A method for uniform distribution control of static pressure load on a heavy-duty workbench
By constructing an orthogonal decoupling matrix and correcting the oil supply flow rate based on oil temperature, the problem of liquid-solid coupling effect in the hydrostatic control system of the heavy-duty workbench was solved, achieving adaptive adjustment and fast response, and improving the stability and accuracy of the system.
Patent Information
- Application Number
- CN202611219975.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-12
- Publication Date
- 2026-11-17
AI Technical Summary
The existing hydrostatic control system for heavy-duty worktables is difficult to achieve adaptive adjustment of oil film stiffness and closed-loop correction of oil supply flow under eccentric heavy load and oil temperature variation conditions. This results in severe fluid-solid coupling between the hydrostatic chambers, leading to unstable control and slow response.
By synchronously acquiring the real-time pressure vector of the static pressure chamber and the dual-axis spatial attitude tilt angle signal, an orthogonal decoupling matrix is constructed using the recursive least squares algorithm and singular value decomposition. Combined with the oil temperature change to correct the oil supply flow, adaptive pressure proportional regulation and differential damping feedforward are realized to generate servo valve control signals to reduce the influence of liquid-solid coupling.
It effectively reduces the mutual influence between static pressure chambers, improves system stability and dynamic response capability, and ensures the attitude stability and control accuracy of the heavy-duty worktable when the load changes.
Smart Images

Figure CN123114086A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of adaptive control technology for industrial control systems, and in particular to a method for uniformly distributing static pressure load on a heavy-duty workbench. Background Technology
[0002] Multi-cavity hydrostatic guideways are typically used as load-bearing components of heavy-duty worktables. They rely on hydraulic oil films to keep the worktable suspended and running smoothly. To ensure that each hydrostatic cavity has sufficient load-bearing capacity, most existing control systems use independent single-loop pressure feedback control or a fixed throttling method to detect and adjust the oil supply pressure of each hydrostatic cavity separately. This type of control can meet the basic support requirements under static or light loads with relatively uniform load distribution. However, as the machining load gradually increases and eccentric loads and constantly changing cutting forces appear, the load borne by each hydrostatic cavity will become significantly uneven. Since the worktable body is rigid, the hydrostatic cavities also affect each other through the compression of the oil film. When the pressure or oil film gap of one hydrostatic cavity changes, the other hydrostatic cavities will also be affected. When a hydrostatic cavity is compressed due to eccentric load and the oil film gap decreases rapidly, traditional single-loop control only increases the oil supply pressure based on the pressure deviation of that hydrostatic cavity. The resulting change in support force will cause the worktable to tilt and further compress the oil film of the opposite hydrostatic cavity, easily causing continuous pressure fluctuations in multiple hydrostatic cavities.
[0003] To avoid instability caused by mutual interference between hydrostatic chambers, existing control systems typically reduce oscillations by lowering the proportional gain. However, a lower proportional gain slows down the system's response to load changes. When the worktable is subjected to cutting impacts, the oil supply adjustment may not keep up in time, increasing the risk of oil film rupture and direct contact with the guide rail surface. Furthermore, the hydraulic oil temperature gradually rises after prolonged operation, and the dynamic viscosity of the hydraulic oil changes. Traditional fixed-parameter control does not correct the oil supply command according to changes in dynamic viscosity; therefore, the same oil supply command may correspond to different actual oil supply volumes at different oil temperatures, affecting the accuracy of pressure regulation. Due to the limitations of traditional hardware structures and control modes, existing technologies for coupling hydrostatic systems... There are also significant shortcomings in the software control and calculation methods for the characteristics. For example, Chinese invention patent CN106777637B discloses a calculation method for the deformation of a hydrostatic turntable that considers the influence of fluid-structure interaction. This technology is based on the theory of circular thin plates and the finite iteration method to solve the elastic deformation and oil film reaction force of the turntable under static eccentric load. However, this method focuses on steady-state analysis in the structural design stage and relies on preset static boundary conditions for offline iterative calculation. It cannot reflect the dynamic coupling relationship between multi-cavity oil film extrusion and biaxial attitude tilt angle under cutting impact, and it cannot be directly applied to real-time dynamic control in industrial control systems. It is difficult to achieve real-time orthogonal decoupling of multi-cavity pressure and adaptive closed-loop compensation of oil film stiffness when heavy load conditions change abruptly.
[0004] Therefore, how to reduce the influence of fluid-solid coupling between the hydrostatic chambers and achieve adaptive adjustment of oil film stiffness and closed-loop correction of oil supply flow under eccentric heavy load and oil temperature change conditions are problems that need to be solved in the existing hydrostatic control of heavy-duty worktables. Summary of the Invention
[0005] To address the problems in the background art, the present invention provides the following technical solution: a method for uniformly distributing static pressure load on a heavy-duty workbench, applied to an industrial control system, comprising:
[0006] Step S1: Synchronously acquire the real-time pressure vector of each static pressure chamber of the heavy-duty workbench and the dual-axis spatial attitude tilt angle signal of the heavy-duty workbench.
[0007] Step S2: Based on the dual-axis spatial attitude tilt angle signal, the change in the oil chamber extrusion geometric clearance of each hydrostatic chamber is calculated. Combined with the real-time pressure vector, the recursive least squares algorithm is used to solve and update the dynamic liquid-solid coupling matrix between each hydrostatic chamber.
[0008] Step S3: Perform singular value decomposition on the dynamic liquid-solid coupling matrix to construct an orthogonal decoupling matrix, and reduce the dimension of the multi-cavity high-dimensional coupling control to an orthogonally distributed pressure equalization component space and attitude correction component space.
[0009] Step S4: Calculate the total load on each static pressure chamber based on the real-time pressure vector, and solve the real-time oil film stiffness by combining the change in the geometric gap of the oil chamber extrusion; In the pressure equalization component space and the attitude correction component space, when the real-time oil film stiffness is lower than the set stiffness reference value, adjust the adaptive pressure proportional coefficient by increasing the value, and superimpose the differential damping feedforward derived from the time change rate of the dual-axis space attitude tilt angle signal to calculate the target oil supply flow rate adjustment command for each static pressure chamber.
[0010] Step S5: Obtain the real-time oil temperature of the hydrostatic oil supply main pipeline, solve the current dynamic viscosity based on the exponential viscosity-temperature relationship, correct the target oil supply flow rate adjustment command using the ratio of the reference dynamic viscosity to the current dynamic viscosity, generate a servo valve control signal and send it to the proportional servo valve drive unit corresponding to each hydrostatic chamber.
[0011] Preferably, in step S1, the synchronous acquisition period of the real-time pressure vector and the dual-axis spatial attitude tilt angle signal is 1ms to 5ms; wherein the total number of each static pressure chamber is greater than or equal to 4, and each static pressure chamber is arranged in a symmetrical structure on the heavy-duty workbench base; step S2 includes the following sub-steps: step S21, obtaining the geometric distribution coordinates of each static pressure chamber in the workbench coordinate system to construct a geometric projection matrix; step S22, solving the basic oil film stiffness under the no-tilt condition based on the continuity equation and the Reynolds equation linearization, and calculating the change in the oil chamber extrusion geometric clearance of each static pressure chamber by combining the geometric projection matrix and the dual-axis spatial attitude tilt angle signal.
[0012] Preferably, the process of updating the dynamic liquid-solid coupling matrix online in step S2 further includes the following sub-steps: Step S23, establishing a local pressure variation differential relationship including the ratio coefficient of liquid elastic modulus to oil cavity volume, gap extrusion flow gain coefficient, and cross-cavity fluid damping coupling coefficient; Step S24, extracting pressure gradient and tilt angle variation data within a 20ms time window, using a recursive least squares algorithm to solve and update each damping coupling element in the dynamic liquid-solid coupling matrix in real time, and simultaneously deriving and solving the real-time eccentric load disturbance vector borne by each hydrostatic cavity based on inverse dynamics.
[0013] Preferably, in step S4, the adjustment logic for the increase of the adaptive pressure proportional coefficient includes: given that the total load and average oil thickness of each static pressure chamber are known, the ratio of the total load to the average oil thickness is determined as the real-time oil film stiffness; when the real-time oil film stiffness is lower than the set stiffness reference value, the difference between the set stiffness reference value and the real-time oil film stiffness is divided by the set stiffness reference value to determine the stiffness deviation rate; the stiffness deviation rate is multiplied by the stiffness compensation sensitive factor and then 1 is added as the gain amplitude adjustment coefficient; the basic proportional gain is multiplied by the gain amplitude adjustment coefficient to determine the adaptive pressure proportional coefficient; when the real-time oil film stiffness is greater than or equal to the set stiffness reference value, the adaptive pressure proportional coefficient is maintained at the basic proportional gain.
