Multi-degree-of-freedom active vibration isolation platform control system for precise instrument
By constructing a generalized pressure characteristic flow tensor model and an asymmetric bias compensation signal, the problems of response lag and insufficient correction accuracy of the fluid active vibration isolation control system in complex environments were solved, and the high-efficiency vibration isolation performance of the multi-degree-of-freedom active vibration isolation platform was realized.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-03-12
- Publication Date
- 2026-04-10
AI Technical Summary
Existing fluid active vibration isolation control systems struggle to simultaneously ensure timely fluid response and precise elimination of micro-frictional resistance when facing complex multi-dimensional oscillator coupling environments. This results in response lag and insufficient correction accuracy, limiting the vibration isolation performance of multi-degree-of-freedom active vibration isolation platforms in extremely demanding and precise detection scenarios.
By constructing a generalized pressure characteristic flow tensor model, data is acquired using multi-dimensional pose sensing and high-frequency flow pressure monitoring components, clock synchronization latching is performed, fluid pressure and mechanical state are analyzed, and a composite control signal with asymmetric bias compensation is generated to achieve independent optimization of macroscopic flow trend control and microscopic friction resistance inversion.
It effectively overcomes the physical hysteresis of fluid transmission, structural nonlinear viscosity, and dry friction coupling interference, improves the correction accuracy and resistance to external disturbances of the multi-degree-of-freedom active vibration isolation platform, and significantly improves the response hysteresis problem under complex working conditions.
Smart Images
Figure CN121832407A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of vibration isolation platform control, in particular to a multi-degree-of-freedom active vibration isolation platform control system for precision instruments. BACKGROUND
[0002] In the fields of cutting-edge industries and scientific research such as precision manufacturing, semiconductor processing, and high-end optical testing, micro vibrations caused by the environment and dynamic disturbances generated by the internal operation of equipment can have a serious adverse effect on the processing accuracy and measurement stability of precision instruments. In order to suppress these vibrations, multi-degree-of-freedom active vibration isolation platforms are widely used, and control systems that use fluid as the execution medium have become an important direction due to their strong carrying capacity and adjustable stiffness.
[0003] However, existing fluid active vibration isolation control systems often have difficulty in accurately coordinating the macro fluid power transmission and micro mechanical friction interference correlation when facing complex multi-dimensional vibrator coupling environments. On the one hand, traditional fluid control loops often have difficulty in overcoming the inherent physical hysteresis phenomenon in the transmission process of fluid medium, and are slow to respond when dealing with high-frequency disturbances that change rapidly in the external environment, making it difficult to plan actions in advance to offset pressure transmission delays. On the other hand, inherent structural stickiness and dry friction are common in the motion mechanical structure of the vibration isolation platform. These microscopic nonlinear resistance factors can seriously damage the linear response expectations of the control system when performing high-precision pose correction at the micro-nano level, resulting in unstable driving force or increased following error.
[0004] Due to the lack of a mechanism for effectively high-dimensional mapping and independent decoupling compensation of the macro pressure change trend of the fluid and the microscopic nonlinear friction of the mechanical motion, existing control methods face the dilemma of being unable to simultaneously consider the timeliness of fluid response and the fine elimination of microscopic frictional resistance, thereby greatly limiting the overall vibration isolation performance and control accuracy of the multi-degree-of-freedom active vibration isolation platform in extreme and demanding precision detection scenarios. SUMMARY
[0005] The present application provides a multi-degree-of-freedom active vibration isolation platform control system for precision instruments to solve the problems raised in the background art. The specific technical problems include how to achieve independent optimization of macro fluid trend control and microscopic friction resistance inversion by constructing and decoupling a generalized pressure feature flow tensor model, and coupling to generate a composite control signal with asymmetric bias compensation capability, to solve the technical problems of existing fluid active vibration isolation platforms that are difficult to simultaneously consider fluid transmission physical hysteresis and structural nonlinear stickiness or dry friction coupling interference, resulting in response lag and insufficient correction accuracy in complex working conditions.
[0006] To achieve the above object, the application provides the following technical scheme: a multi-degree-of-freedom active isolation platform control system for precision instruments, comprising a multi-degree-of-freedom active isolation platform, a multi-cavity liquid container arranged inside the multi-degree-of-freedom active isolation platform, a multi-channel fluid control valve unit, and a controller; the controller is configured to:
[0007] The controller comprises a multi-dimensional pose sensing array and a high-frequency flow pressure monitoring component, and is provided with a clock synchronization locking mechanism, which applies uniform hardware time stamps to the relative displacement and angular acceleration obtained by the multi-dimensional pose sensing array and the oil return back pressure and pressure pulsation data collected by the high-frequency flow pressure monitoring component, and aligns them; the controller synchronously obtains the relative displacement and angular acceleration of each physical axis as current pose state data through the multi-dimensional pose sensing array; at the same time, the oil return back pressure and flow supply side pressure pulsation in each sub-cavity of the multi-cavity liquid container are collected as fluid pressure state data through the high-frequency flow pressure monitoring component; the current pose state data and the fluid pressure state data are strictly aligned by the clock synchronization locking mechanism, effectively eliminating the sampling time delay difference of the multi-dimensional sensing system in space and time distribution, providing a high-precision time reference for subsequent accurate mapping of mechanical domain motion state and fluid domain dynamics state, and being a prerequisite for realizing high-dimensional physical state fusion.
