System and method for improving laser interference measurement precision

Abstract

The invention discloses a system and a method for improving laser interference measurement precision, and belongs to the technical field of laser precision measurement. In order to solve the problems that the temperature still has certain periodic fluctuation even if effective temperature control is carried out during measurement in a laser interference measurement technology, and the fluctuation greatly affects the laser measurement precision, the invention utilizes the characteristic that the temperature variation has an approximately linear deterministic relationship under the coupling action of factors such as thermal stability of multiple devices and the like. A model of data fluctuation generated by spontaneous heating of a device in laser precision measurement is established by utilizing data in a stable state in a precision measurement process in combination with a time sequence prediction model and an integrated learning mode. Through the method, a final model after mutual coupling of all factors can be directly established, the system complexity is effectively reduced, and the final measurement error precision can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of laser precision measurement, and in particular relates to a system and method for improving the accuracy of laser interference measurement. Background Art

[0002] As an important optical measurement technology, laser interferometry has been widely used in scientific research and engineering applications. It is based on the principle of laser interferometry, using the coherence and wave properties of lasers to achieve high-precision measurement of parameters such as the shape, surface features, and motion state of target objects through interference phenomena.

[0003] In the future, the demand for angular distance measurement accuracy in the fields of aerospace satellite attitude correction, intersatellite ranging, quantum radar, etc. has reached 0.1nm, and laser interferometry is the key technology to solve the ultra-precision measurement tasks in the above application scenarios. First of all, in the design and manufacture of spacecraft, laser interferometry technology is widely used to precisely measure the shape and size of spacecraft structures. Through the laser interferometry system, engineers can monitor the deformation of the spacecraft structure in real time to ensure that it has good structural stability and reliability in the space environment. Secondly, laser interferometry technology also plays an important role in the execution of space missions. For example, in satellite docking and spacecraft attitude control, the laser interferometry system can achieve high-precision measurement and monitoring of satellite position, attitude and other parameters to ensure the smooth execution of space missions and accurate docking of targets. In general, laser interferometry technology plays an important role in the aerospace field with its high precision, real-time and wide applicability, and provides key support for the development of aerospace technology and the smooth execution of aerospace missions. With the continuous innovation and improvement of technology, it is believed that this technology will play a more important role in the development of the aerospace field in the future.

[0004] However, under the existing technology, when laser interferometry is used for measurement, the temperature is mostly adjusted by placing multiple semiconductor TEC temperature control devices, or a constant temperature device similar to a constant temperature box that can accommodate the measuring device is established, and the measurement is started when the temperature stability of the measuring system reaches a certain level. However, even if effective temperature control is carried out, the temperature will still fluctuate periodically within a certain period of time, and the impact of temperature changes on the accuracy of laser measurement is more significant, because its changes will cause a chain reaction of changes in various parameters, and they will affect the final data under the coupling effect. Therefore, the final measurement data will also produce a periodic fluctuation that follows the temperature change. The existing technology usually considers the factors affected by temperature separately, such as calculating the impact of device expansion separately, which increases the complexity of the system and the complexity of data processing. Summary of the invention

[0005] In view of the problem that even after effective temperature control during measurement, the temperature of laser interferometer measurement technology still has certain periodic fluctuations, which has a great impact on the laser measurement accuracy, the present invention provides a system and method for improving the accuracy of laser interferometer measurement, the purpose of which is to eliminate the deviation of the final measurement result caused by the common coupling of multiple factors affected by temperature by establishing an error model. The regular drift of the period is found by adopting the self-optimizing time prediction-classification integrated learning method, so as to obtain the actual accurate real measurement accuracy.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A system for improving the accuracy of laser interferometry measurement, comprising a temperature control module, an interferometry module, an error model training unit, and a data processing unit; wherein:

[0008] Temperature control module: used for temperature stability control;

[0009] Interference measurement module: used for static displacement data acquisition under static displacement state;

[0010] Error model training unit: used to perform error model training on the collected and processed static displacement data;

[0011] Data processing unit: used for removing noise from static displacement data and cutting and classifying static displacement data.