[0014] Preferably, in step S3, the process of constructing the orthogonal decoupling matrix and dimensionality reduction mapping includes: performing singular value decomposition on the dynamic liquid-solid coupling matrix to obtain the left singular matrix and the right singular matrix, and taking the transpose of the left singular matrix as the orthogonal decoupling matrix; using the orthogonal decoupling matrix to perform orthogonal linear transformation on the pressure deviation signals of each hydrostatic cavity intertwined coupling, mapping them into four mutually orthogonal technical components; wherein, the first orthogonal component corresponds to the average pressure fluctuation component of the entire cavity, the second orthogonal component corresponds to the attitude overturning moment deviation component around the X-axis, the third orthogonal component corresponds to the attitude overturning moment deviation component around the Y-axis, and the fourth orthogonal component corresponds to the higher-order torsional diagonal deviation component, thereby decomposing the high-dimensional strongly coupled differential equation into an orthogonal combination of independent cavity pressure deviation offsetting and global attitude moment compensation.
[0015] Preferably, in step S5, the process of correcting the target oil supply flow rate adjustment command includes: acquiring the real-time oil temperature measured by the temperature sensor in the hydrostatic oil supply main pipeline; based on the reference temperature and the reference dynamic viscosity at the reference temperature, using the exponential viscosity-temperature relationship to solve the actual dynamic viscosity at the current oil temperature; dividing the reference dynamic viscosity by the actual dynamic viscosity to determine the viscosity compensation coefficient; multiplying the target oil supply flow rate adjustment command of each hydrostatic chamber by the viscosity compensation coefficient to obtain the viscosity-corrected flow control command; converting the viscosity-corrected flow control command into a drive current signal and sending it to the high-frequency response proportional servo valve connected to the oil supply pipeline of each hydrostatic chamber to eliminate the interference of oil temperature rise on the output flow of the servo valve.
[0016] Preferably, a method for uniformly distributing static pressure load on a heavy-duty workbench further includes a dynamic feedforward compensation step for the dead zone of a proportional servo valve: monitoring the motion state of the proportional servo valve core based on the real-time pressure variation rate of each static pressure chamber; calculating a pulse bias signal based on the mechanical hysteresis characteristics of the valve core when the proportional servo valve core crosses the zero flow transition range; superimposing the pulse bias signal onto the servo valve control signal to generate a high-frequency pulse width modulation bias drive signal and outputting it to the proportional servo valve drive unit, thereby eliminating the mechanical response dead zone and time delay of the proportional servo valve during micro-flow regulation.
[0017] Preferably, a method for uniformly distributing static pressure load on a heavy-duty workbench further includes an emergency counter-pressure step for extremely low stiffness: when the real-time pressure of a certain static pressure chamber is lower than the set low-pressure limit threshold due to off-center load impact, and the calculated real-time oil film stiffness is lower than the set safety stiffness lower limit, the amplitude gain of the adaptive pressure proportional coefficient is increased to the set upper limit of the multiplier; at the same time, differential damping feedforward flow is injected according to the rate of change of spatial attitude tilt angle around the X-axis and around the Y-axis to forcibly lift the depressurized static pressure chamber and prevent the oil film from rupturing due to extreme extrusion and causing direct contact between the guide rail metal.
[0018] Preferably, in step S2, the process of solving the eccentric load disturbance vector includes: combining the pressure-bearing area and position coordinates of each hydrostatic chamber, using inverse dynamics to derive and solve the real-time eccentric load disturbance vector borne by each oil chamber under the action of eccentric heavy load and cutting force on the heavy-duty worktable; feeding the real-time eccentric load disturbance vector as a feedforward compensation amount to the calculation of the target oil supply flow rate adjustment command in step S4, so as to offset the interference of time-varying tooling load on the steady-state accuracy of the worktable attitude.
[0019] The beneficial effects of this invention are:
[0020] 1. In the uniform distribution control of static pressure load on heavy-duty worktables, it can reduce the mutual influence between each static pressure chamber, so that the multi-chamber pressure regulation no longer causes repeated oscillations due to the pressure change of a single static pressure chamber. By processing the mutually coupled pressure deviations separately, it can avoid mutual interference between control loops under eccentric loads, reduce the risk of high-frequency oscillation and control instability, and thus maintain system stability without significantly reducing the proportional gain. It can maintain a fast adjustment speed and improve the dynamic response capability of heavy-duty worktables when the load changes.
[0021] 2. This invention can adjust the pressure control intensity in a timely manner according to the change in oil film stiffness, and use the rate of change of the worktable posture tilt angle to suppress the tilting trend in advance. When eccentric heavy load or cutting impact causes oil film contraction and stiffness decrease, the control system can replenish the required flow more quickly, reduce the change in worktable posture and oil film thickness fluctuation, and avoid the oil film from continuing to thin and causing direct contact with the guide rail surface. This can improve the system's ability to resist sudden loads, enable each hydrostatic chamber to quickly restore uniform load, and maintain the stability of the worktable posture.
[0022] 3. This invention can correct the target oil supply flow rate adjustment command according to the dynamic viscosity change caused by the oil temperature in the oil supply pipeline, reduce the impact of oil temperature rise on the actual output flow of the proportional servo valve, so that the control command and the actual oil supply can maintain good consistency under different temperature conditions, avoid the deviation of oil supply flow rate and the distortion of static pressure chamber pressure regulation caused by changes in oil viscosity, thereby improving the control accuracy and stability of the heavy-duty workbench during long-term continuous operation. Attached Figure Description
[0023] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein:
[0024] Figure 1 This is a flowchart of the static load distribution and attitude correction process of the present invention;
[0025] Figure 2 This is a diagram of the multi-cavity collaborative decoupling control logic architecture of the present invention. Detailed Implementation
[0026] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.
[0027] A method for uniformly distributing static pressure load on a heavy-duty workbench includes:
[0028] Step S1: Synchronously acquire the real-time pressure vector of each static pressure chamber of the heavy-duty workbench and the dual-axis spatial attitude tilt angle signal of the heavy-duty workbench.
[0029] Step S2: Based on the dual-axis spatial attitude tilt angle signal, the change in the oil chamber extrusion geometric clearance of each hydrostatic chamber is calculated. Combined with the real-time pressure vector, the recursive least squares algorithm is used to solve and update the dynamic liquid-solid coupling matrix between each hydrostatic chamber.
[0030] Step S3: Perform singular value decomposition on the dynamic liquid-solid coupling matrix to construct an orthogonal decoupling matrix, and reduce the dimension of the multi-cavity high-dimensional coupling control to an orthogonally distributed pressure equalization component space and attitude correction component space.
[0031] Step S4: Calculate the total load on each static pressure chamber based on the real-time pressure vector, and solve the real-time oil film stiffness by combining the change in the geometric gap of the oil chamber extrusion; In the pressure equalization component space and the attitude correction component space, when the real-time oil film stiffness is lower than the set stiffness reference value, adjust the adaptive pressure proportional coefficient by increasing the value, and superimpose the differential damping feedforward derived from the time change rate of the dual-axis space attitude tilt angle signal to calculate the target oil supply flow rate adjustment command for each static pressure chamber.
[0032] Step S5: Obtain the real-time oil temperature of the hydrostatic oil supply main pipeline, solve the current dynamic viscosity based on the exponential viscosity-temperature relationship, correct the target oil supply flow rate adjustment command using the ratio of the reference dynamic viscosity to the current dynamic viscosity, generate a servo valve control signal and send it to the proportional servo valve drive unit corresponding to each hydrostatic chamber.
[0033] Preferably, in step S1, the synchronous acquisition period of the real-time pressure vector and the dual-axis spatial attitude tilt angle signal is 1ms to 5ms; wherein the total number of each static pressure chamber is greater than or equal to 4, and each static pressure chamber is arranged in a symmetrical structure on the heavy-duty workbench base; step S2 includes the following sub-steps: step S21, obtaining the geometric distribution coordinates of each static pressure chamber in the workbench coordinate system to construct a geometric projection matrix; step S22, solving the basic oil film stiffness under the no-tilt condition based on the continuity equation and the Reynolds equation linearization, and calculating the change in the oil chamber extrusion geometric clearance of each static pressure chamber by combining the geometric projection matrix and the dual-axis spatial attitude tilt angle signal.
[0034] Preferably, the process of updating the dynamic liquid-solid coupling matrix online in step S2 further includes the following sub-steps: Step S23, establishing a local pressure variation differential relationship including the ratio coefficient of liquid elastic modulus to oil cavity volume, gap extrusion flow gain coefficient, and cross-cavity fluid damping coupling coefficient; Step S24, extracting pressure gradient and tilt angle variation data within a 20ms time window, using a recursive least squares algorithm to solve and update each damping coupling element in the dynamic liquid-solid coupling matrix in real time, and simultaneously deriving and solving the real-time eccentric load disturbance vector borne by each hydrostatic cavity based on inverse dynamics.