[0008] A generalized pressure characteristic flow tensor model is constructed, and the generalized pressure characteristic flow tensor model is updated according to the current pose state data and the fluid pressure state data, specifically comprising:
[0009] The controller maps the current pose state data in multiple dimensions into a system Jacobian space state matrix;
[0010] The controller corrects the system Jacobian space state matrix by introducing an anti-singular least squares regularization factor in advance to construct a regularized Jacobian matrix, wherein the anti-singular least squares regularization factor is a non-negative dynamic anti-singular decay factor;
[0011] The fluid pressure state data is converted into a corresponding pressure gradient feature vector;
[0012] The regularized Jacobian matrix and the pressure gradient feature vector are fused by using tensor product operation to generate and update a generalized pressure characteristic flow tensor model for mapping the current external mechanical vibration and internal fluid coupling state, and the kinematics state of the mechanical domain and the dynamics state of the fluid domain are mapped to the same high-dimensional space by tensor product;
[0013] The process constructs a regularized Jacobian matrix by introducing a non-negative dynamic anti-singularity damping factor, and generates and updates a generalized pressure characteristic flow tensor model by using tensor product operation to fuse pressure gradient characteristic vectors, which can avoid the singular value dilemma in matrix operation solution under multi-dimensional complex state, realize high-dimensional feature fusion of external mechanical space state and internal fluid medium characteristics, and effectively capture the deep coupling mechanism between mechanical system and control medium.
[0014] In each control cycle, based on the updated generalized pressure characteristic flow tensor model, the flow trend flow data and the microscopic partial differential friction resistance deviation data are extracted, which specifically include:
[0015] The controller uses a pre-constructed high-low frequency orthogonal separation operator to perform frequency domain decoupling on the updated generalized pressure characteristic flow tensor model.
[0016] The low-frequency domain distribution parameters with a frequency lower than a preset basic threshold are analyzed and defined as flow trend flow data to reflect the macroscopic flow direction momentum of the fluid;
[0017] The microscopic disturbance parameters with nonlinear high-frequency characteristics are analyzed and defined as microscopic partial differential friction resistance deviation data after stripping;
[0018] The process uses a high-low frequency orthogonal separation operator to perform frequency domain decoupling to extract flow trend flow data and microscopic partial differential friction resistance deviation data, realizes data separation of macroscopic flow trend control and microscopic friction resistance inversion, reduces the feature coupling interference between fluid dynamics and microscopic friction disturbance, and enables the controller to independently optimize the system variables of different frequency bands and different physical properties.
[0019] The flow trend flow data is operated by a preset algorithm, and a low-frequency advance action instruction is input to the multi-channel fluid control valve unit, which specifically includes:
[0020] The controller predicts the target chamber internal pressure at the next steady state based on the flow trend flow data;
[0021] A feedforward time constant is introduced to offset the physical delay of pipeline fluid transport for time dimension translation processing, and the advance action timing is calculated;
[0022] The low-frequency advance action instruction is generated and issued according to the advance action timing and the target chamber internal pressure;
[0023] The process establishes a feedforward time constant and calculates an advance action timing to issue a low-frequency advance action instruction to the multi-channel fluid control valve unit, which pre-compensates for the inherent physical lag phenomenon in the internal fluid medium transport process, significantly improves the action lag problem caused by the response dead zone of the execution end of the traditional active vibration isolation platform at the macro level, and improves the timeliness of flow trend flow dynamic regulation.
[0024] Low-frequency lead-action commands are used to change the basic back pressure action plane of the multi-cavity fluidized bed and to construct the expected trajectory of the ideal fluid response under the basic back pressure action plane, specifically including:
[0025] On the basic back pressure action datum, the preset stiffness parameters and preset damping parameters are input into the standard second-order oscillatory damping equation to depict the expected trajectory of the ideal fluid response as a reference standard baseline.
[0026] This process describes the expected trajectory of the ideal fluid response on the basic back pressure action datum by combining the standard second-order oscillatory damping equation, and sets a standard smooth response baseline with deterministic reference for the macroscopic action of the multi-channel fluid control valve unit, which helps to reduce the risk of sudden pressure change in the chamber during the multi-chamber action switching process.
[0027] Extracting the actual transient frequency difference generated by a multi-cavity fluidized bed when following the expected trajectory of an ideal fluid response, specifically including:
[0028] The controller uses a state observer to calculate the actual transient pressure slope continuous function during the response process of the multi-cavity fluidized bed;
[0029] The envelope of the actual transient pressure slope continuous function is compared with that of the reference standard baseline, and differential transformation is performed to extract the corresponding phase retention angle and amplitude distortion, which together form the actual transient frequency difference.
