period, the sliding window autocorrelation operator is invoked to perform autocorrelation transformation and time-domain difference integral calculation on the interference envelope, separating the characterization of the first phase. Lubrication branch and the first Flow resistance state parameters of local fluid impedance characteristics of lubrication branch ; where subscript For lubrication branch channel index, This is the channel index for the next adjacent lubrication branch. Assign the phase difference value for the time-sharing activation of the actuators to two adjacent paths. The flow resistance scalar value is extracted through the integration of the autocorrelation function. Specifically, the process of obtaining the axis excitation signal from the axis motion data feedforward in the CNC interpolation buffer is as follows: The control unit reads the position target command sequence of each motion axis in the CNC system interpolation buffer in real time within the next 200ms. The control unit performs first-order and second-order time difference on the position target command sequence at a fixed time step of 5ms, thereby calculating the predicted velocity characteristic value and predicted acceleration characteristic value of each motion axis at each time node in the future. The control unit retrieves the preset friction pair lubrication consumption model, multiplies the predicted velocity characteristic value by the corresponding velocity consumption weight coefficient, and multiplies the predicted acceleration characteristic value by the corresponding acceleration consumption weight coefficient. The sum of the two values ​​is used to obtain the predicted lubrication consumption gradient of each motion axis in the future time sequence. Finally, the control unit performs multi-axis weighted superposition of the predicted lubrication consumption gradient of each motion axis and maps it to generate a discrete digital excitation sequence that changes continuously with time, which is the axis excitation signal used as the feedforward control reference.

[0057] The control unit runs a boundary comparison program to measure the flow resistance state parameters. A boundary comparison is performed with the preset reference impedance range, and the flow resistance state parameters are determined. Exceeding the upper limit of the reference impedance range When calculating the flow resistance state parameters Resistance deviation from the reference impedance range The smoothing filter module is used to process the flow resistance state parameters. The noise is processed and the drag deviation is calculated over five consecutive control cycles using a median filter module. The median, specifically, involves sorting the resistance deviations calculated over five consecutive control cycles in discrete ascending order, removing the maximum and minimum values, and extracting the arithmetic mean of the remaining three cycles as the stable median of the filtered output. This median is used to forcibly offset transient flow pressure fluctuations caused by continuous high-frequency reciprocating commutation excitation through discrete damping characteristics. The original fluctuation variable is replaced by the median of the filtered output. This median is used as the control variable for pulse width adjustment. The driving pulse width of the corresponding branch in the control parameter matrix is ​​gradually increased in a preset fixed increment. By increasing the driving pulse width, the single lubricating oil injection amount of the corresponding friction pair is increased. This causes the pressure wave phase interleaving and aliasing caused by long-distance fluid mass transfer delay and pipeline elastic expansion to undergo amplitude attenuation in the next sampling cycle, suppressing signal fluctuations caused by multi-axis high-frequency reciprocating motion, and thus improving the flow resistance state parameters. Converging to within the reference impedance range; where the subscript Indicates the maximum value. For the current cycle's flow resistance deviation scalar value, in a continuous high-frequency alternating cutting control cycle, the overlapping interference signal in the pressure feedback manifold is converted into a recognizable discrete-channel flow resistance data stream through parameter modulation during the phase shift overlap period. The drive pulse width of the multi-channel actuator varies with the flow resistance state parameters. The real-time feedback generates asymmetric step correction. The single-point pressure sensing channels at the ends of the friction pairs of the spindle bearing, guide rail slider, and lead screw are in a discrete closed-loop response state. The fluid pressure at the end of the system pipeline matches the lubricating oil consumption rate of the friction pairs stably. In actual operation, the predicted lubrication consumption gradient on which the aforementioned reconstructed control parameter matrix is ​​based refers to the theoretical lubricating oil consumption rate per unit time of each moving axis friction pair under a specific motion state. Its quantification standard is determined in the following way: The control unit retrieves the preset discrete empirical oil quantity mapping table stored in the non-volatile memory according to the current working condition label fed back by the CNC system. When the working condition label refers to the high-speed continuous cutting mode, the internal control bus reads the current axis motion speed characteristics of the spindle bearing, guide rail slider, and lead screw pair in real time. The absolute value of the operating speed is directly multiplied by a constant fluid physical wall loss coefficient of 0.005 mL / (m / s). Simultaneously, the follow-up acceleration characteristics of the current moving axis are read, and the absolute value of the acceleration is directly multiplied by a constant contact surface dynamic friction shear correction coefficient of 0.012 mL / (m / s2). The product of the velocity term and the product of the acceleration term are then accumulated in a digital register using a one-dimensional discrete scalar addition to filter out intermediate non-public black box components, thereby determining the lubrication consumption gradient of a single axis. Finally, the lubrication consumption gradients of all moving axes, including the spindle bearing, guide rail slider, and lead screw pair, within the machining center are linearly recombined according to spatial weights to obtain the predicted lubrication consumption gradient representing the global state evolution, providing a definite discrete data stream support for the adaptive correction of the control parameter matrix.

