Positioning system for pulsed ion beam processing optical element

Abstract

The invention relates to the technical field of pulsed ion beam processing, and discloses a positioning system for a pulsed ion beam processing optical element, which can accurately predict nonlinear structure deformation in a complex thermal environment and eliminate compensation deviation caused by a traditional linear model. A multi-sensor data fusion mechanism effectively suppresses the interference of local measurement noise on a control decision, and improves the reliability of a compensation strategy. And a closed-loop control system realizes real-time verification and dynamic optimization of the compensation effect, and machining precision reduction caused by error accumulation is avoided. The submicron executing mechanism ensures that the thermal drift compensation amount is accurately converted into platform pose adjustment, the high-precision optical element machining requirement is met, and the problem that a traditional positioning system is incomplete in model input due to the single temperature field information collection dimension is solved. Meanwhile, the dynamic characteristics of the temperature-displacement coupling relation are verified through high-precision displacement data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of pulsed ion beam processing, in particular to a positioning system for pulsed ion beam processing of optical elements. BACKGROUND

[0002] Pulsed ion beam is widely used in the terminal shaping stage of high-end optical devices due to its non-contact, uniform material removal and high processing precision. To ensure the accurate maintenance and dynamic adjustment of the spatial position and attitude of the optical element during processing, the positioning system as the core of support and regulation directly affects the final processing quality and system stability, and is an indispensable key subsystem for realizing high-precision ion beam processing.

[0003] However, in the prior art, the positioning system for pulsed ion beam processing of optical elements generally uses a thermal drift compensation model based on linear relationship or empirical function, which is difficult to effectively cope with the nonlinear thermal displacement caused by material inhomogeneity, structural thermal response lag, multi-source thermal disturbance and other factors in the actual processing process. The existing compensation method has obvious shortcomings in modeling accuracy, dynamic response ability and environmental adaptability, often leading to positioning deviation accumulation, compensation delay or miscompensation when the thermal steady state has not been established or the environmental temperature fluctuates. In addition, the current system generally lacks adaptive learning and online optimization ability for historical thermal drift behavior, limiting its application in complex surface and high repeatability scenarios.

[0004] Therefore, we propose a positioning system for pulsed ion beam processing of optical elements to solve the above problems. SUMMARY

[0005] The present application aims to provide a positioning system for pulsed ion beam processing of optical elements to solve the above problems in the prior art. The positioning system for pulsed ion beam processing of optical elements generally uses a thermal drift compensation model based on linear relationship or empirical function, which is difficult to effectively cope with the nonlinear thermal displacement caused by material inhomogeneity, structural thermal response lag, multi-source thermal disturbance and other factors in the actual processing process. The existing compensation method has obvious shortcomings in modeling accuracy, dynamic response ability and environmental adaptability, often leading to positioning deviation accumulation, compensation delay or miscompensation when the thermal steady state has not been established or the environmental temperature fluctuates. In addition, the current system generally lacks adaptive learning and online optimization ability for historical thermal drift behavior, limiting its application in complex surface and high repeatability scenarios.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a positioning system for processing optical components with pulsed ion beam, comprising a multi-point sensing module, a nonlinear modeling module, a data fusion module, an intelligent control module, a micro-motion execution module, and an effect evaluation module; The multi-point sensing module is used to collect temperature and displacement data of key parts of the positioning system in real time. The data covers the thermal state and structural deformation information of multiple key points in the positioning system, providing basic data support for subsequent compensation strategies. The nonlinear modeling module is used to establish a nonlinear mapping relationship model between temperature change and structural displacement, accurately describing the nonlinear characteristics of heat-displacement interaction. The data fusion module is used to jointly process, filter, and optimize temperature and displacement data from different sensors to ensure data consistency, accuracy, and reliability. The intelligent control module generates a thermal drift compensation control strategy in real time based on the temperature and displacement information output by the data fusion module and the prediction model established by the nonlinear modeling module. The micro-motion execution module is used to drive the positioning platform to perform sub-micron level displacement and attitude adjustment based on the compensation commands output by the intelligent control module. The effect evaluation module is used to evaluate and provide feedback on the execution effect after thermal drift compensation is implemented in real time, and to determine whether the compensation has achieved the expected positioning accuracy.

[0007] Preferably, the multi-point sensing module includes a temperature detection unit and a displacement detection unit; The temperature detection unit is used to collect temperature change information such as ambient temperature, structural temperature rise, and heat conduction near the processing area in real time, and output multi-dimensional temperature field data to provide basic input variables for the thermal-displacement model. The displacement detection unit is used to perform nanometer-level measurements of key parameters such as spatial displacement and angular attitude of optical elements, and outputs structural response data in real time. This data, together with temperature data, is used for nonlinear modeling and compensation judgment.