[0030] This process extracts the actual transient frequency difference by comparing the state observer with the envelope and performing differential transformation. It can characterize the phase stagnation state and amplitude distortion degree of the multi-cavity fluid mass container during actual operation, which deviates from the expected trajectory of the ideal fluid response. This provides a sensitive and reliable trigger evaluation index for determining whether the system is trapped in micro-mechanical nonlinear disturbance stagnation.
[0031] Using the actual transient frequency difference as the activation condition, the reverse calculation of the microscopic partial differential frictional resistance deviation data is triggered to generate a resistance deviation array for characterizing structural viscosity and dry friction, specifically including:
[0032] The controller determines whether the actual transient frequency difference falls within the preset safety tolerance dead zone.
[0033] When it is determined that the actual transient frequency difference falls within the safety tolerance dead zone, micro-correction is suppressed, and operation is maintained based solely on low-frequency advance action commands;
[0034] When the actual transient frequency difference exceeds the safety tolerance dead zone, a high-frequency correction activation interruption request is issued. The controller calls the preset continuous nonlinear extended friction dynamics equation, which includes the pre-sliding displacement state and the macroscopic sliding state. The microscopic partial differential friction resistance deviation data is substituted in reverse as the disturbance to be eliminated to perform inversion analysis, quantifying the current comprehensive friction resistance distribution vector on the guide rail surface with multiple degrees of freedom, thereby constructing a hysteresis deviation array. The continuous nonlinear extended friction dynamics equation is specifically associated with internal friction state variables reflecting the pre-sliding displacement, contact surface stiffness parameters, microscopic damping parameters, and viscous hysteresis mapping coefficients. The controller establishes an internal dynamic inversion mechanism for friction force through the continuous nonlinear extended friction dynamics equation, transforming the unmeasurable microscopic disturbance into a controllable compensation quantity.
[0035] This process uses the actual transient frequency difference as the activation condition and calls the continuous nonlinear extended friction dynamics equation to implement the internal dynamic inversion mechanism. It effectively combines the pre-sliding displacement state and the macroscopic sliding state, and reverse-reasons and quantifies the nonlinear viscosity and dry friction coupling interference resistance of the precision instrument structure, which is originally difficult to measure directly, into a controllable hysteresis deviation array, thus realizing a highly numerical characterization of microscopic nonlinear disturbances.
[0036] The hysteresis deviation array is used as a nonlinear compensation term and coupled to the output of the low-frequency lead action command through a preset weighting matrix to generate a final composite correction control signal with asymmetric control derivative characteristics. This signal drives the multi-channel fluid control valve unit to perform microscopic and macroscopic parallel correction actions, specifically including:
[0037] As the absolute value of the actual transient frequency difference increases, the local gain weight of the corresponding nonlinear compensation term in the preset weight allocation matrix is increased proportionally.
[0038] The controller incorporates a direction-determining step bias function into the nonlinear compensation term. Through asymmetric bias scaling parameters, the first-order scaling rise slope of the generated final composite bias correction control signal when driving high-pressure charging is greater than the attenuation recovery slope when driving the corresponding low-pressure cavity discharge, thus forming asymmetric control derivative characteristics.
[0039] This process couples the stagnation bias array as a nonlinear compensation term to the output and mixes it with a direction-determining step bias function to form a final composite correction control signal with asymmetric control derivative characteristics. Based on macroscopic fluid advance control and microscopic friction inversion, the nonlinear weighted aggregation of the signal is completed. By using the asymmetric filling and discharging characteristics where the first-order scaling rise slope is greater than the decay recovery slope, the physical hysteresis of fluid transmission and the interference of structural nonlinear viscosity or dry friction coupling are overcome. Under complex working conditions, the response lag and error of the active correction action are improved, effectively enhancing the correction accuracy and robustness of the overall control system of the multi-degree-of-freedom active vibration isolation platform.
[0040] Compared with the prior art, the beneficial effects of the present invention are:
[0041] By acquiring current pose and fluid pressure data, and fusing the regularized Jacobian matrix and pressure gradient eigenvectors using tensor product operations, a generalized pressure characteristic flow tensor model was constructed and updated. Frequency domain decoupling was performed on the updated generalized pressure characteristic flow tensor model using high- and low-frequency orthogonal separation operators, effectively achieving independent optimization of macroscopic fluid quality trend control and microscopic frictional resistance inversion. On one hand, by introducing a feedforward time constant to compensate for the physical delay in pipeline fluid transport, the timing of advance action was calculated, thereby generating a low-frequency advance action command, which effectively compensated for the physical hysteresis phenomenon during the execution of the multi-channel fluid control valve unit. On the other hand, using the extracted actual transient frequency difference as an activation condition, the continuous nonlinear extended friction dynamics equation was called for backward calculation, transforming the microscopic partial differential frictional resistance deviation data into a resistance deviation array characterizing structural viscosity and dry friction. Finally, this was coupled as a nonlinear compensation term to the output of the low-frequency advance action command, synthesizing a final composite correction control signal with asymmetric control derivative characteristics.