[0058] Example 2: This experiment was conducted on a multi-axis linkage high-frequency hydraulic flow simulation test platform to test the accuracy of detecting the local fluid impedance of discrete friction pairs. The total length of the main fluid transmission pipeline of the test platform was set to 12m, and it was equipped with five discrete branch lubrication reversing solenoid valves to simulate the distribution spacing of the friction pairs of the spindle bearing, guide rail slider, and lead screw in a multi-axis linkage machining center. A single high-frequency pressure sensing unit was connected to the end of the main pipeline. The high-frequency pressure sensing unit has a pressure measurement resolution of 0.01MPa and a full-range measurement accuracy of 0.1%, and its reference acquisition frequency is set to 1000Hz. The data acquisition process was accompanied by... A 50Hz continuous mechanical excitation interference is applied by an external vibrator, while a pressure pulsation component with an amplitude of 0.05MPa is injected into the internal circulation loop of the pipeline to counteract parasitic high-frequency background noise interference under high-frequency cutting conditions. The causal relationship for determining the reference acquisition frequency parameters of the high-frequency pressure sensing unit lies in the fact that the bandwidth of the transient shock wave in fluid transmission is the key parameter for determining the sampling frequency. Since the transmission of fluid pressure waves in the flexible pipe wall is accompanied by high-frequency attenuation and phase drift, the selection of its sampling frequency is constrained by the sound velocity of the pipeline fluid and the distribution of higher-order harmonics. Its causal adjustment lies in the distortion-free resolution of waveform details and the control unit. The balance between the data flow processing load of the discrete bus within the unit is addressed when the upper harmonic cutoff frequency of the high-frequency shock wave increases. To cope with spectral aliasing under the Nyquist sampling theorem constraint, the reference acquisition frequency parameter tends towards the upper limit of its range. Under the current forward reversal operation, the sampling frequency determined by the above logic is specifically fixed at 1000Hz. Further clarification of its physical mechanism lies in the fact that although long-distance flexible pipelines have a significant low-pass filtering attenuation effect on high-frequency transient shock waves, this invention actively constructs a phase shift overlap period with microsecond-level hardware delay by controlling two adjacent solenoid valves. This actively excited pressure interference physically produces a low-pass filter. The nonlinear interference envelope waveform of frequency modulation; the characteristic frequency of the fluid interlacing interference signal has been successfully shifted to the low-frequency envelope band, and its main characteristic energy distribution is in the frequency domain below 150Hz; therefore, the sampling frequency of 1000Hz not only fully meets the Nyquist sampling theorem requirement for distortion-free reconstruction of the low-frequency envelope signal, but also the resolution of acquiring 1 pressure sampling point every 1ms can accurately capture the amplitude evolution characteristics and phase slip trajectory of the slowly varying interference envelope, thereby achieving precise decoupling of the flow resistance state of the spatially distributed channel under the engineering premise of ensuring that the discrete bus data processing load inside the control unit is not overloaded.

[0059] The test samples were divided into the present invention sample group, the partially missing control group, and the out-of-range control group. During the continuous simulated cutting operation for 3600s, each test sample group shared the same original input physical boundary conditions, namely, the initial oil supply pressure reference value of the main pipeline was set to 2.0MPa, and the static impedance reference value of each branch solenoid valve was set to 1.2MPa. Under the drive of the multi-axis linkage interpolation command, the multi-channel distribution actuator started the time-division opening and closing action. At this time, in the acquisition of the original data of the set reference, due to the mass transfer delay of the long-distance main pipeline and the elastic dissipation of the pipe wall low-pass filter, the initial undemodulated total pressure waveform collected by the high-frequency pressure sensing unit at the end of the main pipeline showed an overlapping and mixed state. The peak-to-peak value of its original waveform was nonlinearly distorted and misaligned between 1.85MPa and 2.45MPa, and was mixed with high-frequency fluid shear self-excited oscillation, which made it impossible to directly divide the channel image of the waveform discrete sequence through the simple time window truncation method.