[0008] Preferably, the nonlinear modeling module includes a data preprocessing unit and a model building unit; The data preprocessing unit is used to clean, denoise and normalize the raw temperature and displacement data from the multi-point sensing module, perform time synchronization, feature extraction and outlier removal, and construct the processed dataset into a training input format. The model building unit is used to employ machine learning algorithms to automatically select modeling structures and parameters based on different working conditions or structural conditions, and output a predictive model for subsequent control modules to make real-time compensation decisions.

[0009] Preferably, the data fusion module includes a multi-source data fusion unit and a state self-correction unit; The multi-source data fusion unit is used to receive temperature and displacement data from multiple sensors, and to perform comprehensive processing on the spatiotemporal distributed data based on weighted averaging, Kalman filtering and Bayesian fusion algorithms, and output the fused data. The state self-correction unit is used to monitor the drift trend or abnormal fluctuation of the sensor output, automatically identify potential distorted data or inconsistent states, and dynamically correct the fusion strategy.

[0010] Preferably, the intelligent control module includes a compensation strategy generation unit and an adaptive control unit; The compensation strategy generation unit is used to receive the predicted displacement deviation data output by the nonlinear modeling module, integrate the current state information, calculate the required spatial displacement compensation amount in real time, and construct the corresponding control command sequence, including multi-dimensional compensation parameters for position adjustment and attitude change. The adaptive control unit is used to automatically select the control model according to different operating conditions and dynamically adjust the control parameters.

[0011] Preferably, the micro-motion execution module includes a precision drive unit and an attitude feedback unit; The precision drive unit is used to receive control commands and perform fine-tuning operations in the XYZ directions and attitude angles; The attitude feedback unit is used to monitor the spatial attitude changes of the platform in real time.

[0012] Preferably, the effect evaluation module includes a compensation error detection unit and a model feedback update unit; The compensation error detection unit is used to calculate the residual error and determine whether the current compensation effect meets the preset accuracy requirements. The model feedback update unit is used to correct the model or fine-tune the control parameters according to the error change trend.

[0013] Compared with the prior art, the beneficial effects of the present invention are: 1. This application can accurately predict nonlinear structural deformation under complex thermal environments, eliminating compensation deviations caused by traditional linear models. The multi-sensor data fusion mechanism effectively suppresses the interference of local measurement noise on control decisions, improving the reliability of the compensation strategy. The closed-loop control system enables real-time verification and dynamic optimization of the compensation effect, avoiding the decrease in machining accuracy caused by error accumulation. Submicron-level actuators ensure that thermal drift compensation is accurately converted into platform pose adjustment, meeting the requirements of high-precision optical component machining.

[0014] 2. This application solves the problem of incomplete model input caused by the single dimension of temperature field information acquisition in traditional positioning systems. Simultaneously, it verifies the dynamic characteristics of the temperature-displacement coupling relationship through high-precision displacement data. The multidimensional temperature field data output by the temperature detection unit accurately reflects the spatiotemporal distribution characteristics of different heat sources, while the nanometer-level measurement accuracy of the displacement detection unit can capture initial minute deformation trends. The synergistic effect of both enables the nonlinear modeling module to establish a more accurate thermally induced displacement prediction model, thereby improving the reliability and environmental adaptability of thermal drift compensation decisions. Attached Figure Description

[0015] Figure 1 This is a system flowchart of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some 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.

[0017] Example 1: Please refer to Figure 1 A positioning system for processing optical components with pulsed ion beam includes a multi-point sensing module, a nonlinear modeling module, a data fusion module, an intelligent control module, a micro-motion execution module, and an effect evaluation module. The multi-point sensing module is used to collect temperature and displacement data of key parts of the positioning system in real time. The data covers the thermal state and structural deformation information of multiple key points in the positioning system, providing basic data support for subsequent compensation strategies. The nonlinear modeling module is used to establish a nonlinear mapping relationship model between temperature change and structural displacement, accurately describing the nonlinear characteristics of heat-displacement interaction; The data fusion module is used to jointly process, filter, and optimize temperature and displacement data from different sensors to ensure data consistency, accuracy, and reliability. The intelligent control module generates a thermal drift compensation control strategy in real time based on the temperature and displacement information output by the data fusion module and the prediction model established by the nonlinear modeling module. The micro-motion actuator module is used to drive the positioning platform to perform sub-micron level displacement and attitude adjustments based on the compensation commands output by the intelligent control module. The effect evaluation module is used to evaluate and provide feedback on the implementation effect of thermal drift compensation in real time, and to determine whether the compensation has achieved the expected positioning accuracy.

[0018] In this embodiment, the positioning system for pulsed ion beam machining of optical components generally adopts a thermal drift compensation model based on linear relationships or empirical functions. Such models are ill-suited to addressing the nonlinear thermally induced displacement problems caused by material inhomogeneity, structural thermal response hysteresis, and multi-source thermal disturbances during actual machining. Existing systems often experience accumulated positioning deviations, compensation delays, or miscompensation when thermal steady-state conditions are not yet established or when ambient temperature fluctuates, leading to decreased machining accuracy. For example, when machining complex curved surface optical components, traditional systems cannot adapt to nonlinear deformations caused by local temperature gradients, resulting in a mismatch between the compensation strategy and actual thermal drift.