[0035] Preferably, in step S4, the adjustment logic for the increase of the adaptive pressure proportional coefficient includes: given that the total load and average oil thickness of each static pressure chamber are known, the ratio of the total load to the average oil thickness is determined as the real-time oil film stiffness; when the real-time oil film stiffness is lower than the set stiffness reference value, the difference between the set stiffness reference value and the real-time oil film stiffness is divided by the set stiffness reference value to determine the stiffness deviation rate; the stiffness deviation rate is multiplied by the stiffness compensation sensitive factor and then 1 is added as the gain amplitude adjustment coefficient; the basic proportional gain is multiplied by the gain amplitude adjustment coefficient to determine the adaptive pressure proportional coefficient; when the real-time oil film stiffness is greater than or equal to the set stiffness reference value, the adaptive pressure proportional coefficient is maintained at the basic proportional gain.
[0036] Preferably, in step S3, the process of constructing the orthogonal decoupling matrix and dimensionality reduction mapping includes: performing singular value decomposition on the dynamic liquid-solid coupling matrix to obtain the left singular matrix and the right singular matrix, and taking the transpose of the left singular matrix as the orthogonal decoupling matrix; using the orthogonal decoupling matrix to perform orthogonal linear transformation on the pressure deviation signals of each hydrostatic cavity intertwined coupling, mapping them into four mutually orthogonal technical components; wherein, the first orthogonal component corresponds to the average pressure fluctuation component of the entire cavity, the second orthogonal component corresponds to the attitude overturning moment deviation component around the X-axis, the third orthogonal component corresponds to the attitude overturning moment deviation component around the Y-axis, and the fourth orthogonal component corresponds to the higher-order torsional diagonal deviation component, thereby decomposing the high-dimensional strongly coupled differential equation into an orthogonal combination of independent cavity pressure deviation offsetting and global attitude moment compensation.
[0037] Preferably, in step S5, the process of correcting the target oil supply flow rate adjustment command includes: acquiring the real-time oil temperature measured by the temperature sensor in the hydrostatic oil supply main pipeline; based on the reference temperature and the reference dynamic viscosity at the reference temperature, using the exponential viscosity-temperature relationship to solve the actual dynamic viscosity at the current oil temperature; dividing the reference dynamic viscosity by the actual dynamic viscosity to determine the viscosity compensation coefficient; multiplying the target oil supply flow rate adjustment command of each hydrostatic chamber by the viscosity compensation coefficient to obtain the viscosity-corrected flow control command; converting the viscosity-corrected flow control command into a drive current signal and sending it to the high-frequency response proportional servo valve connected to the oil supply pipeline of each hydrostatic chamber to eliminate the interference of oil temperature rise on the output flow of the servo valve.
[0038] Preferably, a method for uniformly distributing static pressure load on a heavy-duty workbench further includes a dynamic feedforward compensation step for the dead zone of a proportional servo valve: monitoring the motion state of the proportional servo valve core based on the real-time pressure variation rate of each static pressure chamber; calculating a pulse bias signal based on the mechanical hysteresis characteristics of the valve core when the proportional servo valve core crosses the zero flow transition range; superimposing the pulse bias signal onto the servo valve control signal to generate a high-frequency pulse width modulation bias drive signal and outputting it to the proportional servo valve drive unit, thereby eliminating the mechanical response dead zone and time delay of the proportional servo valve during micro-flow regulation.
[0039] Preferably, a method for uniformly distributing static pressure load on a heavy-duty workbench further includes an emergency counter-pressure step for extremely low stiffness: when the real-time pressure of a certain static pressure chamber is lower than the set low-pressure limit threshold due to off-center load impact, and the calculated real-time oil film stiffness is lower than the set safety stiffness lower limit, the amplitude gain of the adaptive pressure proportional coefficient is increased to the set upper limit of the multiplier; at the same time, differential damping feedforward flow is injected according to the rate of change of spatial attitude tilt angle around the X-axis and around the Y-axis to forcibly lift the depressurized static pressure chamber and prevent the oil film from rupturing due to extreme extrusion and causing direct contact between the guide rail metal.
[0040] Preferably, in step S2, the process of solving the eccentric load disturbance vector includes: combining the pressure-bearing area and position coordinates of each hydrostatic chamber, using inverse dynamics to derive and solve the real-time eccentric load disturbance vector borne by each oil chamber under the action of eccentric heavy load and cutting force on the heavy-duty worktable; feeding the real-time eccentric load disturbance vector as a feedforward compensation amount to the calculation of the target oil supply flow rate adjustment command in step S4, so as to offset the interference of time-varying tooling load on the steady-state accuracy of the worktable attitude.
[0041] Example 1: The rotary table of a current heavy-duty CNC machine tool has four rectangular hydrostatic chambers arranged symmetrically in a square. The first, second, third, and fourth hydrostatic chambers are sequentially set on the table base with a distribution radius of 0.8m. The effective pressure-bearing area of a single hydrostatic chamber is 0.12m², and the rated design oil film thickness is... The real-time pressure of the first hydrostatic chamber is 0.04 mm when the worktable is subjected to the combined action of eccentric load from a large casting and high-speed cutting force. The real-time pressure in the second static pressure chamber rises to 7.2 MPa. The real-time pressure of the third static pressure chamber is 6.8 MPa. The real-time pressure of the fourth static pressure chamber dropped to 3.2 MPa. When the pressure drops to 2.8 MPa, the worktable simultaneously tilts along two axes, including the axis around the worktable. Attitude tilt angle of the axis It is 0.00015 rad, around Attitude tilt angle of the axis It is 0.00008 rad, around Rate of change of the axis's attitude tilt angle It is 0.002 rad / s, around Rate of change of the axis's attitude tilt angle It is 0.001 rad / s, subscript The current sampling time is indicated. Subscripts 1 to 4 correspond to the first to fourth static pressure chambers, respectively. The eccentric load causes the oil film in the high-pressure static pressure chamber to shrink and the oil film in the low-load static pressure chamber to thicken and reduce the stiffness of the oil film. The static pressure chambers are also coupled with the extruded oil film through the worktable structure. When using single-chamber independent pressure feedback control, the pressure adjustment of one static pressure chamber will change the attitude of the worktable and further change the oil film thickness of other static pressure chambers, which can easily cause continuous pressure fluctuations in multiple chambers.
[0042] The industrial control system acquires the real-time pressure of the first to fourth static pressure chambers with a synchronization acquisition period of 2ms, and constructs a real-time pressure vector according to the chamber number. Simultaneously acquire dual-axis spatial attitude tilt angle signals. Real-time oil temperature of the hydrostatic oil supply main pipeline superscript Represents the transpose of a matrix or vector, with subscripts. This represents the oil in the hydrostatic supply main pipeline. The central processing unit arranges the position coordinates of each hydrostatic chamber in the worktable coordinate system into a geometric projection matrix according to the order of the first to fourth hydrostatic chambers. Each row of this matrix corresponds to one hydrostatic chamber, and the two columns represent the position coordinates of that hydrostatic chamber around the worktable. Axis and winding The clearance projection coefficients under axis tilt are determined by the positive direction of the worktable coordinate system. Under small tilt conditions, the geometric projection matrix is multiplied by the dual-axis spatial attitude tilt angle signal to obtain the change in the oil cavity extrusion geometric clearance of each hydrostatic chamber. This change is the local discrete deviation value of the physical oil film thickness of each hydrostatic chamber caused by the tilt of the worktable. The central processing unit subtracts the change in the oil cavity extrusion geometric clearance of each hydrostatic chamber from the rated design oil film thickness of 0.04mm to obtain the current actual oil film thickness of each hydrostatic chamber. Then, the arithmetic mean of the current actual oil film thickness of all hydrostatic chambers is calculated to derive the average oil thickness of 0.00004m, which is the sole basis for subsequent stiffness calculations. This eliminates the calculation gap between the discrete geometric change and the global average oil thickness. Among them, the change in the oil cavity extrusion geometric clearance of the first hydrostatic chamber... The change in the geometric clearance of the oil chamber in the second static pressure chamber is 0.000184m. for Change in the geometric clearance of the oil chamber in the third static pressure chamber for Change in the geometric clearance of the oil chamber in the fourth static pressure chamber It is 0.000056m.
[0043] The central processing unit extracts the pressure-time gradient and dual-axis attitude tilt angle variation data of each hydrostatic chamber within a 20ms time window. The attitude tilt angle variation data is then converted into the corresponding change in the hydraulic chamber extrusion geometric clearance using a geometric projection matrix. The pressure-time gradient is obtained by dividing the real-time pressure difference between adjacent sampling times by the sampling period. Subsequently, using the pressure-time gradient of each hydrostatic chamber as the observation output, and the oil supply flow command, the change in hydraulic chamber extrusion geometric clearance, and the pressure difference between adjacent hydrostatic chambers as the regression input, a recursive least squares algorithm is used to solve and update the 4×4 dynamic liquid-solid coupling matrix online. The first dynamic liquid-solid coupling matrix The row corresponds to the first The pressure change in the static pressure chamber, the first Column corresponding to the first The influence of each static pressure chamber on the pressure change; the diagonal elements of the matrix represent the pressure response of each static pressure chamber itself, and the off-diagonal elements represent the damping coupling elements between different static pressure chambers.