[0060] In the internal feature decoupling stage, some missing control groups, due to the removal of the phase modulation mechanism of the control pulse flow, have zero phase difference between the working periods of adjacent distributor actuators. The interference pressure envelope from the main pipeline feedback is in a disordered, coherent, and constructive state, causing the demodulation program to fail to obtain phase difference constraints. This results in intermodulation distortion in the intermediate flow resistance value output by the sliding window transformation, and the variance of the flow resistance calculation at each friction pair node increases to 0.34. Impedance identification between channels is in a deadlock state. In contrast, the present invention's sample group utilizes phase shift modulation of the control parameter matrix to generate a pulse waveform sequence that controls the time-division multiplexing of the distributor actuators, causing a preset reference overlap phase difference during the working periods of adjacent electro-hydraulic valves. During the demodulation process, the reference overlap phase difference During the corresponding phase-shift overlap period, the sliding window autocorrelation operator is invoked to perform autocorrelation transformation and time-domain difference integral calculation on the interference envelope, outputting an intermediate demodulated feature data stream, and decoupling and extracting the characterization of the first phase. Lubrication branch and the first Flow resistance state parameters of local fluid impedance characteristics of lubrication branch This transforms the discrete branch resistance information, originally superimposed within the total pressure signal, into discrete feature variables that can be directly used for boundary arbitration; where the subscript... For lubrication branch channel index, This is the channel index for the next adjacent lubrication branch. Assign the phase difference value for the time-sharing activation of the actuators to two adjacent paths. To extract the scalar value of flow resistance through autocorrelation function integration, the specific logic for phase modulation of the control shaft excitation signal and the control parameter matrix in actual operation is as follows: The control unit runs a hardware timer counting loop, and the control parameter matrix stores the reference opening phase angle and pulse width basic time steps corresponding to each discrete lubrication branch channel; during phase modulation, the control unit uses the shaft excitation signal that changes continuously with time generated by the aforementioned steps as a dynamic modulation factor, linearly maps the current channel's shaft excitation signal intensity value to a phase shift increment, and algebraically superimposes it with the corresponding reference opening phase angle in the control parameter matrix, thereby dynamically fine-tuning the opening start time of the solenoid valve drive signal of each channel; after this phase modulation processing, the opening control edges of two adjacent multi-channel actuators in the generated pulse waveform sequence are given a time phase difference based on adaptive sliding with shaft excitation, thereby ensuring that a nonlinear pressure interference signal containing definite phase characteristics and carrying branch impedance information can be excited in the fluid main pipeline.

[0061] In the final output and specific working window empirical phase, the data exhibited a nonlinear degradation trend and a saturation trend, when the baseline overlap phase difference of the control group exceeded the range. When the value is reduced to below the lower limit of the range, i.e., set to 2°, the trailing interlacing phase difference of the transient unloading pressure wave in the pipeline is lower than the fluid sound velocity propagation time constant, causing the characteristic overlapping interval of the demodulation operator to collapse, resulting in the flow resistance state parameter. The state determination error jumped to 45.3%, resulting in a state identification fault, and when the reference overlap phase difference... When the value is increased to above the upper limit of the range, i.e., set at 45°, the data shows that although the decoupling accuracy of the flow resistance parameter is within the stable range of 0.02MPa, the excessively long overlap time of the distributed actuators lengthens the total cycle of the distributed lubrication network, leading to periodic under-lubricated dry friction at the ends of the friction pairs. This causes a sudden increase in the surface temperature of local components of 18.4°C. This indicates that exceeding the operating range will cause technical problems such as deteriorated mass transfer efficiency and lubrication time discontinuity. In contrast, the flow resistance parameters of the present invention within the preset value range under continuous operating conditions... The identification accuracy remained at 98.6% when determining the flow resistance state parameters. Exceeding the upper limit of the preset reference impedance range At that time, the control unit calls the median filter module to calculate the resistance deviation over five consecutive control cycles. The median is used to correct the driving pulse width of the corresponding branch, ultimately guiding the friction coefficient of the friction pair surface to stably converge to 0.015, confirming the system-level stability of the method of this invention in maintaining multi-channel independent lubrication adaptive correction under underactuated constraints; where the subscript Indicates the maximum value. This represents the scalar value of the flow resistance deviation for the current cycle.

[0062] Example 3: When the distributed lubrication control system is applied to a multi-axis vertical machining center subjected to alternating transient thermal stress and strong mechanical vibration, the spindle bearing and high-speed rolling guide pair exhibit nonlinear fluid consumption characteristics at discrete movement speeds. The distributed lubrication network is connected to discrete reversing solenoid valves by a fluid transmission main pipeline. The total length of the fluid transmission main pipeline is set to 12m, and a single pressure conversion unit for single-point status feedback is provided at the end. When the fluid is transported inside the flow channel, friction mass transfer resistance and low-pass filtering elastic dissipation of the pipe wall are generated. This causes the residual reflected unloading wave generated by the closing of the previous stage reversing solenoid valve to lag into the fluid transmission main pipeline. It also causes waveform distortion and time overlap with the loading step wave excited by the opening of the next stage reversing solenoid valve in the local confluence topology cavity. As a result, an intertwined and superimposed pressure interference envelope is formed at the acquisition end of the pressure conversion unit. This causes the discrete total pressure data corresponding to the time-division isolated detection on the time axis to produce multi-channel characteristic adhesion. It is impossible to directly identify the transient flow resistance state of each friction pair discrete node. There is a hidden danger that local lubrication branches may cause abnormal blockage or under-lubrication, leading to component damage.