[0019] This application proposes a positioning system comprising a multi-point sensing module, a nonlinear modeling module, a data fusion module, an intelligent control module, a micro-motion execution module, and an effect evaluation module. The multi-point sensing module collects real-time temperature and displacement data from key locations, covering the thermal state and structural deformation information of multiple key points. The nonlinear modeling module establishes a nonlinear mapping model between temperature changes and structural displacement. The data fusion module jointly processes, filters, and optimizes temperature and displacement data from different sensors. The intelligent control module generates a thermal drift compensation control strategy based on the fused data and predictive model. The micro-motion execution module drives the positioning platform to perform sub-micron-level displacement and attitude adjustments. The effect evaluation module provides real-time evaluation and feedback on the compensation effect.

[0020] The system comprises the following modules: a multi-point sensing module, a device that synchronously acquires temperature and displacement fields through a distributed sensor network (specifically, a combination of a high-precision thermocouple array and a laser interferometer), used to capture spatial heterogeneity during heat conduction; a nonlinear modeling module, a computational unit that constructs thermodynamic coupling relationships based on machine learning algorithms (specifically, a deep neural network used to train a thermal-displacement response model), used to characterize the nonlinear relationship between the thermal expansion coefficient of a material and temperature; a data fusion module, a processing system that integrates multi-source heterogeneous data (specifically, a Kalman filter and Bayesian inference algorithm), used to eliminate sensor measurement errors and improve data confidence; an intelligent control module, a decision-making unit that generates dynamic compensation strategies (specifically, a model predictive control algorithm), used to adjust the compensation amount based on real-time thermal conditions; a micro-motion execution module, an electromechanical drive device that achieves nanometer-level positioning (specifically, a piezoelectric ceramic actuator and a flexible hinge mechanism), used to accurately correct the platform's pose; and an effect evaluation module, a feedback unit that monitors compensation accuracy online (specifically, an optical encoder and an inertial measurement unit), used to verify whether the residual error meets the accuracy threshold.

[0021] Temperature and displacement data are synchronously acquired through a distributed sensor network, eliminating blind spots inherent in traditional single-point monitoring. After data cleaning and feature extraction, a thermodynamic coupling model is trained using a neural network, reflecting the dynamic relationship between temperature gradient and structural deformation. Multi-sensor data is spatiotemporally registered and probabilistically fused to generate consistent thermal state characterization parameters. The control strategy calculates multi-dimensional compensation commands based on the displacement deviation output by the prediction model and sends them to the actuator. The piezoelectric drive device achieves sub-micron-level positioning adjustment through the principle of micro-displacement superposition, while attitude feedback signals are transmitted back to the evaluation module in real time. Residual error analysis results are used to update model parameters, forming a closed-loop control link from data acquisition to compensation verification.

[0022] Compared to existing technologies, traditional solutions use linear regression models to describe the heat-displacement relationship, which cannot handle the hysteresis effect caused by sudden temperature changes. This solution uses nonlinear modeling to accurately characterize the exponential characteristics of the coefficient of thermal expansion with temperature, effectively solving the compensation inaccuracy problem caused by thermal response delay. Existing systems rely on data from a single temperature sensor, making them susceptible to local measurement noise; this solution improves the robustness of thermal state sensing through multi-source data fusion. Conventional open-loop compensation strategies lack a real-time feedback mechanism; this solution uses an online evaluation module to form a closed-loop correction, significantly reducing the risk of error accumulation.

[0023] This application can accurately predict nonlinear structural deformation under complex thermal environments, eliminating compensation biases caused by traditional linear models. The multi-sensor data fusion mechanism effectively suppresses the interference of local measurement noise on control decisions, improving the reliability of the compensation strategy. The closed-loop control system enables real-time verification and dynamic optimization of the compensation effect, avoiding the decrease in machining accuracy caused by error accumulation. Submicron-level actuators ensure that thermal drift compensation is accurately converted into platform pose adjustment, meeting the requirements of high-precision optical component machining.

[0024] Example 2: Please refer to Figure 1 The multi-point sensing module includes a temperature detection unit and a displacement detection unit; The temperature detection unit is used to collect temperature change information such as ambient temperature, structural temperature rise, and heat conduction near the processing area in real time, and output multi-dimensional temperature field data to provide basic input variables for the thermal-displacement model. The displacement detection unit is used to perform nanometer-level measurements of key parameters such as spatial displacement and angular attitude of optical components, and outputs structural response data in real time. It is used in conjunction with temperature data for nonlinear modeling and compensation judgment.

[0025] In this embodiment: the multi-point sensing module includes a temperature detection unit and a displacement detection unit; the temperature detection unit is used to collect temperature change information such as ambient temperature, structural temperature rise, and heat conduction near the processing area in real time, and output multi-dimensional temperature field data to provide basic input variables for the thermal-displacement model; the displacement detection unit is used to perform nanometer-level measurements of key parameters such as spatial displacement and angular attitude of optical elements, and output structural response data in real time, which, together with the temperature data, is used for nonlinear modeling and compensation judgment.