[0042] The present invention effectively overcomes the technical problems of existing fluid active vibration isolation platforms in complex working conditions, such as response lag and insufficient correction accuracy caused by difficulty in taking into account the physical hysteresis of fluid transmission, nonlinear viscosity of structure, and dry friction coupling interference. It significantly improves the correction performance and anti-external disturbance capability of the overall control system of multi-degree-of-freedom active vibration isolation platform. Attached Figure Description
[0043] Figure 1 This is a schematic diagram of the overall core process of the present invention;
[0044] Figure 2 This is a schematic diagram of the model construction and dual-track decoupling process of the present invention;
[0045] Figure 3 This is a schematic diagram of the feedforward control and actual transient frequency difference extraction process of the present invention;
[0046] Figure 4 This is a schematic diagram of the dynamic inversion and final composite correction control signal generation process of the present invention;
[0047] Figure 5 This is a comparative test diagram of the correction performance of the multi-degree-of-freedom active vibration isolation platform of the present invention. Detailed Implementation
[0048] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0049] Next, please refer to Figure 1 This invention provides a technical solution: a multi-degree-of-freedom active vibration isolation platform control system for precision instruments, comprising a multi-degree-of-freedom active vibration isolation platform, a multi-cavity fluid mass container disposed inside the multi-degree-of-freedom active vibration isolation platform, a multi-channel fluid control valve unit responsible for fluid distribution, and a controller as the core hub; wherein, in this embodiment, the multi-degree-of-freedom active vibration isolation platform relies on its internal fluid actuator for support and posture adjustment; in addition, the multi-degree-of-freedom active vibration isolation platform control system for precision instruments also includes a multi-dimensional posture sensing array and a high-frequency flow pressure monitoring component respectively connected to the controller for real-time feedback of physical state.
[0050] The controller updates the generalized pressure characteristic flow tensor model based on the current pose state data and fluid pressure state data to extract fluid trend flow data and microscopic partial differential frictional obstruction deviation data. The former is used to input low-frequency advanced action commands and construct the expected trajectory of ideal fluid response, extract the actual transient frequency difference as the activation condition to trigger reverse calculation, and generate the obstruction deviation array. The obstruction deviation array is used as a nonlinear compensation term and coupled through a preset weight allocation matrix to generate the final composite correction control signal with asymmetric control derivative characteristics. This signal drives the multi-channel fluid control valve unit to perform microscopic and macroscopic parallel correction actions, which significantly improves the correction fidelity and anti-interference stability of the multi-degree-of-freedom active vibration isolation platform.
[0051] In this embodiment, the controller is configured to perform the following overall dynamic correction control steps:
[0052] Please see Figure 2First, during system initialization and data acquisition, the controller acquires the current pose and fluid pressure data of the multi-degree-of-freedom fluid actuator. To ensure time and operating condition alignment of multi-source data, the controller incorporates a clock synchronization latching mechanism. This mechanism assigns a unified hardware timestamp to the relative displacement and angular acceleration data acquired by the multi-dimensional pose sensor array, as well as the return oil back pressure and pressure pulsation data acquired by the high-frequency flow pressure monitoring component, achieving microsecond-level alignment. Displacement and flow pressure are chosen as input parameters because they directly reflect the kinematic response of the mechanical domain and the fundamental dynamics of the fluid domain, respectively, forming the basis for achieving machine-fluid coupling control. Specifically, the controller synchronously acquires the relative displacement and angular acceleration of each physical axis through the multi-dimensional pose sensor array as the current pose data. Simultaneously, it acquires the return oil back pressure and supply-side pressure pulsation within each sub-cavity of the multi-cavity fluid mass container in real time through the high-frequency flow pressure monitoring component, serving as the fluid pressure data.
[0053] Subsequently, the controller constructs a generalized pressure characteristic flow tensor model and updates it based on the current pose state data and fluid pressure state data. In specific calculations, the controller maps the current pose state data in multiple dimensions to the system Jacobian space state matrix. ,
[0054] In this process, to prevent the system's Jacobian space state matrix from being affected by the coplanarity of multiple degrees of freedom (leading to the loss of local degrees of freedom of the platform). Singularity occurs; the controller pre-emptively introduces a singularity-preventing least-squares regularization factor. For the system's Jacobian space state matrix Make corrections and construct a regularized Jacobian matrix. The corrected formula is as follows: ,in, The non-negative dynamic anti-singularity decay factor is extremely small (by system calibration, its value is a predetermined minimum constant when the system is far from the singularity). Approaching 0, and increasing exponentially as it nears singularities to ensure the matrix is full rank. To be related to the system's Jacobian space state matrix Identity matrices of the same order.
[0055] Convert fluid pressure state data into corresponding pressure gradient feature vectors. Next, tensor product operations are used to fuse the Jacobian space state matrix of the system. With pressure gradient eigenvector Generate and update the generalized pressure characteristic flow tensor model to map the current coupling state between external mechanical vibration and internal fluid. The specific formula is as follows:
[0056] ,in, This represents a multi-degree-of-freedom displacement vector in generalized coordinates. Represents the tensor product operator. The fundamental constant parameters characterizing fluid viscosity (preset range of values is 100%) (Based on the dynamic viscosity of the working fluid as per factory calibration). The standard identity matrix is used; the technical purpose of this formula is to map the kinematic state of the mechanical domain and the dynamic state of the fluid domain to the same high-dimensional space through tensor product, thereby achieving the underlying fusion of physical state quantities.