[0063] The control unit retrieves digital process control commands and reads the discrete pressure sequence from the pressure transformation unit. It inputs the time-domain discrete pressure feature array acquired by the pressure transformation unit into a buffer memory. A preset sliding window width is retrieved to truncate a discrete window slice of the discrete pressure sequence. The truncated current window discrete sequence is input into the autocorrelation analysis operator. Within a preset phase overlap period, the time-domain multiplication integral of the discrete pressure sequence and its own delay sequence is calculated to generate the autocorrelation function envelope curve. The peak time difference of the autocorrelation function envelope curve is input into the discrete-time differential integration module to filter out periodic excitation noise and fluid self-excited turbulence disturbances. The independent index is then calculated. Flow resistance state parameters of local fluid characteristics in road lubrication branches and the flow resistance state parameters The input state determination module uses the upper limit of the reference impedance stored in non-volatile memory. Continuous comparison is performed to determine the current channel's flow resistance status parameters. The impedance is greater than the upper limit of the reference impedance for three consecutive control cycles. When calculating the flow resistance state parameters Deviation from the upper limit of the reference impedance resistance deviation The corresponding lubrication branch's drive pulse width update value for the next working cycle is calculated based on the formula, which is expressed as follows: ,in, For the updated allocation actuator drive pulse width, The preset reference drive pulse width, For adaptive adjustment coefficients, This represents the resistance deviation measured in the current sample, where the subscript... For lubrication branch channel index, For the next adjacent lubrication branch channel index, the subscript is... This represents the maximum value, due to the adaptive adjustment coefficient. The monotonically increasing characteristic of the driving pulse width update value causes the driving pulse width update value to change with the drag deviation. As the value increases, the corresponding actuator extends its opening time to offset the pressure loss caused by the dissipation of elastic volume in the main pipeline. Specifically, the process of calculating the differential rate of change of the autocorrelation curve and performing time integration to separate the flow resistance state parameters is not a simple calculation of the difference between the two endpoints of the complete autocorrelation function, but rather a noise reduction and decoupling mechanism using characteristic window integration and extreme value mapping: within the time period corresponding to the reference overlapping phase difference, the demodulation program divides the generated autocorrelation function envelope curve into a first 1 / 2 upper swing window and a second 1 / 2 lower swing window along the time axis; then, in the first 1 / 2 upper swing window, the slope difference between adjacent discrete points is calculated to obtain the differential rate of change sequence, and time numerical summation integration is performed on the differential rate of change sequence within this window interval to extract the main characteristic integral scalar reflecting the residual energy release of the previous lubrication branch; similarly, in the second 1 / 2 lower swing window... The same difference and time integration operations are performed internally to extract the secondary characteristic integral scalar reflecting the step loading of the subsequent lubrication branch. Finally, the data processing unit introduces a preset dimension reduction mapping rule to perform normalized weighted product superposition of the primary characteristic integral scalar and the secondary characteristic integral scalar. Specifically, the data processing unit retrieves a preset two-dimensional fixed weight matrix in the non-volatile memory, directly multiplies the primary characteristic integral scalar by a first weight allocation coefficient of 0.65, and directly multiplies the secondary characteristic integral scalar by a second weight allocation coefficient of 0.35. The two product results are then summed using a baseline algebra. Subsequently, the summation result is subjected to a division physical dimension offset mapping using the system's pre-calibrated zero-load pipeline initial resting pressure mean value, thereby forcibly reducing the dimension and eliminating the interference of DC bias and transient fluctuations between the two endpoints. Finally, dimensionless flow resistance state parameters that can independently characterize the resistance characteristics of a specific branch are obtained.