[0012] A method for improving the accuracy of laser interferometry comprises the following steps:

[0013] Step S1: constructing a measuring device, and placing the entire set of measuring devices in a space environment equipped with a temperature control device to ensure that the temperature of the space environment where the measuring device is located is controllable and the temperature does not change suddenly;

[0014] Step S2: setting the temperature control device at a specified temperature to keep the temperature constant within a certain range while ensuring that the temperature control device operates effectively;

[0015] Step S3: The target to be measured is placed in a stable state, and distance data is collected by laser interferometry technology;

[0016] Step S4: Display the collected data in the time domain, find the period that follows the temperature change, and cut it into sections according to the characteristic points of the change period. After cutting, each section of the measured data presents a data relationship with time as the horizontal axis and the vertical axis approximately linearly changing with respect to time;

[0017] Step S5: preprocessing the cut measurement data, using the time prediction model to perform fitting prediction on the preprocessed measurement data, and establishing a corresponding data model;

[0018] Step S51: Perform Fourier transform on the collected time domain data to obtain frequency domain data, obtain the power of each frequency through an algorithm, and display the frequency point with maximum power;

[0019] Step S52: removing the obvious power interference signal frequency by means of a notch filter;

[0020] Step S53: removing salt and pepper noise from the measured shift data processed by the notch filter through median filtering;

[0021] Step S54: removing white noise from the measured data after the median filtering by mean filtering;

[0022] Step S55: Obtain the best fitting parameters by automatic adjustment.

[0023] Step S6: There are multiple linear prediction models with the same change trend in the data model. The error fitting models of the same type are integrated through ensemble learning to obtain error models under various trends. The optimal parameters of the time prediction ARMIA model are automatically selected by the stepwise regression method to perform the best fitting prediction on the processed data.

[0024] Furthermore, the fitting prediction process of step S6 is as follows:

[0025] Step S61: Select an initial model;

[0026] Step S62: Change one of the parameters and calculate a new AIC value. The calculation formula is:

[0027] AIC=2k-2ln(L)

[0028] In the formula, AIC represents the goodness of fit, the smaller the better, k represents the number of parameters, and L represents the likelihood function;

[0029] Step S63: Compare the AIC values ​​of the current model and the new model, and select the model with the smaller AIC value as the current optimal model;

[0030] Step S64: Repeat steps S62 to S64 until a preset stop condition is reached.

[0031] Step S7: The error fitting models of the same type in step S6 are integrated by means of ensemble learning, and the time prediction error model with relatively good consistency for each data template is finally obtained by the method in the Scikit-learn library;

[0032] Step S8: Cut the newly measured data, cut the newly measured data, and subtract the error models obtained in step S7 according to the corresponding change trends to obtain the final measurement data after error compensation.

[0033] Furthermore, in step S2, when the temperature control device operates effectively, the temperature shows periodic changes.

[0034] Furthermore, in step S4, after the measurement data is cut, the phenomenon that each segment of data is locally amplified to present a sinusoidal oscillation can be ignored when the overall relationship is approximately linear.

[0035] Furthermore, in step S6, when establishing the error model, it is necessary to ensure that the object being measured is in a stable state, so that the change of the measurement data under the influence of temperature under the overall coupling of various factors can be measured.

[0036] Furthermore, the measurement environment in step S8 should be consistent with the measurement environment in step S4.

[0037] Compared with the prior art, the present invention has the following advantages:

[0038] 1. The present invention adopts a machine learning-like method to automatically iterate and obtain the model parameters of time prediction, reducing the amount of calculation and complexity of traditional mathematical modeling.

[0039] 2. The present invention combines a self-optimizing time prediction classification integrated learning hybrid algorithm, continuously collects raw data and conducts training and learning to optimize the model, and finally can effectively reduce the measurement error.

[0040] 3. When the measurement process is significantly affected by a certain environmental parameter, and this environmental parameter will cause changes in many other parameters, the influence of this parameter on each change factor will no longer be considered separately, but it will be directly modularized as a whole, thereby simplifying the complexity of the system.

[0041] Improving measurement accuracy is of great significance to future fields such as aerospace satellite attitude correction, intersatellite ranging and quantum radar. In terms of satellite attitude correction, precise measurement can realize real-time monitoring and adjustment of satellite attitude, ensuring that it maintains stable and accurate positioning in space, thereby improving the success rate of communication and observation missions. In the field of intersatellite ranging, precise measurement can realize the distance measurement of interstellar space, provide reliable data support for interstellar exploration and navigation, and promote human exploration of the farther universe. In terms of quantum radar, high-precision measurement can realize the rapid acquisition of information such as target position and speed, greatly improve the detection range and accuracy of the radar system, and help to cope with aerospace safety challenges in complex environments. Therefore, improving measurement accuracy will not only promote the development of aerospace technology, but also deepen human understanding of the universe and promote continuous progress in the aerospace field. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is a system structure diagram of the present invention;

[0043] Figure 2 Establishing a process for fitting the model in the present invention;

[0044] Figure 3 It is the standard deviation of some measured data, the data after filtering and other processing, the fitting curve, and the standard deviation after residual processing. DETAILED DESCRIPTION

[0045] In order to gain a deeper understanding of the present invention, we will provide a comprehensive and detailed description of the present invention. However, the present invention has multiple implementations and is not limited to the specific examples listed herein. The presentation of these examples is intended to deepen the comprehensive understanding of the disclosure of the present invention.