[0026] A temperature detection unit refers to a sensor array capable of simultaneously acquiring multi-dimensional temperature parameters. Specifically, it can be implemented using a distributed thermocouple array combined with an infrared thermal imager. Through multi-point placement, it covers the processing environment, mechanical structure, and heat source conduction paths, forming three-dimensional temperature field distribution data. This unit captures the dynamic changes of different heat sources, providing complete temperature input variables for nonlinear modeling of thermally induced displacement.

[0027] The displacement detection unit refers to a spatial displacement measurement device with nanometer-level resolution. Specifically, it can be implemented using a laser interferometer combined with a capacitive displacement sensor, acquiring six-degree-of-freedom displacement data of optical elements through multi-axis synchronous measurement. This unit provides fundamental data support for verifying the dynamic characteristics of the thermo-displacement coupling relationship by correlating high-precision displacement data with the time series of the temperature field.

[0028] The temperature detection unit continuously collects data on ambient temperature fluctuations, temperature gradients of the mechanical structure, and thermal radiation conduction in the ion beam processing area during the processing, forming a multi-dimensional temperature field dataset containing spatial distribution characteristics. For example, arranging a thermocouple array at the edge of the processing area can capture the temperature attenuation law of the heat source conduction path, which, combined with the two-dimensional temperature distribution map obtained by the infrared thermal imager, constitutes a complete characterization of the source of thermal disturbance. The displacement detection unit monitors the sub-nanometer displacement of optical components using a laser interferometer. For example, interferometer mirrors are arranged in the X / Y / Z axes to measure the micrometer-level displacement shift caused by thermal expansion during processing in real time. After the temperature and displacement data are synchronized in time, a dynamic mapping relationship between temperature gradient changes and structural deformation can be established, providing a composite dataset with spatiotemporal correlation for subsequent nonlinear model training.

[0029] Traditional systems typically use a single temperature sensor to monitor local temperature rise, failing to distinguish the differences in the effects of ambient temperature fluctuations, structural heat conduction, and processing heat sources, resulting in a lack of information about the source of thermal disturbances in the model's input variables. This solution utilizes a distributed temperature sensor array combined with thermal imaging technology to simultaneously acquire the temperature field distribution characteristics of ambient temperature, structural temperature rise gradient, and heat source conduction paths, enabling the model to identify the contribution weights of different heat sources to displacement deformation. Furthermore, existing displacement detection methods often employ micrometer-level optical grating rulers, which struggle to capture the initial, minute displacements of thermally induced deformation. In contrast, this solution uses a laser interferometer that achieves nanometer-level displacement resolution, and combined with multi-axis synchronous measurement technology, it can accurately quantify the multidimensional composite deformation caused by thermal expansion.

[0030] This application addresses the problem of incomplete model input caused by the single dimension of temperature field information acquisition in traditional positioning systems. Simultaneously, it verifies the dynamic characteristics of the temperature-displacement coupling relationship using high-precision displacement data. The multidimensional temperature field data output by the temperature detection unit accurately reflects the spatiotemporal distribution characteristics of different heat sources, while the nanometer-level measurement accuracy of the displacement detection unit captures initial minute deformation trends. The synergistic effect of both enables the nonlinear modeling module to establish a more accurate thermally induced displacement prediction model, thereby improving the reliability and environmental adaptability of thermal drift compensation decisions.

[0031] Example 3: Please refer to Figure 1 The nonlinear modeling module includes a data preprocessing unit and a model building unit; The data preprocessing unit is used to clean, denoise and normalize the raw temperature and displacement data from the multi-point sensing module, perform time synchronization, feature extraction and outlier removal, and construct the processed dataset into a training input format. The model building unit is used to automatically select the modeling structure and parameters according to different working conditions or structural conditions using machine learning algorithms, and outputs a predictive model for subsequent control modules to make real-time compensation decisions.

[0032] In this embodiment: the nonlinear modeling module includes a data preprocessing unit and a model building unit; the data preprocessing unit is used to clean, denoise and normalize the raw temperature and displacement data from the multi-point sensing module, perform time synchronization, feature extraction and outlier removal, and construct the processed dataset into a training input format; the model building unit is used to train a thermal-displacement nonlinear prediction model using machine learning algorithms, automatically select the modeling structure and parameters according to different working conditions or structural conditions, and output the prediction model for the subsequent control module to make real-time compensation decisions.