[0064] As the drive pulse width in the actuator drive control register is incremented and rewritten, the single-injection flow rate of the lubricating agent at the end of the corresponding lubrication branch pipeline experiences a discrete step increase. The axial deformation volume loss of the lubrication hose in the next working cycle is compensated by physical flow. The phase lag deviation of the residual reflected unloading wave is reduced to below 5% within four consecutive control cycles. The pressure interference envelope at the single-point pressure sensing channel of the main pipeline recovers to the preset timing phase interval, and the flow resistance state parameters... Reconverged to the upper limit of the reference impedance Within the established safety range, the dynamic friction coefficient of the friction pair components remains stably maintained between 0.012 and 0.015 during high-frequency cutting cycles under varying operating conditions. The distributed control loop establishes independent closed-loop responses and adaptive adjustments for the multi-channel discrete node states under underactuated physical constraints. In actual control loop operation, to eliminate the dimensional mismatch between the multi-channel composite characteristics input by the aforementioned formula and the single-channel controlled object, the system introduces a differential state decoupling and corresponding pulse width mapping mechanism: the control unit continuously stores the inherent impedance reference values ​​recorded when each independent channel operates individually in the previous control cycle; when the flow resistance state parameters containing the overlapping characteristics of two adjacent lubrication branches are calculated during the phase shift overlap period, the control unit subtracts the inherent impedance reference value of the (i+1)th channel stored in the previous cycle from the total composite characteristics, thereby separating the i-th single channel due to coking in the current cycle. The independent resistance deviation caused by blockage; the control unit takes the decoupled single-channel resistance deviation as a determined independent variable, substitutes it into the aforementioned drive pulse width update formula for product step operation, and finally accurately rewrites the updated drive pulse width output to the single-channel control register of the i-th channel actuator, realizing the corresponding mapping from multi-dimensional composite detection to single-value independent closed-loop control. Under continuous high-frequency cutting conditions, the inner wall of long-distance lubrication pipelines gradually deposits due to coking of lubricating oil or the gradual adhesion of impurity particles, resulting in the pipeline flow cross-sectional area exhibiting a first-order continuously slow-decreasing fluid dynamic resistance evolution law on the time axis. The system provides a method for calculating insufficient oil supply warning by means of resistance evolution trend. The specific operation steps are as follows: In the multi-axis reciprocating motion of the machining center, the pressure acquisition unit acquires the main pipe pressure data at a reference sampling frequency of 1000Hz. The data demodulation module decouples and calculates the characteristic of the i-th channel actuator in each control cycle. Flow resistance state parameters of local fluid impedance characteristics of lubrication branch As a dynamic input for trend warning; the control unit continuously reads the flow resistance status parameters of the current control cycle and the previous 99 control cycles. Arranged chronologically into a discrete sequence of 100 elements. The data processing unit uses a backward first-order time-domain difference operator to calculate the resistance time variation sequence. The difference in flow resistance state parameters between adjacent control cycles is used to calculate the flow resistance rate parameter at the corresponding moment by dividing the difference by the control cycle time interval. The control unit will input the flow resistance rate parameter. Compared to the preset rate of change limit stored in non-volatile memory Comparison, when the flow resistance rate parameter Exceeding the preset rate of change limit for 5 consecutive control cycles When the system detects a risk of insufficient oil supply due to progressive blockage in the lubrication branch, the control unit writes a high-level signal to a specific pin of the underlying digital output register, outputting an insufficient oil supply warning signal to the external CNC system. Under normal, unobstructed flow conditions, the flow resistance rate parameter... The flow resistance rate parameter stabilized below 0.001 due to progressive coking in the pipeline. If the rate of change exceeds the preset upper limit of 0.020 for five consecutive control cycles, an insufficient fuel supply warning is triggered, and the above flow resistance status parameters are checked. For the first time in the current control cycle The scalar value of the local flow resistance of the lubrication branch is generated by time-domain integration of the sliding autocorrelation function to serve as the basic data characterizing the flow state of the pipeline, and the resistance time variation sequence. This is a discrete dataset consisting of flow resistance state parameters from the current control cycle and the previous 99 control cycles arranged chronologically. It is used to provide a time window for trend analysis and flow resistance rate parameters. The rate of change of the flow resistance scalar over time in the resistance time-varying sequence between adjacent control cycles is used to quantify the severity of the resistance increase, with a preset upper limit for the rate of change. The critical threshold for the allowable rate of change of pipeline progressive coking resistance is determined experimentally under non-blocking baseline conditions and serves as the criterion boundary for triggering the insufficient oil supply warning.

[0065] Example 4: In the commissioning scenario of the first run and calibration of a newly assembled CNC machine, random installation discrete errors arise from the physical length, bending radius, curvature of multiple flexible pipelines within the distributed lubrication network, and the initial zero-point offset of the pressure transformation unit. This causes a low-frequency shift in the static pressure baseline within the main fluid transmission pipeline, leading to phase decoupling waveform distortion in the time-domain pressure data sequence extracted from subsequent high-frequency cutting cycles. The control unit sends full-channel synchronous drive level pulses via the bus to drive the multi-channel actuators to operate in a time-sharing cycle for 20 cycles under zero mechanical load. The pressure transformation unit at the end of the main pipeline continuously records the discrete pressure response curve under pure pipeline flow circulation at a sampling frequency of 1000Hz. The captured waveform amplitude is input into a discrete integral filter to calculate the numerical mean to remove transient vibration components. The initial pressure drift vector representing the static pressure drop characteristics of the current installation topology is then calculated. ; Subscript Indicates the zero-point state. This is a static pressure characteristic scalar value.

[0066] When the system detects the initial pressure drift vector through the pressure transformation unit When the calculation deviation is less than 0.01 MPa for five consecutive cycles, it indicates that the fluid flow field inside the pipeline has reached a stable state. The control unit then runs the discrete zero-point subtraction algorithm and writes the initial pressure drift vector into the initial amplitude register of the corresponding channel in the control parameter matrix. , the initial pressure drift vector The parameters are directly injected as feedforward variables into the boundary comparison program under subsequent high-frequency cutting cycles to differentially offset and eliminate the pressure drop caused by the static expansion of the flexible pipeline and the temperature drift of the sensor in the time-domain pressure data sequence. This ensures that the DC bias component of the feature array of the demodulation program stably converges to zero when extracting the flow resistance characteristics of the time-domain pressure data sequence during the phase shift overlap period. The flow resistance state parameters of the discrete friction pair nodes within the distributed lubrication network are also considered. The identification accuracy was restored to the preset 98.6%, and the initial calibration control quantities of the distributed lubrication control system were rewritten; among them, the subscripts... For lubrication branch channel index, It is a characteristic quantity of the channel flow resistance state.