[0044] The central processing unit continuously records pressure and tilt angle data from 10 sampling points within a 20ms time window. Based on fluid continuity and oil compressibility, the pressure variation rate of each static pressure chamber is directly decomposed into three physically measurable engineering flow components: the oil injection flow rate from the proportional servo valve, the oil film extrusion flow rate from the compressed oil chamber, and the cross-chamber overflow flow rate leaking through the sealing gaps of adjacent static pressure chambers. The recursive least squares algorithm does not need to introduce any undefined intermediate abstract coefficients. It directly uses the proportional servo valve drive current, the change in the geometric gap of the compressed oil chamber, and the pressure difference between adjacent oil chambers as physical input vectors. It solves and directly updates the values of 16 specific numerical elements in the 4th order square matrix in real time. The four diagonal elements represent the rate of self-excited pressure change of each static pressure chamber under valve injection and oil film extrusion, while the twelve off-diagonal elements represent the rate of mutual excitation pressure influence of the sealing gaps of adjacent oil chambers under pressure difference. The recursive least squares algorithm updates the coefficients according to the sampling order, thereby obtaining the dynamic liquid-solid coupling matrix at the current sampling time. .
[0045] The central processing unit also performs inverse dynamics calculations on the eccentric loads and cutting forces acting on the worktable, taking into account the bearing area and position coordinates of each static pressure chamber. During the calculations, the vertical resultant force on the worktable and the surrounding forces are considered. Axis, winding The moment balance relationship of the shaft is used as a constraint to obtain the load disturbances borne by the first to fourth hydrostatic chambers respectively, and to form a real-time eccentric load disturbance vector according to the chamber number order. Each component of the vector corresponds one-to-one with each static pressure chamber. When calculating the target oil supply flow rate adjustment command, the central processing unit converts each component into the flow feedforward compensation amount of the corresponding static pressure chamber based on the pressure-flow response relationship in the differential relationship of local pressure change, so as to offset the chamber pressure deviation caused by eccentric load and cutting force changes.
[0046] For the dynamically updated fluid-solid coupling matrix The central processing unit performs singular value decomposition: ,in, It is a left singular matrix. It is a singular value matrix. The central processing unit takes the left singular matrix, which is the right singular matrix. The transpose of the matrix forms the orthogonal decoupling matrix: To ensure that each orthogonal component corresponds to the actual force direction on the worktable, the central processing unit arranges the rows of the orthogonal decoupling matrix according to the projection relationship between each singular vector and the four pressure distribution directions, and unifies the positive direction of each row. The four pressure distribution directions are, in order, the average pressure direction of the entire cavity changing in the same direction in the four static pressure chambers, the direction around the pressure distribution direction, and the direction of the pressure distribution direction. The two sides of the shaft that generate the overturning moment change direction in opposite directions, around The opposing directions of the overturning moment generated by the shaft and the opposing directions of the torsion of the two sets of diagonal hydrostatic chambers, along with the difference between the real-time pressure and the corresponding target pressure in each hydrostatic chamber, are used to form a pressure deviation signal according to the chamber number, which is then processed by an orthogonal decoupling matrix. After orthogonal linear transformation, the data is mapped to the pressure equalization component space and the attitude correction component space. The pressure equalization component space is defined as a one-dimensional control subspace composed of the first orthogonal component, specifically carrying the arithmetic mean fluctuation data of the pressure changes in the four static pressure chambers of the entire cavity in the same direction, and is bound to the baseline adjustment of the overall vertical suspension height of the worktable. The attitude correction component space is defined as a three-dimensional control subspace composed of the second, third, and fourth orthogonal components, carrying the torque deviation around the X-axis, torque deviation around the Y-axis, and diagonal torsional deviation data respectively, and is bound to the attitude correction of the worktable's dual-axis tilt angle and angular deformation. The first orthogonal component is the average pressure fluctuation component of the entire cavity, and the second orthogonal component is the pressure equalization component... The overturning moment deviation component of the shaft attitude, the third orthogonal component is around The shaft attitude overturning moment deviation component, the fourth orthogonal component is a high-order torsional diagonal deviation component. This fourth orthogonal component corresponds to the statically indeterminate torsional angle deformation data between the diagonal static pressure chambers of the worktable. The central processing unit multiplies the calculated value of the fourth orthogonal component by the diagonal adjustment gain coefficient to derive the forward flow compensation offset between the first and third static pressure chambers, and the reverse flow compensation offset between the second and fourth static pressure chambers. These are then superimposed on the drive current register of the proportional servo valve of the corresponding static pressure chamber. The diagonal torsional deformation is eliminated by injecting differential flow into the diagonal static pressure chambers to drive the worktable. The closed-loop output shows the physical equilibrium state after eliminating the diagonal torsion. Thus, the four orthogonal components are used for pressure regulation and attitude regulation respectively. The damping coupling effect between different static pressure chambers is corrected by the off-diagonal elements of the dynamic liquid-solid coupling matrix.
[0047] The central processing unit calculates the instantaneous average pressure across all four static pressure chambers based on their real-time pressure. subscript This represents the average value across the entire cavity under this operating condition. The pressure is 5 MPa. Considering the effective pressure-bearing area of 0.12 m² for a single static pressure chamber, the total load-bearing capacity of the entire chamber is calculated as follows: N, the average oil film thickness is calculated based on the current oil film thickness of each hydrostatic chamber, and its value is: m, and the ratio of the total load to the average oil film thickness is determined as the real-time oil film stiffness. ,get for N / m, this value represents the equivalent relationship between the total load and the average oil film thickness at the current operating point. The influence of fluid compressibility and flow resistance on the dynamic pressure response is determined by the dynamic fluid-solid coupling matrix. The corresponding elements are characterized and participate in subsequent decoupling and flow regulation without changing the real-time oil film stiffness. The above determination method.
[0048] When real-time oil film stiffness Greater than or equal to the set stiffness reference value ,Right now for When N / m, the adaptive pressure proportional coefficient Maintain basic proportional gain In this embodiment for m³ / (s·Pa), subscript This indicates proportional pressure adjustment; the subscript 0 represents the basic setting value. Representing the reference value, when calculating the target oil supply flow rate adjustment command for each static pressure chamber, it is first determined by the orthogonal decoupling matrix. The pressure deviation signal is transformed into four orthogonal components, and proportional adjustments are made in the pressure equalization component space and the attitude correction component space respectively; for the surrounding... Axis and winding The attitude correction component of the axis is then superimposed with the corresponding attitude tilt angle change rate and differential damping gain. The calculated differential damping feedforward quantity, subscript This indicates differential damping adjustment, and the resulting orthogonal component control quantity is... The data is inversely mapped to the first to fourth static pressure chambers and superimposed with the real-time eccentric load disturbance vector. The calculated flow feedforward compensation amount is used to form the target oil supply flow rate adjustment command for each static pressure chamber.
[0049] In the calculation of the first static pressure cavity, the two projection coefficients in the first row of the geometric projection matrix are denoted as follows: and ,in m, m, based on the positive mapping result and taking into account the flow feedforward compensation amount corresponding to the real-time eccentric load disturbance vector, the pressure proportional adjustment flow component of the first static pressure chamber is: m³ / s, differential damping gain for m³ / (s·rad / s), combined with the biaxial attitude tilt rate of change, the differential damping feedforward of the first hydrostatic chamber is obtained as follows: m³ / s, the sum of the two, the target oil supply flow rate adjustment command for the first static pressure chamber for m³ / s, where This indicates the amount of oil supply flow rate adjustment.
[0050] At this time, the real-time oil temperature of the hydrostatic oil supply main pipeline is 48℃, based on the reference dynamic viscosity at 40℃. Using Pa·s as a reference, the current dynamic viscosity at 48℃ was calculated according to the exponential viscosity-temperature relationship. Pa·s, where the subscript 0 indicates the value at the reference temperature. The central processing unit will use the ratio of the reference dynamic viscosity to the current dynamic viscosity as the viscosity compensation coefficient, and will then adjust the target oil supply flow rate accordingly. Multiplying by this viscosity compensation factor yields the viscosity-corrected flow control command. for m³ / s, or 0.07488 L / min, where the subscript This represents the output after viscosity correction. The central processing unit, according to the flow-current calibration relationship of the proportional servo valve, sends the flow control command... It is converted into driving current, generates servo valve control signal, and sends it to the proportional servo valve drive unit corresponding to the first static pressure chamber.
[0051] When the off-center load impact reduces the real-time pressure in the third static chamber to 0.5 MPa, and the real-time oil film stiffness... Down to When N / m Below the set stiffness reference value The adaptive pressure proportional coefficient is adjusted by increasing according to the following formula: ,in, As the stiffness compensation sensitivity factor, this embodiment takes 0.8, which is substituted into the basic proportional gain. Setting stiffness reference values and real-time oil film stiffness Then, adaptive pressure proportional coefficient Increase to With a gain of m³ / (s·Pa), the proportional gain is increased by 40% compared to the basic proportional gain. The central processing unit recalculates the control quantities of each orthogonal component according to the increased adaptive pressure proportional coefficient, superimposes the differential damping feedforward quantity and the flow feedforward compensation quantity corresponding to the real-time eccentric load disturbance vector, and then obtains the flow control command for each static pressure chamber through orthogonal decoupling matrix inverse mapping and dynamic viscosity correction. Each flow control command is converted into a servo valve control signal and sent to the corresponding proportional servo valve drive unit.