[0033] Data cleaning refers to identifying and deleting invalid or erroneous data through preset rules. This can be achieved using threshold filtering or logical verification methods to eliminate abnormal data points caused by sensor signal loss or transmission errors. Noise reduction involves suppressing high-frequency noise interference using filtering algorithms. This can be achieved using wavelet transform or moving average methods to improve the signal-to-noise ratio. Normalization maps sensor data with different dimensions to a unified numerical range. This can be achieved using max-min scaling or standardization methods to eliminate the impact of dimensional differences on model training. Time synchronization aligns the timestamps of multi-source sensor data. This can be achieved using interpolation or clock synchronization protocols to ensure the temporal correlation between temperature changes and displacement responses. Feature extraction selects key variables from the raw data. This can be achieved using principal component analysis or correlation analysis methods to reduce the negative impact of redundant features on model complexity. Outlier removal identifies and removes data points that deviate from the normal distribution. This can be achieved using box plots or isolated forest algorithms to avoid abnormal samples interfering with model training. Machine learning algorithms refer to data-driven models that construct nonlinear mapping relationships. These can be implemented using neural networks or support vector machines to capture complex nonlinear patterns in heat-displacement interactions. Automatic selection of modeling structure and parameters refers to dynamically adjusting the model architecture based on changes in operating conditions. This can be achieved using Bayesian optimization or genetic algorithms to adapt to differences in thermal response characteristics under different processing environments.

[0034] After cleaning and denoising, the raw temperature and displacement data are synchronized to ensure the correspondence between temperature acquisition and displacement measurement times, eliminating timing misalignment caused by differences in sensor sampling frequencies. Normalization converts temperature and displacement signals to the same dimensional range, preventing weight bias in the model due to differences in input magnitudes. Feature extraction filters variables strongly correlated with thermally induced displacement from multidimensional data, such as key parameters like structural temperature gradient and heat transfer rate, reducing the computational complexity of model training. Outlier removal identifies and removes data points deviating from normal operating conditions through statistical distribution analysis, such as displacement abrupt changes caused by transient electromagnetic interference. The preprocessed data is input into the model building unit, where machine learning algorithms learn the nonlinear mapping relationship between temperature change and displacement response through training, for example, using long short-term memory networks to capture the hysteresis effect of heat conduction. The modeling structure is automatically selected according to different operating conditions; for example, a linear regression model is used under steady-state thermal fields, switching to a nonlinear neural network model under transient thermal fields, achieving dynamic adaptation.

[0035] Existing technologies typically employ fixed threshold filtering or single filtering algorithms to process raw data, which cannot effectively address the issues of multi-source noise coupling and feature redundancy. Furthermore, they rely on empirical formulas to establish linear heat-displacement models, making it difficult to adapt to nonlinear responses under complex operating conditions. This proposed solution eliminates noise interference and extracts key features through multi-level data preprocessing, and combines this with machine learning algorithms to construct a dynamically adjustable nonlinear prediction model, significantly improving the model's adaptability to different heat source distributions and structural materials.

[0036] This application addresses the problem of decreased model prediction accuracy caused by noise interference, temporal misalignment, and outliers in temperature and displacement data. By enhancing the model's ability to capture nonlinear thermally induced displacement through feature selection and dynamic modeling, the positioning system can maintain high-precision compensation even when the processing environment temperature fluctuates or the structural thermal response characteristics change.

[0037] Example 4: Please refer to Figure 1 The data fusion module includes a multi-source data fusion unit and a state self-correction unit; The multi-source data fusion unit is used to receive temperature and displacement data from multiple sensors, and to perform comprehensive processing on the spatiotemporal distributed data based on weighted averaging, Kalman filtering and Bayesian fusion algorithms, and output the fused data. The state self-calibration unit is used to monitor the drift trend or abnormal fluctuation of sensor output, automatically identify potential distorted data or inconsistent states, and dynamically correct the fusion strategy.

[0038] In this embodiment, the data fusion module includes a multi-source data fusion unit and a state self-correction unit. The multi-source data fusion unit receives temperature and displacement data from multiple sensors, performs comprehensive processing on the spatiotemporal distribution data using weighted averaging, Kalman filtering, or Bayesian fusion algorithms, and outputs the fused data. The state self-correction unit monitors the drift trend or abnormal fluctuations in the sensor outputs, automatically identifies potentially distorted data or inconsistent states, and dynamically corrects the fusion strategy.

[0039] The multi-source data fusion unit refers to the unit that performs spatiotemporal alignment and joint processing of heterogeneous data collected by multiple sensors. Specifically, it can be implemented using a Kalman filter algorithm based on timestamp synchronization, for example, by establishing a time-series model of the sensor data to eliminate spatiotemporal misalignment caused by transmission delays. This unit solves the problem of inconsistent data spatial distribution caused by differences in sensor installation locations by fusing the confidence weights of different sensors through algorithms.

[0040] The state self-calibration unit refers to a unit that performs online monitoring of the sensor health status and dynamically adjusts the fusion strategy. Specifically, it can be implemented using a sliding window statistical analysis method, such as calculating the mean square error of the sensor output values ​​to determine if they deviate from the normal operating range. This unit continuously evaluates the reliability of each sensor's data and automatically reduces the weight of the corresponding data in the fusion process when abnormal fluctuations are detected, preventing overall data distortion caused by a single sensor failure.