[0067] Example 5: In the commissioning of a high-speed, heavy-duty multi-axis linkage machining center equipped with a distributed lubrication network, the control unit establishes a pre-calibration procedure for the internal control parameter matrix and the preset reference impedance range. The spindle bearing, guide rail slider, and lead screw of the machining center are connected to the main pressure feedback manifold by separate reversing solenoid valves. A single high-frequency pressure sensor for single-point status feedback is installed at the end of the main pressure feedback manifold. This high-frequency pressure sensor has a pressure measurement resolution of 0.01 MPa. Under zero-load conditions, the control unit reads the status data in the CNC system interpolation buffer through the industrial fieldbus to confirm that the system is in a non-cutting resting state window, and reads the current main pipeline fluid pressure from the high-frequency pressure sensor as the static reference pressure. The static reference pressure is calibrated to 2.00 MPa. The timing and counting module inside the control unit sequentially opens each independent lubrication branch according to the timing code sequence in a time-sharing cyclic operation mode. Taking the first lubrication branch as an example, the control unit sends a single-pulse drive command to open the corresponding reversing solenoid valve, and the drive pulse width is within the reference time window. With a time limit of 100ms, the reversing solenoid valves of the remaining branches remain closed. At the high-frequency pressure sensor, the pressure acquisition unit continuously captures the transient pressure response waveform generated when fluid is replenished to the corresponding branch at a sampling frequency of 1000Hz, and sends a time-domain pressure data sequence containing continuous pressure sampling points to the demodulation module. The demodulation module retrieves a preset sliding window width, extracts discrete window slices from the time-domain pressure data sequence, and uses the time-domain sliding autocorrelation operator within a 100ms reference time window. Within the time domain, the integral of the time-domain multiplication of the time-domain pressure data sequence with its own delay sequence is calculated to generate the autocorrelation function envelope curve. The discrete time-domain differential integration module calculates the differential rate of change of the autocorrelation function envelope curve and performs a time integral on the differential rate of change to filter out periodic excitation noise and separate the dimensionless flow resistance state parameters characterizing the resistance properties of the lubrication branch. The corresponding lubrication branch operates continuously for 20 cycles under normal flow conditions, and the control unit calculates the flow resistance state parameters during these 20 cycles. The average value of the values ​​is used to obtain the initial flow resistance value of the lubrication branch, which is calculated to be 1.00.

[0068] The control unit determines the reference impedance range for the corresponding branch based on a preset allowable deviation of 0.20. The lower limit of this reference impedance range is determined by subtracting the preset allowable deviation of 0.20 from the initial flow resistance value of 1.00, resulting in a lower limit of 0.80. The upper limit of the reference impedance range is determined by adding the preset allowable deviation of 0.20 to the initial flow resistance value of 1.00, resulting in an upper limit of 1.20. This establishes the reference impedance range for the independent lubrication branch as 0.80 to 1.20, where 0.80 corresponds to the absolute lower limit, 1.00 corresponds to the normal median value, and 1.20 corresponds to the absolute upper limit. To verify the adaptability of the boundary check procedure under fault conditions, a flow-limiting and pressure-regulating valve is artificially connected to the first lubrication branch to simulate a coking and blockage state in the pipeline. The control unit restarts the time-sharing drive process, causing the demodulation module to process the high-frequency sampling sequence and calculate the current flow resistance parameters during the corresponding activation period. Increased to 1.35; the boundary comparison procedure compared the flow resistance state parameters. Determine the flow resistance state parameters based on the reference impedance range. If the value exceeds the upper limit of 1.20, the identification module determines that the lubrication branch has a coking blockage abnormality, and the control unit determines the flow resistance status parameter accordingly. The resistance deviation from the reference impedance range is used to rewrite the drive pulse width parameter of the lubrication branch in the control parameter matrix, widening the corresponding drive pulse width from the reference time window of 100ms to 150ms. This incremental rewriting of the drive pulse width parameter enables the lubrication branch to obtain a larger single lubricating oil injection volume in the next cycle, and uses the transient high pressure of fluid dynamics to apply in-situ flushing to the pipeline.