[0052] The proportional servo valve drive unit adjusts the valve opening according to the servo valve control signal. When the first static pressure chamber is subjected to high pressure and the table posture changes rapidly, the proportional servo valve corresponding to the first static pressure chamber increases the compensation flow, which slows down the change in oil film thickness in the high pressure area, reduces the risk of excessive oil film contraction and direct metal-to-metal contact with the guide rail. The dynamic liquid-solid coupling matrix and orthogonal decoupling matrix distribute the damping coupling effect between each static pressure chamber to the corresponding orthogonal components, reducing high-frequency pressure oscillations in multi-chamber closed-loop regulation. The flow control command is corrected according to the ratio of the reference dynamic viscosity to the current dynamic viscosity, so that the proportional servo valve still supplies oil at the target flow rate when the oil temperature changes, thereby maintaining the load distribution of each static pressure chamber and the dual-axis posture of the table.
[0053] Example 2: The current heavy-duty CNC machine tool rotary table test bench has four rectangular hydrostatic chambers arranged symmetrically in a square. The distribution radius of each hydrostatic chamber on the table base is 0.8m, the effective pressure-bearing area of a single hydrostatic chamber is 0.12m², and the rated design oil film thickness is... The thickness is 0.04 mm. The test bench uses a pressure sensor with a measurement range of 0.0 MPa to 20.0 MPa, a resolution of 0.1 kPa, and a sampling period of 2 ms. It also uses a dual-axis tilt sensor with an angular resolution of 0.00001 rad, an oil temperature sensor with a temperature measurement accuracy of 0.1℃, and a proportional servo valve with a frequency response bandwidth of 150 Hz. To simulate electromagnetic interference and mechanical vibration during heavy-duty cutting, Gaussian white noise with a signal-to-noise ratio of 25 dB and 50 Hz power frequency interference harmonics are superimposed in the acquisition channel. The sampling period is determined according to the upper cutoff frequency of the measured signal and the calculation load of the processing unit. During the test, the bias load is gradually increased from 50 kN to 500 kN. The bias load causes the oil film in the high-pressure static pressure chamber to shrink and the oil film in the low-load static pressure chamber to thicken. At the same time, cross-cavity fluid damping coupling is generated between each static pressure chamber.
[0054] The control system acquires the raw pressure signals of the first to fourth static pressure chambers according to a synchronous acquisition cycle of 2ms, forming a real-time pressure vector. Simultaneously acquire dual-axis spatial attitude tilt angle signals. Real-time oil temperature of the hydrostatic oil supply main pipeline Based on the collected data, the 4×4 dynamic liquid-solid coupling matrix is updated online using a recursive least squares algorithm. and real-time eccentric load disturbance vector Then, the dynamic liquid-solid coupling matrix Perform singular value decomposition, and construct an orthogonal decoupling matrix using the transpose of the left singular matrix. The pressure deviation signals of each hydrostatic chamber are mapped to the pressure equalization component space and the attitude correction component space. The control system calculates the real-time oil film stiffness based on the real-time pressure vector. When the real-time oil film stiffness Below the set stiffness reference value ,Right now When N / m, the adaptive pressure proportional coefficient is increased according to the decrease in stiffness. The target oil supply flow rate adjustment command is corrected based on the ratio of the reference dynamic viscosity to the current dynamic viscosity.
[0055] Four parallel sample groups were set up for the experiment. Control group 1 used single-chamber independent pressure feedback control; control group 2 used a fixed decoupling matrix without real-time adaptive adjustment of oil film stiffness and dynamic viscosity correction; control group 3 used a dynamic liquid-solid coupling matrix and real-time adaptive adjustment of oil film stiffness, but without dynamic viscosity correction; the sample group of this invention simultaneously used online updating of the dynamic liquid-solid coupling matrix, orthogonal decoupling matrix mapping, real-time adaptive adjustment of oil film stiffness and dynamic viscosity correction. The minimum oil film thickness and attitude tilt angle of the four sample groups were recorded under each working condition. At the same time, the real-time oil film stiffness, adaptive pressure proportional coefficient and adjustment convergence time of the sample group of this invention were also recorded.
[0056] Under operating condition 1, the offset load is 50 kN and the real-time oil temperature is 40℃. Under this condition, the minimum oil film thickness of control group 1 is 35.8 μm, and the attitude tilt angle is 0.00003 rad; the minimum oil film thickness of control group 2 is 37.1 μm, and the attitude tilt angle is 0.00002 rad; the minimum oil film thickness of control group 3 is 38.1 μm, and the attitude tilt angle is 0.00001 rad. The real-time oil film stiffness calculated for the sample group of this invention... for N / m, higher than the set stiffness reference value Adaptive pressure proportional coefficient Keep as m³ / (s·Pa); After completing orthogonal decoupling and dynamic viscosity correction, the minimum oil film thickness is 38.2 μm, the attitude tilt angle is 0.00001 rad, and the convergence time is adjusted to 8.2 ms.
[0057] Under operating condition 2, the offset load was increased to 120 kN, and the real-time oil temperature was raised to 42°C. The minimum oil film thickness of control group 1 decreased to 30.2 μm, and the attitude tilt angle increased to 0.00007 rad; the minimum oil film thickness of control group 2 was 33.4 μm, and the attitude tilt angle was 0.00005 rad; the minimum oil film thickness of control group 3 was 35.2 μm, and the attitude tilt angle was 0.00003 rad. The real-time oil film stiffness calculated for the sample group of this invention... for N / m, adaptive pressure proportional coefficient Still With a flow rate of m³ / (s·Pa), the minimum oil film thickness is 35.6 μm after flow rate adjustment, the attitude tilt angle is 0.00002 rad, and the convergence time is 9.1 ms.
[0058] Under operating condition 3, the offset load was 200 kN, and the real-time oil temperature was 45°C. The minimum oil film thickness of control group 1 was 22.4 μm, and the attitude tilt angle was 0.00014 rad; the minimum oil film thickness of control group 2 was 27.8 μm, and the attitude tilt angle was 0.00009 rad; the minimum oil film thickness of control group 3 was 31.0 μm, and the attitude tilt angle was 0.00005 rad. The real-time oil film stiffness of the sample group of this invention... for N / m, which is not lower than the set stiffness reference value. Adaptive pressure proportional coefficient Keep as m³ / (s·Pa); After adjustment, the minimum oil film thickness was 32.1 μm, the attitude tilt angle was 0.00003 rad, the adjustment convergence time was 10.3 ms, and the four sample groups maintained oil film support under three bias loads of 50 kN, 120 kN and 200 kN.
[0059] Under operating condition 4, the offset load was increased to 280 kN and the real-time oil temperature was increased to 48°C. The minimum oil film thickness of control group 1 decreased to 14.1 μm, with an attitude tilt angle of 0.00022 rad; the minimum oil film thickness of control group 2 was 21.3 μm, with an attitude tilt angle of 0.00014 rad; and the minimum oil film thickness of control group 3 was 26.1 μm, with an attitude tilt angle of 0.00008 rad. The real-time oil film stiffness calculated for the sample group of this invention... for N / m, lower than the set stiffness reference value The control system will then adapt the pressure proportional coefficient accordingly. Increase to After orthogonal decoupling, adaptive stiffness adjustment, and dynamic viscosity correction, the minimum oil film thickness is 29.5 μm, the attitude tilt angle is 0.00004 rad, and the adjustment convergence time is 11.2 ms.
[0060] Under operating condition 5, the offset load is 350 kN, the real-time oil temperature is 50℃, the minimum oil film thickness of control group 1 is 8.2 μm, and the attitude tilt angle is 0.00032 rad; the minimum oil film thickness of control group 2 is 14.5 μm, and the attitude tilt angle is 0.00019 rad; the minimum oil film thickness of control group 3 is 20.8 μm, and the attitude tilt angle is 0.00011 rad. The real-time oil film stiffness of the sample group of this invention... Down to N / m, the control system will adapt to the pressure proportional coefficient Increase to The minimum oil film thickness obtained was 27.2 μm, the attitude tilt angle was 0.00005 rad, and the convergence time was adjusted to 12.1 ms.
[0061] Under operating condition 6, the offset load is 420 kN, the real-time oil temperature is 52℃, the minimum oil film thickness of control group 1 is 3.1 μm, and the attitude tilt angle is 0.00045 rad; the minimum oil film thickness of control group 2 is 9.2 μm, and the attitude tilt angle is 0.00026 rad; the minimum oil film thickness of control group 3 is 17.1 μm, and the attitude tilt angle is 0.00015 rad. The real-time oil film stiffness calculated for the sample group of this invention... for N / m, adaptive pressure proportional coefficient Increased accordingly With a target oil supply flow rate of m³ / (s·Pa), the minimum oil film thickness is 24.8 μm, the attitude tilt angle is 0.00006 rad, and the adjustment convergence time is 12.8 ms.