[0041] The multi-source data fusion unit first performs time synchronization processing on the raw data from the temperature detection unit and the displacement detection unit, for example, by using interpolation algorithms to unify data from different sampling frequencies to the same time base. Then, a Bayesian fusion algorithm is used to calculate the posterior probability distribution of each data source based on a pre-defined sensor noise model, generating fused temperature field distribution data and displacement change trends. The state self-calibration unit monitors the output stability of each sensor in real time during the data fusion process. For example, when the reading of a temperature sensor continuously exceeds three standard deviations of the historical mean, a fusion strategy adjustment mechanism is triggered, dynamically reducing the weight coefficient of that sensor from the initial value of 0.3 to 0.1, while simultaneously increasing the weight ratio of other adjacent sensors, thereby suppressing the impact of abnormal data on the fusion results.

[0042] Traditional methods typically employ fixed-weight data averaging algorithms, which cannot effectively handle data abrupt changes caused by sensor dynamic drift or sudden interference. This solution introduces a state self-correction mechanism, which can automatically identify and correct the fusion strategy when sensors experience progressive drift. For example, if a displacement sensor experiences thermal expansion error due to localized temperature rise during manufacturing, its data weight is reduced to maintain overall fusion accuracy. Furthermore, existing multi-source data fusion technologies often neglect spatiotemporal synchronization issues, while this solution eliminates the spatiotemporal mismatch caused by dispersed sensor layouts through timestamp alignment and spatial interpolation.

[0043] This application addresses the inconsistency in the spatiotemporal distribution of multi-sensor data. For example, during pulsed ion beam processing, when the temperature field changes rapidly due to heat source movement, it maintains the temporal consistency and spatial rationality of the fused data. Simultaneously, it effectively reduces the risk of data distortion caused by sensor output drift or transient interference. For instance, when a cooling system malfunction causes a sudden local temperature change, it avoids erroneous compensation operations by dynamically adjusting the fusion strategy. This further enhances the reliability of data fusion in complex thermal environments, providing highly consistent input data for subsequent thermal drift compensation control.

[0044] Example 5: Please refer to Figure 1 The intelligent control module includes a compensation strategy generation unit and an adaptive control unit; The compensation strategy generation unit is used to receive the predicted displacement deviation data output by the nonlinear modeling module, integrate the current state information, calculate the required spatial displacement compensation amount in real time, and construct the corresponding control command sequence, including multi-dimensional compensation parameters for position adjustment and attitude change. The adaptive control unit is used to automatically select the control model and dynamically adjust the control parameters according to different operating conditions.

[0045] In this embodiment: the intelligent control module includes a compensation strategy generation unit and an adaptive control unit; the compensation strategy generation unit is used to receive the predicted displacement deviation data output by the nonlinear modeling module, integrate the current state information, calculate the required spatial displacement compensation amount in real time, and construct the corresponding control command sequence, including multi-dimensional compensation parameters for position adjustment and attitude change; the adaptive control unit is used to automatically select the control model according to different working conditions and dynamically adjust the control parameters.

[0046] The compensation strategy generation unit refers to the logical unit that generates multi-dimensional compensation parameters based on predicted displacement deviation data. Specifically, it can be implemented using a real-time optimization algorithm combined with a multivariable control model. This involves fusing displacement deviation and real-time state information to generate composite compensation commands that include position and attitude adjustments. The adaptive control unit refers to the execution unit that adjusts the control strategy according to changes in operating conditions. This can be implemented using fuzzy control or model predictive control algorithms. It dynamically switches control models by online identification of operating parameters such as processing load and ambient temperature.

[0047] The compensation strategy generation unit receives predicted displacement deviation data from the nonlinear modeling module. Combining this data with the temperature distribution and platform load information in the current processing state, it calculates the translational compensation and rotational compensation angle of the optical element in the spatial coordinate system using a multivariate optimization algorithm. This generates a control command sequence containing XYZ axis displacement and attitude adjustment parameters. The adaptive control unit monitors operating parameters such as ambient temperature fluctuations and ion beam processing power changes in real time, triggering a control model switching mechanism. For example, it switches to a feedforward-feedback composite control model when the temperature changes rapidly, and adopts a proportional-integral-derivative control model during the steady-state processing stage. Simultaneously, it dynamically adjusts the control gain parameters based on error feedback.

[0048] Traditional thermal drift compensation systems typically employ fixed proportional-integral-derivative parameters or a single feedforward control model, which cannot cope with nonlinear response differences caused by sudden temperature changes or variations in processing load. This solution utilizes a compensation strategy generation unit to calculate multi-dimensional compensation parameters, resolving the residual angular deviation problem caused by traditional methods that only perform single-axis translational compensation. The adaptive control unit overcomes the response lag defect of fixed control parameters under dynamic thermal disturbances through operating condition identification and model switching mechanisms.