[0069] In the first lubrication branch, the bypass pressure relief valve is manually opened to simulate a pipeline rupture and leakage. The demodulation module calculates the current flow resistance parameters based on the time-domain pressure discrete characteristic sequence during the corresponding opening period. Reduced to 0.75; at this parameter, the boundary comparison procedure compares the flow resistance state parameter. Determine the flow resistance state parameters based on the reference impedance range. If the value is below the lower limit of 0.80, the identification module determines that the lubrication branch has a damaged or leaking abnormality. The control unit directly corrects the drive pulse width parameter of the lubrication branch in the control parameter matrix to 0ms to stop the oil supply to the lubrication branch, and sends a leakage alarm signal to the CNC system via the fieldbus. After the machining center enters the continuous high-frequency reversing cutting condition, the heat conduction of the machine tool spindle and cutting causes the ambient temperature at the sub-oil supply bus to systematically rise from 20℃ to 60℃, causing the viscosity of the lubricating oil to drop significantly and causing the global flow resistance parameter to drift towards a lower value. According to the falling edge slope of the standard response waveform... The specific steps for calculating the fluid viscosity variation coefficient and performing translation correction are as follows: The control unit captures the start and end points of the attenuation of the standard response waveform during the unloading phase, calculates the ratio of the pressure drop change to the time span during this period, and thus determines the actual falling edge slope under the current operating condition; the control unit retrieves the reference falling edge slope stored at the standard temperature of 20℃, divides the current actual falling edge slope by the reference falling edge slope, and uses this ratio as the dimensionless fluid viscosity variation coefficient; during boundary follow-up translation, the boundary comparison program sets the upper and lower limits of the originally preset reference impedance range. Multiplying by the fluid viscosity variation coefficient respectively, a product-based follow-up translation correction is achieved, where the judgment boundary scales proportionally with fluid viscosity degradation. This eliminates the false alarm problem of global flow resistance drift caused by temperature rise. When the CNC system state machine recognizes the G00 rapid positioning command and the resting period condition lasts longer than 2 seconds, the control unit drives the standard verification branch in the multi-channel actuator to open for 100ms to output a test pulse. The pressure acquisition unit captures the corresponding standard response waveform, and the control unit calculates the fluid viscosity variation coefficient based on the falling edge slope of the standard response waveform. The boundary comparison program utilizes... The fluid viscosity variation coefficient is applied to the judgment boundary in real time through product scaling and translation, dynamically shifting the original reference impedance range of 0.80 to 1.20 to the range of 0.50 to 0.75. The identification module sorts the flow resistance state parameters obtained from decoupling and reconstruction within three consecutive lubrication control cycles and extracts the median to generate a filter contribution parameter. When the filter contribution parameter exceeds the impedance threshold boundary after translation three times consecutively, the control unit rewrites the action parameter matrix and outputs a fault isolation command, so that the dynamic friction coefficient of the friction pair component surface is stably maintained between 0.012 and 0 in the high-frequency cutting cycle under varying working conditions.Within the 015 range, specifically, the standard calibration branch is a closed-loop calibration circuit independent of the conventional lubrication circuits connecting all friction pairs. One end is connected to the pressure feedback main pipe, and the other end is directly connected to the lubrication tank via a return oil line. To maintain its constant impedance under extreme thermal interference conditions where the external ambient temperature rises sharply from 20°C to 60°C, this standard calibration branch integrates a high-precision mechanical throttle valve. The throttle cone and valve body of this valve are made of Invar alloy, a material with an extremely low coefficient of thermal expansion, thus eliminating the expansion of the valve port's geometric cross-sectional area caused by thermal expansion and contraction. Simultaneously, because this branch directly returns oil to the tank, its piping is extremely short and made of rigid stainless steel, eliminating the dissipation effect caused by the radial elastic expansion of the flexible pipe wall. This ensures that its inherent local fluid impedance remains absolutely constant under temperature rise conditions, guaranteeing that the output test pulse is entirely affected only by the fluid's own physicochemical viscosity variations.

[0070] The above description is only a few preferred embodiments of the present invention and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present invention is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, technical solutions formed by replacing the above-mentioned features with the technical features with similar functions disclosed in the embodiments of the present invention.

Claims

1. A distributed lubrication control method with single-point state feedback, characterized in that, Includes the following steps: Step S101: Retrieve the preset control parameter matrix from the internal control unit and obtain the axis excitation signal obtained by feedforward deduction of axis motion data from the CNC interpolation buffer. Control the axis excitation signal and the control parameter matrix to perform phase modulation to generate a pulse waveform sequence that controls the time-division multiplexing actuator to start. Step S102: The pulse waveform sequence control timing and counting module is used to drive the multi-channel distribution actuator to start in a time-sharing manner, so that the working period of two adjacent multi-channel distribution actuators generates a preset reference overlapping phase difference, and a pressure interference signal is excited in the pressure feedback manifold of fluid transmission. Step S103: Using the pressure transformation unit on the pressure feedback manifold, pressure interference signals are acquired at a sampling frequency of 1000Hz to generate a time-domain pressure data sequence containing continuous pressure sampling points. Step S104: Input the time-domain pressure data sequence into the demodulation program, perform sliding window autocorrelation transformation and time-domain differential integration calculation within the time period corresponding to the reference overlapping phase difference, and separate the dimensionless flow resistance state parameters that characterize the resistance characteristics of each lubrication branch. Step S105: The flow resistance state parameter is compared with the preset reference impedance range through the boundary comparison program. When the flow resistance state parameter exceeds the reference impedance range, the driving pulse width of the corresponding branch in the control parameter matrix is ​​adaptively corrected.

2. The distributed lubrication control method with single-point state feedback according to claim 1, characterized in that, Before comparing the flow resistance state parameters with the reference impedance range in step S105, when the operating condition is detected to meet the preset rest period conditions, the standard verification branch in the multi-way distribution actuator is controlled to open for a preset time to output test pulses and acquire the corresponding response waveforms. The fluid viscosity variation coefficient is calculated based on the slope of the falling edge of the response waveform. The fluid viscosity variation coefficient is then used to shift and correct the reference impedance range in order to compensate for the flow resistance drift caused by changes in ambient temperature.