[0062] Under operating condition 7, the offset load was further increased to 480 kN, and the real-time oil temperature was maintained at 52°C. The minimum oil film thickness of control group one decreased to 0.8 μm, and the attitude tilt angle increased to 0.00058 rad; the minimum oil film thickness of control group two was 6.1 μm, and the attitude tilt angle was 0.00033 rad; the minimum oil film thickness of control group three was 15.2 μm, and the attitude tilt angle was 0.00018 rad. The real-time oil film stiffness of the sample group of this invention... Down to N / m, the control system will adapt to the pressure proportional coefficient Increase to The adjusted minimum oil film thickness is 22.3 μm, the attitude tilt angle is 0.00007 rad, and the convergence time is 13.5 ms.
[0063] Under operating condition 8, an offset load of 500 kN was applied, and the real-time oil temperature was maintained at 52°C. In control group one, the minimum oil film thickness dropped to 0.0 μm, and the tilt angle was 0.00065 rad. The oil film lost support and triggered instability protection. Control group two used a fixed decoupling matrix, with a minimum oil film thickness of 4.8 μm and a tilt angle of 0.00038 rad. Control group three did not undergo dynamic viscosity correction; its actual oil supply flow rate deviated from the command by 25.0%, with a minimum oil film thickness of 14.1 μm and a tilt angle of 0.00021 rad. The real-time oil film stiffness calculated for the sample group of this invention... for N / m, the control system will adapt to the pressure proportional coefficient Increase to The amplitude of the gain is m³ / (s·Pa), which is 55.8% relative to the basic proportional gain. After dynamic viscosity correction, the minimum oil film thickness is 21.1 μm, which is 7.0 μm higher than the control group. The attitude tilt angle is 0.00008 rad, and the adjustment convergence time is 14.2 ms.
[0064] The continuous changes under various operating conditions show that the real-time oil film stiffness... Higher than the set stiffness reference value At that time, adaptive pressure proportional coefficient Maintaining the baseline value; starting from the 280kN operating condition, if the real-time oil film stiffness is lower than the set stiffness reference value, the control system gradually increases the adaptive pressure proportional coefficient as the stiffness decreases, and the dynamic fluid-solid coupling matrix... and orthogonal decoupling matrix The dynamic viscosity correction is used to correct the pressure coupling between the static pressure chambers, and to compensate for the change in actual oil supply flow caused by the rise in oil temperature. The above data were obtained under the condition of superimposed Gaussian white noise and power frequency interference harmonics, reflecting the role of each control link in the process of increasing bias load and real-time oil temperature rise.
[0065] Example 3: Real-time oil temperature of the hydrostatic oil supply main pipeline when the rotary table of a heavy-duty CNC machine tool is continuously performing heavy cutting. The current dynamic viscosity is raised to 58°C. The pressure drops to 0.018 Pa·s. The worktable simultaneously withstands a sudden off-center load impact of 520 kN and cutting torque fluctuations. The real-time pressure of the first hydrostatic chamber... The real-time pressure of the third static pressure chamber rises to 7.8 MPa. When the pressure drops to 0.4 MPa, the oil film thickness in the third hydrostatic chamber decreases to 0.012 mm. The high-frequency vibration noise caused by cutting impact is superimposed on the sensor signal. The recursive least squares algorithm updates the dynamic fluid-solid coupling matrix within a fixed 20 ms sliding window. At this time, data saturation and convergence lag are likely to occur.
[0066] The control system acquires the real-time pressure of the first to fourth static pressure chambers with a 2ms synchronous acquisition period and constructs a real-time pressure vector. Simultaneously acquire dual-axis spatial attitude tilt angle signals Within a 20ms sliding window, the central processing unit extracts the pressure-time gradient and biaxial attitude tilt angle variation data for each hydrostatic chamber. It converts the attitude tilt angle variation data into the corresponding change in the oil chamber extrusion geometric clearance and calculates the variance of the pressure variation in each hydrostatic chamber. When this variance exceeds 0.25 MPa², the forgetting factor of the recursive least squares algorithm is reduced from 0.98 to 0.85, increasing the weight of the pressure-time gradient and the change in the oil chamber extrusion geometric clearance within the current sliding window, and updating the 4×4 dynamic liquid-solid coupling matrix online. and real-time eccentric load disturbance vector .
[0067] After completing the matrix update, the central processing unit processes the dynamic liquid-solid coupling matrix. Perform singular value decomposition, and take the transpose of the left singular matrix to form an orthogonal decoupling matrix. And according to the average pressure of the whole cavity, the surrounding Axis attitude overturning moment, around The directional orthogonal components of the shaft attitude overturning moment and higher-order torsional diagonal deviation are arranged, and the pressure deviation signals of each hydrostatic chamber are decoupled by an orthogonal decoupling matrix. After transformation, the components are adjusted in the equal pressure component space and the attitude correction component space to reduce the cross-cavity fluid damping coupling interference between the hydrostatic cavities during strong cutting impact.
[0068] During the proportional servo valve switching process, the central processing unit adjusts the flow rate based on the real-time pressure change rate of the third static pressure chamber and the target oil supply flow rate. Monitor the valve core movement status; before switching, the absolute value of the pressure change rate in the third static pressure chamber is greater than 15.0 MPa / s; when the target oil supply flow rate adjustment command is issued... When a switch occurs where the positive and negative signs cross zero, the control system begins to determine whether the valve core has entered the zero flow transition range. During system initialization, the static dead zone current range and oil chamber pressure response lag time of the proportional servo valve under different valve opening degrees are measured in advance, and the correspondence between flow command reversal, valve core mechanical overlap and pressure response is saved.
[0069] When the valve core enters the overlapping area, the proportional servo valve temporarily stops outputting flow, and the pressure change rate of the third static pressure chamber decreases accordingly. The central processing unit performs a logical AND operation on the gradient of the real-time pressure change rate and the reversing signal of the target oil supply flow adjustment command. If the absolute value of the pressure change rate drops sharply from above 15.0 MPa / s to below the change rate threshold and lasts for 2 ms, it is determined that the valve core has entered the zero flow transition range, provided that the target oil supply flow adjustment command has crossed the zero point.
[0070] The central processing unit calculates the pulse bias signal based on the pre-calibrated valve core mechanical hysteresis bias, using the reference dynamic viscosity at 40°C. Pa·s and current dynamic viscosity at 58°C Pa·s corrects the target oil supply flow rate adjustment command of the third static pressure chamber, resulting in a viscosity-corrected flow control command. The pulse bias signal is first recorded as the equivalent pulse bias flow rate according to the flow rate calibration relationship of the proportional servo valve. Its pulse bias amplitude is taken as 8.5% of the rated maximum flow rate, and the pulse duration is taken as 3ms; then, according to the flow-current calibration relationship of the proportional servo valve, the equivalent pulse bias flow rate is... This is converted into a pulse bias current and superimposed on the viscosity-corrected flow control command. The converted servo valve control signal generates a high-frequency pulse width modulation bias drive signal, which is sent to the proportional servo valve drive unit corresponding to the third static pressure chamber.
[0071] Real-time pressure of the third static pressure chamber The calculated real-time oil film stiffness when the pressure drops to 0.4 MPa for N / m, lower than the set stiffness reference value Adaptive pressure proportional coefficient According to the decrease in stiffness, the basic proportional gain is applied. Adjusting m³ / (s·Pa) upwards, stiffness compensation sensitive factor Set the stiffness reference value to 0.8. for N / m, in this embodiment, the low-pressure limit threshold is set to 0.5MPa, and the lower limit of the safety stiffness is set to N / m; When the real-time pressure of the third static pressure chamber is lower than 0.5MPa, and the real-time oil film stiffness Further down to below When N / m, the adaptive pressure proportional coefficient will be used. The amplitude modulation gain is limited to the set upper limit, and the adaptive pressure proportional coefficient is used. The maximum value is the base proportional gain 1.60 times, that is m³ / (s·Pa).
[0072] Within the pressure equalization component space and the attitude correction component space, the central processing unit operates according to the adjusted adaptive pressure scaling factor. Calculate the pressure regulation amount and wrap it around Axis attitude tilt rate of change and around Axis attitude tilt rate of change Multiply by the differential damping gain respectively The differential damping feedforward quantity is obtained by taking m³ / (s·rad / s) and then passing it through an orthogonal decoupling matrix. After inverse mapping, the pressure regulation, differential damping feedforward, and real-time eccentric load disturbance vector are... The corresponding flow feedforward compensation is superimposed to form the target oil supply flow rate adjustment command for the third static pressure chamber. After dynamic viscosity correction and current conversion, it is output to the proportional servo valve drive unit.