[0057] Within each control cycle, based on the updated generalized pressure characteristic flow tensor model The system extracts flow trend data and microscopic partial differential frictional obstruction deviation data; the controller employs a pre-constructed high- and low-frequency orthogonal separation operator. Frequency domain decoupling is performed on the updated generalized pressure characteristic flow tensor model, specifically including:
[0058] On the one hand, low-frequency domain distribution parameters with frequencies below a preset baseline threshold are analyzed and defined as flow trend data to reflect the macroscopic flow momentum of the fluid, as shown in the formula. ,in To separate the obtained fluid trend flow data to reflect the macroscopic flow momentum of the fluid, This is a low-pass quadrature filter operator with a set cutoff frequency of 5Hz. For frequency variables;
[0059] On the other hand, the microscopic perturbation parameters with nonlinear high-frequency characteristics are analyzed and, after being stripped away, defined as microscopic partial differential frictional resistance deviation data, the formula is as follows: ,in This refers to the microscopic partial differential frictional resistance deviation data extracted from the sample.
[0060] The purpose of this frequency domain decoupling technique is to accurately separate the complex coupled model into low-frequency feedforward basis that guides the macroscopic trend and perturbation characteristics that reveal the microscopic resistance.
[0061] Please see Figure 3 After separating the dual-track data (fluid-quality trend data) and microscopic partial differential frictional resistance deviation data After that, the controller performs macro-level feedforward control, specifically including:
[0062] Fluid quality trend data Based on a pre-defined algorithm, a low-frequency advance action command is input to the multi-channel fluid control valve unit; during this process, the controller predicts the next steady-state pressure within the target chamber based on fluid quality trend data. Subsequently, a feedforward time constant was introduced to compensate for the physical delay in fluid transport in the pipeline. (The range is [0.005, 0.02] s, depending on the empirical value of the hydraulic pipeline length test) Time-dimension translation is performed to calculate the advance action timing; based on the advance action timing and the pressure in the target chamber, a low-frequency advance action command is generated. And issued; the formula is as follows:
[0063] , The macroscopic flow proportional gain coefficient (based on the valve flow characteristic curve) (between adjustment) For the current moment; based on this, change the basic back pressure action datum of the multi-cavity fluidized bed; on this basic back pressure action datum, set the preset stiffness parameters. and preset damping parameters Input the standard second-order oscillatory damping equation and plot the expected trajectory of the ideal fluid response, using the standard baseline as a reference. The formula is expressed as follows:
[0064] ,in For the Laplace operator, For the reason The system's inherent frequency is determined. For the reason The damping ratio is determined; the technical purpose of this formula is to provide a highly smooth theoretical fluid response reference baseline without physical friction damage for subsequent micro-error comparisons.
[0065] The preset algorithm in the above process is defined as the time-biased characteristic momentum mapping algorithm. First, the controller uses the transfer function model to analyze the flow trend data. By performing flow-momentum mapping, the target chamber pressure required to achieve steady-state correction in the next sampling period is predicted. Subsequently, a preset feedforward time constant is introduced. The pressure target is shifted over time to generate a leading time reference that compensates for the physical delays in hydraulic pipeline transport; finally, a preset flow rate proportional gain coefficient is applied. The above predictions and the translated composite pressure characteristics are linearly mapped into a voltage command signal that drives the multi-channel fluid control valve unit. .
[0066] However, microscopic variations inevitably exist during actual fluid filling or releasing. Therefore, the system extracts the actual transient frequency difference generated by the multi-cavity fluidized bed as it follows the expected trajectory of the ideal fluid response. The controller uses a state observer to calculate the actual transient pressure slope continuous function during the response process of the multi-cavity fluidized bed. The envelope of the variable is compared with the aforementioned reference standard baseline using differential transformation to extract the corresponding phase retention angle and amplitude distortion. These two variables together constitute the actual transient frequency difference. The formula is as follows:
[0067] .
[0068] Please see Figure 4 To avoid micro-oscillations in the control system, the actual transient frequency difference is used here. As an activation condition, it triggers the microscopic partial differential frictional resistance deviation data. By performing backward calculations, a hindrance deviation array is generated to characterize structural viscosity and dry friction. The controller contains adaptive dead-time detection logic, specifically including:
[0069] When the actual transient frequency difference is determined When the signal falls within the preset safety tolerance dead zone, micro-corrections are suppressed, and operation is maintained solely based on low-frequency lead-action commands; only when the actual transient frequency difference... When the distance exceeds the safety tolerance dead zone, a high-frequency correction activation interrupt request is issued. After this mechanism is triggered, the controller calls a preset continuous nonlinear extended friction dynamics equation that includes the pre-sliding displacement state and the macroscopic sliding state, and then applies the microscopic partial differential friction resistance deviation data. Substituting the disturbance to be eliminated into the inverse analysis, the current comprehensive frictional resistance distribution vector on the guide rail surface with multiple degrees of freedom is quantified, thereby constructing a hysteresis deviation array. The formula for the continuous nonlinear extended friction dynamics equation is expressed as follows:
[0070] ,in To reflect the internal friction state variables of the pre-sliding displacement, For the contact surface stiffness parameters, For micro damping parameters, For viscous hysteresis mapping coefficient; the technical purpose of this formula is to establish an internal dynamic inversion mechanism for friction, transforming unmeasurable microscopic disturbances into controllable compensation quantities.