[0046] The present invention provides a system and method for improving the accuracy of laser interferometry measurement, which compensates for the measurement error caused by temperature change during laser measurement. First, the measurement data of the measured object in a stable state is used to fit the training error model, and then the model is brought into the new measurement data for error compensation. The specific implementation method and steps are as follows:

[0047] A system for improving the accuracy of laser interferometry measurement, comprising a temperature control module, an interferometry module, an error model training unit, and a data processing unit; wherein:

[0048] Temperature control module: used for temperature stability control;

[0049] Interference measurement module: used for static displacement data acquisition under static displacement state;

[0050] Error model training unit: used to perform error model training on the collected and processed static displacement data;

[0051] Data processing unit: used for removing noise from static displacement data and cutting and classifying static displacement data.

[0052] A method for improving laser interferometry accuracy comprises the following steps:

[0053] Step S1: constructing a measuring device, and placing the entire set of measuring devices in a space environment equipped with a temperature control device to ensure that the temperature of the space environment where the measuring device is located is controllable and the temperature does not change suddenly;

[0054] Step S2: setting the temperature control device at a specified temperature to keep the temperature constant within a certain range while ensuring that the temperature control device operates effectively;

[0055] Step S3: The target to be measured is placed in a stable state, and distance data is collected by laser interferometry technology;

[0056] Step S4: Display the collected measurement data in the time domain, and find the period of fluctuation following the temperature change. In this embodiment, the fluctuation period is about 1.5 hours, and the fluctuation is approximately triangular wave in the long term and linear in the short term. The measurement data is cut into sections according to the peaks and troughs of the triangular wave; after cutting, each section of the data presents a change with time as the horizontal axis and the vertical axis approximately linear with respect to time;

[0057] Step S5: preprocessing the data of each segment after cutting, using the time prediction model to fit and predict the preprocessed measurement data, and establishing a corresponding data model;

[0058] Step S51: Perform Fourier transform on the collected time domain data to obtain frequency domain data, obtain the power of each frequency through an algorithm, and display the frequency point with maximum power;

[0059] Step S52: remove the maximum power interference signal frequency through a notch filter, where the frequency is set to around 14 Hz and the notch filter BW is set to 5;

[0060] Step S53: The data processed by the notch filter is subjected to median filtering to remove salt and pepper noise, and the window order of the median filter is set to 1601;

[0061] Step S54: the data after the median filtering is subjected to mean filtering to remove white noise, and the window order of the mean filter is set to 1600;

[0062] Step S55: obtaining the best fitting parameters by automatic adjustment;

[0063] Step S6: There are multiple linear prediction models with the same change trend in the data model. The error fitting models of the same type are integrated through ensemble learning to obtain error models under various trends. The optimal parameters of the time prediction ARMIA model are automatically selected by the stepwise regression method to perform the best fitting prediction on the processed data. In this example, the error fitting data models for the temperature during the callback, the rising and falling processes in the controlled process are established respectively, such as Figure 3 As shown, the original data of a descending process is intercepted here. The standard deviation of the original data is 3.412. The standard deviation of the data obtained after data processing is 1.446. This data is then used for fitting and prediction. Figure 3 The residual processing shows that the fitting effect is good; the fitting prediction process is as follows:

[0064] Step S61: Select an initial model;

[0065] Step S62: Change one of the parameters and calculate a new AIC value. The calculation formula is:

[0066] AIC=2k-2ln(L)

[0067] In the formula, AIC represents the goodness of fit, the smaller the better, k represents the number of parameters, and L represents the likelihood function;

[0068] Step S63: Compare the AIC values ​​of the current model and the new model, and select the model with the smaller AIC value as the current optimal model;

[0069] Step S64: repeating steps S62 to S64 until a preset stop condition is reached;

[0070] Step S7: The error fitting models of the same type in step S6 are integrated by means of ensemble learning, and the time prediction error model with relatively good consistency for each data template is finally obtained by the method in the Scikit-learn library;

[0071] Step S8: subtract the error model obtained in step S7 from the newly measured data to obtain the final measurement data after error compensation.