[0073] After receiving the high-frequency pulse width modulation bias drive signal, the proportional servo valve drive unit drives the valve core past the zero flow transition range and replenishes oil to the third static pressure chamber under the conditions of 58℃ oil temperature and 520kN off-center load impact. The oil film thickness in the third static pressure chamber recovers from 0.012mm to 20.5μm within 15.8ms, and the biaxial spatial attitude tilt angle of the worktable converges to within 0.00007rad; orthogonal decoupling matrix Synchronously reduce the pressure oscillations caused by cross-cavity fluid damping coupling, so that each static pressure cavity can resume bearing capacity and maintain the stability of the worktable posture.
[0074] Example 4: This example combines Figures 1 to 2 This paper describes a method for uniformly distributing static pressure load on a heavy-duty workbench, such as... Figure 1As shown, step S1 is executed by synchronously acquiring the real-time pressure vectors of each hydrostatic chamber of the heavy-duty worktable and the dual-axis spatial attitude tilt angle signal of the heavy-duty worktable. Then, based on the dual-axis spatial attitude tilt angle signal, the change in the oil chamber extrusion geometric clearance of each hydrostatic chamber is solved. Combined with the real-time pressure vector, the dynamic liquid-solid coupling matrix between each hydrostatic chamber is solved and updated using a recursive least squares algorithm, executing step S2. Subsequently, singular value decomposition is performed on the dynamic liquid-solid coupling matrix to construct an orthogonal decoupling matrix, reducing the dimensionality of the multi-chamber high-dimensional coupling control to an orthogonally distributed uniform pressure component space and attitude correction component space. Step S3 is executed, then the total load on each static pressure chamber is calculated based on the real-time pressure vector, and the real-time oil film stiffness is solved by combining the change in the geometric clearance of the oil chamber extrusion. When the stiffness is lower than the reference value, the adaptive pressure proportional coefficient is increased and the differential damping feedforward is superimposed. The target oil supply flow rate adjustment command is calculated and step S4 is executed. Finally, the real-time oil temperature is obtained and the current dynamic viscosity is solved based on the exponential viscosity-temperature relationship. The target oil supply flow rate adjustment command is corrected by using the ratio of the reference to the current dynamic viscosity. The servo valve control signal is generated and sent to the corresponding proportional servo valve drive unit to execute step S5.
[0075] like Figure 2 As shown, in the static pressure load uniform distribution control method of heavy-duty workbench, real-time pressure vector and dual-axis spatial attitude tilt angle signal are obtained by synchronously collecting real-time parameters. Based on this, the change in the geometric gap of oil chamber extrusion is calculated. Combined with the real-time eccentric load disturbance vector and the cross-cavity fluid damping coupling coefficient, the dynamic liquid-solid coupling matrix is solved and updated using the recursive least squares algorithm. At the same time, the orthogonal decoupling matrix is constructed by using singular value decomposition, and the control dimension is reduced to the pressure uniform component space and the attitude correction component space. Combined with the real-time oil film stiffness control adaptive pressure proportional coefficient, the target oil supply flow rate adjustment command is corrected based on the exponential viscosity-temperature relationship.
[0076] Example 5: Before the heavy-duty CNC machine tool rotary table control system is put into operation, the central processing unit performs geometric position and fluid property baseline calibration. During calibration, the table is in a stationary state with no off-center load and no cutting. The oil supply pressure of the hydrostatic oil supply main pipeline is adjusted to 5MPa to keep the table suspended. The initial pressure of the first to fourth hydrostatic chambers is collected simultaneously. Record the distribution radius of each static pressure chamber on the worktable base. The rated design oil film thickness is 0.8m. The static zero-position deflection angle measured by the dual-axis tilt sensor in this state is 0.04 mm. The central processing unit uses this as a reference offset from the subsequently acquired dual-axis spatial attitude tilt angle signal. The zero-position deviation caused by sensor installation is subtracted to correct the zero-position deviation.
[0077] After completing the attitude baseline calibration, the real-time oil temperature of the hydrostatic oil supply main pipeline will be monitored. The temperature was gradually increased from 30°C to 60°C, and the current dynamic viscosity was measured at different temperatures. And fit it according to the exponential viscosity-temperature relationship: Among them, reference temperature At 40℃, the reference dynamic viscosity is... The viscosity-temperature sensitivity coefficient of the oil is 0.032 Pa·s. The reference temperature is 0.028℃⁻¹. Reference dynamic viscosity and oil viscosity-temperature sensitivity coefficient Saved to memory, it is used during operation to calculate the current dynamic viscosity based on the real-time oil temperature, and the target oil supply flow rate adjustment command is corrected using the ratio of the reference dynamic viscosity to the current dynamic viscosity.
[0078] During the calibration process of the proportional servo valve in the zero-flow transition range, the worktable is kept unloaded and suspended. The central processing unit controls the drive current to gradually increase from 0mA until the proportional servo valve spool produces a detectable flow response. The forward dead zone current during the forward movement of the spool is recorded. and the reverse dead zone current during reverse motion Based on the dead zone current in both directions and the time when the oil chamber pressure begins to respond, the amplitude of the bias compensation current corresponding to the pulse bias signal when the valve core reverses is determined. During operation, when the target oil supply flow regulation command switches between positive and negative signs across zero, and the valve core enters the zero flow transition range based on the real-time pressure change rate, the central processing unit generates a high-frequency pulse width modulation bias drive signal according to the calibration result.
[0079] To determine the pressure fluctuation threshold used in updating the dynamic fluid-solid coupling matrix using the recursive least squares algorithm, the control system continuously collected 100 sets of static pressure chamber pressure data with a sliding window of 20ms under the unbiased rotation state of the worktable. Based on each set of pressure data, the unbiased base variance caused by Gaussian background noise and mechanical vibration was calculated. The value is 0.05 MPa². Five times the base variance is determined as the pressure fluctuation threshold, i.e., 0.25 MPa². When the pressure variation variance within the sliding window is not greater than 0.25 MPa², the forgetting factor of the recursive least squares algorithm remains at 0.98. When the pressure variation variance exceeds 0.25 MPa² due to off-center cutting impact, the forgetting factor is reduced to 0.85, and the weights of the pressure time gradient and tilt angle variation data within the current sliding window are increased to update the damping coupling elements in the dynamic fluid-solid coupling matrix. This completes the calibration of the control system's attitude zero position, fluid properties, zero flow transition range of the proportional servo valve, and pressure fluctuation threshold.
[0080] Example 6: When the rotary table is undergoing factory calibration and sensor installation deviation correction, the four-chamber hydrostatic system remains in a non-eccentric load state. The loading device applies a step-incremental load of 100kN to 300kN at an eccentric radius of 0.8m from the center of the table. The central processing unit synchronously collects the real-time pressure of the first to fourth hydrostatic chambers to form a real-time pressure vector. And acquire dual-axis spatial attitude tilt angle signals. A geometric projection matrix is constructed based on the position coordinates of each static pressure chamber in the worktable coordinate system. Then, linear regression is performed using the measured attitude tilt angle and the change in the geometric clearance of the oil chamber in each static pressure chamber to correct the projection coefficients in the matrix. After correction, the first static pressure chamber along... shaft and The projection factors in the axial directions are respectively m and m.
[0081] Under non-tilting conditions, the effective pressure-bearing area of a single static pressure chamber The rated design oil film thickness is 0.12 m². The thickness is 0.04 mm. The central processing unit establishes the relationship between pressure, flow rate, and oil film thickness under this operating condition based on the continuity equation and the Reynolds equation, and performs a linearized solution at the rated operating point to obtain the basic oil film stiffness. for N / m, setting the stiffness reference value Take the base oil film stiffness 93.75% of N / m is used to determine whether the real-time oil film stiffness has entered the range where the adaptive pressure proportional coefficient needs to be increased.
[0082] To determine the stiffness compensation sensitivity factor The value range of is within the real-time oil film stiffness. for Numerical gradient tests were conducted under eccentric loading and low stiffness conditions with N / m, and the stiffness compensation sensitivity factor was considered. When the value is 0.4, the corresponding adaptive pressure proportional coefficient is... for With a static pressure of m³ / (s·Pa), the oil film rise rate in the low-load static pressure chamber is relatively slow, and the minimum oil film thickness drops to 13.5 μm; when the stiffness compensation sensitive factor When the value is 1.4, the corresponding adaptive pressure proportional coefficient is... for When the output flow of the proportional servo valve is m³ / (s·Pa), overshoot occurs, and the static pressure chamber pressure generates a low-frequency, low-amplitude oscillation at a frequency of 18.5Hz.
[0083] Stiffness compensation sensitive factor When the value is set between 0.7 and 0.9, the static pressure chamber pressure can converge without overshoot; among which, When the value is 0.8, the adaptive pressure proportional coefficient for With a damping ratio of m³ / (s·Pa) for dynamic cavity pressure response of 0.707 and a minimum oil film thickness maintained above 21.5 μm, this embodiment uses a stiffness compensation sensitive factor. The value range is determined to be 0.7 to 0.9, with 0.8 being the preferred value.