[0071] For multi-degree-of-freedom active vibration isolation platforms, the actual frictional behavior under minute displacements includes not only macroscopic Coulomb friction or viscous friction, but also the elastoplastic hysteresis effect caused by the microscopic protrusion deformation of the contact surface during the pre-sliding stage. Therefore, in the continuous nonlinear extended friction dynamics equations, an internal frictional state variable reflecting the pre-sliding displacement is specifically introduced. In this real-time inversion analysis, the controller first captures the relative motion velocity of the multi-degree-of-freedom guide surface through a multi-dimensional pose sensor array, and then establishes information about the internal friction state variables accordingly. The first-order nonlinear differential update equation; the internal friction state variable The dynamic characterization of the average elastic deformation process of the micro-contact surface (similar to countless micro-spring bristles) is as follows: the platform undergoes elastic bending in the initial stage of extremely small motion, and gradually enters the unloading or saturation state of macro-slip when the displacement or velocity increases to a threshold.
[0072] Accurately obtaining internal friction state variables After taking its differential time derivative, the system will extract the microscopic partial differential frictional resistance deviation data. The input is sent to the controller; at this point, the controller will call three key tribological characteristic calibration parameters stored at the system's underlying layer to perform a mechanical reconstruction of friction-induced losses, specifically including:
[0073] Contact surface stiffness parameters : Used to quantify the geometric tangential springback capability generated by the microscopic deformation of the guide rail, the controller will With the internal friction state variable at this time By multiplying, the pure rigid rebound force component of the micro spring can be accurately estimated;
[0074] Micro damping parameters Used to quantify the internal material hysteresis damping caused by the rate of microscopic deformation, the controller correlates it with the internal friction state variable. Multiply the time derivative terms to extract the internal damping force component caused by microscopic motion hysteresis;
[0075] Viscous mapping coefficient This is used to characterize the fluid viscous shear stress effect of a multi-cavity fluid cavity under high-frequency micro-excitation, which varies nonlinearly with velocity domain. Unlike traditional tribodynamic models, this embodiment innovatively uses... Direct modulation of microscopic partial differential frictional resistance deviation data To replace the simple macroscopic velocity viscosity term, the equivalent fluid-structure interaction aliasing resistance caused by nonlinear high-frequency disturbances in the fluid domain is derived.
[0076] Finally, the controller executes an internal dynamic inversion mechanism, performing a physical-level tensor superposition of the microscopic pure rigid rebound force component calculated from the contact surface stiffness parameters, the microscopic internal damping force component calculated from the microscopic damping parameters, and the fluid-structure interaction aliasing drag component extracted jointly from the viscous hysteresis mapping coefficient and microscopic deviation data. Through this inversion logic, the high-frequency microscopic disturbances (i.e., the co-existing time-varying attenuation caused by frictional disturbances and high-frequency fluid pulsations), which were originally unmeasurable in conventional mechanical sensing, are robustly and accurately transformed into a controllable compensation force vector in a multi-axis coordinate system, i.e., the successfully output hysteresis deviation array. This compensation mechanism eliminates the nonlinear crawling phenomenon of active vibration isolation systems when commutating at extremely small amplitudes or crossing dead zones from the underlying mechanism, thereby improving the absolute static platform fidelity of precision instruments.
[0077] The end control link performs fusion output, and the controller will impede the deviation array. As a nonlinear compensation term, it is used through a preset weight allocation matrix. Coupled to low-frequency advance action command The output terminal is mixed with a direction-determining step bias function. This generates a final composite correction control signal with asymmetric control derivative characteristics. This drives the multi-channel fluid control valve unit to perform microscopic and macroscopic parallel correction actions; in this fusion step, the weighted modulation features include the following:
[0078] With actual transient frequency difference As the absolute value increases, the preset weight allocation matrix is adjusted proportionally upwards. The controller incorporates the local gain weights corresponding to the nonlinear compensation term. Furthermore, to address the asymmetry in damping characteristics between fluid outflow and inflow processes, a direction-determining step bias function is introduced into the nonlinear compensation term. This results in the generated final composite correction control signal. The first-order scaling slope during high-pressure charging is significantly greater than the attenuation recovery slope during discharge to the corresponding low-pressure chamber. This asymmetric control derivative characteristic enables the multi-degree-of-freedom active vibration isolation platform to efficiently suppress small excitation forces with a fast unilateral response. The formula is as follows:
[0079] ,in Is it following The function matrix that proportionally increases the local gain weight as the absolute value increases, i.e., the preset weight allocation matrix; For direction-determining step bias functions, when fluid flows into the high-pressure chamber (velocity) When the value is ) Flowing into the low-pressure chamber (velocity) When the value is ) The technical purpose of this generative formula is to achieve asymmetric bias scaling parameters (the direction-determining step bias function in the formula). By setting two asymmetric values (1.5 and 0.8) for the fluid flow direction, the first-order expansion ramp slope during high-pressure charging is significantly greater than the attenuation recovery ramp during discharge, thereby achieving efficient unilateral fast-response suppression of small excitation forces by the multi-degree-of-freedom active vibration isolation platform.