[0049] This application can generate compensation commands that include coordinated adjustment of position and attitude based on real-time predicted displacement deviation, effectively eliminating multi-dimensional thermal deformation errors; by dynamically adjusting the control model and parameters, it improves the compensation stability of the system under complex working conditions such as temperature changes and processing load fluctuations, and avoids compensation delays or malfunctions caused by insufficient environmental adaptability.

[0050] Example 6: Please refer to Figure 1 The micro-motion execution module includes a precision drive unit and an attitude feedback unit; The precision drive unit is used to receive control commands and perform fine-tuning operations in the XYZ directions and attitude angles; The attitude feedback unit is used to monitor the spatial attitude changes of the platform in real time.

[0051] In this embodiment: the micro-motion execution module includes a precision drive unit and an attitude feedback unit; the precision drive unit is used to receive control commands and perform fine-tuning operations in the XYZ directions and attitude angles; the attitude feedback unit is used to monitor the spatial attitude changes of the platform in real time.

[0052] A precision drive unit is an actuator that converts electrical signals into mechanical displacement through an electromechanical conversion device. Specifically, it can be implemented using a piezoelectric ceramic driver or a voice coil motor. A high-resolution encoder enables closed-loop control of the displacement, ensuring sub-micron level precision in fine-tuning of the XYZ directions and attitude angles. An attitude feedback unit is a sensing device used to capture changes in the platform's spatial attitude. Specifically, it can be implemented using an optical interferometer or a capacitive displacement sensor. A multi-degree-of-freedom measurement system collects the platform's position and angle data in real time, providing feedback signals for dynamic evaluation of the compensation effect.

[0053] After receiving compensation commands from the intelligent control module, the precision drive unit generates micron-level displacement adjustments in the XYZ translational degrees of freedom and the rotational degrees of freedom around the axis through a multi-axis linkage mechanism, compensating for multidimensional deformation errors caused by thermal drift. The attitude feedback unit synchronously collects the adjusted spatial attitude data of the platform and transmits the real-time monitoring results to the effect evaluation module via a data bus, forming an execution-feedback closed-loop control loop. In this process, the high-precision execution capability of the precision drive unit and the real-time monitoring function of the attitude feedback unit work together to avoid the accumulation of positioning deviations caused by actuator backlash errors or environmental disturbances, while also solving the multi-degree-of-freedom coupling error problem caused by neglecting angular deviations in traditional systems.

[0054] Traditional positioning systems typically only support single-axis linear displacement compensation, lacking multi-degree-of-freedom collaborative adjustment capabilities and failing to integrate real-time attitude monitoring, resulting in the inability to identify and correct angular deviations. This solution, by introducing a multi-axis precision drive unit and a multi-degree-of-freedom attitude feedback unit, constructs a closed-loop execution system covering translational and rotational degrees of freedom. This system can eliminate angular deviations while compensating for displacement, effectively suppressing the impact of multi-dimensional deformation coupling on positioning accuracy.

[0055] This application achieves precise execution of submicron level displacement and attitude adjustment, and dynamically monitors the execution effect through a real-time feedback mechanism to avoid the problem of decreased positioning accuracy caused by the accumulation of execution errors or lack of feedback, thus ensuring the spatial position and attitude stability of optical components in complex thermal environments.

[0056] Example 7: Please refer to Figure 1 The effect evaluation module includes a compensation error detection unit and a model feedback update unit; The compensation error detection unit is used to calculate the residual error and determine whether the current compensation effect meets the preset accuracy requirements. The model feedback update unit is used to correct the model or fine-tune the control parameters based on the error change trend.

[0057] In this embodiment: the effect evaluation module includes a compensation error detection unit and a model feedback update unit; the compensation error detection unit is used to calculate the residual error and determine whether the current compensation effect meets the preset accuracy requirements; the model feedback update unit is used to correct the model or fine-tune the control parameters according to the error change trend.

[0058] The compensation error detection unit is a module that calculates the difference between the compensated actual displacement and the target displacement in real time. Specifically, it can use differential algorithms or least squares methods to quantify and analyze the residual error, and trigger an accuracy warning by setting an error threshold. The model feedback update unit is a module that dynamically adjusts the prediction model or control parameters based on error data. Specifically, it can use incremental learning algorithms or parameter adaptive optimization algorithms to iteratively update the model weights or control gains online to eliminate systematic biases.

[0059] After each compensation action, the compensation error detection unit calculates the residual error value by comparing the difference between the target displacement and the actual displacement, and then compares this error value with a preset accuracy threshold. If the error exceeds the allowable range, an error warning signal is generated. The model feedback update unit identifies the error change trend based on the time series of error data, for example, by using a sliding window statistical method to distinguish between random fluctuations and trend deviations, and then dynamically adjusts the parameters of the prediction model or optimizes the gain coefficient in the control strategy. By periodically executing error detection and model updates, a closed-loop feedback mechanism is formed, enabling the compensation strategy to adapt to the effects of environmental changes or system aging.