3. The distributed lubrication control method with single-point state feedback according to claim 1, characterized in that, In step S105, when the flow resistance state parameter exceeds the upper limit of the reference impedance range, the driving pulse width of the corresponding branch in the control parameter matrix is ​​gradually increased according to the resistance deviation of the flow resistance state parameter from the reference impedance range by a preset step increment.

4. The distributed lubrication control method with single-point state feedback according to claim 1, characterized in that, Step S104 includes the following sub-steps: Step S1041, within the time period corresponding to the reference overlapping phase difference, the time-domain pressure data sequence is extracted by a moving window, the autocorrelation function of the pressure interference signal within the moving window is calculated, and the autocorrelation change curve is obtained; Step S1042, the differential change rate of the autocorrelation change curve is calculated, and the differential change rate is integrated over time to separate the flow resistance state parameters.

5. The distributed lubrication control method with single-point state feedback according to claim 1, characterized in that, Step S103 includes the following sub-steps: Step S1031, the pressure conversion unit continuously acquires the fluid pressure in the pressure feedback manifold at a sampling frequency of 1000Hz and filters out noise in the preset frequency band to obtain continuous pressure data points; Step S1032, the continuous pressure data points are combined according to the acquisition time sequence to construct an interference waveform and generate a time-domain pressure data sequence.

6. The distributed lubrication control method with single-point state feedback according to claim 3, characterized in that, When the driving pulse width of the corresponding branch in the control parameter matrix is ​​gradually increased according to the preset step increment, the flow resistance state parameter is reduced by the smoothing filter module, and the median filter module is used to calculate the median of the resistance deviation in multiple consecutive control cycles to hedge against fluctuations.

7. The distributed lubrication control method with single-point state feedback according to claim 1, characterized in that, When determining the reference impedance range during the calibration phase, each independent lubrication branch is turned on in sequence. The initial flow resistance value of each independent lubrication branch under normal flow conditions is obtained using the pressure conversion unit. Based on the initial flow resistance value plus and minus the preset allowable deviation, the reference impedance range corresponding to each independent lubrication branch is determined.

8. The distributed lubrication control method with single-point state feedback according to claim 1, characterized in that, When using the resistance evolution trend to provide early warning of insufficient oil supply, the flow resistance state parameters of each independent lubrication branch are continuously recorded in each control cycle to generate a resistance change sequence over time. The rate of change of the change sequence over time is calculated to obtain the flow resistance rate parameter; when the flow resistance rate parameter of a certain independent lubrication branch continuously exceeds the preset upper limit of the rate of change, the corresponding insufficient oil supply warning signal is output.

9. The distributed lubrication control method with single-point state feedback according to claim 1, characterized in that, In step S105, when the current resistance state parameter is lower than the lower limit of the reference impedance range, the drive pulse width of the corresponding branch in the control parameter matrix is ​​corrected to zero, and a leakage alarm signal is output.

10. A distributed lubrication control system with single-point state feedback, used to implement the distributed lubrication control method with single-point state feedback as described in claim 1, characterized in that, The system includes: The signal modulation module is used to retrieve the preset control parameter matrix inside the self-control unit and obtain the axis excitation signal derived from the axis motion data feedforward of the CNC interpolation buffer. It controls the axis excitation signal to be phase-modulated with the control parameter matrix to generate a pulse waveform sequence for controlling the time-division control module to start, and transmits the pulse waveform sequence to the timing and counting module. The timing and counting module is used to receive pulse waveform sequences and use the pulse waveform sequences to drive the distribution control module to start in a time-division manner, so that the working periods of two adjacent distribution control modules generate a preset reference overlapping phase difference, so as to generate a pressure interference signal in the pressure feedback main pipe. The distribution control module, connected to the discrete lubrication pipeline, is used to activate in a time-sharing manner under the drive of the timing and counting module; The pressure acquisition and transformation module is used to acquire the pressure interference signal in the pressure feedback manifold at a sampling frequency of 1000Hz, generate a time-domain pressure data sequence containing continuous pressure sampling points, and output the time-domain pressure data sequence to the data demodulation module. The data demodulation module is used to receive time-domain pressure data sequences and, under the control of the demodulation program, perform sliding window autocorrelation transformation and time-domain differential integration calculation within the time period corresponding to the reference overlapping phase difference. It separates the dimensionless flow resistance state parameters that characterize the resistance characteristics of each lubrication branch and inputs the flow resistance state parameters to the boundary comparison correction module. The boundary comparison correction module is used to receive the flow resistance state parameters and, under the control of the boundary comparison process, compare the flow resistance state parameters with the preset reference impedance range. When the flow resistance state parameters exceed the reference impedance range, the module adaptively corrects the driving pulse width of the corresponding branch in the control parameter matrix and feeds back the correction result to the signal modulation module.

Images