[0080] Please see Figure 5In the figure, the dashed line represents the expected trajectory of the ideal fluid response, which has the theoretical characteristic of smoothness. The existing fluid active vibration isolation system (dotted line) is affected by the physical delay of fluid transport in the pipeline, as well as the nonlinear viscosity and dry friction of the structure. Its response curve has obvious phase lag and obvious nonlinear creep phenomenon in the commutation region. In contrast, the actual response curve (solid line) under the action of the final composite correction control signal with asymmetric control derivative characteristics generated in this embodiment pre-compensates the physical lag through low-frequency advance action command and eliminates friction interference by using the compensation term generated by microscopic partial differential friction blocking deviation data. As can be seen from the figure, the response curve of this embodiment has a significantly improved first-order scaling slope when driving high-pressure charging and maintains extremely high absolute static plateau fidelity at the smallest amplitude commutation, which verifies the significant improvement of the present invention in terms of correction accuracy and anti-interference stability.
[0081] As can be seen from the above description, the multi-degree-of-freedom active vibration isolation platform control system for precision instruments provided in this embodiment has the following technical effects:
[0082] This embodiment provides a multi-degree-of-freedom active vibration isolation platform control system for precision instruments. By precisely separating the complex coupling model into a low-frequency feedforward basis that guides the macroscopic direction and a disturbance characteristic that reveals the microscopic resistance, and establishing an internal dynamic inversion mechanism for friction, it successfully transforms the unmeasurable microscopic disturbance into a controllable compensation quantity. This compensation mechanism eliminates the nonlinear crawling phenomenon of the active vibration isolation system when commutating at extremely small amplitudes or crossing dead zones from the underlying mechanism, improving the absolute static platform fidelity of precision instruments. At the same time, by generating a final composite correction control signal with asymmetric control derivative characteristics, the first-order scaling rise slope during high-voltage charging is significantly greater than the attenuation recovery slope during discharge, thereby achieving efficient unilateral fast-response suppression of small excitation forces by the multi-degree-of-freedom active vibration isolation platform, ultimately significantly improving the correction fidelity and anti-interference stability of the multi-degree-of-freedom active vibration isolation platform.
[0083] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.
Claims
1. A multi-degree-of-freedom active vibration isolation platform control system for precision instruments, characterized in that, It includes a multi-degree-of-freedom active vibration isolation platform, a multi-cavity fluid-mass container disposed inside the multi-degree-of-freedom active vibration isolation platform, a multi-channel fluid control valve unit, and a controller; the controller is configured to: Acquire the current pose state data of the multi-degree-of-freedom active vibration isolation platform and the fluid pressure state data of the multi-cavity fluidized bed; Construct a generalized pressure characteristic flow tensor model, and update the generalized pressure characteristic flow tensor model based on the current pose state data and the fluid pressure state data; Within each control cycle, based on the updated generalized pressure characteristic flow tensor model, fluid trend flow data and microscopic partial differential frictional obstruction deviation data are extracted. The fluid trend flow data is processed by a preset algorithm, and a low-frequency advance action command is input to the multi-channel fluid control valve unit. The low-frequency advance action command is used to change the basic back pressure action datum of the multi-cavity fluid container and construct the ideal fluid response expected trajectory under the basic back pressure action datum. Extract the actual transient frequency difference generated by the multi-cavity fluidic cavity when it follows the expected trajectory of the ideal fluid response; Using the actual transient frequency difference as an activation condition, the microscopic partial differential frictional obstruction deviation data is reverse-calculated to generate an obstruction deviation array for characterizing structural viscosity and dry friction. The stagnation deviation array is used as a nonlinear compensation term and coupled to the output of the low-frequency advance action command through a preset weight allocation matrix to generate a final composite correction control signal with asymmetric control derivative characteristics, thereby driving the multi-channel fluid control valve unit to perform microscopic and macroscopic parallel correction actions.
2. The multi-degree-of-freedom active vibration isolation platform control system for precision instruments according to claim 1, characterized in that, The controller includes a multi-dimensional pose sensing array and a high-frequency flow pressure monitoring component, and is equipped with a clock synchronization latching mechanism to align the relative displacement and angular acceleration data captured by the multi-dimensional pose sensing array and the back pressure and pressure pulsation data collected by the high-frequency flow pressure monitoring component with a unified hardware timestamp. The controller synchronously acquires the relative displacement and angular acceleration of each physical axis through the multi-dimensional pose sensing array as the current pose state data; at the same time, it collects the return oil back pressure and supply side pressure pulsation inside each sub-cavity of the multi-cavity fluid mass container in real time through the high-frequency flow pressure monitoring component as the fluid pressure state data.