[0060] Traditional positioning systems typically only perform model calibration in the initial stage, lacking the ability to continuously monitor and dynamically correct the compensation effect, leading to the gradual accumulation of errors over long-term operation. This solution, however, through real-time error detection and online model updates, can promptly correct compensation deviations caused by changes in thermal response characteristics or external disturbances, avoiding the problem of decreased positioning accuracy due to error accumulation.

[0061] This application enables real-time monitoring and dynamic optimization of the compensation effect, effectively suppressing the accumulation of residual errors and improving the long-term stability of the positioning system under complex working conditions. Through a closed-loop feedback mechanism, the compensation strategy can adaptively adjust, overcoming the insufficient adaptability of traditional methods due to model fixation, and ensuring the continuous and reliable operation of the high-precision machining process.

[0062] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

[0063] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A positioning system for pulsed ion beam processing of optical elements, characterized in that: It includes a multi-point sensing module, a nonlinear modeling module, a data fusion module, an intelligent control module, a micro-motion execution module, and an effect evaluation module; The multi-point sensing module is used to collect temperature and displacement data of key parts of the positioning system in real time. The data covers the thermal state and structural deformation information of multiple key points in the positioning system, providing basic data support for subsequent compensation strategies. The nonlinear modeling module is used to establish a nonlinear mapping relationship model between temperature change and structural displacement, accurately describing the nonlinear characteristics of heat-displacement interaction. The data fusion module is used to jointly process, filter, and optimize temperature and displacement data from different sensors to ensure data consistency, accuracy, and reliability. The intelligent control module generates a thermal drift compensation control strategy in real time based on the temperature and displacement information output by the data fusion module and the prediction model established by the nonlinear modeling module. The micro-motion execution module is used to drive the positioning platform to perform sub-micron level displacement and attitude adjustment based on the compensation commands output by the intelligent control module. The effect evaluation module is used to evaluate and provide feedback on the execution effect after thermal drift compensation is implemented in real time, and to determine whether the compensation has achieved the expected positioning accuracy.

2. The positioning system for pulsed ion beam processing of optical elements according to claim 1, characterized in that: The multi-point sensing module includes a temperature detection unit and a displacement detection unit; The temperature detection unit is used to collect temperature change information such as ambient temperature, structural temperature rise, and heat conduction near the processing area in real time, and output multi-dimensional temperature field data to provide basic input variables for the thermal-displacement model. The displacement detection unit is used to perform nanometer-level measurements of key parameters such as spatial displacement and angular attitude of optical elements, and outputs structural response data in real time. This data, together with temperature data, is used for nonlinear modeling and compensation judgment.

3. The positioning system for pulsed ion beam processing of optical elements according to claim 2, characterized in that: The nonlinear modeling module includes a data preprocessing unit and a model building unit; The data preprocessing unit is used to clean, denoise and normalize the raw temperature and displacement data from the multi-point sensing module, perform time synchronization, feature extraction and outlier removal, and construct the processed dataset into a training input format. The model building unit is used to employ machine learning algorithms to automatically select modeling structures and parameters based on different working conditions or structural conditions, and output a predictive model for subsequent control modules to make real-time compensation decisions.

4. The positioning system for pulsed ion beam processing of optical elements according to claim 3, characterized in that: The data fusion module includes a multi-source data fusion unit and a state self-correction unit; The multi-source data fusion unit is used to receive temperature and displacement data from multiple sensors, and to perform comprehensive processing on the spatiotemporal distributed data based on weighted averaging, Kalman filtering and Bayesian fusion algorithms, and output the fused data. The state self-correction unit is used to monitor the drift trend or abnormal fluctuation of the sensor output, automatically identify potential distorted data or inconsistent states, and dynamically correct the fusion strategy.

5. A positioning system for pulsed ion beam processing of optical elements according to claim 4, characterized in that: The intelligent control module includes a compensation strategy generation unit and an adaptive control unit; The compensation strategy generation unit is used to receive the predicted displacement deviation data output by the nonlinear modeling module, integrate the current state information, calculate the required spatial displacement compensation amount in real time, and construct the corresponding control command sequence, including multi-dimensional compensation parameters for position adjustment and attitude change. The adaptive control unit is used to automatically select the control model according to different operating conditions and dynamically adjust the control parameters.

6. A positioning system for pulsed ion beam processing of optical elements according to claim 5, characterized in that: The micro-motion execution module includes a precision drive unit and an attitude feedback unit; The precision drive unit is used to receive control commands and perform fine-tuning operations in the XYZ directions and attitude angles; The attitude feedback unit is used to monitor the spatial attitude changes of the platform in real time.

7. A positioning system for pulsed ion beam processing of optical elements according to claim 6, characterized in that: The performance evaluation module includes a compensation error detection unit and a model feedback update unit; The compensation error detection unit is used to calculate the residual error and determine whether the current compensation effect meets the preset accuracy requirements. The model feedback update unit is used to correct the model or fine-tune the control parameters according to the error change trend.

Images