[0084] The differential damping gain is determined based on the ratio of the table rotational inertia to the hydrostatic oil film extrusion damping coefficient, and is used to compensate for the inertial overshoot caused by the tilting motion of the table. The oil viscosity-temperature sensitivity coefficient is obtained by exponential fitting of the measured hydraulic oil dynamic viscosity data in the range of 40℃ to 60℃, with a fitting correlation coefficient of over 0.99. The pressure fluctuation threshold is obtained by continuously collecting 100 sets of hydrostatic chamber pressure data under no-load static conditions, calculating the Gaussian background noise variance based on each set of data, and using 5 times this variance as the threshold for triggering the recursive least squares algorithm forgetting factor adjustment.
[0085] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention 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 the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the protection scope of the present invention.
Claims
1. A method for uniformly distributing static pressure load on a heavy-duty workbench, characterized in that, include: Step S1: Synchronously acquire the real-time pressure vector of each static pressure chamber of the heavy-duty workbench and the dual-axis spatial attitude tilt angle signal of the heavy-duty workbench. Step S2: Based on the dual-axis spatial attitude tilt angle signal, the change in the oil chamber extrusion geometric clearance of each hydrostatic chamber is calculated. Combined with the real-time pressure vector, the recursive least squares algorithm is used to solve and update the dynamic liquid-solid coupling matrix between each hydrostatic chamber. Step S3: Perform singular value decomposition on the dynamic liquid-solid coupling matrix to construct an orthogonal decoupling matrix, and reduce the dimension of the multi-cavity high-dimensional coupling control to an orthogonally distributed pressure equalization component space and attitude correction component space. Step S4: Calculate the total load on each static pressure chamber based on the real-time pressure vector, and solve the real-time oil film stiffness by combining the change in the geometric clearance of the oil chamber. Within the pressure equalization component space and attitude correction component space, when the real-time oil film stiffness is lower than the set stiffness reference value, the adaptive pressure proportional coefficient is increased and adjusted, and the differential damping feedforward derived from the time change rate of the dual-axis space attitude tilt angle signal is superimposed to calculate the target oil supply flow rate adjustment command for each static pressure chamber. Step S5: Obtain the real-time oil temperature of the hydrostatic oil supply main pipeline, solve the current dynamic viscosity based on the exponential viscosity-temperature relationship, correct the target oil supply flow rate adjustment command using the ratio of the reference dynamic viscosity to the current dynamic viscosity, generate a servo valve control signal and send it to the proportional servo valve drive unit corresponding to each hydrostatic chamber.
2. The method for uniformly distributing static pressure load on a heavy-duty workbench according to claim 1, characterized in that, In step S1, the synchronous acquisition period of the real-time pressure vector and the dual-axis spatial attitude tilt angle signal is 1ms to 5ms; the total number of each static pressure cavity is greater than or equal to 4, and each static pressure cavity is arranged in a symmetrical structure on the heavy-duty workbench base; step S2 includes the following sub-steps: step S21, obtain the geometric distribution coordinates of each static pressure cavity in the workbench coordinate system to construct a geometric projection matrix; Step S22: Based on the continuity equation and the linearization of the Reynolds equation, the basic oil film stiffness under the no-tilt condition is solved, and the change in the oil chamber extrusion geometric clearance of each hydrostatic chamber is calculated by combining the geometric projection matrix and the biaxial spatial attitude tilt angle signal.
3. The method for uniformly distributing static pressure load on a heavy-duty workbench according to claim 1, characterized in that, The process of updating the dynamic liquid-solid coupling matrix online in step S2 further includes the following sub-steps: Step S23, establishing the differential relationship of local pressure variation including the ratio coefficient of liquid elastic modulus and oil cavity volume, the gap extrusion flow gain coefficient and the cross-cavity fluid damping coupling coefficient; Step S24, extracting the pressure gradient and tilt angle variation data within a 20ms time window, using the recursive least squares algorithm to solve and update each damping coupling element in the dynamic liquid-solid coupling matrix in real time, and simultaneously deriving and solving the real-time eccentric load disturbance vector borne by each hydrostatic cavity based on inverse dynamics.
4. The method for uniformly distributing static pressure load on a heavy-duty workbench according to claim 1, characterized in that, In step S4, the adjustment logic for the increase of the adaptive pressure proportional coefficient includes: given that the total load and average oil thickness of each static pressure chamber are known, the ratio of the total load to the average oil thickness is determined as the real-time oil film stiffness; when the real-time oil film stiffness is lower than the set stiffness reference value, the difference between the set stiffness reference value and the real-time oil film stiffness is divided by the set stiffness reference value to determine the stiffness deviation rate; the stiffness deviation rate is multiplied by the stiffness compensation sensitive factor and then 1 is added as the gain amplitude adjustment coefficient; the basic proportional gain is then multiplied by the gain amplitude adjustment coefficient to determine the adaptive pressure proportional coefficient; when the real-time oil film stiffness is greater than or equal to the set stiffness reference value, the adaptive pressure proportional coefficient is maintained at the basic proportional gain.
5. The method for uniformly distributing static pressure load on a heavy-duty workbench according to claim 1, characterized in that, In step S3, the process of constructing the orthogonal decoupling matrix and dimensionality reduction mapping includes: performing singular value decomposition on the dynamic liquid-solid coupling matrix to obtain the left singular matrix and the right singular matrix, and taking the transpose of the left singular matrix as the orthogonal decoupling matrix; using the orthogonal decoupling matrix to perform orthogonal linear transformation on the pressure deviation signals of each hydrostatic cavity intertwined coupling, mapping them into four mutually orthogonal technical components; wherein, the first orthogonal component corresponds to the average pressure fluctuation component of the whole cavity, the second orthogonal component corresponds to the attitude overturning moment deviation component around the X-axis, the third orthogonal component corresponds to the attitude overturning moment deviation component around the Y-axis, and the fourth orthogonal component corresponds to the higher-order torsional diagonal deviation component, thereby decomposing the high-dimensional strongly coupled differential equation into an orthogonal combination of independent cavity pressure deviation offsetting and global attitude moment compensation.
6. The method for uniformly distributing static pressure load on a heavy-duty workbench according to claim 1, characterized in that, In step S5, the process of correcting the target oil supply flow rate adjustment command includes: acquiring the real-time oil temperature measured by the temperature sensor in the hydrostatic oil supply main pipeline; using the exponential viscosity-temperature relationship based on the reference temperature and the reference dynamic viscosity at the reference temperature to solve the actual dynamic viscosity at the current oil temperature; dividing the reference dynamic viscosity by the actual dynamic viscosity to determine the viscosity compensation coefficient; multiplying the target oil supply flow rate adjustment command of each hydrostatic chamber by the viscosity compensation coefficient to obtain the viscosity-corrected flow control command; converting the viscosity-corrected flow control command into a drive current signal and sending it to the high-frequency response proportional servo valve connected to the oil supply pipeline of each hydrostatic chamber to eliminate the interference of oil temperature rise on the output flow of the servo valve.
7. The method for uniformly distributing static pressure load on a heavy-duty workbench according to claim 1, characterized in that, The method for uniformly distributing static pressure load on a heavy-duty workbench also includes a dynamic feedforward compensation step for the dead zone of a proportional servo valve: monitoring the motion state of the proportional servo valve core based on the real-time pressure variation rate of each static pressure chamber; and calculating the pulse bias signal based on the mechanical hysteresis characteristics of the valve core when the proportional servo valve core crosses the zero flow transition zone. The pulse bias signal is superimposed on the servo valve control signal to generate a high-frequency pulse width modulation bias drive signal, which is then output to the proportional servo valve drive unit to eliminate the mechanical response dead zone and time delay of the proportional servo valve during micro-flow regulation.
8. The method for uniformly distributing static pressure load on a heavy-duty workbench according to claim 1, characterized in that, The heavy-duty workbench static pressure load uniform distribution control method also includes an extremely low stiffness emergency counter-pressure step: when the real-time pressure of a certain static pressure chamber is lower than the set low pressure limit threshold due to the impact of the off-center load, and the calculated real-time oil film stiffness is lower than the set safety stiffness lower limit, the amplitude gain of the adaptive pressure proportional coefficient is increased to the set upper limit of the multiplier; at the same time, differential damping feedforward flow is injected according to the rate of change of the spatial attitude tilt angle around the X-axis and around the Y-axis to forcibly lift the depressurized static pressure chamber and prevent the oil film from rupturing due to extreme extrusion and causing direct contact between the guide rail metal.
9. The method for uniformly distributing static pressure load on a heavy-duty workbench according to claim 1, characterized in that, In step S2, the process of solving the eccentric load disturbance vector includes: combining the pressure-bearing area and position coordinates of each hydrostatic chamber, using inverse dynamics to derive and solve the real-time eccentric load disturbance vector borne by each oil chamber under the action of eccentric heavy load and cutting force on the heavy-duty worktable; feeding the real-time eccentric load disturbance vector as a feedforward compensation amount to the calculation of the target oil supply flow rate adjustment command in step S4, so as to offset the interference of time-varying tooling load on the steady-state accuracy of the worktable attitude.
Citation Information
Patent Citations
A calculation method for the deformation of a hydrostatic turntable considering the effects of fluid-structure interaction.
CN106777637B