[0072] The contents not described in detail in the specification of the present invention belong to the prior art known to the professional and technical personnel in the field. Although the illustrative specific embodiments of the present invention are described above to facilitate the understanding of the present invention by the technical personnel in the field, it should be clear that the present invention is not limited to the scope of the specific embodiments. For the ordinary technical personnel in the field, as long as various changes are within the spirit and scope of the present invention defined and determined by the attached claims, these changes are obvious, and all inventions and creations using the concept of the present invention are protected.

Claims

1. A system for improving the accuracy of laser interferometry, characterized in that: The system includes a temperature control module, an interference measurement module, an error model training unit, and a data processing unit; wherein: Temperature control module: used for temperature stability control; Interference measurement module: used for static displacement data acquisition under static displacement state; Error model training unit: used to perform error model training on the collected and processed static displacement data; Data processing unit: used for removing noise from static displacement data and cutting and classifying static displacement data.

2. A method for improving the accuracy of laser interferometry, characterized in that: The method comprises the following steps: Step S1: constructing a measuring device, and placing the entire set of measuring devices in a space environment equipped with a temperature control device to ensure that the temperature of the space environment where the measuring device is located is controllable and the temperature does not change suddenly; Step S2: setting the temperature control device at a specified temperature to keep the temperature constant within a certain range while ensuring that the temperature control device operates effectively; Step S3: The target to be measured is placed in a stable state, and distance data is collected by laser interferometry technology; Step S4: Display the collected data in the time domain, find the period that follows the temperature fluctuation, and cut it into approximately linear segmented data; Step S5: preprocessing the cut measurement data, using the time prediction model to perform fitting prediction on the preprocessed measurement data, and establishing a corresponding data model; Step S6: There are multiple linear prediction models with the same change trend in the data model. The error fitting models of the same type are integrated through ensemble learning to obtain error models under various trends. The optimal parameters of the time prediction ARMIA model are automatically selected by the stepwise regression method to perform the best fitting prediction on the processed data. Step S7: The error fitting models of the same type in step S6 are integrated by means of ensemble learning, and the time prediction error model with relatively good consistency for each data template is finally obtained by the method in the Scikit-learn library; Step S8: Cut the newly measured data, cut the newly measured data, and subtract the error models obtained in step S7 according to the corresponding change trends to obtain the final measurement data after error compensation.

3. The method for improving the accuracy of laser interferometry according to claim 2, characterized in that: The specific operation of the data preprocessing work in step S5 is: Step S51: Perform Fourier transform on the collected time domain steady state data to obtain frequency domain data, obtain the power of each frequency through an algorithm, and display the frequency point with maximum power; Step S52: removing the obvious interference signal frequency by means of a notch filter; Step S53: removing salt and pepper noise from the measured data processed by the notch filter through median filtering; Step S54: removing white noise from the measured data after the median filtering by mean filtering; Step S55: Obtain the best fitting parameters by automatic adjustment.

4. The method for improving the accuracy of laser interferometry according to claim 2, characterized in that: The fitting prediction process of step S6 is as follows: Step S61: Select an initial model; Step S62: Change one of the parameters and calculate a new AIC value. The calculation formula is: AIC=2k-2ln(L) In the formula, AIC represents the goodness of fit, k represents the number of parameters, and L represents the likelihood function; Step S63: Compare the AIC values ​​of the current model and the new model, and select the model with the smaller AIC value as the current optimal model; Step S64: Repeat steps S62 to S64 until a preset stop condition is reached.

5. The method for improving the accuracy of laser interferometry according to claim 2, characterized in that: In step S2, when the temperature control device operates effectively, the temperature changes periodically.

6. The method for improving the accuracy of laser interferometry according to claim 2, characterized in that: In the step S4, after the measurement data is cut, the phenomenon of local amplification of each segment of data presenting sinusoidal oscillation is ignored, and the overall relationship is approximately linear.

7. The method for improving the laser interferometry accuracy according to claim 2, characterized in that: In step S6, when establishing the error model, it is necessary to ensure that the object being measured is in a stable state so that the change of the measurement deviation data affected by temperature under the overall coupling of various factors can be measured.

8. The method for improving the accuracy of laser interferometry according to claim 2, characterized in that: The measurement environment in step S8 should be consistent with the measurement environment in step S4.

Images