3. The multi-degree-of-freedom active vibration isolation platform control system for precision instruments according to claim 1, characterized in that, The construction and updating process of the generalized pressure characteristic flow tensor model specifically includes: The controller maps the current pose state data in multiple dimensions to a system Jacobian space state matrix. The controller pre-corrects the system's Jacobian space state matrix by introducing an anti-singularity least squares regularization factor to construct a regularized Jacobian matrix, wherein the anti-singularity least squares regularization factor is a non-negative dynamic anti-singularity decay factor. The fluid pressure state data is converted into a corresponding pressure gradient feature vector; By using tensor product operations to fuse the regularized Jacobian matrix and the pressure gradient eigenvector, the generalized pressure characteristic flow tensor model used to map the current coupling state between external mechanical vibration and internal fluid is generated and updated. The kinematic state of the mechanical domain and the dynamic state of the fluid domain are mapped to the same high-dimensional space through tensor product.
4. The multi-degree-of-freedom active vibration isolation platform control system for precision instruments according to claim 1, characterized in that, The extraction process of the fluid trend data and the microscopic partial differential frictional obstruction deviation data specifically includes: The controller employs a pre-constructed high- and low-frequency orthogonal separation operator to perform frequency domain decoupling on the updated generalized pressure characteristic flow tensor model; Low-frequency domain distribution parameters with frequencies below a preset basic threshold are analyzed and defined as fluid quality trend flow data to reflect the macroscopic flow momentum of the fluid. The microscopic perturbation parameters with nonlinear high-frequency characteristics were analyzed and defined as microscopic partial differential frictional obstruction deviation data after being stripped away.
5. The multi-degree-of-freedom active vibration isolation platform control system for precision instruments according to claim 1, characterized in that, The process of generating the low-frequency advance action command specifically includes: The controller predicts the next steady-state pressure within the target chamber based on the fluid trend flow data; By introducing a feedforward time constant to offset the physical delay in pipeline fluid transport and performing a time dimension shift, the timing of advance action can be calculated. Based on the aforementioned advance action timing and the pressure within the target chamber, a low-frequency advance action command is generated and issued.
6. The multi-degree-of-freedom active vibration isolation platform control system for precision instruments according to claim 1, characterized in that, The process of constructing the expected trajectory of the ideal fluid response specifically includes: On the basic back pressure action base surface, preset stiffness parameters and preset damping parameters are input into the standard second-order oscillatory damping equation to depict the expected trajectory of the ideal fluid response, which serves as a reference standard baseline.
7. The multi-degree-of-freedom active vibration isolation platform control system for precision instruments according to claim 6, characterized in that, The extraction process of the actual transient frequency difference specifically includes: The controller uses a state observer to calculate the actual transient pressure slope continuous function during the response process of the multi-cavity fluidized bed. The actual transient pressure slope continuous function is compared with the reference standard baseline by performing an envelope comparison and differential transformation to extract the corresponding phase retention angle and amplitude distortion, which together form the actual transient frequency difference.
8. The multi-degree-of-freedom active vibration isolation platform control system for precision instruments according to claim 1, characterized in that, The generation process of the hindrance bias array specifically includes: The controller determines whether the actual transient frequency difference falls within the preset safety tolerance dead zone range; When it is determined that the actual transient frequency difference falls within the safety tolerance dead zone, micro-correction is suppressed, and operation is maintained solely based on the low-frequency advance action command. When the actual transient frequency difference exceeds the safety tolerance dead zone range, a high-frequency correction activation interruption request is issued. The controller calls the preset continuous nonlinear extended friction dynamics equation that includes the pre-sliding displacement state and the macroscopic sliding state. The microscopic partial differential friction resistance deviation data is substituted in reverse as the disturbance to be eliminated to perform inversion analysis, quantify the current comprehensive friction resistance distribution vector on the guide rail surface with multiple degrees of freedom, and thus construct the resistance deviation array.
9. The multi-degree-of-freedom active vibration isolation platform control system for precision instruments according to claim 8, characterized in that, The continuous nonlinear extended friction dynamics equation is specifically associated with internal friction state variables reflecting pre-sliding displacement, contact surface stiffness parameters, micro-damping parameters, and viscous hysteresis mapping coefficients; the controller establishes an internal dynamic inversion mechanism of friction force through the continuous nonlinear extended friction dynamics equation, transforming unmeasurable micro-interferences into controllable compensation quantities.
10. The multi-degree-of-freedom active vibration isolation platform control system for precision instruments according to claim 1, characterized in that, The process of generating the final composite correction control signal specifically includes: As the absolute value of the actual transient frequency difference increases, the local gain weight corresponding to the nonlinear compensation term in the preset weight allocation matrix is increased proportionally. The controller incorporates a direction-determining step bias function into the nonlinear compensation term. Through asymmetric bias scaling parameters, the first-order scaling rise slope of the generated final composite bias correction control signal when driving high-pressure charging is greater than the attenuation recovery slope when driving the corresponding low-pressure cavity discharge, thus forming asymmetric control derivative characteristics.